My carrier freq is 1.575GHz.PRN code is generated by a 168bit length
LFSR.So the code tends to be a random sequence.Clock to LFSR is 1
MHz.So I wanted to know that how the FM spectrum look like.here the
modulating signal will be like sqaure wave with randomly varying duty
cycle.Frequency spectrum will be a discrete one.I wanted to know the
spacing between each spectum line.The spacing must be equal to
modulating freq.here is it the 1MHZ clock period or the period of the
PRN sequence which is very large.So will the spectrum have definite
shape at an instant or keep on changing.
I think this one of the way of generating white noise with digital
method.
regards,
raghavendra

Posted by Allan Herriman●July 17, 20032003-07-17

On 16 Jul 2003 21:11:23 -0700, raghurash@rediffmail.com (Raghavendra)
wrote:

>My carrier freq is 1.575GHz.PRN code is generated by a 168bit length
>LFSR.So the code tends to be a random sequence.Clock to LFSR is 1
>MHz.
>So I wanted to know that how the FM spectrum look like.
>here the
>modulating signal will be like sqaure wave with randomly varying duty
>cycle.

For a small modulation index (NBFM) each side of the spectrum at the
output of the VCO will look like the spectrum of the LFSR output with
a -20dB/decade slope applied.
Things get a little harder to analyse for WBFM.

>Frequency spectrum will be a discrete one.

Will it? The period of the modulating signal is much, much greater
than the lifetime of the universe. It is not possible to resolve the
spectrum into discrete lines. For all intents and purposes, the
spectrum is continuous.

>I wanted to know the
>spacing between each spectum line.

2.67e-045 Hz, which is essentially 0 Hz.

>The spacing must be equal to
>modulating freq.here is it the 1MHZ clock period or the period of the
>PRN sequence which is very large.

It's the period of the PRN sequence.
The 1MHz clock merely positions the nulls in the sin(x)/x shape due to
the zero order hold effect from the flip flop at the output of the
LFSR.

>So will the spectrum have definite
>shape at an instant or keep on changing.

A spectrum never has a definite shape at an instant. Think about the
definition of the FT - the integral has limits from -infinity to
+infinity.
If by "spectrum" you mean "what I see on a spectrum analyser," it will
have a definite, unchanging shape.

>I think this one of the way of generating white noise with digital
>method.

100% of our work is cutting-edge DSP
We are a small, high tech, sonar-processing company that develops state-of-the-art sonar, signal processing software systems used in naval aircraft.
Tasks:
* 60-75% signal processing algorithm development
* 25% real time code development in C++ Full relocation will be provided for the right candidate (e.g. relocation package up to $20,000).