I'm currently working on a proof-of-concept to work out the estimated locations (longitude and latitude) of outdoor beacons with unknown locations to begin with.
The solution will use a mobile receiver (phone) to pick up signals within an area covered by large amount of beacons (>500-1000). By walking through the outdoor area, the mobile would gather RSSI signal strengths of surrounding beacons and also its GPS coordinates.
I'm trying to work out if I can tag the GPS locations of surrounding beacons by reversing the 2D trilateration / multi-lateration method as detailed here:
https://www.101computing.net/cell-phone-trilateration-algorithm
From the 2D-trilateration formula, instead of working out the receiver coordinate from the distance of 3 known points, I would want to use 6 known receiver's GPS locations to work out the 3 unknown points (x, y) pairs.
I understand there would be limitations with volatile signal strength / high noise / reflection issues. I would like to know if there are other efficient algorithms to achieve the goal of this proof-of-concept? Any related research work you can guide me to to further develop on?
I would also like to explore if I can obtain additional sensors' readings from the mobile (accelerometer / gyroscope / compass orientation), would that help to solve the problem efficiently?