Free Spce Opticl Communiction ginst Chnnel Fding Prem.S A.Elngovn ME Applied Electronics, Aruni Engineering College,Tiruvnnmli, Tmil Ndu-606603, Indi Assistnt Professor/ECE, Aruni Engineering College,Tiruvnnmli, Tmil Ndu-606603, Indi Abstrct Communiction in free spce opticl chnnel is impired predominntly by multi-pth fding. The rndom fluctution in signl level, known s fding, severely ffects the qulity nd relibility of wireless communiction. This cn be eliminted by pplying MIMO technology. The use of multiple ntenns t the trnsmitter nd receiver in wireless systems, populrly known s MIMO (multiple-input multiple-output) technology, hs rpidly gined in populrity over the pst decde due to its powerful performnce-enhncing cpbilities. In this project slow-fding chnnel model is considered nd the outge probbility is derived s performnce mesure. Incresing the number of trnsmitters nd receivers results in lower probbility of outge for given SNR. And in the presence of mislignment fding the diversity gin depends only on the mislignment vrince. Thus the grph is plotted between SNR nd outge probbility for SISO x 4 x 4 nd 6 x 6 respectively nd results re nlyzed. Keywords: Signl to Noise Rtio (SNR), MIMO, Free- Spce Opticl (FSO).. Introduction The use of multiple ntenns t the trnsmitter nd receiver in wireless systems, populrly known s MIMO (multiple-input multiple-output) technology, hs rpidly gined in populrity over the pst decde due to its powerful performnce-enhncing cpbilities. Communiction in wireless chnnels is impired predominntly by multi-pth fding. Multi-pth is the rrivl of the trnsmitted signl t n intended receiver through differing ngles nd/or differing time delys nd/or differing frequency (i.e., Doppler) shifts due to the scttering of electromgnetic wves in the environment. (due to dely spred) nd/or time (due to Doppler spred) through the rndom superposition of the impinging multipth components. This rndom fluctution in signl level, known s fding, cn severely ffect the qulity nd relibility of wireless communiction. Additionlly, the constrints posed by limited power nd scrce frequency bndwidth mke the tsk of designing high dt rte, high relibility wireless communiction systems extremely chllenging. MIMO technology constitutes brekthrough in wireless communiction system design. The technology offers number of benefits tht help meet the chllenges posed by both the impirments in the wireless chnnel s well s resource constrints. In ddition to the time nd frequency dimensions tht re exploited in conventionl single-ntenn (singleinput single-output) wireless systems, the leverges of MIMO re relized by exploiting the sptil dimension (provided by the multiple ntenns t the trnsmitter nd the receiver). The dvntges of multiple-input multiple-output (MIMO) systems hve been widely cknowledged, to the extent tht certin trnsmit diversity methods (i.e., Almouti signling) hve been incorported into wireless stndrds. Although trnsmit diversity is clerly dvntgeous on cellulr bse sttion, it my not be prcticl for other scenrios. Specificlly, due to size, cost, or hrdwre limittions, wireless gent my not be ble to support multiple trnsmit ntenns. Exmples include most hndsets (size) or the nodes in wireless sensor network (size, power). Consequently, the received signl power fluctutes in spce (due to ngle spred) nd/or frequency Fig.. MIMO cellulr system 37
. MIMO Technology MIMO technology tht help chieve such significnt performnce gins re rry gin, sptil diversity gin, sptil multiplexing gin nd interference reduction. These gins re described in brief below. Arry gin Arry gin is the increse in receive SNR tht results from coherent combining effect of the wireless signls t receiver. The coherent combining my be relized through sptil processing t the receive ntenn rry nd/or sptil pre-processing t the trnsmit ntenn rry. Arry gin improves resistnce to noise, thereby improving the coverge nd the rnge of wireless network... Sptil diversity gin As mentioned erlier, the signl level t receiver in wireless system fluctutes or fdes. Sptil diversity gin mitigtes fding nd is relized by providing the receiver with multiple (idelly independent) copies of the trnsmitted signl in