Recursive bayesian filtering for multitarget track-before-detect in passive radars

This paper presents a Bayesian algorithm for joint detection and tracking in a multitarget setting. Raw measurements are processed using the track-before-detect (TBD) framework. We first establish a Bayesian recursion, which propagates a probability of target existence along with a target state probability density per delay/Doppler bin. In order to handle the nonlinearity of the observation model obtained for orthogonal frequency division multiplexing (OFDM)-based passive radar, a suitable Gaussian mixture implementation is proposed.