line plots

For the below mentioned code the plots for the Figures:1,3,5,6, is not getting ploted for entire range of values x-axis. and I cannot see the Fig 7 at all.

Kindly help and guide.

Thank You.

clear; clc; close all;

%% --- Common Setup ---

set(0, 'defaultAxesFontName', 'Times New Roman');

set(0, 'defaultTextInterpreter', 'tex');

set(0, 'defaultAxesFontSize', 11);

%% --- Simulation Parameters ---

c = 3e8; % Speed of light [m/s]

fc = 5.9e9; % Carrier frequency [Hz]

lambda = c / fc; % Wavelength [m]

N_realizations = 100; % Monte Carlo realizations

N_paths = 20; % Number of NLOS subpaths per cluster

L = 5; % Number of clusters

sigma_CSI = 0.1; % Std. dev. of CSI phase error [rad]

G_RIS = 5; % RIS amplification gain

K = 10; % Rice factor

sigma_a = 0.1; % Acceleration variance [m/s^2]

sigma_tau_base = [40e-9, 60e-9, 80e-9, 100e-9, 120e-9]; % Base delay spread per scenario [s]

sigma_xi = 4; % Shadowing std. dev. [dB]

v_max = 200 / 3.6; % Max speed [m/s]

T_values = linspace(0.5e-3, 5e-3, 20); % Feedback latency [ms]

M_T = 4; % Number of transmit antennas

V_values = round(linspace(8, 80, 20)); % Number of RIS elements, integer values

P_max = 10^(43/10) / 1000; % Max BS power [W]

P_v = 0.01; % RIS element power [W]

P_circuit = 5; % Circuit power [W]

chi_max = 2; % Max RIS amplification

% Mobility scenarios

scenarios = {'Dynamic Acc.', 'Steady Acc.', 'Uniform Vel.', 'Mobile-to-Static', 'Static-to-Static'};

v0 = [10, 10, 10, 10, 0]; % Initial velocities [m/s]

a = [6, 3, 0, 0, 0]; % Accelerations [m/s^2]

kappa = [60, 50, 40, 30, 20]; % TCF decay rates [1/s]

times = linspace(0, 4, 20); % 20 points for full coverage

kappa_scaling = 0.2; % Scaling for Dynamic Acc.

chi_l = 0.8 * ones(1, L); % Cluster gains

tau_0 = [150e-9, 180e-9, 200e-9, 220e-9, 250e-9]; % PDP decay constants [s]

% Geometry parameters

d_tx_rx_base = [40, 60, 80, 100, 120]; % Base Tx-Rx distance per scenario [m]

d_tx_ris = 100; % Distance Tx-RIS [m]

d_ris_rx = 100; % Distance RIS-Rx [m]

theta_AoA = rand(N_paths, L) * pi; % Random AoA per cluster [rad]

theta_AoD = rand(N_paths, L) * pi; % Random AoD per cluster [rad]

phi_AoA = rand(N_paths, L) * 2 * pi; % Random azimuth AoA [rad]

t = linspace(0, 0.04, 100); % Time difference for TCF [s]

delta_f = linspace(0, 1e6, 100); % Frequency separation for FCF [Hz]

tau = linspace(0, 700e-9, 50); % Delay for PDP [s]

%% --- Monte Carlo Simulation ---

sigma_D_scenarios = nan(length(scenarios), length(times));

Tc_scenarios = nan(length(scenarios), length(times));

Bc_scenarios = nan(length(scenarios), length(times));

EE_scenarios = nan(length(scenarios), length(times));

PL_LOS_scenarios = nan(length(scenarios), length(times));

PL_NLOS_scenarios = nan(length(scenarios), length(times));

G_RIS_scenarios = nan(length(scenarios), length(times));

v_x = nan(length(scenarios), length(times)); % Velocity [m/s]

sigma_tau_x = nan(length(scenarios), length(times)); % Delay spread [ns]

d_tx_rx_x = nan(length(scenarios), length(times)); % Tx-Rx distance [m]

d_total_x = nan(length(scenarios), length(times)); % Total cluster path distance [m]

for s = 1:length(scenarios)

for ti = 1:length(times)

sigma_D_temp = 0;

Tc_temp = 0;

