www.gusucode.com > LTE仿真Matlab源码 > LTE_sim_batch_IA.m
% Basic batch simulation script
% (c) 2009 by INTHFT
% www.nt.tuwien.ac.at
clear
%clear global
%close all
%clc
%% DEBUG level
global DEBUG_LEVEL
DEBUG_LEVEL = 4;
%% SNR setting
% SNR_30percent = [-7, -5, -3, -1, 1, 3, 3, 7, 9, 11, 13, 14.5, 16, 17.75, 19.5];
% SNR_stepsize = 1;
% SNR_window = 0.25;
% power = [];
% noise = [];
Simulation_type = 'IA'; %'SUSISO'
%'MUSISO'
%'SUMIMO'
%'MUMIMO'
%'SUSISO_quick_test'
%'SUSISO_BLER_curves_batch'
%'SUSISO_best_cqi'
%'SUMIMO_quick_test'
%'winner_model_example'
%'wsa_2010_michal'
% counti = 1;
% channel_estimation_error_freq_depend = zeros(72,14,1,2,500);
% Hsave = zeros(72,1000,32);
N_Ue = 1;
N_Bs = 3;
tx_mode = 6;
N_rx = 2;
N_tx = 2;
channel_type = 'VehA';
% SNR_vec = -10:2.5:30;
SNR_vec = 15;
SIR_vec = 0;
connection_table = logical(eye(3));
IA_type = 'closed_form'; % Type of IA; closed_form, min_WLI, or max_SINR
IA_streams = [1 1 1]; % Number of spatial streams
IA_thresh = 1e-5; % IA threshold for stopping iterative algorithms
IA_max_iterations = 1000;
%IA_freq_granularity_vec = [1 2 3 4 6 12 24 36 72];
%IA_freq_granularity_vec = [1 12 36 72];
IA_freq_granularity_vec = 1; % Spectral IA granularity;
% 1 ... alignment is calclulated for each subcarrier
% 12 ... alignment is calclulated for each resource block
% etc.
IA_time_granularity = 14; % IA time granularity;
% 1 ... alignment is calclulated for each OFDM symbol
% 7 ... alignment is calculated for each slot
% 14 ... alignment is calcualted for each subframe
% etc.
receiver = 'ZF';
channel_estimation_method = 'PERFECT';
IA_sigma_H2_E2_ratio_vec = Inf; % Channel measurement error;
% Inf ... perfect channel knowledge
scheduler_type = 'round robin';
scheduler_assignment = 'static';
user_speed_vec = (0:10:50)./3.6;
% user_speed_vec = 30/3.6;
filtering = 'FastFading';
%% Actual simulations
for user_speed_i = 1:length(user_speed_vec)
user_speed = user_speed_vec(user_speed_i);
for IA_freq_granularity_i = 1:length(IA_freq_granularity_vec)
IA_freq_granularity = IA_freq_granularity_vec(IA_freq_granularity_i);
for IA_sigma_H2_E2_ratio_i = 1:length(IA_sigma_H2_E2_ratio_vec)
IA_sigma_H2_E2_ratio = IA_sigma_H2_E2_ratio_vec(IA_sigma_H2_E2_ratio_i);
for SIR_i = 1:length(SIR_vec)
Power_diff = SIR_vec(SIR_i)*ones(N_Bs,1); % difference of the relative signal power from one eNodeB to the other
for cqi_i = 9
N_subframes = 10;
LTE_load_parameters; % Single User Multiple Input Multiple Output
LTE_params.max_HARQ_retransmissions = 0;
LTE_params.channel_noise = 0;
LTE_params.ICI_est_type = 'PERFECT';
LTE_params.show_plots = false;
LTE_sim_main
% Code to generate the output filename
% output_filename = ['LTE_IA_' num2str(tx_mode) '_' num2str(N_Bs) '_' num2str(N_tx) 'x' num2str(N_rx) '_' num2str(IA_streams, '%i') '_' IA_type '_' num2str(SIR_vec(SIR_i)) '_' num2str(cqi_i) '_' num2str(N_subframes) '_' channel_type '_' LTE_params.UE_config.receiver]
% save(fullfile([output_filename '.mat']));
end
end
% throughput(IA_sigma_H2_E2_ratio_i) = mean(sum(simulation_results.cell_specific.throughput_useful,3)/LTE_params.Tsubframe/1e6);
end
% IA_channel_error_vec(IA_freq_granularity_i) = IA_channel_error;
% throughput(IA_freq_granularity_i) = mean(sum(simulation_results.cell_specific.throughput_useful,3)/LTE_params.Tsubframe/1e6);
end
throughput(user_speed_i) = mean(sum(simulation_results.cell_specific.throughput_useful,3)/LTE_params.Tsubframe/1e6);
end
% plot(IA_sigma_H2_E2_ratio_vec,throughput,'bo-')
% plot(IA_freq_granularity_vec,throughput,'bo-');
% grid on;
throughput
plot(user_speed_vec*3.6,throughput,'bo-');
grid on;
% N_Ue = 1;
% N_Bs = 3;
% tx_mode = 2;
% N_rx = 2;
% N_tx = 2;
% channel_type = 'PedA';
% SNR_vec = -10:2.5:30;
% SIR_vec = Inf;
% connection_table = logical(eye(3));
