Weak Signal Detection Algorithm Based on Particle Filtering

Abstract

In the traditional particle filter algorithm, there exist problems such as poor robustness against noise and clutter, and difficulty in balancing the detection probability and false alarm probability. In the problem of detecting weak signals, due to the noise issue, it is more prone to missed detections and false detections. Meanwhile, if there is a large amount of background clutter around the weak signal, it will lead to a decrease in tracking accuracy. To address the problems arising in the signal processing under continuous clutter scenarios and improve the weak signal detection performance of the particle filter, this paper proposes an improved particle filter algorithm. Firstly, by modifying the likelihood ratio function, the influence of clutter on the detection probability is restricted. Secondly, the Bernoulli filter is used to set the detection probability of the signal, which significantly reduces the missed detection probability. Additionally, different initial particle distributions are adjusted with or without prior probability, further enhancing the performance of the filter. Simulation results show that the detection probability of weak signals in the clutter background using the improved particle filter algorithm has been significantly enhanced compared to the processing before the improvement.

Keywords

• |particle filter|likelihood ratio|weak signal|clutter suppression|target detection