Bc_temp = 0;

EE_temp = 0;

PL_LOS_temp = 0;

PL_NLOS_temp = 0;

G_RIS_temp = 0;

d_total_temp = 0;

v_x(s, ti) = v0(s) + a(s) * times(ti) + s * times(ti) * 0.1; % Increased offset

sigma_tau_x(s, ti) = sigma_tau_base(s) * (1 + 0.6 * times(ti)); % Increased scaling

d_tx_rx_x(s, ti) = d_tx_rx_base(s) * (1 + 0.6 * times(ti)); % Increased scaling

valid_realizations = 0;

for r = 1:N_realizations

[sigma_D_r, Tc_r, Bc_r, EE_r, PL_LOS_r, PL_NLOS_r, G_RIS_r, d_total_r] = compute_metrics(t, delta_f, tau, v0(s), a(s), times(ti), kappa(s), tau_0(s), N_paths, L, d_tx_rx_x(s, ti), d_tx_ris, d_ris_rx, ...

theta_AoA, theta_AoD, phi_AoA, fc, c, G_RIS, sigma_CSI, sigma_a, lambda, s == 1, kappa_scaling, K, chi_l, sigma_tau_x(s, ti), sigma_xi, M_T, V_values(ti), P_max, P_v, P_circuit, chi_max, v_max, T_values(ti));

if all(isfinite([sigma_D_r, Tc_r, Bc_r, EE_r, PL_LOS_r, PL_NLOS_r, G_RIS_r, d_total_r]))

sigma_D_temp = sigma_D_temp + sigma_D_r;

Tc_temp = Tc_temp + Tc_r;

Bc_temp = Bc_temp + Bc_r;

EE_temp = EE_temp + EE_r;

PL_LOS_temp = PL_LOS_temp + PL_LOS_r;

PL_NLOS_temp = PL_NLOS_temp + PL_NLOS_r;

G_RIS_temp = G_RIS_temp + G_RIS_r;

d_total_temp = d_total_temp + d_total_r;

valid_realizations = valid_realizations + 1;

end

end

if valid_realizations > 0

sigma_D_scenarios(s, ti) = sigma_D_temp / valid_realizations + s * 20; % Offset

Tc_scenarios(s, ti) = Tc_temp / valid_realizations + s * 0.1e-3; % Offset in ms

Bc_scenarios(s, ti) = Bc_temp / valid_realizations + s * 1e3; % Offset in Hz

EE_scenarios(s, ti) = EE_temp / valid_realizations * (1 + 0.2 * s); % Scaling

PL_LOS_scenarios(s, ti) = PL_LOS_temp / valid_realizations + s * 3; % Offset in dB

PL_NLOS_scenarios(s, ti) = PL_NLOS_temp / valid_realizations + s * 3; % Offset in dB

G_RIS_scenarios(s, ti) = G_RIS_temp / valid_realizations + s * 0.8; % Offset in dB

d_total_x(s, ti) = d_total_temp / valid_realizations + s * times(ti) * 3; % Increased offset

end

end

end

%% --- Figure: Doppler Spread vs. Velocity ---

figure('Position', [100 100 900 500]);

colors = {'r-o', 'b--s', 'g:^', 'm-.d', 'k-*'};

for s = 1:length(scenarios)

plot(v_x(s, :), sigma_D_scenarios(s, :), colors{s}, 'LineWidth', 2, 'MarkerSize', 8); hold on;

end

grid on;

xlabel('Velocity [m/s]');

ylabel('Doppler Spread, \sigma_D [Hz]');

legend(scenarios, 'Location', 'northwest', 'FontSize', 10);

title('Doppler Spread vs. Velocity for Different Mobility Scenarios');

set(gca, 'Box', 'on', 'FontSize', 11, 'XLim', [0 35]);

%% --- Figure: Coherence Time vs. Time ---

figure('Position', [1050 100 900 500]);

for s = 1:length(scenarios)

plot(times, Tc_scenarios(s, :) * 1000, colors{s}, 'LineWidth', 2, 'MarkerSize', 8); hold on;

end

grid on;

xlabel('Time, t [s]');

ylabel('Coherence Time, T_c [ms]');

legend(scenarios, 'Location', 'northwest', 'FontSize', 10);

title('Coherence Time vs. Time for Different Mobility Scenarios');

set(gca, 'Box', 'on', 'FontSize', 11, 'XLim', [0 4]);

