IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 2, NO. 1, MARCH 2018 1

Interference Sources in Congested Environmentsand Its Effects in UHF-RFID Systems: A Review

Josep-Ignasi Cairó , Jordi Bonache, Member, IEEE, Ferran Paredes, Member, IEEE,and Ferran Martín, Fellow, IEEE

Abstract—In scenario where radio frequency identification(RFID) readers become increasingly common in hand helddevices, the radios are prone to several interferences not onlyfrom external radio sources but also from the plurality of portabledevices that may become more common over time. For thatreason it is of interest to well understand how these radiointerferences may be influencing a UHF-RFID transceiver work-ing according to EPCglobal Class-1 Gen-2. In particular, inthis paper, the combination of interference coming from theself-radio, from other radio systems, such as mobile phone orother RFID reader, is analyzed, and such effects are combinedwith the appearance of multiple tag antennas interfering eachother. A method based on simulation using tag antenna designis presented to evaluate inter-tag interference in a variety ofcases. For a better understanding analytic examples are presentedto compute such interference interactions within the RFIDsystem.

Index Terms—UHF-RFID, mobile phone, read-range, ACPR,IIP3, tag antenna.

I. INTRODUCTION

SSYSTEMS using Radio Frequency Identification (RFID)are becoming very popular and wide spread, as the

radio technology improves, system integration into the RadioFrequency Integrated Circuit (RFIC) is increasing, antenna ismore compact, the price lowers and the performance improves.For those reasons, the possibility to integrate the radio inmobile or portable devices is feasible, and this opens thepossibility for its use in even more different situations andenvironments [2]. The number of applications where RFIDis deployed is increasing, and so the number of readertransceivers and Tag antennas.

Due to the longer read range and small antenna size, theUHF-RFID is more attractive as compared to other RFIDsystems. The UHF-RFID specifications are defined by theEPCglobal Class-1 Gen-2 [3], which implements the ISO18000-6C, the ISO 29143 air-interface protocol for mobileRFID interrogators, and ISO 18000-6A/B for operation indirect mode [4]. Details such as spurious emission limitationsare determined by the local regulations such as ETSI [5] in

Manuscript received July 7, 2017; revised January 13, 2018; acceptedFebruary 2, 2018. Date of publication February 15, 2018; date of currentversion May 11, 2018. (Corresponding author: Josep-Ignasi Cairó.)

The authors are with CIMITEC, Departament d´ Enginyeria Electrònica,Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain (e-mail:josepignasi.cairo@e-campus.uab.cat).

Digital Object Identifier 10.1109/JRFID.2018.2806738

Europe and FCC [6] in USA, which are to be considered forthe coexistence of devices.

There are studies devoted to explain the causes and effectsof interference for RFID. Some very important limitations areenvironmental caused such as type of materials close to thetags (metal, water) that highly affect local permittivity, andother caused by multiple reflections. Others not related to envi-ronment are for example self-leakage, which happens whencontinuous carrier is sent by the reader while receiving cir-cuit is activated and may desensitize the receiver. Due to itsstrong influence, some authors [7] are proposing methodolo-gies for its compensation within the own chip. Local oscillatorphase noise, is also influencing read range, since it will leak tothe receiver passing through the power amplifier and isolator,overshadowing thermal noise component from the receiver.This effect can be minimized [8]. Thanks to the short timedifference between RFID transmit and backscattering signal,the range correlation effect can further reduce such unwantedinfluence. In addition other important causes of range reduc-tion are the ones caused by other radios, in adjacent channelsor also propagating strong signals in nearby frequency bands.

From our knowledge, there is no work dealing with a com-bination of possibilities for interference issues that nowadaysRFID readers are prone to. So in this work, several sources ofinterference that may affect the RFID systems when coexistingwith various sort of other radios in the UHF band, influencingin the read range are listed and analyzed in detail. Also analyt-ical expressions for such influences are detailed, introducingnew aspects for such evaluation such as the mutual tag antennainfluence or new radio sources such as LTE (4G) that are pen-etrating in the UHF band. All of them are analyzed and someanalytic examples are presented for better understanding.

II. SELF-LEAKAGE (SELF-JAMMER) NOISE GENERATION

In this first section, a detailed analysis is done on the causesthat, several sort of either internal and external interferencemay cause to the radio system, influencing read range wheninventorying surrounding tag antennas.

A. Self-Leakage (Self-Jammer) Range Reduction

In a RFID system, since the reverse modulated signal that areader receives back from the tags is very weak [9], and due tothe continuous carrier wave that reader needs to transmit to thetag to ensure energy supply, this can be coupled through thesame PCB design or via reflections by nearby close objects

2469-7281 c© 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

2 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 2, NO. 1, MARCH 2018

Fig. 1. PCB component distribution for the UHF transmitter and mixerreceiver input from a reader based on AS3993 from AMS.

