1.

A random process is given as X (t) =At, where A is uniformly

distributed random variable on (-2, 3). Check if the given RP is WSS
RP or not by using any language (MATLAB, JAVA….) with clear
simulation result?

A random process X(t) is Wide-Sense Stationary (WSS) if:

1. Mean is constant: E[X(t)] does not depend on t.

2. Autocorrelation is a function of ∣t1−t2∣ only: Rxx(t1, t2) = E[X(t1)

X(t2)] depends only on the time difference.

For X(t) = A⋅t, we need to check:

 E[X(t)] = E[A]⋅t (Does this depend on time?)

 Rxx(t1,t2)=E[A⋅t1⋅A⋅t2]

Explanation:

1. Set Parameters: Define the limits for the uniform distribution of AA,
the number of samples T, and the time vector t that ranges from 0 to
10.

2. Simulate random variable A: Generate T random values from a

uniform distribution between A_{min} and A_{max}.

3. Calculate X(t)=A⋅t: Compute the matrix Xt, where each element in

the matrix is the result of multiplying a corresponding value of A by the
time vector t.

4. Calculate the mean of X(t): Find the mean value of each column in
Xt (i.e., the mean across all realizations for each time point).

5. Plot Mean of X(t): Create a plot of meanX against t, with proper titles
and labels.

6. Calculate Autocorrelatione: Compute the correlation of Xt, which

measures how values of X(t) at different times relate to each other.

7. Plot Autocorrelation: Create a heatmap of the autocorrelation

matrix, using t as both axes, with a color bar to represent the
covariance values.

MATLAB Code:
% Parameters

A_min = -2; A_max = 3; % Uniform distribution for A

T = 1000; % Number of samples

t = linspace(0, 10, 100); % Time vector

% Calculate random variable A

A = (A_max - A_min) * rand(1, T) + A_min;

% Calculate X(t) = A * t

Xt = A' * t; % Create a matrix with each row being A*t

% Calculate mean of X(t)

meanX = mean(Xt, 1);

% Plot Mean of X(t)

figure;

plot(t, meanX);

title('Mean of X(t) over time');

xlabel('Time t');

ylabel('Mean of X(t)');

% Autocorrelation matrix of X(t)

autocorr = corr(Xt');

figure;

imagesc(t, t, autocorr);

colorbar;
title('Autocorrelation of X(t)');

xlabel('Time t_1');

ylabel('Time t_2');

SIMULATION

Therefore

 The mean of X(t) should vary linearly with t, as E[X(t)] = E[A] ⋅t, which
is not constant.

show it does not only depend on the time difference ∣t1−t2∣.

 The autocorrelation depends on both t1 and t2, but we compute it to

Since the mean is time-dependent, this process is not WSS.

2. Given a random process X(t)=cos(wt+θ) is uniformly distributed

in the interval(−π/2,π/2), then check this RP if it is time dependent by
using any language with clear simulation?
A random process is time-dependent if its statistical properties (such as
mean,and autocorrelation) change over time.

For a uniform distribution over (a, b), the mean and autocorrelation are fixed.
However, checking if a process is time-dependent requires ensuring that
these properties do not change with time.

Explanation:

1. Set Parameters:

o Set the frequency w as 2π (a constant).

o Define T, the number of samples, and generate a time vector t

ranging from 0 to 10 with 100 points.

2. Generate Random θ:

o Generate T random samples from a uniform distribution in the

interval (−π/2,π/2).

3. Calculate X(t)=cos(wt+θ)

o For each value in t, calculate X(t) using the formula cos(wt + θ) ,

where θ is an array of random values.

4. Plot X(t):

o Create a plot with t on the x-axis and the computed X(t) on the y-
axis. Label the plot and axes accordingly.

5. Compute Mean and autocorrelation:

o Compute the mean of X(t) across the samples (along rows for
each time value).

o Compute the autocorrelation of X(t) across the samples.

6. Plot Mean and autocorrelation:

o Create a subplot with two graphs:

 One graph showing the mean of X(t) over time.

 Another graph showing the variance of X(t) over time.

o Add appropriate titles and labels for both plots.

MATLAB Code:
% Parameters

omega = 1; % Constant frequency

num_samples = 1000; % Number of samples

num_time_points = 100; % Number of time points

t = linspace(0, 10, num_time_points); % Time vector

% Generate uniformly distributed random phase theta in the

interval (-pi/2, pi/2)

theta_samples = unifrnd(-pi/2, pi/2, [num_samples, 1]);

% Calculate the random process X(t)

X_t = zeros(num_samples, num_time_points);

for i = 1:num_samples

X_t(i, :) = cos(omega * t + theta_samples(i));

end

% Mean calculation

mean_X_t = mean(X_t, 1);

