1 GSM Principles and Call Flow

1.1 GSM Frequency Band AllocationGSM cellular system can be divided into GSM900M andDCS1800M according to frequency band, with carrier frequencyinterval of 200 KHz and up and down frequencies as follows:

Table 1-1 GSM frequency allocation

Frequency band(MHz), Bandwidth(MHz), Frequency number,Carrier frequency number (pair)GSM900 Up 890–915 Down 935–960, 25, 1–124, 124DCS1800 Up 1710–1785 Down 1805–1880, 75, 512–885, 374

“Up” and “down” are classified according to base station. Basestation transmitting - mobile station receiving is “down”;mobile station transmitting - base station receiving is up.With the expanding services, GSM protocol adds EGSM(expanded GSM frequency band) and RGSM (expanded GSMfrequency band including railway service) to the originalGSM900 frequency band. The frequency band allocation is asfollows:

Table 1-2 EGSM/RGSM frequency allocationFrequency band(MHz), Bandwidth (MHz), Frequency number,Carrier frequency number (pair) EGSM Up 880–915 Down 925–960, 35, 0–124 , 975–1023, 174RGSM Up 876–915 Down 921–960, 40, 0–124, 955–1023, 199

1.2 Multiple Access Technology and Logical Channel

1.2.1 GSM Multiple Access TechnologyIn cellular mobile communications system, since many mobilesstations communicate with other mobiles stations through one

base station, it is necessary to distinguish the signals fromdifferent mobile stations and base stations for them to identifytheir own signals. The way to this problem is called multipleaccess technology. There are now five kinds of Multiple accesstechnology, namely: Frequency Division Multiple Access(FDMA), Time Division Multiple Access (TDMA), Code DivisionMultiple Access (CDMA), Space Division Multiple Access(SDMA), and polar division multiple access (PDMA).GSM multiple access technology focuses on TDMA, and takesFDMA as complement. The following only introduces FDMAand TDMA technologies.I. FDMAFDMA divides the whole frequency band into many single radiochannels (transmitting and receiving carrier frequency pairs).Each channel transmits one path of speech or controlinformation. Any subscriber has access to one of thesechannels under the control of the system.Analog cellular system is a typical example of FDMAapplication. Digital cellular system also uses FDMA, but not thepure frequency allocation. For example, GSM takes FDMAtechnology.II. TDMATDMA divides a broadband radio carrier into several timedivision channels according to time (or timeslot). Eachsubscriber takes one timeslot and sends or receives signals onlyin the specified timeslot. TDMA is applied in digital cellularsystem and GSM.GSM adopts a technology combined with FDMA and TDMA.

1.2.2 TDMA FrameThe basic conception of GSM in terms of radio path is burst.Burst is a transmission unit consists of over one hundred ofmodulation bits. It has a duration limit and takes a limited radiofrequency. They are exported in time and frequency windowwhich is called slot. To be specific, in system frequency band,

central frequency of slot is set in every 200 KHz (in FDMA). Slotoccurs periodically in each 15/26 ms, which is about 0.577 ms(in TDMA).The interval between two slots is called timeslot. Itsduration is used as time unit, called burst period (BP).Time/frequency map illustrates the concept of slot. Each slot isexpressed as one little rectangle with 15/26ms length and 200KHz width. See Figure 1-1. Similarly, the 200 KHz bandwidth inGSM is called frequency slot, equal to radio frequency channelin GSM protocol.Burst represents different meaning in different situation.Sometimes it concerns time – frequency “rectangle” unit, andsometimes not. Similarly, timeslot sometimes concerns timevalue, and sometimes means using one of every eight slotsperiodically.Using a given channel means transmitting burst with aparticular frequency at particular time, that is, a particular slot.Generally, the slot of a channel is not continuous in time.

Physical channel combines frequency division multiple accessand time division multiple access together. It consists oftimeslot flow that connects base station (BS) and mobilestation (MS).The position of these timeslots in TDMA frame isfixed. Figure 1-2 shows the complete structure of TDMA frame,including timeslot and burst. TDMA frame is a repetitive“physical” frame in radio link.One TDMA frame consists of eight basic timeslots, about60/13≈4.615ms in total. Each timeslot is a basic physicalchannel with 156.25 elements, coving 15/26≈0.557ms.There are two kinds of multiframes, consisting of 26 and 51continuous TDMA frames respectively. Multiframes are appliedwhen different logical channels are multiple used in onephysical channel.The 26 multiframe, with a period of 120 ms, is used in trafficchannel and associated control channel. Among the 26 bursts,24 are used in traffic and 2 are used in signaling.

The 51 multiframe, with a period of 3060/13≈235.385 ms, isspecially used in control channel.Many multiframes together form a super frame. Super frame isa continuous 51×26TDMA frame, that is to say, a super frameconsists of fifty-one 26 TDMA multiframes or twenty-six 51TOMA multiframes. The period of super frame is 1,326 TDMAframes, or 6.12 s.Many super frames together form a hyper frame.A hyper frame consists of 2,048 super frames with a period of12,533.7s, or 3 hours and 28’ 53’’ 760’’’. It is used in encryptedvoice and data. Each period of hyper frame consists of2,715,648 TDMA frames numbered from 0 to 2,715,648. Theframe number is transmitted in sync channel.

1.2.3 BurstBurst is the message layout of a timeslot in TDMA channel,which means each burst is sent to a timeslot of TDMA frame.Different message in the burst determines its layout.There are five kinds of bursts:Normal burst: used to carry messages in TCH, FACCH, SACCH,SDCCH, BCCH, PCH and AGCH channelsAccess burst: used to carry message in RACH channelFrequency correction burst: used to carry message in FCCHchannelSynchronization burst: used to carry message in SCH channelDummy burst: transmitted when no specific messagetransmission request from system (In cells, standard frequencysends message continuously)Each kind of burst includes the following elements:Tail bits: Its value is always 0 to help equalizer judge start bitand stop bit to avoid lost synchronization.Information bits: It is used to describe traffic and signalinginformation, except idle burst and frequency correction burst.Training sequence: It is a known sequence, used for equalizerto generate channel model (a way to eliminate dispersion).

Training sequence is known by both transmitter and receiver. Itcan be used to identify the location of other bits from the sameburst and roughly estimate the interference situation oftransmission channel when the receiver gets this sequence.Training sequence can be divided into eight categories innormal burst. It usually has the same BCC setting with cells, butwhen accessed to burst and synchronization bust, trainingsequence is fixed and does not change with cells. For example,in access burst, training sequence is fixed (occupying 41 bits).The 36-bit message digit of the random access burst includesBSIC information of the cell. BSIC settings of the same BCCHshould be different, in order to avoid mis-decoding of randomaccess burst from neighboring cells into local access.Guard period: It is a blank space. Since each carrier frequencycan carry a maximum of eight subscribers, it is necessary toguarantee the non-overlapping of each timeslot intransmission. Although timing advance technology (introducedlater) is used, bursts from different mobile stations still showlittle slips; therefore, protection interval is adopted to allowtransmitter to fluctuate in a proper range in GSM. On the otherhand, GSM requires protection bits to keep constanttransmission amplitude of the effective burst (exceptprotection bits) and properly attenuate the transmissionamplitude of mobile station. The amplitude attenuation of twosequential bursts as well as proper modulation bit stream canreduce the interference to other RF channels.The following is a detailed introduction to the structure andcontent of burst:Access burstIt is used for random access (channel request from network andswitchover access).It is the first burst that the base station needs in uplinkmodulation.Access burst includes a 41-bit training sequence, 36-information bit, and its protection interval is 68.25 bits. There is

only one kind of training sequence in access burst. Since thepossibility of interference is rather little, it is unnecessary toadd extra kinds of training sequences. Both training sequenceand protection interval are longer than normal bursts in orderto offset the bug of timing advance ignorance in the first accessof mobile station (or switch over to another BTS) and improvedemodulation ability of the system.Frequency correction burstIt is used for frequency synchronization in mobile station, equalto an unmodulated carrier. This sequence has 142 constant bitsfor frequency synchronization. Its structure is pretty simplewith all constant bits being 0. After modulated, it becomes apure sine wave. It is used in FCCH channel for mobile station tofind and modulate synchronization burst of the same cell.When mobile station gets the frequency through this burst, itcan read the information of following bursts (such as SCH andBCCH) in the same physical channel. Protection interval and tailbit are the same with that of normal burst.Synchronization burst:With a 64-bit training sequence and two 39-bit informationfields, synchronization burst is used for time synchronization ofmobile station in SCH channel. It belongs to downlink. Since it isthe first burst required to be modulated by mobile station, itstraining sequence is relatively long and easy to be detected.Normal burstIt has two 58-bit groups used in message field. To be morespecific, two 58-bit groups are used to transmit subscriber dataor voice together with two stealing flags. Normal burst is usedto describe whether the transmitted is traffic information orsignaling information. For example, to distinguish TCH andFACCH (when TCH channel is used as FACCH channel totransmit signaling, the stealing flag of the 8 half bursts shouldbe set to 1. It has no other use in channels except in TCHchannel, but can be regarded as the extension of trainingsequence and always set to 1.Normal burst also includes two 3-

bit tails and a protection interval of 8.25 bits. The only bug isthat the receiver has to store the preceding part of burst beforemodulation. Normal burst has a total of 26 bits, 16 of which areinformation bits. In order to get 26 bits, it copies the first fivebits to the end of the training sequence and the last five bits tothe head of the training sequence. There are eight kinds of suchtraining sequence (these eight sequences have the leastrelevancy with each other). They correspond to different basestation color code (BCC, 3 bits) respectively to distinguish thetwo cells using the same frequency.

Dummy burst:This kind of burst is sometimes sent by BTS without carryingany information. Its format is the same with normal burst. Theencrypted bits are changed into mixed bits with certain bitmodel.

1.2.4 Logical ChannelIn real networking, each cell has several carrier frequencies andeach frequency has eight timeslots, proving eight basic physicalchannels. Logical channel carries out time multiplexing in onephysical channel. It is classified according to the type ofinformation in physical channel. Different logical channeltransmits different type of information between BS and MS,such as signaling and data service. GSM defines different bursttype for different logical channel.In GSM, logical channel is divided into dedicated channel (DCH)and common channel (CCH), or traffic channel (TCH) andcontrol channel (CCH) sometimes.

I. TCHTCH carries coded voice or subscriber data. It is divided into fullrate TCH (TCH/F) and half rate TCH (TCH/H) with 22.8 bit/s

information and 11.4 Kbit/s information respectively. Using halfof the timeslots in TCH/F can get TCH/H. A carrier frequencycan provide eight kinds of TCH/F or sixteen kinds of TCH/H.Voice channel types are as follows:Enhanced full rate speech TCH (TCH/EFS)Full rate speech TCH (TCH/EFS)Full rate 9.6 Kbit/s TCH (TCH/F9.6)Full rate 4.8 Kbit/s TCH (TCH/F4.8)Full rate ≤2.4 Kbit/s TCH (TCH/F2.4)II. CCHCCH is used to transmit signaling or synchronous data. It mainlyconsists of broadcast channel (BCCH), common control channel(CCCH), and dedicated control channel (DCCH).III. BCCHFrequency Correction Channel (FCCH)It carries the information for frequency correction in mobilestation. Through FCCH, mobile station can locate a cell anddemodulate other information in the same cell, and recognizewhether this carrier frequency is BCCH or not.Sync Channel (SCH)After FCCH decoding, mobile station has to decode SCHinformation. This information contains mobile station framesynchronization and base station identification. Base stationidentification code (BSIC) occupies six bits, three of which arePLMN color codes ranging from zero to seven, and the otherthree are base station color codes (BCCs) ranging from zero toseven.Reduced TDMA frame (RFN) occupies 22 bits.BCCHGenerally, each BTS has a transceiver containing BCCH in orderto broadcast system information to mobile station. Systeminformation enables mobile station to work efficiently in nullstate.IV. CCCHPaging Channel (PCH)

PCH is a downlink channel used to page mobile station. Whenthe network wants to communicate with a certain mobilestation, it sends paging information marked as TMSI or IMSIthrough PCH to all the cells in LAC area according to the currentLAC registered in mobile station.Access Grant Channel (AGCH)AGCH is a downlink channel used for base station to respondthe network access request of mobile station, that is, toallocate a SDCCH or TCH directly. AGCH and PCH share thesame radio resource. Keep a fixed number of blocks for AGCHor just borrow PCH when AGCH requires without keepingspecial AGCH block (AGB).Random Access Channel (RACH)RACH is an uplink channel used for mobile station to requestSDCCH allocation in random network access application. Therequest includes the reason to build 3-bit (call request, pagingresponse, location update request and short message request)and 5-bit reference random number for mobile station toidentify its own access grant message.V. DCCHStand-alone Dedicated Control Channel (SDCCH)SDCCH is a bi-directional dedicated channel used to transmitinformation of signaling, location update, short message,authentication, encrypted command, channel allocation, andcomplementary services. It can be divided into SD/8 and SD/4.Slow Associated Control Channel (SACCH)SACCH works with traffic channel or SDCCH to transmitsubscriber information and some specific information at thesame time. Uplink mainly transmits radio measurement reportand the first layer head information; downlink mainly transmitspart system information and the first layer head information.The information includes quality of communications, LAI, CELLID, BCCH signal strength in neighboring cells, NCC limit, celloptions, TA, and power control level.Fast Associated Control Channel (FACCH)

FACCH works with TCH to provide signaling information with arate and timeliness much higher than that provided by SACCH.There is another control channel called cell broadcast channel(CBCH) besides the three control channels mentioned above. Itis used in downlink and carries short message service cellbroadcast (SMSCB) information. CBCH uses a physical channelsame as SDCCH.VI. Channel CombinationLogical channel is mapped to physical channel according tocertain rules. The channel combinations specified in GSMprotocol are as follows:TCH/F + FACCH/F + SACCH/TFTCH/H(0,1) + FACCH/H(0,1) + SACCH/TH(0,1)TCH/H(0,0) + FACCH/H(0,1) + SACCH/TH(0,1) + TCH/H(1,1)FCCH + SCH + BCCH + CCCH (main BCCH)FCCH + SCH + BCCH + CCCH + SDCCH/4(0..3) +SACCH/C4(0..3)(BCCH combination)BCCH + CCCH(BCCH extension)SDCCH/8(0. .7) + SACCH/C8(0. .7)VII. Uncombined BCCH/SDCCH and Combined BCCH/SDCCHPaging information transmits in the timeslot 0 of BCCH.Timeslot 0 has the following sub channels:Broadcast channel (BCH): FCCH, SCH, BCCHCCCH: PCH, AGCHDCCH (combined BCCH/SDCCH): SDCCH, SACCH, CBCH ( if usingcell broadcast)Physical channel timeslot 0 is made of multiframes logically.Each multiframe is 235.4 ms in length. Multiframe has differentchannel configurations, such as combined BCCH/SDCCH anduncombined BCCH/SDCCH. Different configuration has differentpaging capacity.Uncombined BCCH/SDCCHEach frame of Uncombined BCCH/SDCCH can have nine pagingblocks. The timeslot 0 of BCCH carrier frequency does not haveSDCCH channel or CBCH channel.

Combined BCCH/SDCCHEach multiframe of combined BCCH/SDCCH can have threepaging blocks. The timeslot 0 of BCCH carrier frequencycontains four SDCCH subchannels (no CBCH) or three SDCCHand one CBCH subchannel.The configuration of combined BCCH/SDCCH has a greatinfluence on paging capacity. Each multiframe has only threepaging blocks instead of nine in uncombined BCCH/SDCCH,which means the paging capacity of cells with combinedBCCH/SDCCH is only one third of that of cells with uncombinedBCCH/SDCCH.

1.3 Data TransmissionRadio channel has totally different characteristics fromwired channel. Radio channel has a strong time-varyingcharacteristic. It has a high error rate when the signal isinfluenced by interferences, multipath fading, or shadowfading. In order to solve these problems, it is necessary toprotect the signals through a series of transformation andinverse transformation from original subscriber data orsignaling data to the information carried by radio waveand then to subscriber data or signaling data. Thesetransformations include channel coding and decoding,interleaving and de-interleaving, burst formatting,encryption and decryption, modulation anddemodulation.1.3.1 Voice CodingModern digital communication system usually uses voicecompression technology. GSM takes tone and noise fromhuman throat as well as the mouth and tongue filtereffect of acoustics as voice encoder to establish a model.The model parameters transmit through TCH channel.Voice encoder is based on residual excited linearprediction encoder (REIP) and its compression effect is

strengthened through long term predictor (LTP). LTPimproves residual data encoding by removing the vowelpart of voice.Voice encoder divides voice into several 20 ms voiceblocks and samples each block with 8 kHz, so each blockhas 160 samples. Each sample is quantified throughfrequency A 13 bits (frequency μ 14 bits). Since thecompression rates of frequency A and frequency μ aredifferent, add three and two “0” bits to the quantificationvalues respectively, and then each sample gets 16 bitsquantification value. Therefore, 128 Kbit/s data flow isobtained after digitizing but before encoding. This dataflow is too fast to transmit in radio path and has to becompressed in encoder. With full speed encoder, eachvoice block is encoded into 260 bits to form a 13 Kbit/ssource coding rate. Next is channel coding. With 20 msas a unit, 260 bits are output after compression encoding,so the encoding rate is 13Kbit /s.Compared with the direct coding transmission of voice intraditional PCM channel, the 13kbps voice rate of GSMis much lower. More advance voice encoder can reducethe rate to 6.5kbps (half rate encoding).1.3.2 Channel CodingChannel coding is used to improve transmission qualityand remove the influence of interferential factors onsignals at the price of increasing bits and information.The basic way of coding is adding some redundantinformation to the original data. The added data iscalculated on the basis of original data with certain rules.The decoding process of receiving end is judging andcorrecting errors with this redundant bit. If the redundantbit of received data calculated with the same way isdifferent from the received redundant bit, errors must

have occurred in transmission. Different code is used indifferent transmission mode. In practice, several codingschemes are always combined together. Common codingschemes include block convolutional code, errorcorrecting cyclic code and parity code.In GSM, each logical channel has its own coding andinterleaving mode, but the principle is trying to form aunified coding structure.Encode information bit into a unified block codeconsisting of information bits and parity check bits.Encode block code into convolutional code and formcoding bits (usually 456 bits).Reassemble and interleave coding bits and add a stealingflag to form interleaving bits.All these operations are based on block. The block sizedepends on channel type. After channel coding, allchannels (except RACH and SCH) are made of 464-bitblock, that is, 456 coded information bits plus 8-bitheader (header is used to distinguish TCH and FACCH).Then these blocks are reinterleaved (concerningchannel).In TCH/F voice service; this block carries one speechframe of information. In control channel, this blockusually carries one piece of information. In TCH/H voiceservice, speech information is transmitted by a block of228 coded bits block.For FACCH, each block of 456 coded information bits isdivided into eight sub blocks. The first four sub blocksare transmitted by even bits of the four timeslotsborrowed from the continuous frames of TCH, and therest four sub blocks borrows odd bits of the fourtimeslots from the four continuous frames delayed fortwo or four frames after the first frame. Each 456 coded

bit block has a stealing flag (8 bits), indicating whetherthe block belongs to TCH or to FACCH. In the case ofSACCH, BCCH or CCCH, this stealing flag is dummy.The synchronous information in Downlink SCH and therandom access information in uplink use short coded bitblocks transmitted in the same timeslot.In TCH/F, a 20ms speech frame is encoded into 456-bitcode sequence. The 260 bits of the 13 Kbit/s 20msspeech frame can be divided into three categories: 50most import bits, 132 important bits and 78 unimportantbits. Add 3 parity check bits to the 50 most importantbits, and these 53 bits together with 132 important bitsand 4 tail bits are convolutionally encoded ( with 1/2convolutional coding rate ) into 378 bits, plus the 78unimportant bits, and the 456 bits code sequence isobtained.In BCCH, PCH, AGCH, SDCCH, FACCH and SACCH,data is transmitted by Link Access Procedure on the Dmchannel (LAPDm). Each LAPDm frame has 184 bits,together with 40 bits error correcting cyclic code and 4tail bits, through 1/2 convolutional coding rate, and the456 bits code sequence is obtained.Each SCH contains 25-bit message field. Among them,19 bits are frame number and 6 bits are BSC number.These 25 bits plus 10 parity check bits and 4 tail bits are39 bits. Through 1/2 rate convolutional coding, 78 bitsare obtained, which occupy an entire SCH burst. .RACH message only has 8 bits, including 3-bit setupcause message and 5-bit discrimination symbol. On thebasis of these 8 bits, add 6 bits of color code (obtainedthrough the MOD 2 of the 6-bit BSIC and 6-bit paritycheck code), plus 4 tail bits to get 18 bits. Through 1/2rate convolutional coding, 36 bits are obtained, which

occupy an entire RACH burst.。1.3.3 InterleavingIf speech signal is modulated and transmitted directlyafter channel coding, due to parametric variation ofmobile communication channel, the long trough of deepfeeding will affect the succeeding bits, leading to errorbit strings. That is to say, after coding, speech signalturns into sequential frames, while in transmission, errorbits usually occur suddenly, which will affect theaccuracy of continuous frames. Channel coding onlyworks for detection and correction of signal error or shorterror string. Therefore, it is hoped to find a way toseparate the continuous bits in a message, that is, totransmit the continuous bits in a discontinuous mode soas to change the error channel into discrete channel.Therefore, even if an error occurs, it is only about asingle or very short bit stream and will not interrupt thedecoding of the entire burst or even the entireinformation block. Channel coding will correct the errorbit under such circumstances. This method is calledinterleaving technology. Interleaving technology is themost effective code grouping method to separate errorcodes.The essence of interleaving is to disperse the b bits into nbursts in order to change the adjacent relationshipbetween bits. Greater n value leads to better transmissionperformance but longer transmission delay. Therefore,these two factors must be considered in interleaving.Interleaving is always related to the use of channel. GSMadopts secondary interleaving method.After channel coding, The 456 bits are divided into eightgroups; each group contains 57 bits. This is the firstinterleaving, also called internal interleaving. After first

interleaving, the continuity of information in a group isbroken. As one burst contains two groups of 57-bit voiceinformation, if the two-group 57 bits of a 20 ms voiceblock after first interleaving are inserted to the sameburst, the loss of this burst will lead to 25% loss of bitsfor this 20 ms voice block. Channel coding cannotrestore so much loss. Therefore, a secondaryinterleaving, also called inter-block interleaving, isrequired between two voice blocks.

After internal interleaving, the 456 bits of a voice blockB are divided into eight groups. Interleave the first fourgroups of voice block B (B0, B1, B2, and B3) with thelast four groups of voice block A (A4, A5, A6, and A6),and then (BO, A4), (B1, A5), (B2, A6), and (B3, A7)form four bursts. In order to break the consistency ofbits, put block A at even position and block B at oddposition of bursts, that is, to put B0 at odd position andA4 at even position. Similarly, interleave the last fourgroups of block B with the first four groups of block C.Therefore, a 20 ms speech frame is inserted into eightnormal bursts after secondary interleaving. Theses eightbursts are transmitted one by one, so the loss of one burstonly affects 12.5% voice bits. In addition, as these burstshave no relations with each other, they can be correctedby channel coding.The secondary interleaving of control channel (SACCH,FACCH, SDCCH, BCCH, PCH, or AGCH) is differentfrom voice interleaving which requires three voiceblocks. The 456-bit voice block is divided into eightgroups after internal interleaving (the same as that ofvoice block), and then the first four groups areinterleaved with the last four groups (the same

interleaving method as that of voice block) to get fourbursts.Interleaving is an effective way to avoid interference, butit has a long delay. In the transmission of a 20 ms voiceblock, the delay period is (9*8)-7=65 bursts (SACCHoccupying one burst), which is 37.5 ms. Therefore, MSand trunk circuit have echo cancellers added to removethe echo due to delay.1.3.4 EncryptionSecurity is a very important feature in digitaltransmission system. GSM provides high securitythrough transmission encryption. This kind of encryptioncan be used in voice, user data, and signaling. It is usedfor normal burst only and has nothing to do with datatype.Encryption is achieved by XOR operation of poisonrandom sequence (generated through A5 algorithm ofencryption key Kc and frame number) and the 114information bits of normal burst.The same poison random sequence generated atreceiving end and the received encryption sequencetogether produce the required data after XOR operation1.3.5 Modulation and DemodulationModulation and demodulation is the last step of signalprocessing. GSM modulation adopts GMSK technologywith BT being 0.3 at the speed of 270.833 Kbit/s andViterbi algorithm. The function of modulation is to add acertain feature to electromagnetic wave according to therules. This feature is the data to transmit. In GSM, thephase of electromagnetic field bears the information.The function of demodulation is to receive signals andrestore the data in a modulated electromagnetic wave. Abinary numeral has to be changed into a low-frequency

modulated signal first, and then into an electromagneticwave. Demodulation is the reverse process ofmodulation.

1.4 Timing advanceSignal transmission has a delay. If the MS moves away from BTSduring calling, the signal from BTS to MS will be delayed, so willthe signal from MS to BTS. If the delay is too long, the signal inone timeslot from MS cannot be correctly decoded, and thistimeslot may even overlap with the timeslot of the next signalfrom other MS, leading to inter-timeslot interference.Therefore, the report header carries the delay value measuredby MS. BTS monitors the arrive time of call and send commandto MS with the frequency of 480 ms, prompting MS the timingadvance (TA) value. The range of this value is 0–63(0–233 us),and the maximum coverage area is 35km. The calculation is asfollows:1/2×3.7us/bit×63bit*c=35km3.7us/bit is the duration per bit (156/577); 63bit is themaximum bit for time coordination; c is light velocity(transmission rate of signal); 1/2 is related to the round-trip ofsignal.According to the preceding description, 1bit to 554 m, due tothe influence of multi-path transmission and the accuracy ofMS synchronization, TA error may be about 3 bits (1.6km).Sometimes a greater coverage area is required, such as incoastal areas. Therefore, the number of channels that each TRXcontains must be reduced. The method is to bind odd and eventimeslots, so there are only four channels (0/1, 2/3, 4/5, and6/7) for each TDMA frame in extended cell. Allocate channels 0,2, 4, and 6 to MS. Within 35 KM around BTS, the TA value of MSis in the normal range 0-63; for the area beyond 35 KM, TAvalue stays at 63. This technology is called extended celltechnology. The maximum value of TA in BTS measurementreport is 63+156.25=219.25 bit, so the maximum radius of

coverage area is:1/2×3.7us× (63+156.25) ×3×108m/s=120km

In real scheme, in order to improve the utilization of TRX, bothcommon TRXs and dual timeslot TRXs can be included. BCCHmust be in dual timeslot TRX to receive random access fromany area. The calls within 35 km are allocated to common TRX;the calls within 35 km–120 km and the switched in calls areallocated to dual timeslot TRX. If the system detects theswitched in call is within 35km, it will switch over this call tocommon TRX. If the MS in conversation goes beyond 35 km, anintra-cell switchover will be carried out. Therefore, both thecapacity requirement for remote areas and the coveragerequirement for local areas can be satisfied.

1.5 System InformationSystem information is sent to MS from network in broadcastform. It informs all the MSs within the coverage area oflocation area, cell selection and re-selection, neighbor cellinformation, channel allocation and random access control. Byreceiving system information, MS can quickly and accuratelylocate network resources and make full use of all kinds ofservices that network provides. There are 16 types of systeminformation: type1, 2, 2bis, 2ter, 3, 4, 5, 5bis, 5ter, 6, 7, 8, and13.System information is transmitted on BCCH or SACCH. MSreceives system information in different mode from differentlogic channel.In idle mode, system information 1– 4, 7, and 8 are transmittedon BCCH ;In communication mode, system information 5 and 6 aretransmitted on SACCH;The content of system information is as follows:System information 1：cell channel description + RACH controlparameter, transmitted on BCCH

System information 2： frequency description of neighbor cell+ RACH control information + network color code (NCC)permitted, transmitted on BCCH, used for cell re-selectionSystem information 2bis： Extended neighbor cell BCCHfrequency description + RACH control information, transmittedon BCCH, used for cell re-selection.System information 2ter： Extended neighbor cell BCCHfrequency description, transmitted on BCCH, used for cell re-selection.System information 3： Cell identity + location area identity(LAI) + control channel description + cell selection + cellselection parameter + RACH control parameter, transmitted onBCCH.System information 4： LAI + cell selection parameter + RACHcontrol parameter + CBCH channel description + CBCH mobileconfiguration, transmitted on BCCH.System information 5： Neighbor cell BCCH frequencydescription, transmitted on SACCH channel, used for cellhandover.System information 5bis： Extended neighbor cell BCCHfrequency description, transmitted on SACCH channel, used forcell handover.System information 5ter： Extended neighbor cell BCCHfrequency description, transmitted on SACCH channel, used forcell handover.System information 6： Cell Global Identification (CGI) + celloption＋NCC Permitted, transmitted on SACCH.System information 7： cell re-selection parameterSystem information 8： cell re-selection parameterBCCH is a low-capacity channel, every 51 multiframes ((235 ms)have only four frames (one information block) to transmit a 23byte LAPDm message.Each information unit contains:Cell channel description contains all the frequencies used in

this cell.RACH control information contains parameters such as MaxRetrans, TX_integer, CBA, RE, EC, and AC CN.Neighbor cell BCCH frequency description contains the BCCHfrequency that the neighbor cell uses.Allowed PLMN is used to provide NCC Permitted that MSmonitors on BCCH TRX.Control channel description contains parameters such as MSATTACH/DEATTACH allowed Indicator ATT, BS-AG-BLKS-RES,CCCH-CONF, BA-PA-MFRMS, and T3212.Cell selection contains parameters such as power control(PWRC) indication, discontinuous Transmission (DTX)indication, and RADIO-LINK-TIMEOUT.Cell selection parameter contains parameters such as cell re-selection hysteresis, MS-TXPWR-MAX-CCH, and RXLEV-ACCESS-MIN.CBCH channel description contains channel type and TDMAdeviation (the combination mode of dedicated channel),timeslot number (TN), training sequence code (TSC), hoppingfrequency channel indication H, mobile allocation index offset(MAIO), hopping frequency sequence number (HSN) andabsolute radio frequency channel number ( ARFCN).CBCH mobile configuration contains the relationship betweenhopping channel sequence and cell channel description.Cell re-selection parameter contains CELLRESELIND, cell barqualify (CBQ), cell reselection offset (CRO), temporary offset(TO), and penalty time (PT).

1.6 Cell Selection and Re-Selection1.6.1 Cell SelectionWhen a MS is switched on, it tries to contact GSM PLMN thatthe SIM permits and select a proper cell to extract controlchannel parameters and other system information. This processis called cell selection.The priority levels of cells include normal, low, and barred. Lowpriority level cell is selected when there is no proper normal

cell.A proper cell means:The cell belongs to the selected network;The cell is not barred;The cell is not in the national prohibited roaming location area;The path loss between MS and BTS is under the limit set bynetwork.The priority level of a cell is determined by CELL_BAR_QUALIFY(CBQ) and CELL_BAR_ACCESS (CBA).1.6.2 Cell Selection ProcessTo perform cell selection and re-selection, MS requires all thefrequencies monitored to stay at the unweighted average valueof Relev RLA_C.I. Cell Selection When MS Storing No BCCH InformationMS searches all RF channels (at least 30 channels for 900 M, 40for 1800 M, and 40 for PSC1900) in the system to obtain theRelev of each RF channel, and calculate the RLA_C based on atleast five samples in three to five seconds, and then arrangethese levels in descending order to select the proper BCCH. MSselects the cells with normal priority first. If the proper cellshave low priority, MS will select the cell with the highest Relev.MS has already decoded and identified all these frequencies bynow. If there is no proper cell, MS will keep on searching. Ittakes a maximum of 0.5 s to synchronize a BCCH TRX and 1.9 sto read the synchronized BCCH TRX data, except that it takesn*1.9s(n>1)to obtain the system information.II. Cell Selection When MS Storing BCCH InformationIf MS stores the BCCH frequency list of the former selectednetworks, MS will perform measurement sampling procedure(only for the stored BCCH TRX) according to this list. If the cellselection within this list fails, common cell selection will beperformed. If all the cells have low priority level, MS will selectthe cell with the highest Relev. MS has already decoded andidentified all these frequencies by now. When a 900 M MSenters the 900/1800 network, MS will probably choose 900 M

network and ignore the priority level, because the MS stores allthe 900 M frequency information in BCCH frequency list.III. Cell Selection CriteriaParameter C1 is the path loss criteria for cell selection, C1 ofthe service cell must exceed 0, the formula is as follows:C1= RLA_C - RXLEV_ACCESS_MIN- MAX((MS_TXPWR_MAX_CCH- P), 0) (2-1)For DCS 1800 cells:C1 = RLA_C - RXLEV_ACCESS_MIN- MAX((MS_TXPWR_MAX_CCH + POWER OFFSET- P), 0)In the formula:RLA_C: Average value of RelevRXLEV_ACCESS_MIN: Minimum Relev that MS allowsMS_TXPWR_MAX_CCH: Maximum transmit power on controlchannelP: Maximum transmit power of MSPOWER OFFSET：Power offset related toMS_TXPWR_MAX_CCH used by DCS1800 cells.1.6.3 Down Link FailureDownlink failure criteria are based on DSC. When a mobilephone stays in a cell, DSC is initialized to an integer most closeto 90/N ( N is BS_PA_MFRMS, range value: 2–9). Each timewhen mobile phone successfully decodes a message on itspaging subchannel, DSC increases by 1, but DSC cannot exceedthe initial value; when decoding fails, DSC decreases by 4.When DSC<=0, downlink failure occurs. Down signaling linkfailure will lead to cell re-selection.1.6.4 Cell Re-Selection ProcessIn cell re-selection, mobile phone will synchronize and read theinformation from six BCCH TRXs (in BA list) with strongestsignals outside the service area. For multi-frequency mobilephones, the TRXs with strongest signals may be in differentfrequency bands.In idle mode, mobile phone monitors all the BCCH TRXs in BAlist and averages each Relev from BCCH TRX within 5 s to Max

{5, ((5 * N + 6) DIV 7) * BS_PA_MFRMS / 4} s. N is the numberof BCCH TRXs outside service area in BA list. Each RLA_Crequires at least five level measurement samples and has to beupdated from time to time. Service area samples the Relev atleast once for each paging block to mobile. RLA_C is calculatedby averaging the level samples received from 5s to Max {5s, fiveconsecutive paging blocks of that MS}.Each RLA_C update is followed by the update of the six BCCHTRXs outside the service area in BA list. And the latter updatemay be even faster.Mobile phone decodes all the BCCH data in a service cell everyother 30 s and the BCCH data blocks related to cell re-selectionparameters of the six BCCH TRXs with strongest signals everyother five minutes. When the mobile phone detects that a newBCCH TRX becomes one of the six TRXs with strongest signals,this BCCH TRX data should be decoded within 30 s. Mobilephone checks the BSICs of the six BCCH TRXs with strongestsignals to make sure they are in the same cell. If the BSIC of aTRX is changed, the MS will regard the TRX as new TRX andreread the BCCH data.MS will re-select a neighbor cell as service cell under certaincondition. This condition includes several factors, such asRLA_C, cell restriction (decided by cell_bar andcell_bar_qualify), and access state of the neighbor cell.Cell re-selection adopts C2 algorithm. The calculation formula isas follows:When PENALTY TIME is not 11111C2=C1+CELL_RESELECT_OFFSET–TEMPORARY_OFFSET*H(PENALTY_TIME–T);When PENALTY_TIME is 11111C2=C1-CELL_RESELECT_OFFSET.When X>0, function H(x) =0; when X≤O, function H(x) =1.T is a timer; its initial value is 0. When a cell is included in thesix neighbor cells with strongest signals by MS, the timer T ofthis cell begins to time; when a cell is excluded from the six

neighbor cells with strongest signals by MS, T will be reset.CELL_RESELECT_OFFSET adjusts the value of C2.After T starts, TEMPORARY_OFFSET will modify the C2algorithm according to the defined value before the penaltytime in order to avoid a micro cell or a cell with small coveragearea is selected by a fast moving MS. If the defined penaltytime is out, the temporary offset will be ignored. Penalty timecan avoid the frequent cell re-selection in those coverage areaslike express highway.These parameters in C2 algorithm works only whenCELL_RESELECTION_INDICATION is activated. Otherwise, MSwill ignore the setting of CELL_RESELECT_OFFSET,TEMPORARY_OFFSET, and PENALTY_TIME, under suchcircumstances, C2=C1.Cell re-selection will be triggered under the followingconditions:The C2 value of a certain cell (belonging to the same locationarea with the current cell) exceeds that of the current cell by 5seconds successively;The C2 value of a certain cell (belonging to different locationarea from the current cell) exceeds the sum of the C2 value ofthe current service cell and cell selection hysteresis value by 5seconds successively;The current service cell is barred;MS detects downlink failure;The C1 value of the service cell is less than 0 for 5 secondssuccessively.

1.7 Frequency HoppingWith the ever growing traffic volume and the limitedfrequency resource, frequency reuse is more and moreaggressive. Therefore, the problem of how to reducefrequency interference becomes more and moreremarkable. The essence of anti-interference is to fully

utilize the current spectrum, time domain, and spaceresources. The key measures include frequency hopping,discontinuous transmission (DTX), and power control.Frequency hopping also can effectively reduce theinfluence of fast fading.1.7.1 Types of Frequency HoppingGSM radio interface uses slow frequency hopping (SFH)technology. The difference between slow frequencyhopping and fast frequency hopping is that the frequencyof latter changes faster than frequency modulation. InGSM, the frequency remains the same during bursttransmission. Therefore, GSM frequency hoppingbelongs to slow frequency hopping.In frequency hopping, the carrier frequency is controlledby a sequence and hops with time. This sequence isfrequency hopping sequence. Frequency hoppingsequence is a sequence of frequencies decided byhopping sequence number (HSN), mobile allocationindex offset (MAIO) and frame number (FN) through acertain algorithm in the mobile allocation containing Nfrequencies. The N channels of different timeslots canuse the same hopping sequence. The different channelsof the same timeslot in the same cell adopt differentMAIO.Frequency hopping can be divided into frame hoppingand timeslot hopping according to time domain and RFhoping and baseband hopping according toimplementation mode.Frame hopping: the hopping frequency changes once ineach TDMA frame period. Each TRX can be regarded asa channel. The TCH of BCCH TRX cannot join in thefrequency hopping in a cell. The hopping TRX shouldhave a different MAIO. Frame hopping is an exception

of timeslot hopping.Timeslot hopping: the timeslot frequency of each TDMAframe changes once. The TCH of BCCH TRX can join inthe frequency hopping, which happens in basebandhopping.RF hopping: both transmission and reception of TRXjoin in the frequency hopping. The number hoppingfrequencies can exceed the number of TRXs in the cell.Baseband hopping: each transceiver works at a fixedfrequency. TX does not join in frequency hopping.Frequency hopping is performed through the handover ofbanseband signal. Therefore, the number of hoppingfrequencies cannot exceed the number of TRXs in thecell.The two frequency hopping modes above are based onBTS. As for MS, since each MS has only one TRX unit,RF hopping is the only mode.I. Baseband HoppingThe system has multiple baseband and TRX processingunit. Each TRX processing unit has a fixed workingfrequency; each baseband processing unit processes oneline of service information and sends the processedinformation to the TRX unit with bus topology in timesequence according to frequency hopping rule. This kindof frequency hopping is called “baseband hopping”.In baseband hopping, each transceiver works with a fixedfrequency. The bursts on the same speech path are sent toeach transceiver. Baseband hopping is based on thehandover of baseband signals. Since the transceiver ofeach BTS has a fixed working frequency, bothbroadband combiner and cavity combiner can beadopted. The number of TRXs decides the maximumnumber of frequency hopping. The problem for baseband

hopping is that if one TRX board fails, the correspondingcode word will be lost, thus affecting all the calls underhopping mode in the cell.II. RF HoppingUnder this mode, each line of service information isprocessed by fixed baseband unit and frequency bandunit. The working frequency of frequency band unit isprovided by frequency combiner. Under the control ofcontrol unit, frequency can be changed according tocertain rules. In RF hopping, the frequencies used by aTRX to handle all the bursts of a call come from thefrequency change of combiner, instead of the handoverof baseband signals. The number of TRXs is not limitedby carrier frequency. As the working frequency of TRXchanges, which means the frequency of the input port tocombiner changes, only broadband combiner can beadopted. This kind of broadband combiner leads to about3dB insertion loss in two-in-one combination and theloss is greater in the link insertion of multi-combiner.GSM protocol does not specify which kind of frequencyhopping is used in GSM BTS. The mode of frequencyhopping can be decided by operators according to theequipments.1.7.2 Frequency Hopping AlgorithmThe parameters related to frequency hopping algorithmare as follows:CA: cell allocation, the collection of frequencies used bya cellFN: TDMA frame number, broadcasted on sync channel.FN (0–2715647) synchronizes BTS with MSMA: mobile allocation, the collection of radiofrequencies used for MS frequency hopping. It is a subsetof CA. MA contains N frequencies, 1≤N≤64.

MAIO: mobile allocation index offset, (0–N-1). Duringcommunication, the radio frequency at air interface is anelement of MA. Mobile allocation index (MAI, 0–N-1) isused to determine the element of MA. That is to say, theactual frequency used is decided by MAI. MAIO is theinitial offset of MAI and it is used to avoid thecontention of frequency by several channels at the sametime.HSN: hopping sequence number (0–63). It determinesthat the hopping sequence with concentrated frequenciesis adopted in frequency hopping. When HSN=0, thehopping is cyclic hopping; when HSN≠0, the hopping israndom hopping.The proper setting of parameters is based on theunderstanding of the use of each parameter in hoppingalgorithm and the hopping theory. The proper settingensures the healthy working state of the system.Remarks: For the cyclic hopping in discontinuoustransmission (DTX), the number of hopping frequenciesshould avoid N mod 13 = 0, because under suchcondition, the probability of transmission andmeasurement of SACCH frame at the same frequency israther high, and the harms are obvious.When HSN=0, S equals the frame number, in othercases, S is only related to frame number and frequencyhopping number. When HSN is fixed and frame numberis the same, S must be the same. Therefore, as the TRXsof each sync cell have the same frame number, differenthopping groups in sync cells can adopt the same HSN. Aproper configuration of MAIO can avoid the inter-cell orintra-cell frequency collision within the same BTS. Theaggressive frequency reuse adopts this theory.1.7.3 Benefits of Frequency Hopping

In GSM, frequency hopping has two benefits: frequencydiversity and interference averaging.I. Frequency DiversityFrequency hopping can reduce the influence of signalstrength change due to multipath transmission. Thiseffect equals that of frequency diversity. In mobilecommunications, Rayleigh fading leads to the greatchange of radio signal in a short time. This kind ofchange is related to frequency: a more independentfading accompanies a greater frequency difference. The200 KHz interval generally ensures the independence ofinter-frequency fading, while the 1 MHz interval canfully guarantee this kind of independence. Throughfrequency hopping, all the bursts containing the codeword of the same speech frame are protected from thedamage of Rayleigh fading in the same way.Statistics shows that frequency hopping gain is related toenvironmental factors, especially to the moving speed ofMS. When the MS moves at a high speed, the locationdifference between two bursts on the same channel isalso affected by other kinds of fading. The higher thespeed is, the lower the gain will be. Frequency diversitybenefits a lot to a large number of MSs moving at lowspeed.Frequency hopping gain is also related to the number offrequencies. When the number of frequencies decreases,the hopping gain falls. The relationship between thenumber of frequencies and hopping gain can beexplained in this way: frequency hopping is pseudospectrum spread, and the hopping gain is the processinggain after transmission frequency band spread. The basicway to test frequency hopping gain is to calculate thedifferences between different C/I at different hopping

frequencies under the same FER. These C/I differencesare the frequency hopping gain.II. Interference AveragingFrequency hopping provides the diversity of interferenceon transmission channel, so that all the bursts containingthe code word of the same speech frame are protectedfrom the damage of interference in the same way.Through error correction coding and interleaving of thesystem, the original data can be restored from the restpart of the received flow. The hopping gain is obtainedonly when the interference is in narrowband distribution.If the interference is in broadband distribution, all thebursts will be destroyed and the original data cannot berestored. Therefore, no hopping gain is obtained. Thecommon interference after frequency hopping can beregarded in narrowband distribution.In frequency hopping, error rate tends to increase in thetest, but we feel the conversation quality improves. It isbecause although the error rate increases, the influence ofinterference is homogenized in frequency hopping, thespeech restoring ability improves because of theinterleaving and de-interleaving before. In GPRS dataservices, frequency hopping can be harmful when thedata rate is rather high (CS4).