spce, frequency or time. With n incresing number of independent copies (the number of copies is often referred to s the diversity order), the probbility tht t lest one of the copies is not experiencing deep fde increses, thereby improving the qulity nd relibility of reception. A MIMO chnnel with M T trnsmit ntenns nd M R receive ntenns potentilly offers M T M R independently fding links, nd hence sptil diversity order of M T M R.. Sptil multiplexing gin MIMO systems offer liner increse in dt rte through sptil multiplexing, i.e., trnsmitting multiple, independent dt strems within the bndwidth of opertion. Under suitble chnnel conditions, such s rich scttering in the environment, the receiver cn seprte the dt strems. Furthermore, ech dt strem experiences t lest the sme chnnel qulity tht would be experienced by single-input single-output system, effectively enhncing the cpcity by multiplictive fctor equl to the number of strems. In generl, the number of dt strems tht cn be relibly supported by MIMO chnnel equls the minimum of the number of trnsmit ntenns nd the number of receive ntenns, i.e., min [M T M R ]. The sptil multiplexing gin increses the cpcity of wireless network...3 Interference reduction nd voidnce Interference in wireless networks results from multiple users shring time nd frequency resources. Interference my be mitigted in MIMO systems by exploiting the sptil dimension to increse the seprtion between users. For instnce, in the presence of interference, rry gin increses the tolernce to noise s well s the interference power, hence improving the signl-to-noise-plus-interference rtio (SINR). Additionlly, the sptil dimension my be leverged for the purposes of interference voidnce, i.e., directing signl energy towrds the intended user nd minimizing interference to other users. Interference reduction nd voidnce improve the coverge nd rnge of wireless network. In generl, it my not be possible to exploit simultneously ll the benefits described bove due to conflicting demnds on the sptil degrees of freedom. owever, using some combintion of the benefits cross wireless network will result in improved cpcity, coverge nd relibility..mimo in Cellulr Networks In cellulr wireless communiction network, multiple users my communicte t the sme time nd (or) frequency. The more ggressive the reuse of time nd frequency resources, the higher the network cpcity will be, provided tht trnsmitted signls cn be detected relibly. Multiple users my be seprted in time (timedivision) or frequency (frequency-division) or code (codedivision). The sptil dimension in MIMO chnnels provides n extr dimension to seprte users, llowing more ggressive reuse of time nd frequency resources, thereby incresing the network cpcity. Figure is the schemtic of cell in MIMO cellulr network. A bse-sttion equipped with L ntenns communictes with P users, ech equipped with M ntenns. The chnnel from the bse-sttion to the users (the downlink) is brodcst chnnel (BC) while the chnnel from the users to the bse-sttion (the uplink) is multiple-ccess chnnel (MAC). The set of rte-tuples (R,R,..R P ) tht cn be relibly supported on the downlink or uplink constitutes the cpcity rte region for tht link. Recently, n importnt dulity hs been discovered between the rte regions for the downlink nd uplink chnnels. In order to understnd the possible gins from MIMO technology in multi-user environment, consider the uplink of cellulr MIMO system where ll the users simultneously trnsmit independent dt strems from ech of their trnsmit ntenns, i.e., ech user signls with sptil multiplexing. To the bsesttion, the users combined, pper s multi-ntenn trnsmitter with PM ntenns. Thus the effective uplink chnnel hs dimension of L PM. This effective chnnel will hve considerbly different structure from the w MIMO single user chnnel due to pth-loss nd shdowing differences between users. 38
