13
BROADCASTING DIVISION Application Note Fading Channel Simulation in DVB Products: TV Test Transmitter SFQ Option Fading Simulator SFQ-B11 7BM05_1E

Fading Channel Simulation in DVB - Rohde & Schwarzcdn.rohde-schwarz.com/pws/dl_downloads/dl_application/application... · Fading, Channel Simulation in DVB 1 Introduction Fading is

  • Upload
    others

  • View
    42

  • Download
    0

Embed Size (px)

Citation preview

BROADCASTING DIVISION

Application Note

FadingChannel Simulation in DVB

Products:

TV Test Transmitter SFQOption Fading Simulator SFQ-B11

7BM05_1E

BROADCASTING DIVISION

2

Contents

Fading, Channel Simulation in DVB

1 Introduction ..................................................................................... 3

2 Basic Elements of Fading ..................................................... 32.1 CONSTANT PHASE ..................................................... 32.2 PURE DOPPLER .......................................................................... 42.3 RICE Fading ................................................................ 52.4 RAYLEIGH Fading .......................................................................... 72.5 LOG NORMAL Fading ................................................................ 72.6 Gaussian Channel ........................................................................... 8

3. Channel Simulation with SFQ Fading Simulator Option B11 ..... 93.1 Definition of Real Conditions ..................................................... 93.2 RF Levels of Fading Profiles ..................................................... 103.3 VALIDATE 100 .......................................................................... 103.4 EASY 3 EASY, 3 km/h ................................................................ 103.5 REGULAR TU50 REGULAR REDUCED TYPICAL URBAN, 50 km/h 113.6 RED HT100 REDUCED HILLY TERRAIN, 100 km/h ........... 113.7 RED6 DVB-T REDUCED DVB-T ANNEX B, 6 PATHS ........... 113.8 ET 50 EQUALIZATION TEST, 50 km/h ................................ 123.9 DIFFICULT RA250 DIFFICULT RURAL AREA, 250 km/h ........... 12

4 Outlook ..................................................................................... 13

BROADCASTING DIVISION

3

Fading, Channel Simulation in DVB

1 IntroductionFading is known from shortwave transmission,where the received field strength level maystrongly vary due to atmospheric disturbances. Inanalog TV, the term "fading" is practicallyunknown. Rather, one talks of "antenna shadows"or "ghost images". The effect in question,however, is fading, ie constant reflection of theeletromagnetic waves emitted by the TVtransmitter by walls of buildings, mountain slopesand similar reflecting natural or artificialobstacles. In analog TV, fading is of minorimportance since the effects thereof can beeliminated almost completely through thedirectivity and exact orientation of the Yagi roofantenna for stationary TV reception at home.

Fading effects can also be observed in analogcable TV, for example in a block of flats linked tothe cable network with one or several antennasockets in every flat. If the taps for the socketsare not match-terminated, reflections withconstant level and constant phase arise whichmay cause level reductions of several dB atexactly calculable points in the cable.

Moreover, the reception of TV signals broadcastvia satellite can be impaired by fading. A knownphenomenon is flickering of the received picture,produced by planes flying past or a drop inreceive field strength caused by an approachingthunderstorm.All the above receive conditions have one thingin common: reception is stationary with a directline of sight to the TV transmitter.

Looking at receive conditions in DVB, the effectsin cable and satellite reception (DVB-C andDVB-S) are found to be similar as in analogreception. In these two modes reception isstationary, too.Terrestrial transmission (DVB-T) not onlyprovides for stationary operation but also forportable and mobile reception. This considerablyaccentuates the effects of fading.

In this application note we investigate fadingeffects in DVB, with the emphasis on those inDVB-T signals.

2 Basic Elements of Fading

2.1 CONSTANT PHASE

The first basic element of fading is reflection.Reflections or echoes occur at all obstacles in thepropagation path of waves. Echoes are describedby the reflected level and the phase shift causedby reflection. The level and phase shift dependon the reflecting material.Example:Reflection of a wave at a plane metallic surfacewill produce no level loss but a phase shift of180°.Reflection at the wall of a building, however, willproduce a large level loss and an undefinedphase shift.

Fig. 1 Echoes

The determining parameters are:- path loss (dB),- delay relative to direct path (ns),- phase angle (deg).

This type of fading is referred to as "constantphase".Looking at a single carrier of the COFDM signal,the effects on the receive signal can be seen onlyfrom the resulting amplitude. Examining the fullspectrum of a DVB signal, however, revealsvalleys in the spectrum depending on the numberof paths with path loss, delay and phase.

