Using historical data to correct GPS data in a network of moving things

A system comprising:

• model generation circuitry operable to process a respective series of positioning data samples received from each of a plurality of satellite based positioning system receivers of respective vehicles mobile within a service area of a wireless network comprising a plurality of vehicles to generate one or more statistical models of historical positions of the plurality of vehicles, wherein the series of positioning data samples from each positioning system receiver is received via a respective mobile access point of a plurality of mobile access points of the network of vehicles, and wherein each of the generated one or more statistical models describes a path of the plurality of vehicles based on an aggregate of the series of positioning data samples;
• and
• positioning circuitry operable to:
• receive a current positioning data sample from a particular one of the plurality of positioning system receivers;
• and
• compensate for error in the current positioning data sample from the particular one of the plurality of positioning system receivers using the one or more statistical models to determine a most-probable location of the particular one of the plurality of positioning system receivers.