Eb=sum (x ^ 2)/1000
We can obtain N0=Eb/10 ^ (3/10)
Then I add white gaussian noise:
Y=x + SQRT (N0) * randn (1100)
The simulation results and theoretical gap
Later found out that the unilateral of N0 just white gaussian noise power spectral density, not the noise of power, confused here,,,
Excuse me each a great god, and how the white noise added to satisfy the Eb/N0=3 db?
CodePudding user response:
Why don't you directly in awgn function