Comparison of Coding Schemes over FRBS Aided AOFDM Systems

Transcription

1 Journal of etwor and Innovatve Computng ISS Volume (03) pp MIR Labs, Comparson of Codng Schemes over FRBS Aded AOFDM Systems Atta-ur-Rahman, Ijaz Mansoor Quresh, Muhammad ahr aseem Baran Insttute of Informaton echnology, PMAS-AA Unversty, Rawalpnd, Pastan Ar Unversty, Islamabad, Pastan Abstract Adaptve technques are becomng demand of almost every modern communcaton system. In ths technque, transmsson parameters are adapted n such a way that the overall throughput of the system s maxmzed and certan constrants are satsfed. In ths paper a smlar constraned optmzaton problem s focused and a fuzzy rule based system (FRBS) s proposed for ts soluton. he transmsson parameters beng optmzed are practcal channel code, modulaton symbol and the transmt power. FRBS suggests the optmal code rate and modulaton symbol whle water-fllng algorthm suggests the power vector accordng to channel condtons at Orthogonal Frequency Dvson Multplexng (OFDM) sub-channels, so that throughput can be maxmzed whle bt error rate (BER) and power reman constraned. Comparson of two channel codes namely Product codes and Convolutonal code s gven whle Quadrature Ampltude Modulaton (QAM) as Modulaton scheme s used under supervson of FRBS. Comparson of both schemes s shown by smulatons. Keywords-component; OFDM; FRBS; BER; ACMP, Modulaton Par, Product s, MIDA, Water-fllng I. IRODUCIO Adaptve communcaton s one of the mportant features of many 3 rd Generaton (3G) and 4 th Generaton (4G) systems. In ths regard, transmtter optmally selects varous transmsson parameters le forward error correcton code (FECC), modulaton sze and power etc wth respect to the varyng channel condtons. If channel condtons are poor then a relatvely lower code rate, and smaller modulaton symbol and better power can be used. Smlarly a comparatvely hgh code rate, a larger modulaton symbol and a moderate power can be used. OFDM systems have been used n almost all wreless standards le IEEE80. [] and IEEE80.6 [] etc. In OFDM a hgh data stream s dvded nto many relatvely low data streams by tang Inverse Fast Fourer ransform (IFF), and then these streams are modulated over orthogonal subchannels. he orthognalty of subcarrers and addton of a sutable cyclc prefx (CP) maes the system less vulnerable to nter symbol nterference (ISI). In lterature many adaptve power and bt loadng technques have been proposed. he dea of adaptve modulaton has been proposed n late 90s; however consderaton of a practcal channel code and other parameters s a recent approach. Kallet, n 989 [3], proposed adaptve modulaton for OFDM systems. Chow et al [4] nvestgated same technque for Gaussan slowly varyng dspersve channel and found t qute applcable. he same dea of adaptve modulaton for wdeband rado channel was nvestgated by Cyzlw n 996 [5]. In 009, Shastr et al [6] proposed adaptve modulaton usng Fuzzy Logc nterface. A turbo coded adaptve modulaton scheme was proposed by Hanzo et al [7], where there sgnal to nose rato (SR) thresholds were used for adaptng modulaton. d bt and power loadng problem for sngle antenna OFDM systems, was addressed by L et al, Low Densty Party Chec s (LDPC) were examned [8] orgnally motvated by [9]. A common observaton n above cted wor was that n these systems one of two parameters was fxed whle other was adaptve. hat s ether modulaton was adaptve wth a fxed channel code [7] [8] [9] or channel code was adaptve and modulaton was fxed [0]. Atta-ur-Rahman et al. proposed a fuzzy rule base system (FRBS) for adaptve codng and modulaton n OFDM systems n []. In ths paper FRBS chooses the best modulaton code par