Re: [Discuss-gnuradio] GPS/DSSS question

[Krzysztof Kamieniecki]

The FFT is an interesting idea, I will investigate.

I read a series of articles about Costas Loops by Eric Hagemann.
The collection also includes articles by different authors about different DSP 
techniques.
http://www.chipcenter.com/dsp/colarch.html

I plan to use his idea of using arctan instead of the I/Q multiplier. At the point where the loop 
would be in the GPS receiver, the sample rate is down to 12Ks/s so it should be no problem.


Deary, Bryce D. (BDEARY) wrote:

> A second option for finding the signal is to use a FFT to do the correlation
> instead of doing an actual correlation.  This often lets you get very close
> in both frequency and time (code phase) before turning on the FLL.  Also
> remember that the PLL used in GPS is typically a costas loop instead of a
> full PLL because once you stop the phase rotation with the FLL and drive all
> the power into I(in-phase) by forcing Q(quadrature) to zero in your loop,
> the in-phase will continually flip sign due to the 20 ms data bits. Normal
> PLLs hate this.
>
>
> -----Original Message-----
> From: Brent Ledvina [mailto:address@hidden 
> Sent: Saturday, August 16, 2003 6:39 PM
> To: Krzysztof Kamieniecki
> Cc: GNU-RADIO DISCUSS MAILING LIST
> Subject: Re: [Discuss-gnuradio] GPS/DSSS question
>
>
> Your algorithm should be modified in the following way:
>
> 1. Detect the code phase offset and carrier doppler shift using your 
> acquisition algorithm (by finding the peak signal power at a particular 
> Doppler offset).
> 2. Track the code offset using a delay-locked loop (DLL)
> 3.  Track the carrier frequency using a FLL or the frequency and phase 
> using a PLL.
>
> An good text on GPS receiver design using software receivers is 
> Fundamentals of Global Positioning Receivers: A Software Approach by 
> James Tsui, Wiley and Sons, 2000.
>
> Regards,
> Brent
>
>
> On Saturday, August 16, 2003, at 08:14 PM, Krzysztof Kamieniecki wrote:
>
>
> > I've been working on my GPS software, and I have a question as to the
> > validity of one of my algorithms.
> >
> > Simplified Background:
> >
> >  Each GPS satellite broadcasts a DSSS(Direct Sequence Spread
> > Spectrum) signal. The chipping rate is 1023M/s and the PRN code is 
> > 1023 chips long.
> >
> >  To lock onto a satellite signal you have to:
> >    1. Find (and track) the correct carrier frequency (because of
> > Doppler / Relativity effects, Receiver effects, etc...)
> >    2. Find (and track) the correct chip offset
> >    3. Find (and track) the correct chipping rate (same problems at 
> > carrier frequency, but this may not be a noticeable effect)
> >
> >
> > My Algorithm:
> > 1. Look for signal peak by trying 1023 different chipping offsets at
> > frequencies separated by 200Hz
> > 2. When a peak signal is found, measure the signal coming out of DSSS 
> > integrators.
> > 3. Adjust carrier NCO based on measured signal (frequency and phase 
> > offset)
> > 4. Activate carrier PLL
> >
> > So the question is, Is step 3 valid thing to do, or do I need a more
> > complicated FLL/PLL?
> >
> > --
> > Krzysztof Kamieniecki
> > mailto:address@hidden
> >
> >
> >
> > _______________________________________________
> > Discuss-gnuradio mailing list
> > address@hidden 
> > http://mail.gnu.org/mailman/listinfo/discuss-gnuradio
>
>
>
>
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/discuss-gnuradio
>
>
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/discuss-gnuradio

--
Krzysztof Kamieniecki
mailto:address@hidden