Fig. 2. Interferences caused by other RFID radios in the nearby.

into the receiver, this can lead to sensitivity problems. Thecause is the fact the weak tag modulated information will bereceived by the reader, while the receiver can be saturatedby the before mentioned transmit carrier leakage or strongreflected power. This will impose limitations in demodulatingthe signal, so it will have strong effect in the reading range.Fig. 1 shows in an RFID reader provided by AMS, the path forRF signal and the coupling paths that need to be maintained asshort as possible. Size and component orientation need to becarefully considered in addition to GND vias and groundingplanes.

Gen-2 protocol specifies that, when the TX launches com-mand information, the receiver is in non-working state; afterthe command information is sent, a continuous carrier waveis transmitted and the receiver circuit starts to work simul-taneously. For a transmit power of 0.5 W (27 dBm) and acirculator with 22 dB isolation (antenna circulator or direc-tional coupler), the carrier leakage will be 5 dBm to thereceiver input, which would saturate the receiver first stage(Ssj in Fig. 2). Such receiver desensitizing signal is known assimply carrier leakage, or leaking carrier. The interference pro-duced by the transmitter in the receiver can be one of the bigproblems in the radio performance, degrading the radio rangeconsiderably.

In case the direct current (DC) offset due to TX leakageto the receiving antenna is removed by the baseband BandPass Filter (BPF), there is also the phase noise of the TX-LOleakage on the receiver bandwidth, that can not be removed bythe filter and may have much stronger level than thermal noisepresent at receiver input, overshadowing the noise in a mannerthat the reverse link interrogation range mainly depends on thephase noise of the TX leakage (Fig. 2).

To evaluate such influence, one must include the non-linearities effects of the receiver into the self-jammer signal, sothe output noise will include such effects when evaluating thesignal quality. Let us name Aisol the isolation between the cir-culator ports and the output power of the power amplifier (PA)Pout (in our example 22 dB). The leakage signal from the out-put of the PA to mixer input, assuming a receiver with mixer asa first stage instead of typical LNA first stage, is SSJ(t) or self-jammer. The third order Input Intercept Point (IIP3) is usedto investigate the effects of the receiver non-linearities intothe SSJ(t) signal and also an interference signal SI(t) that maybe present at the receiver input as well as at an interferencefrequency fI . Describing the output of a receiver as:

ymx,out(t) = α1x(t) + α3x3(t) (1)

with α1 representing the small signal gain and α3 < 0 thenonlinear behavior of the receiver, with x(t) the input signalwith desired receive signal sRX(t) and defined as:

x(t) = sRX(t) + sSJ(t) + sI(t) = ARX · cos[(ωc ± ωRX)t]

+ ASJ · cos(ωct + θSJ) + AI · cos[(ωc ± ωI)t + θI]

(2)

it follows that after substituting (2) into (1) and removing thecomponents out of the band, rejected by the internal IC-LPF,the minimum amplitude of the output in-band signal at theoutput of the receiver ymxout (when phases θSJ = θI = 0), willbe depending upon the amplitude signals of the self-jammer(ASJ) and the amplitude of received signal (ARX), assumingASJ � AI ;

|ymxout | = α1ARX

(1 − 9

4

∣∣∣∣α3

α1

∣∣∣∣A2SJ

)(3)

So the signal-to-noise ratio at the receiver output is [7]:

SNRRX = 10log

( |Sout|2NRX

)= 10log

((α1ARX)2

)

− NRX(dB) + 20log

(1 − 9

4

∣∣∣∣α3

α1

∣∣∣∣A2SJ

)(4)

where NRX is the noise power at the receiver, and the Sout theoutput signal power. The last term shows the receiver sensitiv-ity degradation, which is directly related to the non-linearitylevel of the receiver (represented here with the third levelintercept point of the receiver IIP3), which corresponds to:

IIP3 = 10log

(4

3

∣∣∣∣α1

α3

∣∣∣∣)

(5)

The non-linear performance of the mixer as first stage willlimit the performance regarding sensitivity degradation. Forhigh IIP3, the radio will be less sensible to self-jammer.When combining expressions (4) and (5), we can obtain thatthe receiving power including the effect of self-jammer willbecome:

RRX = SNR + NRX(dB) − 20log

(1 − 3 · a2

SJ

iip3

)(6)

As can be seen in Fig. 3, the range degradation is shownfor a radio with maximum read range of 14 m, for the case

CAIRÓ et al.: INTERFERENCE SOURCES IN CONGESTED ENVIRONMENTS AND ITS EFFECTS IN UHF-RFID SYSTEMS: REVIEW 3

Fig. 3. Read range reduction caused by self-jammer effects, and improvedeffects caused by Range Correlation (upper curves).

of three different radio specifications of IIP3. For low IIP3(-10 dBm) the degradation in range can be very high evenwith small self-jammer interference.

