Hindawi Publishing CorporationJournal of Electrical and Computer EngineeringVolume 2013 Article ID 402514 9 pageshttpdxdoiorg1011552013402514

Research ArticleRadiation Mitigation for Power Line Communications UsingTime Reversal

Amilcar Mescco1 Pascal Pagani2 Michel Ney2 and Ahmed Zeddam1

1 Orange Labs Networks 2 Avenue Pierre Marzin 22200 Lannion France2 Telecom Bretagne Microwave Department Technopole Brest-Iroise 29238 Brest France

Correspondence should be addressed to Amilcar Mescco amesccogmailcom

Received 4 August 2012 Revised 16 January 2013 Accepted 21 January 2013

Academic Editor Moises V Ribeiro

Copyright copy 2013 Amilcar Mescco et alThis is an open access article distributed under theCreativeCommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Power Line Communication (PLC) is the response for nowadays high demand of multimedia services in domestic environmentnot only for its fast and reliable transfer characteristics but also for its flexible low cost implementation since the PLC technologyuses the existing electrical network infrastructure and the omnipresent outlets throughout the homeThe transfer of such a high bitrate through the mains network generates acceptable radiated emission regulated by international standards but the increment inspeed for new generation PLCmay cause higher levels of emissionsThis paper explains the use of the TimeReversal (TR) techniquetomitigate radiated emissions from PLC systemsThismethod was probed experimentally in real electrical networks with excellentresults in 40 of the observations the Electromagnetic Interference (EMI) generated by PLC transmission could be reduced bymore than 3 dB and this EMI mitigation factor could increase to more than 10 dB in particular configurations

1 Introduction

With the increasing demand for both high data rate appli-cations and reliable links for command and control systemsPower Line Communication (PLC) has emerged in recentyears as an attractive communication technique [1]Themainadvantage of this technology is its ability to benefit from theexisting electrical network infrastructure for the transmissionof electromagnetic signals Hence it becomes possible tobuild large communication networks without the need forinstalling new wires

In the home or office environment indoor PLC usesthe Low Voltage (LV) infrastructure The presence of severalelectrical outlets in each room of the house allows ubiquitouscoverage of the communication network In addition therelatively short distance between different outlets allows thesystem to operate within a limited attenuation Current BroadBand (BB) in-home systems primarily operate in the fre-quency range from 2MHz to 30MHz However recent spec-ifications such as IEEE 1901 [2] or ITU-T G9960 [3] allowsignal transmission at higher frequencies up to 100MHz Onthe other hand Narrow Band (NB) PLC systems are deployed

at frequencies below 500 kHz in both indoor and out-door configurations using LV or Medium Voltage (MV)infrastructures [4] These systems allow the transmission ofcommand and control information over longer distances forSmart Grid applications The ITU-T G9955 standard is anexample of such NB PLC systems [5]

LV or MV electrical wires were not initially designed topropagate communication signals at frequencies above 1 kHzAs a consequence the communication channel between thetransmitter and the receiver is a difficult channel generatingattenuation and multiple propagation paths The channelcapacity is hence limited and signal processing needs to beoptimized so as tomaximize the offered data rate andQualityof Service (QoS)

This paper focuses on one of the main limitations relatedto the PLC technology namely the generation of uninten-tional radiated signal This phenomenon is mainly due tothe unbalanced nature of the electrical network [6] Thevariation of the impedances of the loads connected to thenetwork as well as the unequal length of the phase andneutral wires (due to single-phase switches) converts thedifferential PLC signal into common mode current flowing

2 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35 40Frequency (MHz)

PSD

lim

it (a

vera

ge) (

dBm

Hz)

minus95

minus90

minus85

minus80

minus75

minus70

minus65

minus60

minus55

minus50

Figure 1 Example of PSD limit (average) for North America

through the network Consequently the copper wires usedfor transmitting the useful signal act as an antenna and partof the transmitted power is radiated This not only results instronger signal attenuation at the receiver but also leads toElectromagnetic Compatibility (EMC) issues as the radiatedsignal may interfere with other existing services such asamateur radio (HAM) or Short Wave (SW) broadcastingThe impact of PLC transmission on EMC has been studiedfor example within the ICT FP7 project OMEGA [7] andthrough the ETSI Specialist Task Force 410 [8]

In order to avoid interference between PLC systems andother users of the spectrum regulation authorities imposestrict emissionmasks for the transmission of electromagneticsignals on the power lines In the USA the Federal Commu-nicationsCommission (FCC)Part 15 [9] specifies amaximumlevel of radiated field for carrier-current systems (includingPLC) leading system specifications to define constrainedpower transmission masks Figure 1 represents an example ofPower Spectral Density (PSD) limits provided in the IEEE1901 standard for North America The observed notches aredefined to protect specific systems such as HAM bandsIn Europe CENELEC is currently developing a draft onregulation standard applying to in-home PLC systems [10]

Regardless of the regulation limits in place the researchpresented in this paper focuses on the mitigation of unin-tentional radiation due to PLC systems Several attempts tosolve this problem have been presented in the literatureReference [11] presents amethod to reduce radiated emissionsby applying an auxiliary signal cancelling the electromagneticfield on a given point in space Simulations demonstratedgood performance with the drawback that the ElectroMagnetic Interference (EMI) could only be mitigated at asingle location The authors in [12] used additional hardwareconnected at the wall outlets in order to reduce asymmetrieson power lines at the cost of an increased complexity of thePLC network

In our approach we tried to simultaneously reach twocomplementary goals by the means of digital signal pro-cessing First we intend to focus the transmitted signal atthe receiver location The power gain linked to this energyfocalization allows in turn relaxing the required power level

at the transmitter hence generating less Electro MagneticInterference (EMI) Second we target to reduce the level ofenergy dissipated at any location except the intended receiverIn particular it is desirable to minimize the level of radiatedpower from the electrical wires These two benefits alreadyappear as features of a known technique in the field ofwireless transmission Time Reversal (TR) [13] Experimentalinvestigations conducted using Ultra Wide Band (UWB)radio waves demonstrated both the focusing and interferencemitigation properties of this technique [14 15]

In this paper we present for the first time an experimentalanalysis of TR as a mean to mitigate radiation effects forwired signal transmission Our investigation focuses onHighFrequency (HF) BB PLC signals but could be extended to NBPLC and other wired transmission systems such as DigitalSubscriber Line (DSL) access The paper is organized asfollows Section 2 presents the concepts of the TR techniqueand its application to wired systems Section 3 details theexperiment conducted to assess the merits of TR for BB PLCand the results are statistically analyzed in Section 4 Finallyconclusions are drawn in Section 5

2 Time Reversal for PowerLine Communications

21 Time Reversal for Wireless Transmission The TR tech-nique also known as phase conjugation in the frequencydomain was first used in the fields of acoustics [16 17]More recently this concept has been successfully extendedto electromagnetic waves where the rich multipath channelprovides excellent conditions for its application [13] Thebasic concept of TR is simple Let 120575(120591) be an ideal Diracimpulse emitted by a transmitter (Tx) antenna (Figure 2(a))By definition at any receiver (Rx) location r

0 the received

signal is given by the Channel Impulse Response (CIR)ℎ(120591 r0) The CIR is composed of multiple echoes reflectingthe multiple propagation paths of the propagation channel

TR uses this Channel State Information (CSI) at the Txto prefilter the signal to be transmitted More specificallythe CIR ℎ(120591 r0) is time reversed and normalized to serve asan input filter for the signal to be transmitted (Figure 2(b))Physically each delayed echo constituting theTRfilter travelsamong other multiple paths through its original propagationpath As a result themultiple echoes sumup coherently at thereceiver hence focusing the received energy in time

Mathematically applying TR leads for any Rx situatedat an arbitrary location r to the equivalent perceived CIRℎTR(120591 r) (note that this formulation holds for a real valuedCIR) [18]

ℎTR (120591 r) =ℎ (minus120591 r

0)

radicint1003816100381610038161003816ℎ (120591 r0)

10038161003816100381610038162119889120591

otimes ℎ (120591 r) (1)

where the symbol otimes denotes time domain convolutionFormulating (1) in the frequency domain leads to

119867TR (119891 r) =119867lowast(119891 r0)

radicint1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162119889119891

times 119867 (119891 r) (2)

Journal of Electrical and Computer Engineering 3

120575(120591)

Tx Rx

ℎ(120591 r0)

(a)

Tx Rx

ℎ(minus120591 r0) ℎ(120591 r0)⨂ℎ(minus120591 r0)

(b)

Figure 2 Transmission over an ideal multipath propagation channel (a) of a Dirac impulse (b) of the time-reversed CIR

where119867(119891 r) represents the complex valued Channel Trans-fer Function (CTF) and the superscript lowast denotes the com-plex conjugate operation For this reason TR is sometimescalled frequency domain phase conjugation

Two conclusions can be drawn from (1) and (2) First atthe intended Rx location r

0 the perceived CIR simplifies to

ℎTR (120591 r0) =1

radicint1003816100381610038161003816ℎ (120591 r0)

10038161003816100381610038162119889120591

times 119877ℎ(120591 r0) (3)

where 119877ℎ(120591 r0) denotes the time domain autocorrelation of

the function ℎ(120591 r0) Similarly the perceived CTF simplifies

to

119867TR (119891 r0) =1

radicint1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162119889119891

times1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162 (4)

In the time domain the effect of the TR filter is to trans-form the CIR into its autocorrelation For a rich multipathenvironment the autocorrelation of the CIR presents a largepeak at 120591 = 0 with reduced side echoes Experimentalstudies demonstrated that the resulting channel is less spreadin time [14 15] hence reducing the possible IntersymbolInterference (ISI) In the frequency domain the perceivedCTF is proportional to the square of the magnitude of theactual CTF Besides the fact that TR provides a real valuedCTF which could be exploited at the receiver this also leadsto a significant gain in terms of Rx power due to a betterexploitation of the frequency selective nature of the channelIt was demonstrated in [18] that the application of TR in a flatchannel (ie without frequency domain power decay) underRayleigh fading leads to a gain of 3 dB in the total receivedpower This gain was increased to 5 dB when considering thefrequency domain power decay observed in practical UWBradio channels

The second conclusion drawn from (1) and (2) is that forany other location r different from r

0 TR creates a mismatch

between the Tx filter and the channel This is particularlyobservable in the frequency domain representation of TRgiven in (2) The perceived CTF corresponds to the productof two independent CTF119867(119891 r) and119867lowast(119891 r

0) with possibly

very different frequency fading structures More preciselyminima of the first CTF can happen randomly at maxima ofthe second CTF Hence averaging over all frequencies thetotal received power at untargeted locations is reduced Inwireless TR analysis this effect is called spatial focusing andis generally assessed as the ratio between the maximum ofℎTR(120591 r) and the maximum of ℎTR(120591 r0) for a given distancerminus r0 Spatial focusing factors of minus10 dB have been reported

in [18 19]

22 Extension of Time Reversal to Wired Transmission Asobserved through the study of TR for wireless transmissionthe TR scheme provides two main features namely anincrease of the Rx power at the intended Rx location and adecrease of the Rx power at any other locationThese featuresare highly desirable in the context of wired transmissionwhere the level of Tx power is constrained by the uninten-tional radiation from the wires causing possible EMI to othersystems

Based on this observation we conducted experimentalstudies to analyze the potential of TR to mitigate unwantedemissions for PLC systems The main principles of theextension of TR to PLC transmission can be explained withthe help of Figure 3

We assume an intended transmission between a Tx PLCmodem and an Rx PLCmodem over an LV indoor electricalnetwork Different experimental investigations reported thePLC channel as a rich multipath propagation channel due tothe multiple branches present in a classical electrical networkand to the impedance mismatch occurring at the networkterminations (outlets) and nodes [20ndash23] This similarity ofthe PLC channel with wireless channels suggests promisingresults when applying TR to PLC

With reference to Figure 3 the Rx modem is situated atthe intended location r

0 and the Tx modem is situated at the

origin By applying TR filtering at the Tx the Rx power willbe increased at location r

0 hence the Rx modem will benefit

from an increased Signal-to-Noise Ratio (SNR) This powerincreased can also in turn be applied as a reduction of the Txpower to achieve similar performance At other outlets in thenetwork situated for example at locations r

1or r2 the Rx

power will be reduced This effect can be further exploited inthe design of multiuser transmission schemes

For our purpose of radiation mitigation let us nowconsider a location r

3 situated at any point in space in the

vicinity of the electrical network The level of radiated fieldat this location can be evaluated for instance by the meansof an equivalent transfer function 119867(119891 r

3) between the Tx

modem and an ideal antenna situated at location r3 By the

virtue of the TR scheme (2) the perceived transfer functionat location r

3after applying TR will be proportional to the

product119867lowast(119891 r0) times 119867(119891 r3) As the functions119867lowast(119891 r0) and119867(119891 r

3) are not correlated their frequency fading structures

are different In particular the deep notches due to frequencyselective fading do not appear at the same frequencies Asa result the product 119867lowast(119891 r

0) times 119867(119891 r

3) will provide more

average attenuation when compared to 119867(119891 r3) alone and

therefore the total power radiated at location r3will be

reduced A similar observation is made in previous studies

4 Journal of Electrical and Computer Engineering

Tx Rx r0

r1

r2

r3

Figure 3 Principle of the extension of TR to wired transmission

Signal generator Oscilloscope

Biconical antenna

Balun1 Balun2

Mains

Tx Rx

119904(119905) 119910(119905)

ℎ(120591 r0)

ℎ(120591 r3)

Position r0

Position r3

Figure 4 Equipment used in the experimental setup

dedicated to wireless channels [14 15 18] Hence TR appearsas an efficient method to mitigate EMI for wired communi-cations

The theoretical basis for the application of TR to wiredtransmission being set wewill nowdescribe the experimentalassessment of this method in the next Sections

3 Experimental Setup

31 Equipment In order to experimentally assess the use ofTR as a method to mitigate EMI for wired communicationwe used the experimental setup presented in Figure 4 In thissetup a signal generator Tektronix AWG7082C was used as ageneric Tx A Digital Sampling Oscilloscope (DSO) LeCroyWaveRunner 715Zi-A was used to sample the received signalat Rx The signal generated at Tx is denoted 119904(119905) and thesignal received at the DSO is denoted 119910(119905) Two UniversalPLC Couplers were used as baluns to couple the Tx and Rxsignals with the power lines The Tx coupler is situated at theorigin and the Rx coupler is connected to a plug at locationr0 These couplers were developed within the ETSI Specialist

