WIMAX, New DevelopmentsdFrequency bands andequency stepChannel bandwidth anfrequency FDD or half duplex FDD HFDDPeak to Average Power Ratio(PAPR) of wavefFDM modulation has many well-known interests for mobileunications: robustnessto multi-path, simple equalization and good spectral efficiency Unfortunately, OFDMaveforms are characterized by high PAPR It is sometimes noted Peak to Mean Powerbefore the pabe definedgnated on rF or baseband signals with3 db difference betweenber of carriers(N), constellation size (M), the shaping filter and the oversampling rat(L) For L=l and M=22n, the baseband PAPRPAPR=10log(N)+10log/2VMFor N=1024rs this value is grean 30 dB but this maximum is not used infor transmitter dal, the PaPr calculated on eaclOFDM symbol, is a random value and its value depends on the statistical distribution ofal amplitudes

It can be analyzed with its Complementary Cumulative Distributioneedsa given threshold The PAPR value that is used in practice is an Effective Power Ratio(EPIassociated with a given probability a(typically a =10-) EPR is the threshold that isCCDF(EPR)= Pr(PAPR 2 EPRMany expree been proposed to approximate the probability distribution of thePAPR of OFDM signals ForxpressionsN√xebe observed in Figure l that the probability that theratio is greater than 12 dBely 10-3 for4 carriers for a full oFDM WiMAX signal with a 1024 FFtproximately 12 dB For the transmitter design this value of 12 dB is used instead of theeoretical maxin

Mobile WiMAX Handset Front-end: Design Aspects and challengesoriginalSLMPTSg 7 The PAPR improvement due to use of PAPR reductionFT size equal to 25623 Integrationn-Chip(Soc)es are developed Their choices shape the design of elementary circuitsThree constraintsessentialarchitecture transceiverImprovingCircuit size and costsTo tackle these challenges, the silicon integration and the packaging must be taken intosilicon SoC technology combines collectively processed componendesignmpetitive prices for major productioThe overall reliability is also improved

But todayThe system must be divided into functions This techniquco-design The chiconstraintsbut the unit cost is higher The reliability is lower and thdecrease depending on the chosen connection technMobile WimaX standard enables a variable channed thereby offers very flexible deployment As a result of multi-wideband ophigh performance RF front-end has to be employed One of the most challenging

WIMAX, New Developmentsto design in a radio front-end is undoubtedly the frequency synthesizer The frequencynthesizer acts as a Local Oscillator(LO) generationo translate basebandes the overaequirements relates to the WiMAX RF architecture WiMAX considers three RFhitectures, TDD, FDD and Half FDD(HFDD) Since the tDd meequency band for the uplink and downlink direction, only one local oscillator is requiredontrary to the TDD, the FDD mode requires two separate synthesizers for rx and tx duecircuitry, higher power consumption and most importantly, to higher costs ofhe tDd architecture is mthe mobilethe wiMAX standardOFDM scheme, which is employed in WiMAX, has become very popular inefficiently combat selective fading resulting from multipath propagation(Cimini, 1985)achieved by parallel leorthogonal sub-carriers equally spaced by Af 1/Ta, where Tu is the useful symbol lengthength of the Fr interval) In the ideal case, all sub-carriers are orthogonal, in other words,th each other

Thereforeused in frequency-dry close together, the OFDM signal becomes less resilient to frequencyDoppler shift and LO phaseluschallik, 1995):(ZThe phase noise introduced by the frequency synthesizer can be interpreted as a parasiphase modulation of the Lb-carriers as a result of the frequency translation (up and down conversions) Thisparasitic phase modulation causes significant degradation of the OFDM signal and may leadharmful as the number of sub-carriers increases in a given bandwidth, becausehe length of the OFDM symbol Tu gets longer, making thephase noise However, longer symbols increase spectral efficiency of an OFDM system, andrs have to beere exist two different effects of the Lo phase noise on the OFDMfrom the loaffects all sub-carriers in the sameand can be observed as arotation of allconstellation points in the 1/Q plane(Muschallik, 1995) Other phenon is ICl, whichresults from the adjacent LO noise contributionCorrelation between the integrated phase noise (phase jitter) and the bEr degradation hasal,2005);( Armada,2001)iminate this impact, phase noise of the OFDM synthesizer has to be optimized accordingphase noise or phase jitter(for a given BER)channel spacing thatponds to the channel raster WiMAX channel profiles aresummarized in tableIMAX Forum, 2008) The channel frequency step given by the

