【信道估计】基于图着色的试点分配GC-PA实现大规模MIMO网络上行链路中的导频去污染和信道估计附matlab代码

最新推荐文章于 2024-05-17 21:37:37 发布

天天Matlab代码科研顾问

最新推荐文章于 2024-05-17 21:37:37 发布

阅读量511

点赞数 9

分类专栏：信号处理文章标签：开发语言 matlab

本文链接：https://blog.csdn.net/m0_57702748/article/details/136097268

版权

信号处理专栏收录该内容

370 篇文章 53 订阅

订阅专栏

本文提出了一种基于图形着色的试点分配(GC-PA)方法，用于减轻多单元大规模MIMO系统中的细胞间干扰(ICI)。通过构建干涉图并有效分配有限的试点资源，仿真结果展示了该方案在减少潜在干扰方面的性能提升。

摘要由CSDN通过智能技术生成

✅作者简介：热爱科研的Matlab仿真开发者，修心和技术同步精进，代码获取、论文复现及科研仿真合作可私信。

🍎个人主页：Matlab科研工作室

🍊个人信条：格物致知。

更多Matlab完整代码及仿真定制内容点击👇

智能优化算法神经网络预测雷达通信无线传感器电力系统

信号处理图像处理路径规划元胞自动机无人机

🔥 内容介绍

一个大规模的多输入多输出(MIMO)系统利用大量基站(BS)天线为一组用户服务,由于细胞间干扰(ICI)而受到飞行员污染。在这封信中,提出了一个基于图形着色的试点分配(GC-PA)方案,以减轻多单元大规模MIMO系统的试点污染。具体而言,利用衰落通道的大规模特征,首先构建了一个干涉图来描述所有用户的潜在ICI关系。然后,在有限的试点资源下,拟议的GC-PA方案旨在通过在干涉图中有效地向用户分配试点来减少潜在的ICI。通过仿真验证了该方案的性能增益.