BROADCASTING DIVISION

4

Fig. 2 Constant phase

The blue line in Fig. 2 shows the originalspectrum without fading, whereas the yellow linerepresents the spectrum for CONSTANT PHASEwith a second path. The parameter values are asfollows:

Path 1 Path 2Path loss 2 dB 6 dBDelay 0 ns 300 nsPhase 0° 70°

The valleys in the spectrum occur at intervals of∆f = 3.3 MHz, which corresponds to thereciprocal of the path delay (∆f = 1/ delay).The depth L in dB of the valleys is obtained asfollows:

+−−∗=

−−

)110log()110log(20L 20

|2L1L|

20

|L2L1|

dB

(1)where L = depth of valleys

L1 = loss of path 1L2 = loss of path 2L1 ≠ L2 . If L1 = L2, the depth L = ∞.

With more than 2 paths, calculation of the depthis very complex.Example with 2 paths:

L1 = 2 dB, loss of path 1L2 = 6 dB, loss of path 2L = 12.91 dB

The reflection phase determines wherecancellations occur in the spectrum.An adaptive equalizer (also referred to aschannel estimation) in the receiver compensatesfor spectral valleys within its dynamic range toensure undisturbed demodulation even formultipath reception.

The carrier phase further determines the rotationof the demodulated constellation diagram. Theconstellation diagram of path 2 is rotated 70°compared with that of path 1. This rotation, too,should be compensated by the channelestimation.

2.2 PURE DOPPLER

The second basic element is the Doppler effect,ie the frequency shift resulting from themovement of the receiver relative to the site of aDVB-T transmitter.

Fig. 3 Doppler effect

The frequency shift ∆fD caused by the Dopplereffect is obtained by means of the followingformula:

∆f v fcDOPPLER = ∗ ∗cos( )ϕ (2)

wherev = vehicle velocityf = carrier frequency of transmitterc = speed of light (300 000 km/s)ϕ = angle between direction of motion and line of sight to transmitter 0 < ϕ < π

There are three cases:

1. The vehicle is moving towards the transmitter:ϕ = 0°, ie cos ϕ = 1:the transmit frequency increases by ∆fD.

2. The vehicle is moving away from thetransmitter:ϕ = 180°, ie cos ϕ = -1:the transmit frequency decreases by ∆fD.

Transmitter

ϕ

ϕ

f

v

f

fdfd∆ ∆

without Fading

with FadingConstant Phase

BROADCASTING DIVISION

5

3. The vehicle is driving around the transmitter incircles: ϕ = 90°, ie cos ϕ = 0:the transmit frequency remains unchanged,ie ∆fD = 0.

Fig. 4 Case with constant ϕ = 90°

If the vehicle moves towards the transmitter froma long distance, passes the transmitter and thenmoves away from it, the carrier frequency f isDoppler-shifted through the range f - ∆fD ≤ f ≤ f +∆fD, as illustrated by Fig. 3. The shift from f to f +∆fD takes place when the vehicle is movingtowards the transmitter.

From (2), the following can be seen:The Doppler shift ∆fDoppler is directly proportionalto the carrier frequency f of the transmitter. Thelower the frequency f, the smaller the Dopplereffect in DVB-T, since the OFDM carrier offsetremains constant. This is to be demonstrated bytwo examples:

The vehicle is driving towards the transmitter at aconstant velocity of v = 140 km/h.

Example 1:Carrier frequency f = 50.5 MHz (first DVB-Tchannel in band I VHF, 7 MHz bandwidth)

v = 140 km/h= 140 / 3.6 = 38.9 m/s

ϕ = 0°The Doppler shift in this case is:

∆f DOPPLER 38.9 50.5

300 6.54817 Hz= ∗

∗∗ =

106

106 1

Despite the high velocity, the frequency shift isvery small relative to the carrier offset of 1116 Hzin the 8k mode.

Example 2Carrier frequency f = 858 MHz (currently lastDVB-T channel in band V UHF, 8 MHzbandwidth)

v = 140 km/h= 140 / 3.6 = 38.9 m/s

ϕ = 0°Here the Doppler shift is:

∆f DOPPLER 38.9 858 106

300 106 1 111.254 Hz= ∗∗

∗∗ =

In this case a significant shift is obtained, whichis however of little importance in PUREDOPPLER as the frequency shift in the DVB-Tchannel is

995.0f858854f DopplerDoppler ∗∆=∗∆

at the lower end of the channel and

005.1f858862f DopplerDoppler ∗∆=∗∆

at the higher end of the channel, and the receiverPLL should have no problems handling such asmall offset by means of the channel estimation.

