Decorrelation Analysis of Differential LEO Range Corrections

This paper evaluates the residual error of a differential positioning system utilizing Low Earth Orbit (LEO) satellites as navigation beacons. For existing global navigation satellite systems (GNSSs), differential positioning has proven to be an effective error mitigation strategy (Misra and Enge, 2012). This effectiveness is due to the correlation of errors between receivers that are separated by time and distance. Errors that exhibit this type of correlation are called “common mode”. A differential positioning network can greatly reduce common mode ranging errors by transmitting error estimates from a monitoring station to a user receiver. This paper aims to quantify common mode errors between users of a theoretical differential LEO (DLEO) positioning network. Simulation results show that DLEO users will experience more distance-based decorrelation errors compared to traditional DGPS users, therefore demanding shorter baselines to maintain performance. Latency-induced errors are also shown to be larger for DLEO systems, meaning that the broadcast corrections from such a system must be refreshed more frequently than DGPS corrections. For a variety of different scenarios, these errors are evaluated with Iridium as a stand-in for a future LEO GNSS. A simulation framework is presented which allows for the prediction of decorrelation errors given models of error sources (atmospheric, ephemeris, etc.). The models indicate that DLEO error could grow by as much as 2 cm km and 1.7 mm s2.