基于MATLAB的OFDM系统仿真及性能分析-基于MATLAB的OFDM系统仿真及性能分析.pdf

所需积分/C币:22 2019-08-13 10:59:48 368KB PDF

基于MATLAB的OFDM系统仿真及性能分析-基于MATLAB的OFDM系统仿真及性能分析.pdf 基于MATLAB的OFDM系统仿真及性能分析
表1仿真参数 OFDM Simulation 子载波数 200 :999 = 〓 位数/符号 符号数/载波 训练符号数 循环前缀长度 T/4 调制方式 QDPSK 10 多径信道数 2、3、4 信道最大时延 7(单位数据符号) 图4给出高斯信道和多径瑞利衰落信道下系统的误码 No CP and Channel Equalization F 5::: ::: o CP 率曲线,该曲线表明OFDM系统仿真过程正确。图5给出高斯 Channel Equalization 信道下循环前缀和信道估计对误码率影响曲线,该曲线表明 SNR(dB 高斯信道下循环前缀和信道估计对系统的影响并不显著,在 这样的信道条件下可以去除信道估计器以降低接收机的复 图5高斯信道下循环前缀和信道估计对误码率影响曲 杂度。图6给出多径瑞利衰落信道下分别利用 L SMMSE信 道估计方法的误码率对比曲线,该曲线表明MMSE估计法明 OFDM Simulation 显优于LS估计法,但其复杂度较高,在误码率满足的条件 下,可以采用简单的LS估计法降低接收机复杂度 OFDM Simulation -Pat No Estimatlon ° LS Channe| Estimation Path o MMSE Channel Estimation -+ SNR(dB) o AWGN Channel 2-Path Rayleigh Fading Channel+ AWGN Channel r- 3-Path Rayleigh Fading Channel+ AWGN Channe! -i a 4-Path Rayleigh Fading Channel+ AWGN Channel 图6多径瑞利衰落信道下两种信道估计误码率对比曲 8 SNr(dB) carrier count =200 图4高斯信道和多径瑞利衰落信道下系统的误码率曲线 bits per sym bol= 2 bols per carrier =50 SNR input(snr =) 结论 baseband out length= carrier count sym bols per carrier 本文针对目前的研究热点OFDM技术进行计算机仿真* bits per sym bo l 研究,在OFDM仿真模型的基础上用 MA TLAB语言编写出 carriers=(l: carrier count+(ior(FFbi- length/4) OFDM发送、信道及接收整个系统,在系统仿真正确的前提 floor( carrier count/2)) 下,对循环前缀、信道估计等性能改善方法进行进一步仿真 con juga te carriers= iFFT bin length- carriers+2; 验证,得到了预期的理想结果。今后,作者将在这一仿真基础 %信号发射 上进行OHM系统的FPGA实现。 baseband out= round (rand(1, ba seband out length)) 附录:OHDM系统仿真主要M文件 convert matrix re shape( baseband out, bits per sym bo clear all length( ba seband out)/bits per sym bo l) for k =1:( length( baseband out)/bits per symbol) fp rinf(OFM仿真n\n) modulo ba seband() IFFt bin length =1024 for i= 1: bits per symbol 166 o1994-2007ChinaAcademicJournalElectronicPublishingHouse.Allrightsreservedhttp://www.cnki.net modul baseband (k) modul ba seband() d1=4; al=0. 2; d2= 5; a2=0. 3; d3=6: a3=0. 4 convert matrix (i, k)*2(bits per symbol i) 4=0.5; end copal=zeros( size(Tx data)) for i=1+d1: length(Tx da ta carrier matrix =reshape(modulo baseband, carrier count sym bols per carrier) c. copal(i)= al Tx data(i-d1) % QDPSK调制 cop y2 =zeros( size( tx data)) carrier matrix =[ zeros(l, carrier count); carrier matrix for i= 1 d2 length(Tx da ta) for i=2:( symbols per carrier 1) copal (i)=a2* Tx data(i-d2) carrier matrix (i, :)=rem( carrier matrix(i,: )+ carrier matrix(i-1,: ), 2 bits per symbol cop y3= zeros( size(Tx data)) r 1 +d3: length( tx da ta) carrier matrix carrier matrix *((2*pi/(2/ bits per copy (i)=a3* Tx data(i-d3) [x,Y] po cart( carrier matrix, ones( size (carrier cop y4= zeros( size( tx data)); matrix, 1), size carrier matrix, 2))) for i=1+ d4: length(Tx da ta) comp