le vWiMAX standard is 250 kHz, however, the LO has to deliver the required frution of 125 khz as a result of rasterlapping for different channel bandwidthWiMAX Forum, 2008) Another directive parameter for design isning rangetching time between rx and txhas to be performed agilely, with respect to settling time requirements of the standarder, the local frequefulfil theghtest signal purity requirements that can be expressed in terms of the integrated phasetput The integrated phase noise is to be less than 1 deg rmcing)to y2 thef the phase noise can start from as lotfew hundred heenefiting from the ease of integration and flexibility, fractional-N Phase Locked lop ponets imposed by the WiMAx whilst at theFractionaL-N PLLand hence improve the phase noise performance by a factor of 20log(N)compared to thenventionalPLL (Keliu Sanchez-sine2 High-Speed Frequency synthesizer Design Exampleapplications because they allow thearison Phase Frequency Detector(PFD) frequency2005) Hisher PFD frequenop bandwidth is to be 01fpFD)quency automatically leads to wider loop bandwidth(as trmance compared to the integer-N PLL

Moreover, since the Mobile WiMAX standardfines the channel rastertion as 125 kHz, the PFD frequency of an integer-Nnthesizer would have to be as low as 125 kHz and hence, the division factor n would(Valenta et2006) This woulethe in-band phase netribution by up to 48 dBmpared to the fractional-N synthesizer with a reference frequ32 MHz (whereN=119, assuming the same phase noise performance of both reference clockIn the next example, we consider a fractional-N Charge Pump(CP) frequency synthesizer asHz channeplified model is depicted in Figure8 This model includes a tri-state PFD that produces output tp and donon sigiroportional to the phase and frequency difference between the reference and the feedbackmnnr地F(1bttpis clocked by frf, the lower by fain Signals up and down are used to switch the currentd tune the vco with tuning gain of 125 MHz/V and tuning range of 3 4-388 GHz, Theactional division is achieved by altering the division value between twoand N+l, hence the average division beca fraction Howerads to a sawtooth phaseswitching between N andthe other hand, it introduces quantization noise into

WIMAX, New DevelopmentsCP/PFD K=lo/2TSwitched Loop Filter F(s)C32 MHZQ34-3,88GHz(5)Fig 8 General model of the fractional-N charge pump synthesizerThe loop filter is the key component of the PLL and it characterizes the dynamicchoosing the proper loop filtequirements set by the Mobile WiMAX standard, ae filter has been chosen thebandwidth and the settling time has to be taken intoaccount for optimal loop filter design A simplified trade-off presented by( Crawford, 199is BPLL=4/ tock, where tioc is the settling time Applied to WiMAX, the PLL would call for atast 200 kHz loop bandwidth in order to settle within 20 us

However, such a wide PLLintegration and phase jitter(Crowley, 1979) The loop filter is switched to the fast wideband mode during the frequransition and then after a certaind, is shifted back to the normalnarrowband value To understand the switching principle, let us have a look at the PLLcontrol theory and the PLL linearized model The effect of a closed feedback loop on thepin can be described by the closed loop transfer function T(s)asKKPowt (sPi(sfactor 1/N Kd is thof the CP/PFD detector and equals to lo/2n, Kow is the vCo gainMHz/V and F(s)refers to the transdance of the second order loop filter(see Figure 8