%% --- Figure: Coherence Bandwidth vs. Delay Spread ---

figure('Position', [2000 100 900 500]);

for s = 1:length(scenarios)

plot(sigma_tau_x(s, :) * 1e9, Bc_scenarios(s, :) / 1e3, colors{s}, 'LineWidth', 2, 'MarkerSize', 8); hold on;

end

grid on;

xlabel('Delay Spread, \sigma_\tau [ns]');

ylabel('Coherence Bandwidth, B_c [kHz]');

legend(scenarios, 'Location', 'northwest', 'FontSize', 10);

title('Coherence Bandwidth vs. Delay Spread for Different Mobility Scenarios');

set(gca, 'Box', 'on', 'FontSize', 11, 'XLim', [0 300]);

%% --- Figure: Energy Efficiency vs. Feedback Latency ---

figure('Position', [100 650 900 500]);

for s = 1:length(scenarios)

plot(T_values * 1e3, EE_scenarios(s, :), colors{s}, 'LineWidth', 2, 'MarkerSize', 8); hold on;

end

grid on;

xlabel('Feedback Latency, T [ms]');

ylabel('Energy Efficiency, EE [Mbits/Joule]');

legend(scenarios, 'Location', 'northwest', 'FontSize', 10);

title('Energy Efficiency vs. Feedback Latency for Different Mobility Scenarios');

set(gca, 'Box', 'on', 'FontSize', 11, 'YScale', 'log', 'XLim', [0.5 5]);

%% --- Figure: Path Loss (LOS) vs. Tx-Rx Distance ---

figure('Position', [1050 650 900 500]);

for s = 1:length(scenarios)

plot(d_tx_rx_x(s, :), PL_LOS_scenarios(s, :), colors{s}, 'LineWidth', 2, 'MarkerSize', 8); hold on;

end

grid on;

xlabel('Tx-Rx Distance, d_{tx-rx} [m]');

ylabel('LOS Path Loss [dB]');

legend(scenarios, 'Location', 'northwest', 'FontSize', 10);

title('LOS Path Loss vs. Tx-Rx Distance for Different Mobility Scenarios');

set(gca, 'Box', 'on', 'FontSize', 11, 'XLim', [0 300]);

%% --- Figure: Path Loss (NLOS) vs. Total Cluster Path Distance ---

figure('Position', [2000 650 900 500]);

for s = 1:length(scenarios)

plot(d_total_x(s, :), PL_NLOS_scenarios(s, :), colors{s}, 'LineWidth', 2, 'MarkerSize', 8); hold on;

end

grid on;

xlabel('Total Cluster Path Distance, d_{total} [m]');

ylabel('NLOS Path Loss [dB]');

legend(scenarios, 'Location', 'northwest', 'FontSize', 10);

title('NLOS Path Loss vs. Total Cluster Path Distance for Different Mobility Scenarios');

set(gca, 'Box', 'on', 'FontSize', 11, 'XLim', [0 500]);

%% --- Figure: RIS Gain vs. Number of RIS Elements ---

figure('Position', [100 1200 900 500]);

for s = 1:length(scenarios)

plot(V_values, G_RIS_scenarios(s, :), colors{s}, 'LineWidth', 2, 'MarkerSize', 8); hold on;

end

grid on;

xlabel('Number of RIS Elements, V');

ylabel('RIS Gain [dB]');

legend(scenarios, 'Location', 'northwest', 'FontSize', 10);

title('RIS Gain vs. Number of RIS Elements for Different Mobility Scenarios');

set(gca, 'Box', 'on', 'FontSize', 11, 'XLim', [8 80]);

%% --- Function Definition ---

function [sigma_D, Tc, Bc, EE, PL_LOS, PL_NLOS, G_RIS, d_total] = compute_metrics(t, delta_f, tau, v, a, t_eval, kappa_s, tau_0_s, N_paths, L, d_tx_rx, d_tx_ris, d_ris_rx, theta_AoA, theta_AoD, phi_AoA, fc, c, G_RIS, sigma_CSI, sigma_a, lambda, is_dynamic_acc, kappa_scaling, K, chi_l, sigma_tau, sigma_xi, M_T, V, P_max, P_v, P_circuit, chi_max, v_max, T)

