注册 登录  
 加关注
   显示下一条  |  关闭
温馨提示!由于新浪微博认证机制调整,您的新浪微博帐号绑定已过期,请重新绑定!立即重新绑定新浪微博》  |  关闭

Mr.Right

不顾一切的去想,于是我们有了梦想。脚踏实地的去做,于是梦想成了现实。

 
 
 

日志

 
 
关于我

人生一年又一年,只要每年都有所积累,有所成长,都有那么一次自己认为满意的花开时刻就好。即使一时不顺,也要敞开胸怀。生命的荣枯并不是简单的重复,一时的得失不是成败的尺度。花开不是荣耀,而是一个美丽的结束,花谢也不是耻辱,而是一个低调的开始。

网易考拉推荐

特征向量约束的LCMP-DL波束形成器  

2012-03-16 14:10:09|  分类: 学习 |  标签: |举报 |字号 订阅

  下载LOFTER 我的照片书  |
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Figure 6.61
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Example 6.7.11
% LCMP ,eigenvector constraints, with diagonal loading.

clear all
close all

N = 10;
n = (-(N-1)/2:(N-1)/2)';
ud = 0.1;                       %constraint region
BWNN = 4/N;
signalRange = 0:BWNN/2/200:BWNN/2;
eigNum = 3;

SNR = 10^(20/10);
INR = 10^(30/10);
Vi1 = exp(j*n*0.3*pi);
Vi2 = exp(-j*n*0.3*pi);

t = -ud:1/500:ud;
H = exp(j*n*pi*t)*diag(sinc(-(N/2)*t)./(sinc(-.5*t)));
H = transpose(H);
P = transpose(1/500*trapz(H)); %Ud = Bc

for m = 1:N
   for k = 1:N
      Q(m,k) = 2*ud*sinc(ud*(m-k));    
   end
end
[U,S,V] = svd(Q);
C = U(:,1:eigNum);
f = inv(S(1:eigNum,1:eigNum))*C'*P;

m = 1;
for LNR = 10.^([0 10 20 30]/10)
   
   k = 1;
   for ua = signalRange
      Vs = exp(j*n*pi*ua);
      Sn = eye(N) + INR*Vi1*Vi1' + INR*Vi2*Vi2';
      Ss = SNR*Vs*Vs';
      Sx = Ss + Sn + LNR*eye(N);
      W = inv(Sx)*C*inv(C'*inv(Sx)*C)*f;
      SINRo = (W'*Ss*W)/(W'*Sn*W);
      SINRi = SNR/(2*INR+1);
      A(m,k) = SINRo/SINRi;
      k = k + 1;
   end
   m = m + 1;
end


A = 10*log10(abs(A));

plot(signalRange/BWNN,A(1,:),'-');            %BWnn = 4/N
hold on
plot(signalRange/BWNN,A(2,:),'--');            %BWnn = 4/N
hold on
plot(signalRange/BWNN,A(3,:),'-.');            %BWnn = 4/N
hold on
plot(signalRange/BWNN,A(4,:),':');            %BWnn = 4/N


grid
xlabel('{\itu}_{\ita} /{\itBW}_{\itNN}','Fontsize',14)
ylabel('Array gain (dB)','Fontsize',14)
h=legend('{\itLNR}=0 dB','{\itLNR}=10 dB','{\itLNR}=20 dB','{\itLNR}=30 dB');
set(h,'Fontsize',12)
title('LCMP\_DL, SNR=20dB, INR=30dB, U_i=+/-0.30, 3 Eigen Constraints, Bd(u)=Bc(u), ud=0.1')   
axis([0 0.25 20 50])
特征向量约束的LCMP-DL波束形成器 - 阿英 - Mr.Right
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Figure 6.62
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Example 6.7.11
% LCMP ,eigenvector constraints, with diagonal loading.