The signl-to-noise rtio is defined s SNR = PP/σσ nd the chnnel gin = N N n= m= ccounts for the combined effects of tmosphere nd mislignment fding where Fig. Distributed MIMO: multiple users cooperte to form virtul ntenn rry tht relizes the gins of MIMO in distributed fshion. owever, with rich scttering nd L PM, we cn expect tht the sptil signtures of the users re well seprted to llow relible detection. Using multi-user ZF receiver will llow perfect seprtion of ll the dt strems t the bse-sttion, yielding multi-user multiplexing gin of PM. The use of more complex receivers for multi-user detection nd the ssocited performnce trde-offs. A similr thought experiment cn be pplied for the downlink, where the bse-sttion exploits the sptil dimension to bem informtion intended for prticulr user towrds tht user nd steers nulls in the directions of the other users, thus completely eliminting interference. 3. Modules Explntion 3. Chnnel Model This section explins the MIMO FSO system nd chnnel fding. The model considers MIMO FSO system with M trnsmitters (lsers) nd N receivers (pertures). In ll cses, intensity modulted PAM signling with direct detection is considered. The received N vector y = [y,..., y N ] T is given by where nd p = p re independent rndom vribles representing the tmospheric nd time vrying mislignment (pointing) fding respectively between trnsmitter m nd receiver n. In the wek turbulence regime the chnnel gin due to tmospheric turbulence is well modeled by = e where XXmn is Gussin rndom vrible. Assume tht ll XXmn re modeled s independent nd identiclly distributed (i.i.d) rndom vribles. For rdil displcement of RRmn in the receiver plne between the center of trnsmitter bem mm nd the center of perture nn, the loss due to mislignment is p R / w A0e where AA 0 is the equivlent receiver re nd ww is the equivlent bem wist t receiver. X 3. Diversity Gin of MIMO FSO Chnnels These section three different mislignment scenrios will be nlyzed depending on the rndom displcements XX nd YY Symmetric Mislignment in XX nd YY Directions The displcements XX nd YY hve i.i.d Gussin distributions with zero men nd vrince. Defining = ww/(σσss) the pdf of TT = RR /ww is given by where is n M x N chnnel mtrix where the entry 0 represents the chnnel gin from trnsmitter m to receiver n with m=,...,m nd n =,...,N nd ( ) T is the trnspose opertor. The vector x = [x,...,x M ] T is the trnsmitted set of symbols nd z = [z,...,z N ] T is noise vector of independent components modeled s signl independent white nd Gussin distributed. nd hence 39
3.3 Simultion Results where, nd. Substituting ss = vv + BB the outge probbility cn be simplified to Substituting the results in the symptotic probbility of outge given s In this section, we consider Gussin-bem of wvelength λλ = 550 nm, bem wist wwoo = cm, nd rdius of curvture FFoo = 0 m t the trnsmitter. The bem propgtes distnce LL = km through turbulent medium chrcterized by CC nn. Ech receiver hs circulr perture of rdius = 5 cm. The spcing between trnsmitters (s well s between receivers) is set to dd = 0 cm which is typicl vlue for commercil system nd results in independent XXmn s shown. Typicl mislignment vrince of σσ ss = 0. m is considered nd rte RR 0 = bits/chnnel-use is considered 0 0 Numericl Simultion Closed-form ut o Probbility 0 - of outge P Unidirectionl Mislignment In this scenrio XX ~ (0, ) nd YY = 0. The probbility density functions of TT = XX /ww is given by 0-4 0-6 The outge probbility is No Mislignment For comprison, the scenrio when XX = 0 nd YY = 0, i.e., no mislignment, is lso considered. The chnnel gin is given by modified version of the bove sttisticl chnnel model s eqution U 0 = A N N 0 n= m= e X w U = Pm P nd GGoo is Gussin with n men vrince nd The outge probbility is given s 0-8 0 5 0 5 0 5 (Log 0 (SNR)) 4. Conclusions A novel generlized sttisticl model for MIMO FSO chnnels impired by tmospheric nd mislignment fding is developed. The derived model is utilized to study the outge probbility of FSO chnnels nd the diversity gin t high signl-to-noise rtio. Closed-form expressions for the outge probbility re derived tking into ccount different mislignment fding scenrios. It is shown tht, in the presence of tmospheric nd mislignment fding, the diversity gin depends only on the mislignment prmeters nd is independent of both the number of trnsceivers nd tmospheric fding prmeters. Contrrily, when tmospheric fding is the only chnnel impirment, i.e., no mislignment fding, the diversity gin depends on the number of trnsceivers. Thus lrger diversity gin cn be chieved by incresing the number of trnsmitters nd receivers. In ll cses incresing the number of trnsmitters nd receivers decreses the outge probbility for given SNR. owever, in order to hve independent chnnels gins it is required to sufficiently increse the spcing between receivers which is often prcticlly difficult. 40