lex carrier matrix comp lex(X, y) copal (i)=a4* Tx data (i- d4 %加训练序列 end train ing symbols =[ljj1-1-j-j-11jj1-1-j- Tx da ta Tx data copyl cop y2 -11jj1-1 Tx signal power= var( Tx da ta linear SNR= 10( SNR/10) 11jj1-1-j-j-11jj1-1-j-j-11jjl-1-j noise sigma= Tx signal power/linear SNR; j-11j 1 jj1- noise scale fac tor= sqrt( noise sigma) 1-j-j-11jj1-1-j-j-11jjl-1-j-j-11 noise randn(1, length(Tx data))* noise scale fac to r jj1- 1-l-j-j RX Data= Tx da ta nois 11jj1 %信号接收 j-1 RX Data matrix re shape(rx data, IFFT bin length,4+ training symbols= cat(l, train ing symbols, train ing symbols) sym bols per carrier 1) training symbols= cat(1, training symbols, training symbols); Rx spectrum= fft(RX Data matrix); comp lex carrier matrix cat(l, tra in ing sym bo ls, comp lex Rx carriers =Rx spectrum(carriers, carrier matrix) Rx train ing symbols Rx carriers((1: 4),: ) IFFT modu lation zeros(4 symbols per carrier 1, RX carriers= RX carriers((5: 55), : iFFt b in length) %信道估计 IFFT modulation(:, carriers)=comp lex carrier matrix; Rx train ing sym bo Is RX tra ining sym bols /tra in ing IFFT modu lation(:, con juga te carriers) ls conj( comp lex carrier matrix) Rx train ing sym bols deno= Rx train ing sym bo Is 2 tme wave matrix ifft( FFT modu lation Rx train ing sym bols deno tme wave m atrix tme wave matrix Rx train ing sym bols deno(1, :)+ r 1 1: 4+ symbols per carrier 1 Rx train ing sym bols deno (2, ) w endowed tme wave matrix(i,:) real( tme wave Rx training sym bols deno(3, :) matrix (i,: )) Rx train ing sym bols deno (4, Rx train ing sym bols nume =Rx tra in ing symbols(1,:)+ ofdm modulation re shape(w indowed tme wave matrix, Rx train ing sym bols(2, :)+ Rx tra in ing sym bols(3,: ) 1, IFFT bin length*(4+ sym bols per carrier+ 1)) Rx training sym bols (4, Tx data ofdm modulation Rx train ing sym bols nume %信道 conj(rx tra in ing sym bols nume) 167 2019 994-2007ChinaAcademicjOurnalElectronicPublishingHouse.Allrightsreservedhttp://www.cnki.net Rx train ing sym bo ls Rx train ing sym bols nume for i= bits per sym bob -1: 1 /Rx train ing sym bo ls deno Rx train ing sym bo Is 2 cat(1, Rx tra in ing sym bo ls, Rx binary matrix(i, )=rem(RX serial sym bols, 2) Rx train ing sym bo ls) RX serial sym bols= floor(RX serial sym bo ls/2) Rx tra in ing sym bo Is 4 cat(l, Rx train ing sym bols 2 Rx tra in ing sym bo ls 2) Rx binary matrix (i, )=RX serial sym bo ls, Rx train ing sym bo ls 8 cat(l, Rx train ing sym bols 4 Rx tra in ing sym bo ls 4) end Rx training sym bo ls 16 baseband in re shape(rx binary matrix, 1 cat(l, Rx training symbols 8, Rx tra in ing sym bols 8) size(rx binary matrix, 1)* size(rx binary matrix, 2)) Rx tra in ing sym bo Is 32 %误码率计算 cat(l, Rx train ing sym bols 16, Rx train ing sym bols 16) bit errors= find( ba seband in baseband out RX tra ining sym bo ls 48 bit error count size(bit erors, 2) cat(1, Rx training sym bols 32, Rx tra ining sym bols 16); to tal bits= size(baseband out, 2) Rx train ing sym bo ls 50 bit error ra te= bit error count/to tal bits cat(1, Rx tra in ing sym bols 48, Rx tra in ing sym bols 2) fp rinf(%f\n, bit eror ra te) Rx train ing sym bo ls cat(l, Rx tra ining sym bols 50 Rx train ing sym bo ls) 参考文献 Rx carriers=Rx train ing sym bols *Rx carriers [1] Erich Cosby Orthogonal Frequency D ivision Multip lexing Rx phase angle(rx carriers)*(180/pi) (OFDM) Tutorial and Analysis[M]. Northern V irgin ia Cen ter, phase negative find(rx phase < 0) Rx phase(phase negative [2] Mingqi Li,, Q icong Peng, Yubai Li, Perfo mance Eva lua tion of rem(rx phase(pha se nega tive)+ 360, 360) MC-DS-CDMA Systems in Multipath Fading Channels[J]. 0 -7803-7547-5/02,正EE,2002 Rx decoded phase= diff(rx phase) [3] A Pe led, A Ruiz Frequency dom ain data tran m ission using phase negative= find(rx decoded phase <0); reduced compu ta tional comp lexity algorithms[ C]. In Pmc Rx decoded phase(p hase negative IEEE Int Conf Acoust, Speech, Signal Processing, 1980. rem(rx decoded phase(phase negative)+ 360, 360) 964-967 QDPSK解调 [4 R van Nee O FDM W ire less Multmedia Commun ications[M] b 360/2 bits per sym bo l Rrasad R. artech House. 1998. Ita phase= base phase /2 [5]周正兰,等.OFDM及其链路级平台的 Smulink实现[J中国 Rx decoded sym bols= zeros( size(rx decoded phase, 1) 数据通信,2003,(10):90-92 size(rx decoded phase, 2) [6]尹泽明,等.精通 MA ILAB6[M].清华大学出版社,2002 [7]蔡涛,等译.无线通信原理与应用[M].电子工业出版社 for i= 1:(2 bits per center phase base p ha se [8]丁玉美,等.数字信号处理M]西安电子科技大学出版社,2003 p lus delta= center phase delta 作者简介] m inus delta= center phase- delta phase, 吕爱琴(1978-),女(族),江苏人,西安电子科 decoded= find((rx decoded phase < plus delta)& 技大学计算机学院在校硕士研究生,主要研究方 (RX decoded phase m inus delta)) 向为无线通信仿真设计,多载波CDMA等。 Rx decoded symbols( decoded)=i 朱明华(1965-),男(汉族),四川人,北邮博士 硕导,中科院上海微系统与信息技术研究所副研 RX Serial sym bols re shape (rx decoded sym bols, 1 究员,主要研究方向为无线通信仿真设计。 size (rx decoded sym bo ls, 1)* size (rx decoded sym bo ls 田玉敏(1966-),女(汉族),河南人,西安电子科技大学计算机外设 2)); 所,博士,教授,主要研究计算机图形图像处理,网络安全等。 168 o1994-2007ChinaAcademicJournalElectronicPublishingHouse.Allrightsreservedhttp://www.cnki.net

...展开详情
试读 5P 基于MATLAB的OFDM系统仿真及性能分析-基于MATLAB的OFDM系统仿真及性能分析.pdf
img

关注 私信 TA的资源

上传资源赚积分,得勋章
    最新推荐
    基于MATLAB的OFDM系统仿真及性能分析-基于MATLAB的OFDM系统仿真及性能分析.pdf 22积分/C币 立即下载
    1/5
    基于MATLAB的OFDM系统仿真及性能分析-基于MATLAB的OFDM系统仿真及性能分析.pdf第1页
    基于MATLAB的OFDM系统仿真及性能分析-基于MATLAB的OFDM系统仿真及性能分析.pdf第2页

    试读已结束,剩余3页未读...

    22积分/C币 立即下载 >