Raspberry Pi 4: How I built a software-defined radio system to track passing planes
One of the things about being in lockdown here in London is that I've had time to work on various projects that have been sat on the back burner. So in the last three months or so, I've set up a much improved microphone and camera system for podcasting and streaming, built a new time-lapse rig for the sky that passes by, practiced my wildlife photography on the local suburban critters, and built and set up a software-defined radio system to track passing aircraft.
Aircraft over London on a July afternoon, as tracked by my ADS-B receiver
Developer
That last one is perhaps the more interesting for ZDNet's readers, as it's how you can start to feed aircraft location data into a growing network of tracking stations that uses the GPS locations most aircraft transmit to build a global map of air traffic. This approach, ADS-B, is intended to make it easier for aircraft to locate each other and for air traffic control to keep track of them. The fact that the data is there for anyone to capture and use is just a side-effect.
SEE: Hiring Kit: Python developer (TechRepublic Premium)
Picking up the ADS-B signals is easy enough: they're broadcast on a known frequency, 1090MHz, and have to deliver data in a set format. If you've got the right receiver, that binary data can be quickly translated into coordinates that can be plotted on a map, using callsigns to look up aircraft type and owner. You've probably seen services like Flightradar24 and Flightaware in the news, offering ADS-B information to anyone with a web browser.
Those services get their data from a network of volunteers, with their own ADS-B receivers. Usually based around open source software, their base stations are easy enough to build and deploy, using software-defined radios to receive ADS-B signals. Most are based around a popular chipset that was originally developed to add USB digital TV receivers to PCs, with a set of APIs that made it easy to adapt for other uses.
Setting up an ADS-B base station is now as easy as getting a Raspberry Pi, plugging in a receiver, and downloading and configuring a ready-to-run system image. My system is based around a 2GB Raspberry Pi 4 in a heatsink case, with a Flightaware ADS-B radio stick, a 1090MHz desktop antenna, and a copy of Flightaware's Piaware software distribution.
Based on Raspbian, Piaware is a headless Linux distribution with all the software you need to work with a radio stick, including the Dump1090 ADS-B Mode S decoder. Dump1090 comes with a built-in web app that maps currently tracked aircraft and can also be used to send data to other apps and servers, helping feed data into larger public ADS-B networks.
You can get kits that include all the parts you need to set up and run an ADS-B receiver from hobbyist electronics stores, for around £100 (including the Raspberry Pi). The basic antenna that comes with these kits is good for a start, especially if you have a way of getting it outside. You're likely to want to invest in a larger antenna for better, longer range, reception so make sure your budget is ready for an extra £30 to £60, depending on the length of the antenna.
Configuring a receiver is easy enough. You'll need your GPS location to 4 decimal places, as well as the approximate height of your antenna. You'll also need to configure network connections, either WiFi or wired ethernet. Most of this can be done by editing the configuration files on the Piaware microSD card, letting you treat the whole system as an appliance. With the configuration data in place, you can boot your server and leave it running. I've additionally configured mine to use SSH, so I can log into it from my desktop PC to update Raspbian and the Piaware stack as needed.
ADS-B isn't the only way to track aircraft. Not all aircraft, especially the military, transmit their flight data. However most that don't do at least have transponders that broadcast an ID and an altitude on known frequencies. If you have enough receivers in known positions, and enough compute power, you can use basic physics to identify where an aircraft is. This approach is known as multilateration, or, as it's more commonly known, MLAT.
If you've got a receiver set up, it's easy enough to join one or more networks to share the signals you receive and help build a live map of aircraft all around the world, providing both ADS-B data and the raw transponder information that can generate MLAT data. In return for your data you'll get access to the networks' subscription services which give better and more detailed outputs than the free commercial services – and of course, without any advertising.
SEE: Raspberry Pi: Now computer-vision apps get boost with Khronos OpenVX API
I've initially joined two networks, one of the larger public services with nearly 20,000 base stations around the world, and one a private one with a focus on identifying unusual and military aircraft across the UK, and living near the Thames I'm close to a major low-altitude helicopter throughway across London and the city's heliport, where most of the traffic is only trackable via MLAT, as well as under the westbound route into London Heathrow.
My antenna is currently taped to the window ledge of the back bedroom, with the Raspberry Pi and receiver on the wall next to the window, using WiFi to connect to the rest of my network and on out to the wider world of ADS-B data. It's seen aircraft as far away as the coast in Norfolk, and tracked military aircraft that are using closed airports for training, as well as the regular RAF and Army helicopter traffic up and down the river, and even the joint flypast by the French and British display teams. It's been an interesting project, and one that looks to keep paying off in the months and years to come, especially when I get that larger antenna fitted.