Improved FPGA implementation of particle filter for radar tracking applications

Abstract

Particle filter has been considered as great advancement for radar tracking in nonlinear and non-Gaussian scenarios. But the huge computation and complexity limited its application in real-time systems. In this paper, we improve the architectures of sampling importance resampling particle filter (SIRF) in FPGA on the Xilinx VIRTEX-5 platform and then apply it to the bearings-only tracking system. Not only the whole design is simulated by the Modelsim, but also is realized in hardware, where the data are transferred using UART. The improved architectures could provide a fine real-time characteristic in radar tracking system.