Mobile WiMAX Handset Front-end: Design Aspects and challengesF(s)scaR2The angular open loomagnitude of the loop gain reaches unity ThisB are defined at theG(sH-1(0 dB), whereS) K, Kom F(s)g K-m F(S)G(sHAnd then, the open loop phase margin Be at the crossover frequency @e readsf zero and theop filter transferOc is increased by factor a in orderbandwidth and thereby decrease the settling time This adiustment isreducing the value of T2 and Ti by the factor a with help of a parallel resistor R,=R2/(a-1)the product of all elements in(7) has to be increased byfactor of a2 as the angular frequency e in()is in the power of two

This can be done b∞和四〔→4為Cs)H20085I0M1(k100k1M10Nquency Offset (Hz)Frequency (z)g 9 Phase noise performance ()and the open loop gain with the phase()at 359 GHzThe blue line corresponds to behaviour during the frequency transition(wideband modelop by factor a In this example, we consie fast wideband mode the cp current is

WIMAX, New Developments16·the dumping resistance by a factor of 4 (by using aarallelRs) The PLL open-loopR2 CiC2/(C +C2)))are all increased by a factor of 4 while the loop stabilitythe constant phase margin in FigurFigureerformance and the resulting phase jitter of the dual bandwidth adjustmentPhase jitter of both loop configurations has been calculated as follows:of theenoise at the PLL output, fi and f2 correspond to the integrationsettlinghat has been achieved by means of loop filter switching Thetiling time has dropped from 88 us to 32 us(settling time within 100 Hz)3522040608100120140Time (us)Tme(嗎sFig 10 Transient responses of the fast PLL for two cases: wideband mode enabled/ disabled(blue/red line respectively) Plot b)displays the absolute frequency3

6 GHZ33 Hybrid PLL Approach for Frequency SynthesisThe product of all elements in equation(7) can be changed not only by increasing the Iop butalso by simultaneous altering of the feed backn factor n and the lo, however, btering the division factor in the feedback path, the output frequency will shift as well Toe hybrid fractionager PLL approach has to be considered(Memmler et al, 2000)oungho et al, 2008) The fractional-N wideband modetching during the frequency transition and then, after settling, the integer-N narrowbandde is turned on This approach brings a new degree of flexibility and alleviates CEby reducing the division ratio Narticularn, the bandaidbe switched only by altering the feedback divisioctor and the dumping resistance while keeping theactional-N mode, the additional dividers (r-labelled blocks in Figure 11nd the reference frequency as wethe feedback signal is applied directly to the PFDAfter settling, the integer-N mode is enabled by switching the loop bandwidth to the normal

Mobile WiMAX Handset Front-end: Design Aspects and challengesband value and by activating two additional dividers The additional division ratiois chosen such that the pfdApplied to the previous example where a 32 MHz reference is used, the division256, in order to achieve 125 kHz resolution The wideband fractionalby switching the resistance R, and by disabling both dividers This adjustment results inoosting thebandwidth by factor 16 while keeping the lo constant The stabilityffected since R R2/(a-1), where256 and at thetime the ratio LGp/N is increasedl/(N/a2)The hybrid approach inherits the speed performance from the fractional-N PLL and at thethe design simplicity ofger- N PLL (the A2 modulator can be replaced byduring the transieeriod) Moreover, a highbandwidth enlargement is possible compared to thenventional bandwidth switching(where only lop and r are altered On the other hand, thedrawback of the hybrid architecture is evident: the phase noise performance of such ands to the perfinteger-N synthesizer, which hampared to the fractional-N architectureue to the high degree of flexibility and integrability, the hybrid PLL approach is a verypromising choice for multi-standard and multi-band transceivers, where different standardspose differentments in terms of phasetiling time or channel raster(Valenta et al

, 2009)CP/PFDISwitched lI32 MHZ34388GHPrescalerN/NFig 11 Functional scheme of the hybrid PLL synthesizercircuits in theGallium Arsenide(GaAs), which allows high operating frequency and high output power toachieved This technology is generally used for powerfiers and switches Theanufacturing cost is higher compared to silicon Using this technology requires a SiPertain conditions, they can achieve good performances, up to tens of GHz A finalinteresting technthe sige that may eventually replace the gaAs for some fune