1.8 Discontinuous Reception and Discontinuous Transmission1.8.1 Discontinuous Reception and Paging ChannelIn idle mode, if MS selects a cell as its service cell, it begins toreceive the paging information from this cell. But in order toreduce power consumption, discontinuous reception (DRX) isintroduced in GSM. Each user (IMSI) belongs to a paging groupand each paging group corresponds to a paging subchannel. MScan calculate which group it belongs to based on the last three

digits of its IMSI and the configuration of paging channel in thislocation area, and then locate the paging subchannel of thispaging group. In fact, in idle mode, MS just listens to the paginginformation from the system on its subchannel (MS alsomonitors the Relev of BCCH carrier frequency in non-servicearea during this period of time) and ignores the information onother paging subchannels. Some of the hardware equipmentsare even switched off to save the power of MS. But MS mustcomplete the required task of network informationmeasurement within a specified time.Through DRX, MS can receive the broadcast short messagesthat the users want to know with less power consumption, thusextending the service time. BSC has to send schedulingmessages to support DRX at MS. One scheduling messagecontains lots of broadcast short messages to be sent soon. Thetime that all broadcast short messages of a schedulinginformation takes is a scheduling cycle. Scheduling informationcontains the description of all short messages to be broadcastin order and also indicates the position of the messages inscheduling cycle. Through scheduling messages, MS can findthe broadcast short messages it wants quickly so as to reduceits power consumption.The number of paging subchannels of each cell can becalculated based on the configuration type of CCCH,BS_AG_BLKS_RES (the number of blocks belonging to AGCH in51 multiframe), and BS_PA_MFRMS (the number of 51multiframes used as one paging subchannel cycle).When there are three CCCHs in a 51 multiframe, the number ofpaging subchannels is (3- BS_AG_BLKS_RES) ×BS_PA_MFRMSWhen there are nine CCCHs in a 51 multiframe, the number ofpaging subchannels is (9- BS_AG_BLKS_RES)×BS_PA_MFRMSIn addition, the configuration of CCCH parameters has thefollowing principles:The greater the parameter BS_PA_MFRMS, the more thepaging subchannels, and the less the users of each paging

subchannel, but the total capacity of the system remains thesame, because the average delay of the paging information onradio channel increases. When the ratio of retransmissionwaiting is relatively high, BS_PA_MFRMS should be improved toincrease the paging subchannels; when the ratio ofretransmission waiting is relatively low, BS_PA_MFRMS shouldbe reduced to shorten the paging delay.The capacities of paging subchannels of all cells in a locationarea should be the same, because the paging message of alocation area must be sent in all the cells of this location area atthe same time.The longer the cycle of paging channel, the less power the MSin this service area takes. For example, in cities, this cycle canbe defined as 2, which means MS listens to paging messagesonce for every 102 frames. In rural areas, this cycle can bedefined as 4 or 6. The MS with the paging channel cycle of 6consumes 18% less power than the MS with the paging channelcycle of 2. After measuring the system information, MS entersthe rest state and listens to the paging information in thespecified paging blocks only and measures the Relev of BCCH ofneighbor cells at the same time. After 30 s, MS will listen tosystem information again to judge the cell re-selection process.In GSM, CCCH mainly includes AGCH and PCH. Its primaryfunction is to transmit immediate assignment messages andpaging messages. CCCH can be one or several physical channelsand it can also share a physical channel with SDCCH. Thecombination mode of CCCH depends on the parameterCCCH_CONF. The configuration of CCCH_CONF must beconsistent with the actual configuration. It is recommendedthat when there is only one TRX in a cell, the configuration ofCCCH can be a physical channel shared with SDCCH (3 CCCHinformation blocks).When the traffic volume is extremely large, in case one physicaltimeslot is not enough, GSM specification allows theconfiguration of multiple CCCH channels on the TRX besides

BCCH, but these channels must be used in timeslot 0, 2, 4, and6.When CCCH_CONF is confirmed, parameter BS_AG_BLKS_RESactually decides the ratio of AGCH and PCH on CCCH. It isrecommended that this parameter be configured as little aspossible in order to reduce the response time of MS to paging.1.8.2 DTXI. DTX OverviewDuring communication, only 40% time is used for conversation;no useful information is transmitted during the rest 60% time.If all the information is transmitted to network, many of thesystem resources will be wasted, in addition, the interferencewill aggravate. In order to solve this problem, GSM adopts DTXtechnology to stop signal transmission when there is no voicesignal. Therefore, the interference level is reduced and thesystem efficiency is improved.There are two kinds of transmission modes in GSM: normalmode and discontinuous transmission (DTX) mode. In normalmode, noise and voice have the same transmission quality. InDTX mode, the transmission of unuseful messages isprohibited. MS only sends man-made noise signals that aretolerable, which means this noise will not annoy the listenersnor affect the conversation. This kind of noise is called comfortnoise. In DTX mode, 260-bit code is transmitted in every 480ms; in normal mode, 260-bit code is transmitted in every 20ms.Whether the downlink DTX is adopted or not is controlled bynetwork operators of the exchange part. This kind of control isbased on BSC. The control information is transmitted tobaseband processing part through dedicated signaling channel,and then arrives at TC through the inband signaling of TRAUframe to indicate whether downlink DTX is adopted. For somevendors, the downlink DTX can be configured on the basis ofcell.Uplink DTX is configured by network operators of the radio

part. The parameter DTX in system information consists of 2bits.Parameter DTX is contained in “cell option” of information unitand transmitted periodically in the system information of eachcell broadcast. MS decides whether to start DTX function basedon this information.DTX can be used for voice signal transmission andnontransparent data transmission. BCCH TRX does not use thistechnology. The benefits of DTX are listed below:Uplink DTX can save MS batteries and reduce interference.Downlink DTX can save BTS power consumption and reduceinterference and intra-BTS intermodulation.Uplink DTX and downlink DTX used together can improve theintra-frequency ratio of the system. This kind of improvement,when used in aggressive-frequency-reuse cell planning,especially when used with frequency hopping, can greatlyexpand the system capacity.II. Voice Activity DetectionFor voice activity detection (VAD), the source must indicatewhen the transmission is required. When DTX mode isactivated, the encoder must detect the signal is voice or noise.Therefore, the VAD is required. VAD can differentiate voicefrom noise through calculating some signal parameters andthreshold values. This kind of differentiation is based on anenergy rule: the energy of noise is always lower than that ofvoice.VAD generates a group of threshold value in every 20 ms tojudge whether the next 20ms block is voice or noise. When thebackground noise is too loud, the noise signal will be regardedas voice signal to transmit.III. Silence IndicatorThe coding procedure of noise is the same as that of voice.After sampling and quantification, a noise block will be produceby encoder in every 20ms. Like voice block, the coded noiseblock also contains 260 bits, which forms a SID frame. The SID

frame will go through channel coding, interleaving, encryptionand modulation and finally be sent by eight continuous bursts.On TCH, a complete SACCH information block has four 26muliframe cycles (480 ms). In order to differentiate voice frameand SID frame, these eight continuous bursts are arranged atthe beginning of the third multiframe. During other time of the480 ms, no information is transmitted except SACCH timeslot.The SID frame made from the 20 ms noise block is interleavedwith the preceding frame and the following frame; the first SIDframe is interleaved with the preceding voice frame and thefollowing SID frame.IV. MeasurementUplink DTX and downlink DTX are two irrelevant proceduresthat are activated by system parameters respectively. There aretwo kinds of measurement in GSM: full measurement and submeasurement.Global measurement is the average of the level and quality ofthe 104 timeslots in a measurement cycle (four 26multiframes); local measurement is the average of level andquality of 12 timeslots, including eight continuous TCH bursts(for TCH/F, 0-103 TDMA frames as a cycle. The frame numbersof these eight bursts are 52, 53, 54, 55, 56, 57, 58, and 59.when no voice or signaling is transmitted, the descriptor ofcomfort noise they contain is called SID) and four SACCH bursts(0-103 TDMA frames as a cycle, for timeslot 0, the framenumbers of these four bursts are 12, 38, 64, and 90; fortimeslot 1, the frame number is that of timeslot 0 plus 13.similarly, the frame numbers that the eight timeslotscorrespond to can be obtained in this way). In order to achieveuniformity, no matter the uplink DTX or downlink DTX isactivated or not, BTS and MS must complete these two kinds ofmeasurement. Each SACCH measurement report of BTS and MSindicates whether DTX is used in last measurement report time.BSC choose one of the two kinds of measurement based on thisindication.

1.9 Power Control1.9.1 Power Control OverviewPower control is to change the transmission power of MS orBTS (or both) in radio mode within certain area. Power controlcan reduce the system interference and improve the spectrumutilization and prolong the service time of MS battery. Whenthe Relev and quality is good, the transmission power of thepeer end can be reduced to lower the interference to othercalls.In GSM, power control can be used in uplink and downlinkrespectively. The power control range for uplink MS is 20 dB–30dB. Based on the power class of MS (most MSs belongs toclass 4, which means the maximum transmission power is 33dbm), each step can change 2 dB. The downlink power controlrange is decided by equipment manufacturer. Althoughwhether to adopt uplink or downlink power control function isdecided by network operators, all MSs and BTS equipmentsmust support this function. BSS manages the power control inthe two directions.To facilitate BCCH frequency pull-in and the measurement ofRelev (including the Relev of neighbor cell BCCH frequency),GSM protocol specifies that no power control is allowed for thetimeslots in the downlink of BCCH TRX.1.9.2 MS Power ControlThe power control of MS includes two adjustment stages:stable adjustment stage and initial adjustment stage. Stableadjustment is the common way to implement power controlalgorithm. Initial adjustment is used at the beginning of callconnection. When a connection occurs, MS sends signals withnominal power (before receiving power adjustment commend,the nominal transmission power of MS is the maximumtransmission power on BCCH of the cell. If MS does not supportthis power level, it will adopt other power level most close tothis level, such as the maximum power level supported by theclassmark of MS in indication message establishment).

Therefore, MS accesses to network through RACH with themaximum power broadcast on BCCH. When MS power is lowerthan this value, it will transmit with its maximum transmissionpower. The system specifies that the power level of the firstmessage that MS sends on DCH is also this value. The systemcontrol begins after MS receives the power control command inSACCH information block from SDCCH or TCH.Since BTS can support multi-call at the same time, the Rxlevshould be quickly reduced in the new connection. Otherwise,other calls supported by this BTS will deteriorate and the callsin other cells will also be affected. The purpose of initialadjustment stage is to quickly reduce the transmission powerof MS to get the stable MR, so MS can be adjusted according tostable power control algorithm.The required parameters in uplink power control, the expecteduplink Rxlev, and the uplink received quality can be adjustedaccording to the situation of the cell. After receiving a certainnumber of uplink MRs, the system compares the actual uplinkRxlev and received quality obtained by interpolation, filtering,and other methods with the expected values and calculate thepower level that the MS should be adjusted to through powercontrol algorithm. If the calculated power level differs from theoutput power level of MS and meets certain limit conditions(such as step limit of power adjustment and range limit of MSoutput power), the system will send power adjustmentcommand.The command of changing MS power and the required timeadvance will be sent to MS in the layer 1 header of eachdownlink SACCH information block. MS will configure thepower level it uses now in its uplink SACCH information blockand send it to BTS in measurement report. This level is thepower level of the last burst in the previous SACCHmeasurement cycle. When MS receives the power controlinformation in SACCH information block from DCH, it willtransmit with this power level. One power control message

does not make the MS switch to the required levelimmediately. The maximum change rate of MS power is 2 dBfor every 60 ms. For 12 dB, before MS receives the next powercontrol message, it will not end as one SACCH measurementcycle takes 480 ms. In addition, it takes three measurementcycles to send power control message and execute thecommand. Therefore, the power control cycle should not betoo short in order to ensure its accuracy. See Figure 1-10.Figure 1-1 Execution of power control commandThe purpose of uplink power control adjustment is to minimizethe difference between the actual uplink Rxlev and receivedquality and the expected uplink Rxlev and received quality. Thepurpose of interpolation and filtering is to process the lostmeasurement reports and remove temporary nature to ensurethe stability of power control algorithm.

The difference between initial adjustment and stableadjustment is that the expected uplink Relev and receivedquality and the length of filter in initial adjustment are differentfrom that of stable adjustment, and the initial adjustment onlyhas downlink adjustment.1.9.3 BTS Power ControlBTS power control is an optional function. It is similar to MSpower control, but it only uses stable power control algorithm.The required parameters are Rxlev threshold (lower limit), andthe maximum transmission level can be received (upper limit).The Relev is divided into 64 levels ranging from 0 to 63. Level 0is the lowest Rxlev; level 63 is the highest Rxlev.BTS power control is divided into static power control anddynamic power control. Dynamic power control is the finetuning based on static power control. There are six steps (2dB/step) of static power control according to Protocol 0505. Ifthe maximum output power is 46 dBm (40W), the step 6 is 34dBm.Static power control step is defined in the cell distributes list of

data management system, which specifies the maximumoutput power (suppose this value is Pn) of static power control.For step 15 of dynamic power control, the corresponding valuerange is Pn dB–Pn-30dB. When the maximum power controlstill cannot satisfy the requirement, adjust static power controlstep to improve the maximum output power of dynamic powercontrol Pn.1.9.4 Power Control ProcessingI. Measurement Report InterpolationEach measurement report has a sequence number. If networkdetects incontinuous sequence numbers, it means some of themeasurement reports are missing. The network will completethe reports based on interpolation algorithm.The network receives measurement reports n and n+4. Itdetects the sequence numbers are not continuous, so it uses analgorithm to add n+1, n+2, and n+3 to complete the reports.The purpose of measurement report interpolation is to avoidcall loss when the power is too low.II. Measurement Report FilteringNetwork will not judge the state of MS based on only onemeasurement result, because that is too incomprehensive, inaddition, the MS may be fluctuating. Therefore, filtering isrequired. Filtering combines several continuous measurementresults together to determine the state of MS during this periodof time. TA has filters for Rxlev and received quality of uplinkand downlinkThe purpose of measurement report filtering is to removetemporary nature and ensure the algorithm stability.III. Power Control AdjustmentCalculate the power adjustment value based on the differencebetween the Rxlev and the expected value.Power control adjustment based on RxlevPower control module compares the estimate value of Rxlevobtained through pre-processing of measurement report withthe expected value, and calculates the step length of

adjustment. In power control algorithm, variable step is oftenused for quick power control.Power control adjustment based on received qualityPower control module compares the estimate value of receivedquality obtained through pre-processing of measurementreport with the expected value, and calculates the step lengthof adjustment. When the received quality is bad, improve thetransmit power; when the received quality is good, reduce thetransmit power. This kind of power control adopts fixed step.Comprehensive decision for power controlConsider both Rxlev and received quality and adopt differentpower control strategies in different conditions to keep thestability and efficiency of power control algorithm.When the received quality requires the improving of transmitpower while the Rxlev requires the reducing of it, the systemwill make a comprehensive decision to perform no powercontrol adjustment, because bad received quality and goodRxlev represent strong network interference. Under suchcircumstances, improving transmit power will further increasethe interference.

1.10 Immediate Assignment ProcedureThe purpose of immediate assignment is to establish a radioconnection (RR connection) between MS and system at Uminterface.1.10.1 Network Access License and Random Access RequestThe request of MS for channel assignment is controlled by itsown access level and the access grant level broadcast in cell.Each MS has one access level of the ten levels from 0 to 9. Inaddition, it may also have one or several levels of the fivespecial access levels from l1 to 15. Access level is stored in SIMcard. BCCH system information broadcasts access levels andspecial access levels that the network grants and theinformation that whether all MSs allow emergency call or allow

special access levels only. If the mobile originated call is notemergency call, the MS can access to network only when itbelongs to the granted access level or granted special accesslevel. If the mobile originated call is emergency call, the MS canaccess to network only when all the MSs in the cell allowemergency call or it belongs to the granted special access level.When an MS wants to establish connection with the network, itsends a channel request to network through RACH channel.Channel request information contains 8-bit useful signalinginformation, among which 3 bits–6 bits are used as the minimalindicator of access cause. The system processes differentchannel requests based on this rough indication. Itdifferentiates the granted calls from the denied calls andassigns proper channels for the granted calls. This kind ofprocess is especially useful when the network is overload andthe flow control is required. Since the channel capacity islimited, this indicator cannot transfer all the information fromMS, such as the detailed cause of channel request, user identityand the features of mobile equipment. These kinds ofinformation are sent in the following SABM messages. The 8-bitinformation also contains the random discriminator sent by theMS and the immediate assignment command (it containsinformation about the assigned channel). Immediateassignment command carries the discriminator sent by theprevious MS. MS compares this discriminator with its owndiscriminator and judges whether it is the message for itselffrom network. Since there are at most 5 bits in the 8 bitsinformation carrying discriminator, only 32 MSs can bedifferentiated at the same time. Further discrimination of theMSs requires the response information at Um interface.Channel request information belongs to internal information ofBSS.In GSM, RACH is a kind of ALOH. In order to reduce the collisionon RACH during MS access to network and improve theefficiency of RACH channel and MS access. GSM specifies the

required access algorithm for MS. This kind of algorithm definesthree parameters: Tx_interger T, the maximum retransmissiontimes RET, and parameter S related to T and channelcombination.T represents the number of timeslots between twotransmissions when continuous channel requests are sent. S isan intermediate variable depends on T and the configuration ofCCCH. See the description of this parameter in Chapter 7. RET isthe MS maximum retransmission times allowed in order toavoid access collision. Each time after MS sends access request,T3120 is to receive (or reject) immediate assignment message.MS will retransmit access request for the messages that are notreceived or rejected when T3120 times out under the premisethat RET is not exceeded and restart the T3120. When theretransmission times reaches RET and T3120 times out, T3126will be started to receive (or reject) immediate assignmentmessage. When T3126 times out, cell re-selection will beinitiated.1.10.2 Initial Immediate AssignmentAfter decoding the channel request information, BTS sends achannel required message to BSC. This message containsimportant additional information and the estimation of TA byBTS. After receiving this message, BSC selects a proper channelfor this request and activates the land resources by sending achannel active message to BTS. BTS returns a channel activeacknowledge message to BSC. If BSC receives this message, BTSwill send an immediate assignment command or immediateassignment extended message on CCCH. In order to improvechannel efficiency, GSM introduces the message layout ofimmediate assignment extended that contains the assignmentinformation of two MSs. The immediate assignment messagecontains the assignment information of one MS. According toGSM specifications, MS must identity the immediateassignment (extended) information for the last three channelrequests.

If there is no channel to activate, BSC will send an immediateassignment reject or immediate assignment extended rejectmessage to MS. After receiving the reject message, MS stopsT3120 based on one of the last three channel requests andstarts T3122. During the specified time of T3122, MS has noaccess to network and turns into idle mode. Before T3122times out, MS cannot initiate connection attempt exceptemergency call within the same cell.After receiving immediate assignment message, MS comparesthe received assignment command with the information storedin its channel request and judges whether this message is foritself. If this message matches one of its last three channelrequests, MS will stop T3120 or T3126 and switch to theassigned channel. Then it starts to establish the signaling link byusing Set Asynchronous Balanced Mode (SABM) command.1.10.3 Initial MessageAfter receiving immediate assignment message and decoding it,MS adjusts its configuration of transmission and reception tothe assigned channel and transmits signaling according to theTA value specified by BSS and the initial maximum transmissionpower broadcast in BCCH system information (see thedescription of msTxPwrMaxCCH). MS sends an SABM frame onassigned SDCCH/TCH to establish the asynchronous balancedmode (SAPI=0) that is used to establish signaling message linklayer connection under acknowledgement mode. According toGSM protocol, SABM carries an initial message that containslayer 3 service request information.When two MSs send the same channel requests (which ispossible in high traffic volume area), the two MSs may respondto the same dedicated channel. in order to save this problem,after receiving SABM frame, BTS makes no modification butsends a UA frame (no frame number acknowledgement)containing the same information as that of initial message. Ifthe information of UA frame is different from that of SABMframe, MS will abandon this channel and start reaccess process.

Only the right MS can stay on this channel.SABM frame carries four kinds of initial messages: CM servicerequest (such as call setup, short message, and supplementaryservice), location updating request (generic location updating,periodic location updating, and IMSI attach), IMSI detach, andpaging response. All these messages contain the identity of MS,detailed access cause, and MS classmark (indicating some keyfeatures such as transmission power level, encryptionalgorithm, short message capacity, and frequency capacity).After receiving the initial message, BTS sends an establishindication message to BSC. BSC receives this message andsends complete layer 3 information to MSC to request SCCPconnection to MSC. Layer 3 information carries the causes forCM service request, which includes mobile originated call,emergency call, location updating, and short message service.This information also carries cipher key sequence number, MSidentification number, and some physical information of theMS such as transmit power level, ciphering algorithm, pseudo-synchronization, and short message. After receiving thisinformation, MSC sends connection confirmed message to BSC(if the connection cannot be established, MSC will send SCCPrefused message) to indicate that the signaling link betweenMS and MSC has been established. By this time, MSC cancontrol the transmission properties of RR management; BSSmonitors the transmission quality and prepares for handover.Then the MM connection begins.Authentication or encryption is triggered when required in thefollowing processing.In the immediate assignment process, T3101 starts when BSCsends channel active message to BTS and ends when theestablish indication is received. If T3101 times out beforesignaling channel is established, the activated channel will bereleased.1.10.4 Immediate Assignment FailureIf a failure occurs to the underlaying MS on the new channel

before the establishment of signaling link, the network releasesthe assigned channel of MS. The following processing dependson the failure type and previous actions. If the failure is causedby the mismatch of message field in decision contention and nore-assignment is initiated, the immediate assignment isrestarted.If the failure is caused by other reasons or if the re-assignmenttriggered by the mismatch of message field in decisioncontention is carried out and the assignment still fails, MS turnsinto idle mode and triggers cell re-selection.If the available information is not sufficient to define a channelafter the MS receives immediate assignment message, RRconnection fails.If the assigned frequencies of MS belong to two or more thantwo frequency bands, RR connection fails. If the assignedfrequency of MS is not consistent with the requested frequencybut supported by MS, MS accesses the channel with thefrequency used in channel request. If MS does not support theassigned frequency, RR connection fails.If T3101 times out before the signaling channel is established,network releases the assigned channel. Network cannot tellwhether MS resends the access attempt or not.

1.11 Authentication and EncryptionGSM takes lots of measures to protect the safety of system,such as using Temporary Mobile Subscriber Identity (TMSI) toprotect IMSI, using Personal Identification Number (PIN) toprotect SIM card, authentication through authentication center(AUC) for network access, encryption, and equipment identityregister.Authentication and encryption require a group of threeparameters that generated in AUC. Each client is assigned aMobile Station International ISDN Number (MSISDN) and IMSIwhen registers in GSM network. IMSI is preserved onto SIMcard through SIM printer and SIM printer will generate acorresponding client authentication value Ki that is stored in

SIM card and AUC as permanent information. AUC has apseudo number generator used to generate a random numberRAND. GSM defines algorithm A3, A8, and A5 that are used forauthentication and encryption. In AUC, RAND and Ki togetherproduce a response number SRES through A3 authenticationalgorithm and a Kc through A8 encryption algorithm. RAND, Kc,and SRES form a three-parameter group of client. This group isstored in the data base of this client in HLR. Generally, AUCtransfers five groups of parameters to HLR for automaticstorage. HLR can save ten groups of such parameters. WhenMSC/VLR requests for three-parameter group transfer, HLRsends five groups at the same time for MSC/VLR to use one byone. When there are two groups left, MSC/VLR will request fortransfer again.1.11.1 AuthenticationAuthentication is the process that GSM network checkswhether the IMSI or TMSI from MS at radio interface is valid ornot. The purpose of authentication is to avoid unauthorizedaccess to GSM network and the theft of private information byillegal users. Authentication also provides parameters for MS tocalculate new encryption key.The network initiates authentication procedure in the followingsituations:MS requesting for the change of information in VLR or HLR;Service access, including MS originated call, MS terminated call,MS activation and deactivation, and supplementary services;The first network access after MSC/VLR reboot;Mismatching Cipher key Sequence;Whether to initiate authentication procedure depends on if theKc value of the last service processing stored in networkconsistent with that of the present access stored in MS. Ifconsistent, authentication procedure can be escaped and thisKc value is used directly for encryption; if not, Kc value needs tobe recalculated. MS does not send Kc value to network throughradio path for the sake of privacy. Therefore, Cipher Key

Sequence Number (CKSN) is introduced. CKSN is sent to MS byMSC/VLR through authentication request message during thelast network access. It is stored in both SIM card and MSC/VLR.During the initial access of MS, CKSN is sent to MSC/VLRthrough the initial request message of SABM frame. MSC/VLRcompares it with the last CKSN. If they are not consistent,authentication is required before encryption. If CKSN=0, itmeans no Kc is assigned. Authentication procedure is initiatesand controls by network. MSC/VLR sends an authenticationrequest message to MS to initiate authentication procedureand T3260.I. Authentication Success2) AUTHENTICATION REQUEST contains a RAND (128 bits) and aCKSN. The Ki and RAND together generate a SERS (32 bits)through algorithm A3 and a Kc (64 bits) through algorithm A8.The new Kc replaces the former key and is stored in SIM cardtogether with CKSN.3) MS sends AUTHENTICATION RESPONSE to network. Afterreceiving this message, the network stops T3260 and checks itsvalidity (network compares it with the SERS generated by Kiand RAND through algorithm A3 and check whether they areconsistent or not), and then enters the subsequent procedures,such as encryption.II. Authentication RejectIf authentication fails, it means AUTHENTICATION RESPONSE isinvalid.If the MS uses TMSI, the network will initiate identityprocedure. If the IMSI provided by the MS is different from thatin network, the network will restart the authenticationprocedure; if the IMSI is correct, the network will sendAUTHENTICATION REJECT to the MS.If the MS uses IMSI, the network will send AUTHENTICATIONREJECT directly to MS. After sending AUTHENTICATION REJECTmessage, the network releases all the MM connections underestablishment and restarts the procedure for RR connection

release.After receiving AUTHENTICATION REJECT message, MS sets theroaming disabled flag and deletes information such as TMSI,LAI, and cipher key.If MS receives AUTHENTICATION REJECT message in IMSIDETACH INITIATED state, it stops T3220 after RR connection isreleased. If possible, MS initiates local release procedure afterthe normal release procedure or T3220 timeout; if not (such asthe IMSI detach after switch off), MSRR exits abnormally.If MS receives AUTHENTICATION REJECT message in otherstate, it exits all MM connections and call re-establishmentprocedures, stops T3210 and T3230, sets and starts T3240 toenter WAIT FOR NETWORK COMMAND state and wait for therelease of RR connection; If RR connection is not released afterT3240 timeout, MS will exit RR connection abnormally. Underthe two conditions above, MS enters MM IDLE and NO IMSIstate.1.11.2 EncryptionEncryption occurs in service requests such as location updating,service access, and inter-office handover. It requires thesupport of GSM network equipment (especially BTS), as well asthe encryption ability of MS.I. Signaling Procedure1) MSC sends BSC a Ciphering Mode CMD that containsencryption algorithm, Kc, and whether the MS is required toadd IMEI in Ciphering Mode CMP.2) BSC decides the final algorithm based on the encryptionalgorithm in Ciphering Mode CMD, the encryption algorithmthat BSC allows, and the encryption algorithm that MSsupports, and then inform BTS.3) BSC sends MS Ciphering Mode CMD to inform MS of theselected encryption algorithm.4) After receiving Ciphering Mode CMD, MS starts thetransmission of ciphering mode and sends Ciphering ModeCMP to the system.

5) After receiving the Ciphering Mode CMP from MS, BSCtransfer it to MSC.II. Procedure DescriptionA5 algorithmGSM protocol specifies eight kinds of encryption algorithmfrom A5/0 to A5/7. A5/0 stands for no encryption. Theencryption procedure is initiated by the network. Theencryption information of Cipher Mode CMD specifies therequired encryption algorithm. The algorithm that generatesencrypted code is called A5 algorithm. It calculates by using theKc (64 bits) and the current frame number (22 bits) to generatea 114-bit encryption sequence and then implements XORoperation with the 114-bit burst. Two encryption sequencesare used for uplink and downlink. For each burst, one sequenceis used for MS encryption and BTS decryption, the othersequence is used for BTS encryption and MS decryption.Encryption algorithm selectionWhen MS initiates call request, the SABM frame carriesClassmark 1 or 2 to indicate whether the MS supportsalgorithm A5/1, A5/2, or A5/3, and reports Classmark 3 inCLASS MARK CHANGE to further indicate whether the MSsupports Algorithm A5/4, A5/5, A5/6, or A5/7(In systeminformation, if ECSC=1, MS reports Classmark 3 immediately; ifECSC = 0, the Classmark 3 is reported after CLASSMARKENQUIRY is initiated by the network. Therefore, theconfiguration of ECSC = 1 is recommended when the encryptionis used). MSC sends encryption command based on theconfiguration of secret data. BSC chooses the intersection ofthe encryption algorithm allowed in the command sent by MSC,the encryption algorithm allowed in BSC data configuration,and the encryption algorithm supported in the MS report. Inthe intersection, BSC selects a proper algorithm based on thepriority level of A5/7 > A5/6 > A5/5 > A5/4 > A5/4 > A5/3 >A5/2 > A5/1 > A5/0.Encryption in handover

The HANDOVER REQUEST contains the encryption informationunit that indicates the required encryption algorithm and key. Ifone of the two A interfaces of BSS is in PHASE I, due to thelimitation of ETSI GSM PHASE I protocol (no ciphering modesetting information unit in handover command), the two Ainterfaces match only when they share the same encryptionalgorithm (such as A5/2) to ensure the normal inter-BSChandover. Otherwise, special treatment has to be made to thetarget MSC or target BSC (or the source MSC or source BSC) tochange the handover command for inter-BSC handover.For the interconnection of A-interfaces when the encryption isused, whether special data configuration is required for BSCand MSC must be considered.1.11.3 TMSI ReallocationAfter authentication and encryption, the system sends CMSERVICE ACCEPT or TMSI reallocation command to MS andinitiates T3250.When MS registers in the location area for the first time, thenetwork allocates a TMSI to it. When the MS leaves thislocation area, it releases the TMSI. When the MS receives theTMSI reallocation command, it saves the TMSI and LAI andsends TMSI reallocation complete message. After receiving thismessage, the network stops T3250.If the system cannot identify TMSI of the MS, for example,when the data base error occurs, the MS must provide its IMSI.The identification program is initiated before the TMSIreallocation to request for the IMSI.The identification program sends identity request message tothe MS, after receiving this message, the MS provides its IMSIby sending identity response message to the network. Whenthis procedure is over, authentication, encryption, and IMSIreallocation are implemented if required.1.11.4 Exceptional SituationsI. AuthenticationRR connection failure

If the network detects RR connection failure before receivingAUTHENTICATION RESPONSE, it releases all the MMconnections and terminates all the active MM procedures.T3260 timeoutT3260 is started when MSC sends authentication request toBSC and stops when MSC receives AUTHENTICATIONRESPONSE. If the T3260 times out before the AUTHENTICATIONRESPONSE is received, the network releases RR connection,terminates the authentication procedure and all the active MMprocedures, and then releases all the MM connections andinitiates RR connection release procedure.Unregistered SIM cardIf the SIM card of the MS is not registered, the network sendsAUTHENTICATION REJECT message directly to the MS.II. EncryptionEncryption rejectIf BSS does not support the encryption algorithm specified inCIPHERING MODE CMD, it sends CIPHER MODE REJECTmessage to MSC.If the encryption is initiated in BSS before MSC requests for thechange of encryption algorithm, BSS also sends CIPHER MODEREJECT message to MSC.Un-encrypted MSThe CIPHERING MODE COMMAND message is valid when:–The un-encrypted MS receives CIPHERING MODECOMMMAND message that requires encryption.–The un-encrypted MS receives CIPHERING MODECOMMMAND message that requires non-encryption.–The encrypted MS receives CIPHERING MODE COMMMANDmessage that requires non-encryption.In other cases, CIPHERING MODE COMMAND is consideredwrong. The MS sends RR STATUS message with the cause ofprotocol error and performs no action.III. TMSI ReallocationRR connection failure

If RR connection fails before TMSI reallocation completemessage is received, all the MM connections are released andboth the old and new TMSIs are saved during a certain recoverytime.T3250 timeoutT3250 is started when MSC sends TMSI_ REALL_ CMD messageor LOC UPD ACC message with the new TMSI and stops whenMSC receives TMSI _REALL_COM. If T3250 times out before theTMSI _REALL_COM is received, MSC sends CLEAR COMmessage to release RR connection and terminate TMSIreallocation.

1.12 Location UpdateIn GSM, the paging information cannot be sent in thewhole network due to the capacity limit of the pagingchannel. Therefore, the definition of location area (LA) isintroduced. LAC contains many cells. The paging for theMS is carried out through the paging in all the cellswithin the LA of the MS. The size of the LA is of vitalimportance to the system performance in network design.The registration management for the LA is required sincethe paging for the MS is carried out through the pagingin all the cells within the LA, which brings about thedefinition of location update. Location update is dividedinto generic location update, periodic location update,and IMSI attach.1.12.1 Generic Location Update (Inter-LA LocationUpdate)When the MS moves from one LA to another LA,registration is required. If the LAI stored in the MS isdifferent from the LAI of the current cell, the MSinforms the network to change the location information itstores. This procedure is called generic location update.In idle mode, if cell re-selection occurs when the MS

moves within the LA, the MS will not inform thenetwork immediately but implement cell re-selectionwithout location update or network involvement. If theMS moves to another LA after re-selection, the MSinforms the network of this LA change, which is calledforced registration.According to whether the VLR changes or IMSIinvolves, generic location update is divided into thefollowing types:I. Intra VlR Location UpdateIt is the simplest location update that requires no IMSI. Ithappens in the current VLR without informing the HLR.In the initial message carried by SABM frame, the accesscause is MM LOCATION UPDATING REQUEST thatcarries the MS TMSI and LAI. The generic locationupdating is indicated. MSC receives this message andforwards it to VLR. VLR updates the MS locationinformation and stores the new LAI, and then sends anew TMSI to MS if required (MS uses the former TMSIif no TMSI is carried in the TMSI re-allocationcommand). After receiving the TMSI re-allocationcomplete message, MSC sends location updating acceptmessage and releases the channel. Location updatingcompletes.II. Inter-VLR Location Updating, Sending TMSIAfter the MS enters a cell, if the current LAI is differentfrom the LAI it stores, it sends its LAI and TMSI to VLRthrough MSC in location updating request. VLR deducesthe former VLR based on the LAI and TMSI it receivedand sends a MAP_SEND_IDENTIFICATION to theformer VLR to request for IMSI and authenticationparameter. The former VLR sends the IMSI andauthentication parameters to the current VLR. If the

current VLR cannot obtain the IMSI, it sends MS anidentity request message to request for the IMSI. Afterreceiving the IMSI, VLR sends HLR the locationupdating message that contains the MS identityinformation for the data query and path establishment ofHLR. After receiving this message, HLR stores thenumber of the current VLR and sendsMAP/D_CANCEL_LOCATION to the former VLR ifthe current MSC/VLR has the normal service rights.After receiving this message, the former VLR deletes allthe information about this MS and sends the HLR aMAP/D_CANCEL_LOCATION_RESULT message toconfirm the deletion. The HLR will sendMAP_INSERT_SUBSCRIBER_DATA message toprovide the current VLR with the information it requires(including authentication parameters) after the procedurefor authentication, encryption, and TMSI reallocation isover, and confirm the location updating after receivingthe response from the VLR.III. Inter-VLR Location Updating, Sending IMSIThe procedure is similar with the procedure above buteasier because it requests for authentication parameterfrom the HLR through IMSI directly.1.12.2 Periodic Location updatingThe network and the MS lose contact when:The MS is switched on but moves out of the networkcoverage area (dead zone). The network lost contact withthe MS and regards it still in attach status.The MS sends IMSI detach message and the uplinkquality is bad due to interference, the network may notbe able to decode this message correctly. The MS is stillregarded in attach status.The MS is power off. It cannot inform the network of its

status and the contact is lost.If the paging for MS happens when the contact is lost,the system sends paging information in the LA that theMS registered before. The network cannot receive theresponse from the MS. The system resource is wasted.To solve this problem, the implicit detach timer isintroduced in the VLR for the IMSI status management.In addition, measures are taken in BSS to force the MS toreport its location periodically. Therefore, the network isinformed of the status of MS. This kind of mechanism iscalled periodic location updating. The network sends aperiodic location updating time T3212 to all the users inthe cell through BCCH to force the MS to send locationupdating request with the cause of periodic locationupdating after T3212 times out.Before the T3212 times out, if the timeout value ischanged (for example, the service cell changes and theT3212 timeout value is broadcast), the MS uses the timewhen the change happens as the initial value and keep ontiming.If the T3212 times out when the MS is in NO CELLAVAILABLE, LIMITED SERVICE, PLMN SEARCH,or PLMN SEARCH-NORMAL SERVICE status, thelocation updating is initiated after the MS is out of theseservice status.Periodic location updating ensures the close contactbetween network and mobile users. The shorter updatingperiod leads to better network performance. But thefrequent location updating will increase the signalingflow and reduce the utilization of the radio resources, oreven affect the processing ability of MSC, BSC, andBTS. On the other hand, it will greatly increase thepower consumption of MS and reduce its standby time.

The T3212 setting should be based on comprehensiveconsideration.The procedure for periodic location updating is the sameas that for generic location updating.1.12.3 IMSI Attach and DetachIMSI attach and detach means to attach a binary mark tothe subscriber record in MSC/VLR. The former one ismarked as access granted, and the latter one is marked asaccess denied.When the MS is switched on, it informs the network ofits status change by sending an IMSI ATTACH messageto the network to inform. After receiving this message,the network marks the current user status in the systemdatabase for the paging program.If the current LAI and the LAI the MS stores are thesame, IMSI attach is initiated. The procedure is similar tothe intra VLR location updating only that the locationupdating request message is marked as IMSI attach andthe initial message contains IMSI of the MS.If the current LAI is different from the LAI stored,generic location updating is initiated.When the MS is switched off, the IMSI detach istriggered by a key-press. Only one command is sent toMSC/VLR from the MS. This is an unacknowledgedmessage. After receiving this message, MSC informsVLR to do detach mark to this IMSI while the HLR isnot informed of the no-radio of this user. When thepaging for this user occurs, HLR requests for the MSRNfrom the VLR and is informed of the no-radio of thisuser by this time. Therefore, no paging program isimplemented. The paging message is handled directly,such as playing the record: "The subscriber is poweredoff."

The procedure above is explicit IMSI detach. There isalso implicit detach. The implicit detach happens beforethe implicit detach timer times out. If the contactbetween MS and network is not established, the VLRsets the IMSI status as detach. The implicit detach timeris set longer than the periodic location updating timerT3212 to avoid "abnormal" implicit detach. The implicitdetach is denied during the establishment of radioconnection. The implicit detach timer is reset after therelease of radio connection. Implicit detach timer is alsocalled IMSI delete time.VLR deletes the IMSI marked as detach periodically(The period is adjustable) and reports the user status tothe HLR.1.12.4 Exceptional SituationsI. MSAccess denied because of access level limitMS stays in the service cell and performs the normal cellre-selection procedure without triggering locationupdating. When the current cell allows access or othercell is selected, The MS initiates location updatingimmediately.IMMEDIATE ASSIGNMENT REJECT message isreceived during random accessMS stays in the service cell and starts T3122 based onthe value in the immediate assignment reject message.The normal cell selection and re-selection procedure isperformed. If the cell that the MS stays changes orT3122 times out, the MS initiates location updating.Random access failureIf the random access fails, T3213 is started. After theT3213 times out, the random access procedure isinitiated. If two successive random accesses fail, the

location updating is terminated. For the subsequentprocessing, see the following description.RR connection failure: Location updating procedure isterminated. For the subsequent processing, see thefollowing description.T3210 timeout: Location updating fails. For thesubsequent processing, see the following description.The completion of RR connection is abnormal: Locationupdating fails. For the subsequent processing, see thefollowing description.Location updating reject due to reasons other than #2, #3,#6, #11, #12, or #13: MS waits for the release of RRconnection. For the subsequent processing, see thefollowing description.# 2 (IMSI unknown in HLR)# 3 (Illegal MS)# 6 (Illegal ME)# 11 (PLMN not allowed)# 12 (Location Area not allowed)# 13 (Roaming not allowed in this location area)Subsequent processing: If the T3210 is still timing, stopit; If T3210 times out, RR connection fails. Add 1 to thelocation updating attempt timer. The followingprocessing depends on the LAI (stored and received fromthe service cell) and the value of the location updatingattempt timer.If the location updating status is UPDATED, the storedLAI and the received LAI are the same, and the locationupdating attempt timer is less than 4, MS keeps theUPDATED status. After the release of RR connection,the sub status of MM IDLE becomes NORMALSERVICE. The MS also stores the information about theformer location updating type. The T3211 is started after

RR connection release. After it times out, the locationupdating procedure is started again.If the location updating status is not UPDATED, or thestored LAI is different from the received LAI, or thelocation updating attempt timer is equal to or less than 4,the MS deletes the ciphering key sequence, LAI, TMSIstored in SIM card and sets the location updating statusas NOT UPDATED. After the release of RR connection,the sub status of MM IDLE becomes ATTEMPTING TOUPDATE. After the RR connection release, if thelocation updating attempt is less than 4, T3211 is started.Otherwise, T3212 is started. After the T3211 or T3212times out, the location updating procedure is startedagain.After the sub status of MM IDLE becomesATTEMPTING TO UPDATE, the MS will do thefollowing:If T3211, T3213, or T3212 times out, perform locationupdating.If LA changes, perform generic location updatingIf the cause for the status change is (3), (4), (6) (the causeis not the abnormal release with unknown reason), or (7)(cause “retry in the new cell”), perform location updatingwhen entering the new cell.If the cause for the status change is (5), (6) (the cause isabnormal release with unknown reason), or (7) (the causeis not “retry in the new cell”), location updating is notperformed when entering the new cell.No IMSI detach.Support emergency call requestRespond the paging with IMSIPerform generic location updating triggered by therequest from CM layer (if the location updating

succeeds, the MML connection request will be accepted.For details, see section 4.5.1 of the Protocol 0408).II. Matching Between IMSI Delete Time and T3212If the periodic location updating fails for four times,T3212 will be started for the next update. In the badcoverage area, especially in the area where the uplinkand downlink do not match (downlink is better thanuplink), after the periodic location update fails,Another location updating is initiated after T3212 timesout. Therefore, the T3212 is set to be shorter in the badcoverage area. In addition, if the IMSI delete time is lessthan twice of the T3212, the users stay in the service areabut cannot be called. So the IMSI delete time should bemore than twice of the T3212 and based on LAC.III. NetworkRR connection failureAmong all the sub procedures attached to the locationupdating procedure, if the RR connection fails, it ishandled according to the exception handling of othercommon procedures.If no other common procedure is attached to the locationupdating procedure, the MS location updating isterminated.Protocol errorIf the network detects protocol error after receivingLOCATION UPDATING REQUEST, it sendsLOCATION UPDATING REJECT message to the MSwith the following cause if possible:#96 required IE error#99 IE error or no IE exists#100 Conditional IE error#111 Protocol error, undefined

After sending LOCATION UPDATING REJECT to theMS, the network initiates channel release procedure.

1.13 MS Originating Call FlowThe MS needs to set up a main signaling link to connect to MSCfirst, and then initiates the authentication, encryption, andTMSI reassignment flow.1.13.1 Called Number AnalysisAfter the authentication, encryption, and TMSI reassignmentflow are over, the MS starts the call setup flow.First, the MS sends a SETUP message to the network side. Thismessage contains called number and the required services. TheMSC implements the call proceeding according to the message.When receive the SETUP message, the MSC sends the outgoingcall message SEND_INFO_FOR_O/C_CALL to the VLR. Afterreceive the outgoing call message, the VLR analyzes the itemssuch as called number, the calling party capability, and networkresources capability according to the user information obtainedfrom the HLR during the location updating process, to checkwhether to accept this call request. If a certain item cannot bepassed, the VLR sends the RELEASE COMPLETE message to theMS. The call fails. The MS then proceeds to release the bottomlayer connection and switches to the idle state. If the aboveitems can be passed, the VLR sends the COMPLETE_CALLmessage to the MSC. After receive this message, the MSC sendsthe CALL PROCEEDING message to the MS. It means that thecall request is accepted and the call is set up.