Task Force 410 [24] In order to measure the power densityof the radiated emission received at any arbitrary location r

3

a biconical antenna Schwarzbeck EFS921 was connected to asecond port of the DSO

Mains

Balun1 Balun2

Signal generatorOscilloscope

Coaxial 1

Coaxial 2

Coaxial 3

Coaxial 4

Biconical antennaAttenuator

Amplifier

Figure 5 Experimental setup calibration

32 Calibration With reference to Figure 5 two measure-ment paths can be distinguished Path 1 serves for themeasurement of the CTF and is composed of the followingelements the transmitter coaxial cable 1 a 30 dB amplifierIFIM50 coaxial cable 2 balun 1 mains balun 2 coaxial cable3 a 20 dB attenuator Radial R412720000 and the receiverThebaluns were considered as part of the channelThe calibrationfor measurement path 1 was made by directly connectingcoaxial cable 2 and coaxial cable 3

Path 2 is composed by the transmitter coaxial cable 1a 30 dB amplifier IFIM50 coaxial cable 2 the wire to freespace propagation channel (represented by theCTF119867(119891 r

3))

the biconical antenna coaxial cable 4 and the DSO Thecalibration for this second path was made by connectingcoaxial cable 2 and coaxial cable 4 The gain of the antennawas removed from the measurements by postprocessing

The signal generator and the DSO were synchronized bya direct connection of their 10MHz reference clocks Thesampling rates of both devices were set to 119891

119904= 100MHz

33 Signal Processing After initial system calibration themeasurements are performed in 3 steps

(1) The CTF 119867(119891 r0) is evaluated using a specific Tx

frame (see below) The considered frequency bandextends from 28MHz to 375MHz

(2) The TR filter is generated using the phase and magni-tude of119867(119891 r

0)

(3) The CTFs 119867(119891 r0) and 119867(119891 r

3) as well as the per-

ceived CTFs 119867TR(119891 r0) and 119867TR(119891 r3) are measuredusing a single Tx frame (see below) Measurementsat the location of the Rx outlet (r

0) and at arbitrary

locations in space (r3) are made simultaneously using

two ports of the oscilloscope one connected to theRx balun and the other connected to the biconicalantenna

The Tx signal is generated according to the HomePlugstandard [25] The used frame is called PHY Protocol DataUnit (PPDU) and it is composed by a preamble (including aFrame Control) and a number of payload symbols as shownin Figure 6 Each payload symbol consists of a 3072-sampleOFDMsymbol as defined in theHomePlug specification [25]In order to estimate the CTFs 119867(119891 r) and 119867TR(119891 r) theOFDM symbols were loaded with predefined constellations

Journal of Electrical and Computer Engineering 5

Preamble Symbol 1 Symbol 2 Symbol 3 Symbol 4 Symbol 5 Symbol 6

Figure 6 HomePlug frame

Preamble 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905)

Figure 7 HomePlug frame for calibration and CTF measurementbefore computation of the TR filter

The frame used for the calibration and the initial mea-surement of the CTF before computation of the TR filter isrepresented in Figure 7

The frame used for the measurement of the CTF andEMI after computation of the TR filter is represented inFigure 8 Note that in this frame the TRfilter is applied on the3 last symbols only (represented by the signal 1199041015840(119905)) From thisparticular frame scheme the CTF and EMI can be evaluatedwith and without application of TR quasi-simultaneouslyRecomputing the CTF at this stage also allows monitoringany possible temporal evolution of the channel betweencalibration and measurement

In essence the overall channel estimation process issimilar to the channel sounding procedure defined in theHomePlug AV specification In our experiment the compu-tation of the CTF from the received signal uses a classicalZero Forcing (ZF) channel estimation procedure This issuitable in our experiment involving high levels of Signal-to-Noise Ratio (SNR) at the Rx In practical systems operatingat lower SNR more sophisticated methods such as theMinimum Mean Square Error channel estimation would bemore efficient against noise enhancement Note that we choseto actually implement the TR filter in the time domain atthe transmitter using a programmable waveform generatorin order to measure results as close as possible to a realisticimplementation of the TR scheme in practical PLC modems

In order to estimate channel attenuation continuouslyover the measured frequency band no spectral notcheswere implemented for this study In practical systems a TxPSD mask is defined where the Tx power is notched atpredefined frequencies to protect existing services using thesame spectrumTheHomePlug specification uses windowingin order to better exploit the power allocated within thePSDmask while protecting out-of-band servicesThe presentstudy concentrates on reducing the EMI within the bandeffectively used by PLC systems Therefore the results alsohold for practical systems including notches

34 Measured Environment The measurement campaignwas conducted using 13 different topologies of 230V mainsnetworkswithin the premises ofOrange Labs in LannionThecampaign took place in different rooms of about 5 times 4m2Figure 9 presents a picture of the experimental setup in anexemplar location

The Tx and Rx modems were connected to two outletsin the same room with distances varying between 2m and8m In general the rooms are equipped with several other

Preamble 119904(119905) 119904(119905) 119904(119905) 119904998400(119905) 119904998400(119905) 119904998400(119905)

Figure 8 HomePlug frame for CTF and EMI measurement afterapplication of TR

Figure 9 Picture of experimentation

electrical outlets (between 4 and 10) About half of the outletswere connected to classical office appliances (lamps desktopsetc) For each topology one CTF wasmeasured first withoutapplying TR and then after applying TR filtering In additionfor each topology between 3 and 5 locations were selectedto measure the received electrical field with the help of thebiconical antenna In total 13 CTF and 43 measurements ofthe electrical field were collected for statistical analysis

4 Results and Statistical Analysis

41 Preliminary Results Thepreliminary results for an exem-plar networkwill be presented in two parts first the CTF at r

0

and second the electric field and its associated power densityat r3= r0

For the CTF at r0 we show its attenuation characteristics

in Figure 10Let us first consider the measured CTF (black curve)The

deep notches at some frequencies are due to reflections at theterminations of the network and reflect the multipath natureof the PLC networkWe define the average attenuation beforeTR119867(r

0) in dB as

119867(r0) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867 (119891 1199030)10038161003816100381610038162119889119891)

(5)

where 119891min and 119891max respectively represent the minimumandmaximum sounded frequenciesThe average attenuationcorresponds to the signal attenuation perceived by a receivercapable of exploiting all the power received in the frequencyband from 119891min to 119891max Typically an OFDM system likethe HomePlug AV specification is able to exploit the totalreceived power over a wide frequency band In our examplethe average attenuation is about 11 dB

Let us now focus on the perceived CTF after applyingTR Owing to the mathematical definition of the TR filterTR allocates more power to frequencies showing minimal

6 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35

05

1015202530354045

Frequency (MHz)

Atte

nuat

ion

(dB)

Measured CTFPerceived CTF after TR

minus5

Figure 10 Channel attenuation before and after TR

0 5 10 15 20 25 30 355

10

15

20

25

30

35

40

Frequency (MHz)

Elec

tric

fiel

d (d

BuV

m)

Before TR AR power density minus1199dB(Wm2)After TR AR power density minus1274dB(Wm2)

Figure 11 Electric field before and after TR

attenuation while strongly attenuated frequencies are morepower constrained In particular for all frequencies where theattenuation of the channel119867(119891 r0) is higher than the averageattenuation 119867(r

0) the perceived channel is more attenuated

after applying TR This can be clearly seen in the frequencyrange from 26MHz to 375MHz For the frequencies wherethe attenuation of the channel is less than119867(r

0) the response

of the channel is improved using TR We can see this clearlyin the frequency range from 11MHz to 22MHz If we definethe average attenuation after TR119867TR(r0) in dB as

119867TR(r0) = 10log10 (1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867TR (119891 1199030)10038161003816100381610038162119889119891)

(6)

we observe that the average attenuation after applying TR isabout 75 dB Hence the application of TR provided a gain inthe total received power of 35 dB in this particular example

Note that only the measured CTF (black curve) is phys-ical Therefore the measured CTF will always degrade the

transmission with some degree of channel attenuation Onthe contrary the CTF perceived at Rx after application of TR(red curve) corresponds to a logical channel where the effectsof the measured CTF are combined with the effects of the TRfilter at Tx Hence the perceived CTF may exhibit some gainover a limited frequency range

We now consider the electrical field and its associatedpower density at r

3= r0 The value of the electrical field

119864(119891 r3) in dB120583Vm was computed from the CTF 119867(119891 r

3)

measured between the Tx balun and the antenna connectorassuming an injected PSD 119875feed = minus55 dBmHz and using thefollowing formula [24]

119864 (119891 r3) = 119875feed + 20 log10 (

1003816100381610038161003816119867 (119891 r3)1003816100381610038161003816) + 107 + AF (119891)

(7)

where AF(119891) represents the antenna factor and 107 representsthe conversion from dBm to dB120583V In addition we computedthe average radiated power density 119878(r

3) in dB(Wm2) as

119878 (r3) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864 (119891 r3)10038161003816100381610038162119889119891)

(8)

where119891min and 119891max respectively represent the minimumandmaximum sounded frequencies and119864(119891 r

3) is expressed

in Vm In (8) the term 120 times 120587 provides the value of theimpedance of free space in Ohm

Note that both 119864(119891 r3) and 119878(r

3) can also be computed

after applying TR filtering using the following equations

119864TR (119891 r3)=119875feed + 20 log10 (1003816100381610038161003816119867TR (119891 r3)

1003816100381610038161003816)+ 107 +AF (119891)(9)

119878TR (r3)

= 10log10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864TR (119891 r3)10038161003816100381610038162119889119891)

(10)

An example of radiated emission measurement is givenin Figure 11 For this particular electrical network the mit-igation of radiated emissions is clear in the frequency bandfrom 26MHz to 375MHz and the average radiated powerdensity has reduced by about 75 dB Thus we can observewith this example that the application of TR filtering canreduce significantly the level of undesired radiated power

42 Statistical Analysis In this section we present the statis-tical analysis of the measurement database collected within13 rooms of the office building at Orange Labs in LannionThe total measurement set is composed of 13 CTF and 43measurements of the electrical field

We first computed the channel gain 119866TR observed on theperceived CTF119867TR(119891 r0) after application of TR filtering AsOFDM systems can exploit the total power received over agiven frequency band we computed this gain in dB for thetotal received power as

119866TR = 119867(r0) minus 119867TR (r0) (11)

Journal of Electrical and Computer Engineering 7

1 2 3 4 5 6 70

0102030405

06

0708

091

Gain of channel (dB)

CDF

Figure 12 CDF of the channel gain 119866TR

0 2 4 6 80

010203040506070809

1

CDF

minus2

Reduction of total average power density (dBWm2)

Figure 13 CDF of the EMI reduction coefficient 119877TR

The cumulative distribution function (CDF) of 119866TR isgiven in Figure 12 This parameter shows always a positivegain between 14 dB and 66 dB in our experiment which isin line with results reported in similar wireless experiments[18] In about 60 of cases the channel gain is higher than3 dBThis means that at Rx we have always a better receptionusing TR This channel gain can in turn be used to reducethe injected PSD at Tx hence reducing by the same factor theunwanted EMI

We then computed the EMI reduction coefficient 119877TRcorresponding to the reduction of the undesired radiatedpower due to the application of a TR filter This figure ofmerit of the reduction of radiated signal is computed in dBas follows

119877TR = 119878(r3) minus 119878TR(r3) (12)

The CDF of the EMI reduction coefficient 119877TR is givenin Figure 13 Results show that the simple application of TRreduces the EMI in more than 60 of cases In the bestcase the EMI reduced by more than 7 dB using TR In theworst case the EMI increased by 2 dB The observationsof particular cases indicates that the reduction of EMI ismore effective when the CTFs 119867(119891 r

0) and 119867(119891 r

3) are

highly decorrelatedThis is more likely to happen in complex

0 2 4 6 8 10 120

010203040506070809

1

CDF

Reduction of total average power density + gain of channel (dBWm2)

Figure 14 CDF of the effective EMI mitigation119872TR

electrical network topologies where the rich multipath envi-ronment results in different frequency fading structures fordifferent Rx locations

Observing the statistics of the channel gain 119866TR and ofthe EMI reduction coefficient119877TR an optimal strategy can beproposed in order tominimize EMI for a PLC system IndeedFigure 12 shows that the application of TR provides bettersystem performance due to the reduced channel attenuationThis gives us in turn the flexibility to reduce the level of theTx power to further reduce EMI while keeping the systemperformance constant More precisely when TR is applieda Tx power back-off of 119866TR dB can be applied withoutmodifying the total received power Finally following thispower backoff strategy the effective EMI mitigation factor119872TR can be computed as the sum of the power backoff andthe EMI reduction coefficient

119872TR = 119866TR + 119877TR (13)

The CDF of the effective EMI mitigation factor 119872TR isdepicted in Figure 14 Several conclusions can be drawn fromthis statistical result

(i) First theTRmethod is able tomitigate EMI generatedby PLC transmission in 100 of our experimentalobservations To this respect one can conclude thatthe gain119866TR provided by the application of TR allowsa Tx power backoff that largely compensates for thepossible EMI increment observed in Figure 13

(ii) Second in 40 of the cases the undesired radiatedpower is reduced by more than 3 dB

(iii) Finally in the most favorable configurations a reduc-tion of the EMI by more than 10 dB can be observedSuch configurations correspond to cases where theCTF between the Tx and Rx modem and the EMIspectrum are particularly decorrelated

5 Conclusion

In this paper we proposed for the first time the application ofTR in order to mitigate EMI generated by wired communi-cation systems TR was originally used in the field of wireless

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

2 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35 40Frequency (MHz)

PSD

lim

it (a

vera

ge) (

dBm

Hz)

minus95

minus90

minus85

minus80

minus75

minus70

minus65

minus60

minus55

minus50

Figure 1 Example of PSD limit (average) for North America

through the network Consequently the copper wires usedfor transmitting the useful signal act as an antenna and partof the transmitted power is radiated This not only results instronger signal attenuation at the receiver but also leads toElectromagnetic Compatibility (EMC) issues as the radiatedsignal may interfere with other existing services such asamateur radio (HAM) or Short Wave (SW) broadcastingThe impact of PLC transmission on EMC has been studiedfor example within the ICT FP7 project OMEGA [7] andthrough the ETSI Specialist Task Force 410 [8]