Mobile WiMAX Handset Front-e

nd: Design Aspects and challengesN=1024N=64x in dBFig 1 Distribution for N=64 and N=1024The Peak transmit power is typically 23 dBm for WiMAX subscriber stationsThe porue to distance variation for example) for wimaX the tpc must be monotonic and able toa range of at least 45 dB by steps of 1 dB with a relative accuracy of 05 dBhe Error vector Magnitude or EVM represents the average deformation of a constellationensation of distortions due to rotation, translation and gain The relativeonstellation error and ratio of error magnitude (or power)given for the different QAM mapping and coding rate: QPSK(1/2), QPSK(3/4),16-QAM(1/2), 16-QAM(3/4),64-QAM(1/2), 64-QAM(2/3), 64-QAM(/4)with respective EVMalues in dB equal to-,-18,205,24,26,28 and 30For larger constellations such as 64-QAM, the distance between the constellation points isFor the transmitter, the EVM value is one of the parameters (with ACLR) that specify thenents of the Pa Ires the inband distortion generated by pAarity and it is more or less difficult to fulfil depending on the signal PAPRer, the EVM value specification has consequences on the acceptablnsider a PAPR of 12dB It is also necessary to add some margin to take blocking signals,or example, into account

WIMAX, New DevelopmentsAdjacent Channel Leakage Ratio(ACLR) or Adjacent Channel Power Ratio(ACPRThe ACLR (or equivalently ACPR)parameter is the ratio of the power in the signal channelhannel and for the adjhannelhe ACLr impacts the allowed channel spacing and adjacent channel interference ACIperformance For WiMAX, the ACLR requirements are given by national regulatory bod(such as FCC, ETSI, TTA)and network service providers It depends on regions, frequencyThe ACLR and EVM requirements of WiMAX combined wR value of 12 dB for theery challenging specifications for linearity and efficioftransmitter Toth a high level of back-off Howeverack-off has two drawbacks

The first is the necessary over-sizing of theansmit a WiMAX signal with a maximal power of 23 d Bm using a PA with a back-off of 8 dBresulting poor efficiency,the linear classr amplifiers have a much better effih aclass ab pa for a wimax signal is smaller than 20s, which is much smaller than whatred for constant envelope signals such as GSM signalsBandsAs indicated in Table 1, there are several possible frequency bands whose bandwidthsde: 100 MHz or 200 MHz The number and wideness of the frequency bands havensequences on the RF filtering(spurening range of the RF blocks: broadband PA, LNA and synthesizer For mobile terminalsESTI specithe level of spurious signals should be below -30 dBm forasurement bandwidth equal to 10 kHz, 100 kHz or 1 MHz depending on the frequencyrange spacing from the carriChannel Bandwidth and Channel Frequency StepThe mobile terminal must be able to deal with the differentfrequencies with steps of half the frequency raster(see Chapter 3)MIMO and smartThe WimAX standarossibility to usetechnology takes advantage of dtheplementing MIMO technology in a mobile WiMAX terminal is a challenge becatze of antennas in the considered frequency bands A WiMAX transceiver can containseveral transmit and receive channels: typically l transmit and 2 receive channels

Mobile WiMAX Handset Front-end: Design Aspects and challengesThe Wimax standard includes two sub-standards the first standard is 80216-2004 or FixedWireless MAN Single Carrier (SC) transmission with QPSK, 16-QAM or 64-QAMadulation schemes Either the Time Division Duplexing (TDD) or the Frequency DivisionThe second standard is 80216e-2005 or Mobile WiMAX, which operates in multiple licensedfor fixed and mobile Non-LOS (NLOS) applications) As in the Fixed WiMAX standard,lexing methods (tdd and FDD) can be used airface deMobile WiMAX are followingireless maN-oFdmSingle carrier based Wireless MAN-SCays Time Division Multiple Access(TDMAusing modulation techniques such as BPSK, QPSK, 16-QAM, 64-QAM and 256-QAMrthogonal Frequency Division Multiplexing(OFDM) and OFDM Access (OFDMA) basedWMAN use 256 OFDM or a scalablewith BPSK, QpSK, 16-QAM or 64-QAM