B. Noise Power Due to Phase Noise and Interferer

In a dense reader environment, the EPCglobal Gen-2standard provides for one Adjacent Channel Power RatioACPR = 30 dBch and 60 dBch for second adjacent chan-nel. Considering a receiver with a sensitivity of PRX,min =−80 dBm and a Channel Bandwidth CBW = 200 kHz (lim-ited in EU, and 500 kHz in USA), one can find the phase noiserequirements and noise figure of the receiver. We assume amodulation index m between 0.4 and 0.6 the SNR = 11 dBfor a BER = 10−3 and PSK modulation.

1) Phase Noise (�LO): Noise power at the receiver, con-sidering the phase noise and interferer is; PN(dBc/Hz) =PI + �LO(�ω) + 10log(BW). We know also that SNR mustcomply with the condition of accepting Adjacent ChannelPower Rejection Ratio (ACPR) for the SNRmin.

PRX,min

PI · �LO(�ω) · CBW≥ SNRmin;

�LO(�ω) ≤ PRX,min + ACPR − SNRmin − 10log(CBW) (7)

in which case the phase noise needs to be -115 dBc/Hz.2) Phase Noise Power (σ 2): Such power will influence the

reading range. Evaluation of the reading range will dependon data rate that, in accordance with FM0 modulation EPCGlobal standard can be between rmax = 640 kbps and rmin =40 kbps, to comply with BER = 10−3. If reader transmit poweris PTXrd = +30 dBm, and isolation between transmitter andreceiver is Aisol = 22 dB [12].

1) A typical phase noise power spectral density from [10]is −115 dBc/Hz relative to the CW signal power at theoffset of the subcarrier frequency of 640 kHz. So relativeto the CW signal we have total LO phase noise powerat 640 kbps: N�LO = �LO + 10log[R(bps)] = −115 +10 · log(640 · 103) = −56 dBc.

2) The phase noise power of the input leaking component atthe receiver will be for a transmit power PTX = 30 dBm:Nleak = PTX−Aisol+N�LO = 30 dBm−22 dB−56 dBc =−48 dBm.

TABLE IPHASE NOISE AT THE RECEIVER

3) With a conversion factor of Cpa = 50 dB from phase toamplitude phase noise, the equivalent phase noise at thereceiver will be: σ 2 = Nleak+Cpa = −48 dBm−50 dB =−98 dBm.

In previous section similar phase noise has been also used,so one can consider the possibility to use similar results as theones shown in Fig. 3. In table I we can see different valuesfor the phase noise when having the interference signals atdifferent frequency locations for a typical Aisol + Cpa = 65.

C. Reader Range Improvement Due to the RangeCorrelation on Phase Noise

In an ideal receiver, the noise floor is caused by thermalnoise only. However due to the non perfect isolation betweenTX and RX antennas we can consider the primary performancelimiting on the uplink, the local oscillator (LO) phase noise(�LO). LO leaking from transmitter to receiver, overshadowsthe thermal noise component at the receiver input, that is theleaking power (Pleak = PTX−Aisol) a fundamental limitation interms of reader sensitivity. One can express the local oscillatorsignal with its phase noise as:

xLO(t) = ALO · cos[ωt + �LO(t)] (8)

where ALO is the amplitude of the oscillator. The LO signalamplified by the PA and passed through the circulator to thereceiver, is transmitted to the air, and is also being receivedlater as backscattered from the tag. The backscattered signalxM(t) in the PSK modulation at the receiving antenna can beexpressed as:

xM(t) = cos

[ω

(t − 2r

c

)+ �s(t) + �LO

(t − 2r

c

)](9)

where r is the tag-reader distance, and c is the wave prop-agation velocity, �s(t) denotes the phase signal representingthe binary data (“0” or “1”) of the tag. At the receiver thebackscattering signal and the LO signal are mixed, and theoutput is low-pass filtered. The resulting in-phase basebandsignal will be:

x(t) = cos

[�0 + 2r

c+ �s(t) + ��(t)

]+ n0(t) + nPN(t)

(10)

where �0 + 2rc is the constant phase shift dependent on the

tag-reader distance, ��(t) is the residual phase noise, n0(t)is the thermal noise at the receiver, and nPN(t) is the sum ofadditive phase noise of the receiver, that do not affect the tag’ssignal phase which contains information data, so they are notconsidered. The residual phase term in (10) is given by

��(t) = �LO(t) − �LO

(t − 2r

c

)(11)

4 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 2, NO. 1, MARCH 2018

When having the same LO for TX and RX, the phase noiseof the received signal is correlated with the LO, dependingon the time difference between the two signals. For smalltime difference (RFID case), the effect greatly abbreviatesthe phase noise spectrum at baseband, effect known as rangecorrelation [8]. A quantitative characterization of the relationbetween range and phase noise can be done in the frequencydomain. The PSD of ��(t), S��(f0) at an offset frequency,f0 is given by:

S��(f0) = N�LO(f0) ·∣∣∣1 − e−j2π f0

2rc

∣∣∣2

= N�LO(f0) ·[

4sin2(

2πrf0c

)](12)

So for a low offset frequency, the baseband noise spectrumwill increase proportional to the square of the distance betweentag and reader. The correlation level between the phase noiseof the received tag signal with that of the LO is inversely pro-portional to the time difference between the two signals (verysmall ∼ nsec, for the case of RFID due to short distances),and so the phase noise is reduced by correlation effect. Forthe case of having values for r and f0 of 5 m and 160 kHz,respectively, the value of rf0/c will be on the order of 10−3. Sothe range correlation effect will reduce the PSD dramatically(σ 2

RC = -128 dBm). The improved results in the reading rangecan be obtained by considering Table I results computationand Fig. 3.