In order to avoid interference between PLC systems andother users of the spectrum regulation authorities imposestrict emissionmasks for the transmission of electromagneticsignals on the power lines In the USA the Federal Commu-nicationsCommission (FCC)Part 15 [9] specifies amaximumlevel of radiated field for carrier-current systems (includingPLC) leading system specifications to define constrainedpower transmission masks Figure 1 represents an example ofPower Spectral Density (PSD) limits provided in the IEEE1901 standard for North America The observed notches aredefined to protect specific systems such as HAM bandsIn Europe CENELEC is currently developing a draft onregulation standard applying to in-home PLC systems [10]

Regardless of the regulation limits in place the researchpresented in this paper focuses on the mitigation of unin-tentional radiation due to PLC systems Several attempts tosolve this problem have been presented in the literatureReference [11] presents amethod to reduce radiated emissionsby applying an auxiliary signal cancelling the electromagneticfield on a given point in space Simulations demonstratedgood performance with the drawback that the ElectroMagnetic Interference (EMI) could only be mitigated at asingle location The authors in [12] used additional hardwareconnected at the wall outlets in order to reduce asymmetrieson power lines at the cost of an increased complexity of thePLC network

In our approach we tried to simultaneously reach twocomplementary goals by the means of digital signal pro-cessing First we intend to focus the transmitted signal atthe receiver location The power gain linked to this energyfocalization allows in turn relaxing the required power level

at the transmitter hence generating less Electro MagneticInterference (EMI) Second we target to reduce the level ofenergy dissipated at any location except the intended receiverIn particular it is desirable to minimize the level of radiatedpower from the electrical wires These two benefits alreadyappear as features of a known technique in the field ofwireless transmission Time Reversal (TR) [13] Experimentalinvestigations conducted using Ultra Wide Band (UWB)radio waves demonstrated both the focusing and interferencemitigation properties of this technique [14 15]

In this paper we present for the first time an experimentalanalysis of TR as a mean to mitigate radiation effects forwired signal transmission Our investigation focuses onHighFrequency (HF) BB PLC signals but could be extended to NBPLC and other wired transmission systems such as DigitalSubscriber Line (DSL) access The paper is organized asfollows Section 2 presents the concepts of the TR techniqueand its application to wired systems Section 3 details theexperiment conducted to assess the merits of TR for BB PLCand the results are statistically analyzed in Section 4 Finallyconclusions are drawn in Section 5

2 Time Reversal for PowerLine Communications

21 Time Reversal for Wireless Transmission The TR tech-nique also known as phase conjugation in the frequencydomain was first used in the fields of acoustics [16 17]More recently this concept has been successfully extendedto electromagnetic waves where the rich multipath channelprovides excellent conditions for its application [13] Thebasic concept of TR is simple Let 120575(120591) be an ideal Diracimpulse emitted by a transmitter (Tx) antenna (Figure 2(a))By definition at any receiver (Rx) location r

0 the received

signal is given by the Channel Impulse Response (CIR)ℎ(120591 r0) The CIR is composed of multiple echoes reflectingthe multiple propagation paths of the propagation channel

TR uses this Channel State Information (CSI) at the Txto prefilter the signal to be transmitted More specificallythe CIR ℎ(120591 r0) is time reversed and normalized to serve asan input filter for the signal to be transmitted (Figure 2(b))Physically each delayed echo constituting theTRfilter travelsamong other multiple paths through its original propagationpath As a result themultiple echoes sumup coherently at thereceiver hence focusing the received energy in time

Mathematically applying TR leads for any Rx situatedat an arbitrary location r to the equivalent perceived CIRℎTR(120591 r) (note that this formulation holds for a real valuedCIR) [18]

ℎTR (120591 r) =ℎ (minus120591 r

0)

radicint1003816100381610038161003816ℎ (120591 r0)

10038161003816100381610038162119889120591

otimes ℎ (120591 r) (1)

where the symbol otimes denotes time domain convolutionFormulating (1) in the frequency domain leads to

119867TR (119891 r) =119867lowast(119891 r0)

radicint1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162119889119891

times 119867 (119891 r) (2)

Journal of Electrical and Computer Engineering 3

120575(120591)

Tx Rx

ℎ(120591 r0)

(a)

Tx Rx

ℎ(minus120591 r0) ℎ(120591 r0)⨂ℎ(minus120591 r0)

(b)

Figure 2 Transmission over an ideal multipath propagation channel (a) of a Dirac impulse (b) of the time-reversed CIR

where119867(119891 r) represents the complex valued Channel Trans-fer Function (CTF) and the superscript lowast denotes the com-plex conjugate operation For this reason TR is sometimescalled frequency domain phase conjugation

Two conclusions can be drawn from (1) and (2) First atthe intended Rx location r

0 the perceived CIR simplifies to

ℎTR (120591 r0) =1

radicint1003816100381610038161003816ℎ (120591 r0)

10038161003816100381610038162119889120591

times 119877ℎ(120591 r0) (3)

where 119877ℎ(120591 r0) denotes the time domain autocorrelation of

the function ℎ(120591 r0) Similarly the perceived CTF simplifies

to

119867TR (119891 r0) =1

radicint1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162119889119891

times1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162 (4)

In the time domain the effect of the TR filter is to trans-form the CIR into its autocorrelation For a rich multipathenvironment the autocorrelation of the CIR presents a largepeak at 120591 = 0 with reduced side echoes Experimentalstudies demonstrated that the resulting channel is less spreadin time [14 15] hence reducing the possible IntersymbolInterference (ISI) In the frequency domain the perceivedCTF is proportional to the square of the magnitude of theactual CTF Besides the fact that TR provides a real valuedCTF which could be exploited at the receiver this also leadsto a significant gain in terms of Rx power due to a betterexploitation of the frequency selective nature of the channelIt was demonstrated in [18] that the application of TR in a flatchannel (ie without frequency domain power decay) underRayleigh fading leads to a gain of 3 dB in the total receivedpower This gain was increased to 5 dB when considering thefrequency domain power decay observed in practical UWBradio channels

The second conclusion drawn from (1) and (2) is that forany other location r different from r

0 TR creates a mismatch

between the Tx filter and the channel This is particularlyobservable in the frequency domain representation of TRgiven in (2) The perceived CTF corresponds to the productof two independent CTF119867(119891 r) and119867lowast(119891 r

0) with possibly

very different frequency fading structures More preciselyminima of the first CTF can happen randomly at maxima ofthe second CTF Hence averaging over all frequencies thetotal received power at untargeted locations is reduced Inwireless TR analysis this effect is called spatial focusing andis generally assessed as the ratio between the maximum ofℎTR(120591 r) and the maximum of ℎTR(120591 r0) for a given distancerminus r0 Spatial focusing factors of minus10 dB have been reported

in [18 19]

22 Extension of Time Reversal to Wired Transmission Asobserved through the study of TR for wireless transmissionthe TR scheme provides two main features namely anincrease of the Rx power at the intended Rx location and adecrease of the Rx power at any other locationThese featuresare highly desirable in the context of wired transmissionwhere the level of Tx power is constrained by the uninten-tional radiation from the wires causing possible EMI to othersystems

Based on this observation we conducted experimentalstudies to analyze the potential of TR to mitigate unwantedemissions for PLC systems The main principles of theextension of TR to PLC transmission can be explained withthe help of Figure 3

We assume an intended transmission between a Tx PLCmodem and an Rx PLCmodem over an LV indoor electricalnetwork Different experimental investigations reported thePLC channel as a rich multipath propagation channel due tothe multiple branches present in a classical electrical networkand to the impedance mismatch occurring at the networkterminations (outlets) and nodes [20ndash23] This similarity ofthe PLC channel with wireless channels suggests promisingresults when applying TR to PLC

With reference to Figure 3 the Rx modem is situated atthe intended location r

0 and the Tx modem is situated at the

origin By applying TR filtering at the Tx the Rx power willbe increased at location r

0 hence the Rx modem will benefit

from an increased Signal-to-Noise Ratio (SNR) This powerincreased can also in turn be applied as a reduction of the Txpower to achieve similar performance At other outlets in thenetwork situated for example at locations r

1or r2 the Rx

power will be reduced This effect can be further exploited inthe design of multiuser transmission schemes

For our purpose of radiation mitigation let us nowconsider a location r

3 situated at any point in space in the

vicinity of the electrical network The level of radiated fieldat this location can be evaluated for instance by the meansof an equivalent transfer function 119867(119891 r

3) between the Tx

modem and an ideal antenna situated at location r3 By the

virtue of the TR scheme (2) the perceived transfer functionat location r

3after applying TR will be proportional to the

product119867lowast(119891 r0) times 119867(119891 r3) As the functions119867lowast(119891 r0) and119867(119891 r

3) are not correlated their frequency fading structures

are different In particular the deep notches due to frequencyselective fading do not appear at the same frequencies Asa result the product 119867lowast(119891 r

0) times 119867(119891 r

3) will provide more

average attenuation when compared to 119867(119891 r3) alone and

therefore the total power radiated at location r3will be

reduced A similar observation is made in previous studies

4 Journal of Electrical and Computer Engineering

Tx Rx r0

r1

r2

r3

Figure 3 Principle of the extension of TR to wired transmission

Signal generator Oscilloscope

Biconical antenna

Balun1 Balun2

Mains

Tx Rx

119904(119905) 119910(119905)

ℎ(120591 r0)

ℎ(120591 r3)

Position r0

Position r3

Figure 4 Equipment used in the experimental setup

dedicated to wireless channels [14 15 18] Hence TR appearsas an efficient method to mitigate EMI for wired communi-cations

The theoretical basis for the application of TR to wiredtransmission being set wewill nowdescribe the experimentalassessment of this method in the next Sections

3 Experimental Setup

31 Equipment In order to experimentally assess the use ofTR as a method to mitigate EMI for wired communicationwe used the experimental setup presented in Figure 4 In thissetup a signal generator Tektronix AWG7082C was used as ageneric Tx A Digital Sampling Oscilloscope (DSO) LeCroyWaveRunner 715Zi-A was used to sample the received signalat Rx The signal generated at Tx is denoted 119904(119905) and thesignal received at the DSO is denoted 119910(119905) Two UniversalPLC Couplers were used as baluns to couple the Tx and Rxsignals with the power lines The Tx coupler is situated at theorigin and the Rx coupler is connected to a plug at locationr0 These couplers were developed within the ETSI Specialist

Task Force 410 [24] In order to measure the power densityof the radiated emission received at any arbitrary location r

3

a biconical antenna Schwarzbeck EFS921 was connected to asecond port of the DSO

Mains

Balun1 Balun2

Signal generatorOscilloscope

Coaxial 1

Coaxial 2

Coaxial 3

Coaxial 4

Biconical antennaAttenuator

Amplifier

Figure 5 Experimental setup calibration

32 Calibration With reference to Figure 5 two measure-ment paths can be distinguished Path 1 serves for themeasurement of the CTF and is composed of the followingelements the transmitter coaxial cable 1 a 30 dB amplifierIFIM50 coaxial cable 2 balun 1 mains balun 2 coaxial cable3 a 20 dB attenuator Radial R412720000 and the receiverThebaluns were considered as part of the channelThe calibrationfor measurement path 1 was made by directly connectingcoaxial cable 2 and coaxial cable 3

Path 2 is composed by the transmitter coaxial cable 1a 30 dB amplifier IFIM50 coaxial cable 2 the wire to freespace propagation channel (represented by theCTF119867(119891 r

3))

the biconical antenna coaxial cable 4 and the DSO Thecalibration for this second path was made by connectingcoaxial cable 2 and coaxial cable 4 The gain of the antennawas removed from the measurements by postprocessing

The signal generator and the DSO were synchronized bya direct connection of their 10MHz reference clocks Thesampling rates of both devices were set to 119891

119904= 100MHz

33 Signal Processing After initial system calibration themeasurements are performed in 3 steps

(1) The CTF 119867(119891 r0) is evaluated using a specific Tx

frame (see below) The considered frequency bandextends from 28MHz to 375MHz

(2) The TR filter is generated using the phase and magni-tude of119867(119891 r

0)

(3) The CTFs 119867(119891 r0) and 119867(119891 r

3) as well as the per-

ceived CTFs 119867TR(119891 r0) and 119867TR(119891 r3) are measuredusing a single Tx frame (see below) Measurementsat the location of the Rx outlet (r

0) and at arbitrary

locations in space (r3) are made simultaneously using

two ports of the oscilloscope one connected to theRx balun and the other connected to the biconicalantenna

The Tx signal is generated according to the HomePlugstandard [25] The used frame is called PHY Protocol DataUnit (PPDU) and it is composed by a preamble (including aFrame Control) and a number of payload symbols as shownin Figure 6 Each payload symbol consists of a 3072-sampleOFDMsymbol as defined in theHomePlug specification [25]In order to estimate the CTFs 119867(119891 r) and 119867TR(119891 r) theOFDM symbols were loaded with predefined constellations

Journal of Electrical and Computer Engineering 5

Preamble Symbol 1 Symbol 2 Symbol 3 Symbol 4 Symbol 5 Symbol 6

Figure 6 HomePlug frame

Preamble 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905)

Figure 7 HomePlug frame for calibration and CTF measurementbefore computation of the TR filter

The frame used for the calibration and the initial mea-surement of the CTF before computation of the TR filter isrepresented in Figure 7

The frame used for the measurement of the CTF andEMI after computation of the TR filter is represented inFigure 8 Note that in this frame the TRfilter is applied on the3 last symbols only (represented by the signal 1199041015840(119905)) From thisparticular frame scheme the CTF and EMI can be evaluatedwith and without application of TR quasi-simultaneouslyRecomputing the CTF at this stage also allows monitoringany possible temporal evolution of the channel betweencalibration and measurement

In essence the overall channel estimation process issimilar to the channel sounding procedure defined in theHomePlug AV specification In our experiment the compu-tation of the CTF from the received signal uses a classicalZero Forcing (ZF) channel estimation procedure This issuitable in our experiment involving high levels of Signal-to-Noise Ratio (SNR) at the Rx In practical systems operatingat lower SNR more sophisticated methods such as theMinimum Mean Square Error channel estimation would bemore efficient against noise enhancement Note that we choseto actually implement the TR filter in the time domain atthe transmitter using a programmable waveform generatorin order to measure results as close as possible to a realisticimplementation of the TR scheme in practical PLC modems

In order to estimate channel attenuation continuouslyover the measured frequency band no spectral notcheswere implemented for this study In practical systems a TxPSD mask is defined where the Tx power is notched atpredefined frequencies to protect existing services using thesame spectrumTheHomePlug specification uses windowingin order to better exploit the power allocated within thePSDmask while protecting out-of-band servicesThe presentstudy concentrates on reducing the EMI within the bandeffectively used by PLC systems Therefore the results alsohold for practical systems including notches