2.3 RICE Fading

The third basic element is Rice fading.Rice fading is caused by Doppler-shifted echoeswith a Gaussian distribution, but in addition thereis always a direct path from the Tx antenna to the

Rx antenna. Accordingly, figures 1 and 5illustrate Rice fading.

Fig. 5 Rice fading

1

2

3

Transmitter

ϕ = 90 °

f

f

BROADCASTING DIVISION

6

The direct path 1 between the antennasconsiderably boosts the received field strengthlevel. This path can in mobile reception only beinfluenced by the Doppler effect. In addition tothe direct path, many echoes are received. Sincethe echoes arrive from different directions, theangle ϕ for calculation of the Doppler shift is notconstant in mobile reception. So the spectrum ofRice fading is obtained:

The angle with the lowest probability is at ϕ =90°, because a vehicle will only in rare casesmove around a transmitter in circles.Consequently, the lowest level is found at thefrequency f.

High levels are obtained at frequencies f ± ∆fD

since the vehicle will with great probability movetowards or away from the transmitter for a longtime with ϕ = 0° or ϕ = 180° both at large andshort distances from the transmitter. Betweenthese two angles, the level distribution is in theform of a bell-shaped curve. It results fromDoppler-shifted echoes produced by randomreflection at the surrounding buildings, mountainsand other natural and artificial obstacles. Thelevel curve represents the sum of all theseechoes.

The same spectrum is obtained if the vehiclemoves around the transmitter and changes its

direction at random. The spectral bandwidthresults from the maximum Doppler shift of thesingle echoes, the Rice peak from the Dopplershift in the direct path. The maximum Dopplershift of the echoes and the shift of the Rice peakare not correlated to each other, since the anglesϕ of the receive paths hardly ever coincide.The following spectrum is obtained:

Fig. 6 Rice spectrum of single carrier

Depending on the direction of movement relativeto the transmitter, the level peak is shiftedtowards higher or lower frequencies as a result ofthe Doppler effect. In Fig. 6, the level peak is at afrequency above f, ie the vehicle moves towardsthe transmitter but with a transverse component,because the shift is only about ∆fD/3. The angleof direction ϕ is calculated as follows:

ϕ = arc cos (0.333) = 70°

The parameters relevant for Rice fading are:Parameters of Doppler effectEcho parameters path loss and delayAdditional parameters for direct receive path (DISCRETE COMPONENT):

Power ratio (dB), determines the height of the power peakFrequency ratio, determines the frequency shift relative to ∆fD

LOG NORMAL may be activated in addition(see 2.5, LOG NORMAL Fading).

The received power as a function of time showslevel dips that result from the superposition of allechoes with different levels and phases and fromthe direct path at the receive antenna.

Fig. 7 Rice fadingLevel versus time

The level dips have a maximum depth of 25 dB,which is not very large, because the direct pathalways makes a major contribution to thereceived power.

2 x ∆ fD

Power peakin Rice fading

BROADCASTING DIVISION

7

2.4 RAYLEIGH Fading

The fourth basic element is Rayleigh fading.Rayleigh fading, same as Rice fading, is causedby Doppler-shifted echoes with a Gaussiandistribution, but there exists no direct path fromthe Tx antenna to the Rx antenna.

Fig. 8 Rayleigh fading

The parameters relevant for Rayleigh fading are:Parameters of Doppler effectEcho parameters path loss and delay

Here, too, LOG NORMAL may be activated inaddition (see 2.5, LOG NORMAL Fading).The received power as a function of time showslevel dips produced in this case only by thesuperposition of all echoes with different levelsand phases at the receive antenna.

Fig. 9 Rayleigh spectrum of single carrier

Fig. 10 Rayleigh fadingLevel versus time

The level dips have a depth of 60 dB and morebecause the continuous power componentcontributed by the direct path in Rice fading ismissing here.

2.5 LOG NORMAL Fading

Another standard basic element is log normalfading. The term is derived from "normaldistribution of a logarithmic value". This valueexpresses the signal level variation in dB.

