Re: CPM timing recovery

GNU Radio
1

David,

one way to estimate the rate is to raise the CPM signal to an
appropriate power in order to generate spectral lines that can be easily
tracked.
The precise power is a function of the modulation index of your CPM
modulation. Eg, if you are using full-response CPFSK with h=N/D
(where N,D are relative primes) then you can raise your signal to the
power D in order to generate spectral lines. I don’t remember off the
top off my head the exact frequency of these spectral lines but you can
easily find out the relationship by simple experimentation.

BTW, are you using the cpm.py hierarchical block that is
on the trunk? If yes, I attach a simple python code that
demonstrates the spectral line generation for a 4-CPFSK
with h=1/2.

Achilleas

2

BTW, are you using the cpm.py hierarchical block that is

on the trunk? If yes, I attach a simple python code that
demonstrates the spectral line generation for a 4-CPFSK
with h=1/2.

I definitely see the spectral lines for your case, which I believe is
considered an MSK modulation. Unfortunately, I cannot assume that the
signal will be MSK. When I change the index to another often used CPFSK
index of 1/4, I noticed that the spectral lines go away…

any thoughts?

david

   #self.connect (self.src,self.b2B,self.mod_old,self.fft_old)
3

David,

as I explained in my earlier email,
the power you have to raise your signal is not always 2.
If h = 1/4 you need to raise your signal to the power 4.
In general ig h=N/D, you raise it to D.

Achilleas

David S. wrote:

considered an MSK modulation. Unfortunately, I cannot assume that the
#self.connect (self.b2B,self.mod_new ,self.fft_new)
app = stdgui.stdapp (cpm_graph, “test CPM graph”)
app.MainLoop ()

–

Achilleas A.
Associate Professor
EECS Department Voice : (734)615-4024
UNIVERSITY OF MICHIGAN Fax : (734)763-8041
Ann Arbor, MI 48109-2122 E-mail: [email protected]
URL: http://www-personal.engin.umich.edu/~anastas/

4

On 6/7/07, Achilleas A. [email protected] wrote:

David,

as I explained in my earlier email,
the power you have to raise your signal is not always 2.
If h = 1/4 you need to raise your signal to the power 4.
In general ig h=N/D, you raise it to D.

dang it…

good call…thanks Achilleas that worked great.

david