34 Measured Environment The measurement campaignwas conducted using 13 different topologies of 230V mainsnetworkswithin the premises ofOrange Labs in LannionThecampaign took place in different rooms of about 5 times 4m2Figure 9 presents a picture of the experimental setup in anexemplar location

The Tx and Rx modems were connected to two outletsin the same room with distances varying between 2m and8m In general the rooms are equipped with several other

Preamble 119904(119905) 119904(119905) 119904(119905) 119904998400(119905) 119904998400(119905) 119904998400(119905)

Figure 8 HomePlug frame for CTF and EMI measurement afterapplication of TR

Figure 9 Picture of experimentation

electrical outlets (between 4 and 10) About half of the outletswere connected to classical office appliances (lamps desktopsetc) For each topology one CTF wasmeasured first withoutapplying TR and then after applying TR filtering In additionfor each topology between 3 and 5 locations were selectedto measure the received electrical field with the help of thebiconical antenna In total 13 CTF and 43 measurements ofthe electrical field were collected for statistical analysis

4 Results and Statistical Analysis

41 Preliminary Results Thepreliminary results for an exem-plar networkwill be presented in two parts first the CTF at r

0

and second the electric field and its associated power densityat r3= r0

For the CTF at r0 we show its attenuation characteristics

in Figure 10Let us first consider the measured CTF (black curve)The

deep notches at some frequencies are due to reflections at theterminations of the network and reflect the multipath natureof the PLC networkWe define the average attenuation beforeTR119867(r

0) in dB as

119867(r0) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867 (119891 1199030)10038161003816100381610038162119889119891)

(5)

where 119891min and 119891max respectively represent the minimumandmaximum sounded frequenciesThe average attenuationcorresponds to the signal attenuation perceived by a receivercapable of exploiting all the power received in the frequencyband from 119891min to 119891max Typically an OFDM system likethe HomePlug AV specification is able to exploit the totalreceived power over a wide frequency band In our examplethe average attenuation is about 11 dB

Let us now focus on the perceived CTF after applyingTR Owing to the mathematical definition of the TR filterTR allocates more power to frequencies showing minimal

6 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35

05

1015202530354045

Frequency (MHz)

Atte

nuat

ion

(dB)

Measured CTFPerceived CTF after TR

minus5

Figure 10 Channel attenuation before and after TR

0 5 10 15 20 25 30 355

10

15

20

25

30

35

40

Frequency (MHz)

Elec

tric

fiel

d (d

BuV

m)

Before TR AR power density minus1199dB(Wm2)After TR AR power density minus1274dB(Wm2)

Figure 11 Electric field before and after TR

attenuation while strongly attenuated frequencies are morepower constrained In particular for all frequencies where theattenuation of the channel119867(119891 r0) is higher than the averageattenuation 119867(r

0) the perceived channel is more attenuated

after applying TR This can be clearly seen in the frequencyrange from 26MHz to 375MHz For the frequencies wherethe attenuation of the channel is less than119867(r

0) the response

of the channel is improved using TR We can see this clearlyin the frequency range from 11MHz to 22MHz If we definethe average attenuation after TR119867TR(r0) in dB as

119867TR(r0) = 10log10 (1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867TR (119891 1199030)10038161003816100381610038162119889119891)

(6)

we observe that the average attenuation after applying TR isabout 75 dB Hence the application of TR provided a gain inthe total received power of 35 dB in this particular example

Note that only the measured CTF (black curve) is phys-ical Therefore the measured CTF will always degrade the

transmission with some degree of channel attenuation Onthe contrary the CTF perceived at Rx after application of TR(red curve) corresponds to a logical channel where the effectsof the measured CTF are combined with the effects of the TRfilter at Tx Hence the perceived CTF may exhibit some gainover a limited frequency range

We now consider the electrical field and its associatedpower density at r

3= r0 The value of the electrical field

119864(119891 r3) in dB120583Vm was computed from the CTF 119867(119891 r

3)

measured between the Tx balun and the antenna connectorassuming an injected PSD 119875feed = minus55 dBmHz and using thefollowing formula [24]

119864 (119891 r3) = 119875feed + 20 log10 (

1003816100381610038161003816119867 (119891 r3)1003816100381610038161003816) + 107 + AF (119891)

(7)

where AF(119891) represents the antenna factor and 107 representsthe conversion from dBm to dB120583V In addition we computedthe average radiated power density 119878(r

3) in dB(Wm2) as

119878 (r3) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864 (119891 r3)10038161003816100381610038162119889119891)

(8)

where119891min and 119891max respectively represent the minimumandmaximum sounded frequencies and119864(119891 r

3) is expressed

in Vm In (8) the term 120 times 120587 provides the value of theimpedance of free space in Ohm

Note that both 119864(119891 r3) and 119878(r

3) can also be computed

after applying TR filtering using the following equations

119864TR (119891 r3)=119875feed + 20 log10 (1003816100381610038161003816119867TR (119891 r3)

1003816100381610038161003816)+ 107 +AF (119891)(9)

119878TR (r3)

= 10log10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864TR (119891 r3)10038161003816100381610038162119889119891)

(10)

An example of radiated emission measurement is givenin Figure 11 For this particular electrical network the mit-igation of radiated emissions is clear in the frequency bandfrom 26MHz to 375MHz and the average radiated powerdensity has reduced by about 75 dB Thus we can observewith this example that the application of TR filtering canreduce significantly the level of undesired radiated power

42 Statistical Analysis In this section we present the statis-tical analysis of the measurement database collected within13 rooms of the office building at Orange Labs in LannionThe total measurement set is composed of 13 CTF and 43measurements of the electrical field

We first computed the channel gain 119866TR observed on theperceived CTF119867TR(119891 r0) after application of TR filtering AsOFDM systems can exploit the total power received over agiven frequency band we computed this gain in dB for thetotal received power as

119866TR = 119867(r0) minus 119867TR (r0) (11)

Journal of Electrical and Computer Engineering 7

1 2 3 4 5 6 70

0102030405

06

0708

091

Gain of channel (dB)

CDF

Figure 12 CDF of the channel gain 119866TR

0 2 4 6 80

010203040506070809

1

CDF

minus2

Reduction of total average power density (dBWm2)

Figure 13 CDF of the EMI reduction coefficient 119877TR

The cumulative distribution function (CDF) of 119866TR isgiven in Figure 12 This parameter shows always a positivegain between 14 dB and 66 dB in our experiment which isin line with results reported in similar wireless experiments[18] In about 60 of cases the channel gain is higher than3 dBThis means that at Rx we have always a better receptionusing TR This channel gain can in turn be used to reducethe injected PSD at Tx hence reducing by the same factor theunwanted EMI

We then computed the EMI reduction coefficient 119877TRcorresponding to the reduction of the undesired radiatedpower due to the application of a TR filter This figure ofmerit of the reduction of radiated signal is computed in dBas follows

119877TR = 119878(r3) minus 119878TR(r3) (12)

The CDF of the EMI reduction coefficient 119877TR is givenin Figure 13 Results show that the simple application of TRreduces the EMI in more than 60 of cases In the bestcase the EMI reduced by more than 7 dB using TR In theworst case the EMI increased by 2 dB The observationsof particular cases indicates that the reduction of EMI ismore effective when the CTFs 119867(119891 r

0) and 119867(119891 r

3) are

highly decorrelatedThis is more likely to happen in complex

0 2 4 6 8 10 120

010203040506070809

1

CDF

Reduction of total average power density + gain of channel (dBWm2)

Figure 14 CDF of the effective EMI mitigation119872TR

electrical network topologies where the rich multipath envi-ronment results in different frequency fading structures fordifferent Rx locations

Observing the statistics of the channel gain 119866TR and ofthe EMI reduction coefficient119877TR an optimal strategy can beproposed in order tominimize EMI for a PLC system IndeedFigure 12 shows that the application of TR provides bettersystem performance due to the reduced channel attenuationThis gives us in turn the flexibility to reduce the level of theTx power to further reduce EMI while keeping the systemperformance constant More precisely when TR is applieda Tx power back-off of 119866TR dB can be applied withoutmodifying the total received power Finally following thispower backoff strategy the effective EMI mitigation factor119872TR can be computed as the sum of the power backoff andthe EMI reduction coefficient

119872TR = 119866TR + 119877TR (13)

The CDF of the effective EMI mitigation factor 119872TR isdepicted in Figure 14 Several conclusions can be drawn fromthis statistical result

(i) First theTRmethod is able tomitigate EMI generatedby PLC transmission in 100 of our experimentalobservations To this respect one can conclude thatthe gain119866TR provided by the application of TR allowsa Tx power backoff that largely compensates for thepossible EMI increment observed in Figure 13

(ii) Second in 40 of the cases the undesired radiatedpower is reduced by more than 3 dB

(iii) Finally in the most favorable configurations a reduc-tion of the EMI by more than 10 dB can be observedSuch configurations correspond to cases where theCTF between the Tx and Rx modem and the EMIspectrum are particularly decorrelated

5 Conclusion

In this paper we proposed for the first time the application ofTR in order to mitigate EMI generated by wired communi-cation systems TR was originally used in the field of wireless

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 3

120575(120591)

Tx Rx

ℎ(120591 r0)

(a)

Tx Rx

ℎ(minus120591 r0) ℎ(120591 r0)⨂ℎ(minus120591 r0)

(b)

Figure 2 Transmission over an ideal multipath propagation channel (a) of a Dirac impulse (b) of the time-reversed CIR

where119867(119891 r) represents the complex valued Channel Trans-fer Function (CTF) and the superscript lowast denotes the com-plex conjugate operation For this reason TR is sometimescalled frequency domain phase conjugation

Two conclusions can be drawn from (1) and (2) First atthe intended Rx location r

0 the perceived CIR simplifies to

ℎTR (120591 r0) =1

radicint1003816100381610038161003816ℎ (120591 r0)

10038161003816100381610038162119889120591

times 119877ℎ(120591 r0) (3)

where 119877ℎ(120591 r0) denotes the time domain autocorrelation of

the function ℎ(120591 r0) Similarly the perceived CTF simplifies

to

119867TR (119891 r0) =1

radicint1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162119889119891

times1003816100381610038161003816119867 (119891 r0)

10038161003816100381610038162 (4)

In the time domain the effect of the TR filter is to trans-form the CIR into its autocorrelation For a rich multipathenvironment the autocorrelation of the CIR presents a largepeak at 120591 = 0 with reduced side echoes Experimentalstudies demonstrated that the resulting channel is less spreadin time [14 15] hence reducing the possible IntersymbolInterference (ISI) In the frequency domain the perceivedCTF is proportional to the square of the magnitude of theactual CTF Besides the fact that TR provides a real valuedCTF which could be exploited at the receiver this also leadsto a significant gain in terms of Rx power due to a betterexploitation of the frequency selective nature of the channelIt was demonstrated in [18] that the application of TR in a flatchannel (ie without frequency domain power decay) underRayleigh fading leads to a gain of 3 dB in the total receivedpower This gain was increased to 5 dB when considering thefrequency domain power decay observed in practical UWBradio channels

The second conclusion drawn from (1) and (2) is that forany other location r different from r

0 TR creates a mismatch

between the Tx filter and the channel This is particularlyobservable in the frequency domain representation of TRgiven in (2) The perceived CTF corresponds to the productof two independent CTF119867(119891 r) and119867lowast(119891 r

0) with possibly

very different frequency fading structures More preciselyminima of the first CTF can happen randomly at maxima ofthe second CTF Hence averaging over all frequencies thetotal received power at untargeted locations is reduced Inwireless TR analysis this effect is called spatial focusing andis generally assessed as the ratio between the maximum ofℎTR(120591 r) and the maximum of ℎTR(120591 r0) for a given distancerminus r0 Spatial focusing factors of minus10 dB have been reported

in [18 19]

22 Extension of Time Reversal to Wired Transmission Asobserved through the study of TR for wireless transmissionthe TR scheme provides two main features namely anincrease of the Rx power at the intended Rx location and adecrease of the Rx power at any other locationThese featuresare highly desirable in the context of wired transmissionwhere the level of Tx power is constrained by the uninten-tional radiation from the wires causing possible EMI to othersystems

Based on this observation we conducted experimentalstudies to analyze the potential of TR to mitigate unwantedemissions for PLC systems The main principles of theextension of TR to PLC transmission can be explained withthe help of Figure 3

We assume an intended transmission between a Tx PLCmodem and an Rx PLCmodem over an LV indoor electricalnetwork Different experimental investigations reported thePLC channel as a rich multipath propagation channel due tothe multiple branches present in a classical electrical networkand to the impedance mismatch occurring at the networkterminations (outlets) and nodes [20ndash23] This similarity ofthe PLC channel with wireless channels suggests promisingresults when applying TR to PLC

With reference to Figure 3 the Rx modem is situated atthe intended location r

0 and the Tx modem is situated at the

origin By applying TR filtering at the Tx the Rx power willbe increased at location r

0 hence the Rx modem will benefit

from an increased Signal-to-Noise Ratio (SNR) This powerincreased can also in turn be applied as a reduction of the Txpower to achieve similar performance At other outlets in thenetwork situated for example at locations r

1or r2 the Rx

power will be reduced This effect can be further exploited inthe design of multiuser transmission schemes

For our purpose of radiation mitigation let us nowconsider a location r

3 situated at any point in space in the

vicinity of the electrical network The level of radiated fieldat this location can be evaluated for instance by the meansof an equivalent transfer function 119867(119891 r

3) between the Tx

modem and an ideal antenna situated at location r3 By the

virtue of the TR scheme (2) the perceived transfer functionat location r

3after applying TR will be proportional to the

product119867lowast(119891 r0) times 119867(119891 r3) As the functions119867lowast(119891 r0) and119867(119891 r

3) are not correlated their frequency fading structures

are different In particular the deep notches due to frequencyselective fading do not appear at the same frequencies Asa result the product 119867lowast(119891 r

0) times 119867(119891 r

3) will provide more

average attenuation when compared to 119867(119891 r3) alone and

therefore the total power radiated at location r3will be

reduced A similar observation is made in previous studies

4 Journal of Electrical and Computer Engineering

Tx Rx r0

r1

r2

r3

Figure 3 Principle of the extension of TR to wired transmission

Signal generator Oscilloscope

Biconical antenna

Balun1 Balun2

Mains

Tx Rx

119904(119905) 119910(119905)

ℎ(120591 r0)

ℎ(120591 r3)

