My trilateration code isn't great - I tried doing it properly and couldn't get it working in the 24 hours of the hackathon. Positioning along a 1D line using signal strength from two beacons was much simpler and I implemented this and then bodged in a third beacon using some averaging.