1.13.2 Voice Channel Assignment (Follow-up Assignment)After send the CALL PROCEEDING message to the MS, the MSCactivates the follow-up assignment according to the servicerequest. That is, assign the TCH voice channel to the user. Atthis time, the MSC sends the ASSIGNMENT REQUEST messageto the BSC. This message contains the information such as therequested channel type to request the BSC to assign the TCH

voice channel for the call.After receive the channel request from the MSC, the BSC sendsthe Channel Activation for TCH message to the BTS to activatecorresponding terrestrial resources and start a timer at thesame time if the TCH channel resources are available. If the BTShas prepared the resources such as circuit, the BTS sends theCHANNEL ACTIVATION ACK message to the BSC. If the BSC hasno available resources to assign, it sends the RESOURCEFAILURE message to the MSC. But if the system allows queuing,the BSC sends the QUEUING INDICATION message to the MSCand places the assignment request in the queue and starts thetimer T11. If the T11 times out, the BSC sends the CLEARREQUEST message to the MSC.The immediate assignment request, intra-BSC handover, andinter-BSC handover do not support queuing. Only the TCHresource request (that is, the assignment request and intra-cellhandover) allows queuing. The TCH resource requests in thequeue are assigned with relevant channels in the sequence oftheir priorities. In the length of the queue reaches its thresholdor the timer times out, the request is rejected.When the BSC receives the CHANNEL ACTIVATION ACKmessage from the BTS, the BSC puts the physical information ofthe channel provided by the BTS in the ASSIGNMENTCOMMAND message (this message contains the informationsuch as channel type, voice/data indication, channel rate, voicedecoding algorithm and transparent transmission indicator,assignment priority and CIC). The ASSIGNMENT COMMANDmessage is sent to the MS through the SDCCH channel.

After receive the ASSIGNMENT COMMAND message from theBTS, the MS adjusts the transceiver configuration to the TCHchannel and then sends the SABM message to the BTS throughthe FACCH channel in the way of stolen frame. After the BTSreceives the SABM message, the BTS sends the ESTABLISHINDICATION message to the BSC and then sends an

Unnumbered Acknowledge (UA) to the MS, just as the initialsignaling channel assignment does.After receive the UA, the MS sends the ASSIGNMENTCOMMPLETE message to the BTS through the FACCH channel. Ifthe MS fails to identify the assignment information and fails tooccupy the specified channel due to the radio interface failure,radio interface message failure or interference, or hardwareproblems, the MS returns to the original channel and sends theASSIGNMENT FAILURE to the BTS. If the MS does not receivethe ASSIGNMENT COMMAND sent from BTS or the BTS doesnot receive the response message sent from MS due tointerference or other causes, the system starts thecorresponding timers (such as T3103 or T3107) and when thetimer times out, the channel is released.When receive the ASSIGNMENT COMPLETE message, the BSCsends the ASSIGNMENT COMPLETE message to the MSC. At thesame time, it also sends the RF CHANNEL RELEASE message tothe BTS to release the occupied SDCCH signaling channel. Whenthe BTS releases the signaling channel, it sends the RFCHANNEL RELEASE ACK message to the BSC. After the BSCreceive the message, it considers that the signaling channel is inidle state and can be assigned to other channel requests.For different purposes, the GSM has three different channelassignment flows. They are initial channel assignment, follow-up channel assignment, and handover channel assignment.Initial channel assignment: is mandatory to establish the linktransmission between the MS and the network. For example,process the location updating request.During the establishment of the signaling transmission, if theTCH channel is assigned preferably, this assignment is calledvery early assignment (VEA). After the MSC sends theASSIGNMENT REQUEST message, the BSC does not apply fornew channel but initiate the Mode_Modify flow. After theMode_Modify is complete, the BSC reports the ASSIGNMENTCOMPLETE message to the MSC.

If the SDCCH channel is assigned first, and the TCH channel isassigned when it is needed, and then ASSIGNMENT REQUESTmessage from MSC is sent before the Alerting message, thisassignment is called early assignment (EA).If the SDCCH channel is assigned first and the TCH is assignedafter the called party sends the CONNECT message, Generally,it adopts the EA mode.

If the EA mode is used in the initial assignment, when no SDCCHis available, assign the TCH channel for the channel requestdirectly. The TCH channel replaces the SDCCH channel to sendthe signaling message. Please note that using the TCH channelto transmit the signaling wastes the resources a lot becauseone TCH channel equals eight SDCCH channels. When thissituation is quite serious, add more SDCCH to meet therequirement in time.Follow-up channel assignmentAfter the signaling channel finishes the authentication andencryption process, if there is still voice or data request, thefollow-up channel assignment is triggered to assign a TCHchannel.Handover channel assignmentThis assignment is used to apply for channels due to handoverduring the call process. The system judges whether thehandover occurs in the SDCCH or in the TCH to assigncorresponding channels. The handover flow and theassignment flow in the cell are the same. The only difference isthat the message names are different. Similar to the immediateassignment flow, in the MS assignment flow, the timer T3107starts when the BSC sends the ASSIGNMENT COMMANDmessage to the BTS. After the BSC receives the ASSIGNMENTCOMPLETE message from the BTS, the timer T3107 resets.Generally, the timeout of the timer is caused by the bad radiocoverage. When the timer times out, the MS is considereddisconnected with the network and the resources are released

for other MSs. Based on the statistics, the channel assignmentis generally complete within two seconds. If the BSC does notreceive the ASSIGNMENT COMPLETE message within twoseconds, the assignment fails. But sometime, the networkquality is bad, some messages needs to be sent several times,in this case, the assignment can be extended to five seconds.Generally, if the traffic load of the cell is heavy, set the timer as2 seconds to 5 seconds. If not heavy, set the timer as 10seconds.1.13.3 Call ConnectionAfter receiving the ASSIGNMENT COMPLETE message from theBSC, the MSC sends the Initial Address Message (IAM) thatincludes the information used to establish the route to thecalled network. The MSC will receive the call setup report soon.If succeeds, the MSC receives an ADDDRESS COMPLETEmessage (ACM); if fails because of certain reason (such as busyline or congestion), the MSC receives a RELESASE message fromthe called end.If MSC receives the ACM, MSC sends the ALERTING message tothe MS (MS translates it into ring back tone). This message is aDTAP message. If no answer is received from the called partyand the calling party does not terminate the connection, thenetwork will terminate the call or perform no answer calltransfer after a while.If the called party picks up the phone, MSC receives anANSWER message. The link between the calling party and thecalled party is connected. MSC sends a CONNECT message inthe CC protocol to the MS. After receiving this message, the MSsends a CONNECT ACKNOWLEDGE message in the CC protocolto the system. The system starts charging after receiving thismessage. If the called end is data device, it enters CONNECTstatus directly after receiving the SETUP indication. The callconnection procedure is over and the two parties start theconversation or data transmission service.1.13.4 Call Release

If the calling party hangs up first, the MS sends disconnectmessage to MSC through FACCH. After receiving this message,the MSC sends release message to inform the called party toterminate the communication. The end-to-end connection isover. But the call is not complete, because certain tasks such assending charge indication are performed. When the connectionto the MS is no longer necessary, the system sends a RELEASEmessage to the MS and starts T308. After receiving thismessage, the MS sends a RELEASE COMPLETE message to thesystem and the call is over. The MS stops the T308 afterreceiving the RELEASE COMPLETE message. Similarly, if thecalled party hangs up first, it sends a RELEASE message to thecalling party. The MSC sends the calling party a DISCONNECTmessage after receiving the RELEASE message. If the call isterminated in an abnormal way, this message further indicatesthe cause for that.When the MSC receives the RELEASE COMPLETE message fromthe MS, it sends a CLEAR COMMAND message to BSC to releaseall the signaling links. This message contains the cause for thecall clearance, such as handover complete or location updatingcomplete. The call connection release is over. If the abnormalrelease occurs because of radio link failure or device failure, theBSC sends a CLEAR REQUEST message to the MSC.After receiving the CLEAR REQUEST message, BSC sends aCHANNEL RELEASE message to the MS and starts T3109 toshow that all the lower layer links are released. Meanwhile, itrequires the MS to enter the idle mode. When the MS receivesthe CHANNEL RELEASE message, it removes the uplink signalinglink (to stop sending the measurement report of uplink channelassociated signaling on SACCH). The MS sends DISC message toBTS and starts T3110. After receiving this message, The BTSsends UA to MS and the RELEASE INDICATION to the BSC. Whenthe T3110 times out or the MS receives the UA frame, it entersthe idle mode.

In order to ensure the timely removal of the uplink anddownlink, when the BSC sends the CHANNEL RELEASE messageto the MS for the uplink removal, it also sends a deactivateSACCH (SACCH) to the BTS requiring for the release of thedownlink signaling (to stop the signaling connection betweenthe two parties). After receiving this message, the BTS stops thetransmission of the downlink SACCH frame and sends thedeactivate SACCH acknowledgement to the MSC.After receiving the RELEASE INDICATION message, BSC resetsthe T3109 and starts the T3111, and sends RF CHANNLERELEASE to the BTS (the T3111 is reset at the same time),requiring for the release of TCH resources. When the BSCreceives the RF CHANNLE RELEASE acknowledgement messagefrom the BTS, it sends a CLEAR COMPLETE message to the MSC,indicating that the radio link clearance is over and the channelis available for reallocation.After receiving the CLEAR COMPLETE message, the MSCreleases the SCCP connection by sending RLSD and receivingRLC. The whole MS originating call flow is over.1.13.5 Exceptional SituationsI. No Establish Indication Message Is Received After ChannelActivationThe main causes are:The MS may send many channel requests even if the BSS workswell, which activates many signaling channels. But the MS onlyoccupies one of them. Other channels are released by the BSCafter the T3101 times out as they cannot receive the establishindication from the MS. If the Tx_interger is proper, the causefor this problem is that the uplink reception is normal but thedownlink signal cannot be received by the MS. Under suchcircumstances, the received level and the received quality ofuplink and downlink should be checked. If the MS is not faraway from the BTS but the received level and the receivedquality are bad, check the antenna feeder and the TRX in BTS.Improper configuration of Tx-integer in BSC

The Tx-integer affects the interval of channel request re-sending. Improper Tx-integer only leads to the activation ofmany channels by BSS, but no call will be affected.II. BSC Sending Immediate Assignment RejectIf the BSC sends immediate assignment reject to the MS afterreceiving the channel required message, the usual causes are:No proper signaling channel is available for the MS because ofall channels are busy or the channels are blocked.BTS sends channel activation negative acknowledge afterreceiving the channel activation message.If the BTS sends lots of channel activation negativeacknowledge messages to the BSC, it is usually because thetransmission at Abis interface is not stable, which leads to theinconsistent channel status of the BSC and BTS, or becauseerrors occur in certain board of BTS.III. MSC Sending Disconnect Message Instead of AssignmentRequest to Terminate the CallIn the call connection process, the immediate assignment isfollowed by the assignment procedure. But due to certainreasons, the MSC sends a disconnect message instead of theassignment request message to the MS and then terminatesthe call. Under such circumstances, many complaint phonesfrom users cannot get through. Check the following:The A interface circuit of MSCThe data consistencies of the A interface between the MSC andBSC, especially the circuit pool data.IV. Assignment FailureAfter receiving the assignment request, the BSC sendsassignment failure message instead of assignment complete.The usual causes are:No proper voice channel is available for the MS.BSC has no proper voice channel for the MS because all thevoice channels are busy or the channels are blocked.The cause value carried by the assignment failure message is noradio resource.

The MS voice channel access fails.Under this condition, the assignment failure is reported fromthe MS.Due to the special features of the radio transmission, this kindof assignment failure occurs most frequently and is unsolvable.If the occurrence rate is too high, check the antenna feeder, theBTS board, and the parameters related to channel access in BSCdata configuration.The A interface circuit of BSC fails, for example, the CIC in theassignment request is not available.The hardware of BSC fails.The cause value in the assignment failure message sent by BSCis equipment failure.The transmission at A interface fails.V. Directed RetryAfter receiving the assignment request message from the MSC,if no TCH is available and the BSC allows directed retry, the BSCimplements the handover with the cause value of directedretry to change the service cell of the MS.VI. Exceptional Procedure Due to Call DropCall drop may occur any time during the call flow, which affectsthe following procedures. For example, the call drop occurswhen the BSC receives the assignment request message fromthe MSC. The assignment procedure may be not complete (thechannel may be just assigned and no assignment commandmessage is sent). Under this condition, BSC may send clearrequest message instead of assignment complete message orassignment failure message to the MSC.VII. Exceptional Procedure Due to HangupHang up of the calling party or the called party may occur anytime during the call flow, which affects the followingprocedures. For example, the hangup occurs when the BSCreceives the assignment request from the MSC. Under thiscondition, the call flow may be terminated before the BSCsends assignment complete or assignment failure to the MSC.

This assignment procedure neither succeeds (BSC sendsassignment complete) nor fails (BSC sends assignment failure).VIII. Exceptional procedure because MSC sends clear commandAfter the A interface connect is established, MSC may sendclear command or disconnect message to the BSC during thecall flow, which affects the following procedures. For example,the hang up occurs when the BSC receives the assignmentrequest from the MSC. Under this condition, the call flow maybe terminated before the BSC sends assignment complete orassignment failure to the MSC. This assignment procedureneither succeeds (BSC sends assignment complete) nor fails(BSC sends assignment failure)If it happens many times, analysis the following two factors:The cause value carried in the clear commandThe cause value is usually the call control if the call isterminated in a normal way. Otherwise, the cause value may beprotocol error, equipment failure, or others.The interval between the clear command or disconnectmessage and the last messageThe interval between the clear command or disconnectmessage and the last message indicates whether theexceptional procedure is triggered by timeout.

1.14 MS Originated Call Flow1.14.1 EnquiryAfter the signaling link for the calling end is established,the Initial Address Message with Information (IAI) issend from the calling end to the GMSC. The IAI containsthe MSISDN of the called party. GMSC analyzes theidentification number of the CCS7 of the HLR and sendsthis HLR the SEND_ROUTING_INFORMATIONmessage. After receiving this message, the HLR checksthe user record, and then performs different proceduresand responds the GMSC as follows:Under normal circumstances, the HLR only has the

partial information about the identification of the currentVLR, such as the CCS7 address or the universal mark.To get the routing information for the call, the HLRsends the VLR a PROVIDE ROAMING_ NUMBERmessage that contains the user IMSI information,requiring the VLR to provide a MSRN for this call.When the MSC/VLR receives this message, it selects aroaming number from the idle numbers to temporarilyconnect it to the IMSI, and sends thePROVIDE_ROAMING_NUMBER_RESULT messagewith the MSRN assigned to this call in it to the HLR.When the HLR receives the MSRN, it transfers theinformation by sending aSEND_ROUTING_INFORMATION_RESULT messageto the call originating GMSC. Then the GMSC can findthe VLR with the obtained MSRN and sends the IAI toit. After receiving this message, the MSC restores theIMSI of this user in its memory record with the MSRNand starts the paging for the MS. After the call isestablished, this roaming number is released for anotheruser.If the record of the called party is set as Barring of AllIncoming Calls (BAIC) or Barring of Incoming Callswhen roaming is outside the home PLMN country(BIC_roam) according to the message sent by the VLRand the user is in roaming now, the HLR rejects this call.If the user record is set as Call Forwarding Unconditional(CFU), the HLR sends the MSRN to the original GMSCto analyze this number and redefine the routing.If no VLR number of the user is found and no callforwarding is set, Error message will be sent to theGMSC.1.14.2 Paging

After receiving the IAI from the GMSC, the called MSCsends a SEND_INFO_I/C_CALL message to the VLRand the VLR will analyze the called number and thenetwork resource capacity to check whether thisrequirement is acceptable. If certain item is not accepted,it informs the calling end that the call establishment fails.Under normal circumstances, the VLR sends the MSC aPAGING MAP message that contains the location areaidentification (LAI) and the IMSI or TMSI of the calledparty, informing the MSC to perform the pagingprocedure.When the MSC obtains the LA information of the MSfrom the VLR, it sends all the BSCs in this LA thepaging message that contains the cell list and the TMSIand IMSI information required for paging. The IMSI canbe used in the paging for the MS through the cell pagingchannel. In addition, it is also used to confirm the pagingsubchannel in the discontinuous reception processing.BSC sends the PAGING COMMAND to all the cells inthe LA. This command message contains the pagingchannel group number and the timeslot number (obtainedby the calculation of the last three numbers of the IMSI,the total number of the paging channels, and the totalnumber of the paging timeslots).When the cell receives this paging command, it sends thePAGING REQUEST message on the paging channel.The message contains the IMSI or TMSI of the userpaged.If the called MS detects the paging by decoding thepaging information, it sends a channel request to initiatethe channel allocation process. After receiving theimmediate assignment command from the network, theMS sends the initial message of PAGING RESPOSE on

the channel assigned through the SABM frame, and thenimplements the authentication, encryption, TMSIreallocation, and finally begins the call establishmentprocess.

1.14.3 Call Establishment for the Called PartyAfter the TMSI reallocation is over, the MSC sends theMS a SETUP message that includes all the detailsrequired such as the service type and the calling number.After receiving this message, the called MS confirms theinformation and sends a CALL CONFIRMED messageback if the service is available. The call confirmedmessage carries the parameters that the MS selects, suchas the channel type (full rate TCH or half rate TCH) andthe service type.After receiving the call confirmed message, the MSCsends the assignment command to the BSC for the voicechannel allocation. After the assignment procedure isover, the called MS sends an ALERTING message to thenetwork and a ringing prompt occurs to the called MS.when the MSC receives this message, it sends anAddress Complete Message (ACM) to the calling end.After receiving this message, the calling end makes aring back tone as the originating user prompter.The called user hears the ringing and responds, and thensends a CONNECT message to the MSC. After receivingthis message, the MSC connects all the transmissionlinks. The end-to-end transmission is established.1.14.4 The Influence of Call Transfer to RoutingIn the supplementary services, call transfer has thegreatest influence on call routing. The call transfer ismainly caused by Call Forwarding Unconditional (CFU),Call Forwarding Busy (CFB), Call Forwarding on mobile

subscriber Not Reachable (CFNRc), and Call Forwardingon No Reply (CFNRy). The routing selection for eachfunction is as follows:I. CFUWhen the GMSC sends theSEND_ROUTING_INFORMATION message to theHLR, if the CFU function is available, the HLR sends theSEND_ROUTING_INFORMATION_RESULT messagewith the transfer number in it back to the GMSC for it toredefine the routing.II. CFBWhen the GMSC finds the VMSC/VLR with the MSRNobtained from the HLR, but the called end is busy andthe CFB function is available, the VMSC/VLRimplements the call transfer of the transfer number andsends it to the third party. If the CFB function is notavailable, the GNSC handles the call directly, such asplaying the user bush record.III. CFNRcThe routing selection for this function is based on howthe network decides the called party is not reachable. Theprocessing is different for different criteria.If the last location registration of the called user fails,and the HLR keeps the record of this situation and knowsthe MS is unreachable, it makes the CFNRc decision byitself.If the HLR does not keep the record of this situation, thecall flow continues until the MSC performs the pagingfor the user and gets no response from the user in duetime. The user is decided not reachable. The MSCforwards this call. This kind of situation has manycauses. One of them is that the user enters the dead zoneor the MS is power-off, but the VMSC has not made the

periodic check on the IMSI attached user yet, so it cannotjudge the MS status and the paging fails. Another causeis that the MS is in frequent location updating on theedge of the LA and cannot respond the paging or thechannel request fails, which leads to paging timeout.If the MS is in IMSI detach (the MS is switched off orout of the service area for a long time), because thedetach tag is in the VLR instead of the HLR, the callforwarding can only be initiated by the VMSC/VLR.When the VLR periodically deletes the long-termdetached IMSI and informs the HLR, the HLR need notcontact the VLR.IV. CFNRyIf the paging of the VMSC for the user succeeds and thecalled end sends the ALERTING message to the system,but the called user makes no response in due time and theCFNRy function is activated, the call forwardingprocedure is initiated.V. CW and HOLDCall Waiting (CW) is a supplementary service. When theMSC receives the IAI from the calling end, if the calleduser is in another conversation and the CW function isenabled, the MSC skips the paging procedure anddirectly sends a SETUP message to the MS by using thecurrent signaling mode. When the CW function isenabled, the handover of the two calls can be performed.When the CFB and the CW are enabled at the same time,the CW is initiated first if another call is coming. TheCFB will be initiated when a third call is coming.1.14.5 Exceptional SituationsThis section only analyzes the common abnormalprocedures. For other abnormal procedures, see "MobileOriginating Call Establishment Procedure."

Upon paging failure, the MSC prompts voice informationto the calling party, indicating the called MS is outsidethe serving area or cannot be connected. In this case,trace the signaling on interfaces A and Abis to checkwhether the paging failure is caused by:No PAGING COMMAND at A interfaceNo PAGING COMMAND at Abis interfaceNo PAGING RESPONSE at Abis interfaceNo PAGING RESPONSE at A interfaceI. No Paging Command at A InterfaceThrough signaling tracing over interface A, the MSC isdetected that it has not sent a PAGING message to theBSC. In this case, check the data configuration and MSinformation in the MSC/VLR and HLR on the NSS side.Additionally, power off the called MS, power it on andmake a test call to check whether the MS is normal.Checking user data in VLRWhen an MS is paged, the MSC judges the current stateof the MS by the user data (including MS active state,registered LA, cell information), and decides whether orhow to send the PAGING message.If the MS state has changed (for example, the MS isswitched off, or has entered a different LA) and has notregistered in the network normally or updated user datain VLR, the MS may probably be unable to be paged.In that case, the MS only need to initiate a locationupdating procedure to ensure that the user data in VLR iscorrect. The period of periodic location updating isindicated in system information. On MSC side, there isalso a location updating period (See "Location updatingProcedure"). The two parameters of BSC and MSC mustsatisfy a certain relationship, which requires that MSmust initiate a location updating procedure within the

period specified in MSC.Checking RA- or Cell-Related parameter settings inMSCIf a routing area or cell related parameter is incorrectlyset in the MSC, the transmission of the PAGINGmessage may fail. For example, if a wrong target BSC isselected, the PAGING message that should have beensent to the local BSC will be sent to another BSC.II. No Paging Command at Abis InterfaceUpon receiving the PAGING message from the MSC,the BSC detects that the MSC has not sent PAGINGCOMMAND to the BTS over interface Abis. In thiscase, check the operations and data configuration in theBSC。Checking if flow control is enabledCheck if the system load suddenly increases due tocentralized transmission of short messages or massaccess bursts.Checking relevant data configurationCheck if the CGI information in BSC data configurationis consistent with the LAC information in the PAGINGmessage over A interface. Additionally, if RA- or cell-related parameter is not correctly set in the MSC, forexample, a wrong target BSC is selected, the PAGINGCOMMAND message cannot be successfully sent overAbis interface.Check whether the following parameters in the [Systeminformation table] are correctly set:"BS_AG_BLKS_RES", "CCCH-CONF" and"BS_PA_MFRMS".III. No Paging Response at Abis InterfaceThrough signaling tracing over Abis interface, the BSCis detected that it has not received the Establishment

Indication (PAGING RESPONSE) after sendingPAGING COMMAND to the BTS. In this case, checkthe relevant data configuration and radio signal coverage.Check if there is PCH or AGCH overload due tocentralized short message transmission or mass accessbursts.Check the called MS or SIM in it.Check BTS by making test calls in a different cell.Check data configuration in BSCCheck whether thefollowing parameters in the [System information table]are correctly configured: "BS_AG_BLKS_RES","CCCH-CONF", "BS_PA_MFRMS", "Tx-integer," and"MS MAX retrans". Check the setting for "locationupdating period" in BSC and that in MSCCheck radio signal coverageDue to the problem of radio signal coverage, there mightbe some blind coverage areas. The MS that has entered ablind coverage area cannot receive the PAGINGREQUEST message. In that case, the MS cannot bepaged. Such cases, if any, only exist in partial areas.IV. No Paging Response at A InterfaceThrough signaling tracing at Abis interface, the BSC isdetected that it has received an Establishment Indication(PAGING RESPONSE) message from the BTS but thismessage is not reported over interface A.

1.15 HOAs a key technology in the cellular mobile telecommunicationsystem, handover (HO) can reduce the call drop rate and thenetwork cross interference. The handover procedure consistsof handover trigger, handover preparation and decision, andhandover execution.HO can be divided into synchronous HO and asynchronous HObased on Timing Advance (TA). Synchronous HO means the two

cells are synchronized with each other and the MS can calculatethe new TA (the HO command indicates whether the HO issynchronous or not). Asynchronous HO requires the BTS tocalculate the new TA. When the MS receives the HO commandand requests for the new BTS access, the new BTS informs theMS of the calculated TA. The MS access to the new channel canalso be divided into four types: synchronous, asynchronous,pre-synchronous, and pseudo-synchronous. The first threetypes are required in MS and the last one is optional. Thepseudo-synchronous HO can be performed only when the MSsupports this function. In the pseudo-synchronous HO, thehandover command from the BTS of the original service cellcontains the RTD value (the TA difference between the sourceBTS and the target BTS). The MSC calculates the TA required forthe access to the new BTS based on the RTD value.The HO process involves MS, BTS, BSC, and MSC. According tothe location where the HO happens, the HO can be divided intointra-cell HO and inter-cell HO. To be more specific, intra-cellHO, intra-BTS HO, intro-BSC HO, intra-MSC HO, and inter-MSCHO. The function of each unit is: MS measures the downlinkperformance and the signal strength; BTS monitors thereceived signal level and quality of the uplink and theinterference level of the idle traffic channel; BSC handles themeasurement report and makes the HO decision; MSC decidesthe target cell of the inter-BSC HO.1.15.1 HO PreparationI. Measurement ReportThe HO decision depends on the measurement report (MR)sent by MS through uplink SACCH to the network and the MRof the uplink sent by BTS. These two reports are sent to BSC atthe same time for decision. The system information thatincludes the parameters of the current cell and the neighborcell are sent to the MS under the dedicated mode through thedownlink SACCH. The MS reports the RXLEV and quality, TAvalue, power control, and DTX usage to the network according

to the system information. In addition, the MS also performsthe pseudo-synchronization with the neighbor cell defined bythe system for HO and measures the RXLEV from the BCCH. TheMS measures all the frames except the idle frames that areused to synchronize the neighbor cell and decode SCH. The MSreports the condition of the cell and the six neighbor cells withthe strongest RXLEV it measures during the measurementperiod to the system for the HO decision.Measurement periodThe SACCH measurement period is different if the MS occupiesdifferent channel under the dedicated mode.–If the SACCH is associated with SDCCH, the measurementperiod is 470ms, because a complete SACCH message blockoccupies two 51 multiframes of SDCCH.–If the SACCH is associated with TCH, the measurement periodis 480 ms, because a complete SACCH message block occupiesfour 26 multiframes of TCH.A complete MR consists of four continuous SACCH bursts. Onthe SDCCH, the four bursts are transmitted continuously. Onthe TCH, each 26 multiframe has only one SACCH burst, so acomplete MR requires four 26 multiframes.Figure 1-1 Measurement periodWhether to use DTX or not, the MR has two values: fullmeasurement value and sub measurement value. For details,see the DTX description in Chapter 2.MR processingBTS handles the uplink MR it makes and the downlink MR itcollects from the MS. It obtains the sample values of the RXLEV,RXQUAL, and TA, and then calculates the arithmetical meanvalue and the weighted mean value based on the relatedparameters. When the time is up, the system decides whetherto perform the level handover, quality handover, or distancehandover.II. Neighbor Cell MonitoringTo establish the HO relation with the neighbor cells, the MS

must listen to the standard frequency of the neighbor cellsdefined in the system message. The standard frequency carriesthe synchronous channel and frequency correction channel.One way to decide the received channel is the standardfrequency channel is to confirm that the frequency carries aFCCH. The MS also decodes the SCH that carries the TDMAframe number and BSIC. The MS can only analyze the BCCHstandard frequency of the neighbor cell in the idle timeslot ofthe TCH multiframe. In fact, during the data exchange, theinterval between the end of the reception and the beginning ofthe transmission (about 1 ms) can be used to measure theRXLEV and the RXQUAL, but it is not sufficient to measure thelevel of the neighbor cell. The interval between the end of thetransmission and the beginning of the reception (about 2 ms) issufficient to measure the level of the neighbor cell, but notsufficient to find the FCCH. In the 26 muliframe of TCH, there isalways an idle frame (about 6 ms) available for MS to decodethe FCCH and SCH. But the FCCH of the neighbor cell may notbe found during this timeslot. Therefore, the use of thearithmetic feature of the two numbers 26 and 51 is required.Because these two numbers have no common factor, the FCCHcan be found during the 11 periods. When SACCH is associatedwith SDCCH, although its period is also 51 multiframe, theSDCCH channel assigned to the MS only occupies 1/8 of the 51multiframe. Since there are lots of idle timeslots, the MS cansynchronize the neighbor cell.When the MS receives the SCH, the synchronization isestablished. To translate the message on the downlink CSCH,the MS must know the training sequence of the CSCH. Thetraining sequence is of eight types, matching the BCC 0 to BCC 7of BSIC respectively. The BSIC carried by the SCH can inform theMS of the training sequence number of its service cell.BSIC also enables the MS to differentiate the cells using thesame BCCH frequency. The two cells with the same BCCHfrequency and BSIC must be far from each other. The MS

reports the six neighbor cells with the strongest signals, butdifferentiates them according to the BSIC and frequency itobtains to achieve the pre-synchronization. The MR onlycontains the sequence number of the frequencies in the BA list.Therefore, if a cell shares the same frequency and BSIC with theneighbor cell and its signal is strong enough, error report anddecision of MS may occur, leading to HO failure and call drop.III. Conditions Required for Neighbor Cells to Join in HODecision QueueWhen the BTS receives the report on the neighbor cell from theMS, it checks whether this neighbor cell is qualified to join inthe HO decision queue. The following conditions must be met:RXLEV(n) > RxLevMinCell(n)+ MAX(0,Pa(n)) + OFFSET (2-4)Pa(n)=MS_TXPWR_MAX(n)－MAX_POWER_OF_MSRXLEV(n) is the RXLEV of the neighbor cell; RxLevMinCell(N) isthe minimal access level of the neighbor cell; OFFSET is theoffset of the minimal access level; MS_TXPWR_MAX(n) is themaximal transmit power of MS defined by the system;MAX_POWER_OF_MS is the maximal transmit power the MScan achieve. The unit is dBm.RxLevMinCell(n) and MS_TXPWR_MAX(n) are defined by theHO cell parameters. Under the dedicated mode, the systeminforms the MS by sending the system message throughSACCH. The neighbor cell can be listed in the HO candidate cellsonly when its RXLEV is qualified according to the formulaabove.The defined RxLevMinCell (n) must be higher than theRXLEV_ACCESS_MIN. If it is too low, the threshold for thecandidate cells is reduced, which may lead to HO failure. Thepurpose to define the Pa is to ensure the low power MS canaccess the neighbor cell only when the RXLEV is high enough,thus improving the quality of conversation.1.15.2 HO TypesHO must be performed on time under different conditions toensure the quality of communication. According to the cause of

the HO, it can be divided into Power Budget (PBGT) HO, edgeHO, bad quality (BQ) HO, direct retry, and timing advance (TA)HO.I. PBGT HOPBGT HO is based on path loss. PBGT HO algorithm looks for acell with less path loss to decide whether HO is necessary. Thebiggest difference between the PBGT HO and others is that thetriggering condition is path loss but not receiving power.The formula of PBGT HO is as follows:PBGT (n) > PGBT_Ho_Margin (n) (2-5)PBGT(n) = ( BSTX_MAX - RXLEV_DL - PWR_C_D ) - (BSTX_MAX(n)- RXLEV_NCELL(n) )- ( RXLEV_DL - RXLEV_UL -SENSI_CORRECT)- max ( BSTX_MAX(n)- min(MSTX_MAX(n),P) -BSTX_MAX + min (MSTX_MAX,P) ,0 )BSTX_MAX: The maximum transmit power of BS in service cellBSTX_MAX (n): The maximum transmit power of BS in neighborcellRXLEV_DL: The downlink received signal level in service cellRXLEV_UL: The uplink received signal level in service cellSENSI_CORRECT: The correct factor of MS/BS receiversensitivityRXLEV_NCELL (n): the received signal level of MS from neighborcell nPWR_C_D: the decrease of the transmission power in BTSpower controlP: Max MS Transmission powerMSTX_MAX (n): Max MS transmit power allowed of theneighboring cell nMSTX_MAX: Max MS transmit power allowed of the service cellThe neighbor cell with the biggest PBGT (n) is selected as thetarget cell for HO. The PGBT_Ho_Margin is the defined RXLEVdifference value between the service cell and the neighbor cellwhen the HO is initiated. If this value is too low, it may lead toping-pong handover; if it is too high, HO hysteresis may occurand the HO efficiency is reduced. Since the PGBT_Ho_Margin is

defined for the specific neighbor cell, the traffic load can beadjusted accordingly. For example, when cell A and cell B areadjacent, A is the high-traffic cell and B is the low-traffic cell,the call distribution can be balanced by reducing thePGBT_Ho_Margin from A to B and increasing that from B to A.In fact, this way to balance the call distribution equals thedecrease of the coverage area for cell A and the increase of thecoverage area for cell B.PBGT HO only happens between the peer cells. .II. Edge HOThe uplink/downlink edge HO margin is defined in the HOparameters. When BSC finds in the MRs from the MS and BTSthat the uplink or downlink RXLEV is lower than the edge HOmargin defined, it selects a proper neighbor cell from the MRsas the target cell to initiate HO, thus avoiding the call drop.In the edge HO, the RXLEV of the neighbor cell should be higherthan that of the service cell by a certain value. This value iscalled the edge HO margin. This algorithm is also used to avoidping-pong handover. The edge HO margin should be higherthan the minimal access level of the MS.III. BQ HOThe decision mechanism of BQ HO is similar to that of the edgeHO. When BSC finds in the MRs from the MS and BTS that thebit error rate of the uplink or downlink is higher than the BQHO margin defined, the BQ HO is initiated. To furtherdifferentiate the BQ HO, the interference HO is introduced. Ifthe RXLEV is higher than the defined RXLEV margin of theinterference HO and the RXQUAL is higher than the quality HOmargin, the frequency interference exists. The interference HOwill trigger the intra-cell HO (when the intra-cell HO is available)first to improve the bad conversation quality due tointerference, and then trigger the inter-cell HO. The intra-cellHO is not effective when the frequency hopping is used. Byimproving the interference HO margin, the BQ HO will bemainly performed between cells.

IV. Direct RetryDuring the call establishment, the SDCCH is assigned first andthen is the TCH. If the service cell has no idle TCH, the callattempt usually fails because of TCH congestion. To fully utilizethe radio resources and reduce the congestion, the direct retryfunction is introduced. When the SDCCH is assigned, but noTCH is available, the assignment request is sent in the form ofMR and the call is accessed to the idle speech channel. Afterthe direct retry function is enabled, the queuing function canbe activated to provide enough time for the system to selectthe neighbor cell available for direct retry.V. TA HOTA HO can be used to control the coverage area of the BTS.When the BSC finds the TA value reported by the MS is higherthan the defined margin, the TA HO is initiated. If the TA marginis relatively low, the frequent ping-pong handover may betriggered. Therefore, special attention should be paid to thematching of different kinds of HO.1.15.3 HO Process AnalysisI. Intra-Cell HOIn the real network, sometimes the interference may occur tocertain frequency or a certain TRX fails, leading to the highRXLEV but low RXQUAL or the remarkably low signal level ofTRX. To improve the conversation quality and avoid the calldrop, the intra-cell HO is used.The intra-cell HO is initiated by the RXLEV margin or RXQUALquality. During the conversation, BSC analyzes the MR from theMS and BTS. If the requirement for intra-cell HO margin issatisfied, it sends a CHANNEL ACTIVE message to BTS to initiatethe intra-cell HO. The connection process is similar to the TCHassignment during the call establishment. Because the TCH isalso assigned within the cell, the BTS can indicate the MS toperform the intra-cell HO through HO command or assignmentcommand. When the BSC receives the ASSIGNMENTCOMPLETE/HANDOVER COMPLETE message from the BTS, it

sends MSC the HO PERFOMED message that contains the HOtype. Then the BSC sends a RF CHANNEL RELEASE message toBTS. After receiving the message, the BTS releases the TCHresource and sends a RF CHANNEL RELEASE ACK message back.When the intra-cell HO is enabled, intra-cell HO increases a lot,and the system load also increases. Therefore, if the traffic loadis already heavy, the intra-cell HO function is notrecommended.II. Intra-BSC HOIntra-BSC HO is performed by BSC and no MSC has to beinvolved. To inform MSC that the HO is complete, BSC will senda HO PERFOMED message to MSC.1) The MS sends MR to BTS1 on SACCH at Um interface, andBTS1 forwards the message to the BSC.2) BSC receives the MR. If it decides that the MS should behanded over to another cell, it sends Channel Activation toBTS2 of the target cell to activate the channel.3) BTS2 receives the CHANNEL ACTIVATE. If the channel type iscorrect, it turns on the power amplifier on the specifiedchannel to receive information in the uplink direction, and sendCHANNEL ACTIVATE ACK to the BSC.4) After receiving the CHANNEL ACTIVATE ACK from BTS2, theBSC sends HANDOVER COMMAND to the MS through BTS1 andstarts T3103. The handover command contains all the featureinformation of the transmission on the new channel and thedata required for MS access. It also indicates whether this HO issynchronous or asynchronous.5) After receiving the HANDOVER COMMAND, the MS decidesthe type of it. If it is synchronous HO, the MS sends the targetcell four continuous HANDOVER ACCESS messages on theassigned TCH, and then starts the transmission based on thecalculated. For the synchronous HO, the former TA can beused; for pre-synchronous HO, the TA in the handovercommand is used (If the TA is not provided in the handovercommand, the default value is used); for pseudo-synchronous

HO (MS reported whether this HO is supported or not before),the TA is calculated based on the difference value provided inthe handover command. Please note that the HANDOVERACCESS is send by the access burst. It is the only time when theaccess burst is used on the DCH. It only contains the 8-bit HOreference number obtained from the handover command.Since this reference number is known to the target cell, thetarget cell can check whether the access request is from theexpected MS with this number.The HO reference number is not fully defined in the protocol.During the HO access, if the assigned TCH is on the BCCH, dueto synchronization error and delay or other reasons, the accessburst may offset to the BCCH RACH timeslot. If the 8-bitreference number is the same as a service application number,the system will regard it as a random access by mistake andassign the SDCCH through AGCH, leading to a waste of AGCHand SDCCH. But as the access burst contains the BSICinformation, only the HO access cell will be affected.Since there are more than four HO access bursts, and after thenew BSS assigns a channel to the MS, it will no re-assign thischannel to other MS, even if no reference number is used, thenetwork can find the MS to access and the HO will not beaffected.To further avoid the waste of radio resources, the referencenumber is assigned a fixed value that is different from theapplication number for service type in random access.6) BTS2 receives the HANDOVER ACCESS from the MS, and sendHANDOVER DETECT to the BSC notifying that the HANDOVERACCESS message is received.7) For asynchronous HO, after the BTS2 channel of the targetcell is activated, it waits for the MS access on the assigned DCH(until the T3103 times out). When it detects the handoveraccess from the MS, the BTS2 sends the HO DETECT message tothe BSC and the PHYSICAL INFO that contains the calculated TAto the MS. During the PHYSICAL INFO transmission, the

network initiates T3105. Before receiving the SABM frameresponse from the MS, the BTS2 re-enables the T3105 aftertimeout and resends the PHYSICAL INFO NY1. For asynchronousHO, after receiving the PHYSICAL INFO, the MS sends the SABMto the BTS2; for synchronous HO, the MS sends the SABM tothe BTS2 immediately after sending the HANDOVER ACCESS.8) For asynchronous HO, the MS starts the T3124 when sendingthe HANDOVER ACCESS message for the first time and stopsthe T3124 after receiving the PHYSICAL INFO. For details, seethe parameter description section.9) After receiving the first SABM, BTS2 sends BSC the EST IND toinform it of the radio link establishment. When the networkreceives this message, it sends an ESTABLISHE INDICATIONmessage to the BSC to show that the data link layer isestablished. Meanwhile, it also sends the UA response frame tothe MS. after receiving the UA response, the MS regards thatthe signaling answer mode is established with this cell.10) The MS sends HANDOVER COMPLETE to the BTS2, and BTS2forwards it to the BSC. Then it sends the target cell aHANDOVER COMPLETE message that only contains thehandover complete indication but no other information. TheMS stops considering the possibility to return to the formerchannel only when this message is sent. If the MS does notreceive the PHYSICAL INFO from the target cell or the UAresponse frame, it sends a HANDOVER FAILURE message on thesource channel.11) After receiving the HANDOVER COMPLETE message, theBSC stops the T3103 and sends MSC the HANDOVERPERFORMED that contains the handover type. Meanwhile, theBSC initiates the local release for the former channel of BTS1.When the target cell receives the handover complete messagefrom the MS, it forwards it to the BSC. After receiving thismessage, the BSC sends the RF CHANNEL RELEASE message toinform the source cell to release the former TCH. When thesource cell receives this report, it sends a RF CHANNEL RELEASE

ACK to indicate the radio channel is released and available foranother assignment.III. Intra MSC HOCompared with the intra-BSC HO procedure, the procedure forthe inter-BSC HO only has several A interface signaling added.1) When the MS has to be handed over to the cell where theBSC2 belongs to, the BSC1 sends a HO REQUIRED message thatcontains cell ID of the target cell group and the source cell andthe HO cause to the MSC and starts T7 at the same time.2) After the MSC receives this message, if it shares the sameLAC with the target cell, it searches the BSC of the target cell(BSC2) and sends the BSC2 a HANDOVER REQUEST messagethat contains the information of the target cell and the sourcecell, transmission mode, encryption mode, classmark, and thechannel type required. When the BSC2 receives this message, itsends MSC a CC message to indicate that the connectionbetween the MSC and its SCCP is established for transmissionof the information from the A interface.3) After the new channel is activated, the BSC2 sends the MSC aHO REQUEST ACK to indicate that the channel is available. Thismessage carries the HO command with the information aboutthe resource allocation in it to show that the local end is readyfor HO.4) After receiving the HO REQUEST ACK, the MSC sends a HOCOMMAND to the BSC1. BSC1 stops the T7 and starts the T8,and forwards the HO COMMAND to the MS and starts T3103,informing the MS to access the new channel. This commandcontains the cell ID, channel type, and HO reference.5) After receiving the HO COMPLETE from the BSC2, MSC sendsa CLEAR COMMAND to the BSC1. This command contains theclear cause (such as HO clear). BSC1 stops T8 and T3103, andreleases the former channel. Meanwhile, it sends a CLEARCOMPLETE message to the MSC.

T3103 is started when BSC sends the HO command and cleared

when the BSC receives the HO COMPLETE (INTRA BSC) or CLEARCOMMAND (INTER BSC). The T3103 should be set less than T8.During the HO, the BSC provides the time for TCH both in thesource cell and the target cell according to the T3103. Whenthe T3103 is timing, two channels are reserved. The longest HO(INTER MSC) may take about five seconds, so the T3103 can beset to five seconds. If it is set too long, the system resourceswill be wasted.If the target cell and the source cell are not in the same LA, alocation updating will be performed at the end of each call.IV. Inter-MSC HOThe procedure for inter-MSC HO is shown in Figure 1-26.1) When MSCa receives the HANDOVER REQUIRED messagefrom the BSC, if it finds that the LAC of the preferred target cellis not in the local LAC list, it queries the remote LAC list thatcontains the routing address of the neighbor MSC/VLR.2) When the target MSCb is found, the MSCa sends a PREPAREHANDOVER message that contains the HANDOVER REQUEST toit.3) After receiving the PREPARE HANDOVER message, the MSCbsends the VLRb an ALLOCATE_HO_NUMBER message torequest for HO number (HON) assignment. The HON indicatesthe routing between MSCa and MSCb.4) VLRb selects an idle HON and sends it to MSCb through theSEND HO REPORT message.5) MSCb establishes a SCCP link to the target BSC and sends aHANDOVER REQUEST message to BSCB. Then the BSC activatesthe channel of the target cell. After receiving the channelactivation response from the target cell, the BSC sends MSCb aHANDOVER REQUEST ACK message that contains the HOcommand.6) After receiving this message, MSCb sends a PREPAREHANDOVER ACK message that contains the HANDOVERREQUEST ACK and the HON to the MSCa.7) MSCa receives this message and sends an IAM to MSCb. The

IAM contains the HON assigned by VLRb for MSCb to identifywhich speech channel is reserved for the MS. MSCb sends aSEND HO REPORT RESP message to the VLRb anytime after itreceives the IAM.8) After MSCa receives the ACM from the MSCb, it sends theHO command to the MS. Then the MS will perform the HOaccess to the target cell.9) After receiving the HO access message from the MS, MSCbsends MSCa a PROCESS ACCESS SIGNALLING message toindicate that the HO is detected.10) When the target cell receives the HANDOVER COMPLETEmessage from the MS, it informs the MSCb. Then the MSCbsends a SEND END SIGNAL REQ message to MSCa to inform itthe HO is complete. After the HO-DETECT or HO-COMPLETE isreceived, the connection between MSCa and MSCb isestablished. MSCb will release the HON.11) When MSCa receives the HO complete message, it sends aclear command to the former BSC to release the channelresource. The inter-MSC HO is complete. To avoid thePSTN/ISDN contradiction of the MSCa and MSCb, MSCb mustsend an answer signaling when receiving theHO-DETECT/COMPLETE.12) MSCa controls the call until it is cleared. When MSCa clearsthe MS call, it also clears the call control function of MSCa andsends a MAP-SEND-END-SIGNAL message to release the MSCbMAP resource.MSCb sends a HO failure indication to the MSCa if the MSCbcannot identify the target cell, the HO to the target cell is notallowed, the target cell has no radio channel available, or thedata error occurs. The MSCa will perform the HO to thesecondary cell or terminate the HO.V. Subsequent Inter-MSC HOAfter the MSCb receives the HO request, it checks this targetcell belongs to MSCb and performs the inter-MSC HO. After theHO is complete, it informs the MSC.