for a gven set of channel condtons at dfferent subchannels n OFDM. Product codes were used as FECC and QAM as modulaton scheme. In ths scheme power dstrbuton was assumed fxed. Scheme was examned and compared wth HYPERLA/ standard and results shown n terms of smulatons. In [], same authors examned the role of FRBS for IEEE 80.n (WIFI) standard. In ths paper, Convolutonal codes wth constraned length 3 were used as forward error correcton code and QAM as modulaton scheme. Power dstrbuton was consdered fxed over all subchannels. he sgnfcance of proposed scheme was shown over varous schemes of same area le the one proposed by Bocelmann et al. n [3]. In whch a Bsecton method was used for selecton of optmum modulaton code par (MCP). Dynamc Publshers, Inc., USA

2 84 Atta-ur-Rahman et al. In [4], same authors proposed an FRBS based adaptve codng, modulaton and power scheme, n whch all three parameters were adapted. FRBS was used for adaptng code rate and modulaton scheme whle Water-fllng prncple was used for adaptng the power vector for all subcarrers at OFDM system. In ths paper Product codes were used as FECC and QAM and modulaton scheme. In ths paper, we have presented the comparson of two FRBS based adaptve codng and power schemes. In these schemes FRBS s used for adaptng modulaton and code rate whle Water-fllng prncple s used for adaptng the power vector. In one scheme Product codes are used for FECC whle n other scheme Convolutonal codes was used. Rest of the paper s organzed as follows. System model s gven n secton ; secton 3 contans a bref ntroducton of Product s and ther decodng; coded modulaton and smulatons for varous modulaton code pars n gven n secton 4; secton 5 contans rate optmzaton crtera and cost functon to be optmzed, n secton 6 and 7 Fuzzy Rule Base System s and water-fllng algorthm are explaned respectvely, smulaton results of proposed scheme are depcted n secton 8 whle secton 9 concludes the paper. II. SYSEM MODEL he system model consdered s OFDM equvalent baseband model wth number of subcarrers. It s assumed that complete channel state nformaton (CSI) s nown at both transmtter and recever. he frequency doman representaton of system s gven by y h. p. x z ;,,..., () where y, h, p, x and z denote receved sgnal, channel coeffcent, transmt ampltude, transmt symbol and the Gaussan nose of subcarrer,,...,, respectvely. he overall transmt power of the system s P total p and the nose dstrbuton s complex Gaussan wth zero mean and unt varance. It s assumed that sgnal transmtted on the th subcarrer s propagated over an ndependent non-dspersve sngle-path Raylegh Fadng channel and where each subcarrer faces a dfferent amount of fadng ndependent of each other. Hence, the channel coeffcent of th subcarrer can be expressed as: j h e ;,,..., () where s Raylegh dstrbuted random varable of th subcarrer, and the phase s unformly dstrbuted over 0,. much shorter consttuent bloc codes are used nstead of one long bloc code. Bascally these are matrx codes where rows are encoded by one bloc code whle columns are encoded by another bloc code. hs arrangement enhances ther error correcton capablty snce errors are corrected row-wse as well as column-wse. Also these codes are burst error correctng codes snce a row-wse burst can easly be corrected column-wse and vce versa. Snce burst error n rows wll become sngle error for column code and vce versa. Consder two lnear bloc codes A and A wth parameters [ n,, d] and [ n,, d ] respectvely, where n, and d ;, are the length, dmenson and mnmum Hammng dstance d mn of the code A (,) respectvely. A wll be used as row code whle A wll be used as column code. he rates of