% Autocorrelation calculation

R_X = zeros(num_time_points, num_time_points);

for i = 1:num_time_points

for j = 1:num_time_points

R_X(i, j) = mean(X_t(:, i) .* X_t(:, j));

end

end
% Plotting Mean

figure;

subplot(1, 2, 1);

plot(t, mean_X_t, 'b', 'LineWidth', 2);

title('Mean of Random Process X(t)');

xlabel('Time (t)');

ylabel('Mean');

grid on;

% Plotting Autocorrelation

subplot(1, 2, 2);

imagesc(t, t, R_X);

colorbar;

title('Autocorrelation of X(t)');

xlabel('t_1');

ylabel('t_2');

axis xy; % To display the correct orientation of the axes

grid on;

% Adjust layout

sgtitle('Mean and Autocorrelation of X(t) = cos(\omega t + \Theta)');

SIMULATION
Therefore

Since the uniform distribution has not constant mean and

autocorrelation , meaning the random process is time-dependent.

3. We have a random process X(t)=Acos(wt + θ) then prove if this

RP is stationary or not by using any language with clear and neat
simulation result?

For a process X(t), to check if it's stationary, we typically compute the mean,
and autocorrelation over time. If any of these properties change with time,
the process is not stationary.

Explanation:

1. Calculate X(t)=Acos(wt + θ):

o Compute the values of X(t) for each time value t by evaluating

the cosine function with frequency w and phase shift θ.

2. Compute Mean and Variance:

o Mean: Calculate the mean of X(t) for each time point (across all
realizations/samples).

o Variance: Calculate the variance of X(t) for each time point.

3. Plot Mean and Variance:

o Create a figure for plotting.

o Subplot 1: Plot the mean of X(t) against t. Set the title and axis
labels accordingly.

o Subplot 2: Plot the variance of X(t) against t. Set the title and
axis labels accordingly.

MATLAB Code:

% Parameters% Parameters

A = 1; % Amplitude

w_c = 2 * pi; % Frequency (1 Hz)

num_samples = 1000; % Number of samples for B

time_vector = linspace(0, 10, 100); % Time vector from 0 to 10

seconds

% Generate random samples of B

B_samples = unifrnd(-pi/2, pi/2, [num_samples, 1]); % B uniformly

distributed in (-90, 90)

% Initialize X(t)

X_t = zeros(length(time_vector), num_samples);

% Compute X(t) for each sample of B

for i = 1:num_samples

X_t(:, i) = A * cos(w_c * time_vector + B_samples(i));

end

% Calculate mean and variance over the samples

mean_X = mean(X_t, 2);

autocorr = corr(X_t);

% Plot results

figure;

% Mean plot

subplot(2, 1, 1);

plot(time_vector, mean_X);

title('Mean of X(t)');
xlabel('Time (s)');

ylabel('Mean');

grid on;

% Variance plot

subplot(2, 1, 2);

plot(autocorr);

title('AUTOCORRELATION of X(t)');

xlabel('Time (s)');

ylabel('autocorr');

grid on;

SIMULATION
Therefore

 The process X(t) will have a mean and variance that changes with
time, indicating that the process is not stationary.

4. Write a simulation code for messages receiving in to your mobile

phone per a day by giving an example?

This is a typical example of a Poisson process, where the number of

messages received per day follows a Poisson distribution.

Explanation:

1. Set Parameters:

o Set the average number of messages per day λ as 5.

o Set the number of days to simulate for one year (365 days).

2. Generate Poisson-Distributed Messages:

o Use the Poisson distribution to simulate the number of messages

received each day. Each day has a Poisson-distributed number of
messages with a mean of λ=5 .

3. Plot Histogram:

o Create a histogram of the number of messages received per day.

o Display the histogram with 30 bins.

o Set the plot's title and labels for the x and y axes.

4. Display Statistics:

o Compute and print the mean of the number of messages

received per day.

o Compute and print the variance of the number of messages

received per day.

MATLAB Code:

% Parameters

lambda = 50; % Average number of messages per day (Poisson rate)

days = 1; % Number of days to simulate

num_simulations = 1000; % Number of simulations

% Simulate the number of messages received in one day

messages_received = poissrnd(lambda, num_simulations, days);

% Calculate statistics

mean_messages = mean(messages_received);

var_messages = var(messages_received);

% Display results

fprintf('Mean number of messages received in a day: %.2f\n',

mean_messages);

fprintf('Variance of messages received in a day: %.2f\n',

var_messages);

% Plot histogram of the results

figure;

histogram(messages_received, 'Normalization', 'probability');

title('Histogram of Messages Received in One Day');

xlabel('Number of Messages Received');

ylabel('Probability');

grid on;

% Plotting the probability mass function (PMF)

x = 0:max(messages_received);

pmf = poisspdf(x, lambda);

figure;

bar(x, pmf, 'FaceColor', [0.2 0.6 1]);

title('Poisson Probability Mass Function for Messages Received');

xlabel('Number of Messages');

ylabel('Probability');
grid on;

SIMULATION

Therefore

 The number of messages received per day is modeled as a Poisson

process with an average of 5 messages per day. The distribution will
show the count of how many days had 0, 1, 2, etc., messages.