Kindly find in the uploaded image:

- r(t): the equation of the received signal (shape 800x2) consists of:

p: probe transmission power (arbitrary constant)

h(t) : channel gain coefficient that we want to get (shape 800*2)

x(t): gmsk modulated input signal (shape 800*2)

n(t): AWGN noise (shape 800*2)

N.B. 800x2: 800 is the number of symbols (100) times number of samples (8), 2 is the number of channels (real and imaginary). Therefore the input signal before modulation is 100x1

- f(h): the PDF that we should use to get h(t)

Following your answer, should I write the pdf with a symbol variable h?

I mean the pdf is beta*exp(-beta*h)*(1+(1/alpha)*(0.5*(beta^2)*(h^2)-2*beta*h+1)). Beta and Alpha have a calculated values while using your proposed function but h is not what should I do?