Systems and methods for learning wireless transceiver locations and updating a spatially-dependent path-loss model

An indoor positioning system determines the location of a mobile device by comparing measured signal strengths to a database determined during offline calibration. Automated creation, maintenance, and repair of the database are facilitated by accurately characterizing indoor radio signal propagation. Systems and methods for generating spatially-dependent path-loss models are disclosed. In one variation, a computer-implemented method of generating a spatially-dependent path-loss model involves dividing a coordinate-plane representing an indoor environment into non-overlapping tiles; obtaining transmit and received signal strengths of radio signals generated by and measured at a number of wireless access points positioned throughout the indoor environment; calculating vectors which represent the traversal distances of the tiles by each of the radio signals; and solving a system of path-loss equations relating the transmit signal strengths, the received signal strengths, and the distance vectors to determine values for tile-specific path-loss coefficients for each of the tiles within the indoor environment.