The channel bandwidth is variable in all of theabove mentioned modes and itDue to a very large variety of multiple variables that can be employed, the WiMAXanisation has established certification profiles, in order to promote the compatibility andinteroperability of wireless communication products The basic radio channel features areammarized in table 1 and will be considered in the rf designData flow rates depend on the modulatiode and the channel bandwidthof 73, 19 Mbit/s for a band of 20 MHz and 256using a modulape 64-QAM Aspower requirements, the minimuis-50 dBm for WMAN-SCa and -45 dBm forWMAN-OFDM and OFDMA, Maxilower at emission must coassification as follows(transmit power for QPSK), (WiMAX FClass2:23≤PdBm)≤P(dBm)≤30Class 4: 30 s P(dBbe characterized by the EVM For example, WiMAX OFDMAn EVM less than 3 16%(for 64-QAM(/4))(IEEE, 2005)

Frequency Channel Frequency ChannelFFT DuplexBandwidthSize g mode23-242502305-23202502345-2362496-269250(200TDD1024TDTable 1 Mobile WiMAX profiles defined by the WiMAX Forum(WiMAX Forum, 2008)2 Front-End Architecture Challenges21 Architectures for wiMAXgnal characteristics are

In this section we focus only on the transmitter where designnallenges are more critical in terms of power, size and consumption A WiMAX transmittedeband OFDM (the bandwidth can be up to 100 MHz) and high PAPR signal (or highdynamic range)in the typical range of 20 dB(29 dB theoretical maximum) The linearizationf the transmittmandatory because power amplification of the WiMAX signalintroduces Non-Linearities(NLs)in amplitude and phase, as illustrated in Figure 2AMAM(conversion)Fig 2 Non Linearity effects of compression and conversion in high power amplificationf architecture for WiMAX requires a careful study of lineahniques and their performance with wideband and high dynamic range signals Thereeral linearization techniques depending on the PAPR of the signal, the added

Mobile WiMAX Handset Front-end: Design Aspects and challengesmplexity and the increase in size and consumption of the system thathe d2007)Miteria characterize the linearization techniquessuch as static/ dynamic processing, adaptability, frequency(digital, baseband, IF or RF),WiMAX application, we basically classify these techniques in three groups: (@)correctiontion of necombination of theoften dedicated to wideband signalExamples of correction techniques are feed-back(A), feed-forward(B)and the anticipationechnique principle of pre-distortion (C)(see Figure 3) Their common point is to addarchitecture considerations here do not include the modulator nor the baseband signalrocessing This needfully derived model of the pa non -linear effects (volterraseries,Wiener or Saleh model etc ) Adaptability to the signal amplitude can bepensate for the lack of accuracy in the NL effects model and memory effectsof the Pa (also temperature drift compensation can be considered)(Baudoin et al

, 20端=HAK(A2-K1)Fig 3 Principles of feed-back (A), feed-forward (B)and pre-distortion(C)namIcThe feed-back can be realised on the amplitude polar feed-backQ quadrature components of the signal(Cartesian feed-back) and both are dedicated tonearization Feed-forward (B)requires a significant increase in signalrocessing and rf blocks in the transmitter, with the hypothesis of a precise matchingtween NLs and reconstructed transfer functions The improvement in linearity will bepaid for in terms of consumption and size (integration) Advantages are thetability and possibility to process wideband signals The most interesting is pre-distortion(C)because of its flexibility: the anticipation can be done in the digital part and so providedaptability of the technique, but this needs a feed-back loop The digital pre-distortion

WIMAX, New Developmentsts a non-negligible additional consumption of a Digital Signal Processor(DsP)andk up table The signal is widened in frequency because of the expensivn-linear law of the pre-distorter (as for IPx theory on a modulated signal spectrume-distortion have been made with OFDM signals in( Baudoin et al, 200high efficiency switched mode RF PAs with constant envelopegnals, avoiding AM/AMAM/PM(Raab et al, 2003);(Diet et al, 2003-2004)Theselete modification of the architecture and its elements specificaRF and power RF

