channel coefficients constant

[Ennis]

Is the value of channel coefficients same?

Please help me. Very urgent

 

[Pulkit Varshney]

of which channel...please elaborate your question

 

[Ennis]

of which channel...please elaborate your question

Does the channel coefficients of a real-time wireless channel vary? Is it possible for the channel coefficients in a channel emulator to change or will it be constant always?
Thank you so much in advance. Thank you...

 

[Pulkit Varshney]

yes they change and how fast they change that depends upon coherence time of channel.

 

[Ennis]

yes they change and how fast they change that depends upon coherence time of channel.

Thank you so much... What is CFI in channel emulation?

 

[ehbas]

Hi I have a related question

suppose a 3-path Rayleigh Frequency Selective fading channel which coherence time is T_c. does it mean that within a T_c time duration, channel coefficients are constant?
and is there any model for changing channel coefficients with time?

 

[Pulkit Varshney]

yes, coherence time implies the same. i.e. channel coefficients are constant during coherence time
i think jakes model is there for that

 

[ehbas]

one other question about channel coefficients:

suppose a Rayleigh fading channel with just one path between transmitter and receiver (a frequency flat fading). Is channel coefficient for this single path Complex or Real.
off course I know that if the channel coefficient is complex, then it's magnitude is of Rayleigh distribution and it's phase is of uniform distribution.

I have simulated BPSK in frequency flat Rayleigh fading with both complex and real coefficient. I assumed that we don't know the channel coefficient at the receiver (no CSI).
the results showing that in frequency flat Rayleigh fading, the channel coefficient should be real random variable of Rayleigh distribution.

here is my code: (to change the coefficient to real uncomment the 2 commented lines)

clear;
clc;
avg_gama_b = [0         5     10    15     20     25     30    ];
limit                =  [1e2  1e3  1e3   1e4    1e4 1e4   1e5  ];
power_dist = 1;
power_dist_pmt = sqrt ( power_dist ./2 );
omega = sum(power_dist);
omega_db = 10*log10(omega);
eb_by_n0 = avg_gama_b - omega_db;
sim_point = length(eb_by_n0);
ser = zeros(1, sim_point);                     %vector of BER's

for j = 1:sim_point
    sym_error = 0;
    for o=1:limit(j)
        b = rand();
        if b > 0.5,            b = 1;        else            b = -1;        end
        s = b;
        signal_power = 10*log10(sum( s .^ 2 ));
        noise_power = power(10, (signal_power - eb_by_n0(j))/10);
        sigma_n = sqrt(noise_power);
        n = (sigma_n/sqrt(2)) * (randn() + 1i * randn());
        alpha = power_dist_pmt .* (randn() + 1i * randn());
%         alpha = abs(alpha);
%         n = sigma_n * randn();
        r = n + alpha*s;
        if real(r) > 0,            b_hat = 1;        else            b_hat = -1;        end
        if b ~= b_hat
            sym_error = sym_error +1;
        end
    end
    ser(j) = sym_error / limit(j);
end
ser_matlab = berfading(eb_by_n0, 'psk', 2, 1);
figure, semilogy( avg_gama_b, ser,'-sb', eb_by_n0, ser_matlab,'-or'), xlabel('average SNR'), ylabel('Error Probability'),
title(['BPSK in  Rayleigh'] );
grid on;
hleg1 = legend('my BPSK BER', 'MATLAB BPSK BER');
set(hleg1,'Location','SouthWest');
set(hleg1,'Interpreter','none');