% Initialize

h = zeros(N_paths * L + 1, length(tau)); % CIR

f_D = zeros(N_paths * L + 1, 1); % Doppler shifts

h_eff = zeros(M_T, 1); % Effective channel

h_direct = zeros(M_T, 1); % Direct channel

r_MT = [0, 0, 0]'; % MT position

r_MR = [max(d_tx_rx, 1e-3), 0, 0]'; % MR position, ensure non-zero

r_C = rand(3, L) * 150 + [d_tx_ris; 0; 0]; % Cluster positions, increased range

v_MT = [v + a * t_eval, 0, 0]'; % MT velocity

v_MR = [v + a * t_eval, 0, 0]'; % MR velocity (simplified)

v_C = randn(3, L) * sigma_a; % Cluster velocities

lambda_CSI = besselj(0, 2 * pi * (v_max * fc / c) * T); % CSI correlation

% LOS component

tau_LOS = d_tx_rx / c;

u_LOS = (r_MR - r_MT) / norm(r_MR - r_MT);

f_D_LOS = dot(v_MR - v_MT, u_LOS) / lambda;

h_LOS = sqrt(K / (K + 1) / (4 * pi * d_tx_rx / c)) * exp(1j * 2 * pi * f_D_LOS * t_eval) * exp(1j * normrnd(0, sigma_CSI));

h(1, :) = h_LOS * exp(-1j * 2 * pi * fc * (tau - tau_LOS));

f_D(1) = f_D_LOS;

h_direct = h_LOS * ones(M_T, 1); % Simplified MIMO

% NLOS components (RIS-reflected)

idx = 2;

H_r_b = sqrt(1 / (4 * pi * d_tx_ris / c)) * exp(1j * randn(M_T, V)); % BS-RIS channel

h_r = sqrt(1 / (4 * pi * d_ris_rx / c)) * exp(1j * randn(V, 1)); % RIS-MR channel

Theta = diag(chi_max * exp(1j * 2 * pi * rand(V, 1))); % RIS matrix

d_total = 0;

for ell = 1:L

tau_NLOS = (norm(r_C(:, ell) - r_MT) + norm(r_MR - r_C(:, ell))) / c;

d_total_ell = norm(r_C(:, ell) - r_MT) + norm(r_MR - r_C(:, ell));

if d_total_ell < 1e-3

d_total_ell = 1e-3; % Prevent zero distance

end

d_total = d_total + d_total_ell / L;

for n = 1:N_paths

u_T = [cos(phi_AoA(n, ell)) * cos(theta_AoA(n, ell)), sin(phi_AoA(n, ell)) * cos(theta_AoA(n, ell)), sin(theta_AoA(n, ell))]';

u_R = [cos(phi_AoA(n, ell)) * cos(theta_AoA(n, ell)), sin(phi_AoA(n, ell)) * cos(theta_AoA(n, ell)), sin(theta_AoA(n, ell))]';

f_D_NLOS = dot(v_MT - v_C(:, ell), u_T) / lambda + dot(v_MR - v_C(:, ell), u_R) / lambda;

h_NLOS = sqrt(chi_l(ell)^2 / (K + 1) * G_RIS / (4 * pi * d_total_ell / c)) * exp(1j * 2 * pi * f_D_NLOS * t_eval) * exp(1j * normrnd(0, sigma_CSI));

h(idx, :) = h_NLOS * exp(-1j * 2 * pi * fc * (tau - tau_NLOS));

f_D(idx) = f_D_NLOS;

idx = idx + 1;

end

end

h_eff = h_direct + H_r_b * Theta * h_r; % Effective channel

% Doppler Spread

sigma_D = sqrt((K / (K + 1)) * (f_D(1) - mean(f_D))^2 + sum((chi_l.^2 / (K + 1)) .* mean(reshape((f_D(2:end) - mean(f_D)).^2, N_paths, L), 1)));

% Coherence Time

if is_dynamic_acc

kappa_s = kappa_s * (1 + t_eval * kappa_scaling);

end

Tc = log(2) / kappa_s;

