The Average of the Matrix G with the dimension (M,n) that I'm repeating 3 times in a loop so I'll have 3 matrices of G (M,n) and I want to take the Average of these 3 matrices so I can get only one matrix of G(M,n) represent the Average of them)

clc; clear;
M=5; n=3; freq = 28e9; lambda = physconst('LightSpeed')/freq; d_BS=lambda/2; %Inter-antenna separation atthe BS. d_IRS=lambda/2; %Inter-element separation at the IRS.
G = zeros(M,n);
for l=1:3 
    theta_2=-pi+(2*pi*rand); %AoA [-pi,pi]
    phi_2=(-pi+(2*pi*rand))/2; %AoA [-pi/2,pi/2]
    theta_1=-pi+(2*pi*rand); %AoD [-pi,pi]
    phi_1=(-pi+(2*pi*rand))/2; %AoD [-pi/2,pi/2]
    for mi=1:M
        for ni=1:n
            e = exp(1i*2*pi/lambda*((mi-1)*d_BS*sin(theta_1)*sin(phi_1)+(ni-1)*d_IRS*sin(theta_2)*sin(phi_2)));
            Beta = 1; % path loss factor
            G(mi,ni) = 1/sqrt(2).*sqrt(Beta).*e
            
        end
    end 
end

[Rena Berman]

(Answers Dev) Restored edit

Answer 1

[DGM]

Something like this

M=5;
n=3;
freq = 28e9;
lambda = physconst('LightSpeed')/freq;
d_BS=lambda/2; %Inter-antenna separation atthe BS.
d_IRS=lambda/2; %Inter-element separation at the IRS.

G = zeros(M,n,3); % allocate big enough for all copies
for l=1:3
    theta_2=-pi+(2*pi*rand); %AoA [-pi,pi]
    phi_2=(-pi+(2*pi*rand))/2; %AoA [-pi/2,pi/2]
    theta_1=-pi+(2*pi*rand); %AoD [-pi,pi]
    phi_1=(-pi+(2*pi*rand))/2; %AoD [-pi/2,pi/2]
    for mi=1:M
        for ni=1:n
            e = exp(1i*2*pi/lambda*((mi-1)*d_BS*sin(theta_1)*sin(phi_1)+(ni-1)*d_IRS*sin(theta_2)*sin(phi_2)));
            Beta = 1; % path loss factor
            G(mi,ni,l) = 1/sqrt(2).*sqrt(Beta).*e; % write just this page
        end
    end
end
G = mean(G,3) % average along dim3

Answer 2

[Sulaymon Eshkabilov]

in your exercise, the loop is not a good one. You can use vectorization instead. E.g.:

      theta_2=-pi+(2*pi*rand(3,1)); %AoA [-pi,pi]
      phi_2=(-pi+(2*pi*rand(3,1)))/2; %AoA [-pi/2,pi/2]
      theta_1=-pi+(2*pi*rand(3,1)); %AoD [-pi,pi]
      phi_1=(-pi+(2*pi*rand(3,1)))/2; %AoD [-pi/2,pi/2]
      M_theta2=mean(theta_2);
      M_phi2=mean(phi_2);
      M_theta1=mean(theta_1);
      M_phi1=mean(phi_1);