The subsequent HO is the handover of MSCb to other MSCafter an inter-MSC HO is complete. The target MSC can be theformer MSCa or the new MSCb’. The circuit switch happens inthe MSCa for both situations. After the subsequent HO iscomplete, the connection between MSCa and MSCb is released.The procedure for the subsequent HO with circuit switch is asfollows:MSCb is handed over back to MSCa

1) MSCb sends MAP PREPARE SUBSEQUENT HANDOVERrequest to MSCa. This message contains MSCa number, targetcell ID, and all the information in HO REQUEST.2) MSCa is the call control MSC. It can search the idle channelimmediately without target HO number routing.3) After the radio channel is assigned, MSCa sends a MAPPREPARE SUBSEQUENT HANDOVER response back.4) If the TCH is busy, BSSa sends a QUEUING INDICATION toMSCb (optional). MSC sends MSCb the MAP FORWARD ACCESSSIGNALLING request that contains the subsequent TCHassignment result (HO REQUEST ACK or HO FAILURE). If theradio channel cannot be assigned or the error occurs to thetarget cell ID, or the target cell ID does not match the targetMSC number according to the HO REQUEST, a MAP PREPARESUBSEQUENT HANDOVER response that contains the HOFAILURE information in it is sent to the MSCb. MSCb keeps theconnection to the MS.5) If the MSCa is successfully assigned, and the MAP PREPARESUBSEQUENT HANDOVER response is sent to MSCb. The MSCbrequests the handover of the MS to the new cell of the MSCaby sending a HO command.6) After receiving the HO complete message, MSCa releases thecircuit connection to MSCb.7) MSCa must send a proper MAP message to terminate theMAP procedure for MSCa and MSCb during the basic HO. WhenMSCb receives the MAP SEND END SIGNAL response message,

it releases the BSSb resources.

MSCb is handed over to MSCb'Note 1: This message can be sent anytime after the IAM isreceived.

1) MSCb receives the HO request and finds that the target celldoes not belong to the MSCb. It sends a PREPARE SUBSHANDOVER to the MSCa. This message contains the MSCb’ ID,target cell ID, and all the information in HO REQUEST. MSCa willinitiate a basic HO to MSCb’.2) If the MSC can be found in the MSCa LAC list and remote LAClist (it contains information about other MSC), after the HON isprovided by the VLRb’ and the MSCb’ channel is activated,3) MSCa sends a MAP PREPARE SUBSEQUENT HANDOVERresponse message to the MSCb. This message contains the HOREQUEST ACK from the BSSb’ and the BSSMAP information thatmay be special.4) After receiving this message, MSCb sends the HO commandto the MS. After the access succeeds, if the MSCa receives theMAP SEND END SIGNAL REQUEST (it contains the HOCOMPLETE information of the BSSb’) from the MSCb’, the HO iscomplete and the connection between MSCa and MSCb isreleased. MSCa also sends the MAP SEND END SIGNALresponse to MSCb to end their MAP conversation. MSCbreceives this message and releases the radio resources.5) After the subsequent HO is complete, the MSCb’ replaces theMSCb. Any subsequent inter-MSC HO is the same as describedabove.The remote LAC list of MSCa must be complete and contain asmany MSCs as possible besides the neighbor MSC. For example,if a user in place A calls another user in place B, the MSC inplace A must contains all the data of the MSCs and cells withinthe area between A and B. Otherwise, the HO cannot beperformed and the call drops.

1.15.4 Exceptional SituationsThe following are some extra exceptional situations on thebasis of what has described before.I. HO Failure Due to CIC ExceptionIf the CIC allocated in the Handover REQ received by BSC ismarked as BLOCK, BSC will respond to MSC with HandoverFailure due to "requested terrestrial resource unavailable".II. HO Failure Due to MS Access FailureIf the BTS cannot decode Handover Access or HandoverCompleted correctly when a MS accesses the new channel, theHO will fail. The MS returns to the old channel, and respondswith a Hanover Failure message.For the intra-BSC handover, if the BSC has not received theHandover CMP message on the new channel, or HandoverFailure message on the old channel at expiry of timer T3103A, itwill consider the call as dropped and send a Clear REQ messageto the MSC on the old channel. Upon receiving the Clear CMDmessage from the MSC, the BSC releases the old channel andnotifies the target cell to release the new channel. If timerT3103B1 or T3103B2 times out, the target cell will release thenew channel.For the inter-BSC handover, if BSC1 has not received theHandover CMP message at expiry of timer T3103B2, it will senda Clear REQ message to the MSC to release the call. If BSC2 hasnot received the Handover DET or Handover CMP message, itwill send a Clear REQ message to the MSC for the samepurpose.

1.16 Call Re-Establishment1.16.1 IntroductionThe re-establishment procedure allows MS to resume aconnection in progress after a radio link failure, possibly in anew cell or in a new location area (re-establishment in a newlocation area initiates no location updating).Whether call re-establishment is allowed depends on thecalling status, the cell's allowance of call re-establishment, and

activated MM connection (MM is in status 6 "MM connectionactivated" or status 20 " Waiting for additional MMconnection" Call re-establishment can only be initiated by MS.GSM protocol does not specify the implementation mode forthe short message service and the independent callsupplementary service. In the other end, no voice is heardduring the call re-establishment.During the radio transmission, a connection may be brokensuddenly because of the great transmission loss due toobstructions such as bridges, buildings, or tunnels. When thecall re-establishment is used, the MS can maintain theconversation by using another cell in a short time, thusimproving the network quality. Call re-establishment can beregarded as the HO initiated by MS to save the interrupted callin the current cell.Call re-establishment is of two types according to the entitythat has the radio link failure first.I. Radio Link Failure Occurs to MS FirstThe MS sends a call re-establishment request in the selectedcell (source cell or target cell). The former channel resource isreleased after the BTS timer times out.II. Radio Link Timeout Occurs to BSS FirstAfter the radio link timer in BTS times out, the BTS sends aradio link failure message to the BSC and BSC activates theSACCH. According to the protocol, the network must handle thecontext for a while after detecting the lower layer faults for thesuccessful call re-establishment. The implementation mode andduration are decided by the equipment provider. Afterdetecting the radio link failure, the MS selects a neighbor cellwith the highest RXLEV within five seconds and sends thechannel request in the selected cell. This cell should not bebarred and the C1 is over 0. In addition, this cell must permitthe call re-establishment. If all the neighbor cells are notqualified, the call re-establishment is abandoned.

During the call re-establishment, the MS cannot return into theidle mode. If the MS selects a cell in different LA as the targetcell for call re-establishment, it cannot perform locationupdating until the call ends.Under normal circumstances, the call re-establishmentprocedure lasts about 4 to 20 seconds. Most users have hungup the phone before the procedure is over. Therefore, the callre-establishment cannot achieve its goal but wastes a lot ofradio resources. For the areas with limited channel resources,the activation of this function is not recommended.1.16.2 Call Re-Establishment Procedure1) After the MM connection failure indication is reported to theCM entity, if the MS receives at least one request for MMconnection re-establishment from CM, it will initiate the call re-establishment procedure. If several CM entities request for re-establishment, only one re-establishment procedure will beinitiated.2) After the CM sends the request for the re-establishment ofMM connection, MM sublayer sends a request for theestablishment of RR connection and enters the WAIT FORREESTABLISH state. This request includes an establishmentcause and a CM re-establishment request. When the RRsublayer indicates a RR connection is established (the CM re-establishment request message has been sent through the Uminterface), the MM sublayer starts T3230 and indicates to allthe CM entities that the MM connection is under construction.The MM sublayer stays in WAIT FOR REESTABLISH state.The CM Re-establishment Request message contains the MSidentity (IMSI or TMSI), Classmark 2, and encrypted sequencenumber.Whether the CM entity can request for re-establishmentdepends on protocol discriminator (PD).3) After receiving the CM re-establishment request, thenetwork analyzes the request type and starts the MM programor RR program. The network can start the classmark enquiry

program to obtain more information about the MS encryptionability. The network can also decide to perform theauthentication procedure or ciphering mode setting procedure.4) When the RR sublayer indicates the ciphering mode settingprocedure is over or the CM SERVICE ACCEPT message isreceived, the MM connection is re-established. The T3230stops and informs all the CM entities related to the re-establishment to enter the MM CONNECTION ACTIVE state.5) If the network cannot connect the re-establishment requestto the current MS call, it sends the CM SERVICE REJECT with thereject cause to the MS.The reject cause (value) includes unidentifiable call (#38),unidentifiable IMSI (# 4), unauthorized ME (# 6), networkfailure (#17), congestion (#22), unsupported service (#32), andtemporary service failure (#34)。6) After receiving the CM SERVICE REJECT, the MS stops T3230and releases all MM connections and RR connections. If thereject cause if #4, the MS deletes the TMSI, LAI, and CKSN inSIM card, and changes the status from “updating” into “noupdating”, and then enters the “WAIT FOR NETWORKCOMMAND” state. The location updating will be initiated afterthe RR release.If the reject cause is #6, the MS deletes the TMSI, LAI, and CKSNin SIM card, and changes the status from "updating" into“roaming inhibit”. The SIM is regarded invalid until the MS isswitched off or the SIM card is pulled out.1.16.3 Exceptional SituationsI. Re-Establishment Prohibition or FailureWhen MM connection is established, the MM layer may sendan indication to the CC layer. If the MM layer is disconnected,the connection may be re-established through CC request.If the re-establishment is not allowed, and the call is initiatedwithin the establishment or clearing period, the CC layer shallrelease MM connections.If re-establishment is unsuccessful, MM connections shall be

released, and a release indication shall be sent to the CC layer.II. RR Connection FailureIf random access failure or RR CONNECTION FAILURE isdetected by the MS, the MS will stop timer T3230, abort thecall re-establishment procedure, and release all MMconnections.If RR CONNECTION FAILURE is detected by the MSC, the MSCwill abort the call re-establishment procedure and release allMM connections.III. T3230 Time-outIf the T3230 times out, the MS will stop call re-establishmentand release MM and RR connections.1.16.4 SM ProcedureShort messages can be transmitted either on SDCCH or SACCH.A short message procedure can be classified into short messagecalling procedure and called procedure. For details, seeGSM03.40 protocol.1.16.5 Short Message Procedure on SDCCH When MS is callingI. Signaling ProcedureII. Procedure DescriptionThe random access, immediate assignment, authentication,and encryption procedures of short message procedure onSDCCH when MS is calling are the same as general procedures.After encryption, the MS sends SABM again, notifying thenetwork side that this user needs short message service (SMS).Then, BSC provides a transparent-transmission channel for MSto exchange short message information with MSC. In thisprocedure, the MSCs of some manufacturers are capable tosend ASS REQ to BSC, requesting it to assign channel for shortmessage transmission. The time for sending ASS REQ is thesame as that for a common call. BSC can provide SMS either byallocating other channels or by using the original SDCCH.Point to Point short messages protocol is divided intoconnection management layer (CM), relay layer (RL), transportlayer (TL) and application layer (AL).

CP_DATA and CP_ACK are the messages on CM layer, CP_DATAis used to transmit the content of RL and AL message, andCP_ACK is the acknowledgement message of CP_DATA.The release procedure after message is sent is the same asgeneral ones.1.16.6 Short Message Procedure on SDCCH When MS is calledI. Signaling ProcedureII. Procedure DescriptionThe paging response and immediate assignment procedures ofshort message procedure on SDCCH when MS is called are thesame as general procedures. For the short message procedurewhen MS is called, after encryption, the BSC sends EST REQ toMS to establish short message connection. When EST CNF isreceived from MS, the connection is successfully established.BSC transparently transmits the short message till the end ofthe transmission.The release procedure after message is sent is the same asgeneral ones.1.16.7 Short Message Procedure on SACCH When MS is callingI. Signaling ProcedureII. Procedure DescriptionThe MS sends CM SERV REQ through FACCH. The MSC respondswith the CM SERV ACC message and establishes CC layerconnection. Then, it establishes RR layer connection on SACCH,and sends the short message.1.16.8 Short Message Procedure on SACCH when MS is calledI. Signaling ProcedureII. Procedure DescriptionThe BSC receives the CP DATA message from MSC, andestablishes an RR layer connection for SMS. Upon reception ofCP ACK from MS, MSC sends the short message.

1.17 CBSCell Broadcast Service (CBS) is similar to paging stationbroadcast information. It means the mobile network

operator broadcasts the public information to the mobileusers within a certain area. The information that the userscan read is called CBS message. It is generated by theCell Broadcast Entity (CBE) and sent to the CellBroadcast Center (CBC) for processing. After theprocessing, it is forwarded to the BSC and broadcast tothe users through CBCH. The MS can only receive theCBS message in idle mode. Unlike the Point to PointShort Message service, the CBS message is broadcastwithout the acknowledgement of the user terminal.CBS includes:- Common public information service, such as weather,news, stock market, exchange rate, and lottery.- Special public information service, such as peoplesearch, traffic navigation, and call charge prompt.- Advertising service, such as information about stores,restaurants, and theaters.1.17.1 CBS MechanismOperators or information providers can define the cellbroadcast area through CBE. The minimal area is a celland the maximal area can be all the cells of the BSCs thatthe CBC connects with. Features such as intervals,duration, and priority levels can also be specified to meetdifferent requirements. The field length of the CBSmessage sent to BSC from CBC must be 82 bytes. If thelength is shorter than 82 bytes, fill codes are added to it.If the length exceeds 82 bytes, the message is broken to amaximum of 15 pages. If the sending fails, the messagemay be sent again and the message with high prioritylevel is sent first. The CBS information is sent to theproper cells through four continuous SMSBROADCAST REQUEST messages or one SMSBROADCAST COMMAND message. Each CBS

message contains 82-byte user information and 6-byteheader. The CBS message can be sent to BTS in the formof SMS BROADCAST REQUEST or SMSBROADCAST COMMAND. For details, see 1.17.2BTS can send the CBCH Load Indication message toBSC and the system will speed up or delay the messagesending according to this message. Although the BSCconsiders the CBCH capacity when sending the messageand the BTS can indicate the status of the current CBCH,when the CBCH LOAD INDICATION mode is enabled,the BTS can send CBCH LOAD INDICATION torequest for immediate broadcast of the m(1－15)SMSCB timeslot message when the CHCB is idle. Afterthe BSC sends the m timeslot message, it sends messagesaccording to its own schedule. If the message volumethat the BTS requests exceeds the volume that the BSCcan provide, the BSC only sends the messages within itsvolume limit. When the CBCH LOAD INDICATIONmode is enabled, the BTS can send CBCH LOADINDICATION to stop the sending of the m(1－15)timeslot message if overload occurs. Then the BSC willcontinue the sending according to its own schedule.CBCH LOAD INDICATION is only used in DRX mode.The CBCH is of two types: basic CBCH and extendedCBCH. They are four continuous multiframes. The TB ofbasic CBCH is 0, 1, 2, or 3; The TB of extended CBCHis 4, 5, 6, or 7. TB = (FN DIV 51) mod (8).For the basic CBCH, the CBS message head is sent onthe multiframe with TB being 0; for the extended CBCH,it is sent on the multiframe with TB being 4. The systemmessage on BCCH indicates whether the CBS isavailable or not. When SMSCB is used, theBS_AG_BLKS_RES is set as 1 or above. When the

CBCH is mapped to the CCCH+SDCCH/4, the numberof BS_AG_BLKS_RES will not be limited by SMSCB.MS recomposes the CBS message and displays it for theuser.MS obtains the CBS message from the CBCH. BTSinforms MS of the short message information during theschedule in the form of bitmap by sending schedulemessage. There are three reception modes for MS onCBCH:- Non-DRX mode. MS reads the first block of allmessage timeslots. The rest blocks will be read if themessage head indicates that the following timeslots areused. If the MS does not support other reception mode,or it does not receive the scheduling for the next messagetimeslot, Non-DRX mode is used.- First DRX mode. If MS receives the scheduling for thenext message timeslot, but the first scheduling messageof the last scheduling period, or all the information of thelast period or even earlier period is not received, firstDRX mode is used.- Second DRX mode. If MS receives the importantinformation of the last scheduling period and reads thefirst scheduling message of the current period, secondDRX mode is used.Whether the network uses DRX to receive the broadcastshort message can be set through the maintenanceconsole in BSC.1.17.2 BSC-BTS Message Transmission ModeA CBS message consists of eighty eight 8-bit bytes.These bytes are divided into four message blocks witheach block containing twenty two 8-bit bytes. Each blockis added by an 8-bit block type, and the length of theblock is twenty three 8-bit bytes. A CBS message

contains four continuous blocks: first block, secondblock, third block, and fourth block.When the SMS BROADCAST REQUEST mode is used,the message is sent to BTS from BSC. The BSC handlesthe queuing, repetition, and short message sending. Italso considers the CBCH capacity and takes charge ofthe SMS segmentation at radio interface. In the SMSBROADCAST REQUEST message, each SMSCBInformation cell carries a complete frame that can betransmitted on CBCH and the layer 2 information thatindicates the radio path. SMSCB Channel Indicator cellindicates the CHCH used for broadcast. If this cell doesnot provide the information, the basic CBCH will beused.

When the SMS BROADCAST COMMAND mode isused, SMS BROADCAST COMMAND message is sentto BTS from BSC. BSC requires the immediate messagesending during the next CBCH time. The defaultbroadcast mode for BTS can also be set through thismessage. In the default broadcast mode, if there is noother message to broadcast, BTS will send the defaultmessage.

In the SMS BROADCAST COMMAND message, theSMSCB message cell contains the information to bebroadcast on CBCH. It has four continuous blocks with amaximum of 88 bytes. BTS segments the message andestablishes the block format. It also adds bytes to theblock if required. SMSCB Channel Indicator cellindicates the CHCH used for broadcast. If this cell doesnot provide the information, the basic CBCH will beused.

4 GSM Parameter Configurationand Adjustment

When operators prepare to construct a mobilecommunication network, they must predict coverageaccording to traffic prediction and local radiopropagation environment. This guides project designof the system and parameter configuration of radionetwork.

The project design includes the following aspects:

Network topology designSelecting the location of base stationFrequency planningCell parameter configuration

The RF planning determines the coverage range of acell, and the serving range of the cell is determinedbased on the combination of RF planning and cellparameter configuration. By this, the MS always enjoysoptimal services and maximum network capacity atthe best cell.

This chapter discusses the meaning and effect ofimportant parameters in GSM radio communication.Mastering the effect and impact of these parametershelps to configure network parameters and optimizethe network in later stages.

In a GSM network, abundant radio parameters areconfigured according to cells or partial areas; however,the parameter configuration might affect neighborareas. Therefore, while configuring and adjustingparameters, you must pay attention to the impact ofconfiguring parameters on other areas, especiallyneighbor areas.

4.1 Network and Cell ID4.1.1 Cell Global ID

I. Definition

GSM is a global cellular mobile communication system.To ensure that each cell corresponds to a unique IDglobally, the GSM system numbers the following items:

Each GSM network in each countryEach location areaEach base stationEach cell

Numbering the previous items aims as follows:

An MS can identify the serving network sothat the MS can select a network in anyenvironment.The network can obtain the precise locationof the MS so that the network can processvarious service requests involving the MS.

The MS can report information aboutneighbor cells to the network during calling toavoid call drop.

The cell global identity (CGI) is a major networkidentity parameter. CGI consists of location areaidentity (LAI) and cell identity (CI). LAI includes mobilecountry code (MCC), mobile network code (MNC), andlocation area code (LAC).

The system transmits CGI information through systeminformation (SI) transmitted by cell broadcast. Whenan MS receives SI, it demodulates SI for CGIinformation. The MS judge whether to camp on thecell according to the MCC and MNC. It also judgeswhether the current location area changes todetermine updating location. While updating location,the MS reports LAI information to the network so thatthe network can know the location area of the MS.

II. Format

The CGI is MCC-MNC-LAC-CI, with details as follows:

MCC consists of three decimal digits,ranging from 000 to 999.MNC consists of two decimal digits, rangingfrom 00 to 99.LAC ranges from 0 to 65535CI ranges from 0 to 65535.

III. Configuration and Influence

As a globally unique mobile identity, the MCC isuniformly distributed and managed by internationaltelecommunication union (ITU). The MCC for China is460 (decimal).

The MNC is uniformly distributed by statetelecommunication management organs. Now twoGSM networks exist in China. The MNC for ChinaMobile is 00. The MNC for China Unicom is 01.

The method for coding LAC is ruled by each countryaccordingly. This caters for China also (refer to GSMsystem from Ministry of Information Industry). At theearly stage of network construction, the LAC is codedand distributed. The LAC is seldom changed in the laterstages.

The coverage areas related to the LAC is vital in thenetwork. You can configure it as great as possible.

No special restriction is on the distribution of CI. The CIranges from 0 to 65535 (decimal). It must be ensuredthat two equivalent CIs exist in the same location area.This is determined in the system design. Except forspecial situations (such as constructing base stations),the CI must not be changed during the systemoperation.

IV. Precautions

You must pay attention to the following aspects:

The MNC is unchangeable.While configuring the LAC, you must followrelated regulations. Equivalent LACs must notexist in the state network.Equivalent CIs must not exist in the samelocation area.

4.1.2 Base Station Identity Code

I. Definition

In a GSM network, each base station corresponds to adistributed local color code, called base stationidentity code (BSIC). When the MS receives broadcastcontrol channel (BCCH) carriers of two cells at thesame time, with same channel number, the MSdistinguishes them by BSIC.

In network planning, the BCCH carriers of neighborcells are different in frequency to reduce intra-frequency interference. The cellular communicationsystem features that the BCCH carrier might bereused. Therefore, the BSIC of the cells with the sameBCCH carrier must be different.

The system transmits BSIC on synchronization channel (SCH) ofeach cell. The effect of BSIC is as follows:

The BSIC involves in decoding process ofrandom access channel (RACH) to preventbase stations from connecting to the RACHsent to the neighbor cells by the MS by error.After the MS receives SCH messages, itjudges that it has been synchronous to the cell.Decoding information on the downlinkcommon signaling channel correctly requirestraining sequence code (TSC) used oncommon signaling channel.GSM regulations describe TSC in eight fixedformats, and the sequence number of them is0–7. The cell BCC determines the TSC usedby the common signaling channel of a cell.Therefore the BSIC helps inform the MS ofthe TSC used by the common signalingchannel of the serving cell.

In a call, the MS must measure the level ofBCCH carrier of neighbor cells and report it tothe base station according to regulations toneighbor cell list of BCCH. Meanwhile, theMS must provide measured BSIC of thecarrier in the uplink measurement reports.When the neighbor cells of a cell include twoor more cells with the same BCCH carrier, thebase station can distinguish the cells by BSICto avoid incorrect handover. In a call, the MS must measure signals ofneighbor cells, and sends measurement reportsto the network. The measurement report cancontain information about six neighbor cellsonly, so the MS must be controlled to reportthe cells actually related to handover. The firstthree digits of BSIC (namely, NCC) aims aspreviously mentioned. Operators control theMS to report the neighbor cell informationpermitted by the serving cell NCC bybroadcast parameters NCC permitted.

II. Format

The BSIC is NCC-BCC, with details as follows:

The NCC ranges from 0 to 7.The BCC ranges from 0 to 7.

III. Configuration and Influence

Usually different GSM PLMNs use the same frequencyresource, but, to some degree, their network planningis independent. The neighbor GSM PLMNs usedifferent NCCs according to regulations. This ensures

that the neighbor base stations with same frequencyuse different BSICs.

The BCC is part of the BSIC. It helps identify differentbase stations with same BCCH carrier number in thesame GSM PLMN. The values of BCC must meet theprevious requirements. According to GSM regulations,the TSC of cell BCCH carrier must be same as that ofcell BCC. The equipment providers must ensure theTSC consistency.

IV. Precautions

The neighbor cells or cells nearby using the same BCCHcarrier must use different BSICs. Especially when twoor more cells use the same BCCH carrier in theneighbor cell list of a cell, theses cells must usedifferent BSIC. Pay attention to cells at the borderingareas between provinces and cities, and otherwisecross-cell handover might fail and abundant mistakenaccess problems might occur.

4.2 Paging and Access ControlParameters4.2.1 Number of Access Grant Reserved Blocks(BS_AG_BLK_RES or AG)

I. Definition

The common control channel consists of accessgrant channel (AGCH) and paging channel(PCH).

For different CCCHs, each BCCH multiframe(including 51 frames) contains CCCH messageblocks different number. The CCCH is shared byAGCH and PCH. According g to regulations,partial message blocks on CCCH are especiallyreserved for AGCH. This avoids that the AGCHmessages are blocked when the PCH traffic isgreat.The number of parameter access grant reservedblocks (AG) refers to the number of messageblocks reserved for AGCH on CCCH in eachBCCH multiframe.

II. Format

The AG ranges from 0 to 2 when CCCH sharesphysical channel (CCCH_CONF = 1) with stand-alone dedicated control channel (SDCCH).The AG ranges from 0 to 5 when CCCH does notshare physical channel (CCCH_CONF=0) withstand-alone dedicated control channel (SDCCH).

III. Configuration and Influence

When the channel combination of the cell isfixed, the parameter AG adjusts the ratio ofAGCH and PCH in CCCH. When the PCH isidle, it can send immediate assignment messages.The AGCH does not transmit paging messages.Equipment operators can balance AGCH andPCH by adjusting AG, with the followingprinciples.The principle for AG value is that based on nooverload of AGCH, you must reduce theparameter to shorten the time for MS to respond

to paging, and to improve system serviceperformance. When the immediate assignmentmessages are superior to paging messages to besent, configure AG to 0.The value of AG is recommended as follows:

AG is 1 when the CCCH and SDCCHshare a physical channel. AG is 2 or 3 in other situations.

In network operation, take statistics of overloadsituations of AGCH and adjust AG accordingly.By default the immediate assignment messagesare superior to paging messages to be sent in thenetwork, so you need not reserve a channel forimmediate assignment messages. In thissituation, configure AG to 0.

4.2.2 Frame Number Coding Between IdenticalPaging

Frame number coding between identical pagingis BS_PA_MFRMS (MFR for short).

I. Definition

According to GSM regulations, each MS(corresponding to an IMSI) belongs to a paginggroup (for calculation of paging groups, seeGSM regulation 05.02). Each paging group in acell corresponds to a paging subchannel.According to its IMSI, the MS calculates thepaging group that it belongs to, and thencalculates the location of paging subchannel thatbelongs to the paging group. The MS onlyreceives the signals of the paging subchannel thatit belongs to, and neglects that of other paging

subchannels. In addition, the MS even powers offsome hardware of itself during other pagingsubchannel to lower power cost of itself.The number of paging channel multiframe(MFR) is the number of multiframes used in aperiod of paging subchannel. The MFRdetermines the number of paging subchannelsthat the cell PCH is divided into.

II. Format

The MFR ranges from 2 to 9, which respectivelymeans that the same paging group cycles in aperiod of 2 to 9 multiframes.

III. Configuration and Influence

According to the definition of CCCH, AG, andMFT, you can calculate the number of pagingchannel in each cell.

When the CCCH and SDCCH share aphysical channel, there is (3 - AG) MFRs. When the CCCH and SDCCH share aphysical channel, there is (9 - AG) MFRs.

According to the previous analysis, the greaterthe MFR is, the more the paging channels of thecell are (see the calculation of paging groups inGSM regulation 05.02). Theoretically, thecapacity of paging channels does not increasewith the increase of MFR. The number of buffersfor buffering paging messages on each basetransceiver station (BTS) increases. The pagingmessages are sent more evenly both in time andspace, so it seldom occurs that the pagingmessages overflow in the buffers so call lost

occurs (related to functions by equipmentproviders).However, to enjoy the previous advantages, youwill have a longer delay of paging messages onthe radio channels. The greater the MFR is, thegreater the delay of paging messages in the spaceis, and the lower the average service performanceof the system is. Therefore, the MFR is animportant parameter in network optimization.The following principle caters for configuringMFR:The configured strategy for buffers of eachequipment provider is different, so you mustselect the MFR properly so that the pagingmessages do not overflow on PCH. Based onthis, configure the parameter as small as possible.In addition, you must measurement the overflowsituations of PCH periodically while the networkis running, and adjust MFR accordingly.

IV. Precautions

Any paging message of the same location areamust be sent to all cells in the location areas atthe same time, so the PCH capacity of each cellin the location area must be equivalent or close toeach other. Otherwise, you must consider smallerPCH capacity as the evidence for designinglocation area.

4.2.3 Common Control Channel Configuration(CCCH-CONF)

I. Definition

The CCCH includes AGCH and PCH. It sendsimmediate assignment messages and pagingmessages. In each cell, all traffic channels(TCHs) share CCCH. According to the TCHconfiguration and traffic model of the cell, theCCCH can be one or more physical channels. Inaddition, the CCCH and SDCCH share a physicalchannel. The combination methods for CCH aredetermined by CCCH parameter CCCH_CONF.

II. Format

The CCCH_CONF consists of three bits, with the codingmethods listed in CCCH configuration codingCCCH configuration coding

CCCH_CONF

MeaningNumber of CCCH messageblocks in a BCCH multiframe

000One physical channel forused for CCCH, not sharedwith SDCCH

9

001One physical channel forused for CCCH, sharedwith SDCCH

3

010Two physical channels forused for CCCH, not sharedwith SDCCH

18

100Three physical channelsfor used for CCCH, notshared with SDCCH

27

110Four physical channels forused for CCCH, not sharedwith SDCCH

36

III. Configuration and Influence

When the CCCH and SDCCH share one physicalchannel, the CCCH has the minimum channelcapacity. When the CCCH and SDCCH do not

share a physical channel, the more physicalchannels that the CCCH uses, the greater thecapacity is.The CCCH_CONF is determined by theoperators based on combination of cell trafficmodel and paging capacity of the location areawhere a cell belongs to. It is determined insystem design, and adjusted in networkexpansion. According to experiences, when thepaging capacity in the location area is not highand cell has one or two carriers, it isrecommended that the CCCH uses one physicalchannel and share it with SDCCH (incombination CCCH methods). This spares aphysical channel for paging. Otherwise, themethod that CCCH and SDCCH do not share onephysical channel is used.When the cell TRX exceeds 6 and CCCHOVERLOAD occurs in the cell, it isrecommended that the CCCH uses two or morebasic physical channel and does not share themwith SDCCH.

IV. Precautions

The CCCH_CONF must be consistent with theactual configuration of cell CCCH. In addition,you must consider the influence on the accessgrant reserved blocks.

4.2.4 Extended Transmission Slots (TX_INTEGER)

I. Definition

In a GSM network, a random access channel(RACH) is an ALOH. To reduce the conflicting

times on RACH when an MS accesses thenetwork, and to increase RACH efficiency, GSMregulations (sections 3.3.1.2 of 04.08) prescribethe compulsory access algorithm for MS. Thealgorithm defines three parameters as follows:

Extended transmission slots T Maximum retransmission times RET TIt is the number of slots between two sendingwhen the MS keeps sending multiple channelrequest messages. SIt is related to channel combination, and is anintermediate variable of access algorithm. It isdetermined by T and CCCH configuration.

II. Format

The value of T is from 3 to 12, 14, 16, 20, 25, 32,and 50.The value of S ranges as listed in Values of S

Values of S

T

S in different CCCH combination methods

The CCCH and SDCCHdoes not share a physicalchannel

The CCCH and SDCCH share aphysical channel

3, 8, 14, 50 55 41

4, 9, 16 76 52

5, 10, 20 109 58

6, 11, 25 163 86

7, 12, 32 217 115

III. Configuration and Influence

To access the network, the MS must originate animmediate assignment process. To begin theprocess, the MS sends (RET + 1) channel requestmessages on RACH. To reduce conflicts onRACH, the time for MS to send channel requestmessages must meet the following requirements:

The number of slots (not including slotsfor sending messages) between originatingimmediate assignment process by MS andsending the first channel request messages israndom. Its range is {0, 1, …, MAX (T, 8) -1}. When the MS originates the immediateassignment process, it takes a value from therange according to even distributionprobability. The number of slots (not including slotsfor sending messages) between a channelrequest message and the next is from {S, S +1, …, S + T - 1} according to evendistribution probability.

According to previous analysis, the greater the Tis, the larger the range of intervals between onechannel request message and the next, and theless the RACH conflicting times is. The greaterthe S is, the greater the interval between onechannel request message and the next, the lessthe RACH conflicting times is, and the moreefficiently the SDCCH is used. However, theincrease of T and S leads to longer time for MSto access the network, so the access performanceof the whole network declines. Therefore youmust configure T and S properly.

S is calculated by MS according to T andcombination of CCH. You can configure T freelyand sends it to MS by system information.Usually, you need configure T properly to makeT + S as small as possible (to reduce the time forMS to access the network); meanwhile you mustensure an effective assignment of SDCCH toavoid overload (for all random access requests,the system does not distinguish whether they arefrom the same MS, but assigns a SDCCH). Inoperation, you can adjust the value according totraffic measurement of cell immediateassignment.

4.2.5 Minimum Access Level of RACH

I. Definition

The minimum access level of RACH is the levelthreshold for the system to judge whether there isa random access request.

II. Format

The minimum access level of RACH ranges from0 to 63 (corresponding to –110 dBm to –47dBm).The unit is level grade value.

III. Configuration and Influence

When the access burst level of RACH is greaterthan the threshold, the BTS judges that there isan access request. The BTS, together with theparameter random access error threshold,determines whether the random access burst isvalid. To configure the parameter properly, you

must combine actual sensitivity of the basestation and the parameter minimum receivedlevel permitted for MS to access. This preventsthe MS from failing in calling though there aresignals. The access burst level of RACH affectscall drop rate and access range (coverage), so youmust pay attention to the influence on access ofMS.

4.2.6 Random Access Error Threshold

I. Definition

GSM protocols prescribe that by relativity ofjudgment training sequence (41 bits) the systemcan judge whether the received signals are therandom access signals of MS.

II. Format

The value ranges from 0 to 255. Therecommended value is 180.

III. Configuration and Influence

The random access error threshold defines therelativity of training sequence. If the smaller it is,the more errors of random access signalspermitted by the network are, the easily the MSrandomly accesses the network, and the greaterthe report error rate is. If the greater the randomaccess error threshold is, the smaller the reporterror rate is, and the more difficult the access tothe network is when signals are weak. Seeprotocol 0408, 0502.

The system requires the random access errorthreshold transferred by current bit of 41 bittraining sequence.

90–100 33

101–120 34

121–140 35

141–160 36

161–175 37

176–195 38

196–221 39

222–243 40

244–250 41

0–89 or251–255

38

The two parameters random access errorthreshold and minimum access level of RACHdetermine the validity of random access burst.

4.2.7 Access Control Class (ACC)

I. Definition

GSM regulations (02.11) prescribe that eachGSM user (common user) corresponds to anaccess class, ranging from class 0 to class 9. Theaccess class is stored in SIM of mobile users. Forspecial users, GSM regulations reserves fivespecial access classes, ranging from class 11 toclass 15. Theses classes are prior to other classesin accessing. Special users might have one ormore access classes (between 11 and 15), whichare also stored in user SIM. Users of class 11 to15 are prior to that of class 0 to 9. However, the

class between 0 and 9 or between 11 and 15 doesnot mean priority.The access class is distributed as follows:

Class 0–9: common users Class 11: users for PLMN management Class 12: users for security departments Class 13: common business departments(in charge of water, gas) Class 14: emergency services Class 15: PLMN staff

Users of class 0–9 have its access rights cateringfor home PLMN and visited PLMN. Users ofclass 11 and 15 have its access rights catering forvisited PLMN only. Users of class 12, 13, and 14have its access rights catering for in the countrywhere home PLMN belongs to.

II. Format

The access control class consists of two parts: Common access control classValue range: a check option, including class 0disabled, …, class 9 disabled.Recommended value: all 0. Special access control classValue range: a check option, including class11 disabled, …, class 15 disabled.Recommended value: all 0.

If a class is configured to 1, it means that accessis forbidden. For example, a common accessclass is configured to 1000000000; commonusers excluding class 0 users can access thenetwork.

III. Configuration and Influence

C0–C15 (excluding C10) are set by equipmentroom operators. Usually these bits are configuredto 1. Proper configuration contributes to networkoptimization as follow:

When installing a base station, starting abase station, or maintaining and testing insome cells, configure C0–C15 (excludingC10) to 1. In this way, different users areprevented from accessing the network, so theinstalling and maintenance is less influenced. During busy hours of cells with hightraffic, congestion occurs, RACH conflictingtime increase, AGCH traffic overloads, andAbis interface traffic overloads. When youconfigure class of some users to 1, you canreduce the traffic of the cell.

4.2.8 Maximum Retransmission Times (RET)

I. Definition

See GSM regulation 04.08. When an MSoriginates an immediate assignment process, itsends a channel request message to the networkon RACH. The RACH is an ALOH, so the MScan send multiple channel request messagesbefore receiving immediate assignmentmessages, to increase access success rate of MS.The maximum retransmission times M (RET) isdetermined by equipment room operators, andsent to MS by SI.

II. Format

The maximum retransmission times consists oftwo bits, with the meanings listed in Coding ofmaximum transmission times M

Coding of maximum transmission times MM maximum transmission times

00 1

01 2

10 4

11 7

III. Configuration and Influence

The greater the M is, the higher the success rateof call attempt is, and the higher the connectionrate is, but the load of RACH, CCCH, andSDCCH increase. In cell with high traffic, if theRET is over great, overload of radio channels andcongestion occur, so the connection rate andradio resource utilization declines sharply. If theRET is over small, the call attempt times of MSreduces, success rate reduces, so the connectionrate reduces. Therefore, proper configuration ofRET for each cell help utilize network radioresources and improve connection rate.For configuration of RET M, refer to thefollowing methods:

For areas with low traffic, such as insuburban or rural areas, configure RET to 7 toincrease the access success rate of MS.

For areas with average traffic, such ascommon urban areas, configure RET to 4.

For microcell with high traffic and of apparentcongestion, configure RET to 1.

4.2.9 Control Class of MS Maximum TransmitPower (MS-TXPWR-MAX-CCH)

I. Definition

MS-TXPWR-MAX-CCH is sent in BCCH SIs. Itaffects behavior of MS in idle mode. It is alsoused in calculating C1 and C2, and determinescell selection and reselection.

C1 = RLA_C - RXLEV_ACCESS_MIN -MAX((MS_TXPWR_MAX_CCH - P), 0) RLA_C: average received level by MS RXLEV_ACCESS_MIN: minimumreceived level permitted for MS to access MS_TXPWR_MAX_CCH: maximumpower level of control channel (control classof MS maximum transmit power) P: Maximum transmit power level of MS

II. Format

The range of MS-TXPWR-MAX-CCH is 0–31.For cells of GSM900 and GSM1800, the dBmvalues corresponding to the control class aredifferent.

In a GSM900 network, the 32 control classof maximum transmit power corresponding to0–31 is as follows:{39, 39, 39, 37, 35, 33, 31, 29, 27, 25, 23, 21,

19, 17, 15, 13, 11, 9, 7, 5, 5, 5, 5, 5, 5, 5, 5, 5,5, 5, 5, 5} In a GSM1800 network, the 32 controlclass of maximum transmit powercorresponding to 0–31 is as follows:{30, 28, 26, 24, 22, 20, 18, 16, 14, 12, 10, 8, 6,4, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 36,34, 32}

Recommended values are 5 for GSM900 and 0for GSM1800.

III. Configuration and Influence

MS-TXPWR-MAX-CCH determines the powerclass used before MS receives power controlmessages. For details, see protocol 0508.The smaller it is, the greater the output power ofMS is. The MS near the base station interfereswith neighbor channels of the cell, so the accessto the network by other MSs and communicationquality are influenced. The greater it is, thesmaller the output power of MS is, and the lowerthe access success rate of MS at cell borders is.You must configure MS-TXPWR-MAX-CCHproperly according to the serving range of thecell.

4.2.10 Power Offset (POWEROFFSET)

I. Definition

When the MS accesses the network and before itreceives the initial power control messages, allGSM900 MSs and type 1 and type 2 DCS1800MSs use MS_TXPWR_MX_CCH of BCCH. Ifthe MS_TXPWR_MX_CCH exceeds the

maximum transmit power of MS, the MS usesthe closest power.The parameter POWEROFFSET is effective totype 3 DCS1800 MSs. When the type 3DCS1800 MS accesses the network, it use totalpower of MS_TXPWR_MX_CCH +POWEROFFSET before receiving the initialpower control message. See protocol GSM0508.

II. Format

The values of 0–3 correspond to 0 dB, 2 dB, 4dB, and 6 dB.The recommended value is 2.

III. Configuration and Influence

The greater the parameter is, the more easily thetype 3 DCS1800 MS accesses the network. Agreat POWEROFFSET enables MS to access thenetwork afar, but does not help control cross-cellinterference, so the network quality is influenced.

4.2.11 IMSI Attach/Detach Allowed

I. Definition

The IMSI detach means that the MS informs thenetwork of itself work state changing fromworking to non-working. Usually it refers towhen the MS powers off or the SIM is taken offMS. After receiving the inform from MS, thenetwork sets the IMSI as in non-working state.The IMSI attach is opposite of IMSI detach. Itmeans that MS informs the network of itselfwork state changing to working. Usually it refersto when the MS powers on or the SIM is put into

MS again. After the MS turns to working stateagain, it detects whether the current locationareas (LAI) is the same as that recorded in MS atlast.

If yes, the MS starts IMSI attach process(this is one of location updating). If no, the MS starts location updatingprocess of cross location area.

After receiving the location updating message orIMSI message from MS, the network sets theIMSI as in working state.The parameter IMSI attach/detach allowed(ATT) is used for informing MS of the IMSIattach/detach process.

II. Format

The value of ATT includes YES/NO. NO meansthat starting IMSI attach/detach process by MS isforbidden. YES means that starting IMSIattach/detach process by MS is compulsory.

III. Configuration and Influence

Usually configure ATT to YES so that thenetwork will not process the proceeding of theMS after the MS powers off. This frees systemresources (such as PCH).

IV. Precautions

The ATT of different cells in the same locationarea must be the same to avoid abnormalitieswhile the MS is called. For example, in a cellwith YES as the value of ATT, when the MSpowers off, it starts IMSI detach process.

Therefore the network records that the MS is innon-working state, so it does not page the MS. Ina cell with No as the value of ATT and the cellbeing different from the one where the MSpowers off, when the MS powers on again in thecell, the MS does not start IMSI attach process.In this situation, the MS cannot be callednormally until it starts location updating process.

4.2.12 Direct Retry (DR)

I. Definition

During the assignment process of call setup,congestion might cause assignment failure. Theassignment failure causes failure of the wholecall. GSM networks has a function to avoid suchfailures, namely, DR. The DR is that the BSSdirectly assign MS to TCH of neighbor cells. Theparameter is used by system to set whether toallow direct retry function.

II. Format

The value of DR includes YES and NO. YESmeans that the system allows directional retry.NO means that the system does not supportdirection retry function.

III. Configuration and Influence

DR improves call success rate. If conditions areready, start DR. On the contrary, DR is that theBSS directly assign MS to TCH of neighbor cellswhen congestion occurs in the cell where the MScamps, so the MS can originates a call in the non-best cell with lowest received level, and extra

interference might be brought about in frequencyreuse networks. Therefore, you must use thefunction properly according to comprehensivenetwork situations.

