LLR for AWGN

Problem

Return the log-likelihood ratio of a BPSK observation y over an AWGN channel with noise variance noise_var. With equally likely symbols transmitted as +1 and -1,

LLR = 2 * y / noise_var

Hint

The LLR is linear in the observation and inversely proportional to the noise power.

Test cases

case 1

in: [0.5,1]

expected: 1

case 2

in: [1,2]

expected: 1

case 3

in: [-0.25,0.5]

expected: -1

case 4

in: [2,1]

expected: 4

Starter codePython

def llr_awgn(y, noise_var):
    # Your code here
    pass

4 test cases · run your code against them in the editor