Log normal describes slow changes in the fadingpath. The fading profiles discussed above are, incontrast to log normal, "fast" profiles, where levelchanges take place after distances as short asthe wavelength of the received signal, ie in theorder of milliseconds. Variations in log normalfading are caused by the vehicle moving indifferent environments, for instance in built-upurban areas, in flat, open terrain, or hilly terrain.The distances over which the relevant parametervalues change are correspondingly large.

The relevant parameters for log normal fadingare:

Local constant in meters:Defines the distance over which fadingconditions do not change.In strongly structured areas, the local constant will assume low values(eg 50 m in street canyons in cities), whereas on broad plains without

vegetationhigh values (>300 m) are to be expected.

Standard deviation:Describes the interval (±1 σ) within which66% of all level changes occur (withGaussian distribution).

2 3

1

BROADCASTING DIVISION

8

The variations of fading conditions as a functionof time further depend on the vehicle velocity.For a tour in the city with v = 50 km/h = 50 / 3.6m/s and a local constant of 50 m, a time constantof

τ = 50 / (50/3.6) = 3.6 swill be obtained, whereas in an open plain withv = 120 km/h = 120 / 3.6 m/s and a local constantof 300 m, the time constant will be

τ = 300 / (120/3.6) = 9 sAccordingly, the time axis for the received powerwill be scaled in seconds and not in millisecondsas with Rice fading and Rayleigh fading.

2xSTD DEV

Fig. 11 Log normalLevel versus time

Fig. 11 shows log normal fading with:Local constant 200 mStandard deviation 5 dBSignal level -5 dBm

The time axis is scaled 2 s/div. Level changesare slow compared with Rice or Rayleigh fading.In mobile reception, ie in a car, bus or train, lognormal fading is encountered, since this type offading always involves the reception of Doppler-shifted echoes (Rayleigh fading), sometimes alsowith a direct line of sight to the transmitter (Ricefading).Log normal fading with Rayleigh fading is alsoreferred to as Suzuki fading.

2.6 Gaussian Channel

Noise is an important quality parameter in anykind of signal transmission. Whenever fading isdescribed, therefore, the Gaussian channel isincluded. It contains none of the above-describedelements for channel simulation. The quality ofthe transmission channel is determined only bywhite noise superimposed on the signal, thisbeing referred to as C/N ratio.

In measurements for determining the quality of atransmission path, compliance with the bit errorratio BER ≤ 2∗10-4 according to Viterbi is appliedas limit criterion. If this limit is not attained bymeans of the predefined profiles of channelsimulation, additive white noise is superimposedon the signal to reach the BER limit value.Fading profile parameters are important factorsin channel simulation, but also the Gaussianchannel, which is defined by the C/N ratio.

BROADCASTING DIVISION

9

3. Channel Simulation with SFQFading Simulator Option B11

3.1 Definition of Real Conditions

The SFQ Fading Simulator option simulates allthe above-described simple fading profiles. Realprofiles however result from more than a singlepath; they are the sum of at least 5 to 10 singleprofiles. In GSM (mobile radio), profiles with upto 20 single profiles are defined. The most well-known are the COST 207 profiles with up to20 fading paths. For DVB-T, however, it hasshown that 6 paths are absolutely sufficient tosimulate fading conditions in all of the threereceive modes:-

stationary: reception with fixed, directionalYagi antenna with defined gain,

portable: reception with rod antenna or "rabbit ear" antenna set up at afixed position (eg park bench)and

mobile: reception with rod antenna in amoving vehicle, eg a car, bus

ortrain.

The optional SFQ Fading Simulator B11,therefore, is capable of generating fading profileswith up to 6 paths. For applications requiringmore paths, a second option B11 can be installedin SFQ for the generation of up to 12 paths.In the MOTIVATE and VALIDATE workinggroups for DVB-T, some profiles have beenrecommended to cater for special receiveconditions:

Fig. 12 Fading profiles

Up to 5 of n proposed fading profiles can bestored in SFQ under FADING PARAMETER.With currently 10 profiles developed by theworking groups, n = 10 is valid at present. Theproposed fading profiles can be called directlyfrom the SETUP menu MODULATION / FADING/ PARAMETERSET.

