Improvement in OFDM using PAPR reduction technique Rasmeet KourDept. of Electronics [email protected] Abstract – This paper portrays critical issue ofhigh PAPR in high speed communication system using Orthogonal Frequency DivisionMultiplexing (OFDM).
OFDM is a special subset of Multicarrier communications,which is based on the principle of transmitting simultaneously many narrow- bandorthogonal frequencies, often called OFDM subcarriers or subcarriers. Ittransmits the large amount of data over radio waves but, the summation of datain time domain OFDM signal results in high PAPR which is the major shortcomingof OFDM system. This high PAPR creates the requirement of PA (Power Amplifier)to be linear in its wide dynamic range and PA’s efficiency is a dependentfactor of its linear dynamic range which decreases as the range increases.
Various techniques have been recommended to reduce PAPR in OFDM. But thereis a hidden cost that has to be paid in terms of payload data rate reduction toreduce high PAPR. This paper enlightens few reduction techniques (Clipping andFiltering, Selective Mapping, Block coding) of PAPR, and their comparativestudy. Keywords: Orthogonal FrequencyDivision Multiplexing (OFDM), Peak to Average Power Ratio (PAPR), Clipping andFiltering, Selective Mapping INTRODUCTIONDue to its intrinsicrobustness against multipath fading channels, Orthogonal frequency divisionmultiplexing (OFDM) has come to the forefront of technology in the field ofWireless communication systems 1. OFDM is a multicarriermodulation technique transmitting large data over radio waves. While sendinghigh rate DataStream, it splits it into number of lower rate streams which aretransmitted over a number of subcarriers in parallel. In a decade, OFDM hasemerged as the standard of choice in a number of important high dataapplications. It finds its application widely being used in Digital TelevisionBroadcasting (such as the digital ATV Terrestrial Broadcasting), EuropeanDigital Audio Broadcasting (DAB) and Digital Video Broadcasting Terrestrial(DVB-T), Wireless Asynchronous Transfer Mode (WATM).
) 2, 3It is based on the principle of transmitting simultaneously many narrow- band orthogonal frequencies, often also called OFDM subcarriersor subcarriers. 1, 4. The number of subcarriers is often noted by N.These frequencies are orthogonal to each other which eliminate the interferencebetween channels. OFDM uses the spectrum much moreefficiently by spacing the channels more closely together. The mathematical anddiagrammatical representation of OFDM in time–domain model is: Where:X(t) – Continuous timebaseband OFDM symbolN – Number of Subcarriersf – SubcarrierspacingT – OFDM symbol period (T = 1/f)Xk – Digitallymodulated data symbol Fig.1.
Baseband OFDM System PAPR- PROBLEM AND REDUCTION REQUIREMENTOFDM is very effectiveand efficient technique in the field of high speed multi-carrier datatransmission. However, OFDM has a drawback too in the sense that it exhibitshigh PAPR (Peak to Average Power Ratio) due to the time-domain superposition ofmany data subcarriers, and thus the resulting time domain signal exhibits theRayleigh-like characteristics and large time–domain amplitude variations5. The output in OFDM is superpositionof multiple subcarriers which may cause some instantaneous power output toincrease and thus become far higher than the mean power of the system. Itrequires high power amplifiers (HPA) to transmit signals with high PAPR 6.These types of PA’s are low efficiency with high cost and, also can lead tononlinear power amplifiers, if the peak power is too high.
PAPR:Insimple, PAPR can be defined as the ratio between the maximum power and theaverage power, and is defined by the equation below: PAPR = =MAX ?Xn?2 / E ?Xn?2 PAPRREDUCTION TECHNIQUESA variety oftechniques have been proposed to reduce PAPR in the past few years and areclassified mainly into two groups – signalscrambling techniques and signaldistortion techniques which are given below 7: A) Signal Scrambling Techniques:· Block Coding Techniques· Block Coding Scheme with ErrorCorrection· Selected Mapping (SLM)· Partial Transmit Sequence (PTS)· Interleaving Technique· Tone Reservation (TR)· Tone Injection (TI) B) Signal Distortion Techniques:· Peak Windowing· Envelope Scaling· Peak Reduction Carrier· Clipping and Filtering Scramblingtechniques scramble the codes to decrease the PAPR. Coding techniques can beused for signal scrambling. Signal scrambling techniques with side informationintroduces redundancy and thus does not result in effective throughput. On theother hand the signal distortion techniques reduce high peaks directly bydistorting the signal prior to amplification.
But they introduce both In-bandand Out-of-band interference and complexity to the system. CLIPPING AND FILTERING – SIGNAL DISTORTION TECHNIQUE FOR REDUCING PAPR INOFDM Clipping being the simple and effective PAPRreduction technique, cancels the signal components that exceed some unchangingamplitude called clip level. Clipping is nonlinear process 8 and causesin-band noise distortion, which causes degradation in the performance of bitBER and out-of-band noise, which decreases the spectral efficiency. However,the clipping introduces signal distortion but this undesirable effect can besuppressed by low pass filtering of clipped signal that unfortunately furtherincreases the PAPR.A method 9 called repeated clipping and filteringthat is based on K-times repetition of the clipping and filteringprocess. Therefore both PAPR and adjacent spectral emissions are reduced,although the PAPR reduction is far from simple clipping case.
But the majordownside of this method is its high complexity. For each frequency domainfiltering, two FFT calculations are necessary. A method named simplifiedclipping and filtering with bounded distortion (SCAFBD) gives almost the samePAPR reduction as repeated clipping and filtering, but the complexity issignificantly reduced. Only 3 FFT’s are required for the PAPR reductionequivalent to iterative method using arbitrary K.METHODDESCRIPTIONFig.2.