III. RANGE REDUCTION CAUSED BY OTHER RADIOS

Affecting the read range in RFID systems are also the influ-ence caused by other radios in close proximity, working atnearby frequency bands or adjacent channels. It is very com-mon to share a communication channel with several radiosystems that are interconnected aiming at inter-changing infor-mation among them in a non-deterministic manner. So thereare time intervals that the radios broadcast to the medium inorder to interrogate for a possible presence of other radios, inthe same band or a band close to our transmitting-band [13].There may be the case where, a radio transmitting higherpower level, may interfere with the receiving radio so theycan incur in:

• Presence of other readers that are requested to transmithigh power, due to low tag sensitivity and may interferewith our radio.

• De-sensitivity, due to the presence of strong signal fromother radio systems (phone, Short Range Devices (SRD),Wireless Sensor Networks (WSN), etc,· · · ) at the receiverthat causes the receiver first stage to saturate and operatewrongly.

A. Mobile Phones As Interference Source

In order to expand the number of frequencies avail-able to mobile operators, the World RadiocommunicationsConference (WRC-07) allocated in 2007 in Region 1 the790-862 MHz frequency band to mobile services. This newband referred to as Digital Divident band, while allowing theadvanced cellular systems that use high speed broadband LTE,

TABLE IIFREQUENCIES AND LTE BANDS

there is a concern on potential interference to low power SRDoperating in adjacent band 863-870 MHz [15].

Considering mobile phones a source of interference, thereexist two mobile phone bands that are close to RFID-UHFsystem, such as 800 Band, that in the case of Spain, is becom-ing the 20 band (B20) for 4G LTE table II (this band hasdisappeared from TDT channels) also in use in other Europeancountries, and the 900 Band, being used by 3G GSM (3grural areas, corresponds to B8). Other bands are being usedin other countries such as LTE band 5 at 850 MHz, used inCorea and Israel and LTE Band 6 of 800 MHz being usedin LTE Japan. Due to the proximity to the UHF bands usedfor the SRD such as our RFID system, there may exist someeffects of interference, that could degrade the behavior of theradio, mostly in the upper adjacent band, because of highlevel of Out-Of-Band (OOB) emissions produced by very widebandwidth LTE signals Fig. 4.

Fig. 4 shows in wide dark how the output spectrum maskof a UE transmitter is divided on three components: occupiedor channel bandwidth (with 99% of the total integrated meanpower), out of band (OOB) and the out of spurious emissiondomain [19]. The figure shows the in-band and out-of-bandpower. A possible case for interference evaluation would betransmitting maximum power (23 dBm) in 16QAM, on all50 resource blocks within a 10 MHz channel centered on857 MHz [11].

With the LTE upmost 10 MHz channel (852 − 862 MHz)radio as interfere, from the spectrum mask it is obtainedinterference power of PI = −23 dBm (RBW = 100 kHz) at theRFID band. For a SNR of 11 dB in the RFID (BER = 10−3),we can obtain from Friis equation:

C

I + N= Srd

PIGrd

(λ

4πr

)2

(13)

In this case for the RFID radio with -80 dBm sensitivity,and reader antenna gain of 6 dBi, the distance where the LTEradio should be located in order not to interfere with the RFIDsystem is minimum 62 m (considering direct line-of-sight). Sofor LTE (4G) radios closer than 62 m from the RFID reader,the effect would be noticeable in tag reading degradation.

In the work done by [14], there are some test performedintroducing a RFID system at 868 MHz which is interferedby a GSM phone working at 880.2 MHz. The GSM radiolocated at 1 m from the RFID reader transmits at 2 W, andchanges its transmit power from 33 dBm (2 W) maximumlevel down to 5 dBm. Table III shows how the radiated fieldby the GSM signal reduces the reading range.

B. Interference Between Readers

In a scenario where the number of RFID readers starts toincrease, it can exist a situation where their interaction and

CAIRÓ et al.: INTERFERENCE SOURCES IN CONGESTED ENVIRONMENTS AND ITS EFFECTS IN UHF-RFID SYSTEMS: REVIEW 5

Fig. 4. LTE spectrum mask and example of radiation power.

TABLE IIIREAD RANGE REDUCTION DUE TO GSM RADIO

mutual interference may become a real problem. In such cases,it is necessary to investigate the influence they may have onthe back-scattering from the tag.