Position r0

Position r3

Figure 4 Equipment used in the experimental setup

dedicated to wireless channels [14 15 18] Hence TR appearsas an efficient method to mitigate EMI for wired communi-cations

The theoretical basis for the application of TR to wiredtransmission being set wewill nowdescribe the experimentalassessment of this method in the next Sections

3 Experimental Setup

31 Equipment In order to experimentally assess the use ofTR as a method to mitigate EMI for wired communicationwe used the experimental setup presented in Figure 4 In thissetup a signal generator Tektronix AWG7082C was used as ageneric Tx A Digital Sampling Oscilloscope (DSO) LeCroyWaveRunner 715Zi-A was used to sample the received signalat Rx The signal generated at Tx is denoted 119904(119905) and thesignal received at the DSO is denoted 119910(119905) Two UniversalPLC Couplers were used as baluns to couple the Tx and Rxsignals with the power lines The Tx coupler is situated at theorigin and the Rx coupler is connected to a plug at locationr0 These couplers were developed within the ETSI Specialist

Task Force 410 [24] In order to measure the power densityof the radiated emission received at any arbitrary location r

3

a biconical antenna Schwarzbeck EFS921 was connected to asecond port of the DSO

Mains

Balun1 Balun2

Signal generatorOscilloscope

Coaxial 1

Coaxial 2

Coaxial 3

Coaxial 4

Biconical antennaAttenuator

Amplifier

Figure 5 Experimental setup calibration

32 Calibration With reference to Figure 5 two measure-ment paths can be distinguished Path 1 serves for themeasurement of the CTF and is composed of the followingelements the transmitter coaxial cable 1 a 30 dB amplifierIFIM50 coaxial cable 2 balun 1 mains balun 2 coaxial cable3 a 20 dB attenuator Radial R412720000 and the receiverThebaluns were considered as part of the channelThe calibrationfor measurement path 1 was made by directly connectingcoaxial cable 2 and coaxial cable 3

Path 2 is composed by the transmitter coaxial cable 1a 30 dB amplifier IFIM50 coaxial cable 2 the wire to freespace propagation channel (represented by theCTF119867(119891 r

3))

the biconical antenna coaxial cable 4 and the DSO Thecalibration for this second path was made by connectingcoaxial cable 2 and coaxial cable 4 The gain of the antennawas removed from the measurements by postprocessing

The signal generator and the DSO were synchronized bya direct connection of their 10MHz reference clocks Thesampling rates of both devices were set to 119891

119904= 100MHz

33 Signal Processing After initial system calibration themeasurements are performed in 3 steps

(1) The CTF 119867(119891 r0) is evaluated using a specific Tx

frame (see below) The considered frequency bandextends from 28MHz to 375MHz

(2) The TR filter is generated using the phase and magni-tude of119867(119891 r

0)

(3) The CTFs 119867(119891 r0) and 119867(119891 r

3) as well as the per-

ceived CTFs 119867TR(119891 r0) and 119867TR(119891 r3) are measuredusing a single Tx frame (see below) Measurementsat the location of the Rx outlet (r

0) and at arbitrary

locations in space (r3) are made simultaneously using

two ports of the oscilloscope one connected to theRx balun and the other connected to the biconicalantenna

The Tx signal is generated according to the HomePlugstandard [25] The used frame is called PHY Protocol DataUnit (PPDU) and it is composed by a preamble (including aFrame Control) and a number of payload symbols as shownin Figure 6 Each payload symbol consists of a 3072-sampleOFDMsymbol as defined in theHomePlug specification [25]In order to estimate the CTFs 119867(119891 r) and 119867TR(119891 r) theOFDM symbols were loaded with predefined constellations

Journal of Electrical and Computer Engineering 5

Preamble Symbol 1 Symbol 2 Symbol 3 Symbol 4 Symbol 5 Symbol 6

Figure 6 HomePlug frame

Preamble 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905)

Figure 7 HomePlug frame for calibration and CTF measurementbefore computation of the TR filter

The frame used for the calibration and the initial mea-surement of the CTF before computation of the TR filter isrepresented in Figure 7

The frame used for the measurement of the CTF andEMI after computation of the TR filter is represented inFigure 8 Note that in this frame the TRfilter is applied on the3 last symbols only (represented by the signal 1199041015840(119905)) From thisparticular frame scheme the CTF and EMI can be evaluatedwith and without application of TR quasi-simultaneouslyRecomputing the CTF at this stage also allows monitoringany possible temporal evolution of the channel betweencalibration and measurement

In essence the overall channel estimation process issimilar to the channel sounding procedure defined in theHomePlug AV specification In our experiment the compu-tation of the CTF from the received signal uses a classicalZero Forcing (ZF) channel estimation procedure This issuitable in our experiment involving high levels of Signal-to-Noise Ratio (SNR) at the Rx In practical systems operatingat lower SNR more sophisticated methods such as theMinimum Mean Square Error channel estimation would bemore efficient against noise enhancement Note that we choseto actually implement the TR filter in the time domain atthe transmitter using a programmable waveform generatorin order to measure results as close as possible to a realisticimplementation of the TR scheme in practical PLC modems

In order to estimate channel attenuation continuouslyover the measured frequency band no spectral notcheswere implemented for this study In practical systems a TxPSD mask is defined where the Tx power is notched atpredefined frequencies to protect existing services using thesame spectrumTheHomePlug specification uses windowingin order to better exploit the power allocated within thePSDmask while protecting out-of-band servicesThe presentstudy concentrates on reducing the EMI within the bandeffectively used by PLC systems Therefore the results alsohold for practical systems including notches

34 Measured Environment The measurement campaignwas conducted using 13 different topologies of 230V mainsnetworkswithin the premises ofOrange Labs in LannionThecampaign took place in different rooms of about 5 times 4m2Figure 9 presents a picture of the experimental setup in anexemplar location

The Tx and Rx modems were connected to two outletsin the same room with distances varying between 2m and8m In general the rooms are equipped with several other

Preamble 119904(119905) 119904(119905) 119904(119905) 119904998400(119905) 119904998400(119905) 119904998400(119905)

Figure 8 HomePlug frame for CTF and EMI measurement afterapplication of TR

Figure 9 Picture of experimentation

electrical outlets (between 4 and 10) About half of the outletswere connected to classical office appliances (lamps desktopsetc) For each topology one CTF wasmeasured first withoutapplying TR and then after applying TR filtering In additionfor each topology between 3 and 5 locations were selectedto measure the received electrical field with the help of thebiconical antenna In total 13 CTF and 43 measurements ofthe electrical field were collected for statistical analysis

4 Results and Statistical Analysis

41 Preliminary Results Thepreliminary results for an exem-plar networkwill be presented in two parts first the CTF at r

0

and second the electric field and its associated power densityat r3= r0

For the CTF at r0 we show its attenuation characteristics

in Figure 10Let us first consider the measured CTF (black curve)The

deep notches at some frequencies are due to reflections at theterminations of the network and reflect the multipath natureof the PLC networkWe define the average attenuation beforeTR119867(r

0) in dB as

119867(r0) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867 (119891 1199030)10038161003816100381610038162119889119891)

(5)

where 119891min and 119891max respectively represent the minimumandmaximum sounded frequenciesThe average attenuationcorresponds to the signal attenuation perceived by a receivercapable of exploiting all the power received in the frequencyband from 119891min to 119891max Typically an OFDM system likethe HomePlug AV specification is able to exploit the totalreceived power over a wide frequency band In our examplethe average attenuation is about 11 dB

Let us now focus on the perceived CTF after applyingTR Owing to the mathematical definition of the TR filterTR allocates more power to frequencies showing minimal

6 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35

05

1015202530354045

Frequency (MHz)

Atte

nuat

ion

(dB)

Measured CTFPerceived CTF after TR

minus5

Figure 10 Channel attenuation before and after TR

0 5 10 15 20 25 30 355

10

15

20

25

30

35

40

Frequency (MHz)

Elec

tric

fiel

d (d

BuV

m)

Before TR AR power density minus1199dB(Wm2)After TR AR power density minus1274dB(Wm2)

Figure 11 Electric field before and after TR

attenuation while strongly attenuated frequencies are morepower constrained In particular for all frequencies where theattenuation of the channel119867(119891 r0) is higher than the averageattenuation 119867(r

0) the perceived channel is more attenuated

after applying TR This can be clearly seen in the frequencyrange from 26MHz to 375MHz For the frequencies wherethe attenuation of the channel is less than119867(r

0) the response

of the channel is improved using TR We can see this clearlyin the frequency range from 11MHz to 22MHz If we definethe average attenuation after TR119867TR(r0) in dB as

119867TR(r0) = 10log10 (1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867TR (119891 1199030)10038161003816100381610038162119889119891)

(6)

we observe that the average attenuation after applying TR isabout 75 dB Hence the application of TR provided a gain inthe total received power of 35 dB in this particular example

Note that only the measured CTF (black curve) is phys-ical Therefore the measured CTF will always degrade the

transmission with some degree of channel attenuation Onthe contrary the CTF perceived at Rx after application of TR(red curve) corresponds to a logical channel where the effectsof the measured CTF are combined with the effects of the TRfilter at Tx Hence the perceived CTF may exhibit some gainover a limited frequency range

We now consider the electrical field and its associatedpower density at r

3= r0 The value of the electrical field

119864(119891 r3) in dB120583Vm was computed from the CTF 119867(119891 r

3)

measured between the Tx balun and the antenna connectorassuming an injected PSD 119875feed = minus55 dBmHz and using thefollowing formula [24]

119864 (119891 r3) = 119875feed + 20 log10 (

1003816100381610038161003816119867 (119891 r3)1003816100381610038161003816) + 107 + AF (119891)

(7)

where AF(119891) represents the antenna factor and 107 representsthe conversion from dBm to dB120583V In addition we computedthe average radiated power density 119878(r

3) in dB(Wm2) as

119878 (r3) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864 (119891 r3)10038161003816100381610038162119889119891)

(8)

where119891min and 119891max respectively represent the minimumandmaximum sounded frequencies and119864(119891 r

3) is expressed

in Vm In (8) the term 120 times 120587 provides the value of theimpedance of free space in Ohm

Note that both 119864(119891 r3) and 119878(r

3) can also be computed

after applying TR filtering using the following equations

119864TR (119891 r3)=119875feed + 20 log10 (1003816100381610038161003816119867TR (119891 r3)

1003816100381610038161003816)+ 107 +AF (119891)(9)

119878TR (r3)

= 10log10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864TR (119891 r3)10038161003816100381610038162119889119891)

(10)

An example of radiated emission measurement is givenin Figure 11 For this particular electrical network the mit-igation of radiated emissions is clear in the frequency bandfrom 26MHz to 375MHz and the average radiated powerdensity has reduced by about 75 dB Thus we can observewith this example that the application of TR filtering canreduce significantly the level of undesired radiated power

42 Statistical Analysis In this section we present the statis-tical analysis of the measurement database collected within13 rooms of the office building at Orange Labs in LannionThe total measurement set is composed of 13 CTF and 43measurements of the electrical field

We first computed the channel gain 119866TR observed on theperceived CTF119867TR(119891 r0) after application of TR filtering AsOFDM systems can exploit the total power received over agiven frequency band we computed this gain in dB for thetotal received power as

119866TR = 119867(r0) minus 119867TR (r0) (11)

Journal of Electrical and Computer Engineering 7

1 2 3 4 5 6 70

0102030405

06

0708

091

Gain of channel (dB)

CDF

Figure 12 CDF of the channel gain 119866TR

0 2 4 6 80

010203040506070809

1

CDF

minus2

Reduction of total average power density (dBWm2)

Figure 13 CDF of the EMI reduction coefficient 119877TR

The cumulative distribution function (CDF) of 119866TR isgiven in Figure 12 This parameter shows always a positivegain between 14 dB and 66 dB in our experiment which isin line with results reported in similar wireless experiments[18] In about 60 of cases the channel gain is higher than3 dBThis means that at Rx we have always a better receptionusing TR This channel gain can in turn be used to reducethe injected PSD at Tx hence reducing by the same factor theunwanted EMI

We then computed the EMI reduction coefficient 119877TRcorresponding to the reduction of the undesired radiatedpower due to the application of a TR filter This figure ofmerit of the reduction of radiated signal is computed in dBas follows

119877TR = 119878(r3) minus 119878TR(r3) (12)

The CDF of the EMI reduction coefficient 119877TR is givenin Figure 13 Results show that the simple application of TRreduces the EMI in more than 60 of cases In the bestcase the EMI reduced by more than 7 dB using TR In theworst case the EMI increased by 2 dB The observationsof particular cases indicates that the reduction of EMI ismore effective when the CTFs 119867(119891 r

0) and 119867(119891 r

3) are

highly decorrelatedThis is more likely to happen in complex

0 2 4 6 8 10 120

010203040506070809

1

CDF

Reduction of total average power density + gain of channel (dBWm2)

Figure 14 CDF of the effective EMI mitigation119872TR

electrical network topologies where the rich multipath envi-ronment results in different frequency fading structures fordifferent Rx locations

Observing the statistics of the channel gain 119866TR and ofthe EMI reduction coefficient119877TR an optimal strategy can beproposed in order tominimize EMI for a PLC system IndeedFigure 12 shows that the application of TR provides bettersystem performance due to the reduced channel attenuationThis gives us in turn the flexibility to reduce the level of theTx power to further reduce EMI while keeping the systemperformance constant More precisely when TR is applieda Tx power back-off of 119866TR dB can be applied withoutmodifying the total received power Finally following thispower backoff strategy the effective EMI mitigation factor119872TR can be computed as the sum of the power backoff andthe EMI reduction coefficient

119872TR = 119866TR + 119877TR (13)

The CDF of the effective EMI mitigation factor 119872TR isdepicted in Figure 14 Several conclusions can be drawn fromthis statistical result

(i) First theTRmethod is able tomitigate EMI generatedby PLC transmission in 100 of our experimentalobservations To this respect one can conclude thatthe gain119866TR provided by the application of TR allowsa Tx power backoff that largely compensates for thepossible EMI increment observed in Figure 13

(ii) Second in 40 of the cases the undesired radiatedpower is reduced by more than 3 dB

(iii) Finally in the most favorable configurations a reduc-tion of the EMI by more than 10 dB can be observedSuch configurations correspond to cases where theCTF between the Tx and Rx modem and the EMIspectrum are particularly decorrelated

5 Conclusion

In this paper we proposed for the first time the application ofTR in order to mitigate EMI generated by wired communi-cation systems TR was originally used in the field of wireless

