Radar target detect using particle filter

A radar target track-before-detect (TBD) algorithm using particle filter (PF) is presented in this paper. System dynamic model and measurement model are established based on a sequence of radar range-Doppler measurements using the new algorithm. Furthermore, a linear extended target model is proposed, which is more capable of describing a maneuvering target than the conventional point target model. The likelihood ratio function of the new model is also derived in this paper. Due to the accumulation of the PF-TBD over time and the effectiveness of the proposed target model, an improved probability of detection for dim target is obtained. The experimental simulations demonstrate that the proposed method is capable of detecting and tracking a target with SNR of 1 dB robustly.