ndvdual codes are R and Rrespectvely gven by, R,, (3) n he product code Φ can be obtaned by codes A,, n the followng manner. Place nformaton bts n an array of rows and columns Encode rows usng code A, whch wll result n an array of n ow encode ncolumns usng code A, whch wll result n nnproduct code. he resultant product code Φ has the parameters [ nn,, dd ] and the rate wll be RR. In ths way long bloc codes can be constructed usng much shorter consttuent bloc codes. hs concept can also be vewed as ' that product code Ω s ntersecton of two codes A and A '. ' Where A s a code represented by all nnmatrces whose ' each row s a member of code A, smlarly A s a code represented by all nnmatrces who s each column s a member of code A. hs can be wrtten as; Φ A A (4) ' ' III. PRODUC CODES AD MODIFIED IERAIVE DECODIG ALGORIHM A. Product s Product codes are serally concatenated codes that were frstly presented by Elas n 954 [5]. he concept of Product codes s qute smple as well as powerful, where

3 Comparson of Codng Schemes over FRBS Aded AOFDM Systems 85 Sr Row Column Product rate Error Correctng Capablty C3 [63,5,5] [63,63,] [3969,33,5] 0.8 C4 [63,36,] [63,63,] [3969,68,] C5 [63,63,] [63,57,3] [3969,359,3] 0.9 C6 [63,57,3] [63,57,3] [3969,349,9] C7 [63,5,5] [63,57,3] [3969,907,5] C8 [63,36,] [63,57,3] [3969,05,33] Fgure. Structure of the Product code B. Modfed Iteratve Decodng Algorthm Iteratve decodng algorthm (IDA) for product codes was orgnally presented by [0] n hs Doctoral thess that s based upon Lst Decodng also desgnated as Maxmum Lelhood (ML) decodng of product codes. ML decodng s an optmum decodng wth an exponental complexty. he teratve decoder was proposed to reduce the complexty of ML decodng, but yet t exhbts a huge complexty. Modfed Iteratve Decodng Algorthm (MIDA) s modfcaton of IDA n whch complexty of basc algorthm s sgnfcantly reduced by usng concept of Syndrome Decodng to overcome the search space radus. he decoder s conssted of two sub-decoders namely row-decoder and column-decoder both placed n successon. Interested readers may vst orgnal paper by the same author [6]. IV. CODED MODULAIO here are two modulaton codng schemes nvestgated. here detals s gven as follows. A. Frst Scheme Codng schemes used for ths framewor are set of product codes. Snce product codes are matrx codes, where rows contan one code and column contans another code. he set of row codes and column codes used n ths paper are lsted n table. All of these codes are BCH codes. So set of code s ntally conssted of four dfferent product codes. hat s C { C }; 8 (5) he error correctng capablty of a bloc code can be found by the followng equaton. d mn t (6) In frst four codes that s C to C4 column code s consdered as rate that s [63, 63, ] whle n last four codes that s C5 to C8 [63, 57, 3] s consdered as column code. he modulaton scheme used for ths experment s Quadrature Ampltude Modulaton (QAM) whch s recommended by many OFDM standards. Followng set of modulaton symbols s used. hat s M {, 4,8,6,3, 64,8} (7) So wth these codng and modulaton sets we have 8 possble modulaton code pars (MCP) by a Cartesan product of the sets C and M. P Cx M {( c, m ) c C, m M } (8) B. Second Schemes j j he codes used n ths scheme are non-recursve convolutonal codes wth code rates taen from the set C {/ 4,/3,/,/3,3/ 4} wth constrant length 3. For decodng, standard soft output Vterb decoder s used. For convenence, message bts length s taen equal to the number of subcarrers and standard OFDM nterleaver/denterleaver are used n smulatons. In ths scheme the symbols taen from Quadrature Ampltude Modulaton (QAM), wth rectangular constellaton. he modulaton symbols are taen from the set M = {,4,8,6,3,64,8}. All of the possble combnatons of modulaton code pars n both schemes are plotted usng the sequence shown n Fg-. Few of these graphs are shown n subsequent fgures. ransmtter Encoder Modulator Fgure. Smulaton Model AWG Channel Recever Decoder Demodulator ABLE I. CODIG PARAMEER Sr Row Column Product rate Error Correctng Capablty C [63,63,] [63,63,] [3969,3969,] 0 C [63,57,3] [63,63,] [3969,359,3] 0.9