Example:

EASY 3USER REGULAR TU50RED HT100VALIDATE100DIFFICULT RA250

The settings valid for the individual profiles arebriefly explained under"PRESET A FADING PARAMETER SET WITHSTANDARD PARAMETERS":

1 EASY 3:EASY, 3 km/h

2 REGULAR TU50:REGULAR REDUCED,TYPICAL URBAN, 50 km/h

3 RED HT100:REDUCED HILLY TERRAIN, 100 km/h

4 VALIDATE100:VALIDATE RECOMMENDATION,100 km/h

5 RED6 DVB-T:REDUCED DVB-T ANNEX B, 6 PATHS

6 ET 50:EQUALIZATION TEST, 50 km/h

7 DIFFICULT RA250:DIFFICULT, RURAL AREA, 250 km/h

If only 6 paths are available in SFQ for fadingsimulation, the profiles including up to 12 pathsare displayed in italics:

8 RED12 DVB-T:REDUCED DVB-T ANNEX B, 12 PATHS

9 TU3 12 PATHS:TYPICAL URBAN, 3 km/h, 12 PATHS

10 TU50 12 PATHS:TYPICAL URBAN, 50 km/h, 12 PATHS

SETUP

HARDWARE INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

FADING PARAMETER PRESET A FADING PARAMETER SET WITH STANDARD PARAMETERS

EASY 3USER REGULAR TU50RED HT100VALIDATE100DIFFICULT RA250

EASY 3 EASY, 3 km/h REGULAR TU50 REGULAR REDUCED, TYPICAL URBAN, 50km/hRED HT100 REDUCED HILLY TERRAIN, 100 km/hRED6 DVB-T REDUCED DVB-T ANNEX B, 6 PATHS VALIDATE100 VALIDATE RECOMMEDATION, 100 km/h ET 50 EQUALIZATION TEST, 50 km/hDIFFICULT RA250 DIFFICULT, RURAL AREA, 250 km/hRED12 DVB-T REDUCED DVB-T ANNEX B, 12 PATHS TU3 12 PATHS TYPICAL URBAN, 3km/h, 12 PATHSTU50 12 PATHS TYPICAL URBAN, 50km/h, 12 PATHS

ACTIVE SET: RED HT 100 / OFF

RED HT100

DIFFICULT RA250 DIFFICULT, RURAL AREA, 250 km/h

BROADCASTING DIVISION

10

The single parameters of the profiles aredisplayed on opening the predefined fadingprofiles. TV Test Transmitter SFQ allows allsingle parameters to be optimized for themeasurements to be performed.

3.2 RF Levels of Fading Profiles

Each path of a fading profile contributes towardsthe total RF output power of SFQ. The levelshould however remain constant to simplifyfading measurements. If the level C at the inputof a receiver changes, the C/N ratio also changesbecause the noise N remains constant. Thismeans that there are no constant conditions atthe input of the DVB-T receiver. Constantconditions are necessary however for obtainingcomparable results in measurements withdifferent fading conditions.SFQ therefore takes into account the levels of allpaths defined by the PATH LOSS parameter, andcorrects the sum level to yield a constant outputlevel, which is displayed.Special cases are the sum levels of profiles withCONSTANT PHASE paths and PURE DOPPLERpaths with identical Doppler shift. Here thenominal value is displayed although the sumlevel may deviate from this value because of thefixed phase relationships in the paths.Despite this, the displayed sum level is correct inaccordance with DVB-T specifications, whichdefine the following: the level with multipathreception is the sum level of the individual paths,the effects of CONSTANT PHASE and PUREDOPPLER being left out of account. TV TestTransmitter SFQ simulates the real conditions ofmobile reception, where the level of C changes,resulting from the varying superposition of pathsas the receiver is moving, while the noise Nremains constant. Still, the receiver must be ableto demodulate such signals correctly. Themovement of the receiver is simulated by varyingthe phase with CONSTANT PHASE and thespeed with PURE DOPPLER. The current state isfrozen and the new receive conditions areevaluated.For such profiles, the level and thus the C/N ratiofor the simulated current position of the receivershould be measured with a spectrum analyzerwith frequency markers, or with thermal powermeters at the output of SFQ with the noiseswitched off.The displayed sum value of C according tospecifications is obtained for the above profilesby measuring the level of C at many pointslocated close together and forming the averageof the results.

For CONSTANT PHASE, for example, the valueof C displayed on TV Test Transmitter SFQ canbe checked as follows:measurement of C at various phases (eg atincrements of 5°) and calculation of averagevalue. The measured average agrees with thedisplayed value of C.

3.3 VALIDATE 100

Fig. 13 VALIDATE 100 profile

This fading profile, which was proposed by theVALIDATE working group, simulates mobilereception with the vehicle moving in hilly terrainat 100 km/h without a direct line of sight to thetransmitter in an MFN (multifrequency network)or towards a transmitter in an SFN (single-frequency network). Six Rayleigh paths aresufficient to simulate the receive conditions. Lognormal is switched off, which can be seen whenscrolling down with the Page Down key F4.