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

4 Journal of Electrical and Computer Engineering

Tx Rx r0

r1

r2

r3

Figure 3 Principle of the extension of TR to wired transmission

Signal generator Oscilloscope

Biconical antenna

Balun1 Balun2

Mains

Tx Rx

119904(119905) 119910(119905)

ℎ(120591 r0)

ℎ(120591 r3)

Position r0

Position r3

Figure 4 Equipment used in the experimental setup

dedicated to wireless channels [14 15 18] Hence TR appearsas an efficient method to mitigate EMI for wired communi-cations

The theoretical basis for the application of TR to wiredtransmission being set wewill nowdescribe the experimentalassessment of this method in the next Sections

3 Experimental Setup

31 Equipment In order to experimentally assess the use ofTR as a method to mitigate EMI for wired communicationwe used the experimental setup presented in Figure 4 In thissetup a signal generator Tektronix AWG7082C was used as ageneric Tx A Digital Sampling Oscilloscope (DSO) LeCroyWaveRunner 715Zi-A was used to sample the received signalat Rx The signal generated at Tx is denoted 119904(119905) and thesignal received at the DSO is denoted 119910(119905) Two UniversalPLC Couplers were used as baluns to couple the Tx and Rxsignals with the power lines The Tx coupler is situated at theorigin and the Rx coupler is connected to a plug at locationr0 These couplers were developed within the ETSI Specialist

Task Force 410 [24] In order to measure the power densityof the radiated emission received at any arbitrary location r

3

a biconical antenna Schwarzbeck EFS921 was connected to asecond port of the DSO

Mains

Balun1 Balun2

Signal generatorOscilloscope

Coaxial 1

Coaxial 2

Coaxial 3

Coaxial 4

Biconical antennaAttenuator

Amplifier

Figure 5 Experimental setup calibration

32 Calibration With reference to Figure 5 two measure-ment paths can be distinguished Path 1 serves for themeasurement of the CTF and is composed of the followingelements the transmitter coaxial cable 1 a 30 dB amplifierIFIM50 coaxial cable 2 balun 1 mains balun 2 coaxial cable3 a 20 dB attenuator Radial R412720000 and the receiverThebaluns were considered as part of the channelThe calibrationfor measurement path 1 was made by directly connectingcoaxial cable 2 and coaxial cable 3

Path 2 is composed by the transmitter coaxial cable 1a 30 dB amplifier IFIM50 coaxial cable 2 the wire to freespace propagation channel (represented by theCTF119867(119891 r

3))

the biconical antenna coaxial cable 4 and the DSO Thecalibration for this second path was made by connectingcoaxial cable 2 and coaxial cable 4 The gain of the antennawas removed from the measurements by postprocessing

The signal generator and the DSO were synchronized bya direct connection of their 10MHz reference clocks Thesampling rates of both devices were set to 119891

119904= 100MHz

33 Signal Processing After initial system calibration themeasurements are performed in 3 steps

(1) The CTF 119867(119891 r0) is evaluated using a specific Tx

frame (see below) The considered frequency bandextends from 28MHz to 375MHz

(2) The TR filter is generated using the phase and magni-tude of119867(119891 r

0)

(3) The CTFs 119867(119891 r0) and 119867(119891 r

3) as well as the per-

ceived CTFs 119867TR(119891 r0) and 119867TR(119891 r3) are measuredusing a single Tx frame (see below) Measurementsat the location of the Rx outlet (r

0) and at arbitrary

locations in space (r3) are made simultaneously using

two ports of the oscilloscope one connected to theRx balun and the other connected to the biconicalantenna

The Tx signal is generated according to the HomePlugstandard [25] The used frame is called PHY Protocol DataUnit (PPDU) and it is composed by a preamble (including aFrame Control) and a number of payload symbols as shownin Figure 6 Each payload symbol consists of a 3072-sampleOFDMsymbol as defined in theHomePlug specification [25]In order to estimate the CTFs 119867(119891 r) and 119867TR(119891 r) theOFDM symbols were loaded with predefined constellations

Journal of Electrical and Computer Engineering 5

Preamble Symbol 1 Symbol 2 Symbol 3 Symbol 4 Symbol 5 Symbol 6

Figure 6 HomePlug frame

Preamble 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905)

Figure 7 HomePlug frame for calibration and CTF measurementbefore computation of the TR filter

The frame used for the calibration and the initial mea-surement of the CTF before computation of the TR filter isrepresented in Figure 7

The frame used for the measurement of the CTF andEMI after computation of the TR filter is represented inFigure 8 Note that in this frame the TRfilter is applied on the3 last symbols only (represented by the signal 1199041015840(119905)) From thisparticular frame scheme the CTF and EMI can be evaluatedwith and without application of TR quasi-simultaneouslyRecomputing the CTF at this stage also allows monitoringany possible temporal evolution of the channel betweencalibration and measurement

In essence the overall channel estimation process issimilar to the channel sounding procedure defined in theHomePlug AV specification In our experiment the compu-tation of the CTF from the received signal uses a classicalZero Forcing (ZF) channel estimation procedure This issuitable in our experiment involving high levels of Signal-to-Noise Ratio (SNR) at the Rx In practical systems operatingat lower SNR more sophisticated methods such as theMinimum Mean Square Error channel estimation would bemore efficient against noise enhancement Note that we choseto actually implement the TR filter in the time domain atthe transmitter using a programmable waveform generatorin order to measure results as close as possible to a realisticimplementation of the TR scheme in practical PLC modems

In order to estimate channel attenuation continuouslyover the measured frequency band no spectral notcheswere implemented for this study In practical systems a TxPSD mask is defined where the Tx power is notched atpredefined frequencies to protect existing services using thesame spectrumTheHomePlug specification uses windowingin order to better exploit the power allocated within thePSDmask while protecting out-of-band servicesThe presentstudy concentrates on reducing the EMI within the bandeffectively used by PLC systems Therefore the results alsohold for practical systems including notches

34 Measured Environment The measurement campaignwas conducted using 13 different topologies of 230V mainsnetworkswithin the premises ofOrange Labs in LannionThecampaign took place in different rooms of about 5 times 4m2Figure 9 presents a picture of the experimental setup in anexemplar location

The Tx and Rx modems were connected to two outletsin the same room with distances varying between 2m and8m In general the rooms are equipped with several other

Preamble 119904(119905) 119904(119905) 119904(119905) 119904998400(119905) 119904998400(119905) 119904998400(119905)

Figure 8 HomePlug frame for CTF and EMI measurement afterapplication of TR

Figure 9 Picture of experimentation

electrical outlets (between 4 and 10) About half of the outletswere connected to classical office appliances (lamps desktopsetc) For each topology one CTF wasmeasured first withoutapplying TR and then after applying TR filtering In additionfor each topology between 3 and 5 locations were selectedto measure the received electrical field with the help of thebiconical antenna In total 13 CTF and 43 measurements ofthe electrical field were collected for statistical analysis

4 Results and Statistical Analysis

41 Preliminary Results Thepreliminary results for an exem-plar networkwill be presented in two parts first the CTF at r

0

and second the electric field and its associated power densityat r3= r0

For the CTF at r0 we show its attenuation characteristics

in Figure 10Let us first consider the measured CTF (black curve)The

deep notches at some frequencies are due to reflections at theterminations of the network and reflect the multipath natureof the PLC networkWe define the average attenuation beforeTR119867(r

0) in dB as

119867(r0) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867 (119891 1199030)10038161003816100381610038162119889119891)

(5)

where 119891min and 119891max respectively represent the minimumandmaximum sounded frequenciesThe average attenuationcorresponds to the signal attenuation perceived by a receivercapable of exploiting all the power received in the frequencyband from 119891min to 119891max Typically an OFDM system likethe HomePlug AV specification is able to exploit the totalreceived power over a wide frequency band In our examplethe average attenuation is about 11 dB

Let us now focus on the perceived CTF after applyingTR Owing to the mathematical definition of the TR filterTR allocates more power to frequencies showing minimal

6 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35

05

1015202530354045

Frequency (MHz)

Atte

nuat

ion

(dB)

Measured CTFPerceived CTF after TR

minus5

Figure 10 Channel attenuation before and after TR

0 5 10 15 20 25 30 355

10

15

20

25

30

35

40

Frequency (MHz)

Elec

tric

fiel

d (d

BuV

m)

Before TR AR power density minus1199dB(Wm2)After TR AR power density minus1274dB(Wm2)

Figure 11 Electric field before and after TR

attenuation while strongly attenuated frequencies are morepower constrained In particular for all frequencies where theattenuation of the channel119867(119891 r0) is higher than the averageattenuation 119867(r

0) the perceived channel is more attenuated

after applying TR This can be clearly seen in the frequencyrange from 26MHz to 375MHz For the frequencies wherethe attenuation of the channel is less than119867(r

0) the response

of the channel is improved using TR We can see this clearlyin the frequency range from 11MHz to 22MHz If we definethe average attenuation after TR119867TR(r0) in dB as

119867TR(r0) = 10log10 (1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867TR (119891 1199030)10038161003816100381610038162119889119891)

(6)

we observe that the average attenuation after applying TR isabout 75 dB Hence the application of TR provided a gain inthe total received power of 35 dB in this particular example

Note that only the measured CTF (black curve) is phys-ical Therefore the measured CTF will always degrade the

transmission with some degree of channel attenuation Onthe contrary the CTF perceived at Rx after application of TR(red curve) corresponds to a logical channel where the effectsof the measured CTF are combined with the effects of the TRfilter at Tx Hence the perceived CTF may exhibit some gainover a limited frequency range

We now consider the electrical field and its associatedpower density at r

3= r0 The value of the electrical field

119864(119891 r3) in dB120583Vm was computed from the CTF 119867(119891 r

3)

measured between the Tx balun and the antenna connectorassuming an injected PSD 119875feed = minus55 dBmHz and using thefollowing formula [24]

119864 (119891 r3) = 119875feed + 20 log10 (

1003816100381610038161003816119867 (119891 r3)1003816100381610038161003816) + 107 + AF (119891)

(7)

where AF(119891) represents the antenna factor and 107 representsthe conversion from dBm to dB120583V In addition we computedthe average radiated power density 119878(r

3) in dB(Wm2) as

119878 (r3) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864 (119891 r3)10038161003816100381610038162119889119891)

(8)

where119891min and 119891max respectively represent the minimumandmaximum sounded frequencies and119864(119891 r

3) is expressed

in Vm In (8) the term 120 times 120587 provides the value of theimpedance of free space in Ohm

Note that both 119864(119891 r3) and 119878(r

3) can also be computed

after applying TR filtering using the following equations

119864TR (119891 r3)=119875feed + 20 log10 (1003816100381610038161003816119867TR (119891 r3)

1003816100381610038161003816)+ 107 +AF (119891)(9)

119878TR (r3)

= 10log10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864TR (119891 r3)10038161003816100381610038162119889119891)

(10)

An example of radiated emission measurement is givenin Figure 11 For this particular electrical network the mit-igation of radiated emissions is clear in the frequency bandfrom 26MHz to 375MHz and the average radiated powerdensity has reduced by about 75 dB Thus we can observewith this example that the application of TR filtering canreduce significantly the level of undesired radiated power

42 Statistical Analysis In this section we present the statis-tical analysis of the measurement database collected within13 rooms of the office building at Orange Labs in LannionThe total measurement set is composed of 13 CTF and 43measurements of the electrical field

We first computed the channel gain 119866TR observed on theperceived CTF119867TR(119891 r0) after application of TR filtering AsOFDM systems can exploit the total power received over agiven frequency band we computed this gain in dB for thetotal received power as

119866TR = 119867(r0) minus 119867TR (r0) (11)

Journal of Electrical and Computer Engineering 7

1 2 3 4 5 6 70

0102030405

06

0708

091

Gain of channel (dB)

CDF

Figure 12 CDF of the channel gain 119866TR

0 2 4 6 80

010203040506070809

1

CDF

minus2

Reduction of total average power density (dBWm2)

Figure 13 CDF of the EMI reduction coefficient 119877TR

The cumulative distribution function (CDF) of 119866TR isgiven in Figure 12 This parameter shows always a positivegain between 14 dB and 66 dB in our experiment which isin line with results reported in similar wireless experiments[18] In about 60 of cases the channel gain is higher than3 dBThis means that at Rx we have always a better receptionusing TR This channel gain can in turn be used to reducethe injected PSD at Tx hence reducing by the same factor theunwanted EMI

We then computed the EMI reduction coefficient 119877TRcorresponding to the reduction of the undesired radiatedpower due to the application of a TR filter This figure ofmerit of the reduction of radiated signal is computed in dBas follows

119877TR = 119878(r3) minus 119878TR(r3) (12)

The CDF of the EMI reduction coefficient 119877TR is givenin Figure 13 Results show that the simple application of TRreduces the EMI in more than 60 of cases In the bestcase the EMI reduced by more than 7 dB using TR In theworst case the EMI increased by 2 dB The observationsof particular cases indicates that the reduction of EMI ismore effective when the CTFs 119867(119891 r

0) and 119867(119891 r

3) are

highly decorrelatedThis is more likely to happen in complex

0 2 4 6 8 10 120

010203040506070809

1

CDF

Reduction of total average power density + gain of channel (dBWm2)

Figure 14 CDF of the effective EMI mitigation119872TR

electrical network topologies where the rich multipath envi-ronment results in different frequency fading structures fordifferent Rx locations

Observing the statistics of the channel gain 119866TR and ofthe EMI reduction coefficient119877TR an optimal strategy can beproposed in order tominimize EMI for a PLC system IndeedFigure 12 shows that the application of TR provides bettersystem performance due to the reduced channel attenuationThis gives us in turn the flexibility to reduce the level of theTx power to further reduce EMI while keeping the systemperformance constant More precisely when TR is applieda Tx power back-off of 119866TR dB can be applied withoutmodifying the total received power Finally following thispower backoff strategy the effective EMI mitigation factor119872TR can be computed as the sum of the power backoff andthe EMI reduction coefficient

119872TR = 119866TR + 119877TR (13)

The CDF of the effective EMI mitigation factor 119872TR isdepicted in Figure 14 Several conclusions can be drawn fromthis statistical result

(i) First theTRmethod is able tomitigate EMI generatedby PLC transmission in 100 of our experimentalobservations To this respect one can conclude thatthe gain119866TR provided by the application of TR allowsa Tx power backoff that largely compensates for thepossible EMI increment observed in Figure 13

(ii) Second in 40 of the cases the undesired radiatedpower is reduced by more than 3 dB

