Re: Galileo Frequency Shifting and Filtering

[David Dima]

All right guys, I got it, it works now. The central frequency should have been 1911.795 MHz instead of 1189.5MHz. The final configuration results as follows:

Thank you very much to everybody!

Best regards,

David

El mar, 28 feb 2023 a las 18:17, Marcus Müller (<mmueller@gnuradio.org>) escribió:

Hi David,

please stay in the habit of keeping the mailing list in CC!

Can you please write this email to the mailing list?

Thanks,
Marcus

On 28.02.23 17:55, David Dima wrote:
> Hi Marcus,
> According to Galileo's signal plan
> (https://gssc.esa.int/navipedia/index.php/Galileo_Signal_Plan
> <https://gssc.esa.int/navipedia/index.php/Galileo_Signal_Plan>), E5b bandwidth is  25MHz,
> but to save memory and resources, I sample at 20MHz because I demonstrated by testing that
> is enough to process the signal without losing information.
>
> Regarding the Xlating filters: if I tune the USRP in the middle of the signals (1164MHz +
> (1215-1164)/2)MHz = 1189.5MHz) , then I have to shift the spectrum +- 15.345 (According to
> Daniel: "The band plan you posted is slightly wrong. The distance between the
> AltBOC sidebands and the centre frequency is 15 * 1.023 MHz.") to be able to record at
> 25MHz each band (E5a and E5b). I am trying to do the following:
> imagen.png
> Kind regards,
> David
>
> El mar, 28 feb 2023 a las 17:25, Marcus Müller (<mmueller@gnuradio.org
> <mailto:mmueller@gnuradio.org>>) escribió:
>
>     Hi David,
>
>     On 28.02.23 17:17, David Dima wrote:
>>     Dear Marcus, all
>>     Thank you for your response. You are right, I will try to better explain the problem:
>>
>>     Well, when I record the signals separately (for example E5b) as shown in the below
>>     picture, and after that I try to process the file containing the IQ samples by means
>>     of GNSS-SDR free-tool, I can parse the messages to generate ephemeris, almanac or
>>     Rinex files. Even if E5b bandwidth is 25MHz, it works sampling at 20 MHz.
>>     imagen.png
>
>
>     You're losing 1/5 of the relevant spectrum. That doesn't mean it *can't* work, it's
>     just unnecessarily bad – honestly, **why** are you using a sampling rate that's not
>     fit to represent the *full* signal? Can you please state *reasons*?
>
>
>>
>>     The issue comes when I try to record both signals at the same time. Taking into
>>     account the suggestions Jim, Daniel and you have given me in the forum, I have made
>>     a couple of changes in the program resulting in:
>>     imagen.png
>>     Now, at least I think I am closer to getting what I need but I am not sure if there
>>     are more mistakes in the flowgraph. So, I have tried to record the signals according
>>     to the above flowgraph but after that I still cannot read any signal from the
>>     satellites whereas if I record them separately I can.
>
>     This makes no sense at all. Where do your "Center Frequencies" in the Freq.-Xlating
>     filters come from?
>
>     Best regards,
>     Marcus
>