Fig. 14 VALIDATE 100 profile LOG NORMALOFF

3.4 EASY 3EASY, 3 km/h

This profile was defined by the participants in theMOTIVATE group and is derived from the DVB-TMobile Profile (SFN) developed by this group.Only two paths with PURE DOPPLER are active.LOG NORMAL is switched off. The user shouldadapt the vehicle speed, the echo delay and the

SETUPHARDWARE

INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: VALIDATE 100 / ON

VALIDATE 100

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASELOG NORMAL LOCAL CONSTANT STD DEVIATION

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONRAYLEIGH2.8 dB0.00µs28.7 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.05µs28.7 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH3.8 dB0.40µs28.7 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.1 dB 1.45 µs28.7 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH2.6 dB2.30µs28.7 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH1.3 dB2.80µs28.7 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

F2=ALL EQUAL F3=PG UP

SETUPHARDWARE

INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: VALIDATE100 / OFF

VALIDATE100

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONP.DOPPLER0.0 dB0.00µs0.8 m/s1.7 Hz0.0 DEGOFF0.0 dB-1.0

ONP.DOPPLER3.0 dB28.00 µs0.8 m/s1.7 Hz0.0 DEGOFF0.0 dB1.0

OFFRAYLEIGH3.8 dB0.40µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

OFFRAYLEIGH0.1 dB 1.45 µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

OFFRAYLEIGH2.6 dB2.30µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

OFFRAYLEIGH1.3 dB2.80µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

F2=ALL EQUAL F4=PG DOWN

BROADCASTING DIVISION

11

level of the second path to match hismeasurement conditions.MOTIVATE has in part defined these values:

Speed Not definedDelay

guard21 T∗

Level 0 dBTable 1

Fig. 15 EASY 3 profile

This profile did not place any special demands onthe receiver during measurements. It allowshowever conclusions to be drawn as to "normal"operation in areas of average structure. Extremeconditions are not covered.

3.5 REGULAR TU50REGULAR REDUCED TYPICAL URBAN, 50 km/h

Fig. 16 REGULAR TU 50 profile

This profile was defined by COST 207 tosimulate typical terrestrial wave propagation in anurban environment. The originally 12 paths werereduced to 6. The echo delays in the paths covera wide range, and the echo levels are relativelyhigh. The vehicle velocity corresponds to the50 km/h allowed in built-up areas. This profile,which places average demands on thedemodulator, is already used in DAB and GSM.Experience has shown that, if receivers canhandle this profile without problems, reception inan urban environment in general is (almost)ensured.

3.6 RED HT100:REDUCED HILLY TERRAIN, 100 km/h

Fig. 17 REDUCED HILLY TERRAIN 100 profile

This profile too was defined by COST 207. It hasonly 6 paths, and simulates conditions for avehicle moving at 100 km/h in hilly terrain. Inaddition to shortterm echoes with low loss, thereare two paths with long reflection times and lowerlevels. All paths have Rayleigh characteristic,and the LOG NORMAL function is switched off.

3.7 RED6 DVB-TREDUCED DVB-T ANNEX B, 6 PATHS

In Annex B of the EN 300 744 DVB-T standard, afading profile for stationary reception with20 paths is described. This profile was adoptedby COST 207. A selection of 6 of the 20 pathsyields the REDUCED DVB-T profile.

Fig. 18 REDUCED DVB-T (6 paths) profile

This profile contains the 6 most important pathsof the 20-path profile defined in EN 300 744 sothat there is no major difference with respect tothe original profile.

All of the six paths are defined as CONSTANTPHASE with average path loss between 3.9 dBand 5.8 dB. The reflection angles are distributedover a wide range, simulating omnidirectionalreception. The resulting spectrum at the site ofreception is shown by Fig. 19.

