A Track-Before-Detect Algorithm Based on Extended H_∞ Particle Filter

This paper addresses the problem of the signal detection and tracking of targets with low signalto- noise ratio, a new track-before-detect method based on extended H_∞ particle filter is proposed. The proposed method overcomes the degeneracy problem of particle filter, because the system’s latest observations are considered by the importance density function of the extended H_∞ particle filter. Therefore, the acquired particles are more convergent to the true samples. The simulation results exhibit the robustness of the proposed method which shows much better performance than track-before-detect based on traditional particle filter.