% IA_type = 'closed_form';
% IA_streams = [1 1 1];
%
% %% Actual simulations
% for SIR_i = 1:length(SIR_vec)
% Power_diff = SIR_vec(SIR_i)*ones(N_Bs,1); % difference of the relativ signal power from one eNodeB to the other
% for cqi_i = 1:15
% N_subframes = 1000;
% LTE_load_parameters; % Single User Multiple Input Multiple Output
% LTE_params.show_plots = false;
% LTE_sim_main
% % Code to generate the output filename
% output_filename = ['LTE_IA_' num2str(tx_mode) '_' num2str(N_Bs) '_' num2str(N_tx) 'x' num2str(N_rx) '_' num2str(IA_streams, '%i') '_' IA_type '_' num2str(SIR_vec(SIR_i)) '_' num2str(cqi_i) '_' num2str(N_subframes) '_' channel_type]
% save(fullfile([output_filename '.mat']));
% end
% end
%
% N_Ue = 1;
% N_Bs = 3;
% tx_mode = 3;
% N_rx = 2;
% N_tx = 2;
% channel_type = 'PedA';
% SNR_vec = -10:2.5:30;
% SIR_vec = Inf;
% connection_table = logical(eye(3));
% IA_type = 'closed_form';
% IA_streams = [1 1 1];
%
% %% Actual simulations
% for SIR_i = 1:length(SIR_vec)
% Power_diff = SIR_vec(SIR_i)*ones(N_Bs,1); % difference of the relativ signal power from one eNodeB to the other
% for cqi_i = 1:15
% N_subframes = 1000;
% LTE_load_parameters; % Single User Multiple Input Multiple Output
% LTE_params.show_plots = false;
% LTE_sim_main
% % Code to generate the output filename
% output_filename = ['LTE_IA_' num2str(tx_mode) '_' num2str(N_Bs) '_' num2str(N_tx) 'x' num2str(N_rx) '_' num2str(IA_streams, '%i') '_' IA_type '_' num2str(SIR_vec(SIR_i)) '_' num2str(cqi_i) '_' num2str(N_subframes) '_' channel_type]
% save(fullfile([output_filename '.mat']));
% end
% end
%
% N_Ue = 1;
% N_Bs = 3;
% tx_mode = 4;
% N_rx = 2;
% N_tx = 2;
% channel_type = 'PedA';
% SNR_vec = -10:2.5:30;
% SIR_vec = Inf;
% connection_table = logical(eye(3));
% IA_type = 'closed_form';
% IA_streams = [1 1 1];
%
% %% Actual simulations
% for SIR_i = 1:length(SIR_vec)
% Power_diff = SIR_vec(SIR_i)*ones(N_Bs,1); % difference of the relativ signal power from one eNodeB to the other
% for cqi_i = 1:15
% N_subframes = 1000;
% LTE_load_parameters; % Single User Multiple Input Multiple Output
% LTE_params.show_plots = false;
% LTE_sim_main
% % Code to generate the output filename
% output_filename = ['LTE_IA_' num2str(tx_mode) '_' num2str(N_Bs) '_' num2str(N_tx) 'x' num2str(N_rx) '_' num2str(IA_streams, '%i') '_' IA_type '_' num2str(SIR_vec(SIR_i)) '_' num2str(cqi_i) '_' num2str(N_subframes) '_' channel_type]
% save(fullfile([output_filename '.mat']));
% end
% end
% shutdown(10)
% simulation_results.cell_specific.BER_uncoded_overall
% simulation_results.cell_specific.BER_coded_overall
% mean(sum(simulation_results.cell_specific.throughput_uncoded,3)/LTE_params.Tsubframe/1e6)
% mean(sum(simulation_results.cell_specific.throughput_coded,3)/LTE_params.Tsubframe/1e6)
% confidence_intervals_color = [0.8 0.8 0.8];
% %% Cell throughtput plot
% cell_throughput_plot_figure = figure;
%
% % Plot total throughput (sum of all streams)
% cell_throughput_coded = sum(simulation_results.cell_specific.throughput_coded,3)/LTE_params.Tsubframe/1e6;
% plot(SNR_vec,mean(cell_throughput_coded,1),'.-b','Markersize',5);
% hold on
% cell_throughput_uncoded = sum(simulation_results.cell_specific.throughput_uncoded,3)/LTE_params.Tsubframe/1e6;
% plot(SNR_vec,mean(cell_throughput_uncoded,1),'.-r','Markersize',5);
% %
% % LTE_plot_confidence_intervals(SNR_vec, confidence_intervals_color ,cell_throughput_plot_figure,'mean', cell_throughput_coded);
% % LTE_plot_confidence_intervals(SNR_vec, confidence_intervals_color ,cell_throughput_plot_figure,'mean', cell_throughput_uncoded);
%
% legend('cell coded throughput','cell uncoded throughput','Location','best');
% xlabel('SNR [dB]');
% ylabel('Throughput [Mbit/s]');
% title('Cell throughput');
% hold off
% grid on