SETUP

HARDWARE INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: EASY 3 / OFF

EASY 3

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONP.DOPPLER0.0 dB0.00µs0.8 m/s1.7 Hz0.0 DEGOFF0.0 dB-1.0

ONP.DOPPLER3.0 dB28.00 µs0.8 m/s1.7 Hz0.0 DEGOFF0.0 dB1.0

OFFRAYLEIGH3.8 dB0.40µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

OFFRAYLEIGH0.1 dB 1.45 µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

OFFRAYLEIGH2.6 dB2.30µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

OFFRAYLEIGH1.3 dB2.80µs28.7 m/s58.0 Hz0.0 DEGOFF0.0 dB0.3

F2=ALL EQUAL F4=PG DOWN

SETUPHARDWARE

INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: VALIDATE100 / OFF

REGULAR TU50

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONRAYLEIGH3.0 dB0.00µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.20µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH2.0 dB0.50µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH6.0 dB 1.60 µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH8.0 dB2.30µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH10.0 dB5.00µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

F2=ALL EQUAL F4=PG DOWN

SETUP

HARDWARE INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: RED HT100 / OFF

RED HT 100

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONRAYLEIGH0.0 dB0.00µs27.8 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH1.5 dB0.10µs27.8 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH4.5 dB0.30µs27.8 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH7.5 dB 0.50 µs27.8 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH8.0 dB15.00µs27.8 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH17.7 dB17.20µs27.8 m/s58.0 Hz0.0 DEGOFF200 m0.0 dB

F2=ALL EQUAL F4=PG DOWN

SETUPHARDWARE

INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: RED6 DVB-T / OFF

RED6 DVB-T

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONC.PHASE3.9 dB0.50 µs27.8 m/s58.0 Hz336.0 DEGOFF200 m0.0 dB

ONC.PHASE4.0 dB1.95 µs27.8 m/s58.0 Hz8.9 DEGOFF200 m0.0 dB

ONC.PHASE4.5 dB3.25 µs27.8 m/s58.0 Hz175.0 DEGOFF200 m0.0 dB

ONC.PHASE5.2 dB 2.75 µs27.8 m/s58.0 Hz127.0 DEGOFF200 m0.0 dB

ONC.PHASE5.3 dB0.45 µs27.8 m/s58.0 Hz340.0 DEGOFF200 m0.0 dB

ONC.PHASE5.8 dB0.85 µs27.8 m/s58.0 Hz36.0 DEGOFF200 m0.0 dB

F2=ALL EQUAL F4=PG DOWN

BROADCASTING DIVISION

12

Fig. 19 Spectrum ofREDUCED DVB-T ANNEX B, 6 PATHS

A static fading profile is obtained, which iscompensated by channel correction in thereceiver. So, there is only the requirement ofsufficient dynamic range, ie ∆ ≈ 15 dB, of theequalizer, but no requirement as to the speed ofequalization in dB/s.

This profile has only CONSTANT PHASE paths.It therefore describes, among other things, thetypical case of constant reflections in cables inthe event of mismatch. An important application,therefore, is simulation of cabling in DVB-C. Thisprofile makes it easy to test set top boxes forDVB-C in the real environment of a domesticcabling system with numerous antenna socketson each floor, each of which is more or lessdistorting the characteristic impedance of 75 Ω.In this case, 6 faulty sockets are simulated by the6 paths. The reflection losses and associateddelays represent the change of characteristicimpedance and the spatial separation of theantenna outlets.

3.8 ET 50EQUALIZATION TEST, 50 km/h

This test profile too was defined by COST 207and adopted for DVB-T.

Fig. 20 ET 50 profile

The profile simulates reception in Rayleigh modewith 0 dB echoes with a delay of t = n ∗ 3.2 µs,with n = 0 to 5 and at 50 km/h driving speed. At acarrier frequency of 626 MHz, for example, thiscorresponds to a Doppler shift of 29 Hz. Thechannel estimation in the DVB-T receiver shouldbe able to handle this profile without anyproblems.

3.9 DIFFICULT RA250DIFFICULT RURAL AREA, 250 km/hFig. 21

DIFFICULT RURAL AREA, 250 km/h profile

This profile is bound to place great demands onchannel correction because of the high velocity.At a carrier frequency of 858 MHz (UHF channelwith the currently highest carrier frequency), thisvelocity results in a Doppler shift of 197.8 Hz.Channel correction is facilitated by the RICE pathwith 0 dB level loss at 0 µs, and a power ratio ofthe Rice level peak to the Doppler-shiftedRayleigh paths of 6.5 dB at a frequency ratio of1. This means that the Rice peak is at the upperend of a Doppler-shifted Rayleigh path, but isvisible in the spectrum only in a single-carriermeasurement.