Block diagram for SCAFBDFirstly the input data are mapped according to theselected constellation (QPSK). Then, Z zero subcarriers are inserted tozero-pad the signal. Resulting data are transformed using IFFT into timedomain. Subsequently, the signal is clipped to level A. The error signalis computed as the difference between the original and clipped signal. Thiserror signal is transformed back into frequency domain by using FFT andmultiplied by the constant ?,corresponding chosen number of clipping and filtering stages.
The modifiederror signal in the frequency domain is then passed through the block ensuringthe distortion bounding (limitation of I (real) and Q (imaginary) part of theerror signal separately to value D).For the simulation 9, the OFDM signal with 64 datasubcarriers modulated by the QPSK has been used. The signal is 3-timesoversampled by zero-padding, normalized to 0dB in the time domain and clippedto A=3dB. The constant ? isset as the equivalent to 3-times repetition of clipping and filtering process.The distortion bound is derived using constant D=10. Fig.3. CCDF function for PAPR SELECTIVE MAPPING (SLM) AS A SCRAMBLING TECHNIQUE FOR REDUCING PAPR INOFDM In SLM technique 10-11, firstly Uindependent vectors are generated.
Each of these containing N random phasesymbols in order to improve the PAPR of the symbol block that is to betransmitted. Each of these vectors is used to modify the phases of the complex basebandinformation symbols in order to randomize the phases in the block around theunit-circle. Therefore, this process produces U new sets of phase-modifiedsymbol all the new symbols carry the same originalinformation as X, but each with a different phase-mask. After passing each ofthe U modified symbols through the IFFT process, the symbol with the best PAPRperformance will then be selected for transmission.
Although PAPR reductions ofseveral dBs can be achieved by using this method but this method invokes theIFFT operation U times per OFDM-block. Due to this, the system complexitysignificantly increases and hence power consumption and time latency.Fig4. Transmitter Scheme for SLM A novel method Post-IFFT AmplitudeTransforming (PIAT) have been proposed which applies the randomizingprocedure directly in the power spectrum and then generate a correspondingamplitude coefficient sequence for the time domain signal. By doing this,multiple IFFT units and huge amount of searching and computational process canbe avoided and thus reducing system complexity. OVERALL COMPARISONBETWEEN REDUCTION TECHNIQUES PARAMETERS/ REDUCTION TECHNIQUE CLIPPING AND FILTERING SELECTIVE MAPPING DISTORTION LESS NO YES POWER INCREASE NO NO RATE LOSS NO YES COMPLEXITY YES YES PROMISING METHOD SCAFBD PIAT CONCLUSIONOFDM is a veryattractive technique for wireless communications due to its spectrum efficiencyand channel robustness. While selecting PAPR reduction technique, variousfactors have to be considered whichincludes PAPR reduction capacity, Power increase in transmit signal, BERincrease at the receiver, loss in data rate, computational complexity increaseand so on. PAPR reduction technique should be carefully chosen according tovarious system requirements as there is no specific PAPR reduction techniquebest suited for all multi carrier transmission.
The method for PAPR reductionin OFDM signal by the time domain clipping with bounded distortion and SLMprovides good PAPR reduction and increase signal noise immunity. When comparedwith the ordinary SLM technique the PIAT technique is able to achieve hugecomplexity reductions while at the same time has a better overall PAPRreduction ability than the SLM technique. Therefore it is a promising candidatefor future high speed wireless applications, especially for battery poweredmobile terminals. REFERENCES1 Nee,R.
V. and R. Prasad, OFDM for Wireless Multimedia Communications,ArtechHouse Publishers, Norwood, MA,January2000 2 Liu,P.
, “A new peak to average power ratioreduction schemes for OFDM system,” Master dissertation, 20043 Min,K.-S., M.-S. Kim, C.
-K. Park, and M. D. Vu, “Design for PCS antenna based onWiBro-MIMO,” Progress In Electromagnetic Research Letters, Vol. 1,77–83, 20084 E.Telatar, “Capacity of multi-antenna Gaussian channels,”European Transactions on Telecommunications, vol. 10, no 3, Dec 19995 S.
Muller, R. Bauml, R. Fischer, and J. Huber, “OFDM with reduced peak-to-averagepower ratio by multiple signal representation”, Annals of Telecommunications,vol. 53, pp. 58–67, February 19976 A.
Ghassemi and T. A. Gulliver, .Fractional Selective mapping using Decimation intime IFFT/FFT, Publication in WCNC 2008 Proceedings IEEE, pp. 543-5477 KUANGYu-jun, TENG Yong, “A new symbolsynchronization scheme for cyclic prefix based systems,” The Journal ofChina Universities of Posts and telecommunications8 Md. Ibrahim Abdullah, Md. Zulfiker Mahmud, Md.
Shamim Hossain, Md. Nurul Islam, ”ComparativeStudy of PAPR Reduction Techniques in OFDM”, ARPN Journal of Systems and Software, VOL. 1, NO. 8, November20119 Josef Urban, ”PAPRReduction in OFDM system by simplified Clipping and Filtering with BoundedDistortion”,Doctoral Degree Programme (2), FEEC BUT10 LinYang and Emad Alsusa, ”Novel Low-Complexity Post-IFFT PAPR Reduction Techniqueby Utilising Amplitude Transforming for OFDM Systems”,Microwave & Communication Systems Engineering Group, School of Electricaland Electronics Engineering,The University of Manchester, Manchester M601QD,England, United Kingdom11 StefanH.Muller and Johannes B.
Huber, .A Comparison of Peak Power Reduction Schemesfor OFDM., In Proc. of The IEEE Global Telecommunications conference GLOBECOM.97, Phonix, Arizona, USA, pp.
1-5, Nov. 1997