Could not use cluster range test because protocol wouldn't allow it. Would like to explore that as an option in the future.
It might be easier to have the local radio device as the moving object and have multiple remote devices set up in known locations. Would be an interesting inversion of the project that would allow one to locate themselves.
Now defunct:
Decided not to use quadcopter because switching over to a powered RC glider means less of a power issue, as the lift doesn't come directly from the rotors spinning.
May be you need to think about another solution. E.g. change in signal strength or direction. May be by shielding the antenna from different directions. Or by difference in time of flight between multiple antennas.