Looking at the spectrum of all COFDM carriers,the Rice peak will not be visible since the

SETUPHARDWARE

INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: ET 50 / OFF

ET 50

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONRAYLEIGH0.0 dB0.00 µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB3.20 µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB6.40 µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB 9.60 µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB12.80 µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB16.00 µs13.9 m/s29.0 Hz0.0 DEGOFF200 m0.0 dB

F2=ALL EQUAL F4=PG DOWN

SETUPHARDWARE

INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: DIFFICULT RA 250 / OFF

DIFFICULT RA 250

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONRICE0.0 dB0.00 µs69.4 m/s198.7 Hz0.0 DEGON6.6 dB1.0

ONRAYLEIGH0.0 dB0.10 µs69.4 m/s198.7 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.20 µs69.4 m/s198.7 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB 0.30 µs69.4 m/s198.7 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.40 µs69.4 m/s198.7 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.40 µs69.4 m/s198.7 Hz0.0 DEGOFF200 m0.0 dB

F2=ALL EQUAL F4=PG DOWN

BROADCASTING DIVISION

13

resolution bandwidth and the span of thespectrum analyzer do not allow this.

Fig. 22 Spectrum of DIFFICULT RURAL AREA250 km/h fading profile

The effect of Rice and Rayleigh fading can beseen clearly. The spectrum appears to be verynoisy compared with the same spectrum withoutfading.

Fig. 23 DVB-T 8k UHF 8 MHz spectrumwithout fading

The DIFFICULT RURAL AREA, 250 km/h profileconstitutes an extreme case of fading. Mobilereception at a velocity as high as 250 km/h,which corresponds to 69.4 m/s, is probablypossible only in high-speed trains such as theICE or TGV, whereas the simulated conditionswill hardly ever be fulfilled by cars on motorways.At a carrier offset of ∆f = 1/224∗10-6 = 4 464 Hzin 2k mode with QPSK modulation or possiblyeven with 16 QAM and a code rate of 2/3, a QEFtransport stream with this profile could just bedemodulated. At higher rates and with 64 QAM,

however, demodulation will be errored to aconsiderable extent.8k mode with a carrier offset of∆f = 1/896∗10-6 = 1116 Hz is certainly notappropriate for QEF reception with this profile.

The profile can easily be adapted for testingmobile reception in a car at a velocity ofapprox. 130 km/h common today on motorways.To this effect, the DIFFICULT RURAL AREA,250 km/h profile is opened and the velocitychanged to 130 km/h. The new Doppler shift ∆fD

for the selected carrier frequency is automaticallycalculated and displayed.The profile designation is changed from

DIFFICULT RURAL AREA, 250 km/hto

USER DIFFICULT RURAL AREA, 250 km/h.The actual values of the fading parameters arestated in the SETUP menu.The instrument is immediately ready formeasurement after this modification.

Fig. 24USER DIFFICULT RURAL AREA, 250 km/hprofile

4 Outlook

Fading profiles are needed to test receivers forimmunity to interference encountered instationary, portable and mobile reception.Proposals to this effect have been developed byvarious working groups, and their current versionfully integrated in TV Test Transmitter SFQ. Tosuit a given application, the fading profiles caneasily be modified on SFQ directly in the profiledescription. SFQ has been designed to cope withall future settings and changes. Any furtherchannel simulation profiles for importantmeasurements developed by the working groupswill be added to the SFQ profile list.Specifications laid down by the known standardcommittees of ETSI, ITU or DVB are howeverstill outstanding.

SETUP

HARDWARE INFOFIRMWARE

TIME/DATE CLOCK

COMMUNI- CATION PRESET

CHANNEL TABLE

FADINGPARAMETER SERVICE

ACTIVE SET: USER DIFFICULT RA 250 / ON

USER DIFFICULT RA 250

PATH STATEPROFILEPATH LOSS DELAYSPEEDDOPPLER FREQUENCYPHASEDISCRETE COMPONENT POWER RATIO FREQUENCY RATIO

PATH 1 PATH 2 PATH 3 PATH 4 PATH 5 PATH 6

ONRICE0.0 dB0.00 µs130 km/h103.3 Hz0.0 DEGON6.6 dB1.0

ONRAYLEIGH0.0 dB0.10 µs130 km/h103.3 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.20 µs130 km/h103.3 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB 0.30 µs130 km/h103.3 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.40 µs130 km/h103.3 Hz0.0 DEGOFF200 m0.0 dB

ONRAYLEIGH0.0 dB0.40 µs130 km/h103.3 Hz0.0 DEGOFF200 m0.0 dB

F2=ALL EQUAL F4=PG DOWN