In the case of RFID readers close to each other, there isthe possibility of interference when a reader transmits a com-mand signal that interferes with the tag reception of anotherreader. Thus signals transmitted from distant readers maybe sufficiently strong to impede accurate decoding of theback-scattered signals from adjacent tags. Let us imagine areader transmitting 1 W output power with antenna gain 6 dBiand a reader located at a distance “r” receiving in a 1 dBiantenna (out of main lobe). In case both are tuned to thesame channel, the power received at 5 m distance would bePRX = −3.5 dBm, enough to block the receiver.

For the situation of being in a multiple reader environ-ment, using the EPC Global Gen.2 dense interrogation mode,with the reader transmitting in the adjacent channel, it wouldtransmit with 1% of the total energy of the principal chan-nel (20 dB down), so at 10 m distance, the received powerby the reader at central channel will be; PI

RX = −9.5 dBm−20 dB = −29.5 dBm, sufficient power such as also to blocktags at 1 or 2 m distance.

IV. RANGE REDUCTION CAUSED BY NEARBY TAGS

One of the largest disadvantages in RFID is the low tagidentification efficiency due to multiple tags being in the prox-imity of the reader. Two causes are identified, one being thefact the multiple tags placed in a shared radiansphere willabsorb individually less radiated power than would do onesingle tag. Another is the tag collision because various tagsoccupy the same interrogation area and channel. In this event,packets may have to be transmitted and retransmitted untileventually they are correctly received. Since passive tags can-not detect collisions it is necessary a kind of anti-collisionprotocol that enables the recognition of tags with some col-lisions, but since frame sizes in the protocol are limited, theframed slotted ALOHA algorithm is limited, and also whenthe number of tags is large, the number of slots required toread tags increases exponentially as the number of tags does.

Some methods are proposed, such as [20], in order to try solv-ing such problem, but always including complexity into thesystem.

A deeper analysis on tag mutual effects is proposed in thisanalysis to better understand the physical causes for tag reduc-tion in backscattering signal, so reducing reading range. Itis considered here the case of having several tag antennasclose to each other, and evaluate their mutual effects, in termsof power influence. Electromagnetic coupling will affect eachother antennas modifying backscattering radiation pattern, sothe matching between antennas and active parts will be influ-enced and the radiation pattern from the antennas will bemodified. In real situations such as retail, large number oftag antennas may be present in close proximity when manyitems or sensors are to be tagged (i.e., shopping basket). Insuch circumstances collective scattering modulation in certaintype of tags may lead to confusion or failures in anti-collisionprocedure in the protocol.

From the optimal design of power antenna transfer in thetag, with matching conditions between the antenna and thetag IC, there can be a deviation from the optimal impedancedesign, so a mismatch will cause a decrease in sensitiv-ity and read range. When dipole antennas are close, mutualimpedance arises at each antenna. In multiple arrayed tagsthere is also antenna interference that causes a reduction inthe radar cross section [21]–[23]. To experimentally visual-ize such effect, one can take measurements by examining thetwo port S-parameters matrix S presenting the power wavereflection coefficient S with the associated power reflectioncoefficient |S|2 Fig. 5:

S = ZL − Z∗a

ZL + Zac =

∣∣∣∣ZL − Z∗a

ZL + Za

∣∣∣∣2

(14)

with ZL as the load being seen by the antenna, Za the antennaimpedance and, |S|2 represents the fraction of the maximumpower available from the generator that is not delivered into theload. For an easy interpretation of mutual impedance effectsbetween the antennas, it is recommended converting to Zparameters, where the mutual effects between tag antennaswill be represented by the measurement of the mutual com-ponents Z12 and Z21, and the equivalent antenna impedance.Considering mutual effects from other antennas, the equiva-lent antenna impedance of a tag when the two tags are closeis expressed as [24]:

Zin = 50

[Z11 − Z12Z21

Z22 + ZIC

]= 50

[Z2

11 + Z11ZIC − Z212

Z11 + ZIC

]

(15)

For the two tags with similar geometry and characteristicsZ will be symmetrical (that is to say, Z11 = Z22, Z12 = Z21).In this case ZIC is the impedance presented by the Tag IC thatshould correspond to ZL.

To show this effect, a tag dipole antenna has beendesigned using an electromagnetic 2.5D solver from Agilent(Momentum). It matches an Impinj Monza 5 IC (Cp =0.825 pF, Rp = 1.8 k and Cin = 0.245 pF).

6 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 2, NO. 1, MARCH 2018

Fig. 5. Effects on reader range of close proximity tags. Also shown RFID-UHF tag design.

Fig. 6. Simulation results for the influence over tag impedance for thehorizontal and vertical close proximity of tags.

The effects on the influence on mutual tags, has been ana-lyzed based on close proximity in the horizontal and verticalplane axis (Fig. 6). In such cases looking at the impedancepresented at the tag antenna terminals, there is a clear decreaseof the impedance by placing both tags together, so at theextreme of 5 mm distance, the impedance shift is from theoriginal Zi = 16 + j133, down to Zf = 9.6 + j9.