4 86 Atta-ur-Rahman et al. Fgure 3: Performance of dfferent QAM schemes usng C as product code from frst scheme Fgure 6. BER comparson of dfferent QAM modulatons usng rate /3 convolutonal code from second scheme Fgure 4: Performance of dfferent QAM schemes usng C4 as product code from frst scheme Fgure 5. BER comparson of dfferent QAM modulatons usng rate /4 convolutonal codes from second scheme Fgure 7. BER comparson of dfferent QAM modulatons usng rate / convolutonal code from second scheme V. RAE OPIMIZAIO In order to maxmze the rate for OFDM system followng constraned optmzaton problem s consdered n frst scheme. max R R log ( M ) s.t, BER and otal otal BER P p P otal (9) where R s the rate of product code used from set C and M s the sze of modulaton symbol used from set M. P s the avalable transmt power and p s power transmtted per subcarrer. BER s target BER that depends upon a specfc qualty of servce (QoS) request or applcaton demand. he

5 Comparson of Codng Schemes over FRBS Aded AOFDM Systems avalable QoS are BER 0,0,0,0 whle s total number of subcarrers n OFDM system. he above cost functon s optmzed by the proposed Fuzzy Rule Base System. Smlarly, n second scheme followng cost functon s consdered. max s.t, n n n R otal QoS n r n n BER BER and p P (0) where rn (log ( M )) n RC, n s bt rate of nth subcarrer, whch s product of code rate and modulaton order used, P s the avalable transmt power and BER s target QoS n BER that depends upon a specfc qualty of servce (QoS) request or applcaton requrement over nth subcarrer, whle s number of subcarrers n OFDM system. VI. FUZZY RULE BASE SYSEM A fuzzy rule base system (FRBS) s proposed, whch s capable of decdng the best modulaton code par (MCP) for the next transmsson nterval, based upon the heurstcs. Fuzzy logc s recommended for the stuatons that are vague, ambguous, nosy or mssng certan nformaton. here are many ways to buld a Fuzzy Rule Base System, we have used table looup scheme for ths purpose. he looup table s gven n fg-8 that s used for creaton. A. Obtanng Graphs Graphs for dfferent combnatons of s and Modulaton schemes are obtaned; few of them are depcted n Fg-3 to 5 accordng to frst scheme and n Fg-6 and 7 accordng to second scheme. B. Data Acquston Data s obtaned from the graphs n terms of nput/output (IO) pars. hs s taen by puttng the horzontal lnes for varous bt error rates and then ponts of ntersecton of these lnes wth the curves are noted and stored n a table. hs process s shown n Fg-8. C. Rule Formulaton Rules for each par are obtaned by the approprate fuzzy set used. hat s by puttng complete par n nput/output set and then a rule generated for each par accordngly. hs process of rule formulaton can be seen n detal n [7]. D. Elmnaton of Conflctng Rule he rules havng same IF part but dfferent HE parts are nown as conflctng rules. hs appears when more than one modulaton code par (MCP) are avalable for gven specfcaton. For example n frst scheme, there s a rule whose HE part contans two dfferent MCP namely, [3, C3] and [3, C4]. ow [3, C3] s best snce ts throughput s 5x0.8=4 b/s/hz whle others have 5x0.57=.58 respectvely. Smlarly, sometme there could be two dfferent pars wth same throughput le [3,C3] and [6,C] both have same throughput that s 4 b/s/hz, then [6,C] wll be chosen snce t exhbts less modulaton/demodulaton and codng/decodng cost. Smlarly, n second scheme, there s a rule whose HE part contans three dfferent MCP namely, [8,/], [6,/3] and [6,3/4]. ow [6,3/4] s best among the rest snce ts throughput s 4x3/4=3 whle others have 3x/=.5 and 4x/3=.67 respectvely. Smlarly, sometme there could be two dfferent pars wth same throughput le [,/] and [4,/4] both have same throughput that s x/=0.5, then [,/] wll be chosen snce t exhbts less modulaton/demodulaton and codng/decodng cost. E. Completon of Looup able Snce n looup table scheme we may not have complete number of IO pars, then those parts are flled by heurstc or expert nowledge. For example, a modulaton code par s suggested by rule for a certan SR and QoS. Snce f a modulaton code par performs for lower SR, then t can easly sustan n hgher SR stuatons. Smlarly, f a MCP performs for a good QoS then t can sustan for poor QoS demands. For example n second scheme, let [8,3/4] be 3 suggested for 5dB SR and BER 0 then ths par can be used for 6-30dB SR and 0 BER cases as well. F. Fuzzy Rule Base Creaton Usng the Looup table n above phase Fuzzy Rule Base s created usng Fuzzy Logc oolbox n MALAB. Further detals are gven ahead. Fgure 8. Process of obtanng IO pars

6 88 Atta-ur-Rahman et al. Components of Fuzzy Rule Base System A. Looup table hs table shows the facts extracted for smulated performance of dfferent codes and modulaton pars n prevous secton. It can be stated as for a gven receved SR and a fxed QoS, whch MCP maxmzes the throughput. In frst scheme receved sgnal to nose rato (SR) s expressed n level to level 9 and Qualty of Servce (QoS) are gven le poor, med, good and hgh that s 3 4 0,0,0,0 respectvely. In second scheme, the receved SR s expressed n level 0 to level 30 and QoS s gven n fve dfferent levels that are BER 0,0,0,0. he nput-output pars needed for desgn of FRBS are of the form; p p p ( x, x ; y ); p,,3... n (0) p p where x represents receved SR, x represents requred p BER (QoS) and y represents the output MCP suggested by FRBS, so the rule format wll be gven as below; {IF ( x s Good and x s L7) HE y s P5} Fgure 9. Looup able for FRBS Creaton n frst scheme he looup table for the frst scheme s shown n Fg-9 whle for the second scheme t s shown n Fg-0. Fgure 0. Looup able for FRBS Creaton n second scheme B. Fuzzy Sets Suffcent numbers of fuzzy sets are used to cover the nput output spaces. here are two nput varables average receved SR and mnus log bt error rate (MLBER) that represents a QoS. he reason tang MLBER s because 3 4 BER of a requred QoS s gven by 0,0,0,0 etc whle the range of fuzzy varable should be equally spaced and quantfable. So to get ths, followng operaton s done frst. MLBER log( BER ) BER 0 q () q MLBER log(0 ) q here s one output varable for modulaton code par MCP. In frst scheme, there are nne, four and eghteen fuzzy sets used for the nput varables SR, MLBER and output varables MCP, respectvely. hese are shown n fgures In second scheme there are thrty one, sxteen and twentyfve fuzzy sets are used for the nput varables SR, MLBER and output varables MCP, respectvely. C. Fuzzfer & De-Fuzzfer Standard trangular fuzzfer s used wth AD as MI and OR as MAX. Standard Center Average Defuzzfer (CAD) s used for defuzzfcaton. D. Rule Base & Inference Engne Rule base contans rules aganst all the IO pars. In frst scheme, there are nne sets (L to L9) for frst nput varable named receved SR and about four sets (low, medum, good and hgh) for nput varable MLBER. Hence there are 36 rules n rule base. hs s shown n Fg- to Fg-3.