An lterntive pproch is to utilize lrge single-perture with the equivlent re of the NN pertures. This pproch provides simple system structure nd reduces the fding vrince vi perture verging. Note tht, in this cse the fding is correlted cross the perture nd hence its vrince is lrger thn tht of system with multiple pertures nd independent fding t ech perture. Consequently, multiple-perture receiver systems chieve improved performnce t the cost of dditionl detector element ACKNOWLEDGEMENT I like to express hert full thnks to Mr.A.Elngovn, ssistnt Professor, Deprtment of ECE, to spend his vluble time to do this project nd lso I wnt to thnk to my college nd my friends for their support. References [] J. M. Khn nd J. R. Brry, Wireless infrred communictions, Proc. IEEE, vol. 85, pp. 65 98, Feb. 997. [] D. Tse nd P. Viswnth, Fundmentls of Wireless Communiction, st edition. Cmbridge University Press, 005. [3] M. Rzvi nd J.. Shpiro, Wireless opticl communictions vi diversity reception nd opticl premplifiction, IEEE Trns. Wireless Commun., vol. 4, pp. 975 983, My 005. [4] X. Zhu nd J. Khn, Free spce opticl communiction through tmospheric turbulence chnnels, IEEE Trns. Commun., vol. 50, pp.93 300, Aug. 00. [5] S. M. Nvidpour, M. Uysl, nd M. Kvehrd, BER performnce of free-spce opticl trnsmission with sptil diversity, IEEE Trns.Wireless Commun., vol. 6, pp. 83 89, Aug. 007. [6] T. A. Tsiftsis,. G. Sndlidis, G. K. Krginnidis, nd M. Uysl, Opticl wireless links with sptil diversity over strong tmospheric turbulence chnnels, IEEE Trns. Wireless Commun., vol. 8, pp. 95 957, Feb. 009. [7] N. Letzepis nd A. G. i Fábregs, Outge probbility of the Gussin MIMO free-spce opticl chnnel with PPM, IEEE Trns. Commun., vol. 57, pp. 368 3690, Dec. 009. [8], Outge probbility of the free-spce opticl chnnel with doubly stochstic scintiltion, IEEE Trns. Commun., vol. 57, pp. 899 90, Oct. 009. [9] S. M. s nd J.. Shpiro, Cpcity of wireless opticl communictions, IEEE J. Sel. Ares Commun., vol., pp. 346 356, Oct. 003. [0] S. G. Wilson, M. Brndt-Perce, Q. Co, nd J.. Leveque, Free-spce opticl MIMO trnsmission with Q- ry PPM, IEEE Trns. Commun., vol. 53, pp. 40 4, Aug. 005. [] S. G. Wilson, M. Brndt-Perce, Q. Co, nd M. Bedke, Opticl repetition MIMO trnsmission with multipulse PPM, IEEE J. Sel. Ares Commun., vol. 3, pp. 90 90, Sep. 005. [] G. A. Koepf, R. Peters, nd R. G. Mrshlek, Anlysis of brust error occurrence on opticl interstellite link (ISL) design, in Proc. SPIE Opt. Tech. Commun. Stellite Applictions, Jn. 986, vol. 66, pp. 9 36. [3] S. Arnon, S. Rotmn, nd N. S. Kopeik, Bem width nd trnsmitter power dptive to trcking system performnce for free-spce opticl communiction, J. Appl. Opt., vol. 36, pp. 6095 60, Aug. 997. [4] S. Arnon, Optimiztion of urbn opticl wireless communiction systems, IEEE Trns. Commun., vol., pp. 66 69, Nov. 003. [5] C. C. Chen nd C. S. Grdner, Impct of rndom pointing nd trcking errors on the design of coherent nd incoherent opticl interstellite communiction links, IEEE Trns. Commun., vol. 37, pp. 5 60, Mr. 989. [6] S. Arnon, Effects of tmospheric turbulence nd building swy on opticl wireless communiction systems, Optics Lett., vol. 8, pp. 9 3, Jn. 003. [7] K. Kisleh, On the probbility density function of signl intensity in free-spce opticl communictions systems impired by pointing jitter nd turbulence, Opt. Eng., vol. 33, pp. 3748 3757, Nov. 994. [8] A. A. Frid nd S. rnilovic, Outge cpcity optimiztion for freespce opticl links with pointing errors, IEEE J. Lightwve Technol.,vol. 5, pp. 70 70, July 007. [9], Diversity gins for MIMO wireless opticl intensity chnnels with tmospheric fding nd mislignment, in Proc. 00 IEEE Globe-Com, pp. 05 09. [0] J. A. Anguit, M. A. Neifeld, nd B. V. Vsic, Sptil correltion nd irrdince sttistics in multiplebem terrestril free-spce opticl communiction link, J. Appl. Opt., vol. 46, pp. 656 657, Sep. 007. 4