The tags influence can be understood as a shadowing effecton the read range. One effect that could be represented byincluding a factor into the Friis equation, such as a GainPenalty (Gp):

PRtag = PTrd Grd

(λ

4πd

)2

GtagLsysGP (16)

here Lsys are the losses in the system and Grd and Gtag are thereader and tag antenna gains. On the other hand, such gainpenalty factor, can be analyzed evaluating the influence thatsuch mismatch produced by proximity tags, would have on thevoltage induced in the tag. Considering the modulation of theincident electromagnetic field in the antenna, by the tag-IC,

Fig. 7. BER degradation considering two tags close vertically 20 mm.

such effect is presented in the back-scattering signal affectedby nearby tags.

The signal to noise ratio at the receiver is proportional tothe energy of the bit so, to the distance between bits in theconstellation diagram. According to dmin = 2V0, will deter-mine the minimum detection threshold to resolve between allthe received bits. Assuming ideal, matched-filter demodulationand Additive White Gaussian Noise (AWGN) with standarddeviation σ at the detector input, the BER for both PSK andASK (OOK) will be determined based on such energy dis-tance between received bits. The effect on the back-scatteringof the tag antennas, can be then introduced by the reflectioncoefficient between load and antenna impedance affected byproximity tags.

The proximity tags affects the impedance values from thedifferent modulation states, Z1 = R1 + jX1 and Z2 = R2 + jX2to Z1’ and Z2’. Since for both ASK and PSK, it is possibleto express the BER depending on the modulation index (m)as; m = |ρ1−ρ2|

2 , being ρ1,2 the reflection coefficient at thetag antenna terminals [25], [26], it will be possible to includesuch proximity tag effects into the BER expression.

BER = 1

2erfc

( |VR1 − VR2|4√

2σ

)

= 1

2erfc

( |V0| · |ρ1 − ρ2|4√

2σ

)= 1

2erfc

( |V0| · m

2√

2σ

)(17)

In ASK modulation, the two possible states, will dependon the modulation depth, which will be conditioned also bythe mismatch between antenna and modulation impedances;|VR1 − VR2| = |V0| · |ρ1 − ρ2| = |V0|2m. Taking as examplethe case where two tags are located at a distance of 20 mm oneon top of the other, the simulation results from the mutual cou-pling, shows that m = 0.54. The results of BER degradationdue to imperfect matching between antennas shows a degra-dation of SNR of 3 dB, due to only this effect (Fig. 7). Suchmodulation effect, will degrade reading range.

V. CONCLUSION

This paper has addressed several important issues to be con-sidered in early stages when designing real RFID systems,mainly related to the high density concentration of wire-less radio sources in the environment. The main sources ofinterferences for RFID have been explained in detail in order

CAIRÓ et al.: INTERFERENCE SOURCES IN CONGESTED ENVIRONMENTS AND ITS EFFECTS IN UHF-RFID SYSTEMS: REVIEW 7

to describe some of the major causes that may degrade theread range of RFID readers. In conclusion, such effects mustbe taken into consideration at early stages of the design phasein order to minimize them and provide high performance RFIDsystems.

REFERENCES

[1] A. E. Abdulhadi and R. Abhari, “Multiport UHF RFID-tag antennafor enhanced energy harvesting of self-powered wireless sensor,” IEEETrans. Ind. Informat., vol. 12, no. 2, pp. 801–808, Apr. 2016.

[2] Z. Rashid, E. Peig, and R. Pous, “Bringing online shopping experienceto offline retail through augmented reality and RFID,” in Proc. 5th Int.Conf. Internet Things (IoT), Seoul, South Korea, Oct. 2015, pp. 45–51.

[3] EPC Radio-Frequency Identification Protocols Class-1 Generation-2UHF RFID Protocol for Communications at 860 MHz—960 MHz,Version 1.2.0, EPCglobal Inc., Brussels, Belgium, Oct. 2008.

[4] ISO/IEC 18000-63:2015, Information Technology—Radio FrequencyIdentification for Item Management—Part 63:Parameters for AirInterface Communications at 860 MHz to 960 MHz Type C, ANSI,Lawrenceville, NJ, USA, Oct. 2015.

[5] Electromagnetic Compatibility and Radio Spectrum Matters (ERM);Radio Frequency Identification Equipment Operating in the Band865 MHz to 868 MHz With Power Levels Up to 2 W; Part 1: TechnicalRequirements and Methods of Measurement, ETSI Standard EN 302208: (V1.1.2), Jun. 2014.

[6] 47 CFR 15.245—Operation Within the Bands 902-928 MHz, 2435-2465 MHz, 5785-5815 MHz, 10500-10550 MHz, and 24075-24175 MHz,Title 47, FCC, Washington, DC, USA, May 2017.

[7] A. Ghahremani, V. D. Razaei, and M. S. Bakhtiar, “A UHF-RFIDtransceiver with a blocker-canceller feedback and +30 dBm outputpower,” IEEE Trans. Circuits Syst. I, Reg. Papers, vol. 60, no. 11,pp. 3043–3054, Nov. 2013.