(iii) Finally in the most favorable configurations a reduc-tion of the EMI by more than 10 dB can be observedSuch configurations correspond to cases where theCTF between the Tx and Rx modem and the EMIspectrum are particularly decorrelated

5 Conclusion

In this paper we proposed for the first time the application ofTR in order to mitigate EMI generated by wired communi-cation systems TR was originally used in the field of wireless

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 5

Preamble Symbol 1 Symbol 2 Symbol 3 Symbol 4 Symbol 5 Symbol 6

Figure 6 HomePlug frame

Preamble 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905) 119904(119905)

Figure 7 HomePlug frame for calibration and CTF measurementbefore computation of the TR filter

The frame used for the calibration and the initial mea-surement of the CTF before computation of the TR filter isrepresented in Figure 7

The frame used for the measurement of the CTF andEMI after computation of the TR filter is represented inFigure 8 Note that in this frame the TRfilter is applied on the3 last symbols only (represented by the signal 1199041015840(119905)) From thisparticular frame scheme the CTF and EMI can be evaluatedwith and without application of TR quasi-simultaneouslyRecomputing the CTF at this stage also allows monitoringany possible temporal evolution of the channel betweencalibration and measurement

In essence the overall channel estimation process issimilar to the channel sounding procedure defined in theHomePlug AV specification In our experiment the compu-tation of the CTF from the received signal uses a classicalZero Forcing (ZF) channel estimation procedure This issuitable in our experiment involving high levels of Signal-to-Noise Ratio (SNR) at the Rx In practical systems operatingat lower SNR more sophisticated methods such as theMinimum Mean Square Error channel estimation would bemore efficient against noise enhancement Note that we choseto actually implement the TR filter in the time domain atthe transmitter using a programmable waveform generatorin order to measure results as close as possible to a realisticimplementation of the TR scheme in practical PLC modems

In order to estimate channel attenuation continuouslyover the measured frequency band no spectral notcheswere implemented for this study In practical systems a TxPSD mask is defined where the Tx power is notched atpredefined frequencies to protect existing services using thesame spectrumTheHomePlug specification uses windowingin order to better exploit the power allocated within thePSDmask while protecting out-of-band servicesThe presentstudy concentrates on reducing the EMI within the bandeffectively used by PLC systems Therefore the results alsohold for practical systems including notches

34 Measured Environment The measurement campaignwas conducted using 13 different topologies of 230V mainsnetworkswithin the premises ofOrange Labs in LannionThecampaign took place in different rooms of about 5 times 4m2Figure 9 presents a picture of the experimental setup in anexemplar location

The Tx and Rx modems were connected to two outletsin the same room with distances varying between 2m and8m In general the rooms are equipped with several other

Preamble 119904(119905) 119904(119905) 119904(119905) 119904998400(119905) 119904998400(119905) 119904998400(119905)

Figure 8 HomePlug frame for CTF and EMI measurement afterapplication of TR

Figure 9 Picture of experimentation

electrical outlets (between 4 and 10) About half of the outletswere connected to classical office appliances (lamps desktopsetc) For each topology one CTF wasmeasured first withoutapplying TR and then after applying TR filtering In additionfor each topology between 3 and 5 locations were selectedto measure the received electrical field with the help of thebiconical antenna In total 13 CTF and 43 measurements ofthe electrical field were collected for statistical analysis

4 Results and Statistical Analysis

41 Preliminary Results Thepreliminary results for an exem-plar networkwill be presented in two parts first the CTF at r

0

and second the electric field and its associated power densityat r3= r0

For the CTF at r0 we show its attenuation characteristics

in Figure 10Let us first consider the measured CTF (black curve)The

deep notches at some frequencies are due to reflections at theterminations of the network and reflect the multipath natureof the PLC networkWe define the average attenuation beforeTR119867(r

0) in dB as

119867(r0) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867 (119891 1199030)10038161003816100381610038162119889119891)

(5)

where 119891min and 119891max respectively represent the minimumandmaximum sounded frequenciesThe average attenuationcorresponds to the signal attenuation perceived by a receivercapable of exploiting all the power received in the frequencyband from 119891min to 119891max Typically an OFDM system likethe HomePlug AV specification is able to exploit the totalreceived power over a wide frequency band In our examplethe average attenuation is about 11 dB

Let us now focus on the perceived CTF after applyingTR Owing to the mathematical definition of the TR filterTR allocates more power to frequencies showing minimal

6 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35

05

1015202530354045

Frequency (MHz)

Atte

nuat

ion

(dB)

Measured CTFPerceived CTF after TR

minus5

Figure 10 Channel attenuation before and after TR

0 5 10 15 20 25 30 355

10

15

20

25

30

35

40

Frequency (MHz)

Elec

tric

fiel

d (d

BuV

m)

Before TR AR power density minus1199dB(Wm2)After TR AR power density minus1274dB(Wm2)

Figure 11 Electric field before and after TR

attenuation while strongly attenuated frequencies are morepower constrained In particular for all frequencies where theattenuation of the channel119867(119891 r0) is higher than the averageattenuation 119867(r

0) the perceived channel is more attenuated

after applying TR This can be clearly seen in the frequencyrange from 26MHz to 375MHz For the frequencies wherethe attenuation of the channel is less than119867(r

0) the response

of the channel is improved using TR We can see this clearlyin the frequency range from 11MHz to 22MHz If we definethe average attenuation after TR119867TR(r0) in dB as

119867TR(r0) = 10log10 (1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867TR (119891 1199030)10038161003816100381610038162119889119891)

(6)

we observe that the average attenuation after applying TR isabout 75 dB Hence the application of TR provided a gain inthe total received power of 35 dB in this particular example

Note that only the measured CTF (black curve) is phys-ical Therefore the measured CTF will always degrade the

transmission with some degree of channel attenuation Onthe contrary the CTF perceived at Rx after application of TR(red curve) corresponds to a logical channel where the effectsof the measured CTF are combined with the effects of the TRfilter at Tx Hence the perceived CTF may exhibit some gainover a limited frequency range

We now consider the electrical field and its associatedpower density at r

3= r0 The value of the electrical field

119864(119891 r3) in dB120583Vm was computed from the CTF 119867(119891 r

3)

measured between the Tx balun and the antenna connectorassuming an injected PSD 119875feed = minus55 dBmHz and using thefollowing formula [24]

119864 (119891 r3) = 119875feed + 20 log10 (

1003816100381610038161003816119867 (119891 r3)1003816100381610038161003816) + 107 + AF (119891)

(7)

where AF(119891) represents the antenna factor and 107 representsthe conversion from dBm to dB120583V In addition we computedthe average radiated power density 119878(r

3) in dB(Wm2) as

119878 (r3) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864 (119891 r3)10038161003816100381610038162119889119891)

(8)

where119891min and 119891max respectively represent the minimumandmaximum sounded frequencies and119864(119891 r

3) is expressed

in Vm In (8) the term 120 times 120587 provides the value of theimpedance of free space in Ohm

Note that both 119864(119891 r3) and 119878(r

3) can also be computed

after applying TR filtering using the following equations

119864TR (119891 r3)=119875feed + 20 log10 (1003816100381610038161003816119867TR (119891 r3)

1003816100381610038161003816)+ 107 +AF (119891)(9)

119878TR (r3)

= 10log10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864TR (119891 r3)10038161003816100381610038162119889119891)

(10)

An example of radiated emission measurement is givenin Figure 11 For this particular electrical network the mit-igation of radiated emissions is clear in the frequency bandfrom 26MHz to 375MHz and the average radiated powerdensity has reduced by about 75 dB Thus we can observewith this example that the application of TR filtering canreduce significantly the level of undesired radiated power

42 Statistical Analysis In this section we present the statis-tical analysis of the measurement database collected within13 rooms of the office building at Orange Labs in LannionThe total measurement set is composed of 13 CTF and 43measurements of the electrical field

We first computed the channel gain 119866TR observed on theperceived CTF119867TR(119891 r0) after application of TR filtering AsOFDM systems can exploit the total power received over agiven frequency band we computed this gain in dB for thetotal received power as

119866TR = 119867(r0) minus 119867TR (r0) (11)

Journal of Electrical and Computer Engineering 7

1 2 3 4 5 6 70

0102030405

06

0708

091

Gain of channel (dB)

CDF

Figure 12 CDF of the channel gain 119866TR

0 2 4 6 80

010203040506070809

1

CDF

minus2

Reduction of total average power density (dBWm2)

Figure 13 CDF of the EMI reduction coefficient 119877TR

The cumulative distribution function (CDF) of 119866TR isgiven in Figure 12 This parameter shows always a positivegain between 14 dB and 66 dB in our experiment which isin line with results reported in similar wireless experiments[18] In about 60 of cases the channel gain is higher than3 dBThis means that at Rx we have always a better receptionusing TR This channel gain can in turn be used to reducethe injected PSD at Tx hence reducing by the same factor theunwanted EMI

We then computed the EMI reduction coefficient 119877TRcorresponding to the reduction of the undesired radiatedpower due to the application of a TR filter This figure ofmerit of the reduction of radiated signal is computed in dBas follows

119877TR = 119878(r3) minus 119878TR(r3) (12)

The CDF of the EMI reduction coefficient 119877TR is givenin Figure 13 Results show that the simple application of TRreduces the EMI in more than 60 of cases In the bestcase the EMI reduced by more than 7 dB using TR In theworst case the EMI increased by 2 dB The observationsof particular cases indicates that the reduction of EMI ismore effective when the CTFs 119867(119891 r

0) and 119867(119891 r

3) are

highly decorrelatedThis is more likely to happen in complex

0 2 4 6 8 10 120

010203040506070809

1

CDF

Reduction of total average power density + gain of channel (dBWm2)

Figure 14 CDF of the effective EMI mitigation119872TR

electrical network topologies where the rich multipath envi-ronment results in different frequency fading structures fordifferent Rx locations

Observing the statistics of the channel gain 119866TR and ofthe EMI reduction coefficient119877TR an optimal strategy can beproposed in order tominimize EMI for a PLC system IndeedFigure 12 shows that the application of TR provides bettersystem performance due to the reduced channel attenuationThis gives us in turn the flexibility to reduce the level of theTx power to further reduce EMI while keeping the systemperformance constant More precisely when TR is applieda Tx power back-off of 119866TR dB can be applied withoutmodifying the total received power Finally following thispower backoff strategy the effective EMI mitigation factor119872TR can be computed as the sum of the power backoff andthe EMI reduction coefficient

119872TR = 119866TR + 119877TR (13)

The CDF of the effective EMI mitigation factor 119872TR isdepicted in Figure 14 Several conclusions can be drawn fromthis statistical result

(i) First theTRmethod is able tomitigate EMI generatedby PLC transmission in 100 of our experimentalobservations To this respect one can conclude thatthe gain119866TR provided by the application of TR allowsa Tx power backoff that largely compensates for thepossible EMI increment observed in Figure 13

(ii) Second in 40 of the cases the undesired radiatedpower is reduced by more than 3 dB

(iii) Finally in the most favorable configurations a reduc-tion of the EMI by more than 10 dB can be observedSuch configurations correspond to cases where theCTF between the Tx and Rx modem and the EMIspectrum are particularly decorrelated

5 Conclusion

In this paper we proposed for the first time the application ofTR in order to mitigate EMI generated by wired communi-cation systems TR was originally used in the field of wireless

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

6 Journal of Electrical and Computer Engineering

0 5 10 15 20 25 30 35

05

1015202530354045

Frequency (MHz)

Atte

nuat

ion

(dB)

Measured CTFPerceived CTF after TR

minus5

Figure 10 Channel attenuation before and after TR

0 5 10 15 20 25 30 355

10

15

20

25

30

35

40

Frequency (MHz)

Elec

tric

fiel

d (d

BuV

m)

Before TR AR power density minus1199dB(Wm2)After TR AR power density minus1274dB(Wm2)

Figure 11 Electric field before and after TR

attenuation while strongly attenuated frequencies are morepower constrained In particular for all frequencies where theattenuation of the channel119867(119891 r0) is higher than the averageattenuation 119867(r

0) the perceived channel is more attenuated

after applying TR This can be clearly seen in the frequencyrange from 26MHz to 375MHz For the frequencies wherethe attenuation of the channel is less than119867(r

0) the response

of the channel is improved using TR We can see this clearlyin the frequency range from 11MHz to 22MHz If we definethe average attenuation after TR119867TR(r0) in dB as

119867TR(r0) = 10log10 (1

119891max minus 119891minint

119891max

119891min

1003816100381610038161003816119867TR (119891 1199030)10038161003816100381610038162119889119891)

(6)

we observe that the average attenuation after applying TR isabout 75 dB Hence the application of TR provided a gain inthe total received power of 35 dB in this particular example

Note that only the measured CTF (black curve) is phys-ical Therefore the measured CTF will always degrade the

transmission with some degree of channel attenuation Onthe contrary the CTF perceived at Rx after application of TR(red curve) corresponds to a logical channel where the effectsof the measured CTF are combined with the effects of the TRfilter at Tx Hence the perceived CTF may exhibit some gainover a limited frequency range

We now consider the electrical field and its associatedpower density at r

3= r0 The value of the electrical field

119864(119891 r3) in dB120583Vm was computed from the CTF 119867(119891 r

3)

measured between the Tx balun and the antenna connectorassuming an injected PSD 119875feed = minus55 dBmHz and using thefollowing formula [24]

119864 (119891 r3) = 119875feed + 20 log10 (

1003816100381610038161003816119867 (119891 r3)1003816100381610038161003816) + 107 + AF (119891)

(7)

where AF(119891) represents the antenna factor and 107 representsthe conversion from dBm to dB120583V In addition we computedthe average radiated power density 119878(r

3) in dB(Wm2) as

119878 (r3) = 10log

10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864 (119891 r3)10038161003816100381610038162119889119891)

(8)

where119891min and 119891max respectively represent the minimumandmaximum sounded frequencies and119864(119891 r

3) is expressed

in Vm In (8) the term 120 times 120587 provides the value of theimpedance of free space in Ohm

Note that both 119864(119891 r3) and 119878(r

3) can also be computed

after applying TR filtering using the following equations

119864TR (119891 r3)=119875feed + 20 log10 (1003816100381610038161003816119867TR (119891 r3)

1003816100381610038161003816)+ 107 +AF (119891)(9)

119878TR (r3)

= 10log10(1

119891max minus 119891minint

119891max

119891min

1

120120587

1003816100381610038161003816119864TR (119891 r3)10038161003816100381610038162119889119891)

(10)