After amplificationth loplification case, while maintaining the efficiency of the architecture Basiction with non Linear Conts (LINC)Envelope Elimination and Restoration(EER) methods(and theirs recent evolutions)(Cox,The LINC principle relies on a decomposition of the modulated signal into tvelope signals as is shown in Figure 4 The decomposition can be computedby combining two Voltage Controlled OscillaCOs)in quadrf instability and additional costs of realisation The amplification of theseAmplifiers(HPAs)at themod n be wide ad often causes signal distortion due to imbalance mismatch Also the HPAhasband because the signal decomposition is a non-linear process, and the phaselation ratio is increasedc(}-0()p(t)-0t)tncs(()+()(t)+(t)Fig 4 LINC decomposition andlificationchnique is(LINC/CALLUM), the default is that efficiencydirectly determined by the recombination sum operation It is very difficult to avoid losses

Mobile WiMAX Handset Front-end: Design Aspects and challenges)+jQ(t=R(t)g 5 Princithe eEr technique(Kahn, 1952)Another decomposition technique was proposed by Kahn in 195basicallyde and pha(EER) This method was first proelope modulated signal (carrying the phase information), enabling theefficiency amplifier(Raab et al, 2003):(Sokal Sokal, 1975):Diet et al, 2005-2008) The difficulty is to reintroduce the amplitude information using the variation of theThis implies a power amplification of therate frequency The recombination can be done with high efficiency switched(saturated)class PA because their output voltage is linearly dependantof the amplitude before the recombination are the two major difficulties of suchlinearizatnanoseconds during the recombination of a 20 MHz OFDM 8021la signal causes spectre-growth of more than 40 dBc(standard limit) at 30 MHz from the carrier frequency(5GHz) Receaal

, 2008-2009) The generation of the amplitude and phase componeof dspspreviously discussed in(Diet et al, 2003), the bandwidths of the envelope and phase signalse widened and make it necessary to design the circuit for three to four times the symbol ratedth on the phase and amplitudebaseband paths in the range of a hundred MHz (as for LINC technique and any other nolipping in frequency anpossible, they are suited for new high data rate standards such as WiMAx, where efficiencyof the emitter and linearization are mandatory Also, the multi-standards and multi-radiconcepts evolved the polar architectures in multiple ways(Diet et al, 2008) For example,the drive signal of the Pa is possible because the amplitudformation modulates the phase RF signal and is restored by the band-pass shape functiof the following blocks: In filter antenna The emitted spectrum is thef quality to be considered carefully, because the Pulse-Width Modulation(PWM)by the advantage of high flexibility of this architecture(Robert et al, 2009) Actual work isfocused on the front end design to provide the highestble with the pa

WIMAX, New Developments如Hheposed by the phase and amplitude coded information The digitallyto rF converter pere key parameters in(Suarez et al, 2008); (Robert et al 2009); Diet et al, 2008)PWM orDAO@D WINHAC hTWiMAX architecture is expected to be wideband and high efficiency due toloment and corresponds to the actual focus in radio-communication research topics Thehigh performance of digital adaptive pre-distortion techniques on OFDM signals should benoted

It is later of interest in a polar/ pre-distorted mixed linearized techni22 PAPR ReductioAs mentioned above, the OFDM signals suffer from the high envelope dynamiharacterized by the PAPR The PAPR reduction methods can be categorized intoeneral groups- the methods introducing signal distortion and the distortionless methodsThe simplest method from the former group is the clipping In its basic form, this methodsulting in out of bandthese effects, the research was directed towards moadvanced methods like repeated clipping and filtering etcMany techniques that do not intthe distortions have been proposed in the past Theased on the creation ofThe one with the lowest papr is selected for trarn In the partial transmitSequences(PTS)(Mauller Huber, 1997)method, the IFFT input symbols are divided intoeral frequency disjoint sub-blocks The output of each sub-blockplied by therotation factor, These factors are optimized in order to find the variant with theThe potential of PAPR improvement can be illustrated on the example resulting from thePTs (4 sub-blocks,e7 for theof FFT size equal to 256