Dear All,

Thank you for your participation in my discussion in advance

I am working on the phase correction step in a receiver. The following algorithm has been already implemented:

•  correct vectors (samples) by using the multiplication by exp (-1i * phI)
•  compute an amplitude and rotate a vector:

A = sqrt(real^2+imag^2);

theta = arctan(imag/real);

• now theta will be the next angle for rotation

Actually, it is the standard algorithm to correct a phase, I think.

The simulation works but it gives me a big error (for me)\" dPhi\' (difference between phases adjacent samples) under low snr

Is there an algorithm to minimize the phase difference of two vectors??

Are you simply averaging the theta values, which have roll-over, as that roll over can cause a significant non-linearity. 

You could average the complex phase (i.e. the original complex signal divided by its amplitude) and then do the arctan step. 

There are a range of choices depending on how you handle the varying amplitude. 

Definitely worth some experimentation to see what is going on for the complex phase addition scenarios.

i have tried...''the original complex signal divided by its amplitude''

it doesnt give me improvement

Ali,

are you phase aligning a sinewave of a local DDS with that of a received complex sinewave? Or are you phase locking to a modulated signal with out an underlying carrier?

I want to make sure we are both looking at same problem.

fred h