An example of radiated emission measurement is givenin Figure 11 For this particular electrical network the mit-igation of radiated emissions is clear in the frequency bandfrom 26MHz to 375MHz and the average radiated powerdensity has reduced by about 75 dB Thus we can observewith this example that the application of TR filtering canreduce significantly the level of undesired radiated power

42 Statistical Analysis In this section we present the statis-tical analysis of the measurement database collected within13 rooms of the office building at Orange Labs in LannionThe total measurement set is composed of 13 CTF and 43measurements of the electrical field

We first computed the channel gain 119866TR observed on theperceived CTF119867TR(119891 r0) after application of TR filtering AsOFDM systems can exploit the total power received over agiven frequency band we computed this gain in dB for thetotal received power as

119866TR = 119867(r0) minus 119867TR (r0) (11)

Journal of Electrical and Computer Engineering 7

1 2 3 4 5 6 70

0102030405

06

0708

091

Gain of channel (dB)

CDF

Figure 12 CDF of the channel gain 119866TR

0 2 4 6 80

010203040506070809

1

CDF

minus2

Reduction of total average power density (dBWm2)

Figure 13 CDF of the EMI reduction coefficient 119877TR

The cumulative distribution function (CDF) of 119866TR isgiven in Figure 12 This parameter shows always a positivegain between 14 dB and 66 dB in our experiment which isin line with results reported in similar wireless experiments[18] In about 60 of cases the channel gain is higher than3 dBThis means that at Rx we have always a better receptionusing TR This channel gain can in turn be used to reducethe injected PSD at Tx hence reducing by the same factor theunwanted EMI

We then computed the EMI reduction coefficient 119877TRcorresponding to the reduction of the undesired radiatedpower due to the application of a TR filter This figure ofmerit of the reduction of radiated signal is computed in dBas follows

119877TR = 119878(r3) minus 119878TR(r3) (12)

The CDF of the EMI reduction coefficient 119877TR is givenin Figure 13 Results show that the simple application of TRreduces the EMI in more than 60 of cases In the bestcase the EMI reduced by more than 7 dB using TR In theworst case the EMI increased by 2 dB The observationsof particular cases indicates that the reduction of EMI ismore effective when the CTFs 119867(119891 r

0) and 119867(119891 r

3) are

highly decorrelatedThis is more likely to happen in complex

0 2 4 6 8 10 120

010203040506070809

1

CDF

Reduction of total average power density + gain of channel (dBWm2)

Figure 14 CDF of the effective EMI mitigation119872TR

electrical network topologies where the rich multipath envi-ronment results in different frequency fading structures fordifferent Rx locations

Observing the statistics of the channel gain 119866TR and ofthe EMI reduction coefficient119877TR an optimal strategy can beproposed in order tominimize EMI for a PLC system IndeedFigure 12 shows that the application of TR provides bettersystem performance due to the reduced channel attenuationThis gives us in turn the flexibility to reduce the level of theTx power to further reduce EMI while keeping the systemperformance constant More precisely when TR is applieda Tx power back-off of 119866TR dB can be applied withoutmodifying the total received power Finally following thispower backoff strategy the effective EMI mitigation factor119872TR can be computed as the sum of the power backoff andthe EMI reduction coefficient

119872TR = 119866TR + 119877TR (13)

The CDF of the effective EMI mitigation factor 119872TR isdepicted in Figure 14 Several conclusions can be drawn fromthis statistical result

(i) First theTRmethod is able tomitigate EMI generatedby PLC transmission in 100 of our experimentalobservations To this respect one can conclude thatthe gain119866TR provided by the application of TR allowsa Tx power backoff that largely compensates for thepossible EMI increment observed in Figure 13

(ii) Second in 40 of the cases the undesired radiatedpower is reduced by more than 3 dB

(iii) Finally in the most favorable configurations a reduc-tion of the EMI by more than 10 dB can be observedSuch configurations correspond to cases where theCTF between the Tx and Rx modem and the EMIspectrum are particularly decorrelated

5 Conclusion

In this paper we proposed for the first time the application ofTR in order to mitigate EMI generated by wired communi-cation systems TR was originally used in the field of wireless

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 7

1 2 3 4 5 6 70

0102030405

06

0708

091

Gain of channel (dB)

CDF

Figure 12 CDF of the channel gain 119866TR

0 2 4 6 80

010203040506070809

1

CDF

minus2

Reduction of total average power density (dBWm2)

Figure 13 CDF of the EMI reduction coefficient 119877TR

The cumulative distribution function (CDF) of 119866TR isgiven in Figure 12 This parameter shows always a positivegain between 14 dB and 66 dB in our experiment which isin line with results reported in similar wireless experiments[18] In about 60 of cases the channel gain is higher than3 dBThis means that at Rx we have always a better receptionusing TR This channel gain can in turn be used to reducethe injected PSD at Tx hence reducing by the same factor theunwanted EMI

We then computed the EMI reduction coefficient 119877TRcorresponding to the reduction of the undesired radiatedpower due to the application of a TR filter This figure ofmerit of the reduction of radiated signal is computed in dBas follows

119877TR = 119878(r3) minus 119878TR(r3) (12)

The CDF of the EMI reduction coefficient 119877TR is givenin Figure 13 Results show that the simple application of TRreduces the EMI in more than 60 of cases In the bestcase the EMI reduced by more than 7 dB using TR In theworst case the EMI increased by 2 dB The observationsof particular cases indicates that the reduction of EMI ismore effective when the CTFs 119867(119891 r

0) and 119867(119891 r

3) are

highly decorrelatedThis is more likely to happen in complex

0 2 4 6 8 10 120

010203040506070809

1

CDF

Reduction of total average power density + gain of channel (dBWm2)

Figure 14 CDF of the effective EMI mitigation119872TR

electrical network topologies where the rich multipath envi-ronment results in different frequency fading structures fordifferent Rx locations

Observing the statistics of the channel gain 119866TR and ofthe EMI reduction coefficient119877TR an optimal strategy can beproposed in order tominimize EMI for a PLC system IndeedFigure 12 shows that the application of TR provides bettersystem performance due to the reduced channel attenuationThis gives us in turn the flexibility to reduce the level of theTx power to further reduce EMI while keeping the systemperformance constant More precisely when TR is applieda Tx power back-off of 119866TR dB can be applied withoutmodifying the total received power Finally following thispower backoff strategy the effective EMI mitigation factor119872TR can be computed as the sum of the power backoff andthe EMI reduction coefficient

119872TR = 119866TR + 119877TR (13)

The CDF of the effective EMI mitigation factor 119872TR isdepicted in Figure 14 Several conclusions can be drawn fromthis statistical result

(i) First theTRmethod is able tomitigate EMI generatedby PLC transmission in 100 of our experimentalobservations To this respect one can conclude thatthe gain119866TR provided by the application of TR allowsa Tx power backoff that largely compensates for thepossible EMI increment observed in Figure 13

(ii) Second in 40 of the cases the undesired radiatedpower is reduced by more than 3 dB

(iii) Finally in the most favorable configurations a reduc-tion of the EMI by more than 10 dB can be observedSuch configurations correspond to cases where theCTF between the Tx and Rx modem and the EMIspectrum are particularly decorrelated

5 Conclusion

In this paper we proposed for the first time the application ofTR in order to mitigate EMI generated by wired communi-cation systems TR was originally used in the field of wireless

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

8 Journal of Electrical and Computer Engineering

transmission as amean to focus the transmitted signal in bothtime and space around the intended receiverWe proposed touse the same property to focus the signal injected in a wiredmedium such as the electrical network in the case of PLCfor instance As a result the energy lost through undesiredradiation is expected to decrease significantly

We presented an experimental setup in order to demon-strate this method The experiment was conducted in thetime domain using signal frames similar to the industrialspecification HomePlug AV In addition the TR filter wasactually implemented using an arbitrary wave generator thusproviding results encompassing the possible drawbacks of apractical implementation

Results demonstrated that on the wired medium TRcould provide a transmission channel gain between 1 dB and7 dB which is similar to the gain observed in wirelesstransmission In addition the application of a TR filter alonecould effectively reduce EMI for 60 of the observationswith a maximum EMI mitigation of 7 dB Finally by com-bining the channel gain and the EMI reduction features weexperimentally demonstrated that TR was efficient in 100of the observed cases to reduce EMI In 40 of the casesthe EMImitigation was larger than 3 dB withmaxima higherthan 10 dB TR is thus seen as a promising technique to helpresolving EMC issues related to PLC and other wired media

In the TR strategy presented in this paper the CTFgain provided by the TR technique was fully dedicated as apower backoff to minimize the EMI Another strategy couldtarget an increase of the offered capacity while maintaining aconstant EMI Further analyses will therefore be dedicatedto analyze the tradeoff between channel throughput increaseand EMI reduction In addition future research will focuson the study of wired TR at higher frequencies and on othermedia such as DSL cables Finally optimal protocols will bedeveloped to practically implement TR in future standardsIn particular the application of TR to multicast or broadcastscenarios involving one Tx modem and several Rx modemscould be further investigated

References

[1] H C Ferreira L Lampe J Newbury and T G Swart EdsPower Line Communications Theory and Applications for Nar-rowband and Broadband Communications Over Power LinesWiley Chichester UK 2010

[2] IEEE 1901-2010 ldquoIEEE standard for broadband over power linenetworks medium access control and physical layer specifica-tionsrdquo 2010

[3] ITU-T G 9960 ldquoUnified high-speed wireline-based home net-working transceiversmdashsystem architecture and physical layerspecificationrdquo 2010

[4] S Galli A Scaglione and Z Wang ldquoFor the grid and throughthe grid the role of power line communications in the smartgridrdquo Proceedings of the IEEE vol 99 no 6 pp 998ndash1027 2011

[5] V Oksman and J Zhang ldquoGHNEM the new ITU-T standardon narrowband PLC Technologyrdquo IEEE Communications Mag-azine vol 49 no 12 pp 36ndash44 2011

[6] M Ishihara D Umehara and Y Morihiro ldquoThe correlationbetween radiated emissions and power line network compo-nents on indoor power line communicationsrdquo in Proceedings of

the IEEE International Symposium on Power Line Communica-tions and Its Applications (ISPLC rsquo06) pp 314ndash318 Orlando FlaUSA March 2006

[7] Seventh Framework ProgrammeTheme 3 ICT-213311OMEGADeliverable D33 ldquoReport on ElectroMagnetic Compatibility ofPower Line Communicationsrdquo 2009

[8] A Schwager W Baschlin et al ldquoEuropean MIMO PLC FieldMeasurements overview of the ETSI STF410 Campaign ampEMI Analysisrdquo in IEEE International Symposium on Power LineCommunications and Its Applications (ISPLC rsquo12) pp 304ndash309Beijing China March 2012

[9] Federal Communications Commission ldquoTitle 47 of the code offederal regulations part 15rdquo 2007

[10] CENELEC Final Draft European Standard FprEN 50561-1ldquoPower line communication apparatus used in low voltageinstallationsmdashRadio disturbance characteristicsmdashLimits andmethods of measurementmdashPart 1 Apparatus for in-home userdquo2011

[11] A Vukicevic M Rubinstein F Rachidi and J L BermudezldquoOn the impact of mitigating radiated emissions on the capacityof PLC systemsrdquo in Proceedings of the IEEE InternationalSymposium on Power Line Communications and Its Applications(ISPLC rsquo07) pp 487ndash492 Pisa Italy March 2007

[12] P Favre C Candolfi and P Krahenbuehl ldquoRadiation anddisturbance mitigation in PLC networksrdquo in Proceedings ofthe 20th International Zurich Symposium on ElectromagneticCompatibility (EMC Zurich rsquo09) pp 5ndash8 Zurich SwitzerlandJanuary 2009

[13] G Lerosey J De Rosny A Tourin A Derode G MontaldoandM Fink ldquoTime reversal of electromagnetic wavesrdquo PhysicalReview Letters vol 92 no 19 Article ID 193904 4 pages 2004

[14] A E Akogun R C Qiu andN Guo ldquoDemonstrating the lever-ages of time reversal in ultra-wideband communications usingtime domain measurementsrdquo in Proceedings of the 51st Inter-national Instrumentation Symposium pp 737ndash742 KnoxvilleTenn USA May 2005

[15] A Khaleghi G El Zein and I H Naqvi ldquoDemonstrationof time-reversal in indoor ultra-wideband communicationtime domain measurementrdquo in Proceedings of the 4th IEEEInternational Symposium on Wireless Communication Systems(ISWCS rsquo07) pp 465ndash468 October 2007

[16] A Derode P Roux and M Fink ldquoAcoustic time-reversalthrough high-order multiple scatteringrdquo in Proceedings of theIEEE International Ultrasonics Symposium vol 2 pp 1091ndash1094Seattle Wash USA November 1995

[17] D R Jackson and D R Dowling ldquoPhase conjugation in under-water acousticsrdquo Journal of theAcoustical Society of America vol89 pp 171ndash181 1991

[18] P Pajusco and P Pagani ldquoOn the use of uniform circular arraysfor characterizing UWB time reversalrdquo IEEE Transactions onAntennas and Propagation vol 57 no 1 pp 102ndash109 2009

[19] C Zhou and R C Qiu ldquoSpatial focusing of time-reversedUWB electromagnetic waves in a hallway environmentrdquo inProceedings of the 38th Southeastern Symposium on SystemTheory pp 318ndash322 Cookeville Tenn USA March 2006

[20] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to100 MHzmdashpart I one-parameter deterministic modelrdquo IEEETransactions on Power Delivery vol 23 no 3 pp 1392ndash14012008

[21] M Tlich A Zeddam F Moulin and F Gauthier ldquoIndoorpower-line communications channel characterization up to 100

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 9

MHz-part II time-frequency analysisrdquo IEEE Transactions onPower Delivery vol 23 no 3 pp 1402ndash1409 2008

[22] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modeling-part I random topology model and efficienttransfer function computationrdquo IEEE Transactions on PowerDelivery vol 26 no 2 pp 891ndash898 2011

[23] A M Tonello and F Versolatto ldquoBottom-up statistical PLCchannel modelingmdashpart II inferring the statisticsrdquo IEEE Trans-actions on Power Delivery vol 25 no 4 pp 2356ndash2363 2010

[24] ETSI TR 101 562-1 V211 ldquoPowerline telecommunications(PLT) MIMO PLT part 1 measurements methods of MIMOPLTrdquo Tech Rep 2012 Chapter 71

[25] HomePlug ldquoHomePlug AV specification version 11rdquo 2007

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of