4.3 Serial Parameters of Cell Selectionand Reselection4.3.1 cell_bar_access

I. Definition

In the SI broadcasted in each cell, a bit indicateswhether the MS is allowed to access the networkin the cell, namely, cell_bar_access.

II. Format

The value of cell_bar_access includes 1 and 0.The value 0 indicates that MS is allowed toaccess the network from the cell. The value 1indicates that the MS is barred to access thenetwork from the cell. Actually whether to allowMS to access the network from the cell isdetermined by both cell_bar_access andcell_bar_qualify.

III. Configuration and Influence

The cell_bar_access is configured by equipment roomoperators. Usually the MS is allowed to access thenetwork from all the cells, so cell_bar_access isconfigured to 0. In special situations, the operators wantsome cell for handover service only, so cell_bar_accessis configured to 1The MS usually works in microcells (you can configurethe priority of cells and reselection parameters to enable

this). When the MS is calling while moving fast, thenetwork force MS to hand over to the base station G. Thesignals of base station G are stronger than microcell basestation in most areas. When the call terminates, the MSjust camps near base station G and at edge of microcellcells, the MS will not reselect a cell according to GSMregulations, therefore the MS cannot return to microcell.

The capacity of base station G is usually small,so the previous phenomenon leads to congestionof base station G. To solve the problem, you canconfigure the cell_bar_access to 1, namely, toforbid MS directly accessing base station G. Inarea A, handover is allowed to base station G.

IV. Precautions

The cell_bar_access is used only in some specialareas. For common cells, it is configured to 0.

4.3.2 cell_bar_qualify

I. Definition

The cell_bar_qualify determines the priority ofcells, namely, it enables MS to select some cellby preference.

II. Format

The value of cell_bar_qualify includes 1 and 0.The cell_bar_qualify and cell_bar_accessdetermine the priority state of cells, as listed inTable 7-1 Cell priorit.

Table 7-1 Cell prioritiescell_bar_qualify

cell_bar_access

Cell selectionpriority

Cell reselection state

0 0 Normal Normal

0 1 Barred Barred

1 0 Low Normal

1 1 Low Normal

An exception is that the cell selection priorityand cell reselection state are normal when thefollowing conditions are met:

The cell belongs to the PLMN which theMS belongs to. The MS is in cell test operation mode. The cell_bar_access is 1. The cell_bar_qualify is 0. The access control class 15 is disabled.

III. Configuration and Influence

The priority of all the cells are usually configuredto normal, namely, cell_bar_qualify = 0. Inmicrocell and dualband networking, operatorsmight want MS to camps on the cell of some typeby preference. In this situation, the equipmentroom operators can configure the priority of thesecells to normal and other cells to low.During cell selection, when the proper cells withnormal as the priority is not present (proper cellsmeans that all parameters meet the conditions forcell selection, namely, C1 > 0, and the cell isallowed to access), the MS will select cells withlow priority.

IV. Precautions

Pay attention to the following aspects: When cell priority is used as a method tooptimize network, the cell_bar_qualify onlyaffects cell selection, without any influence on

cell reselection. You must optimize thenetwork by combining cell_bar_qualify andC2. During cell selection, when the propercells with normal as the priority is not present,the MS will select cells with low priority.Therefore when the level of the cell withnormal priority is low, and cells with lowpriority and high level are present, the MSwill access the network slowly whilepowering on.

4.3.3 Minimum Received Level Allowing MS toAccess (RXLEV_ACCESS_MIN)

I. Definition

To avoid bad communication quality, call drop,and a waste of network radio resources due toMS accessing the network at low received signallevel, GSM regulations prescribe that when anMS accesses the network the received level mustbe greater than the threshold level, namely, theminimum received level allowing MS to access.

II. Format

The value range of RXLEV_ACCESS_MIN isfrom –110 dBm to –47 dBm.

III. Configuration and Influence

The recommended RXLEV_ACCESS_MINneeds to be approximately equal to the receivingsensitivity of MS. The RXLEV_ACCESS_MINaffects cell selection parameter C1, so it is

important to traffic adjustment and networkoptimization.For cells with over high traffic and severecongestion, you can increaseRXLEV_ACCESS_MIN. In this way, the C1 andC2 of the cells decrease, and the effectivecoverage range decreases. You must notconfigure RXLEV_ACCESS_MIN over great,because this might cause non-seamless coverageand complaints for signal fluctuation. It isrecommended that the RXLEV_ACCESS_MINis smaller than or equal to –90 dBm.

IV. Precautions

Except for areas of high density of base stationsand of qualified coverage, adjusting cell trafficby RXLEV_ACCESS_MIN is not recommended.

4.3.4 Additional Reselection Parameter Indicator

I. Definition

The cell selection and reselection by MS dependson the parameters C1 and C2. Whether C2 is thecell reselection parameter is determined bynetwork operators. Additional reselectionparameter indicator (ADDITIONALRESELECT) informs MS of whether to use C2in cell reselection.

II. Format

ADDITIONAL RESELECT consists of 1 bit. InSI3, it is meaningless, and equipmentmanufacturers configure it to N. The MS usesADDITIONAL RESELECT of SI4.

When ADDITIONAL RESELECT isconfigured to N, the meaning is: if the restbytes of SI4 (SI4RestOctets) are present, theMS must abstract and calculate parametersrelated to C2 and related cell reselectionparameter PI. When ADDITIONAL RESELECT isconfigured to Y, the meaning is that the MSmust abstract and calculate parameters relatedto C2 and related cell reselection parameterPI.

III. Configuration and Influence

Cells seldom use SI7 and SI8, so you canconfigure ADDITIONAL RESELECT to N.When cells use SI7 and SI8, and the parameterC2 is used in cell reselection, you can configureADDITIONAL RESELECT to Y.

4.3.5 Cell Reselection Parameter Indicator

I. Definition

The cell reselection parameter indicator(CELL_RESELECT_PARAM_IND) is used ininforming MS of whether C2 is a cell reselectionparameter and whether C2 is present.

II. Format

The value of CELL_RESELECT_PARAM_INDincludes Y and N, with the meanings as follows:

Y: The MS must calculate C2 byabstracting parameters from SIs of cellbroadcast, and set C2 as the standard for cellreselection.

N: The MS must set C1 as the standard,namely, C2 = C1.

III. Configuration and Influence

The equipment room operators determine thevalue of PI. Configure PI to Y if related cells setC2 as the standard for cell reselection; otherwise,configure it to N.

4.3.6 Cell Reselection Offset, Temporary Offset, andPenalty Time

I. Definition

After the MS selects a cell, without great changeof all the conditions, the MS will camp on theselected cell. Meanwhile, it does as follow:

Starts measuring signals level of BCCHcarrier in neighbor cells. Records the 6 neighbor cells with greatestsignal level. Abstract various SI and controlinformation of each neighbor cell from the 6cells.

When conditions are met, the MS hands overfrom the selected cell to another. This process iscalled cell reselection. The conditions include:

Cell priority Whether the cell is barred to access Radio channel level (important)

When the signal level of neighbor cells exceedsthat of the serving cell, cell reselection occurs.The channel level standard used in cellreselection is C2, with the calculation as follows:

1) When PENELTY_TIME ≠ 11111:C2 = C1 + CELL_RESELECT_OFFSET -TEMPORARY_OFFSET * H(PENALTY_TIME - T)Wherein, if PENALTY_TIME - T (x) < 0, thefunction H(x) = 0; if x ≥ 0, H(x) = 1.2) When PENELTY_TIME = 11111:C2 = C1 - CELL_RESELECT_OFFSET

T is a timer, with 0 as the initial value. When acell is listed by MS in the list of cells withmaximum signal level, start T with step of4.62ms (a TDMA frame). When the cell isremoved from the list, the associated T is reset.After cell reselection, the T of original cell worksas PENALTY_TIME. Namely, temporary offsetis not performed on the original cell.CELL_RESELECT_OFFSET (CRO) modifiescell reselecting time C2.TEMPORARY_OFFSET (TO) is supplementedto C2 from starting working of T to theprescribed time.PENALTY_TIME is the time forTEMPORARY_OFFSET having effect on C2.When PENALTY_TIME = 11111, the MS isinformed of using C2 = C1 – CRO.CELL_RESELECT_OFFSET,TEMPORARY_OFFSET, andPENALTY_TIME are cell reselectionparameters.

When the cell reselection parameter PI is1, the MS is informed of receiving values ofthree parameters on BCCH.

If PI is 0, the MS judges that the previousthree parameters are 0, namely C2 = C1.

If the C2 of a cell (in the same location area asthe serving cell) calculated by MS is greater thanthe C2 of the cell where MS camps, and this lastsfor over 5s, the MS reselects to camp on the cell.If the C2 of a cell (in different location area asthe serving cell) calculated by MS is greater thanthe sum of C2 of the cell where MS camps andcell reselect hysteresis, and this lasts for over 5s,the MS reselects to camp on the cell.The interval between two reselections is at least15s, and this avoids frequent cell reselection byMS.C2 is formed on the combination of C1 andartificial offset parameters. The artificial offsetparameters help MS camp on or prevent MSfrom camping on some cell. This balances thetraffic of the network.

II. Format

1) The cell reselection offset (CRO) is indecimal, with unit of dB. It ranges from 0 to63, which means 0 to 126 dB (2 dB as thestep). The recommended value is 0.2) The temporary offset (TO) is in decimal,with unit of dB. It ranges from 0 to 7, whichmeans 0 to 70 dB (10 dB as the step). Therecommended value is 0.3) The penalty time (PT) is in decimal, withunit of second. It ranges from 0 to 31. Thevalue 0 to 30 means 20s to 620s (20s as thestep). The value 31 is reserved for changing

the effect direction of C2 by CRO. Therecommended value is 0.

III. Configurationa and Influence

The previous parameters can be adjustedaccordingly in the following three situations:

1) When the communication quality is baddue to heavy traffic or other causes, changethe parameters to enable MS not camps on thecell (the cell is exclusive from the MS). Forthis situation, configure PT to 31, so TO isineffective. C2 = C1 – CRO. The C2 isartificially lowered. So the probability for MSto reselect the cell decreases. In addition, theequipment room operators can configure CROto a proper value according to the exclusivelevel of the cell by MS. The greater theexclusion is, the greater the CRO is.2) For cells with low traffic and equipmentof low utilization, change the parameters toenable MS to camp on the cell (the cell isprior). In this situation, configure CRO to 0–20 dB according to the priority. The higher thepriority is, the greater the CRO is. TO isconfigured the same as or a little greater thanCRO. PT helps avoid over frequent cellreselection, the recommended value of PT is20s or 40s.3) For cell with average traffic, configureCRO to 0, PT to 11111 so that C2 = C1. Noartificial influence is on the cell.

IV. Precautions

In whatever situations, the CRO must not begreater than 30 dB, because over great CROleads to unstable network, such as complaintsabout signal fluctuation.

4.3.7 Cell Reselection Hysteresis (CRH)

I. Definition

CRH affects cell reselection of cross locationarea. The MS starts cell reselection if thefollowing conditions are met:

The signal level of neighbor cell (indifferent location area) is greater than that ofthe serving cell. The difference between the signal levels ofthe neighbor cell and the serving cell must begreater than the value prescribed by cellreselection hysteresis.

The difference is based on the cell reselectionmethods used by MS. If the MS reselects a cellwith C2, then compare values of C2.

II. Format

CRH is in decimal, with unit of dB. The range is0 to 14, with step of 2 dB. The recommendedvalue is 4.

III. Configuration and Influence

If the original cell and target cell belongs todifferent location areas, the MS must originate alocation updating process after cell reselection.Due to the attenuation feature of radio channels,the C2 of two cells measured at the borderingarea of neighbor cells fluctuates much, so the MS

reselect cells frequently. The interval betweentwo reselections is over 15s, which is rather shortfor location updating. The signal flow of networkincreases sharply, radio resources cannot be fullyutilized.During location updating, the MS cannot respondto paging, so the connection rate decreases.Adjust CRH according to signal flow andcoverage. When signal flow overloads or locationupdating of cross location area is frequent, thecell reselection hysteresis is increased asrecommended. You must avoid abnormalcoverage due to over large location area.

IV. Precautions

Do not configure CRH to 0 dB.

4.4 Parameters Affecting NetworkFunctions4.4.1 Newly Established Cause Indicator (NECI)

I. Definition

In a GSM network, the traffic channel (TCH)consists of full-speed TCH and half-speed TCH.When the network supports half-speed TCH, theMS is informed of whether the area supportshalf-speed TCH by NECI.

II. Format

The value of NECI includes Y and N, with themeaning as follows:

Y means that the area support half-speedTCH. N means that the area cannot support half-speed TCH.

III. Configuration and Influence

Half-speed TCHs enable each carrier to supportmore traffic channel, but you must confirmwhether the system support half-speed TCH.

4.4.2 Power Control Indicator (PWRC)

I. Definition

The PWRC informs MS of whether to takestatistics of downlink level of BCCH carrier slotfor measuring average value when the BCCHfrequency participates in frequency hopping. Thecauses to configuring PWRC are as follows:

GSM regulations allow frequency hoppingchannels to use BCCH (frequency hopping notin BCCH slots) . GSM regulations allow downlink powercontrol over frequency hopping channels. The MS needs signal level of themeasured neighbor cells, so the power of eachslot on BCCH frequency is prohibited tochange. The downlink power control does notinvolve carrier slots for BCCH which includesthe frequency hopping.

For previous causes, when the MS measures theaverage downlink channel level with commonmethods, the measurement result is inaccurate forpower control because the average value includesthe downlink received level of BCCH carriers the

power of which are not controlled, so themeasurement report is inaccurate for powercontrol.To avoid the influence on power control, whenthe MS calculates average received level duringfrequency hopping, the received level obtainedfrom BCCH carrier slot must be removed (seeGSM regulations 05.08).

II. Format

The value of PWRC includes 0 and 1, withmeanings as follows:

When PWRC is 0, the measurement resultby MS includes BCCH carrier. When PWRC is 1, the measurement resultby MS does not include BCCH carrier.

III. Configuration and Influence

The PWRC is usually configured to 0. Configureit to 1 if all the following conditions are met:

Channels have frequency hopping on twoor more frequencies. One of the frequency is BCCH carrierfrequency. The system uses downlink power control.

IV. Precautions

The value of PWRC depends actually on thefollowing parameters:

Whether to use frequency hopping. Whether the hopping frequency includesBCCH carrier.

Whether the system uses downlink powercontrol.

4.4.3 Discontinuous Transmit of Uplink

I. Definition

Discontinuous transmit of uplink (DTXU) refersto the process for MS not to transmit signalsduring silent period (see description about DTXin Chapter 2).

II. Format

Whether the network allows uplink to usediscontinuous transmit (DTX) is set byequipment room operators. DTX ranges from 0to 2, with the following meanings:

0: MS can use DTXU. 1: MS must use DTXU. 2: MS cannot use DTXU.

III. Configuration and Influence

Using uplink DTX affects call quality, but it ishelpful in the following aspects:

Lower interference to radio channels.Due to this, the average call quality ofnetwork is improved. Cut power consumption by MS

For the previous advantages, DTX isrecommended to use.

4.4.4 Discontinuous Transmit of Downlink

I. Defintion

Discontinuous transmit of downlink (DTXD)means the network does not transmit signalsduring silent period.

II. Definition

DTXD is in string, and the range is YES and NO.The meanings are as follows:

YES: Downlink uses DTX. NO: Downlink does not use DTX.

III. Configuration and Influence

Using downlink DTX affects call quality in alimit scale, but it is helpful in the followingaspects:

Lower interference to radio channels.Due to this, the average call quality ofnetwork is improved. Reduce load of base station CPU

Therefore, if possible, you use DTX.

IV. Precautions

According to GSM regulations, downlink DTX isoptional. If the base station equipment supportsDTXD, then use it. However, you must ensurethat voice transcoder is available to supportDTXD.

4.4.5 Call Resetup Allowed

I. Definition

When coverage voids cause radio link failure,consequently call drop, the MS starts to resetupthe call for recovery. Whether resetting up the

call is allowed depends on the parameter callresetup allowed (RE).

II. Format

The values of call resetup allowed are 1 and 0,with meanings as follows:

1: Call resetup is allowed in the cell. 0: Call resetup is forbidden in the cell.

III. Configuration and Influence

When a connected MS passes coverage voids,call drop occurs easily. If call resetup is allowed,the average call drop rate (CDR) is lowered.However, call resetup takes longer time, andmost users disconnects before completion of callresetup. Therefore call resetup is difficult toachieve, and even wastes abundant radioresources. In a word, call resetup is disabled.

4.4.6 Emergency Call Allowed

I. Definition

The following MSs cannot enjoy variousservices:

MS without SIM MS with ACC as one of C0 to C9 and withcell_bar_access

The parameter emergency call allowed (EC)determines whether the MS is allowed foremergency calls, such as police emergency call.

II. Format

EC consists of 1 bit. For the MS with ACC of C0to C9 or without SIM, the EC is NO, meaning

emergency call forbidden. YES meansemergency call allowed. For the MS with ACCof C11 to C15, when both the access control bitand EC are configured to forbidden, it isforbidden for emergency calls.

III. Configuration and Influence

According to the GSM regulations, theemergency number is 112, different from that inChina. The Chinese emergency call cannotfunction as prescribed in GSM regulations. Forinternational roaming users, set 112 toanswerphone to inform users of various specialservice numbers. Therefore, setting emergencycall must be allowed through configuring radioparameters, namely, configure EC to 1.

4.4.7 Early Classmark Sending Control

I. Definition

In a GSM network, the MS classmark marks thefollowing aspects:

Service capacity Supported frequency band Power capacity Encryption capacity

Classmark consists of classmark1, classmark2,and classmark3. A GSM MS. In a GSM network,the MS reports Classmark1 or Classmark2information immediately after ESTIND<CMSERV REQ> (corresponding to L2-SABM atUm interface) is allocated. Classmark3 (CM3)information includes power information ofvarious frequency band of multi-frequency MS.

During handover between different bands, thepower class must be correctly described. Whenthe GSM system pages and transmits BA2 indifferent bands, it must know the CM3 message.In GSM regulation Phase2plus, early classmarksending control (ECSC) is added. ECSC meansthat by SI the system informs MS of reportingClassmark3 after link setup. This avoids queryingprocess by network.

II. Format

The values of ECSC are Y and N, with thefollowing meanings:

Y: The MS reports Classmark3 to thenetwork immediately after link setup. N: The MS is forbidden to report itsClassmark3 to network initiatively.

III. Configuration and Influence

The major information of Classmark3 is fordualband network, so do as follows:

Configure ECSC to N in single frequencyGSM application areas. Configure ECSC to Y in dualband GSMapplication areas.

IV. Precautions

In a dualband network, configure the parameterof all cell to the same value. Configuring theparameter to different values in one or more cellsis forbidden; otherwise, the network qualitydeclines.

4.5 Frequency Hopping Parameters

4.5.1 Frequency Hopping Sequence Number

I. Definition

In a GSM network, the cell allocation (CA)means the set of carriers used by each cell,recorded as {R0, R1, …, Rn - 1}. Wherein, Riindicates the absolute channel number. For eachcommunication process, the set of carriers usedby base station and MS is mobile allocation(MA), recorded as {M0, M1, …, Mn - 1}.Wherein, Mi indicates the absolute channelnumber. Obviously MA is a subset of CA.During a communication process, the airinterface uses a carrier number, one element ofMA. The variable mobile allocation index (MAI)determines an exact element of MA. Accordingto the frequency hopping algorithm in GSMregulation 05.02, the MAI is the TDMA framenumber (RN) or reduced frame number (RFN),frequency hopping sequence number (HSN), andmobile allocation index offset (MAIO).Wherein, the HSN determines two aspects:

Track of frequency points duringfrequency hopping The asynchronous neighbor cells using thesame MA can avoid continuous frequencycollision during frequency hopping by usingdifferent HSNs.

II. Format

HSN is in decimal, ranging from 0 to 63,wherein:

0: cyclic frequency hopping

1–63: pseudo frequency hopping

III. Configuration and Influence

You can choose any HSN in cells usingfrequency hopping, but you must ensure that thecells using same frequency group must usedifferent HSN. The following paragraph is anexception:In an 1X1 network, three cells under a basestation use the same frequency group, but theyare synchronous cells because of same FN.Therefore the three cells use the same HSN. Youmust plan MAIO properly to avoid frequencycollision of the three cells under the same basestation.

4.5.2 Mobile Allocation

I. Definition

The mobile allocation (MA) in the GSM networkindicates a frequency set for frequency hopping.Namely, when the MA of a cell is fixed, thecommunication frequency points of the cellperforms transient in the set by MA according torules.The parameter MA determines all the elements inMA.

II. Format

MA is a set, with all GSM frequency points as itselement, namely:

For GSM900 networks: 1–124 and 975–1023. For GSM1800 networks: 512–885

III. Configuration and Influence

MA is configured according to networkdesigning requirements.

IV. Precautions

Chinese GSM networks do not cover all availablefrequency bands of GSM system, so configureMA in available frequency bands.The number of elements in each MA set cannotexceed 63.The MA cannot include BCCH carriers.The number of MA must not be multiples of 13 ifall the following conditions are met:

Using DTX HSN = 0 (cyclic frequency hopping)

You must avoid SACCH to appear usually at thesame frequency point.

4.5.3 Mobile Allocation Index Offset

I. Definition

During communication, the air interface uses acarrier frequency, one element of MA set. MIOdetermines an exact element of MA set.According to the frequency hopping algorithm inGSM regulation 05.02, the MAI is the TDMAframe number (RN) or reduced frame number(RFN), frequency hopping sequence number(HSN), and mobile allocation index offset(MAIO). MAIO is an initial offset of MAI, and itaims to avoid multiple channels to use the samefrequency carrier in the same time.

II. Format

MAIO ranges from 0 to 63.

III. Configuration and Influence

MAIO is configured by equipment roomoperators.

IV. Precautions

The different cells using same group of MA mustuse consistent MAIO.Using different MAIOs enables different sector

4.6 Distance Control Parameters4.6.1 Call Clearing

I. Definition

Call clearing (CallClearing) means that themaximum allowed distance threshold is clearedbetween MS and base station in talk.

II. Format

CallClearing ranges from 0 to 63, with unit ofTA.

III. Configuration and Influence

Configure CallClearing according to actualcoverage range of a cell. Proper configuration ofCallClearing helps check whether the handoverthreshold of the cell is properly defined,especially for urban cells.If the call is frequently cleared after CallClearingthreshold is defined according to cell radium,probably the handover threshold is improperly

configured. This is due to that the MS cannothand over to the best server cell after exceedingdesigned coverage range.Define CallClearing according to msRangeMax,namely, CallClearing > msRangeMax.In actual network operation, call clearing isunusually performed, because radio link fails dueto over poor coverage before call clearing.Defining CallClearing aims to restrict thedistance between MS and base station and toavoid MSs in allowed coverage range to interfereother MSs, especially in areas with complexlandform.The cell coverage range is irregular, so islandeffect might occur. For this phenomenon, defineCallClearing to clear calls in island areas.

4.6.2 TA Handover Threshold (MSRANGEMAX)

I. Defintion

When the distance between MS and base stationreaches or exceeds MSRANGEMAX, distancehandover is triggered.

II. Format

MSRANGEMAX ranges from 0 to 63, with unitof TA. The reference is 63.

III. Configuration and Influence

MSRANGEMAX must be smaller thanCallClearing, and otherwise the handoverfunction will be actually unavailable. Whileconfiguring MSRANGEMAX, you must adjustthe threshold of other types of handover;

otherwise ping-pong handover occurs. oneoccasion might be as follows:The distance between MS and the serving cellexceeds the threshold, but the signals of targetcell are weaker than that of original cell.Consequently the PowerBudget handover istriggered immediately after distance handover istriggered.

4.6.3 TA Restriction (MS_BS_DIST_USED)

I. Definition

The maximum allowed access distance betweenbase station and MS. If the distance between anMS and base station exceeds the maximumallowed access distance, the MS is forbidden toaccess cells.

II. Format

The range is 0 to 63, with unit of TA. Thereference is 63.

III. Configuration and Influence

For its configuration, refer to the method forconfiguring CallClearing. Adjust the parameterto enable it consistent with the geographiccoverage range of the cell. Set a proper thresholdto filter pseudo RACH requests to avoidunnecessary assigning SDCCH.According to tests, for mountain-mounted basestations, the coverage and interference is difficultto control. If you define the maximum allowedaccess distance to 63, the RACH misjudgmentincreases (the system demodulates interference to

RACH bursts by mistake). Therefore the radioperformance and traffic measurement indexes ofthe cell are affected.

4.7 Radio Link Failure Process andParameters

The radio link failure is detected from uplink anddownlink. The MS completes downlinkdetection, while the base station completes uplinkdetection.

4.7.1 Radio Link Failure Counter (RLC or RadioLink Timeout)

I. Definition

The MS originates call resetup or disconnects byforce if all the following conditions are met:

The voice or data quality is too poor to bereceived. Power control and handover cannot help toimprove the quality.

A disconnection by force actually brings about acall drop, so the MS considers it a radio linkfailure that the voice or data service is actuallytoo poor to be received. GSM regulations providesolutions to the previous problems as follows:Set a counter S in the MS. The initial value of Sis provided at the beginning of talk, and it is thevalue of the parameter radio link failure counter.S changes as follows:

S decreases by 1 if the MS fails indecoding a correct SACCH message when theMS should receive the SACCH message. S increases by 2 if the MS succeed indecoding a correct SACCH message.

S cannot exceed the value for radio link failurecounter. When S equals to 0, the MS originatescall resetup or disconnects by force.

II. Format

The step from 4 to 64 is 4, with unit of SACCHperiod as follows:

For TCH, the SACCH period is 480ms. For SDCCH, the SACCH period is 470ms.

III. Configuration and Influence

The value of the parameter radio link failurecounter affects CDR and utilization of radioresources.Assume that cell A is a neighbor cell to cell Band the bordering coverage is poor. When an MSmoves from P to Q while in talk,

If the radio link failure counter is oversmall, call drop occurs before cross-cellhandover. If the radio link failure counter is overgreat, the network releases related resourcesuntil radio link expires, though the voicequality is too poor when MS camps on cell Bnear P. Therefore, the utilization of radioresources declines.

Proper configuration of radio link failure counteris important, and is related to the actual

situations. To configure radio link failurecounter, refer to the following rules:

Configure it to between 52 and 64 in areaswith over low traffic. Configure it to between 36 and 48 in areaswith low traffic and great coverage radium Configure it to between 20 and 32 in areaswith heavy traffic.

IV. Precautions

Configure radioLinkTimeout to smaller thanT3109. This contributes to success of call resetupand avoids the following situation effectively:Before the MS releases radio resources due toexpiration, the network side completes releasingchannels resources and reallocates resources toother MSs. Therefore two MSs might use thesame slot and this causes interferences even calldrop.

4.7.2 SACCH Multiframe (RLTO_BS)

I. Definition

Refer to the description of radio link failurecounter. A counter is set accordingly to radio linkat base station side for managing radio linkfailures. The solutions vary due to differentequipment providers, but a general method is asfollows:Set a counter S in the base station. The initialvalue of S is provided at the beginning of talk,and it is the value of the parameter radio linkfailure expiration. S changes as follows:

S decreases by 1 if the MS fails indecoding a correct SACCH message when theMS should receive the SACCH message. S increases by 2 if the MS succeed indecoding a correct SACCH message.

S cannot exceed the value for radio linkexpiration of base station. When S equals to 0,the MS originates call resetup or disconnects byforce, as shown in .لم یتم العثور على مصدر المرجع! خطأ .

II. Format

RLT0_BS ranges from 4 to 64.

III. Configuration and Influence

Proper configuration of radio link expiration ofbase station affects CDR and utilization of radioresources. It is related to the actual situations. Toconfigure radio link failure counter, refer to thefollowing rules:

Configure it to between 52 and 64 in areaswith over low traffic. Configure it to between 36 and 48 in areaswith low traffic and great coverage radium Configure it to between 20 and 32 in areaswith heavy traffic. Configure it to a greater value in areaswith apparent voids or where call drop occursfrequently while the MS moves.

IV. Precautions

RLT0_BS and RLC must be consistent.

4.8 Handover and Related Parameters4.8.1 PBGT Handover Threshold (HoMargin)

I. Definition

The PBGT handover threshold is powerhandover tolerance (handover in serving areas).When the signal level of neighbor cell ishoMargin (dB) higher than that of the servingcell, handover occurs. Complex radiopropagation conditions cause fluctuation ofsignal level. Using handover tolerance avoidsfrequent handover at bordering areas. The PBGThandover threshold is similar to HO_MARGIN(GSM 05.08).

II. Format

The PBGT handover threshold ranges from 0 to127, corresponding to –64 dB to +63 dB. Thereference value for suburban areas is 68. Thereference value for urban areas is 70 to 72.

III. Configuration and Influence

The PBGT handover threshold aims to adjusthandover difficulty properly, and to avoid ping-pong handover. If it is configured over great, thehandover is delayed and handover is lessefficient. When it is smaller than 64, the MShands over from the serving cell to the neighborcell with lower level.

4.8.2 Minimum Downlink Power of HandoverCandidate Cells (rxLevMinCell)

I. Definition

It is the minimum allowed access level for a cellto be a neighbor cell. When the cell levelmeasured by MS is greater than the threshold, theBSS list the cell into candidate cell list forhandover judgment.

II. Format

It ranges from –110 dBm to –47 dBm.

III. Configuration and Influence

It is helpful in the following two aspects: It guarantees communication quality.For a common single layer network structure,the value ranges from –90 dBm to –80 dBm. It helps allocate traffic between cellsaveragely.Especially in multi-layer network structure, tomaintain MS in a network layer, you canincrease the level of the cell of the networklayer (such as –70 dBm), and also decreasethat in other cells.

IV. Precautions

You cannot configure rxLevMinCell over great(over –65 dBm) or over small (lower than –95dBm), and otherwise communication quality isaffected.

4.8.3 Handover Threshold at Uplink Edge

I. Definition

If the uplink received level keeps being smallerthan the handover threshold at uplink edge for aperiod, edge handover can be performed.

II. Format

It ranges from 0 to 63, corresponding to –110dBm to –47 dBm. The recommended values areas follows:

Configure it to 25 in urban areas withoutPBGT handover. Configure it to 20 in single site ofsuburban areas. Configure it to 20 in urban areas withPBGT handover

III. Configuration and Influence

When PBGT handover is enabled, thecorresponding edge handover threshold can belowered. When PBGT handover is disabled, andthe edge handover threshold is over low, anartificial cross-cell non-handover occurs.Therefore call drop occurs or intra-frequency andside interference occur due to cross-cell talk.

4.8.4 Handover Threshold at Downlink Edge

I. Definition

If the downlink received level keeps beingsmaller than the handover threshold at downlinkedge for a period, edge handover can beperformed.

II. Format

It ranges from 0 to 63, corresponding to –110dBm to –47 dBm. The recommended values areas follows:

Configure it to 30 in urban areas withoutPBGT handover. Configure it to 25 in single site ofsuburban areas. Configure it to 25 in urban areas withPBGT handover

III. Configuration and Influence

When PBGT handover is enabled, thecorresponding edge handover threshold can belowered. When PBGT handover is disabled, andthe edge handover threshold is over low, anartificial cross-cell non-handover occurs.Therefore call drop occurs or intra-frequency andside interference occur due to cross-cell talk.

4.8.5 Downlink Quality Restriction of EmergencyHandover

I. Definition

If the downlink received quality is lower than thethreshold of downlink quality restriction ofemergency handover, the quality differenceemergency handover occurs.

II. Format

It ranges from 0 to 70, corresponding to RQ(QoS 0 to 7) x 10.The recommended value is 50.

III. Configuration and Influence

When frequency hopping is enabled, the voicequality is better with the same RQ, you canconfigure it to 60 or 70. When emergency

handover occurs, the intracell handover occursfirst. If there are no other candidate cells, and theintracell handover is enabled, the intracellhandover occurs.

4.8.6 Uplink Quality Restriction of EmergencyHandover

I. Definition

If the uplink received quality is lower than it,quality difference emergency handover istriggered.

II. Format

It ranges from 0 to 70, corresponding to RQ(QoS 0 to 7) x 10.The recommended value is 50.

III. Configuration and Influence

When frequency hopping is enabled, the voicequality is better with the same RQ, you canconfigure it to 60 or 70. When emergencyhandover occurs, the intracell handover occursfirst. If there are no other candidate cells, and theintracell handover is enabled, the intracellhandover occurs.

4.8.7 Uplink Quality Threshold of InterferenceHandover

I. Definition

It is the uplink received quality threshold of theserving cell that triggers interference handover.

The interference handover is triggered if all thefollowing conditions are met:

The uplink received level is higher thanthe uplink received power threshold ofinterference handover. The uplink received quality is lower thanthe uplink quality threshold of interferencehandover.

When handover switch is enabled, theinterference handover occurs within the cell bypreference.

II. Format

It ranges from 0 to 70, corresponding to RQ(QoS 0 to 7) x 10.The recommended value is 50.

III. Configuration and Influence

When frequency hopping is enabled, the voicequality is better with the same RQ, you canconfigure it to 60 or 70. When interferencehandover is triggered, select the candidatesaccording to the sorted result. If the serving cellranks first and its intracell handover is enabled,the MS selects the serving cell; otherwise itselects the second candidate cell.

4.8.8 Downlink Quality Threshold of InterferenceHandover

I. Definition

It is the downlink received quality threshold ofthe serving cell that triggers interference

handover. The interference handover is triggeredif all the following conditions are met:

The downlink received level is higher thanthe downlink received power threshold ofinterference handover. The downlink received quality is lowerthan the downlink quality threshold ofinterference handover.

When handover switch is enabled, theinterference handover occurs within the cell bypreference.

II. Format

It ranges from 0 to 70, corresponding to RQ(QoS 0 to 7) x 10.The recommended value is 50.

III. Configuration and Influence

When frequency hopping is enabled, the voicequality is better with the same RQ, you canconfigure it to 60 or 70. When interferencehandover is triggered, select the candidatesaccording to the sorted result. If the serving cellranks first and its intracell handover is enabled,the MS selects the serving cell; otherwise itselects the second candidate cell.

IV. Precautions

The interference handover quality must be betterthan emergency handover quality.

4.8.9 Uplink Received Power Threshold ofInterference Handover

I. Definition

If interference handover occurs due to uplinkquality, the serving cell must reach the minimumuplink received power threshold. If this is met,the system judges that uplink is interfered, sointerference handover is triggered.The interference handover is triggered if all thefollowing conditions are met:

The uplink received level is higher thanthe uplink received power threshold ofinterference handover. The uplink received quality is lower thanthe uplink quality threshold of interferencehandover.

When handover switch is enabled, theinterference handover occurs within the cell bypreference.

II. Format

It ranges from 0 to 63, corresponding to –110dBm to –47 dBm.The recommended value is 25.

III. Configurationa and Influence

When interference handover is triggered, selectthe candidates according to the sorted result. Ifthe serving cell ranks first and its intracellhandover is enabled, the MS selects the servingcell; otherwise it selects the second candidatecell.

4.8.10 Downlink Received Power Threshold ofInterference Handover

I. Definition

If interference handover occurs due to uplinkquality, the serving cell must reach the minimumdownlink received power threshold. If this is met,the system judges that downlink is interfered, sointerference handover is triggered.The interference handover is triggered if all thefollowing conditions are met:

The downlink received level is higher thanthe downlink received power threshold ofinterference handover. The downlink received quality is lowerthan the downlink quality threshold ofinterference handover.

When handover switch is enabled, theinterference handover occurs within the cell bypreference.

II. Format

It ranges from 0 to 63, corresponding to –110dBm to –47 dBm.The recommended value is 30.

III. Configurationa and Influence

When interference handover is triggered, selectthe candidates according to the sorted result. Ifthe serving cell ranks first and its intracellhandover is enabled, the MS selects the servingcell; otherwise it selects the second candidatecell.

4.8.11 Maximum Repeated Times of PhysicalMessages (NY1)

I. Definition

In asynchronous handover process of GSMsystem, when the MS receives handovermessages of the network, it sends handoveraccess messages on the target channel. After thenetwork receives the message, it does as follows:

1) Calculate related RF features.2) Send physical messages (it the channelmessages are encrypted, start encryption anddecryption algorithm) in unit data to MSs.3) Start timer T3105.

If the network does not receive correct layer 2frames sent by MS until expiration of T3105, thenetwork will resend the physical message andrestart T3105. The maximum times for resendingphysical messages is determined by theparameter maximum repeated times of physicalmessages (NY1)

II. Format

NY1 ranges from 0 to 254.The recommended value is 20.

III. Configuration and Influence

When the network receives the handover accessmessages sent by MS, the physical channel(PCH) needs to be synchronous. If thecommunication quality on channels isguaranteed, the MS can receive physicalmessages correctly and send layer 2 frames to thenetwork.If the physical messages are sent multiple times,and the network cannot receive layer 2 frames

sent by MS, the PCH is too poor to communicatenormally. Though link is setup after multipletrials, the communication quality is notguaranteed. This lowers the utilization of radioresources. Therefore configure NY1 to a smallervalue.

IV. Precautions

Configuring NY1 is affected by T3105. If T3105is configured to a short value, then the NY1needs to be increased accordingly.If a handover trial fails before the original cellreceives the HANDOVER FAILURE message,and the T3105 of the target cell expires for Nytimes, the target BTS sends a CONNECTIONFAILURE INDICATION message to the targetBSC. Though the MS might return to the originalchannel, the traffic measurement counters frommultiple vendors will take statistics of connectionfailure.To avoid the previous phenomenon, configureT3105 as follows:Ny * T3105 > T3124 + delta (delta: the timebetween expiration of T3124 and receivingHANDOVER FAILURE message by originalBTS)

4.8.12 Multiband Indicator (multiband_reporting)

I. Definition

In a single band GSM network, when the MSsend measurement reports of neighbor cells to the

network, it needs to report the content of the sixneighbor cells with strongest signals.In a multiband network, operators wish that MSuses a band by preference in cross-cell handover.Therefore the MS sends measurement reportsaccording to signal strength and signal band. Theparameter multiband indicator indicates MS toreport content of multiband neighbor cells.

II. Format

The multiband indicator ranges from 0 to 3, withmeanings as follows:

0: According to signal strength ofneighbor cells, the MS must report six allowedmeasurement reports of neighbor cells withstrongest signals and known NCC, with theneighbor cells in whatever band. 1: The MS must report the allowedmeasurement report of a neighbor cell withknown NCC and with strongest signals at eachband expect for the band used by the servingcell. The MS must also report the neighborcells of the band used by the serving cell inrest locations. If there are other rest locations,the MS must report conditions of otherneighbor cells in any band. 2: The MS must report the allowedmeasurement report of two neighbor cells withknown NCC and with strongest signals at eachband expect for the band used by the servingcell. The MS must also report the neighborcells of the band used by the serving cell inrest locations. If there are other rest locations,

the MS must report conditions of otherneighbor cells in any band. 3: The MS must report the allowedmeasurement report of three neighbor cellswith known NCC and with strongest signals ateach band expect for the band used by theserving cell. The MS must also report theneighbor cells of the band used by the servingcell in rest locations. If there are other restlocations, the MS must report conditions ofother neighbor cells in any band.

III. Configuration and Influence

In multiband networks, it is related to traffic ofeach band. For configuration, refer to thefollowing rules:

If the traffic of each band is approximatelyequal, and operators do not select a bandintentionally, you can configure the multibandindicator to 0 If the traffic of each band is obviouslydifferent, and operators want MS to select aband by preference, you can configure themultiband indicator to 3. For situations between the previous two,configure multiband indicator to 1 or 2.

4.8.13 Permitted Network Color Code (nccpermitted)

I. Definition

During a talk, the MS must report the measuredsignals of neighbor cells to the base station, buteach report includes only six neighbor cells.

Therefore the MS is configured to report thepotential handover target neighbor cells, insteadof reporting unselectively and according to signallevel.To enable previous functions, restrict MS tomeasure the cells with the fixed network colorcode (NCC). The NNC allowed by parameterslist the NCCs of the cells to be measured by MS.The MS compares the measured NCC ofneighbor cells and NCCs set allowed byparameters. If the measured NCC is in the set, theMS reports the NCC to the base station;otherwise, the MS discard the measurementreport.

II. Format

The parameter ncc permitted is a bit mappingvalue, consisting of 8 bits. The most significantbit is bit 7 while the least significant bit is bit 0.Each bit corresponds to an NCC code 0 to 7 (seeGSM regulations 03.03 and 04.08).If the bit N is 0 (N ranges from 0 to 7), the MSneeds not to measure the level of the cell withNCC of N. Namely, it only measures the signalquality and level of the cells corresponding to bitnumber of 1 in NCC and ncc permittedconfiguration.

III. Configuration and Influence

Each area is allocated with one or more NCCs. Inthe parameter ncc permitted of the cell, the localNCC is absolutely and only included. Ifexcluded, abnormal handover and call drop

occur. For normal roaming between areas, theNCC of neighbor areas must be included in theedge cells of an area.

IV. Precautions

Improper configuration of the parameter causesnormal handover and even call drop. Theparameter only affects behaviors of MS.

4.9 Power Control and RelatedParameters4.9.1 Maximum Transmit Power of MS(MSTXPWRMX)

I. Definition

The transmit power of MS in communication iscontrolled by BTS. According to the uplinksignal strength and quality, power budget result,the BTS controls MS to increase or decrease itstransmit power.

Note:In any situation, power control is prior to relatedhandover for BSS. Only when the BSS fails toimprove uplink signal strength and voice qualityto the prescribed level, it starts handover.

To reduce interference between neighbor cells,the power control of MS is restricted. Namely,

the BTS controls MS to transmit power withinthe threshold.MSTXPWRMX is the maximum transmit powerof MS controlled by BTS.

II. Format

MSTXPWRMX ranges from 0 to 31.The dBm values corresponding to GSM900 andGSM1800 cells are different:

The 32 maximum transmit power controlclasses for GSM900 are {39, 39, 39, 37, 35,33, 31, 29, 27, 25, 23, 21, 19, 17, 15, 13, 11,9, 7, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5} The 32 maximum transmit power controlclasses for GSM900 are {30, 28, 26, 24, 22,20, 18, 16, 14, 12, 10, 8, 6, 4, 2, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0, 36, 34, 32}

III. Configuration and Influence

Configuring MSTXPWRMX helps controlinterferences between neighbor cells, because:

If MSTXPWRMX is over great, theinterference between neighbor cells increases. If MSTXPWRMX is over small, the voicequality declines and improper handover mightoccur.

4.9.2 Received Level Threshold of Downlink PowerIncrement (LDR)

I. Definition

When the downlink received level of the servingcell is smaller than a threshold, the network must

start power control to increase the transmit powerof base station and to guarantee communicationquality of MS.The received level threshold of downlink powerincrement defines the downlink received levelthreshold. When the downlink level received byMS is smaller than it, the base station startspower control to increase its transmit power.The parameter N1 means that at lease N1sampling points must be measured before startinghandover algorithm.The parameter P1 means the level of at least P1sampling points in N1 sampling points is smallerthan the threshold prescribed by received levelthreshold of downlink power increment.

II. Format

It ranges from –110 dBm to –47 dBm.N1 ranges from 1 to 32.P1 ranges from 1 to 32.

III. Configuration and Influence

The received level is between –60 dBm and –80dBm in a GSM network, so configure receivedlevel threshold of downlink power increment to –85 dBm.N1 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N1 tobetween 3 and 5.Configure P1 to about 2/3 of N1.

4.9.3 Received Level Threshold of Uplink PowerIncrement (LUR)

I. Definition

When the uplink received level of the servingcell is smaller than a threshold, the network muststart power control to increase the transmit powerof MS and to guarantee communication qualityof MS.The received level threshold of uplink powerincrement defines the uplink received levelthreshold. When the uplink level received by MSis smaller than it, the base station starts powercontrol to increase MS transmit power.The parameter N1 means that at lease N1sampling points must be measured before startinghandover algorithm.The parameter P1 means the level of at least P1sampling points in N1 sampling points is smallerthan the threshold prescribed by received levelthreshold of uplink power increment.

II. Format

It ranges from –110 dBm to –47 dBm.N1 ranges from 1 to 32.P1 ranges from 1 to 32.

III. Configuration and Influence

The received level is between –60 dBm and –80dBm in a GSM network, so configure receivedlevel threshold of uplink power increment to –85dBm.

N1 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N1 tobetween 3 and 5.Configure P1 to about 2/3 of N1.