[8] A. D. Droitcour, O. Boric-Lubecke, V. M. Lubecke, J. Lin, andG. T. A. Kovacs, “Range correlation and I/Q performance benefitsin single-chip silicon Doppler radars for noncontact cardiopulmonarymonitoring,” IEEE Trans. Microw. Theory Techn., vol. 52, no. 3,pp. 838–848, Mar. 2004.

[9] C.-C. Yen, A. E. Gutierrez, D. Veeramani, and D. van der Weide, “Radarcross-section analysis of backscattering RFID tags,” IEEE AntennasWireless Propag. Lett., vol. 6, pp. 279–281, 2007.

[10] D. M. Dobkin, The RF in RFID. Passive UHF RFID in Practice.Amsterdam, The Netherlands: Elsevier, 2008.

[11] LTE User Equipment Coexistence With 862—870 MHz, Research doc-ument, Ofcom, London, U.K., 2012.

[12] R. Chakraborty, S. Roy, and V. Jandhyala, “Revisiting RFID link budgetsfor technology scaling: Range maximization of RFID tags,” IEEE Trans.Microw. Theory Techn., vol. 59, no. 2, pp. 496–503, Feb. 2011.

[13] S. Pourbagheri, M. S. Bakhtiar, and M. Atarodi, “Cellular design fora dense RFID reader environment,” in Proc. IEEE Asia–Pac. Conf.Circuits Syst. (APCCAS), Nov./Dec. 2008, pp. 1124–1127.

[14] D. Arnaud-Cormos, T. Letertre, A. Diet, and A. Azoulay,“Electromagnetic environment of RFID systems,” in Proc. IEEE37th Eur. Microw. Conf. (EuMA), Munich, Germany, Oct. 2007,pp. 1652–1655.

[15] M66 05R1 SE24 LTE Unwanted Emissions vs. Adjacent 863-870 MHzBand, document M66 CEPT ECC SE24 Meeting, CEPT, Ahmedabad,India, 2012.

[16] AS3993 UHF Single Chip Reader EPC Class 1 Gen2 Compatible ENv1, Datasheet AS3993, AMS, Unterpremstätten, Austria, Mar. 2014.

[17] S. Woo and H. Kim, “An empirical interference modeling for linkreliability assessment in wireless networks,” IEEE/ACM Trans. Netw.,vol. 21, no. 1, pp. 272–285, Feb. 2013.

[18] D. Gokhale, S. Sen, K. Chebrolu, and B. Raman, “On the feasibility ofthe link abstraction in (rural) mesh networks,” in Proc. IEEE INFOCOM,Phoenix, AZ, USA, Apr. 2008, pp. 61–65.

[19] LTE; Evolved Universal Terrestrial Radio Access (E-UTRA); UserEquipment (UE) Radio Transmission and Reception, ETSI Standard TS136 101, V10.7.0, 2012.

[20] F. Zheng and T. Kaiser, “Adaptive Aloha anti-collision algorithms forRFID systems,” EURASIP J. Embedded Syst., vol. 2016, pp. 1–14,Apr. 2016.

[21] P. V. Nikitin et al., “Power reflection coefficient analysis for compleximpedances in RFID tag design,” IEEE Trans. Microw. Theory Techn.,vol. 53, no. 9, pp. 2721–2725, Sep. 2005.

[22] P. V. Nikitin and K. V. S. Rao, “Effect of Gen2 protocol parameters onRFID tag performance,” in Proc. IEEE Int. Conf. RFID, Orlando, FL,USA, Apr. 2009, pp. 117–122.

[23] Y. Tanaka, Y. Umeda, O. Takyu, M. Nakayama, and K. Kodama,“Change of read range for UHF passive RFID tags in close proxim-ity,” in Proc. IEEE Int. Conf. RFID, Orlando, FL, USA, Apr. 2009,pp. 338–345.

[24] H. Yohima et al., “Analysis of read range for UHF passive RFID tags inclose proximity with dynamic impedance measurement of tag ICs,” inProc. IEEE Radio Wireless Symp. (RWS), Phoenix, AZ, USA, Jan. 2001,pp. 110–113.

[25] U. Karthaus and M. Fisher, “Fully integrated passive UHF RFIDtransponder IC with 16.7-μW minimum RF input power,” IEEE J.Solid-State Circuits, vol. 38, no. 10, pp. 1602–1608, Oct. 2003.

[26] F. Fuschini, C. Piersanti, F. Paolazzi, and G. Falciasecca, “On theefficiency of load modulation in RFID systems operating in real envi-ronment,” IEEE Antennas Wireless Propag. Lett., vol. 7, pp. 243–246,2008.

