Subject: Re: [GSoC 2026] Self-proposed idea: BeiDou B1C Signal Simulator OOT Module
Hello Sir,
The constellation simulator direction makes much more sense. If the output can actually give a software receiver enough to compute a real PVT fix, that's something genuinely useful for researchers, students, anyone.
So here's how I'm thinking about the expanded scope now
- multi-constellation , with Doppler shifts and time-of-flight delays that match what a receiver at a given location and time would actually see
- Generate proper navigation messages so a receiver can extract ephemeris and compute a position fix
- Validate everything using GNSS-SDR as a software receiver feed the generated I/Q straight in and check if it produces a valid PVT solution. No hardware needed.
One thing I'm unsure about: is it better to do BeiDou B1C really well and completely, or to cover B1C + GPS L1C + Galileo E1 at a slightly shallower level? Since all three share 1575.42 MHz I can see the appeal of multi-constellation.Would love your take on this.
I'll start working on a proper week-by-week draft and share it here soon for feedback.
Best,
Devanshi.
On Thu, Mar 19, 2026 at 12:50 AM Daniel Estévez <daniel@destevez.net> wrote:
Hi Devanshi,
Thanks for sharing this idea. Here is some quick feedback.
You don't say explicitly, but my understanding from what you wrote is
that your idea is to generate a single B1C signal with nominal chip rate
and no Doppler. This does not seem enough work even for a small (90
hour) project, specially if you are already strongly familiar with GNSS
signal generation and the B1C signal. Such kind of simulator is of
fairly limited use, since it only allows to test that a receiver is
capable of acquiring and tracking a single signal in these conditions.
It is more common to have constellation simulators, which are capable of
simulating signals from multiple satellites whose times of flight and
Dopplers are consistent with the signals that a receiver in a given
location would see. A constellation simulator, if well planned out,
might be reasonable for a GSoC project (probably more on the larger side
of the project size spectrum). Constellation simulators can be used to
obtain a PVT solution with a receiver, and test it in reasonably
realistic conditions.
You should detail in your proposal how you plan to test this. It is fine
that the tool itself generates baseband IQ and does not require any
hardware, but you will need to test somehow that the signals you are
generating are correct. Ideally you would use a hardware receiver and an
SDR to transmit your baseband IQ, but this requires you to already have
this hardware. If you don't have the required hardware, using a software
receiver is a good alternative, but you will need to identify such a
software receiver that allows you to check all you generate.
Another way in which you could increase the scope of your proposal and
make it more attractive is by including open service signals from other
constellations (GPS, Galileo, etc.), all of which have ICDs which are
publicly accessible.
Best,
Daniel.
On 18/03/2026 19:22, Devanshi B wrote:
> Hi GNU Radio community,
>
> I am Devanshi , a GSoC 2026 applicant interested in proposing a self-
> directed project.
>
> I have strong familiarity with GNSS signal generation concepts and the
> BeiDou B1C ICD specification, with hands-on experience .
>
> Proposed idea: A GNU Radio OOT module tentatively gr-beidou
> implementing a BeiDou B1C software signal simulator. The module would cover:
>
> - B1C PRN code generation (data and pilot components, per ICD)
> - BOC modulation
> - Basic CNAV-1 navigation message framing
> - Baseband I/Q output usable for receiver testing and simulation
>
> The implementation would be built entirely from the public BeiDou B1C
> ICD with no hardware dependency making it useful for receiver testing,
> education, and research. I plan to model the module structure after gr-
> satellites, which sets a great precedent for clean, well-documented
> signal-level OOT modules.
>
> This fills a genuine gap no open-source GNU Radio B1C signal generator
> currently exists.
>
> if any mentor would be interested in supervising this, or if there is a
> better framing that fits current GNU Radio priorities.
>
> Cyberspectrum is the best spectrum.