📣 部分代码

clear;clc; tic;% system parametersM = 512;    % number of BS antennas ML = 7;     % cellsK = 10;     % usersS = 15;     % number of pilotslambda=0.1;R_Cell = 1000;r_Min = 100;alpha = 3.8;sigma_shadow = 8;% q = 15;rho_ul = 10^(15/10);     % 15 dB                                                        % Test_num = 300;Test_num = 15;UL_SINR_LS_sw = zeros(K, L, Test_num);UL_SINR_MMSE_sw = zeros(K, L, Test_num);UL_SINR_MLE_sw = zeros(K, L, Test_num);UL_SINR_LS_wg = zeros(K, L, Test_num);UL_SINR_MMSE_wg = zeros(K, L, Test_num);UL_SINR_MLE_wg = zeros(K, L, Test_num);UL_SINR_LS_sp = zeros(K, L, Test_num);UL_SINR_MMSE_sp = zeros(K, L, Test_num);UL_SINR_MLE_sp = zeros(K, L, Test_num);UL_SINR_LS_ra = zeros(K, L, Test_num);UL_SINR_MMSE_ra = zeros(K, L, Test_num);UL_SINR_MLE_ra = zeros(K, L, Test_num);% Simulationfor i_test = 1:Test_num        display(i_test);                        % Generate Channel vector        [H, Beta] = F_H_Generate(M, L, K, R_Cell, r_Min, sigma_shadow, alpha);                %% random        N = S;   % the length of the pilot sequence (N>=K)        temp = zeros(L,S);        for i=1:L            temp(i,:) = randperm(S);        end                pilot = F_ZC(N,S);                pilots = zeros(N,K,L);        for l=1:L            for k=1:K                pilots(:,k,l) = pilot(:,temp(i,k));            end        end                    %--- Received Pilot Signal ---%         Y = zeros(M,N,L);   % received uplink training sequences at i-th BS        Y = received_signal(M,N,L,rho_ul,H,pilots);                H_est = CE_aid(M, K, L, H, pilots, rho_ul);                  UL_SINR_LS_ra(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);        %         % ===== LS --> SINR =====%%         H_est = CE_LS(M, K, L, Y, pilots, rho_ul, N);  %         %         UL_SINR_LS_ra(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);%                %         % ===== MMSE --> SINR =====%%         H_est = CE_ideal_MMSE(M, K, L, Y, pilots, rho_ul, N, Beta);  %                 %         UL_SINR_MMSE_ra(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);                % ===== MLE --> SINR =====%        H_est = CE_MLE(M, K, L, Y, pilots, rho_ul, N, Beta);                         UL_SINR_MLE_ra(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);                        %% sprs        [pilots, N] = F_SPRS2(L,K,Beta,lambda);                Y = received_signal(M,N,L,rho_ul,H,pilots);                H_est = CE_aid(M, K, L, H, pilots, rho_ul);                  UL_SINR_LS_sp(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);        %         % ===== LS --> SINR =====%%         H_est = CE_LS(M, K, L, Y, pilots, rho_ul, N);  %         %         UL_SINR_LS_sp(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);%                %         % ===== MMSE --> SINR =====%%         H_est = CE_ideal_MMSE(M, K, L, Y, pilots, rho_ul, N, Beta);  %                 %         UL_SINR_MMSE_sp(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);                % ===== MLE --> SINR =====%        H_est = CE_MLE(M, K, L, Y, pilots, rho_ul, N, Beta);                         UL_SINR_MLE_sp(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);                %% wgcpa        [P, N] = F_WGCPA(L,K,S,Beta);                Y = received_signal(M,N,L,rho_ul,H,P);                H_est = CE_aid(M, K, L, H, P, rho_ul);                  UL_SINR_LS_wg(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,P);        %         % ===== LS --> SINR =====%%         H_est = CE_LS(M, K, L, Y, P, rho_ul, N);  %         %         UL_SINR_LS_wg(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,P);%                %         % ===== MMSE --> SINR =====%%         H_est = CE_ideal_MMSE(M, K, L, Y, P, rho_ul, N, Beta);  %                 %         UL_SINR_MMSE_wg(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,P);                % ===== MLE --> SINR =====%        H_est = CE_MLE(M, K, L, Y, P, rho_ul, N, Beta);                         UL_SINR_MLE_wg(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,P);                %% SPRS+WGC-PA               [pilots, N] = F_SPRS_WGCPA2(L,K,Beta,lambda,S);                 %===== Received Pilot Signal =====%%         Y = zeros(M,N,L);   % received uplink training sequences at i-th BS        Y = received_signal(M,N,L,rho_ul,H,pilots);                H_est = CE_aid(M, K, L, H, pilots, rho_ul);                  UL_SINR_LS_sw(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);        %         % ===== LS --> SINR =====%%         H_est = CE_LS(M, K, L, Y, pilots, rho_ul, N);  %         %         UL_SINR_LS_sw(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);%                %         % ===== MMSE --> SINR =====%%         H_est = CE_ideal_MMSE(M, K, L, Y, pilots, rho_ul, N, Beta);  %                 %         UL_SINR_MMSE_sw(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);                % ===== MLE --> SINR =====%        H_est = CE_MLE(M, K, L, Y, pilots, rho_ul, N, Beta);                         UL_SINR_MLE_sw(:, :, i_test) = F_SINR(L,K,H,H_est,rho_ul,pilots);               end%% UL SINRfigure; Tau = 20;y = 10*log10(reshape(UL_SINR_LS_ra,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'cx-');hold on;% y = 10*log10(reshape(UL_SINR_MMSE_ra,1,L*K*Test_num));% ymin=min(y); % ymax=max(y); % x=linspace(ymin,ymax,Tau); % yy=hist(y,x);% yy=yy/length(y)/(x(2)-x(1));% s=0; % for i=2:length(x) %     s=[s,trapz(x([1:i]),yy([1:i]))]; % end % plot(x,s,'c*-');% hold on;y = 10*log10(reshape(UL_SINR_MLE_ra,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'co-');hold on;y = 10*log10(reshape(UL_SINR_LS_sp,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'bx-');hold on;% % y = 10*log10(reshape(UL_SINR_MMSE_sp,1,L*K*Test_num));% ymin=min(y); % ymax=max(y); % x=linspace(ymin,ymax,Tau); % yy=hist(y,x);% yy=yy/length(y)/(x(2)-x(1));% s=0; % for i=2:length(x) %     s=[s,trapz(x([1:i]),yy([1:i]))]; % end % plot(x,s,'b*-');% hold on;y = 10*log10(reshape(UL_SINR_MLE_sp,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'bo-');hold on;y = 10*log10(reshape(UL_SINR_LS_wg,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'gx-');hold on;% % y = 10*log10(reshape(UL_SINR_MMSE_wg,1,L*K*Test_num));% ymin=min(y); % ymax=max(y); % x=linspace(ymin,ymax,Tau); % yy=hist(y,x);% yy=yy/length(y)/(x(2)-x(1));% s=0; % for i=2:length(x) %     s=[s,trapz(x([1:i]),yy([1:i]))]; % end % plot(x,s,'g*-');% hold on;y = 10*log10(reshape(UL_SINR_MLE_wg,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'go-');hold on;y = 10*log10(reshape(UL_SINR_LS_sw,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'rx-');hold on;% % y = 10*log10(reshape(UL_SINR_MMSE_sw,1,L*K*Test_num));% ymin=min(y); % ymax=max(y); % x=linspace(ymin,ymax,Tau); % yy=hist(y,x);% yy=yy/length(y)/(x(2)-x(1));% s=0; % for i=2:length(x) %     s=[s,trapz(x([1:i]),yy([1:i]))]; % end % plot(x,s,'r*-');% hold on;y = 10*log10(reshape(UL_SINR_MLE_sw,1,L*K*Test_num));ymin=min(y); ymax=max(y); x=linspace(ymin,ymax,Tau); yy=hist(y,x);yy=yy/length(y)/(x(2)-x(1));s=0; for i=2:length(x)     s=[s,trapz(x([1:i]),yy([1:i]))]; end plot(x,s,'ro-');hold on;grid on;%axis([-80 40 0 1]); title('CDF of uplink SINR');legend('Random','MLE:Random','SPRS','MLE:SPRS','WGC-PA','MLE:WGC-PA','SPRS+WGC-PA','MLE:SPRS+WGC-PA');xlabel('SINR(dB)');ylabel('CDF');hold off;toc;

⛳️ 运行结果

🔗 参考文献

X. Zhu, L. Dai and Z. Wang, "Graph Coloring Based Pilot Allocation to Mitigate Pilot Contamination for Multi-Cell Massive MIMO Systems," in IEEE Communications Letters, vol. 19, no. 10, pp. 1842-1845, Oct. 2015, doi: 10.1109/LCOMM.2015.2471304.