Although it was mentioned obliquely before, Glie now exists. It receives the ADS-B data from RTL-SDR in 1090ES band and produces an image. I can hit refresh in the browser and watch airplanes coming in to land at a nearby airport in real time.
This stuff is hardly groundbreaking. Many such programs exist, some are quite sophisticated in interfacing to various mapping, geography, schedule, and airframe information services, as well as in the UI. This one is mainly different because it's my toy.
Actually, the general aim of this project is also different, because unlike most stuff out there, it is not meant to be a surveilance tool, but to provide a traffic awareness readout. No persistent database of any kind is involved. No map either. Instead, I'm going to focus on onboard features, such as relative motion history (so one can easily identify targets on collision course).
But mostly, it's for fun and education. And already Glie is facing a few technical challenges:
- The best orientation for onboard display is "nose-up" (obviously). However, one can only derive a "track-up" orientation from a GPS. This is obviously wrong in case of Glie flying on a helicopter that can fly sideways. It is less obviously wrong in case of a crosswind, but can greate a significant distortion. To get the nose direction, I have to acquire a compass readout, which seems quite challenging. It's not like common airplanes have AHARS sockets under panels.
- I really want the graphics anti-aliased for better looks and finer precision, but I have no clue how to accomlish it. The standard TIS-B symbology essentially requires it, too, so I'm stuck with ad-hoc diamonds.
- The darn FAA split ADS-B in the U.S. into two bands: 1090ES and UAT. The 1090 is a solved problem in RTL-SDR space (although performance of receivers is not excellent and I'm thinking about building semi-hardware receivers in the future). However, UAT has 1 mbit/s data rate and RTL-SDR fails miserably on its face when trying to deal with it. I probably need something like Ettius GNU Radio receiver, which is expensive - $800 and up, last I checked. Unfortunately, it looks increasingly that all the interesting traffic is going to follow UAT route, come the 2020 A.D. (the year of ADS-B manate).
For now I'm going to take it easy and play with what I have, aiming for some kind of a portable system. Perhaps someone develops an open source UAT receiver on a practical platform in the meanwhile.