Multitarget Bernoulli Track-Before-Detect for OTHR

This work considers the application of track-before-detect (TBD) to detect and track dim airborne targets in over-the-horizon radar (OTHR). The problem is challenging with multiple paths being present due to reflections off the ionosphere, which leads to multiple reflections for each target. This is coupled with aliasing in Doppler when the target is fast-moving. This paper develops a sequential Bayesian TBD for OTHR. The key component of this approach is the Bernoulli TBD filter, as the optimal Bayesian TBD filter for joint detection and tracking of a single target. A multi-target particle-based extension is developed by incorporating the logic that treats all nearby targets as contributors to the background interference. The algorithm performance is evaluated using a detailed OTHR data simulator.