% Coherence Bandwidth

FCF = (K / (K + 1)) * exp(1j * 2 * pi * delta_f * tau_LOS);

for ell = 1:L

tau_NLOS = (norm(r_C(:, ell) - r_MT) + norm(r_MR - r_C(:, ell))) / c;

FCF = FCF + (chi_l(ell)^2 / (K + 1)) * exp(1j * 2 * pi * delta_f * tau_NLOS);

end

FCF = abs(FCF) / max(abs(FCF));

Bc = delta_f(find(FCF <= 0.5, 1, 'first'));

if isempty(Bc)

Bc = delta_f(end);

end

% Energy Efficiency (simplified)

w_c = sqrt(P_max / (M_T + 1)) * ones(M_T, 1); % Common stream precoding

w_k = sqrt(P_max / (M_T + 1)) * ones(M_T, 1); % Private stream precoding

sigma_r = 1e-3; % RIS noise variance

sigma_k = 1e-3; % AWGN variance

gamma_c = (abs(w_c' * (lambda_CSI * h_eff + sqrt(1 - lambda_CSI^2) * randn(M_T, 1)))^2) / ...

(abs(w_k' * (lambda_CSI * h_eff + sqrt(1 - lambda_CSI^2) * randn(M_T, 1)))^2 + sigma_r * sum(abs(diag(Theta)).^2) + sigma_k);

R_c = log2(1 + gamma_c);

R_k = log2(1 + gamma_c); % Simplified for one CV

P_total = norm(w_c)^2 + norm(w_k)^2 + sum(abs(diag(Theta)).^2) * P_v + P_circuit;

EE = (R_c + R_k) / P_total;

% Path Loss

PL_LOS = -20 * log10(sqrt(K / (K + 1)) * lambda / (4 * pi * d_tx_rx));

PL_NLOS = 0;

for ell = 1:L

d_total_ell = norm(r_C(:, ell) - r_MT) + norm(r_MR - r_C(:, ell));

if d_total_ell < 1e-3

d_total_ell = 1e-3; % Prevent zero distance

end

PL_NLOS = PL_NLOS + (chi_l(ell)^2 / L) * (-20 * log10(sqrt(1 / (K + 1)) * lambda / (4 * pi * d_total_ell)) + normrnd(0, sigma_xi));

end

% RIS Gain

G_RIS = 10 * log10(mean(abs(h_eff).^2) / mean(abs(h_direct).^2));

end

Answer 1

[Haelee]

Hi Sanjay,

Thank you for sharing your code. I was able to reproduce the issue you mentioned.

1. Regarding Figures 1, 3, 5, and 6 not showing the full range of x values:

By default, MATLAB adjusts the display range to fit your data unless you set it manually. In your code, you set the x-axis limits for each figure with the following line such as:

set(gca, 'Box', 'on', 'FontSize', 11, 'XLim', [0 500]);

This forces the x-axis to display only the range from 0 to 500.

2. About Figure 7 not appearing:

Please check the position values you are using. For Figure 7, you set the Position property to [100 1200 900 500] with this line:

figure('Position', [100 1200 900 500]);

If your monitor's height is less than 1200 pixels, the figure may be created off-screen. You can check your screen size using

get(0, 'screensize')

to make sure the position values fit your display.

Let me know if you have any other questions or need further assistance.

Thanks.

[sanjay]

Hello Haelee,

For Fig 7 , changing "figure position" worked, Thank you for your kind help.

But for others, I removed x-axis limits ("set gca") there is marginal improvement but still the not getting ploted for entire range of values x-axis.

Thank you for your kind help and guidance.

[Haelee]

Hi @sanjay

This is what I see in Figure 5 after removing the x-axis limits. Is this different from the result you expected, or does it look different on your end?

I’m also attaching the code I’ve updated, where I removed the "XLim" settings and added

axis(gca, 'tight')

to make sure the figure automatically fits the data range.

If you still experience issues with this code, it might be related to a corrupted MATLAB preference. In that case, you can try resetting your MATLAB preferences. You can find instructions from "How do I regenerate my MATLAB preferences? "

If you continue to have issues, please let me know (1) which MATLAB release and (2) OS you are using.

[sanjay]

Got it!!

Thank you for you guidance and help!!