Hi Guys,
I am interested in real-time channel estimation of a GSM signal. Since
GSMK is a non-coherent modulation, performing channel estimation is
not easy, at least to me. I have found a paper in which they correct
for the 90 degree phase shift in every symbol (multiply by i^k, k’th
symbol) before they correlate it with a known sequence (mid-amble of
the SCH-packet). I hope you guys understand what I’m talking about.
The absolute value of the main peak of my channel impulse response is
what I expect it to be, however the phase information is totally off.
Could anyone point me to some reference, or explain briefly, how I
could perform a good channel estimation for GSM? Or any idea where else
I should post this question?
Thanks,
Teun van Berkel
Philips Research
Nat.Lab., WY 5.015a
Eindhoven, The Netherlands