clear all
close all

N = 10;
n = (-(N-1)/2:(N-1)/2)';
ud = 0.1;                       %constraint region
BWNN = 4/N;
signalRange = -ud:ud/50:ud;
SNRrange = -10:2:40;
LNRrange = -20:2:50;
eigNum = 3;
%SNR = 10^(20/10);
%INR = 10^(30/10);
Vi1 = exp(j*n*0.3*pi);
Vi2 = exp(-j*n*0.3*pi);


t = -ud:1/500:ud;
H = exp(j*n*pi*t)*diag(sinc(-(N/2)*t)./(sinc(-.5*t)));
H = transpose(H);
P = transpose(1/500*trapz(H)); %Ud = Bc

for m = 1:N
   for k = 1:N
      Q(m,k) = 2*ud*sinc(ud*(m-k));    
   end
end

[U,S,V] = svd(Q);
C  = U(:, 1:eigNum);
f = inv(S(1:eigNum,1:eigNum))*C'*P;


k1 = 1;
for INR = 10.^([10 20 30]/10)
   disp(['loop ' int2str(k1) ' of 3 ...'])
   k2 = 1;
   for SNR = 10.^(SNRrange/10)
      k3 = 1;
      for LNR = 10.^(LNRrange/10)
         Gain(k1,k2,k3) = 0;
for ua = signalRange
   Vs = exp(j*n*pi*ua);
   Sn = eye(N) + INR*Vi1*Vi1' + INR*Vi2*Vi2';
   Ss = SNR*Vs*Vs';
   Sx = Ss + Sn + LNR*eye(N);
   W = inv(Sx)*C*inv(C'*inv(Sx)*C)*f;
   SINRo = (W'*Ss*W)/(W'*Sn*W);
   SINRi = SNR/(2*INR+1);
   Gain(k1,k2,k3) = Gain(k1,k2,k3) + SINRo/SINRi;
          end %end of ua
          k3 = k3 + 1;
       end %end of LNR
       k2 = k2 + 1;
    end %end of SNR
    k1 = k1 + 1;
end %end of INR
 
 
Gain = 10*log10( Gain/length(signalRange) );

for k1 = 1:3
   for k2 = 1:length(SNRrange)
      [A(k1,k2),I(k1,k2)] = max( real(Gain(k1,k2,:)) );
   end
end

for h1 = 1:3
   [x,y] = min ( abs(    A(h1,:)-max(A(h1,:)) +3 ));
   c(h1) = -10 + 2*(y-1);
   d(h1) = A(h1, y);
end

plot(SNRrange,A(1,:),'-',SNRrange,A(2,:),'--',SNRrange,A(3,:),'-.')
hold on
plot(c,d,'-o')

grid
axis([-10 40 0 50])
xlabel('{\itSNR} (dB)','Fontsize',14)
ylabel('Optimal {\itE}(gain) (dB)','Fontsize',14)
h=legend('{\itINR}=10 dB','{\itINR}=20 dB','{\itINR}=30 dB','3 dB points',3);
set(h,'Fontsize',12)
title('LCMP\_DL, SNR=20dB, u_i=+/-0.30, 3 Eigen Constraints, B_d(u)=B_c(u), u_d=0.1')
axis([-10 40 0 50])
特征向量约束的LCMP-DL波束形成器 - 阿英 - Mr.Right
 

  评论这张
 
阅读(431)| 评论(0)
推荐 转载

历史上的今天

在LOFTER的更多文章

评论

<#--最新日志,群博日志--> <#--推荐日志--> <#--引用记录--> <#--博主推荐--> <#--随机阅读--> <#--首页推荐--> <#--历史上的今天--> <#--被推荐日志--> <#--上一篇,下一篇--> <#-- 热度 --> <#-- 网易新闻广告 --> <#--右边模块结构--> <#--评论模块结构--> <#--引用模块结构--> <#--博主发起的投票-->
 
 
 
 
 
 
 
 
 
 
 
 
 
 

页脚

网易公司版权所有 ©1997-2016