7 Comparson of Codng Schemes over FRBS Aded AOFDM Systems 89 Smlarly, n second scheme there are thrty-one sets (L0 to L30) for frst nput varable named SR and about sxteen sets (Q to Q6) for nput varable MLBER. Hence there are 496 rules n rule base. hese membershp functons are shown n Fg-4 to Fg-6. Rule base s complete n a sense that rules are defned for all possble combnatons of nput space. Standard Mamdan Inference Engne (MIE) s used that wll nfer whch nput par wll be mapped on to whch output pont. Fgure 5. Membershp functon of second nput varable QoS Fgure 6. Membershp functon of output varable MCP Fgure. Fuzzy Sets for nput varable average receved SR Fg-7 and Fg-8 show the rule surfaces for frst and second schemes respectvely. A rule surface shows that by varyng SR and QoS the throughput s vared. For the hghest value of SR and lowest value of QoS, throughput of the system approaches to 5bts/s/Hz. Fgure. Fuzzy Sets for nput varable QoS Fgure 7. Rule surface for frst scheme Fgure 3. Fuzzy Sets for output varable Modulaton Par Fgure 4. Membershp functon of frst nput varable SR Fgure 8. Rule surface for second scheme

8 90 Atta-ur-Rahman et al. VII. HE WAER-FILLIG ALGORIHM Water-fllng prncple has been wdely used for multcarrer loadng problems. It s restated here for sae of reference. Maxmze the bt rate Rotal for the entre multchannel transmsson system; through an optmal sharng of the total transmt power P between the sub-channels, subject to the constrant that P s mantaned constant. In contrast to our system ths phenomenon can be wrtten mathematcally as, p K ; () H( f ) Where p s transmt power, s nose varance (power) and H( f ) magntude response at th subchannel respectvely. he choce of constant K depends upon applcaton and t s under desgner control. hat s, the sum of the transmt power and nose varance (power) scaled by nverse of square of channel (subchannel) magntude response must be mantaned constant for each subchannel. hs can also be wrtten as; p K ; (3) ( CR ) Where H( f ) ( CR ) (4) the value of the constant K s calculated analytcally by [8] K Pavg (5) ( CR ) where P avg s the average transmt power. VIII. RESULS A. Frst Scheme In ths secton both schemes are compared wth each other as well as the well-nown adaptve technques. Smulaton parameters are enlsted n table. In fg-9, frst scheme s plotted for varous qualty of servce (QoS) le average BER=0e-, 0e-, 0e-3 and 0e-4. In ths way QoS was fxed ntally then dependng upon the receved sgnal to nose rato (SR), most approprate modulaton code par (MCP) was chosen usng Fuzzy Rule Base System (FRBS), for entre OFDM system, then the product of modulaton rate and code rate so called modulaton-code-product s consdered as throughput s plotted, whle the power consderaton was fxed. In fg-0, adaptve power usng water-fllng prncple s consdered. In fg-, proposed scheme s compared wth the Adaptve Codng scheme proposed by Al-Asary n ths PhD dssertaton [0], n whch HYPERLA/ standard was focused and adaptaton crtera was based upon SR based thresholds. As smulaton results reveal, proposed scheme profoundly performs better than that of proposed by Al- Asary as well as HYPERLA/ standard. Also proposed adaptve power outperforms that of the fxed power. Fgure 9. Comparson of proposed scheme for varous QoS wth fxed power Fgure 0. Comparson of proposed scheme for varous QoS wth adaptve power Fgure. Comparson of proposed scheme wth dfferent schemes

