An open source Global Navigation Satellite Systems software-defined receiver | Hacker News Reader

An open source Global Navigation Satellite Systems software-defined receiver

gnss-sdr.org

104 pointsokket18 comments

I wonder if this would be usable for applications like high-altitude balloon tracking, where commercial GPS receivers cut out above 18km due to ITAR regulations.

Like unsignedint mentioned, the altitude requirement were removed a few years ago, and I think the speed limit was relaxed a bit. Even before then, GPS receivers were only required to stop providing positions when the speed and altitude exceeded the limits. Unfortunately many manufacturers implemented this as an "or" check instead of and. Some receivers (I think ublox does this) achieved compliance by only checking one parameter (always work below certain altitude regardless of speed, never supply position when above that altitude)

I've tried using it for rocket tracking (sdrgps.blogspot.com) without much success. I've been talking to the developers and provided RF for them to play with - maybe they will put a grad student on it at some point. I've had better luck with SoftGNSS which sadly is an offline processor. If I had more spare time I'd port it to realtime c++.

Looks like that limitation is removed.[0]

[0]: https://www.gps.gov/cgsic/meetings/2014/kim.pdf (page 7)

Limitation is to commercial sales of receivers. Private individuals building/operating their own receivers have no limits.

Does anyone know if any of these SDR solutions for hobbyists can get survey grade position solutions?

http://www.rtklib.com/ is an open source implementation of real time kinematic centimeter-level positioning. The link on that page to another SDR GPS project is dead, but https://gnss-sdr.org/docs/tutorials/understanding-data-types... claims support for RTCM and RINEX, which you might be able to feed into RTKLIB.

We've been trying to use it with a hackrf and bladerf and no dice at receiving actual GPS, even with an antenna exactly tuned to 1.57542GHz. Only when we transmit GPS signals ourselves so the signal is much stronger, are we able to obtain any satellites at all.

The software works, but you need a low noise amplifier or something else that you'd probably have to order from America or China (that would have taken too long for our school project).

Are you using a bias-tee to insert power on the center pin of the coax? Your antenna probably has a preamp built in, which needs to be powered otherwise it acts as an attenuator.

I was under the impression that gps antennas were powered...and special.

The receiver has to pull the signal from below the noise floor so needs a little help but, admittedly, my understanding on how it all works is limited.

A sibling comment mentioned using a Bias-T, and that's pretty much always needed unless you're using a COTS Rx (which would generally have one built-in)

It doesn't even have to be a fancy one. In a pinch, once, I use a random-valued inductor and the smallest-valued capacitor I had on-hand, and soldered them to a little PCB I made by slicing up copper-clad board with an x-acto knife. (0603 parts, if I recall.) Worked like a charm!

Don't most external GPS antennas patch include an LNA right on the antenna?

Of course, it's still a very faint signal.

Not really, actually a while ago I ordered a bunch of unamplified ceramic patch antennas because they were cheap. Managed to get fix with a 4x4 array and a uBlox NEO-6M chip; at 5x5 gain was good but the beam too narrow. In either case the signal was still under the noise floor.

What sat-constellation does this hear?

It appears that the eventual objective is to be able to use any GNSS. The page on telemetry decoder blocks[1] seems to indicate that GPS, GLONASS and Galileo signals are currently supported.

[1] https://gnss-sdr.org/docs/sp-blocks/telemetry-decoder/

At this time of writing, GNSS-SDR cannot be built natively on Microsoft Windows. As far as authors’ knowledge, nothing prevents it but their own ignorance on Microsoft’s building tools. Users have reported successful buildings of GNU Radio on Windows systems, so there should be not serious caveats building GNSS-SDR as well. However, GNSS-SDR is not blocked for Windows users. There are several virtualization tools that work well.

Great, let's blow off 95% of the desktop market and see how that goes for us.

Doesn't GNU Radio already support GNSS decoders? If I want a GPL-encumbered GNSS receiver solution, there are plenty of existing options.

The entitlement is strong in this one...

Yes, that's certainly an easy conclusion to draw, isn't it.

Reply on news.ycombinator.com