基于蒙特卡洛的信道LOS,NLOS对5G网络通信性能的影响matlab仿真分析【包含程序操作录像】

%-根据均匀空间泊松分布部署的小小区基站 %密度“λ”的点过程（SPPP）。BS的平均数量 %被认为是“N_points” %-我们将分析重点放在位于 %网络（在原点） %-我们假设基站使用频率ALOHA进行传输，即每个BS %在一个（随机且均匀选择）信道上传输 %“freq_reuse”通道 %-路径损耗：具有LOS和NLOS路径的双斜率，具有瑞利衰落 %指数分布幂~exp（1） clear all; close all; %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% PARAMTERS to be set %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % network geometry parameters lambda = 10/1e6; % intestity of SPPP (tested for values from 1 to 10^4 BSs/sqkm) N_points = 50000; % number of simulated BSs lambda_users = 1e3/1e6; % density fo users (num of users is deterministic N_iteration = 10000; % number of simulation snapshots % channel model parameters LOS_function = 1; % LOS likelihood function: 1 = ~exp(-x^2); 2 = ~exp(-x); 3 = 3GPP function beta_L = 2.09; % LOS attenuation exponent = 2.09 beta_NL = 3.75; % NLOS attenuation exponent = 3.75 K_L_3GPP = 103.8; % LOS attenuation at 1 km K_NL_3GPP = 145.4; % NLOS attenuation at 1 km fading_mean = 1; % fading power (as long as fading is included, the fading mean does not impact the SINR) L = 69/sqrt(log(1/0.5)); % LOS likelihood parameter approx. 82.5m % PHY parameters SIR_threshold = 10^(-8/10); % SIR threshold = -8dB freq_reuse = 1; % Frequency ALOHA reuse factor %%%%% Dependant PARAMETERS edgeLength = sqrt( N_points / lambda); % length of the square edge area = edgeLength * edgeLength ; % area of the SPPP to simulate K_L = 10^( -( K_L_3GPP - beta_L*10*3) / 10); % LOS attenuation at 1m K_NL = 10^( -( K_NL_3GPP - beta_NL*10*3) / 10); % NLOS attenuation at 1m tic % start timer %%%%%%%%%%%%%%%%%%%%%%%%%%%%%% FUNCTIONS %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% PL_LOS = @(x) K_L .* x .^( - beta_L); % LOS path-loss function PL_NLOS = @(x) K_NL .* x .^( - beta_NL); % NLOS path-loss function p_LOS = @(x) exp( -(x/L).^2); % LOS likelihood function ~exp(-x^2) p_LOS_exp = @(x) exp( -x/L); % LOS likelihood function ~exp(-x) p_LOS_3GPP = @(x) 0.5-min(0.5.*ones(size(x)),5*exp(-156./x))+... min(0.5.*ones(size(x)), 5*exp(-x./30)); % 3GPP LOS likelihood function [36.814 - Urban pico-cells] %%%%%%%%%%%%%%%%%%%%%%%% INITIALIZE VARIABLES %%%%%%%%%%%%%%%%%%%%%%%%%% % Define variable to save data for all snapshots SIR_vector = zeros(1,N_iteration); rate_vector = zeros(1,N_iteration); ASE_vector = zeros(1,N_iteration); for n_iter = 1:N_iteration numOfBSs = poissrnd(lambda.*area); % generate number of BSs bs_position = edgeLength*( ( rand(1,numOfBSs) - 0.5) + 1i*(rand(1,numOfBSs) - 0.5) ); % position of the BSs distance_from_BS = abs( bs_position ); % distance of the points from the origin % process only active BSs p_active = 1/freq_reuse; % probability of a BS being an interferer drawVariable_active = rand(1,numOfBSs); % draw random Bernoulli variable for active BSs active_BS_boolean = drawVariable_active < p_active; % find boolean indices of active BSs % determine LOS likelihood depending on the distance switch LOS_function case 1 p_of_having_LOS = p_LOS(distance_from_BS); % prob to have LOS case 2 p_of_having_LOS = p_LOS_exp(distance_from_BS); % prob to have LOS case 3 p_of_having_LOS = p_LOS_3GPP(distance_from_BS); % prob to have LOS otherwise error('No other LOS likelihood functions'); end drawVariable_LOS = rand(size(bs_position)); % draw random Bernoulli variable for LOS/NLOS LOS_boolean = drawVariable_LOS <= p_of_having_LOS; LOS_idx = find(LOS_boolean); % indices of LOS BSs NLOS_idx = find(~LOS_boolean); % indices of NLOS BSs P_rx = zeros(1,numOfBSs); % vector of user's received power from different BSs P_rx(LOS_idx) = PL_LOS(distance_from_BS(LOS_idx)); % user's received power from LOS BSs P_rx( NLOS_idx ) = PL_NLOS(distance_from_BS(NLOS_idx)); % user's received power from NLOS BSs if numOfBSs > 0 [max_Prx, serving_BSs_idx]= max(P_rx); % find index of serving base (the one providing the highest power) inactive_BSs_idx = find( ~active_BS_boolean ); % find indices of inactive BSs interf_effective = P_rx.* exprnd(fading_mean,1,length(P_rx)); % vector of total interference with added exponential fading component interf_effective(unique([serving_BSs_idx inactive_BSs_idx]) ) = 0; % remove serving BS and inactive BSs from interferers rx_pw_tmp = max_Prx .* exprnd(fading_mean,1,1); % add exponential fading to received power from serving BS tot_int_tmp = sum( interf_effective ); % total interference SIR_linear = rx_pw_tmp ./ tot_int_tmp; % compute SIR SIR_vector(n_iter) = SIR_linear; % copy SIR for this snapshot into the SIR vector rate_vector(n_iter) = log2(1+SIR_linear); % copy spectral efficiency (SE) for this snapshot into the SE vector ASE_vector(n_iter) = sum(active_BS_boolean)*log2(1+SIR_linear)/area/freq_reuse; % copy Area Spectral Efficiency (ASE) for this snapshot into the ASE vector end if fix(n_iter/N_iteration*100/10) > fix((n_iter-1)/N_iteration*100/10) disp([ num2str(fix(n_iter/N_iteration*100/10)*10) '% completed']) % display completed simulation perc. end end toc % display results disp(' '); disp(['The coverage is : ' num2str(100*sum(SIR_vector>SIR_threshold)/N_iteration) '%'] ); disp(' '); disp(['The average rate is : ' num2str(mean(rate_vector)) ' bps'] ); disp(' '); disp(['The ASE is : ' num2str(mean( ASE_vector)*10^6) '*10^-6 bps/(Hz*sqm)'] ); disp(' ');