OTHR multipath tracking with uncertain coordinate registration

Multipath probabilistic data association (MPDA) exploits discrete multipath propagation to improve tracking performance for systems such as over-the-horizon radar (OTHR). The original formulation assumed a known ionospheric environment for mapping radar measurements to ground coordinates. We consider an extension of the technique that allows for ionospheric uncertainties affecting the transformation between slant and ground coordinates in addition to the usual sensor noise. A new technique, MPDA for uncertain coordinate registration (MPCR), is described and tested on simulated and real OTHR data. The study assumes a two-layer spherical mirror ionosphere with random fluctuations of the ionospheric heights around known mean values. An extensive Monte Carlo analysis of track errors and track loss is carried out for MPCR under two scenarios and the results are compared with a baseline probabilistic data association (PDA) approach and with MPDA. The results indicate the high reliability of the MPCR approach.