4.9.4 Received Quality Threshold of Downlink PowerIncrement (LDR)

I. Definition

When the downlink received quality of theserving cell is smaller than a threshold, thenetwork must start power control to increase thetransmit power of base station and to guaranteecommunication quality.The received quality threshold of downlinkpower increment defines the downlink receivedlevel threshold. When the downlink qualityreceived by MS is smaller than it, the base stationstarts power control to increase its transmitpower.The parameter N3 means that at lease N3sampling points must be measured before startinghandover algorithm.The parameter P3 means the quality of at least P3sampling points in N3 sampling points is smallerthan the threshold prescribed by received qualitythreshold of downlink power increment.

II. Format

It ranges from 0 to 7, the voice quality grade.N3 ranges from 1 to 32.P3 ranges from 1 to 32.

III. Configuration and Influence

The received quality is 0 to 2 of quality grade ina GSM network, so configure received qualitythreshold of downlink power increment to –85dBm.N3 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N3 tobetween 3 and 5.Configure P3 to about 2/3 of N3.

4.9.5 Received Quality Threshold of Uplink PowerIncrement (LUR)

I. Definition

When the uplink received quality of the servingcell is smaller than a threshold, the network muststart power control to increase the transmit powerof MS and to guarantee communication quality.The received quality threshold of uplink powerincrement defines the uplink received qualitythreshold. When the uplink quality received byMS is smaller than it, the base station startspower control to increase transmit power of MS.The parameter N3 means that at lease N3sampling points must be measured before startinghandover algorithm.The parameter P3 means the quality of at least P3sampling points in N3 sampling points is smallerthan the threshold prescribed by received qualitythreshold of uplink power increment.

II. Format

It ranges from 0 to 7, the voice quality grade.N3 ranges from 1 to 32.P3 ranges from 1 to 32.

III. Configuration and Influence

The received quality is 0 to 2 of quality grade ina GSM network, so configure received qualitythreshold of uplink power increment to 3.N3 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N3 tobetween 3 and 5.Configure P3 to about 2/3 of N3.

4.9.6 Received Level Threshold of Downlink PowerDecrement (UDR)

I. Definition

When the downlink received level of the servingcell is greater than a threshold, the network muststart power control to decrease the transmitpower of base station and to decreaseinterference to radio channels.The received level threshold of downlink powerdecrement defines the downlink received levelthreshold. When the downlink level received byMS is greater than it, the base station startspower control to decrease its transmit power.The parameter N2 means that at lease N2sampling points must be measured before startinghandover algorithm.The parameter P2 means the level of at least P2sampling points in N2 sampling points is greater

than the threshold prescribed by received levelthreshold of downlink power decrement.

II. Format

It ranges from –110 dBm to –47 dBm.N1 ranges from 1 to 32.P1 ranges from 1 to 32.

III. Configuration and Influence

The received level is between –60 dBm and –80dBm in a GSM network, so configure receivedlevel threshold of downlink power decrement to–85 dBm.N2 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N2 tobetween 3 and 5.Configure P2 to about 2/3 of N2.

4.9.7 Received Level Threshold of Uplink PowerDecrement (UUR)

I. Definition

When the uplink received level of the servingcell is greater than a threshold, the network muststart power control to decrease the transmitpower of MS and to decrease interference toradio channels.The received level threshold of uplink powerdecrement defines the uplink received levelthreshold. When the uplink level received by MSis greater than it, the base station starts powercontrol to decrease transmit power of MS.

The parameter N2 means that at lease N2sampling points must be measured before startinghandover algorithm.The parameter P2 means the level of at least P2sampling points in N2 sampling points is greaterthan the threshold prescribed by received levelthreshold of uplink power decrement.

II. Format

It ranges from –110 dBm to –47 dBm.N2 ranges from 1 to 32.P2 ranges from 1 to 32.

III. Configuration and Influence

The received level is between –60 dBm and –80dBm in a GSM network, so configure receivedlevel threshold of uplink power decrement to –60dBm.N2 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N2 tobetween 3 and 5.Configure P2 to about 2/3 of N2.

4.9.8 Received Quality Threshold of Downlink PowerDecrement (UDR)

I. Definition

When the downlink received quality of theserving cell is greater than a threshold, thenetwork must start power control to decrease thetransmit power of base station and to decreasespace interference.

The received quality threshold of downlinkpower decrement defines the downlink receivedquality threshold. When the downlink qualityreceived by MS is greater than it, the base stationstarts power control to decrease transmit powerof MS.The parameter N4 means that at lease N4sampling points must be measured before startinghandover algorithm.The parameter P4 means the quality of at least P4sampling points in N2 sampling points is greaterthan the threshold prescribed by received qualitythreshold of downlink power decrement.

II. Format

It ranges from 0 to 7, the voice quality grade.N4 ranges from 1 to 32.P4 ranges from 1 to 32.

III. Configuration and Influence

The received quality is 0 to 2 of quality grade ina GSM network, so configure received qualitythreshold of downlink power decrement to 0.N4 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N4 tobetween 3 and 5.Configure P4 to about 2/3 of N4.

4.9.9 Received Quality Threshold of Uplink PowerDecrement (UUR)

I. Definition

When the uplink received quality of the servingcell is greater than a threshold, the network muststart power control to decrease the transmitpower of MS and to decrease space interference.The received quality threshold of uplink powerdecrement defines the uplink received qualitythreshold. When the uplink quality received byMS is greater than it, the base station startspower control to decrease transmit power of MS.The parameter N4 means that at lease N4sampling points must be measured before startinghandover algorithm.The parameter P4 means the quality of at least P4sampling points in N4 sampling points is greaterthan the threshold prescribed by received qualitythreshold of uplink power decrement.

II. Format

It ranges from 0 to 7, the voice quality grade.N4 ranges from 1 to 32.P4 ranges from 1 to 32.

III. Configuration and Influence

The received quality is 0 to 2 of quality grade ina GSM network, so configure received qualitythreshold of uplink power decrement to 0.N4 is related to propagation quality of radiochannels within cell coverage range. To reduceinfluence by attenuation, configure N4 tobetween 3 and 5.Configure P4 to about 2/3 of N4.

4.9.10 Power Control Interval (INT)

I. Definition

It takes a period from beginning of power controlto detection of effect of power control. Thereforean interval must exist between continuous twopower controls; otherwise the system becomesunstable and even call drop occurs.The parameter power control interval (INT)configures the minimum interval between twocontinuous times of power control.

II. Format

It ranges from 0 to 31s.

III. Configuration and Influence

According to frame structure of GSM network,configure INT to about 3s.

IV. Precautions

INT cannot be smaller than 1s, and otherwise thesystem becomes unstable.

4.9.11 Power Increment Step (INC)

I. Definition

The INC indicates the power increment of MS orbase station in power control.

II. Format

The range of INC is 2 dB, 4 dB, or 6 dB.

III. Configuration and Influence

The recommended value is 4 dB.

4.9.12 Power Decrement Step (RED)

I. Definition

The RED indicates the power decrement of MSor base station in power control.

II. Format

The range of RED is 2 dB or 4 dB.

III. Configuration and Influence

The recommended value of RED is 2 dB.

4.10 Systematic Important Timers4.10.1 T3101

I. Definition

T3101 is the BSC timer controlling time ofimmediate assignment process.

II. Format

T3101 ranges from 0 to 255s. The recommendedvalue is 3s.

III. Configuration and Influence

In an immediate assignment process, the BSCrequires BTS to provide SDCCH to set upsignaling channel. When the BSC sends achannel activation message, T3101 starts timing.When the BSC receives the setup instruction sentby BTS, T3101 stops timing. When T3101expires, the system releases corresponding

SDCCH resources. Proper configuration ofT3101 reduces congestion due to dualassignment SDCCH effectively.The greater the T3101 is, the longer theinefficient time for using signaling resources is.For example, if the extended transmission delayis improperly configured (usually the sum of Tand S is over small), the MS fails in respondingto the network side, so the MS resends therandom access request message.Therefore, the network side will assign SDCCH(the network cannot distinguish the repeatedsending access request from the first send). Forbetter use of signaling resources, especially inactivating queue function, you must configureT3101 to a smaller value. The minimum intervalfor sending channel activation message andreceiving setup indicator is 600ms. For non-overload BSS, the maximum interval is 1.8s.

4.10.2 T3103

I. Definition

In inter- and intra-BSS handover, the BSCdetermines the time for keeping TCH both inhandover-originated cell and target cell. Whenthe time receives handover completion (intra-BSC) or clearing (inter-BSC) message, T3103stops.

II. Format

T3103 ranges from 0 to 255s. The recommendedvalue is 5s.

III. Configuration and Influence

The following paragraph is an example of inter-BSS handover.When T3103 receives the handover command, itis reset and starts timing. When it receivesclearing command, it is reset. This means thatT3103 reserves two channels when it is timing,one channel for source BSC, and one channel fortarget BSC. If it is over long, two channels areoccupied for a long time and resources might bewasted.According to the tests, if the NSS timer isproperly configured, the handover process occurswithin 5s. Therefore, the recommended value is5s.

4.10.3 T3105

I. Definition

See the protocol 0408 and 0858. When sendingphysical information, the network starts T3105.If the timer expires before receiving any correctframes from MS, the network resends physicalinformation and restarts the T3105. Themaximum repeated times is Ny1.

II. Format

T3105 ranges from 0 to 255, with unit of 10ms.

III. Configuration and Influence

The physical information is sent on FACCH. Thetime for sending four TDMA in a time onFACCH is about 18ms. If the next physical

information is just sent 18ms after the first one,probably the first physical information is stillbeing sent. The minimum time for sendingphysical information continuously and mostquickly is 20ms.

IV. Precautions

T3105 is related to the timer NY1. If T3105 issmall, configure NY1 to a greater value. If ahandover trial fails and the T3105 of the targetcell expires for Ny times before the original cellreceives the HANDOVER FAILURE message,the target BTS sends the CONNECTIONFAILURE INDICATION message to the targetBSC.The counter of target BSC is renewed though MSmight return to the original channel. To avoidthis, the T3105 must meet the following foulard:Ny * T3105 > T3124 + deltaWherein, delta is the time between expiration ofT3124 and receiving HANDOVER FAILUREmessage by original BSC.

4.10.4 T3107

I. Definition

T3107 is a BSC timer, restricting the time forexecuting TCH assignment instruction. It catersfor TCH assignment of intracell handover andchannel assignment of calling.

II. Format

T3107 ranges form 0s to 255s. Therecommended values are as follows:

10s when channel resources are enough. 5s when channel resources are limited.

III. Configuration and Influence

T3107 starts after the BSC sends the ASS_CMDmessage to BTS. It stops after the BSC receivesthe ASS_CMP or ASS_FAIL message sent byBTS. If T3107 expires, the system judges that theMS disconnects to the network, so the occupiedresource is released to other MSs. According tothe measured statistics result of network, thechannel assignment is complete within 2s. If theBSC does not receive ASS_CMP message after2s, the assignment command fails.If the radio link is bad and some informationmust be resent, the process might be prolonged to5s. To avoid premature disconnection, configureT3107 to 10s. In this way, the MS can reuse theoriginal channel when handover or assignmentfails. Therefore the call drop due to intracellhandover decreases or the system service qualityof re-assignment is improved (if the systemsupports re-assignment function). However, thechannel resource might be wasted for severalseconds. When the network capacity is limited,you must save the resource as possible.

4.10.5 T3109

I. Definition

The BSC restricts the releasing resource ofSACCH by T3109.

II. Format

T3109 ranges from 3s to 34s. The recommendedT3109 is as follows:T3109 = a + RdioLinktimeOut x 0.480s, a = 1s or2s.

III. Configuration and Influence

T3109 measures the time for channel releasingindicator after sending MS clearing instructions.It starts after the BSC sends DEACT_SACCHmessage to BTS. It stops after the BSC receivesthe REL_INC message sent by BTS. WhenT3109 expires, the BSC sends the CLEARREQUEST message to MSC.

IV. Precautions

The sum of T3111 and T3109 must be greaterthan RadioLinkTimeOut. If T3109 is over small,the corresponding radio resources are re-allocated before RadioLinkTimeOut is due (radiolink is not released).

4.10.6 T3111

I. Definition

T3111 is a connection release delay timer, usedin deactivation of delayed channel afterdisconnection of major signaling link. T3111aims to spare some time for repeateddisconnections. When BSC receives theREL_IND message sent by BTS, T3111 starts.For time protection, T3111 stops until expirationand the BSC sends the RF_CHAN_REL messageto BTS.

II. Format

T3111 ranges from 0s to 5s.The recommended value is 2s.

III. Configuration and Influence

After the disconnection of major signaling link,T3111 delays the release of channels. It allowsthe base station to retransmit the instruction forreleasing radio channels to MS within delayedtime. After the base station sends a releaserequest massage, the radio resources remain forT3111 time.If the system capacity is small, configure T3111as short as possible. The minimum value ofT3111 is 2s, over five multiples of the time forresending MS the instruction for releasing radiochannel resources. A greater T3111 might be ofno help, but affects congestion of SDCCH andTCH easily.

4.10.7 Parameter T3212

I. Definition

In a GSM network, the causes to locationupdating are as follows:

The MS attach. The MS detects that its location areachanges. The network forces MS to update locationperiodically.The network controls how frequent the MSupdates location, and the period for locationupdating is determined by the parameterT3212.

II. Format

T3212 ranges from 0 to 255, with unit of 6minutes (1/10 hour). If T3212 = 1, it means thatT3212 is 6 minutes. If T3212 = 255, it means thatT3212 is 25 hours and 30 minutes. If T3212 = 0,it means that MS is not required for periodicallocation updating in the cell. The recommendedT3212 is 240.

III. Configuration and Influence

As an important means, the periodical locationupdating enables network to connect to MSsclosely. Therefore, the short the period is, theoverall service performance of the network is.Anyhow frequent periodical location updatingbrings two negative aspects:

The signaling flow of the networkincreases sharply and the utilization of radioresource declines. When the period is overlong, the processing capability of networkelements (NE, including MSC, BSC, andBTS) is directly affected. The MS must transmit signals with greaterpower, so the average standby time isshortened sharply.

Therefore, configure T3212 according toresource utilization in various aspects of network.T3212 is configured by equipment roomoperators. Its value depends on the flow andprocessing capability of each NE. ConfigureT3212 as follows:

Configure T3212 to a greater value (suchas 16 hours, 20 hours, or even 25 hours) inareas with heavy traffic and signaling flow. Configure T3212 to a smaller value (suchas 3 hours or 6 hours) in areas with low trafficand signaling flow. Configure T3212 to 0 in areas with trafficoverrunning the system capacity.

To configure T3212 properly, you mustpermanently measure the processing capabilityand flow of each UE in the running network,such as:

The processing capability of MSC andBSC A interface, Abis interface, and Uminterface The capability of HLR and VLR

If any of the previously listed NEs is overloaded,you can consider increasing T3212.

IV. Precautions

T3212 cannot be over small. Otherwise, thesignaling flow at each interface increases sharplyand the MS (especially handset) consumesincreasing power. If the T3212 is smaller than 30minutes (excluding 0), the network will befiercely impacted.Configuring T3212 of different cells in the samelocation area to the same value is recommended.In addition, the T3212 must be consistent withrelated parameters of switching side (smallerthan the implicit detach timer at switching side).

If the T3212 of different cells in the samelocation area is the same, in the cell reselection,the MS continues to time according the T3212 ofthe original cell. If the T3212 of the original andtarget cell in the same location area is different,the MS uses the T3212 of the original cellmodulo that of the serving cell.According to the actual tests of MS in thenetwork, if the T3212 in the same location area isdifferent, after the MS performs moduloalgorithm based on behaviors of some users, theMS might power on normally. However, the MSfails in originating location updating, so thenetwork identifies it as implicit detach. Now theMS powers on normally, but a user has poweredoff prompt appears when it is called.

4.10.8 T3122

I. Definition

T3122 defines the period that the MS must waitfor before the second trial calling if the first trialcalling fails. It aims to avoid congestion ofSDCCH due to repeated trial calling by MS andto relieve system load.

II. Format

T3122 ranges from 0s to 255s. Therecommended value is 10s.

III. Configuration and Influence

The value of T3122 is included in the immediateassignment reject message. After the MS receivesthe immediate assignment reject message (no

channels for signaling, A interface failure,overload of central processing unit, namely,CPU), it can send new trial calling request afterT3122. T3122 aims to relieve radio signaling andvoice channel resources.T3122 also help avoid systematic overload.When the CPU is overloaded, the systemmultiplies T3122 by a factor (determined byprocessorLoadSupconf) to increase T3122through overload control. In peak load time, youcan manage network access by increasing T3122.Namely, you can increase the interval betweentwo continuous trial callings to relieve networkload.

4.10.9 T3124

I. Definition

T3124 is used in occupation process inasynchronous handover. It is the time for MS toreceive the physical information send by networkside.

II. Format

Configure it to 675ms when the channel type ofassigned channel for HANDOVER COMMANDmessage is SDCCH (+ SACCH). Configure it to320ms in other situations.

III. Configuration and Influence

When the MS sends the HANDOVER ACCESSmessage on the primary DCCH, T3124 starts.When the MS receives a PHYSICALINFORMATION message, the MS stops T3124,

stops sending access burst, activates the PCH insending and receiving mode, and connects to thechannel if necessary.If the assigned channel is a SDCCH (+ SACCH),you must enable MS to receive a correctPHYSICAL INFORMATION message sent bynetwork side in any block. If T3124 expires (onlyin asynchronization) or the low layer link fails inthe new channel before sending theHANDOVER COMPLETE message, the MSproceeds as follows:

1) Deactivate the new channel2) Restart the original channel3) Reconnect to TCH4) Trigger to setup primary signaling link

Then the MS sends the HANDOVER FAILUREmessage on the primary signaling link and returnnormal operation before trial handover. Theparameters for returning the original channel arethose before response to the HANDOVERCOMMAND message (such as in encryptionmode).

4.10.10 T11

I. Definition

T11 is an assignment request queue timer.

II. Format

T11 is determined by equipment room operators.It indicates the maximum queuing delay forassignment request.

III. Configuration and Influence

When the BSC is sending the ASSIGNMENTREQUEST message, no TCHs are available. TheASSIGNMENT REQUEST message must be putto a queue and the BSC sends the QUEUINGINDICATION message to MSC. Meanwhile,T11 starts timing.When the BSC sends the ASSIGNMENTCOMPLETE message (TCH is successfullyassigned) or the ASSIGNMENT FAILUREmessage (TCH is not assigned) to MSC, T11stops timing.If T11 expires, the correspondingASSIGNMENT REQUEST message is removedfrom queue and the BSC sends a CLEARREQUEST message with the cause of no radioresource available to MSC to clear calling.Assignment queuing helps reduce servicerejection times due to congestion, so enabling itis recommended in a network. Anyhow, T11cannot be over great and it must be configuredaccording to customer habits.

4.10.11 T200

I. Definition

T200 is important (both the MS and base stationhave T200) at Um interface in data link layerLAPDm. LAPDm has different channels, such asSDCCH, FACCH, and SACCH, and thetransmission rate of different channel is different,so T 200 must be configured with differentvalues. The type of the channels corresponding toT200 is the value of the T200.

II. Format

Different channels corresponds different valuesof T200. According to the protocol, when SAPI =0 and SAPI = 3, the T200 of corresponding datalink is dependently implemented, depending ondelay of synchronous processing mechanism andprocess in layer 1 and layer 2.

Table 7-1 Value range and default of each type ofT200

T200Minimu

mMaxim

umDefault

T200_SDCCH_SAPI0 50 100 60; /* = 60 * 5 ms */

T200_FACCH_Full_Rate 40 100 50; /* = 50 * 5 ms */

T200_FACCH_Half_Rate 40 100 50; /* = 50 * 5 ms */

T200_SACCH_TCHSAPI0

120 200 150; /* = 150 * 10 ms */

T200_SACCH_TCHSAPI3

120 200 150; /* = 150 * 10 ms */

T200_SACCH_SDCCH 50 100 60; /* = 60 * 10 ms */

T200_SDCCH_SAPI3 50 100 60; /* = 60 * 5 ms */

III. Configuration and Influence

T200 avoids deadlock in sending data in data linklayer. The data link layer changes the physicallink in which error occurs easily to data link withno errors. At the two ends of the data linkcommunication system, a confirm-to-resendmechanism is used. Namely, receiving a messageby the receiver must be confirmed by the sender.If it is unknown that the message is lost, both twoends wait for messages, so the system confronts a

deadlock. Therefore, T200 is used by the sender.When T200 expires, the sender judges that thereceiver fails in receiving the message, so itresends the message.When the sender needs to confirm whether thereceiver has received the message, T200 starts.When the sender receives the response from thereceiver, T200 stops. When T200 expires, theresending mechanism starts. If the senderreceives no response from the receiver aftermultiple resendings, it sends ERRORINDICATION (T200 expiration) to layer 3.

IV. Precautions

T200 must be properly configured to ensure apredictable behavior at Um interface. The rulesfor configuring T200 include:

The potentially-existing lost frames inradio link must be detected as possible. Necessary retransmission of frames muststart at the earliest possible moment. If the response is delayed due to UEfailure, the T200 cannot expire beforereceiving and processing the next frame fromthe opposite end. If T200 expires and no other frames aresent by preference, the related frames must beresent in the message block. T 200 starts immediately after next PH-READY-TO-SEND.

4.10.12 N200

I. Definition

N200 is the resending times after expiration ofT200.

II. Format

To configure N200, follow rules below:1) When SAPI = 0 or 3, N200 depends onthe state and the channel used.When multiframe operation is set up, itensures a common time value for layer 2 linkfailure in all channels. For layer 2 linkestablishment and release, configure N200 to5.2) In timer recovery state, configure N200as below: 5 (SACCH) 23 (SDCCH) 34 (FACCH of full rate) 29 (FACCH of half rate)3) When SAPI is unequal to 0 or 3,configure N200 to 5, as shown in Table 7-2Situations of SAPI unequal to 0 or 3.

Table 7-2 Situations of SAPI unequal to 0 or 3

SAPI ChannelValid

responsedelay

Minimumresending

delay

Maximumresending delay

Tresp Trmin Trmax Note 30 SDCCH MS: 11 51 51

BSS: 32

0FACCH/Fullrate

9 26 39

0FACCH/Halfrate

10 34 44

3 SDCCH MS: 11 51 51 Note 1BSS: 32

3SACCH(withTCH)

25/129 Note2

312 416 Note 2

The TDMA frame is the measurement unit of values in this table,equal to 120/26ms (approximately 4.615ms)

Note 1: It caters for the process without SAPI 0 transmission.Otherwise, it does not have a upper limit due to the priority ofSAPI 0 transmission.Note 2: You can configure it to a greater value only when PCHis unavailable due to SAPI frame transmission if SAPI = 3.Note 3: It caters only for sending monitoring frames that areavailable and without F equal to 1.

III. Configuration and Influence

If the BSC fails in receiving lay 2 responsemessage after multiple resending, it sends theERROR INDICATION message (T200 expires)to layer 3. The BSC takes statistics of ERRORINDICATION message by corresponding trafficmeasurement counter. When T200 or N200 isconfigured to an over small value, call dropoccurs probably due to ERROR INDICATION.

5 GSM Radio Network Optimization

Radio network optimization aims to improve networkperformance and maximize the benefit of the existingnetwork resources through parameter collection, dataanalysis, parameter adjustment, and necessarytechnical means.

From the perspective of carriers, they hope toconfigure the system rationally, utilize network

resources to the maximum, enhance networkeconomic benefit, and reduce operation costs throughnetwork optimization. From the perspective of users,they hope to get satisfactory telecommunicationservices in terms network stability, speech quality, andso on. Therefore, the core task of radio networkplanning and optimization is to seek a balance amongcoverage, capacity, and quality based on rationalinvestment and the limited frequency resources, thusachieving the best rate of investment return.

5.1 Network Optimization ProcedureHereunder details the procedure:

Network information acquisitionYou are required to confirm the actual engineeringparameters and network parameters, survey thelocal radio environment and hot-traffic spots, andunderstand customer requirement.

Data collectionYou are required to collect OMCR traffic statisticsdata and alarm data; drive test data, and theobjective reflection of MS.

Data analysisYou are required to analyze network performance,network parameters, and OMCR traffic statisticsusing network optimization tools.

Network tuningYou are required to tune engineering parametersand network functional parameters.

Network optimization reportA network optimization report must includeoptimization measures, fulfilled networkperformance indexes, and suggestions for networkdevelopment.

5.2 Network Optimization Tools5.2.1 Test MS

Test MS is a daily must for engineers to performnetwork test. The test MS can display the servicecell of a mobile telecommunication network andthe six neighbor cells. It can also be used to testnetwork parameters. The test MS can beconnected to a computer, so it can collect andanalyze data with the help of drive test software.The functions of a SAGEM test MS are listedbelow:

View the IMSI of the SIM card Scan BCCHThe SAGEM test MS can scan the BCCH,and it will provides the Rxlev and BSIC ofthe scanned BCCH in each cell. View network parametersThe SAGEM test MS can display networkparameters of the service cell of and the sixneighbor cells when it is idle or inconversation state. Forced cell selection

The test MS can be forced to reselect thedesignated cell for dialing test and handovertest. Forced handoverThe test MS can be forced to hand over to thedesignated cell to analyze whether thehandover is normal during conversation. Frequency selection GPRS function test (The MSs of SAGEMOT96 support this function)

5.2.2 Drive Test Software

ANT, TEMS, and SAFCO are the GSM drivetest software in common use. Generally, the drivetest software consists of two parts: foregrounddata collection software and background dataanalysis software.

I. Foreground Data Collection Software

The foreground data collection software ismainly responsible for collecting the uplink anddownlink data at the Um interface of the GSMmobile network. If the BTSs and the landformmap of the test areas are imported, the track ofthe drive test, the radio parameters of each testpoint, and the main information of the servicecell, neighbor cell, and the Um interface can beobserved.The functions of foreground data collectionsoftware are listed below:

Test functions

Test functions include call test, scan test, callinterference test, double-networksynchronization test, dialing record test, dual-band comparison test, and MS selection test. Topical testTopical test is responsible for forcedhandover and frequency locking (call) test,BCCH frequency locking (standby) test,forced location update test, and band lockingtest. Interference testInterference aims to locate the BCCH same-frequency and neighbor frequencyinterference and TCH neighbor frequencyinterference within the same network in time. Parameter collectionThe field strength of the service cell and theneighbor cells, bit error ratio, frame errorrate, and various radio parameters of the cellneed to be collected. Geographic navigationThe foreground data collection software canbe used to display the geographic navigationthrough combing the digital map and BTSresources. SQI (Speech Quality Index) test

SQI test aims to evaluate the speech qualityexperienced by mobile users.

Traffic statisticsThe foreground data collection software canbe used for CQT traffic statistics. It enables

the call setup, call duration, and call releaseto be recorded respectively. In addition, it isalso responsible for recording call drop rateand congestion rate. Data recordThe foreground data collection software canrecord and store the test data of the test MSsynchronously, and can record and store thescanned data of the 900MHz and 1800MHzsynchronously. Scanning testScanning test aims to test and record the fieldstrength of the channels of the GSM 900MHznetwork and GSM1800MHz network. Competitive testThe comparison between real-time fieldstrength and speech quality is available. Inaddition, real-time check of cell parameters isallowed.

II. Bachground Data Analysis Software

The background data analysis software cangeographically present the radio network testdata and reflect the distribution of networkparameters on the electronic map visually. It canlocate the problem cell by fully considering thedrive test data, network resource data, digitalfrequency sweep receiver data, and GSMsignaling characteristics, thus guiding engineersto evaluate and optimize the network rationallyand effectively.

The functions of background data analysissoftware are listed below:

Radio coverage evaluation and analysis Interference analysis Neighbor cell analysis Handover analysis Signaling analysis Speech record and analysis

5.2.3 Signaling Analyzer

MA-10 and K1205 are the GSM signalinganalyzer in common use. This section introducesthe functions of the signaling analyzer.MA-10 signaling analyzer can test the Abisinterface signaling messages, the A-G interfacesignaling messages, and the SS#7 interfacesignaling messages of ISUP, TUP, and INAPwhen it is on line. In addition, it can performPCM BERT (bit error rate test) and GSM BERT.Furthermore, it can enable the signaling messagefiles saved during on-line monitoring when it isoff-line.The MA-10 signaling analyzer has five sub-applications. They are listed below:

MA-10 control (It is used to test Abisinterface signaling messages, A-G interfacesignaling messages, and bit errors when it ison line.) MONITOR ABIS offline (It is used toanalyze the Abis interface signaling messageswhen it is off line.)

MONITOR MSC offline (It is used toanalyze the A-G interface signaling messageswhen it is off line.) GSM-BERT offline (It is used to analyzeGSM BERT when it is off line.) PCM-BERT offline (It is used to analyzePCH BERT when it is off line.)

With the help of MA-10 signaling analyzer,network optimization engineers can collect andanalyze Abis interface data and A-interface data,view the whole signaling procedure, and obtainthe measurement report, and then compare theinformation with the downlink signals obtainedfrom drive test. These means can help networkoptimization engineers have an overallunderstanding of the operation of the network. Inthis case, the causes and places for the problems,such as call drop, handover failure, andcongestion can be located.

5.2.4 Network Optimization Software

Good network optimization software can work asa platform for radio projects and maintenancepersonnel. NASTAR, Huawei network planningand optimization tool, can deeply analyze GSMnetwork by integrating OMCR traffic statistics,configuration data, alarm data, engineering data,and so on. In addition, it also provides theinterfaces for network planning, performanceanalysis, alarm analysis, and geographyconditions.The functions of NASTAR are listed below:

Import static traffic statistics data (importthe traffic statistics files of the maintenanceconsole into database) Import static traffic statistics dataaccording to time Import dynamic Huawei traffic statistics Import BSC data configuration Display the geographic view of BTSs Cancel, add, and move the cells based onthe geographic view of BTSs Record spectrum utilization and analyzeidle channel number Check same-frequency BSIC Analyze neighbor cells Measure distance and azimuth angle Display navigation window and legend Query performance analysis tasks incommon use Make performance analysis task Query network optimization tasks incommon use Make network optimization tasks Perform tasks and alarms automatically Import and export tasks Display task result and site view Query traffic statistics inversely from thecell set selected by site view Display the traffic tasks in direct views,including linear figure, column, pie (supportthe display of double y-axis and thesimultaneous display of multiple indexes;support 2D and 3D.)

Export electronic table (Excel) for queryresults.

5.3 Network Performance EvaluationBefore putting network optimization intopractice, you should have an overallunderstanding of the network performance. Thetraffic statistics data, DT (drive test) data, andCQT (call quality test) data are necessary fornetwork performance acquisition.

5.3.1 DT

DT is used to evaluate the connectivity,coverage, call drop, and voice quality for themain roads and transportation backbones inurban areas. The indexes for the roads includeconnected ratio, call drop rate, coverage rate,voice quality, and so on. The indexes fortransportation backbones include call drop rateper kilometer, coverage rate, voice quality,connected rate, and so on.

Connected ratioConnected ratio = total connected

times/attempted calls *100% Call drop rate

Call drop rate = call drop times/totalconnected times *100%

Coverage rateCoverage rate = (≥ -94dBm test road

kilometers)/total test road kilometers*100%

Voice qualityAccording to bit error rate, voice quality canbe divided into 8 classes, from 0 to 7. Eachclass matches its bit error rate range.The calculation of voice quality depends onactual conditions. Generally, the followingmethod is in common use:Voice quality = [Rxqual (class 0)% + Rxqual(class 1)% + Rxqual (class 2)%*1 + Rxqual(class 3)%* + Rxqual (class 4)%*0.8 +Rxqual (class 5)%* + Rxqual (class 6)%*0.5+ Rxqual (class 7)%*0.2 Call drop ratio per kilometer

Call drop ratio per kilometer = (≥ -94dBmtest road kilometers)/total call drop times

5.3.2 CQT

CQT is applied to the important spots in urbanareas. It enables you to experience the networkquality from the perspective of users. Thissection introduces the indexes used to evaluatingthe CQT.

Coverage rateCoverage rate = (≥ -94dBm test

points)/total calling test points)*100% Connected ratio

Connected ratio = total connectedtimes/attempted calls *100%

Call drop rateCall drop rate = call drop times/total

connected times *100%

Voice discontinuity/background noise rateVoice discontinuity/background noise rate =(total voice discontinuity occurrence times +total background occurrence times)/totalconnected times*100% One-way audio/echo/cross-talking rate =(total one-way audio times + total echooccurrence times + total cross-talkingoccurrence times)/total connectedtimes*100%

CQT enables you to use the MOS (mean opinionscore) to evaluate the voice quality from theperspective of people's objective feeling. TheMOS can be divided into five classes, from 1 to5. For the evaluation standard, see Table 5-2 ofChapter 5 GSM Radio Network Planning of GSMRadio Network Planning and Optimization.

5.3.3 Network Operation Indexes

Though DT and CQT can detail networkproblems, they are restricted from test routes andtime. Therefore, DT and CQT cannot test theoverall network. To give an overall evaluationtowards the network, you should collect as morenetwork operation indexes as possible.The indexes evaluating network operation qualityare listed below:

Service access capacity indexesService access capacity indexes include tollnetwork connected ratio, short messageconnected ratio, (GPRS) PDP activationratio, and IP telephone connected ratio.

Service hold capacity indexesService hold capacity indexes include calldrop rate, worst cell ratio, traffic call dropratio, handover success rate, and shortmessage gateway transit success rate.

With the expansion of network scale, the networkstructure becomes ever more complicated. In thiscase, networks with high performance but lowcost are encouraged.The indexes on network utilization can be used toevaluate whether the cost to run a network is low.These indexes include toll circuit utilization rate,traffic channel availability, busiest and idlest cellratio, and so on.Hereunder introduces the methods to calculatethe indexes used to evaluate network operation atthe radio side.

Radio connected ratioRadio connected ratio = (1 – SDCCHcongestion rate)*(1 – TCH congestion rate(all busy))*100% Call drop rate

Call drop rate = TCH call drop/successfulTCH seizures (all busy)

Worst cell ratioA cell with TCH congestion rate higher than5% at busy hours or a cell with TCH calldrop rate higher than 3% is defined as a worstcell. The number of the worst cells varieswith network scales. Traffic call drop ratio

Traffic call drop ratio = total TCH trafficvolume *60%/total TCH call drop times (allbusy) Handover success rate

Handover success rate = successfulhandovers/attempted handovers*100%

Traffic channel availabilityTraffic channel availability = the availabletraffic channels at busy hour/configuredtraffic channels*100%

5.4 Traffic Statistics Index AnalysisAt the network optimization stage, the trafficstatistics indexes are the basis for networkperformance optimization. For networkoptimization, the KPIs, such as congestion rate,call drop rate, and handover success rate, are incommon use. These indexes are the externalrepresentation of network quality. The radiocoverage quality, channel capacity, and cellparameters are the internal factor to affect thenetwork quality. The traffic statistics analysisaims to look into these internal factors throughexternal factors. Since the mobile network is acomplex system, you should consider related DTinformation, signaling messages, and alarminformation for the overall analysis.

5.4.1 General Analysis Method

Traffic statistics analysis is performed from BSCoverall performance to cell performance, fromprimary indexes to secondary indexes.

First you should have a rough understanding ofthe network performance through BSCperformance analysis. Here the indexes such asTHC traffic intensity, TCH call drop rate, TCHcongestion rate, and inter-cell handover successrate should be considered. Attention that inaddition to check the percentages of the indexes,you should also check the absolute numbers ofthe indexes, because the percentages maysometimes hide some cell problems.After having understood the indexes about theoverall network performance, you should analyzethe indexes for each cell if finding abnormalindexes. First you should judge if the abnormalindex is a common phenomenon or it is really anabnormal one. If it is a common phenomenon,you should begin the analysis from theperspective of coverage, capacity, frequencyplanning, and cell parameters. If it is really anabnormal case, you should register thecorresponding traffic sub-items and analyze themin detail. In addition, you should also make anoverall judgment through collecting theinformation about alarm, engineers' operation,and other external causes. If the traffic statisticsanalysis cannot contribute a correct judgment,you should employ DT equipment and signalinganalyzer for help.

5.4.2 High Call Drop Rate Analysis

If the uplink and downlink quality deteriorates toa level that cannot hold normal conversation, theconversation will be disconnected. This is

defined as call drop. Since the user mobility andradio propagation is uncertain, call drop alwaysexists in a mobile network. However,optimization measures can be adopted to reducethe call drop rate.When the call drop rate of the BSC overallperformance is found abnormal, you can checkTCH performance to judge whether the call dropis just a common phenomenon or it is anindividual phenomenon. After that, you canjudge whether the high call drop rate occurs inseveral cells or in all the BTSs. If the call drop isa common phenomenon, you should make anoverall check towards the coverage planning, cellparameter planning, and frequency planning toanalyze whether the link budget meet therequirements, whether the configuration of thepath failure counter is rational, and whether thenetwork interference is too great. In addition, youshould also check the BSC hardware, and thenperform drive test to check the network coverage.If it the abnormality is caused by the severe calldrop in individual cells, you should confirmwhether it is equipment failure that caused thecall drop. Generally, alarm messages are alwayscome together with equipment failure, so you cantake equipment failure as a reference.After the equipment failure is excluded, you cananalyze the call drop rate from the perspective ofinterference, coverage, and handover.

1) Interference is divided into uplinkinterference and downlink interference. You

can analyze the uplink interference accordingto the number of interference bands intowhich the idle TCHs drop. It is normal that theidle TCHs drop into interference band 1 andinterference band 2. For the network withaggressive frequency reuse, it is acceptablethat the idle TCHs drop into interference band3. Here the frequency hopping, PBGThandover and coverage control must beconsidered. If the idle TCHs drop intointerference band 4 or above, you shouldcarefully check the interference. Generally,the interference within the network increaseswith the traffic volume. The increase of theRxqual class can be seen through the Rxqualmeasurement task and Rxlev measurementtask. The poor handover ratio can be seenarising through inter-cell handoverperformance measurement. In addition, thehandover re-establishment failures will resultin more handover failures.2) If the coverage is inadequate or it isunbalance on the uplink and downlink, the calldrop will also be resulted. You can judge ifthe Rxlev is adequate through the mean Rxlevof the power control measurement task andthe power class. If the Rxlev is still low whenthe transmitter power reaches the maximum,there are areas with poor coverage.Meanwhile, you can take the mean Rxqualand Rxlev during call drop as a reference. Thedistribution of TA (timing advance) valuescan help you estimate the radius of subscriber

distribution. Through checking the receivedchannel strength of the neighbor cells, you cananalyze the cell coverage. Generally, drive testis needed for a detailed analysis.If the uplink coverage and downlink coverageare unbalance, RF component failure or cableconnection problem will occur. The pathunbalance can be seen from the path balancemeasurement task, power measurement task,and call drop measurement task. At this time,the alarm information and user complaintalso deserve your attention.

3) Handover failure will prevent the MSfrom moving to the best cell. In this case, calldrop may be resulted. In addition, cross-cellhandover and target cell congestion may causecall drop. To solve this problem, you can addneighbor cell relationship and balance thetraffic within the cells.The high SDCCH call drop rate analysis issimilar to high TCH call drop rate analysis.Acting as the point-to-point signalingchannel, the SDCCH is more sensitive tothe interference than TCH. In this case, thecommon method to reduce the call drop rateis to adjust the access threshold and reduceinterference.

5.4.3 High TCH Congestion Rate Analysis

This section discusses TCH congestion,including the congestion caused by TCH seizure

all busy and the congestion caused by TCHseizure failure.When the congestion rate of the BSC overallperformance is found abnormal, you can find outthe cells with high congestion rate throughchecking the TCH performance statistics. In thiscase, you can discover the problems throughanalyzing each functional sub-item of the TCHperformance statistics of this cell. In addition,you should check whether there is transmissionproblem, clock problem, or hardware problemthrough considering the alarm information.It is a must to analyze the load according to theTCH traffic intensity and the configured TCHcapacity.

1) Check if the THC congestion rate iscaused by TCH seizure all busy throughanalyzing the TCH performance measurementof the cell. If the congestion is caused byheavy traffic, you should predict the realtraffic of the cell and check if other cells canshare the traffic. If it is beyond theoptimization capability to enable other cells toshare the traffic, you should considerexpanding the capacity of the network. Theadjustment measures for traffic balance maynot be consistent with the principle ofminimum radio path loss, so they are appliedto emergent causes only. In most cases, youcan balance the traffic through adjustingcoverage scope, adjusting access threshold,adjusting CRO and handover threshold, orenabling load handover. If the congestion is

not caused by TCH seizure all busy, go onwith the check.2) Check if the TRXs of the congestion cellwork normally. The damage or performancedecline of the uplink channels may prevent theMS from accessing other cells. In this case,many cells will be seized, which will causecongestion. The incoming cell handoverperformance measurement will show thatmany handovers towards this cell are failures.In this case, you should query the statue ofeach TRX within each cell through queryingthe Rxlev performance measurement task orRxqual performance measurement task. Inaddition, you should find out which TRX isrelated to the abnormality through queryingthe uplink and downlink measurement reportsof the same TRX.3) Check if the congestion rate is related tointerference, namely, check if any abnormalityis present from the interference band 1 tointerference band 5 in the traffic statistics. Ifthe interference is present in a cell, the calldrop rate of the cell will be high, and theSDCCH congestion rate will increaseaccordingly. Moreover, the RACH in therandom access performance measurement maybe congested, and the immediate assignmentsuccess rate will decrease.4) Under some conditions, the congestion ofsome cells is a result of large coverage. In thiscase, you should analyze the relationshipbetween TA value and Rxlev through

querying the power control mean level, themean level during call drop, and TA. Inaddition, you should also use drive test todefine the coverage area of the cell. Throughquerying the TCH availability of the neighborcell, you can confirm if the congestion iscaused by neighbor cell failures. Throughquerying path balance performancemeasurement, you can judge if the reason forthe TCH seizure failure is that the downlinkpower is greater than the uplink power.5) Frequent handovers can also cause TCHcongestion. Through querying the ratio of thehandovers to the call seizure successes, youcan check if the ratio is rational. Throughquerying the incoming and outgoing ratio, youcan check if the congestion is caused byirrational handover.

5.4.4 High SDCCH Congestion Rate Analysis

The SDCCH congestion rate is mainly caused byheavy traffic. First you should define if thecongestion is a common phenomenon or if it isjust an individual phenomenon. If it is a commonphenomenon, you should analyze if the locationupdate timer is irrationally set, and then calculatethe SDCCH capacity to see if it meets systemrequirement. If it is just an individualphenomenon, you should analyze it from theperspective of equipment, location area, andinterference.

1) From the perspective of equipment, youshould first check the TRX sound ratio in the

BSC overall performance measurement andthe SDCCH availability in the SDCCHperformance measurement, and then check theTCH activation NACK/TIMEOUT in theTCH performance measurement. After that,you can define if the congestion is caused byboard problem.2) Check the messages for SDCCH bearerlocation update. Irrational location areaplanning will cause frequent location update,which will result in SDCCH congestion. Youare required to analyze of the edge of thelocation is set at the areas with a great numberof subscribers by checking the location areaplanning and actual drive test. In addition, youare also required to check if the locationupdate messages accounts a too largerpercentage of the SDCCH seizure requests atthe edge. The method is to query the ratio ofthe successful SDCCH seizures (locationupdate) to the total SDCCH seizure successesin the SDCCH performance measurement.3) Interference also causes SDCCHcongestion. Especially for the networks inwhich the distance between BTSs is small andthe BCCH frequency is aggressive, the systemmay receive more interference random accesssignals. The network will allocate a SDCCHfor each random access, which causesSDCCH congestion. In this case, theimmediate assignment success rate willdecrease, the paging success rate will

decrease, and the RACH in the random accessperformance measurement may be overloaded.

5.4.5 Low Handover Success Rate Analysis

The analysis for handover success rate is quitecomplicated, because it involves capacity,coverage, clock, signaling, equipment, and evenMS.

