联合哈特莱变换和限幅算法降低OFDM峰均功率比的matlab仿真代码

clear all; K=128; %子载波数 IF=2; % 插值系数 N=256; CR=1.4; %限幅率 QPSK_Set = [1 -1 j -j]; %qpsk编制 for n=1:10000 Index = randint(1,K,length(QPSK_Set))+1; X = QPSK_Set(Index(1,:));%对随机信号进行QPSK调制 %计算PAPR_Orignal y = ifft(X,[],2); Signal_Power0 = abs(y.^2);%计算信号的峰值功率 Peak_Power0= max(Signal_Power0,[],2);%计算信号的平均功率 Mean_Power0= mean(Signal_Power0,2);%计算PAPR PAPR_Orignal(n) =10*log10(Peak_Power0./Mean_Power0); %计算PAPR_Clipping XX = [X(1:K/2) zeros(1,256-K) X(K/2+1:K)]; %zeros(1,N-K)为插入的0数据 y=ifft(XX,[],2); Signal_Power = abs(y.^2); Mean_Power= mean(Signal_Power,2);%计算信号的平均功率 y2=abs(y); y1=y; A=CR*sqrt(Mean_Power); % Clipping for h=1:256; if y2(h)>A; y1(h)=A*y(h)/y2(h); end end % Filtering 限幅后进行滤波 XX = fft(y1,[],2); XX(K/2+(1:256-K))=zeros(1,256-K); %人为的将带外噪声置零，滤波的目的 x = ifft(XX,[],2); % PAPR Compute Signal_Power2 = abs(x.^2); Peak_Power2 = max(Signal_Power2,[],2); Mean_Power2 = mean(Signal_Power2,2); PAPR_Clipping(n) = 10*log10(Peak_Power2./Mean_Power2); %计算PAPR_DHT z=Dht(X); %对采样信号作DHT变换 y = ifft(z,[],2); Signal_Power1 = abs(y.^2); %计算信号的峰值功率 Peak_Power1= max(Signal_Power1,[],2);%计算信号的平均功率 Mean_Power1= mean(Signal_Power1,2);%计算PAPR PAPR_DHT(n) =10*log10(Peak_Power1./Mean_Power1); X1=Dht(X); %对采样信号作DHT变换 y=ifft(X1,[],2); Signal_Power= abs(y.^2); Mean_Power= mean(Signal_Power,2);%计算信号的平均功率 y2=abs(y); y1=y; A=CR*sqrt(Mean_Power); % Clipping for h=1:128; if y2(h)>A; y1(h)=A*y(h)/y2(h); end end Signal_Power3 = abs(y1.^2); Peak_Power3 = max(Signal_Power3,[],2); Mean_Power3 = mean(Signal_Power3,2); PAPR_DhtClipping(n) = 10*log10(Peak_Power3./Mean_Power3); end [cdf0, PAPR0] = ecdf(PAPR_Orignal); [cdf1, PAPR1] = ecdf(PAPR_Clipping); [cdf2, PAPR2] = ecdf(PAPR_DHT); [cdf3, PAPR3] = ecdf(PAPR_DhtClipping); figure(1); semilogy(PAPR0,1-cdf0,'-b+',PAPR1,1-cdf1,'-r*',PAPR2,1-cdf2,'-gs',PAPR3,1-cdf3,'-mo'); legend('Orignal','Clipping','DHT','DhtClipping'); xlabel('PAPR0 [dB]'); ylabel('CCDF (Pr[PAPR>PAPR0])'); xlim([0 13]); grid on;