9 Comparson of Codng Schemes over FRBS Aded AOFDM Systems 9 B. Second Scheme In ths secton we compared both schemes at dfferent qualty of servce. Fg- shows the comparson of both schemes for QoS=0e-. hs graph reveals that second scheme outperforms compare to frst scheme as well as fxed power scheme. Smlarly, n Fg-3 and Fg-4 both schemes are compared wth each other as well as wth adaptve codng and modulaton and fxed power case. In all of these graphs t s apparent that second scheme s sgnfcantly better than frst scheme and fxed power scheme. he reason that second scheme s better than frst s because n frst scheme once we choose a modulaton code par usng fuzzy rule base system, t remans same for all subcarrers. Whle n second scheme each subcarrer may have a dfferent modulaton code par after each transmsson nterval. hs dfference can be notced n equ- 9 and equ-0. Fgure 4. Comparson of both schemes at QoS=0e-4 ABLE II. SIMULAIO PARAMEERS Sr Parameter name Value Codng Schemes Product s, Convolutonal s rates n Frst Scheme, 0.9, 0.8, 0.57 (C to C4) 0.9,0.8,0.7,0.5 (C5 to C8) 3 rates n Second Scheme 0.5, 0.33, 0.5, 0.67, Modulaton Scheme, 4, 8, 6, 3, 64, 8 QAM 5 Bts/symbols n modulaton,, 3, 4, 5, 6, 7 6 otal MCPs 8x7=56 7 OFDM Standard used HYPERLA/ 8 umber of subchannel 63 9 Adaptaton Modulaton, code and power 0 Adaptaton Crtera Fuzzy Rule Base System+Waterfllng Fgure. Comparson of both schemes at QoS=0e- Fgure 3. Comparson of both schemes at QoS=0e-3 IX. COCLUSIOS In ths paper two adaptve codng, modulaton and power schemes are compared. Frst scheme uses Product codes as FECC whle second scheme uses Convolutonal codes; remanng all parameters are ept same. hs comparson was to show that whch codng scheme n contrast to FRBS exhbts better spectral effcency n an adaptve OFDM (AOFDM) envronment. Quadrature Ampltude Modulaton s used as modulaton scheme whle Water-fllng prncple s used for adaptng power n each case. he frst scheme was tested for OFDM HYPERLA/ standard and compared to a smlar wor namely Adaptve Codng for OFDM System [0] and comparson of the scheme s shown by smulatons. Second scheme perform better compare to frst scheme because n second scheme the decson of modulaton code par s based upon ndvdual channel condtons and each subcarrer may have a dfferent MCP. In frst scheme, however, the decson s based on average channel condton and same MCP s used for all subcarrers of OFDM system. Sgnfcance of the proposed schemes s due to the followng factors,

10 9 Atta-ur-Rahman et al.. Wde range of Product codes and convolutonal codes are nvestgated. Both codng schemes are beng used n practcal IEEE standards le HYPERLA/ and IEEE80.n (WIFI). Fuzzy Rule Base System s to choose sutable most combnaton of code and modulaton scheme based upon a specfc Qualty of Servce and average receved channel to nterference nose rato (channel state). 3. Water-fllng prncple n conjuncton wth FRBS to adapt power vector for all subcarrers. 4. A relatvely low complexty decoder (MIDA) of product codes 5. Combned adaptve codng, modulaton and power ACKOWLEDGME hs research wor was supported by Hgher Educaton Commsson (HEC) of Pastan. REFERECES [] IEEE Std 80.6M-004, Part 6: Ar nterface for fxed broadband wreless access systems, Oct 004. [] IEEE Std 80.6Etm-005, Part 6: Ar nterface for fxed and moble broadband wreless access systems, Feb [3] I. Kalet, he multtone channel, IEEE ran. Commun., vol. 37, pp.9 4, Feb [4] P. S. Chow, J. M. Coff, and J. A. C. Bngham, A practcal dscrete multtone transcever loadng algorthm for data transmsson over spectrally shaped channels, IEEE rans. Commun., vol. 48, pp , 995. [5] A.Cyzlw, Adaptve OFDM for wdeband rado channels, Global elecommuncatons Conference, vol., pp , ov 996. [6] K. Seshadr Sastry, Fuzzy logc based Adaptve Modulaton Usng on Data Aded SR Estmaton for OFDM system, Internatonal Journal of Engneerng Scence and echnology (IJES) Vol. (6), 00, [7]. Keller and L. Hanzo, Adaptve modulaton technques for duplex OFDM transmsson. IEEE ransactons on Vehcular echnology, Vol.49, no.5, pp , 000 [8] Y. L and W.E. Ryan, Mutual-Informaton-Based Adaptve Bt- Loadng