Josep-Ignasi Cairó was born in Terrassa(Barcelona), Spain, in 1967. He received theelectronics and telecommunication degrees from thePolitechnical University of Catalonia, Barcelona,Spain, in 1988 and 1992, respectively. He joinedthe Research and Development Department fromthe Group Circutor in 1995, where he wasinvolved in industrial power electronics and RFIDsystems. From 1999 to 2003, he joined PhilipsSemiconductors, Nijmegen, The Netherlands. Hedeveloped GSM power amplifiers using BiCMOS

process and, later, in San Jose, CA, USA, he engaged in the developmentof the first integrated WLAN transceiver, which was awarded by a prizein 2001. Since 2003, he has been with Epson Electronics GmbH, Barcelona,leading the Research and Development Laboratory in RF projects, such asantennas, CMOS device characterization, transceiver and system design, fromnarrow band to ultrawide band. He joined the Institute for Energy Researchof Catalonia, as a Senior Researcher and a Project Leader devoted to energyefficiency, microgrids and smart grid technologies, and communications andrenewable integration activities in 2010. He is currently involved in wirelesscommunications and RFID systems. He holds 27 patents and a long list ofpublications in conferences and relevant journals.

Jordi Bonache (S’05–M’07) was born in Barcelona,Spain, in 1976. He received the physics and electron-ics engineering degrees and the Ph.D. degree in elec-tronics engineering from the Universitat Autònomade Barcelona (UAB), Barcelona, in 1999, 2001,and 2007, respectively. In 2000, he joined theHigh Energy Physics Institute, Barcelona, where hewas involved in the design and implementation ofthe control and monitoring system of the MAGICtelescope. In 2001, he joined the Department ofElectronics Engineering, UAB, where he is currently

an Associate Professor. From 2006 to 2009, he was an Executive Manager withCIMITEC, Drachten, The Netherlands, where he currently leads the researchin RFID and antennas. His current research interests include active and passivemicrowave devices, metamaterials, antennas, and RFID.

8 IEEE JOURNAL OF RADIO FREQUENCY IDENTIFICATION, VOL. 2, NO. 1, MARCH 2018

Ferran Paredes was born in Badalona (Barcelona),Spain, in 1983. He received the telecommunicationsengineering Diploma (specializing in electronics)and the telecommunications engineering degrees,and the Ph.D. degree in electronics engineering fromthe Universitat Autònoma de Barcelona in 2004,2006, and 2012, respectively, where he was anAssistant Professor from 2006 to 2008 and is cur-rently a Research Assistant. His research interestsinclude metamaterial concepts, passive microwavesdevices, antennas, and RFID.

Ferran Martín (M’04–SM’08–F’12) was born inBarakaldo, Spain, in 1965. He received the B.S.degree in physics and the Ph.D. degree fromthe Universitat Autnoma de Barcelona (UAB),Barcelona, Spain, in 1988 and 1992, respec-tively. From 1994 to 2006, he was an AssociateProfessor in electronics with the Departmentd’Enginyeria Electrònica, UAB, where he has beena Full Professor of electronics since 2007. In recentyears, he has been involved in different researchactivities, including modeling and simulation of

electron devices for high frequency applications, millimeter wave and THzgeneration systems, and the application of electromagnetic bandgaps tomicrowave and millimeter wave circuits. He is currently very active inthe field of metamaterials and their application to the miniaturization andoptimization of microwave circuits and antennas. He is the Head of theMicrowave Engineering, Metamaterials, and Antennas Group (GEMMAGroup), UAB, and the Director of CIMITEC, a Research Center onMetamaterials supported by TECNIO (Generalitat de Catalunya). He hasorganized several international events related to metamaterials, includingWorkshops at the IEEE International Microwave Symposium in 2005 and2007 and European Microwave Conference in 2009, and the Fifth InternationalCongress on Advanced Electromagnetic Materials in Microwaves and Optics(Metamaterials) in 2011, where he has acted as the Chair of the LocalOrganizing Committee. He has acted as a Guest Editor for three SpecialIssues on Metamaterials in three International Journals. He has authoredand co-authored over 500 technical conference, letters, journal papers, andbook chapters. He has co-authored the book on Metamaterials entitledMetamaterials With Negative Parameters: Theory, Design, and MicrowaveApplications (Wiley, 2008) and he has authored the book entitled ArtificialTransmission Lines for RF and Microwave Applications (Wiley, 2015). Hehas generated 16 Ph.D. students. He has filed several patents on metama-terials and has headed several Development Contracts. He was a recipi-ent of the 2006 Duran Farell Prize for Technological Research and twoICREA ACADEMIA Awards in 2008 and 2013. He holds the Parc deRecerca UAB Santander Technology Transfer Chair. He is a Reviewer ofthe IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES andthe IEEE MICROWAVE AND WIRELESS COMPONENTS LETTERS, amongmany other journals. He is a member of the IEEE Microwave Theoryand Techniques Society. He serves as an Editorial Board Member of IETMicrowaves, Antennas and Propagation and the International Journal ofRF and Microwave Computer-Aided Engineering. He is also a memberof the Technical Committees of the European Microwave Conference andInternational Congress on Advanced Electromagnetic Materials in Microwavesand Optics (Metamaterials). He has been a fellow of the IET since 2016.