Target localization and track association for over-the-horizon radar with a statistical ionospheric model

Over-the-horizon radar exploits the refractive and multipath nature of high frequency propagation through the ionosphere to achieve wide-area surveillance. For multiple radar dwells, the coordinate registration (CR) process converts the sequences of group delays and azimuths (i.e. slant tracks) from a set of multipath target returns, to a sequence of estimated target locations (i.e. ground tracks). This is performed by associating the slant tracks with ray modes determined using a computational electromagnetic propagation model. In multiple target scenarios, the CR process of assigning the ray modes to slant tracks is complicated by the fact that the slant-track-to-target associations are unknown. Not surprisingly, errors in the estimates of down-range ionospheric parameters can seriously degrade both the accuracy of the target ground track estimates and the probability of correctly associating the slant tracks to targets. This work proposes a method for the problem of joint multiple target ground track estimation and slant track association, or mode linking, with uncertain ionospheric propagation conditions where the slant-track-to-target assignments and slant tracks' ray mode paths are unknown. Modeling down-range ionospheric parameters as random variables with known statistics facilitates maximum likelihood (ML) coordinate registration which is more robust to errors in the measured ionospheric conditions. Maximum a posteriori (MAP) mode linking exploits the statistical dependence between slant tracks on different ray mode paths to provide accurate slant track associations to targets and to propagation paths and thus accurate ground track estimates. The approach uses a Markov model for the temporal correlation between mode linking hypotheses at different revisits to obtain smooth ground tracks. Monte Carlo simulation results indicate that MAP mode linking can potentially provide nearly a 3:1 improvement in ground track accuracy over conventional mode linking with much higher probabilities of correct track associations and ray mode assignments. Results with real OTH radar slant track data for multiple slant tracks from multiple targets and validated against FAA ground truth, confirm the improvement offered by the MAP mode linker.