1) If the handover success rate of all cells islow, you should check the problem from theperspective of handover parameters, A-interface circuit, and BSC clock.2) Filter the cells with poor handover. If anetwork is run by the equipments of differentcarriers, you should check if it interoperabilityproblem by comparing the inter-BSChandover success rata with the intra-BSChandover success rate in the handoverperformance measurement. Generally, theinter-BSC handover success rate is a littlelower than the intra-BSC handover successrate. In addition, you need to monitor thesignaling messages and data configurationbetween BSCs and analyze the radio linkbudget and clock of each carrier.3) Check if any problem is present at theUm interface through comparing the handoversuccess rate and radio handover success rate.The radio handover success rate is equal to orgreater than the handover success rate. If thehandover success rate is far smaller than theradio handover success rate, you shouldanalyze the ground link and capacity. If the

difference between the radio handover successrate and the handover success rate, you needto consider the interference.4) Analyze if it is incoming handover failureor it is outgoing handover failure throughquerying the incoming cell handover successrate and outgoing handover success rate in thehandover performance measurement. Afterthat, analyze the outgoing cell handoverperformance measurement and incoming cellhandover performance measurement of theproblem cell so as to find out the incominghandover failure cells from the outgoing cellperformance measurement. Confirm if thepoor handover is caused by target cellcongestion through analyzing the "incomingcell handover failures", "TCH trafficintensity", and "TCH congestion rate (allbusy)" of all the incoming handover failurecells.5) Check if any equipment fails throughquerying the TRX sound ratio, TCHavailability, and TCH activationNACK/TIMEOUT of the target cell. Analyzeif the TRX performance decreases throughquerying the Rxlev performance measurementof the target cell.6) Check if any ground link equipment failsthrough querying the A-interface failures andthe ground link breaks during TCH seizure.When the microwave is used for thetransmission or during inter-BSC handover,the clock deviation is another cause for poor

handover. And this can be proved by theintra-BSC handover failures. For the cellswhere the clock synchronization isunavailable, the BSIC cannot be decoded, sothe handover can never occur. In this case,you need to check if the clock is normal andanalyze the call drop rate.If these two causes are excluded, you need tomake adjustment from the perspective ofcoverage and interference.To reduce call drop rate and enhancehandover success rate, you can leave amargin for the Rxlev and Rxqual duringhandover. If the Rxlev of a cell is lower than-90dBm during handover, you should checkthe mean Rxlev and TA value of TCH calldrop in the call drop performancemeasurement and analyze drive test to see ifthe coverage distance of the cell is too longand if the signal is not strong enough.For the networks in which better cellalgorithms are enabled, you should check the"attempted handovers (better cell)". It isbetter that the percentage it accounts 60% ofthe handover causes.The interference will also affect the handoversuccess rate. When the interference ispresent, the voice quality will decrease andthe call drop rate will increase.Handover problems are rather complicated.To solve the problems arising in actual work,you are supposed to integrate the methods

introduce above, the signaling analyzer,equipment condition, and drive test intoconsideration.

5.5 Network Problem Solutions5.5.1 Coverage Problems

I. Solutuon Procedures

[Problem: the coverage is becoming smallerafter the BTS is enabled]After a BTS has run for a period of time (forexample, half years), the coverage of the BTSmay become smaller or even dead zone mayappear due to various causes. In this case, thesystem performance will be affected. The shrinkof the coverage is not only related to thetechnical indexes (such as the BTS sensitivityand power), but also related to the engineeringquality, geographic factors, and theelectromagnetic environment.The factor concerning the BTS problems are asfollows:

Transmitter output power decrease Receive sensitivity decrease Antenna azimuth angle change Antenna tilt change Antenna gain change Feeder loss Coupler loss Working band change Propagation environment change

Diversity effect change.You can check the problem according to thefollowing procedure:

1) Check the conditions around the BTSantennaYou are required to check if there are otherantennas (such as micro antenna),decorations, billboard, trees, or glass wallsstanding around the BTS antenna. Thesebarriers may exert a negative effect againstthe antenna reception and transmission, thusaffecting the coverage of the BTS. In thiscase, you can tune the azimuth angel of thecorresponding antenna or change the antennaheight.

2) Check the change of the propagationenvironmentThe change of the propagation environmentof the electromagnetic wave will weaken thesignals received by radio terminals.Especially for mountains, the propagation ofthe electromagnetic waves depends on thereflection of mountains. For example, thechange in the vegetation of the mountain willreduce the coverage of the BTS. In addition,the climate and other natural factors alsohave some effect against the electromagneticwaves. The propagation loss varies withwood intensity, season, and so on. Themaximum loss can reach 30 dB. If newbuildings prevent the propagation of theelectromagnetic waves and weakens the

signals, the areas in the remote cannot becovered, so the subscribers cannot enjoy theservice. Especially the high buildings nearthe BTS have a great effect against thepropagation of electromagnetic waves.

3) Check if there is standing wave alarm andmain diversity reception alarm at the operationand maintenance consoleThis problem can be checked according tothe standing wave alarm messages and thediversity reception alarm messages. If thealarms of this kind occur, you should checkthe corresponding antennas and feeders.

4) Check if the standing wave ratio issmaller than 1.5%The tolerance of the standing wave alarmthreshold of the CDU or EDU is great.Therefore, after checking that the set-toppower is normal, you can further check if thestanding wave is smaller than 1.5%. If thestanding wave ratio is abnormal, you need tocheck if the water has penetrated into theantenna or feeder connector, or if it islightening protector problem.

5) Check if the tower amplifier worknormallyCheck if tower amplifier alarm is present atthe operation and maintenance console.Generally, the problems are the low noiseamplifier was damaged or the water haspenetrated into the amplifier. The amplifieralarm always comes together with the

damage of the low noise amplifier. If thewater has penetrated into the tower amplifier,no alarm will be generated, but the RF loss isgreat. In this case, the receiver sensitivity willdecrease dramatically.

6) Check the engineering parameters(including antenna tilt and azimuth angle)The increase of the antenna tilt or thedeviation of the azimuth angle will reduce thecoverage of the BTS. Therefore, antennasmust be firmly fixed so that they can standstrong wind and storms.

7) Check the set-top output power of thetransceiverFirst you should check if the lines are wellconnected, and then check if the set-toppower is normal. If it not normal, you shouldreplace the problem hardware.

8) Check if the receiver sensitivity is normalCheck if the coverage distance is shortenedby the low receiver sensitivity. In addition,you can monitor the messages at the Abisinterface and find out the relationshipbetween level and bit error rate. After that,you can get the value of the level when thebit error rate is 2%. This means, however,only applies to the situation that when thereceiver sensitivity drops dramatically.

9) Check if the parameters affecting thecoverage are rationally set

10) Check if the high back noise in thecoverage area is caused by interference andpoor electromagnetic environment.

[Coverage problem caused by BTS expansion]If the coverage of the BTS shrinks afterexpansion, in addition to making the previouschecks, you are supposed to check the followingitems.

1) Check if the combiner keeps the samebefore and after expansionThe loss of different combiners varies greatlyTherefore, the combiner configurationdeserves special attention during BTSexpansion. If different combiners are a must,you should fully communicate withcustomers.

2) Check if the antennas are rationallyselectedSuitable antennas must be selected for projectinstallation and network planning so that thebest coverage can be achieved. It must bepointed out that you should use zero-pointfilling antenna or the electrical title antennawhen the antenna height is great. In addition,omni antennas cannot be widely used for thelarge area coverage. In this case, the coverageproblem can be solved by directionalantennas.

3) Check if the installation of the newly-added antennas are qualifiedYou should first check if the design of theantenna height, azimuth angle, and antenna

tilt is qualified. Generally, the importantcoverage areas cannot be bared by tower.Meanwhile, the important coverage areascannot be perpendicular to the diversitydirection of the antenna. In this case, theantenna diversity effect can be excavated tothe maximum. To reduce the coverageshadow caused by the tower, you should payattention to the distance between the antennaand the tower. Moreover, the pole of theomni antenna and the RF part of the antennacannot be overlapped.

4) Check the position of the BCCHtransmitter antennaSince the tower effect is present, the BCCHtransmitter antenna must be installed at a sideof the important coverage area. In this case,the coverage shadow can be avoided. Toprevent the assignment failure caused by theinconsistence of the BCCH coverage andTCH coverage, you can use the concentricchannel allocation algorithm. In addition, theimportant coverage area cannot beperpendicular to the diversity direction of theantenna.

5) Check if the tilts and the azimuth anglesof the directional dual transmitter antennas areconsistent with each otherIf the tilts and azimuth angles of thedirectional dual transmitter antennas areinconsistent, call drop, assignment failure,and handover failure will easily occur. In this

case, the coverage area of the BTS willbecome small. In addition, since the towereffect is present, the BCCH transmitterantenna must be installed at a side of theimportant coverage area. In this case, thecoverage shadow can be avoided. Moreover,the important coverage area cannot beperpendicular to the diversity direction of theantenna.

6) Check the set-top output power of variousTRXs if the scheme for the maximumcoverage is used.When the maximum coverage is pursued, theTRXs are required to be combined in variousways. In this case, the coverage distance ofthe BCCH will be longer than that of theTCH. As a result, the TCH assignment failurewill be caused, so the concentric technologyis needed. The channel assignment failurecaused by low transmit level in the innercircle and the channel congestion in theexternal circle can be avoided if the TAvalues of the inner circle and the externalcircle are correctly set and allocated to theinner circle and external circle according tothe right priority.

[Coverage problems caused by BTS swap orconstruction]

1) Check if the azimuth angle and theantenna height are the same before and afterthe BTS swap

If all the antenna and feeder components arenewly constructed, the old BTS can only beswapped after the new antenna is installed.Therefore, the azimuth angle and the antennaheight may be different from that of the oldantenna. In this case, the coverage area maydecrease. As a result, you should check if theazimuth angle and the antenna height are thesame before and after the bas station swap.

2) Check antenna tilt problems caused bynetwork swapGenerally, the tilt must keep the same. If youneed to control the coverage area due to newBTSs are added to urban areas, you canconsider increasing the tilt.

3) Check if the set-top power of the swapBTS is the same as that of the old BTS.4) Check if the receiver sensitivity of theBTS is normal.5) Check if it is the interference or the poorelectromagnetic environment that makes theback noise of the whole area too high.6) Check if any standing wave alarmmessage or diversity reception alarm messageis generated for antenna and feeder at theoperation and maintenance console.7) Check the parameters that will affectcoverage are rationally set.8) Check if the installation of the antennas isqualified after the BTS has been enabled orswapped.9) Check if the right type of antenna isselected.

10) Check the position of the BCCH TRXtransmitter of the omni dual transmitterantenna.11) Check if the tilts and the azimuth anglesof the two directional antennas keep the sameafter the directional dual transmitter antenna isused.12) Check if the antennas and feeders of thecell are inversely connected.13) Check if the tower amplifier worksnormally.14) Check the set-top power for variousTRXs when the configuration scheme for themaximum coverage is pursued.

II. Problems Affecting Coverage and Solutions

[Antenna water penetration]It is quite accidental that the water penetrates intothe antenna. Water penetration means that thewater enters the RF internal channel. In this case,the voltage standing wave ratio of the antennawill increase; the antenna loss will increase, thecoverage area will decrease; or event the poweramplifier will be disabled.[Antenna passive intermodulation]The passive intermodulation of the antenna andvarious connectors will cause interference. Theexclusive method can be used for the check. Thatis, you can connect the antenna feeders of theneighbor cells where there is no interference tothe test cell. If any problem is found, you shouldchange the antenna.[Improper antenna selection]

Generally, if the antenna height exceeds 50m andif the first zero point under the main antennabeam is not filled, the "shadow under tower" mayoccur. That is, the area under the tower cannot becovered by signals. In this case, you should selectthe antenna with zero point filling function.If three-sector directional antennas are used forvast coverage, the antennas must have a highgain and their half power angle must be greaterthan 90 degrees. If the half power angle is small,the gain of the two neighbor sectors will be low.In this case, the coverage radius is small.If the antenna tilt is great, the all mechanical tiltantenna is not a suitable choice. In this case, youshould select the fixed "electrical tilt +mechanical tilt" antenna or the "continuousadjustable electrical tilt (0 to 10 degrees) +mechanical tilt" antenna.As the frequency reuse becomes moreaggressive, the front-to-back ratio of the antennamay not meet the requirement of a single BTS orseveral BTSs. Therefore, you should select theantennas with greater front-to-back ratio.[Tower effect against Omni antenna radiation]

The tower effect against omni antenna radiation deservesenough attention. It is hard to estimate the damage of theomni antenna directional diagram caused by the tower.The damage varies greatly with the distance between thetower and the antenna.If the antenna is installed on the tower and metal tube,you should pay special attentions to the following items:

The metal tube and the effect radiationpart of the antenna cannot be overlapped. Take measures to avoid installing thewhole antenna on the metal tube. If the antenna is installed on the tower,make sure that the distance between theantenna and the nearest end of the tower isgreater than 6 wavelengths. The omni dual transmitter technology isnot recommended. The antenna must be perpendicular to 1/8of the half power beam width at least.

[Directional antenna installation problem]Two problems may occur for directional antennainstallation:

The antenna is inversely or wronglyconnected. The azimuth angles and the tilts of thetransmitter antenna and the receiver antennaare inconsistent and or the error is great.Engineering causes are the explanations ofthe two problems. Generally, the error scopeof the azimuth angle cannot exceed 5degrees, and that of the tilt cannot exceed 0.5degrees. If the error is too great, the coverageof the transit antenna and that of the receiverantenna will be different. In this case, it ishard to make calls the coverage edges. Problems concerning the diversity distancebetween the transit antenna and the receiverantenna or the isolation between the antennasand tower.

The coverage of the antenna will be affectedif the diversity distance between thetransmitter antenna and the receiver antennaor the isolation between the antennas and thetower is not great enough. For GSM 900MHzsystem, the diversity distance between thetransmitter antenna and the receiver antennais required to be greater than 4m. For GSM1800MHz system, it is required to be greaterthan 2m. The antenna mount must be at least1.5m away from the tower. Meanwhile, theantenna mount must be installed within the45-dregree protection areas of the lighteningprotector. There are shadows in coverage areas.When installing a directional antenna, youshould make sure that there is no shadowwithin the coverage area. Generally, if thereare huge barrier, such as high buildings andmountains, around the BTS, shadows mayappear. If you intend install the BTS on theroof of a high building, you should install itat the edges of the building so as to avoid theshadow. Since the environment around theroof is quite complex, the antenna heightmust be great enough. In this case, however,you should consider the ability of the antennato stand the wind and storm.

[Omni antenna installation problem] The radiator of the omni antenna is barredby antenna pole.

The coverage will be affected if the radiatorof the omni antenna is barred by antennapole. Generally, there is a jacket installed atthe bottom of the omni antenna and the jacketis used to connect the omni antenna and theantenna pole. From the perspective ofinstallation, the top of the jacket must be atthe same level with or higher than the top ofthe pole; otherwise the radiation will beaffected. The problems concerning antennadiversity distance and isolation betweenantenna and tower.If the antenna diversity distance or theisolation between antenna and tower is notgreat enough, the coverage will be poor. Ifthe antenna diversity distance is too small, itwill reduce diversity gain. In this case, thereceiver sensitivity will reduce. Though thetower effect against the omni antennaradiation is unavoidable, you can increase theisolation between the antenna and the towerto reduce the effect.It is suggested that the isolation between theomni antenna and the tower is greater than2m, the horizontal diversity distance of the900MHz omni antenna is greater than 4m,and the horizontal diversity distance of the1800MHz antenna is greater than 2m. The omni antenna is not perpendicular tothe horizontal plane.

If the omni antenna is not perpendicular tothe horizontal plane, the antenna directionaldiagram will be distorted in the coveragearea. In this case, the coverage of the antennawill be affected.It is suggested that installation plane of theantenna mount be perpendicular to thehorizontal plane. If the mount extendsbeyond the tower, make sure that the mountis still in the protection areas of the lighteningprotector. Generally, the areas 45-dregeeunder the lightening protector top are theprotection areas.

[Connection problems of antenna and feeder,combiner and splitter, and CDU]

If various connectors of the antenna andfeeder system are not connected according torequirement, the performance of the antennaand feeder system will be affected. In thiscase, the coverage area of the BTS will alsobe affected. Water penetration occurs at the variousconnectors of the antenna and feeder system.If water has penetrated into the connector andfeeder, the standing wave ratio will increase.In this case, the coverage area will beaffected. Various connectors are not tightened.If the connectors for set-top jumpers, for thecables from TRX boards to combiner andsplitter, and for various RF cables are nottightened, both the reception performance

and the transmit performance of the systemwill decrease. In this case, the coverage areaand the conversation quality will be affected. The transmitter antenna and the receiverantenna are inversely connected due toinconsistent configuration of the set-topjumper and data. The connection between the jumper andfeeder is not tight, which results in high lossand standing wave ratio. In this case, thecoverage will be affected and interference willbe caused.

[Tower amplifier problem] Water penetration will increase the loss,deteriorate the standing wave ratio, anddecrease the receiver sensitivity. The damage of the LNA (it is in the toweramplifier) will decrease the gain or evendecrease the gain to a negative value. The input end and the output end of thetower amplifier are inversely connected. Inthis case, the tower amplifier will be short-circuited. If the short circuit lasts for a longtime, the front module will be damaged.

[BTS front module problem] Isolator problem Duplexer and other filter damage Standing wave ratio error alarm LNA (low noise amplifier) damage Low TRX or amplifier output power

[Parameter configuration problem]

The parameters affecting coverage are listedbelow:

TRX power class Tower amplifier attenuation coefficient MS maximum transmit power controlpower MS minimum Rxlev RACH minimum access threshold

III. Coverage Cases

Case 1: Use down tilt omni antenna toimprove coverage[Problem description]In a suburban area, the omni antenna with a gainof 11dBi is used for the BTS. This coveragedistance can reach 9km in plain environment.However, the coverage in the area near the BTSis poor. The Rxlev in the small town 800-1400maway from the BTS is about -90dBm.[Problem analysis and solution]On-site survey shows that the antenna height istoo great. The height of the tower on which theantenna is installed 50m. Moreover, the tower isestablished on a small mountain, so the town is120m below the antenna. The first judgment isthat the phenomenon of "shadow under tower"has been caused.Further analysis of the collected data finds thatomni antenna is used for the BTS. The antennagain is 11dBi, and the vertical half power angel is7 degrees. If the valid antenna height is 120m,the half power points of the antenna major lobe

are scattered in the area about 2000m away fromthe BTS. Therefore, this town is not in thecoverage area of the BTS.Through checking the fluctuation of the Rxlevaccording to the drive test map, engineers foundthat this town locates within the radiation area ofa zero power point of the BTS. However, thetown is too far away from the mountains around,so it cannot get the signals reflected by themountains. Therefore, the Rxlev in this town isquite slow.

After having replaced the antenna with an omni antennawith 5 degrees of the down tilt angle, engineers retestedthe Rxlev and found that it increased by 15-20 dB in theareas 3km within the BTS. In some areas, the Rxlev isincreased by 30 dB. Therefore, the coverage has beenimproved remarkably.Case 2: Improper installation of omni antenna haseffect against the coverage

[Problem description]A new BTS has been enabled for a local network.Users complain that the coverage area becomesmaller after that. For the low narrow areas 2kmaway from the BTS, the Rxlev is already lowerthan -90dBm.[Problem analysis and solution]Through surveying the environment around theBTS, engineers found that the major transmitterantenna and the diversity receiver antenna areinstalled in a plane parallel to the road.Apparently, this kind of installation does notmeet the criteria.

The correct way is to install the major transmitterantenna and the diversity receiver antenna in a planeperpendicular to the road. In addition, the majortransmitter antenna must be located at one side of theroad.

Case 3: Improper configuration of data causespoor coverage[Problem description]During the optimization for a place, engineersfound that the signals at a section of the road inthe suburban area are quite poor. The measuredRxlev is -95dBm.[Cause analysis]This section locates in the suburban area and isabout 3km away from the urban area. There is noabrupt change in terms of landform within thissection. Theoretically, the Rxlev here should beabout -80dBm, so the difference between thetheoretical Rxlev and the measured Rxlev isgreat. According to the frequency sweep test, thestrength of the Fa signal is about -95dBm, andthe strength of the Fb signal is about -80dBm.For this section, it is covered by three cells of theBTS A and BTS B that are installed in the urbanarea (the BCCH frequencies are Fa and Fb). Inaddition, a cell of the BTS C installed at theremote suburban area also covers the section (theBCCH frequency is Fc).Through checking data, engineers found that theFb is not included in the neighbor channelnumbers of the A-3 cell in the BA1. When theMS moves from the urban area to the suburban

area, it will choose A-3 cell to camp on, becausethe Fb is not configured in the neighbor channelnumbers. In this case, the MS cannot reselect theB-3 cell to camp on. In the cell neighborrelationship list, the A-3 cell and B-3 cell cannotwork as the neighbor cell for each other, and theFb is not configured in the neighbor channelnumbers of the A-3 cell listed in BA2. Therefore,in conversation mode, the MS cannot keep theconversation in A-3 cell. When it arrives at thissection, it cannot hand over to the B-3 cell.Therefore, the signals are poor, so is the voicequality.[Solution]Enable the A-3 cell and B-3 cell to work asneighbor cell for each other.Case 4: Irrational BTS swap affects coverage[Problem description]In an urban area, a BTS must be swapped for thebuilding on which the BTS was installed were tobe moved. Considering that coverage for thescenic spot 2km away (the scenic spot locatesbehind a hill) is poor, so engineers intended toinstall the BTS on the top of the hill. On the topof the cell, the whole city and the scenic spot canbe seen. However, after the BTS swap, userscomplaint that there were no signals in the indoorenvironment of the cells near the site where theold BTS was installed.[Cause analysis]The buildings of the resident area are denselydistributed and the average height is 8m. Before

the BTS swap, the cell used for this area is only100m away, and the antenna height is 15m.Therefore, the indoor conversation quality can beguaranteed. After the BTS swap, however, thecell used this area is 1.8km away, and theantenna height is 30m. In this case, the signalsare quite weak when arriving at the bottom ofresident area, though the signals falling at the topof the building is good. To solve this problem,you can only increase the output power of thetransmitter antenna or increase the antenna gain.However, the coverage is still not to users'satisfactory even increase the antenna height to30m. Therefore, when swapping or constructinga BTS at the densely populated area, you shouldpay attention to the following items:

It is suggested that the distance betweenthe BTS and the resident area is equal to orsmaller than 150m, otherwise the coverage forthis area will become weak dramatically. The antenna of the swapped BTS cannotbe too great. If the BTS is installed amongresident buildings, the antenna height issuggested to be 7-10m. If the BTS is installedbeyond the resident buildings and thebuildings are high, you can increase theantenna height accordingly. You can solve the problems concerningcross area coverage can through controllingthe power class of the BTS, tuning theazimuth angle of the antenna, or tuning the tiltangle of the antenna.

[Solution]

According to on-site survey, engineers found thatthe indoor signals of this area are too poor tohold the conversation. This area can be seenclearly from the tower on which the BTS isinstalled. The distance between the BTS and thearea is only 1.8km, and between them are vastfarmlands. To solve this problem, you canattempt to tune the azimuth angle and the tiltangle of the antenna. If the coverage is not yetimproved, you can use the following methods:

Replace the common antenna used for thiscell (its gain is 15dBi) with the high-gainantenna used for the scenic spot (its gain is18dBi) There are 4 TRXs in this cell, all in SCUmode, replace the SCU mode with the dual-CDU mode.

After the above methods are done, the antennagain for this area can be increased by 6dB. Afterthe antenna replacement, you need to tune theantenna tilt for the best coverage.Through retesting the indoor signal level,engineers found that it increased by 6-12dB. Andeven the common MS can keep normalconversation.

5.5.2 Interference Problems

Interference is a key factor affecting networkperformance, including conversation quality, calldrop, handover, congestion, and so on.

I. Interference Sources

In the mobile telecommunication system, when the BTSis receiving the signals from a remote MS, it will notonly be interfered by other telecommunicationequipments, but also it will be interfered by the otherBTSs and MSs within the system.Hereunder introduces the interference sources affectingthe GSM system.

Intra-network interferenceIf the frequencies are improperly planned, orthe frequency reuse is too aggressive, intra-frequency interference or neighbor cellinterference will be caused. Repeater interferenceAt the early stage of network construction,repeaters are widely used for extending thecoverage distance of the network. However,if the repeaters are improperly planned, thenetwork will be interfered.If the repeaters are not installed according torequirement, that is, there is not enoughisolation left between the donor antenna andthe subscriber antenna, the BTS to which therepeaters attach will interfered.For the repeaters enabling broadband non-linear amplifier, the intermodulation indexesare far greater than that required in theprotocols. In this case, the greater the poweris, the greater the intermodulation will be.Therefore, the BTS near the repeaters will beinterfered. Interference from other big-powertelecommunication equipments

These equipments include radar, analog BTS,and other telecommunication equipmentsusing the same band. Hardware problemsTRX problem: If the performance of the TRXdecreases, the system may be interfered.CDU problem or splitter problem: Activeamplifier is used in the CDU splitter andsplitter module. When any problem occurs,the system may also be interfered.Stray and intermodulation: If the out-bandstray of the power amplifier or the TRX ofthe BTS go beyond requirement, or theisolation of the transmission and thereception of the CDU duplexer is too small,the connection channel will be interfered.Meanwhile, the passive equipments, such asthe feeder and the antenna, will generateintermodulation.

II. Interference Positioning and Elimination

[Positioning and elimination procedure]1) Find out the interference cell according toKPIIf the call drop rate, handover success rate,traffic volume, congestion rate, andinterference band of a cell deteriorate to a badlevel abruptly, it means that interference mayexist in the cell.In this case, you can also check the historicalrecord of operations made in this cell. Forexample, check if the hardware and software

of the BTS has been added or increased andif the data of the BTS has been modified.Generally, the appearance of interference isrelated to these operations.If these parameters are not adjusted, theinterference may be from the hardware itselfout outside factors. In this case, you aresuggested to check if it is hardware problem.If it is not, you should check outside factors.

2) Check OMC alarmSometimes high call drop rate, low handoversuccess rate, and high congestion rate may berelated to equipment problems. In this case,you can check OMC alarm records. Theserecords are related to the deterioration ofthese indexes.

3) Check frequency planningIf the interference is doubt in a cell, you cancheck the frequency planning for the cell andthe neighbor cells of the cell. For this check,you are required to make clear thedistribution of the antennas, find out theazimuth angle of each cell, draw thetopology, and mark the BCCH/TCH channelnumbers. Meanwhile, you are also required tocompare the planned channel numbers withthe configured channel numbers in the BSC.According to the accurate frequency planningtopology, you can make sure if the intra-frequency interference or neighbor frequencyinterference is present in the network.

4) Check cell parameter configuration

The cell parameters, such as CRO, threshold,handover duration, neighbor cell relationship,and so on, may have interference against thesystem.If the CRO is set to a great value, the MSmay be guided to an idle cell whose level islower than its surrounding cells. Once theconversation is started but the C/I cannotmeet the threshold requirement (12dB),interference will be caused.If neighbor cells are missing, the MS cannothand over to a cell with better signal leveland quality. In this case, the interference willalso be generated. If the handover thresholdand the P/N are too great, the handoversbetween cells are unavailable. If the P/N istoo small, however, it will result in frequenthandover. In this case, both the call drop rateand the system load will be increased.

5) Drive testDrive test is an effective method to positionthe interference. There are two drive testmethods: idle mode test and dedicated modetest.For idle mode test, the test equipment can testthe signal level of both the signal level andthe neighbor cells. In addition, the testequipment can also perform the frequencysweep test for the designate channel numbersor bands. In this case, the interference causedby cross-cell coverage signals can bediscovered.

For dedicated mode test, the test equipmentscan test the signal level of the service cell andneighbor cells, the Rxqual, the TA, and so on.If the Rxlev is equal to or greater then -80dBm and the Rxqual is equal to or greaterthan 6 in an area, it can be confirmed that theinterference exists in the area. Some testequipment can display the FER (frame errorrate). Generally, if the FER is greater or equalto 25%, the conversation will not becontinuous. That is, the interference exists.

6) Interference eliminationYou can eliminate the interference accordingto the above checked results, and thenevaluate the elimination through KPI anddrive test.

[Hardware problem positioning andelimination]When the interference is doubted in a cell, youshould first check if the BTS where the celllocates works normally. In the remote end, youshould check if there is antenna alarm, TRXalarm, or BTS clock alarm generated. In the nearend, you should check if there is antennaproblem, water penetration, feeder (jumper)damage, CPU problem, TRX problem, wrongjumper connection or clock problem occurred.

Antenna performance declineAntenna a passive component and its damageprobability is small. However, if the antennais damaged or its performance declines, thevoice quality will become poor.

Antenna connector problemGSM RF signals are micro wave signals. Ifthe connections between TRX, CDU, feeder,and antenna have any problem, both thestanding wave ratio and the intermodulationwill increase. In this case, the interferencewill be resulted. Inverse antenna connectionThe inverse antenna connection is acommonly seen problem. If the antenna isinversely connected, the channel numbersused by the cell and the planned channelnumbers are completely inconsistent. In thiscase, intra-frequency interference, inter-frequency interference, and handoverdifficulty will be resulted. Especially for thenetworks that have inadequate frequencyresource, the inverse antenna connection hasgreat effect against network quality. Jumper problemMany jumpers locate between antennas, sothey are often wrongly connected. In thiscase, high call drop rate will be resulted. TRX problemIf TRX problems occur, the interference willincrease, the coverage distance area willdecrease, and the access is difficult. Clock failureIf the clock deviation is too great, it is hardfor the MS to lock the frequencies of theBTS, so the handover failure always occurs,

or the MS cannot camp on any cell of theBTS. In addition, if the clock deviation is toogreat, the BTS cannot understand the signalsof the BTS, which will result in bit errors.However, the clock failure will not reallyintroduce interference, but it is thetransmission errors that make the voicequality decrease. ConclusionAny problem concerning the TRX, CDU,feeder, antenna, jumper, and connector maycause interference or call drop. Therefore, ifinterference appears, you should check thehardware of the BTS. In addition, BTS clockfailure will also cause interference and calldrop.It is easy to solve the hardware problemsthrough changing the boards or adjustingtraffic data. If there is spectrum analyzeravailable, you can position the problem moreefficiently. Especially when the interferenceappears without any modification of networkdata, you should focus on checking thehardware.

[Intra-Network Interference]The intra-network interference is mainly fromintra-frequency interference and neighbor cellinterference. When C/I is smaller than 12dB orthe C/A is smaller than -6dB, the interference isunavoidable. However, the aggressive frequencyreuse technology will increase of the occurrenceprobability of interference.

Same-frequency and neighbor frequencyinterferenceIn GSM system, the frequency reuse isunavoidable. When the frequency reusedistance of two cells using the samefrequency is smaller than cell radius, same-frequency interference will be easily caused.Past experiences show that the frequencyreuse must be avoided in many cases.

The interference against the uplink channelnumbers can be judged by the interference banddata in the traffic statistics.For the interference against the downlink channelnumbers, the existing drive test equipments canbe indirectly used to measure if the same-frequency interference is present. First youshould lock the test MS in the service cell andenable make the MS work in conversation modeduring drive test. If you find that the Rxlev in anarea is high but the Rxqual is low, it is likely thatthe same-frequency is present in this area.

Interference caused by cross coverageIn a properly designed network, each cellcovers the areas around the BTS only and theMS camps on or holds conversation in thenearest cell. Cross coverage means that thecoverage of a cell is too large and the cell cancover the areas under the control of otherBTSs. If cross coverage occurs, irrationaltraffic absorption, interference, call drop,congestion, and handover failure may arise.

Interference caused by aggressivefrequency reuseCapacity and quality always contradicts toeach other. In urban areas, the aggressivefrequency reuse technology must be used forthe number of subscribers in urban areas aregreat. In this case, the network quality willsurely decrease. In the areas where BTSs areirrationally distributed, the aggressivefrequency reuse technology may cause thecollision of same frequency and neighborfrequencies. Interference caused by repeaterIt is convenient to use repeater for specialcoverage. However, if a repeater is notqualified or it is not properly installed, it willcause interference. Interference caused by outsideenvironmentOutside environment, such as TV station,big-power radio station, micro wave, radar,high voltage wire, analog BTS, and so on,will cause interference.

III. Interference Cases

Case 1: Interference cause by antennaperformance decline[Problem description]There are 5 BTSs in a county. The configurationtype is S4/4/4 or S6/6/6. The interference band 5reaches 15 according to the TCH performance

measurement of the most cells. There is no alarmfound at the OMC.[Problem positioning and solution]

1) Through monitoring and registering theinterference band traffic statistics for theproblem cells all day, engineers found that theinterference band 5 mostly appeared at daytime, and it seldom appeared at early morning.2) Through sending the idle BURSTS of allthe BTSs, engineers found that theinterference bands of these cells appeared inthe early morning. If the sending of these idleBURSTS stopped, these interference bandsdisappeared. Therefore, it can be proved thatthe interference came from the network. It isnot related to other telecommunicationequipments.3) The frequencies and other data were notadjusted before the interference appeared, sothe interference is not related to the frequencyplanning.4) Through surveying the RXM testinterface of the CDU using the spectrumanalyzer during the traffic peak at day time,engineers found that the broadbandinterference was strong and the back noisewas rising.5) There was no interference in one cell, butthe interference in another two cells wasstrong. Through replacing the antenna feederof the cell with no interference with theantenna feeders of the cells with interferenceand sending idle BURSTS, engineers found

that the interference went with the antennafeeder. Therefore, it can be decided that theproblem occurred at the antenna and feedersystem.6) Through changing the antenna, engineersfound that the interference went with theantenna. Therefore, the problem is likelypresent at the antenna.7) Through replacing the antenna with dualpolarization antenna, engineers found that thestrong interference disappeared immediately.Through replacing the old antenna of anotherBTS with a new one, engineers found that theinterference also disappeared.

Case 2: Call drop caused by intra-networkinterference[Problem description]

Customers in a place complaint that call drop happenfrequently.[Problem analysis and solution]

1) Through a careful test, engineers foundthat there were 12 channel numbers gatheringat the call drop spot and Rxlev reached -73dBm. When the MS seized channel number11, the interference from channel number 112caused the call drop.2) Through testing the CGI of channelnumber 12 using test MS, engineers found thatthis channel number was one of the BCCHnumber of D3.3) Through surveying BTS D, engineersfound that the antenna of D3 is installed at the

top of a building. In addition, a house made ofglass was found 8m away and 4m under theantenna. Engineers tested that the signalstrength near the antenna was about -45dBm,and the signal strength at near the glass was -30dBm, which was beyond the expectation ofengineers. In fact, the cause was that thesignals reflected by the glass were reflected tothe call drop spot.4) It is suggested to change the antennainstallation place and channel number. Youshould interchange the channel number 111and channel number 114 of BTS A andincrease the down tilt angle of A3 cell. Inaddition, to avoid the interference caused bychannel number 111 after the interchange, youshould adjust the direction of channel number113 of C1 cell.5) Test shows that everything is normal afterthe adjustment. The channel number 113 ofBTS C has no effect against channel number114. And the call drop disappears.

Case 3: Interference caused by repeater[Problem description]Users in an area complaint that the MS cannotseize a channel to hold conversation, or the noiseis great after channel seizure and the channel andthe MS signal is strong. Two BTSs are installedin this area. The antenna azimuth angle of cell1rightly directs to the north. Before usercomplaint, the BTS in this area ran normally andthe network indexes met the requirement. Afterthe problem arisen, the traffic volume of the two

BTSs dropped sharply from the perspective oftraffic statistics indexes. In addition, the trafficvolume of cell1 and cell3 also dropped sharply.Though the signals for the conversation werestrong, the voice quality was quite poor.According to traffic statistics, the interferencebands of the four cells were of level 3, level 4,and level 5, and 95% of the channels wereinterfered. In addition, other channels wereinterfered to some extent. However, no alarmmessages were generated at the OMC.[Problem analysis and solution]

1) According to user feedbacks, the possiblereasons include transmission problem, antennafeeder problem, hardware problem, intra-network interference, and outside interference.2) The uplink interference signals in thenorthwest direction might strong. Therefore,cell1, cell2, and cell3 of the two BTSs wereinterfered, in which cell1 and the cell3 wereseriously interfered.3) Through on-site dialing test, engineersfound that it was hard to make calls in theareas covered by cell1 and cell3. Even if a callwas put through, the voice quality was quitepoor. In addition, the voice was discontinuousand the interference was strong. Throughusing MS to call a fixed phone, engineersfound it was hard to hear the voice clearly. Onthe contrary, they could hear the voice fromthe fixed phone clearly. This has proved theabove analysis. That is, the interference mightbe from the outside, or the standing wave

problem was occurring at the antenna (fromthis perspective, it can be judged that theinterference existed on the uplink only).4) Through using antenna feeder analyzer toperform on-site test, engineers found noproblem was existing at any BTS. A newrepeater was found in this area, and it waslocated two kilometers away from the BTS inthe northwest direction. Moreover, theinterference appeared just when the repeaterwas enabled. On-site test found that the BTSbecame normal state once the repeater shutdown, and the interference bands also becamenormal, so did the call. If the repeater wasenabled, however, it was hard to make callsand the interference was strong. At last, theagreement to shut down the repeater wasreached. After that, the conversation becamenormal.

Case 4: Microwave interference[Problem description]During network maintenance, through analyzingBSC traffic statistics, engineers found that thecall drop rate of the cell2 and cell3 of a S2/2/2BTS arisen abruptly, and the value even reached20% at some time.[Problem analysis and solution]

1) Through checking BSC traffic statistics,engineers found that the number of idle TCHswas increasing at the interference bands 3-5around 8:30. Around 10:00, the idle TCHswere found at the interference band 4 and

interference band 5. Around 22:00, the idleTCHs were found at the interference band 1.Therefore, it could be judged that theinterference existed.2) Because the BTS ran normally, theproblems cannot be related to frequencyplanning.3) According to the TRX managementmessages, engineers found that theinterference existed at the four boards of thecell2 and cell3 of the BTS. Because theprobability for the four boards to be damagedsimultaneously is quite small, the TRXproblem can be excluded. However, one boardwas changed in case of abnormal conditions,but the interference was not eliminated.4) Through checking all the BSC trafficstatistics data, engineers found that cells of theBTSs near the BTS and the cells of the cell2and cell3 of the BTS were interfered to someextent. In addition, engineers also found thatthe SDCCHs (16 SDCCHs in total) of theseriously-affected cells were seized atsometimes. However, the number ofsubscribers determined that the probability forall the SDCCHs to be seized simultaneously isquite small. Therefore, it could be judged thatthe uplink was interfered by outside factors.However, the interference might be related todirection only.5) To further position the problem,engineers interchanged the jumpers of cell1and cell3 at the set top. In this case, the

interference was found at cell1, but theinterference was disappearing from cell3.Therefore, the interference was not related tochannel number.6) Because the interference was not relatedto channel number, it might be the big-powersignals that caused the interference.7) Through using the spectrum analyzer toperform the measurement at the outputinterface of the BTS splitter, engineers foundthat the big-power signals existed at the904MHz channel number (it has an interval of5M between the used channel number. For theBTS where the interference was strong, thesignal level can reach as high as about -25dBm. For other BTSs, the signal level wasabout -50dBm. Therefore, it could be judgedthat it was this signal that affected the BTS.8) Through using the spectrum analyzer toscan the areas near the BTS, engineers foundthat there was a microwave antenna outputtingbig-power at the channel number 904.9) The interference disappeared after themicrowave equipment was shut down.

5.5.3 Call Drop Problems

For the GSM network, call drop is users' majorworry and the call drop rate is an important indexevaluating network quality.

I. Call Drop Resasons and Solutions

i) Call drop due to coverage reasons[Reason analysis]

Discontinuous coverage (dead zone)For a single BTS, the quality of the signals atthe edge of the station is quite poor, so theMS cannot hand over to another cell. In thiscase, the call drop occurs.If the landform of the coverage areas iscomplex or fluctuates greatly, or the radiotransmission environment is complex, thesignals will be barred. In this case, thecoverage is discontinuous and call drop willoccur. Poor indoor coverageIf the buildings in an area are denselypopulated, the signal attenuation is great. Andif the walls of the buildings are thick, thepenetration loss is great and the indoor signallevel is low. In this case, the call drop mayeasily occur. Isolated island effectAs shown in .لم یتم العثور على مصدر المرجع! خطأ , the servicecell forms an isolated island due to variousreasons (for example, the power is too great).In this case, the MS still seizes the signals ofthe service cell A after moving to cell C, butthe cell A does not define the neighbor cell C.At this time, if the MS still performs thehandover according to the neighbor cell Bprovided by neighbor cell A, it cannot find asuitable cell. In this case, the call drop willoccur. Small coverage

If the coverage is too small, the hardwareequipment of a cell may fail. For example,the antenna is barred or the TRX failureoccurs (the power amplifier part).

[Judgment methods]First you should find out the areas where thecoverage is inadequate according to usercomplaints, and then you should perform thedrive test in a larger scope to check if the signallevel and the handover are normal and if the calldrop exists. In addition, you can employ thetraffic statistics recorded at the OMC to checkthe BSC overall call drop rate and find out thecell with great call drop rate. Furthermore, youcan still make the analysis and judgment byreferring to other traffic statistics items.Hereunder lists some ones:

Power control performance measurement(to check if the mean uplink and downlinksignal strength is too low) Rxlev performance measurement (to checkif the ratio of the low Rxlev is too great) Cell performance measurement/inter-cellhandover performance measurement (to checkif the level class and the mean Rxlev are toolow) Call drop performance measurement (tocheck if the signal level is too low during calldrop and if the TA value is normal before calldrop) Defined neighbor cell performancemeasurement (to position the cell with lowmean signal level)

Undefined neighbor cell performancemeasurement (to check if the undefinedneighbor cells with high signal level exist) Power control performance measurement(to measure the greatest TA value between theMS and BTS)

[Solutions]1) Check the areas where the coverage isinadequateYou can find out the area where the coverageis inadequate through drive test. For anisolated BTS or the BTSs installed inmountain areas that cannot form seamlesscoverage, you can add BTSs to these areasfor seamless coverage. Or you can improvethe coverage through other means. Forexample, you can enhance the maximumtransmit power of the BTS, change theantenna azimuth angle, change the antennatilt, change the antenna height, and so on. Inaddition, you should also analyze if the calldrop is caused by landforms. Generally, calldrop can easily occur at tunnels, big shoppingmarket, underground railway entrance,underground parking lot, and low-lyingplaces. In this case, you can use the microcell to solve the coverage problem.

2) Ensure indoor call qualityTo ensure indoor call quality, you shouldmake sure that the outdoor signals are strongenough. To strength the outdoor signals, youcan increase the maximum BTS transmit

power, change the antenna azimuth angle,change the antenna tilt angle, and change theantenna height, and so on. If the indoor callquality is still not improved remarkably, youcan consider adding BTSs. For improving theindoor coverage of office buildings andhotels, you can consider using the indoorantenna distribution system.

3) For the cells having no neighbor cells,you can configure the neighbor cells for thecell so as to reduce the call drop rate. Toeliminate the isolation island effect, you canreduce the tilt angle of the BTS.4) Eliminate hardware problems

You can check if there are hardware problemsand if the coverage area is too small throughdrive test. If the call drop rate of a cell arisesdramatically but all other indexes are normal,you should check if the neighbor cells of this cellwork normally. (Generally, the downlinkproblems may occur. For example, TRXproblem, diversity unit problem, and antennaproblem are commonly seen. If the uplink fails,the outgoing handover failure rate of the old cellwill be high.)ii) Call drop due to handover reasons[Reason analysis]

Irrational parameter configurationIf the signal level at the cross-area of twocells is quite low, the level of the handovercandidate cell is too low, and if the handoverthreshold is too low, some MSs will hand

over to the neighbor cell when the signallevel of the neighbor cell is higher than thatof the service cell. If the signal level of theneighbor cell deteriorates dramatically justafter the handover, the call drop will occur ifno suitable cell is available for the handover. Incomplete neighbor cell definitionIf the neighbor cell definition is incomplete,the MS will hold the conversation in theexisting cells until it moves beyond the edgesof the cell but cannot hand over to a strongercell. In this case, the call drop will occur. Neighbor cells with same BCCH and sameBSIC exist. Traffic congestionIf the traffic is unbalance, no TCH will beavailable in the target cell. In this case, thehandover failure will occur. BTS clock lost synchronizationIf the BTS lost synchronization, thefrequency offset will go beyond therequirement. In this case, the call drop willoccur if handover fails. T3103 expiry

The T3103 will be started when the networksends a handover command. Upon the receptionof the message to complete the handover or themessage to remove the command, the T3103 willstop. T3103 is used to hold the channel longenough for the MS to return to the old channel. Ifthe T3103 is set to a too small value, the MS

cannot necessarily return to the old channel. Inthis case, call drop may occur during handover.[Judgment methods]You can judge if the cells with low handoversuccess rate, frequent re-establishment failures,and high call drop rate through analyzing trafficstatistics indexes. After the judgment, you canfind out what causes the handover. For example,the uplink and downlink Rxlev can cause thehandover; the uplink and downlink Rxqual cancause the handover; power budget can causehandover; call direct retry can cause handover;and also handover can be initiated by trafficreasons.To check if the BTS clock runs normally, youcan check if the any alarm is generated for theBTS clock. If necessary, you must correct theBTS clock to eliminate clock problem. You cancheck if there is handover problem through drivetest. If there is a problem cell, you shouldperform drive near the cell for several times.Hereunder lists the indexes concerning call drop:

Inter-cell handover performancemeasurement (frequent handover failures,frequent re-establishment failures) Inter-cell handover performancemeasurement (frequent handovers, high re-establishment rate) Undefined neighbor cell performancemeasurement (the undefined neighbor celllevel and the number of measurement reportgo beyond the standard)

Outgoing cell handover performancemeasurement (find out the reasons for lowoutgoing cell handover from the handovertarget cell) Low incoming cell handover success rate;the cell handover parameters are improperlyset; the target cell is congested. TCH performance measurement (thehandover times are not proportional to theTCH call seizure successes; the handoverhappens too frequent)

[Solution]1) Check the parameters affecting thehandover. For example, you can check thehierarchical and level setting, each handoverthreshold, each handover hysteresis, handovertime, handover duration, the minimum accesslevel of the handover candidate cell, and soon.2) If the call drop is caused by unbalancetraffic volume or if the call drop occurs due tono handover channel is available at the targetBTS, you can solve the problem by adjustingthe traffic volume. For example, you canadjust the project parameters, such as antennatilt and antenna azimuth angle, to control thecoverage scope of a cell. To balance the trafficvolume, you can use CRO to guide the MS tocamp on other idle cells, or you can set thehierarchical and level priority to guide the MSto hand over to the idle cell. In addition, youcan solve the problem by expanding the TRXdirectly.