Algorthms for LDPC-d OFDM, IEEE rans. on Wreless Communcatons, vol. 6, no. 5, pp , May 007 [9] G. Care, G. arcco, and E. Bgler, Bt-Interleaved d Modulaton, IEEE rans. on Informaton heory, vol. 44, no. 3, pp , 998. [0] O. Al-Asary, Codng and teratve decodng of concatenated multlevel codes for the Raylegh fadng channel, n Doctoral thess n Rado communcaton systems, Stocholm, Sweden: KH Informaton and Communcaton echnology, 006. [] Atta-ur-Rahman, Quresh I.M. and Muzaffar M.Z. Adaptve Codng and Modulaton for OFDM Systems usng Product s and Fuzzy Rule Base System. Internatonal Journal of Computer Applcatons (IJCA), Vol. 35(4), pp.4-48, December 0. [] Atta-ur-Rahman, Quresh I.M., Mal A.., A Fuzzy Rule Base Asssted Adaptve Codng and Modulaton Scheme for OFDM Systems, J. Basc Appl. Sc. Res. Vol. (5), pp , 0. [ISI Indexed Journal] [3] C. Bocelmann, D. W ubben and K. D. Kammeyer, Rate Enhancement of BICM-OFDM wth Adaptve d Modulaton va a Bsecton approach. IEEE 0th Worshop on Sgnal Processng Advances n Wreless Communcatons, SPAWC '09. pp , 009. [4] Atta-ur-Rahman, Quresh I.M., Muzaffar M.Z., aseem M.., Adaptve Resource Allocaton for OFDM Systems usng Fuzzy Rule Base System and Water-fllng Prncple. IEEE Conference on Soft Computng for Pattern Recognton (SoCPaR ), pp. 8-86, December 0-3, Brune Darussalam. [5] P. Elas, Error-free codng, IEEE transacton on Informaton heory, vol. 4, pp. 9-37, 954. [6] Atta-ur-Rahman, Ghour, S.A. Adeel, H., Waheed, A. "Performance of Modfed Iteratve Decodng Algorthm for Product codes", Internatonal Conference on Computatonal Aspects of Socal etwors (CASo), pp. 47-5, 9- Oct. 0, Salamanca, Span. [7] Wang, L.X. A Course n Fuzzy Systems and Controls. (Chapter ). Prentce Hall Publcatons, 997. [8] Atta-ur-Rahman, Quresh I.M., Mal A.., Adaptve Resource Allocaton n OFDM Systems usng GA and Fuzzy Rule Base System, World Appled Scences Journal, Vol. 8(6), pp , 0. [ISI Indexed Journal] Authors Bographes Dr. Atta-ur-Rahman has receved hs BS degree n Computer Scence from Unversty of the Punjab Lahore, Pastan n 004; MS degree n Electronc Engneerng from Internatonal Islamc Unversty Islamabad, Pastan n 008 and PhD degree n Electronc Engneerng from ISRA Unversty, Islamabad Campus, Pastan n 0. Currently he s worng as Assstant Professor as well as Deputy Drector (Academcs) at Baran Insttute of Informaton echnology, Rawalpnd, Pastan. Hs research nterests nclude Dgtal/Wreless Communcatons, Dgtal Sgnal Processng, Informaton and Codng heory, Softcomputng, Artfcal Intellgence, Evolutonary Computng and Fuzzy & Hybrd Intellgent Systems. Dr. Ijaz Mansoor Quresh has receved hs BE degree n Avonc Engneerng from ED Unversty Karach, Pastan. Frst MS degree n Electrcal Engneerng from Mddle East echncal Unversty (MEU), Anara, urey and second MS degree n Hgh Energy Physcs from Syracuse Unversty, ew Yor, USA. He earned hs PhD degree n Hgh Energy Physcs from Unversty of oronto, oronto. He has more that twenty-seven year post PhD experence n teachng and

11 Comparson of Codng Schemes over FRBS Aded AOFDM Systems 93 research at dfferent ntutons of good repute n Pastan. More than twelve PhD has been produced n hs supervson. Currently he s Professor at Electrcal Engneerng departments, Ar Unversty Islamabad, Pastan. Hs research nterests nclude Dgtal/Wreless Communcatons, Dgtal Sgnal Processng, Informaton and Codng heory, Soft and Evolutonary Computng. Mr. Muhammad ahr aseem receved the B.S. degree n Computer Scence from Unversty of the Punjab, Pastan and M.S. degree n Electronc Engneerng from Internatonal Islamc Unversty, Islamabad, Pastan n 005 and 008, respectvely. Currently, he s PhD research student at ISRA Unversty Islamabad Campus, Pastan. Hs research nterests nclude Dgtal Watermarng, Dgtal Sgnal Processng and Informaton Securty.