3) Calibrate the problem BTS clock toenable the synchronization of the clock.

iii) Call drop due to interference reasons[Reason analysis]If the MS receives strong same-frequencyinterference signals or strong neighbor frequencyinterference signals in the service cell, the biterror rate will deteriorate. In this case, the MScannot demodulate the BSIC code of theneighbor cells accurately, or it cannot receive themeasurement report from the MS correctly. As aresult, the conversation will be interfered, the callquality will become poor, and call drop willoccur.[Judgment methods]The interference may be from the network itselfor the outside network, or it may exist in theuplink signals or downlink signals. The followingmethods can be used to position the interference.

Find out the cells might be interferedthrough checking traffic statistics. Perform the call drive test for the areasthat might be interfered and check the uplinkand downlink interference according to usercomplaint. You can find out if there is a placewhere the signal is strong but the call qualityis poor through drive test tools. In addition,you can use a test MS to perform dialing testto check if a channel number is interfered. Check the frequency planning to see ifsame-frequency interference and neighbor

frequency interference occur in the area wherethe frequency is improperly planned. Adjust the channel numbers that might beinterfered to see if the interference can beavoided or reduced. Exclude the interference caused byequipment. If the previous methods fail to eliminatethe interference, you can use the spectrumanalyzer to scan the frequencies to find out theinterfered channel number and the interferencesource.

Hereunder lists several traffic statistics indexesused for interference analysis:

Interference bandYou can check the uplink interferencethrough analyzing the interference band inthe traffic statistics. If an idle channel appearsat the interference bands 3-5, the interferenceis present. If it is intra-network interference,it will increase as the traffic volume grows.Generally, if it is outside interference, it isnot related to traffic volume. It must bepointed out that the interference bands arereported to the BSC by the BTS TRXchannel (when in idle mode) through RFresource indication messages. If the currentchannel is busy and cannot report RFresource indication message, you mustconsider the traffic volume for the measuringthe interference bands. Rxlev performance measurement

The Rxlev performance measurementprovides the matrix relationship between thesignal level and quality. If the signal level ishigh but the quality is poor, it means that theinterference (same-frequency interference,intra-frequency interference, and outsideinterference) is present at the channelnumbers of the TRX board. Poor quality handover ratioThe cell performance measurement, inter-cellhandover performance measurement, or theoutgoing cell handover performancemeasurement records the outgoing handoverattempt times. If the frequent handover iscaused by poor signal quality, it means thatthe interference is present. Rxqual performance measurement

It is related to the mean Rxlev and Rxqualduring call drop.

Call drop performance measurementIt records the mean Rxlev and Rxqual

during call drop. Frequent handover failures and frequentre-establishment failures

It means that the interference may bepresent in the target cell.

[Solutions]1) Check the interfered road and thedistribution of signal quality through dive test.As far as the actual conditions are concerned,you can adjust the BTS transmit power and

antenna tilt of the related cells or adjust thechannel number planning to avoid theinterference.2) Use DTX technology, frequency hoppingtechnology, power control, and diversitytechnologyThese technologies can be used to reduce thesystem noise and enhance anti-interferencecapacity of the system. DTX is divided intouplink DTX and downlink DTX. In this case,the transmit time can be reduced and theinterference level of the system can also bereduced. However, you should adjust theDTX according to the actual radioenvironment and the neighbor cellrelationship. When signals received by theMS are poor, the use of the DTX will resultin call drop. If the downlink DTX is enabled,the BTS will increase its transmit power afterthe call is established. During theconversation, however, the BTS will reduceits transmit power. In this case, theinterference against other BTSs will bereduced. If the interference is present near theBTS, the downlink DTX will deteriorate theconversation quality. When the BTS reducesits transmit power, the conversation qualitywill decrease or the call drop may even occurin the areas where the Rxlev is low but theinterference signal is strong.

3) Solve the equipment problems, such asthe self-excitation of TRX boards and theantenna demodulation interference.

4) Exclude the outside interference.iv) Call drop due to antenna feeder reasons[Reason analysis]

Engineering problem may be one of thereasons. For example, if the transmit antennasbetween two cells are inversely connected, thelevel of the uplink signal will be far poorerthan that of the downlink signal. In this case,the call drop, one-way audio, and calldifficulty will be found in the areas far awayfrom the base station. If polarization antennas are used, a cellhad two sets of antennas. If the tilt angles ofthe two antennas are inconsistent with eachother, the call drop will occur.If a directional cell has a master antenna anda diversity antenna, the BCCH and theSDCCH of the cell may be transmittedthrough the two antennas respectively. If thetilt angles of the two antennas are different,the coverage scope of the two antennas willbe different. In this case, the MS can receiveBCCH signals but cannot seize the SDCCHwhen starting a call. Thus the call drop isresulted. If the azimuth angles of the two antennasare inconsistent with each other, call drop mayalso occur. That is, the MS can receive theSDCCH signals, but it may be assigned withthe TCH. In this case, the call drop will occur. The problems concerning antenna feederwill also cause call drop. For example, if the

antenna is damaged, or water penetrates intothe antenna, or connector problem is present,the transmitter power and the receiversensitivity will decrease. In this case, the calldrop will occur. To confirm the problem, youcan check the standing wave ratio.

[Problem positioning and solution]1) Check if any alarm concerning thecombiner, CDU, tower amplifier, and standingwave is generated and check if the BTSboards are normal in the OMC.2) Analyze if the path balance is realizedaccording to traffic statistics.3) Further analyze if the path balance isrealized through monitoring the messages sentacross the Abis interface.4) Perform drive test and dialing test.During drive test, you can check if the BCCHnumbers of the service cell are consistent withthe planned ones, namely, if the transmitantenna of the antenna is correctly installed.5) Check and test the on-site BTSs. Here theinstallation of the azimuth angle and the tiltangle of the antenna must be checked. Inaddition, you should also check if the feederand jumper are correctly connected, if there isconnector problem, and if the feeder isdamaged. Furthermore, you must still check ifthe standing wave is normal.6) Judge if it is BTS hardware that causespath unbalance and call drop. To solvehardware problem, you can either change thecomponents that may have problems or

disable other TRXs in the cell. To find out theproblem TRX, you can position the problemthrough dialing test. Once a problem hardwarecomponent is found, you must replace it witha sound one. If no sound one is available, youmust shut down the problem hardwarecomponent to prevent it from affectingnetwork quality.

Hereunder lists several traffic statistics items forpath balance analysis:

Path balance measurement (to analyze ifthe path balance is realized) Call drop performance measurement (toanalyze the uplink and downlink level andquality during call drop) Power control performance measurement(to analyze mean Rxlev on the uplink anddownlink)

v) Call drop due to transmission reasonsIf the transmission quality across the Abisinterface and A-interface may be not good andstable, call drop may occur. The followingmethods can be used to solve the problem:

1) Check the transmission alarm and boardalarm and analyze if there is transmissionintermittent and board failure.2) Check the transmission channel, test thebit error rate, check 2M connectors, and checkif the equipment grounding is rational toensure stable transmission quality and reducecall drop rate.

3) Check the traffic statistics to see if thefrequent call drop is caused by transmissionproblems. Especially you should check TCHperformance measurement, because it canindicate if the A-interface failures during TCHseizure is normal, if the TCH utilization isnormal, and if the ground link call drop timesare great.

vi) Call drop due to parameter reasonsHere you should focus on checking if theparameters related to call drop are irrationallyset. If the following parameters are notirrationally set, the call drop may be resulted.

Radio link failure counterThis parameter acts on the downlink. Whenthe MS fails to decode the SACCH, it willuse this parameter to decide when todisconnect the call. If this parameter is set toa too small value, the radio link failures willeasily occur and cause call drop. For deadzones or the areas where the call dropfrequently happens, you are recommended toset this parameter to a greater value.When changing the radio link failure counter,you should change the corresponding T3109.The T3109 must be set to a value greatenough for the MS to detect a radio linkfailure. For example, if the value of the radiolink failure counter is 16 (about 8 seconds),the value of T3109 must be greater than 8seconds (you can set T3109 to 9 seconds or10 seconds).

SACCH multiframe numberThis parameter acts on the uplink. The BTSuses this parameter to notify the radio linkconnection failure message to BSS. The BSSside judges the radio link failure according tothe bit error rate on the uplink SACCH. Ifthis parameter is set to a too small value, theradio link failure will happen frequently andthe call drop rate will be resulted. Access control parametersThe access control parameters include theMinimum RACH Rxlev, RACH busythreshold, and so on. If the access controlparameters are irrationally set, the call dropwill be easily resulted. T3101, T3107T3101 is started when the BSC sends aCHANNEL ACTIVATE message to theBTS. It stops when an ESTABLISHINDICATION message is received. Thistimer monitors the immediate assignmentprocedure. If T3101 expires, the allocatedchannels will be removed.T3107 is started when the BSC sends anASSIGNMENT COMMAND message to theBTS. Once the BSC receives theASSIGNMENT COMPLETE message fromthe BTS, this timer will reset. T3107 is usedto hold the channels long enough so that theMS can return to the old channel. Or it canalso be used by the MS to release a channel.

If the two timers are set to a too small value,the system will not have enough time to sendthe ASSIGNMENT COMPLETE message tothe BSC. In this case, the call drop will occurif the timer expires. T200; N200T200 is an important timer used for theLAPDm (Link Access Procedure on the Dmchannel). It prevents the deadlock fromoccurring when the data is transferred acrossthe data link layer. In GSM system, themessages transferred across radio interfacescan be divided into two types: the messagesneeding opposite acknowledgement and themessages not needing oppositeacknowledgement.For the messages needing oppositeacknowledgement, a T200 must be startedonce the message is sent. If the oppositeacknowledgement is not received after aperiod of time, the message should beretransmitted. In this case, the timer must berestarted. If the retransmission times exceedthe maximum allowed times, the messagewill no longer be retransmitted and the linkwill be released. That is, this call drops. N200is the maximum retransmission timesallowed. T200 and N200 have different typesdepending on channel types (TCH full rate,TCH half rate, and SDCCH) and servicetypes (signaling and messages). The givenchannel type and service type matches a pairof T200 and N200.

The call drop rate can be reduced if themessage is retransmitted as early as possiblebefore the opposite acknowledgement isreceived. That is, the value of T200 must beset as small as possible and the value of N200must be set as great as possible. However, theT200 cannot be set to a too small value andthe N200 cannot be set to a too large value. Ifthe opposite party has acknowledged that thelink had been removed, the retransmissionwill make nonsense.Therefore, to reduce the call drop rate, youcan adjust the T200 and N200 according toactual radio conditions.

II. Call Drop Cases

Case 1: Call drop caused by frequencyhopping collision[Problem analysis]A BTS uses 1 x 3 RF frequency hopping. Aftercapacity expansion, the TCH allocation failurerate is still high due to radio link problems. Inaddition, the TCH call drop rate and incominghandover failure rate are high. The SDCCH calldrop rate is normal.[Problem positioning and solution]Because high call drop rate and high incominghandover failure rate come together with theTCH allocation rate, it can be judged that theproblem may arise during TCH assignment or thechannel numbers or timeslots seized by the callare interfered or unstable. Because the SDCCH

call drop rate is normal, it can be judged that theprobability for the BCCH carriers and BCCHnumbers to the interfered are small, but the non-BCCH carriers and non-BCCH numbers may begreatly interfered.Through checking the hardware, antenna feeder,and transmission, engineers found no problem.According to drive test, engineers found that thesignal level was high but the quality was poor.Through on-site dialing test, engineers found thatthe conversation quality was poor. Throughchecking engineering parameters, engineersfound that the MAIO of the new carrier was thesame as that of the old carrier. Therefore, it canbe judged that the call drop was caused by thefrequency hopping collision. After modifying theMAIO, engineers found that call drop ratebecame normal.Case 2: Call drop caused by isolated islandeffect[Problem description]Users complained that call drop always occurredabove the fifth floor of a building.[Problem analysis]

1) Through on-site test, engineers found thatcall drop and noise existed here. As far as thetest MS was concerned, it was always in theservice area of the other BTS (hereundercalled BTS B) other than the local BTS(hereunder called BTS C) before the call drop.2) It is estimated that the service cellbelongs to BTS B, which is 3-4 kilometers

away from the building. Therefore, it can bejudged that the signals from the BTS B arereflected signals. As a result, an area similar toan isolated island is formed.3) Through checking data configuration,engineers found that only the cell 2 of BTS Ahas the neighbor cell relationship with BTS B.Therefore, when the MS is using the signals incell 2 of the BTS B, if the signals in cell 3 ofBTS A were strong, and if the cell 2 of BTS Bhas no neighbor cell relationship with the cell3 of the BTS A, the handover cannot beperformed.The signals from the cell 2 of BTS B arereflected many times. Therefore, whensignals (from BTS B) received by the MSbecame poor dramatically, emergenthandover may be initiated. In this case,however, either the cell 2 or cell 3 of the BTSA is not an ideal candidate cell for the cell 2of BTS B. As a result, the MS may hand overto other BTS (hereunder called BTS C), butthe MS cannot receive the signals from BTSC. Therefore, call drop occurs.

[Solution]You are recommended to change the data in theBA1 (BCCH) list, BA2 (SACCH) list, andneighbor cell relationship list. For example, youcan configure the cell 3 of BTS A as the neighborcell of cell 2 of BTS B. To eliminate the isolatedisland effect, you should also further optimize theengineering parameters. After that, the call dropproblem can be solved.

Case 3: Reduce call drop rate throughoptimizing handover parameters[Problem description]The drive test in an area found that the call droprate at a cave near the BTS high because thehandover cannot be performed in due time.[Problem analysis and solution]The cave is near the BTS. The signal level of thetarget cell is about -80dBm in the cave, but thesignal level of the old cell drops below -100dBm.The downlink power of the two cells outside thecave is good, so the handover cannot be initiated.However, the signal level deterioratesdramatically in the cave, so the call drop occursbefore the measurement time is arriving.To reduce the call drop rate, you can optimizeand adjust the handover parameters:

1) If no ping-pong handover is present andthe conversation is continuous, you can makethe PBGT handover happen as easily aspossible.2) Set the threshold to trigger the emergenthandover rationally so that the emergenthandover can be triggered before call drop.

For the parameter modification, see

Handover parameter optimization

Parameter NameBeforeModification

AfterModification

PBGT handover measurementtime

5 3

PBGT handover duration 4 2

PBGT handover threshold 72 68

Uplink quality threshold foremergent handover

70 60

Minimum downlink power forhandover candidate cell

10 15

Case 4: Call drop caused by clock problem[Problem description]The cell A of an 1800MHz network has beencutover. After the establishment of a cell at siteB, the calls made in the cell handing over to theGSM900 MHz cell that shares the same BTS sitedrops in the GSM900 MHz cell. And the calldrop rate is great.[Problem analysis and solution]Engineers find that the clock of the GSM900MHz BTS and that of the GSM1800 MHz BTSare asynchronous. When the calls established onthe GSM1800 MHz cell intend to hand over tothe GSM900 MHz cell, the drive test data showsthat the FER arises dramatically first, and thengradually disappears to none. If the handover isfrom a GSM900 MHz cell to a GSM1800 MHzcell, this phenomenon is also present. Throughmonitoring signaling, engineers find that theconversation held several seconds before the calldrop is just process for call re-establishment.However, the test MS shows that the call hasbeen handed over to the GSM900 MHz cell.Therefore, it can be judged that the clocks areseriously asynchronous. To solve this problem,the carrier concerned and the GSM900equipment provider cooperate with each other onclock calibration. After that, the abnormal call

drop disappears. Therefore, for dual-bandhandover, the clock of the GSM900 MHz BTSand that of the GSM1800 MHz BTS must besynchronous.

5.5.4 Handover Problems

The MS is always moving during conversation.To ensure channel quality, the MS must measurethe quality of the channels of the surroundingchannels without stop, and then send themeasurement report to the BSC through theservice cell. The BSC will perform radio linkcontrol according to the signal level and qualitycontained in the measurement report. If the MSmoves to another cell, the new cell will replacethe old cell to ensure the continuity of theservice. The handover enables each cell to form aseamless network.

I. Handover Problem Positioning Steps

1) Find out if the problem occurs at anindividual cell or all cells and find out thecharacteristics of the problem cells. Forexample, if the cells are the neighbors cell of acell, or if they are co-BSC cells, or if they areco-MSC cells.If the handover between two cells fails, youshould focus on checking if the data of thetwo cells is correctly configured. In addition,you should also check the hardware of thetwo cells.If the problem is found in all the neighborcells of a cell, you should focus on checking

of the data of this cell is correctly configured.In addition, you should also check thehardware of the cell.If the problem is found in all the cells underthe same BSC, you should focus on checkingthe data configuration between the BSC andMSC.If the problem is found in all the cells underthe same MSC, the cooperation between thelocal exchange and the opposite exchangemay fail. For example, the signaling isincompatible and the timer setting isirrational.

2) Check if the data has been modifiedbefore handover problems occur.If the problem is found in an individual cell,you should focus on checking if the dataconfiguration for this cell has been modified.If the problem is found in all the cells underthe same BSC, you should focus on checkingthe data configuration for the local BSC andthe opposite MSC has been modified.If the problem is found in the cells under thesame MSC, you should check if the dataconfiguration for the opposite MSC has beenmodified.

3) Check if it is the hardware failure thatcauses the handover problem.4) Register the related traffic statistics items,such as the handover performancemeasurement and TCH performancemeasurement.

Check if the TCH seizure of the problemcell is normal. For example, if the calldrop rate is high.

Check if the outgoing handover successrate is normal.

Find out the causes for the handoverfailure.

Check if the radio handover success rateis normal.

5) Perform drive test for the problem celland analyze the drive test signaling. Check if the uplink and downlink of the

problem cells are balanced, becauseunbalanced path may cause handoverproblem (BTS problem may cause theunbalance).

Check if the measurement report for theproblem cell contains correct neighbor celllist.

Check if a call can hand over from aproblem cell to a neighbor cell correctlyand check if it can hand over from aneighbor cell to the problem cell.

Analyze if the signaling procedure for thehandover is normal.

II. Handover Problem Analysis Methods

i) Handover cannot be initiatedIf the MS is in a cell where the signal is poor, itcannot hand over to another cell. In this case, youshould consider of the handover conditions aremet and if there is an outgoing cell available.Hereunder details the possible reasons:

The handover threshold is set to a lowvalueFor edge handover, the handover triggeringcondition is that the Rxlev must be smallerthan the handover threshold. If the edgehandover threshold is set to a too small value,the signal level of the neighbor cells will befar higher than that of the service cell. In thiscase, the handover cannot be initiated. As aresult, the conversation quality will beaffected, or even the call drop will beresulted. The setting of the handoverthreshold depends on the coverage scope ofthe cell. You can change the size of theservice area of the cell through changing thehandover threshold. Neighbor cell relationship is not setThough the signal level in the neighbor cellsof the service cell is high, the MS will notreport the neighbor cells if the neighbor cellrelationship is not set. In this case, the MScannot hand over to a neighbor cell. Throughperforming cell reselection or dialing test,you can check the neighbor cell list reportedby the MS. If the MS has moved to the majorlobe of a cell but this cell is not found in theneighbor cell list, you should check if thecorrect neighbor cell relationship has beencorrectly set. During the drive test, you canuse another MS to scan the BCCH numbersto check if the strong BCCH numbers are inthe service cell or in the neighbor cell list.

Handover hysteresis is irrationally setIf the difference between the signal level ofthe handover candidate cell and that of theservice cell is greater than handoverhysteresis, the cell can work as a target cell.If the hysteresis is set to a too great value, thehandover is hard to be initiated. The best measurement time "N" and "P"are irrationally setDuring normal handover, the MS uses N-Prules to list the handover candidate cells in acertain order. If a candidate cell is the bestcell within P seconds out of N seconds, it willbe treated as the best cell.When there are two cells become the best cellalternately, the MS may find it hard to selecta best cell through N-P rule, which makes thehandover difficult. In this case, you canadjust the values of N and P and reduce themeasurement time to make the handoverdecision more sensitive to level change.If the landform and the ground objects of theservice cell are quite complicated, the signalsreceived by the moving MS will fluctuategreatly. In this case, the handover candidatecell cannot meet N-P rule, which will makethe handover difficult.

ii) Handover problem caused by hardwarefailureIf the data configuration for the problem celland the neighbor cells has not been modifiedrecently but the handover problems occur

abruptly, you should first consider if theproblems are caused by BTS hardwareequipment.If the cells sharing the same base station withthe cell have similar problem, you shouldconsider if the problem is caused by thecommon hardware of the cells.If the problem is found in only one cell underthe base station, you should consider if it isthe hardware of this cell that causes theproblem. For example, if some of the carriersare damaged. To test the problems of thiskind, you can disable some of the carriers. Ifthe handover success rate returns to normalstate after a carrier is disabled, you can checkif the problem is present at this carrier or ifthe CDU and antenna feeder part related tothis carrier fails. If signals of a cell on theuplink and downlink are seriouslyunbalanced, frequent handover will be causedand the handover success rate will decrease.To check if the signaling flow of the cell isnormal and if the uplink Rxqual anddownlink Rxqual are good, you can monitorthe messages sent across the Abis interface. Ifthe Rxqual is poor, it means that thehardware equipment of the fails or seriousinterference is present in the cell. In this case,the signaling exchange is unavailable and thehandover problem will occur.

iii) Handover problem caused by irrationaldata configuration

For stand-alone networking mode, if theoutgoing MSC or incoming MSC handover isabnormal, you should check if the signalingcooperation of the two MSCs is correct. Inaddition, you should also check if the dataconfiguration for the opposite MSC and thelocal MSC has been modified recently. For co-MSC networking mode, if thehandover is performed within the BSCs ofdifferent providers and the inter-BSChandover is abnormal, you should first checkif the signaling cooperation between the BSCsis normal, and then check if the dataconfiguration for the BSCs has been modified. If the abnormal handover is found at a cellonly, you need to analyze the abnormalhandover according to actual conditions.If the incoming handover of a cell isabnormal, you need first check if theincoming handovers to this cell is abnormal.Generally, when the handover is abnormal,the handover success rate is low, or even thehandover does not occur.If all the incoming handovers to this cell isabnormal, you should check if the dataconfiguration for this cell is correct. Here thedata configuration includes both the dataconfigured for this cell and the dataconfigured for other cells but is related to thiscell. For example, the CGI of this cell may becorrectly configured, but it may be wronglyconfigured in other cells.

If there is only one incoming handover to acell is abnormal but other incominghandovers to this cell are normal, in additionto checking if the data configuration for thiscell is correct, you should also check if thedata configuration for the neighbor cells iscorrect. Furthermore, you should also checkif the hardware equipment of the cell isnormal.The methods to analyze the abnormaloutgoing handovers are similar to themethods to analyze the abnormal incominghandovers. Check the timers (such as T3105, Ny1,T3103, and T3142) related to the handover.

T3105 indicates the interval for continuousPHYSICAL INFORMATION to be sent to theMS. The network will start T3105 for the sendingof the PHYSICAL INFORMATION. If the timerfails before receiving any correct frame from theMS, the network will resend the PHYSICALINFORMATION and restart the timer. A pieceof PHYSICAL INFORMATION can be sent Ny1times to the maximum. Here the product of Ny1and T3105 must be greater than the sum ofT3124 and delta ("delta" indicates the intervalbetween the expiry of T3124 and the reception ofthe Handover Failure message of the old BSC),otherwise the MS cannot perform successfulhandover.T3124 is a timer waiting for the PHYSICALINFORMATION from the network side during

asynchronous handover. When sending theHANDOVER ACCESS message on the DCCHfor the first time, the MS will start T3124. Uponreceiving a piece of PHYSICALINFORMATION, the MS will stop T3124. If thechannel type allocated in the HANDOVERCOMMAND message is SDCCH (+SACCH),the T3124 is set to 675ms. For other cases, theT3124 is set to 320ms.

III. Handover Cases

Case 1: No handover candidate cell isavailable due to CGI error[Problem description]The handover in an area is abnormal. When theMS moves from cell A to cell B, the signals incell B are stronger than that of cell A, but thehandover does not happen. After the MS movesfrom cell B to cell C, the MS hands over fromcell A to cell C.[Cause analysis]If a cell can work as a service cell and can handover to other cells, but the incoming handover isunavailable, you can check if the CGI, BSIC,BCCH number of the cell are correct.[Problem solution]

1) Use the test MS to lock the BCCHnumbers of cell B. The call is normal. The MScan hand over to any other cell by force.2) Make a call after locking the BCCHumber of any neighbor cell of cell B, and thenforce the MS to hand over to cell B, but the

handover does not happen, because nohandover command is seen in the drive testsoftware.3) The handover procedure requires the MSdetecting the neighbor cell signals andreporting the detected signals to BSC with ameasurement report. Upon receiving themeasurement report, the BSC must make thehandover decision. If the handover conditionsare met, the BSC should activate the TCH ofthe service cell, and then send a handovercommand to the MS.4) If the signals of cell B are far strongerthan that of cell A and the handoverconditions are met (the PBGT handoverthreshold is 70), but no handover command issent, it means that errors occur during theactivation of the target cell TCH.5) If the cell B works as the target cell butthe TCH cannot be activated, the data may beincorrectly configured for the cell. In thiscase, the BSC that contains the cell cannotfind the target cell, so the TCH cannot beactivated and no handover command can befound in the service cell.6) The CGI error is found in cell B throughdata checking. The handover is normal afterthe CGI is changed to correct value.

Case 2: Unbalanced path causes low handoversuccess rate[Problem description]

The incoming BSC handover success rate is quitelow for the two cells under a BTS, ranging from10% to 30%.[Cause analysis]Generally, if the data problems, such as CGIerror or intra-frequency interference, exists and ifthere is dead zones in heavy-traffic areas, or if itis hard for the MS to access the cell due to pooruplink signals, the incoming BSC handoversuccess rate is low.[Solution]

1) The cell data is found normal.2) Through checking traffic statistics items,engineers found that all incoming cellhandover success rates were low.3) Through drive test, engineers found thatfrequent handover attempts were made in thearea 2km away from the BTS, but thehandover always failed. Even if a successfulhandover was made, call drop occurredimmediately. During the handover, engineersfound that the downlink level was about -85dBm. Engineers made 10 dialing tests withfrequency locked, all the originating callsfailed. For the answering calls, they can beconnected but cannot be called out.4) It is estimated that the CDU uplinkchannel loss is great, or the jumpers areincorrectly connected at the BTS top. In thiscase, the uplink signals will be poor, whichcauses the problem.

5) After changing the CDU, engineers foundthat the incoming handover success rateincreased to 95%.

Case 3: Improper antenna planning causeslow handover success rate[Problem analysis]The handover success rate among the three cellsunder a BTS is quite low according to trafficstatistics. For the handover from cell1 to cell3and the handover from cell2 to cell3, the successrate is lower than 30%.[Cause analysis]Generally, low handover success rate is causedby board failure, handover data error, orimproper antenna planning.[Solution]

1) The BTS hardware is normal and noalarm concerning handover parameters isgenerated, so the hardware problem andparameter setting problem can be excluded.2) The BTS locates at the eastern side of asouth-north road and is 700m away from theroad. The azimuth angles of the three cells are0°, 80°and 160°. They three cells direct to thetwo directions and the open resident areaslying under a hill in the east respectively.Among the three cells, the down tilts of twocells are 7°. To make the coverage asspecified as possible, engineers concentratedthe antenna azimuth angles of the three cellsin design. In this case, however, the cells ofthe BTS are seriously overlapped in the east.

For the areas just in the west, the coverage isprovided by the side lobes and back lobes ofthe three cells. Therefore, when the MS ismoving on this road, first it is covered bycell1. When it moves to the west, the signalsof the three cells are poor and fluctuatinggreatly. In addition, since the handovermeasurement time and the handover durationis set to a small value, the handover is rathersensitive, and that's why the frequenthandover failure occurs.3) After setting the azimuth angles of thethree cells to 60°, 180°, and 350°, engineersfound that the handover success rate of amongthe three cells increased to 95%.

Case 4: Problems concerning the cooperationof different carriers' equipment cause lowoutgoing BSC handover success rate[Problem description]There is a dual-band network in which theGSM900 MHz network and the GSM1800 MHznetwork are stand-alone. After the two networkscompleted cell reselection and handoverparameter setting, engineers found that the dual-band handover success rate was low; especiallythe success rate of the handover from theGSM1800 MHz network to the GSM900 MHznetwork was low, ranging from 60% to 80%.However, the success rate of the handover fromthe GSM900 MHz network to the GSM1800MHz network was higher than 92%.[Cause analysis]

For a dual-band network, if the problemsconcerning the cooperation of different carriers'equipment are found, you must know the dataconfiguration of the equipment. For example, ifthe equipment supports Phase 2+ and EFR.[Solution]

1) Through using signaling analyzer toanalyze the message flowing across the A-interface and E-interface, engineers found thatthe MSC of the GSM1800 MHz networkwould send back a Handover Reject messageto the BSC of the GSM1800 MHz networkwhen the BSC sent a Handover Requiredmessage to the MSC.2) The MSC of the GSM1800 MHz networksent a Prepare Handover message to the MSCof the GSM900 MHz network. Uponreceiving the message, the MSC of theGSM900 MHz network sent back an Abortmessage.3) Because the success rate of the handoverfrom GSM900 MHz network to the GSM1800MHz network was high, engineers found thatthe voice version carried in the PrepareHandover message (from the GSM900 MSCto GSM1800 MSC) is half rate version 1, butthe voice versions carried in the PrepareHandover message (from GSM1800 MSC toGSM900 MSC) are full rate version 1, fullrate version 2, and half rate version 1, whichbelong to PHASE 2+. However, MSC ofprovider A does not support the PHASE 2+,so the handover failure is caused.

4) Through modifying the MSC data of thecircuit MSC data at the A-interface andselecting the full rate version 1 only, engineersfound that the voice versions carried in thePrepare Handover message (from GSM1800MSC to GSM900 MSC) are full rate 1 and 2.After that, the dual-band handover successrate was greatly increased.

5.5.5 Congestion Problems

This section introduces the methods to handleSDCCH congestion and TCH congestion, inwhich TCH congestion indicates SDCCH seizureall busy. The TCH congestion has two cases. Oneis TCH seizure all busy. For this case, the realchannels cannot be allocated to the MS, so theMS will fail to request the channels. The otherone is that the TCH assignment fails after anassignment is sent due to various reasons.

I. Congestion Problem Solutions

Congestion caused by heavy trafficYou can check if the SDCCH traffic andTCH traffic are normal through viewingtraffic statistics. If the congestion is causedby heavy traffic, the most efficient method tosolve the problem is to expand the capacity ofthe network. In addition, you can adopttraffic sharing technologies to ease thecongestion. For example, you can modify theCRO, enable direct retry or load handoverfunction. SDCCH congestion caused by burst traffic

If the SDCCH congestion rate is high and thetraffic is heavy but the TCH traffic is normal,the SDCCH congestion may be caused byburst traffic. The SDCCH congestion alwaysoccurs at BTSs along railways and tunnelexits, because the BTSs are installed inremote places and the capacity of a BTS issmall. As a result, when the train moves fastalong the railways or stops at a railwaystation, most of MSs failing to capture anetwork will perform location update, whichwill result in SDCCH congestion. In addition,when short messages are sent at aconcentrated time, the SDCCH congestionwill also occur easily. SDCCH congestioncannot be completely avoided, but somemeasures can be taken to ease the congestion.For example, you can add the number ofSDCCHs, or enable the dynamic conversionbetween SDCCH and TCH. Congestion caused by TRX problemsWhen a carrier configured in a multi-TRXcell cannot provide services, the channelcongestion will also occur. To solve theproblem, you should replace the problemTRX with a sound one. If the TRX problemcannot be positioned, you should check if theantenna feeder connection is correct and ifthe antenna standing wave is normal. If yes,recheck the TRX where the problem may begenerated. Congestion caused by interference

The interference present across the radiointerfaces will also cause congestion. In thiscase, you should solve the interferenceproblem first. Channel assignment failure caused byinconsistent coverageIf the concentric technology is not used, thetransmit power of the TRXs within the samecell will be inconsistent, which will result ininconsistent coverage. In this case, thechannel assignment failure will easily occur.To position this problem, you can check ifthe connection between the splitter andconnector and if the connection betweenCDU and SCU are correct.When a cell uses multiple transmitterantennas, inconsistent coverage will beinconsistent, which will result in channelassignment failure. To solve this problem,you should make the coverage of eachtransmitter antenna as consistent as possiblethrough engineering adjustment.In addition, if the transmitter antenna and thereceiver antenna of a cell is not in the sameplane or the antenna tilt angles areinconsistent, the channel assignment failurewill also be caused. In this case, you cancalibrate the antennas to solve the problem. Congestion caused improper dataconfigurationIf the congestion is caused by improperlocation area planning, you can reduce the

SDCCH congestion rate by planning thelocation area properly.If the congestion is caused by the problemsconcerning SDCCH dynamic allocation, youcan reduce the SDCCH congestion rate byenabling the SDCCH dynamic allocationfunction.For dual-band network, you can properly setthe parameters (such as CRO, CBA, and cellreselection hysteresis) to reduce the SDCCHcongestion rate.If the timers, such as T3101, T3103, T3107,T3122, T3212, and T3111, are not properlyset, SDCCH congestion will also be caused.

Hereunder are the solutions to the previousproblems.

You can ease the congestion caused bySDCCH dual allocation through reducing theT3101 to a smaller value. If the T3101 is setto a large value, the radio resources will beseized for a long period of time. To fully usethe radio resources, therefore, you can reducethe T3101 value.You can save the TCH resources throughreducing the T3103 and T3107 to a rationalvalue. Generally, T3103 and T3107 are set toabout 5 seconds.The T3122 must be stopped once the MSreceives an IMMEDIATE ASSIGN REJECTmessage. Only after the T3122 expires, theMS can send a new channel request message.If the MS sends channel request messages

frequently, the RACH load and CCCH loadwill increase. To solve this problem, you canincrease the T3122 to a larger value.T3212 stands for the time limit value forperiodical location update. You can ease theSDCCH load by increasing the T3212 to alarger value.T3111 is related to release latency. It is usedfor the deactivation of the latency channelsafter the major signaling link breaks. T3111can be initiated during either TCH release orSDCCH release. The value of T3111 must beconsistent with that of the T3110 at the MSside. Generally, it should be 2 seconds. If theT3111 is set to a large value, great SDCCHcongestion rate may be caused.

II. Congestion Cases

Case 1: SDCCH congestion caused by wrongLAC configuration[Description]A BTS is configured as S1/1/1. It is found thatthe SDCCH congestion rate for 2 cells reachesas high as 8%.[Problem analysis and solution]

1) Through checking the measurementindexes for TCH and SDCCH, engineersfound that the TCH traffic was not heavy. Thetraffic volume for each cell during busy hoursis lower than 2.2Erl. However, the requests forSDCCH seizure are great, reaching 3032times during busy hours. The SDCCH traffic

reaches 1.86Erl, and the congestion ratereaches 8%.2) The main reasons for SDCCH seizureinclude the messages sent before call setup,the messages sent during handover, thelocation update messages sent under the idlemode, and other short messages.3) The TCH traffic is normal, the requestsfor TCH seizure (including handover) arenormal (318 times), and the handover requestsare also normal (146 times). Therefore, theSDCCH congestion may be caused by a largenumber of location update messages or shortmessages.4) The LAC of the BTS is 0500, and theLACs of other cells of the surrounding cellsare 0520. After changing the LAC of the BTSto 0520, engineers found that the requests forSDCCH seizure during busy hours were 298,the SDCCH traffic was 0.27Erl, and thecongestion rate reduced to 0.

Case 2: SDCCH congestion caused by burstlocation updates[Problem description]The radio connected ratio of a local network islower than average level. According to trafficstatistics analysis, it is found that the SDCCHcongestion happened at several BTSs.[Problem analysis and solution]

1) Through analyzing traffic statistics,engineers found that the SDCCHs of thecongested cells were seized for 300 to 400

times during busy hours. Here the BTS wasconfigured as S1/1/1 and each cell wasconfigured with 8 SDCCHs. Therefore, theSDCCHs can be seized by 400 times, but theSDCCHs were congested for tens of timesduring busy hours.2) As far as the registered traffic statisticsitems were concerned, most of the SDCCHseizures were caused by location update.Taking the BTS location into consideration,engineers found that most of the BTSs wereinstalled at the intersections of two railways.Therefore, it might be the burst locationupdate that caused SDCCH congestion.3) To verify if it was the burst locationupdate that caused the congestion, engineersregistered the traffic statistics items in 5minutes and found that most of the locationupdate happened within the five seconds.Through querying the train time table,engineers found that there were 4 to 5 trainspassing by within the five seconds. When thetrains passed the intersections, a large numberof location updates were generated in a shorttime. In this case, the congestion was caused.

Therefore, if the BTSs are installed at the railwayintersections, you are suggested to enable theSDCCH dynamic allocation function andconfigure a suitable margin for the SDCCH.Case 3: Great TCH congestion rate caused bythe inconsistent tilt angles of two antennasunder the same cell[Problem description]

It is found that the TCH congestion rate of a cellis great (greater than 5%) according to trafficstatistics.[Problem analysis and solution]

1) Through checking BSC traffic statistics,engineers found that it was the TCH seizurefailure that directly caused the great TCHcongestion rate.2) Generally, TCH seizure failure is causedby TCH assignment failure. Throughmonitoring Abis interface, engineers foundthat most of the TCH assignment failureoccurred at the No.4 TRX and No.5 TRX, andthe probability for the assignment failure ratefor the No.4 TRX was near that for the No.5TRX.3) Through checking the antenna feederpart, engineers found that the tile anglecorresponding to the transmitter antennas ofthe two TRXs were too great (it is 10 degreesthan that of the antenna for BCCH, becausethe antenna nuts were found loosen.Therefore, if the MS is far away from theBTS, it can receive the BCCH signals butcannot receive the TCH signals. If the TCH isassigned to the MS when the MS starts a call,the TCH seizure failure will occur.4) To solve the problem, you can enable thetilt angel of the TCH antenna and that of theBCCH antenna to be consistent with eachother. In this case, the TCH congestion ratecan be reduced to 2% or lower.

Case 4: High TCH congestion rate caused bydownlink interference[Problem description]A cell of BTS is responsible for covering a largearea of sea surface along the coast. According tothe registered traffic statistics items, engineersfound that when the traffic volume was lowerthan 1Erl, the TCH congestion of the cell reached10% at some time. However, no alarm wasgenerated. All the interference bands fell withinthe interference band1 and the hardware and RFconnections were normal.[Problem analysis and solution]Because all the interference bands fell within theinterference band1, the uplink interference isimpossible. Considering that the coveragedistance reaches 60 to 70 kilometers and multiplenormal cells are present along the coast, theprobability for the channel numbers of thedownlink areas and the cell to be interfered ishigh. Through modifying the channel numbers ofthe cell, engineers found that the TCH congestionrate was improved. Through further optimizingthe channel numbers, engineers found that theTCH congestion rate was lower than 1%.Therefore, it can be judged that the 10% ofcongestion rate is caused by the downlinkinterference of some areas.

5.5.6 Other Problems

I. Subscriber Is Not in Service Area

When a subscriber is not in service area, the MSworks as the called party when the signalsreceived by the MS are good and the callingparty hears a voice saying that the subscriber isnot in the service area. If the coverage, operation,and data configuration of a network are good, theoccurrence probability for the problem must belower than 1%, otherwise it is other causes thatresult in the problem.Generally, if a subscriber is not in the servicearea, the following causes may be present:

Coverage problemIf the subscriber complaint happens at cell edges,the problem may be related to coverage. In thiscase, the probability for the calling party failureis equal to the probability for the "subscriber isnot in service area".

Parameter settingIf the subscriber is not in service area, theparameters concerning the paging, access, andimmediate assignment may be not properly set.In this case, you can check if the messagesconcerning RACH overload, PCH overload, andSDCCH overload are generated through queryingtraffic statistics and alarms. If yes, the subscribermay not in the service area.

System capacity limit or overloadIf system capacity limit or overload occur (forexample, HDB overload, CPU overload, orcapacity overload during busy hours), the systemmay fail to read the subscriber information, orsubscribers cannot access the radio network. In

this case, the subscriber may not in the servicearea. To solve this problem, you can takemeasures to expand the system capacity.

Transmission problemIf the links between systems (such as the LAPDlink of the Abis interface and the links of theeach network entity) and the links within asystem (such as the link among the modules ofBCS/MSC) are not stable, the messages sentthrough these links may be missing. In this case,subscribers may not in the service area. Toposition the problem, you can check the alarms.

Equipment causesIf the designs concerning MSC and BSC areincomplete, the probability of "subscriber is notin service area" will increase.

MS causesIf the RF parts or the software parts of the MShave problems, for example, the receptioncapability of the MS is poor; the frequencydeviation goes beyond the requirements definedin the protocols; and the dual-band performanceof the MS is poor, this problem may also occur.

II. Signal Fluctuation

Signal fluctuation indicates that change of theMS signal strength. The following factors maycause signal fluctuation.

Radio wave propagationThe strength of the signals received by the MS isthe amplitude of the sum of the vectors of variouspropagation paths. Because the propagation

environment is ever-changing, the attenuation ofradio channels is ever-changing. Therefore, evenif the MS does not move, the strength of thereceived signals will change.

Cell reselection and handoverWhen the MS moves from the old cell to the newcell through handover or cell reselection, thesignals of the MS will fluctuate because thesignal strength of the old cell and that of the newcell are inconsistent.

Antenna shakeWhen the antenna shakes, the antenna gain willchange, so the signals will also fluctuate.

Location update or channel assignmentoccurs at non-BCCH TRX

If the SDCCH is assigned to a non-BCCH TRXduring location update and power control isavailable on SDCCH, the signal strength mayfluctuate.

III. Voice Discontinuity

Voice discontinuity stands for pauses or wordsloss occurs in conversation. If the voicecontinuity is remarkable, the conversation qualitywill be affected.The following factors may cause voicediscontinuity

Frequent handoversOnly hard handover is available in GSM system.Therefore, when the MS hands over from thesource channel to the target channel, thedownlink frames may loss at the Abis interface.

As a result, the voice continuity is unavoidableduring handover. Generally, frequent handoversmay occur at cell edges or during cross coverage.In this case, the voice discontinuity will becomea headache of subscribers. To avoid the frequenthandovers, you can adjust the antenna tilt andheight and configure the data, such as uplink anddownlink quality threshold and restrictionproperly.

Radio link interferenceRadio link interference will increase the bit errorrate, which will cause voice discontinuity. Inaddition, the signals always fluctuate greatly atcell edges, so the conversation quality at the celledges is bad.

Poor network coverageIf the network coverage is poor, the Rxlev andRxqual will become poor, thus the conversationquality will be affected.

BTS transmission problemBTS transmission problems will affectconversation quality. For connectors, you shouldcheck of the connection among the connector isgood. For optical transmission, you should checkif the optical headers are clean and if thetransmission errors are great. For micro wavetransmission, it may be affected by weather. Ifthere is large amount of dust in the equipment,the conversation quality may also be affected. Ifboth micro wave transmission and opticaltransmission are used, you should pay attention

to the cooperation of the transmission impedanceat the equipment interfaces.

TRX board failureHardware problem will result in poorconversation quality. In this case, you shouldreplace the problem hardware with the soundone.