Parallelized Monte Carlo Optimization with Efficient Pipelining on an FPGA for Onboard Ballistic Target Tracking

초록

This study addresses the challenge of accurately estimating high-speed ballistic targets in complex environments. To achieve precise target tracking in nonlinear and non-Gaussian noise conditions, Monte Carlo optimization (MCO) is employed. While increasing the number of samples in MCO enhances target estimation accuracy, it simultaneously escalates the computational burden, which in turn significantly extends processing time, making real-time application difficult. To overcome this limitation, we propose a parallelized MCO algorithm implemented on a field-programmable gate array (FPGA). The proposed approach achieves a reduction in computation time by up to 2.95× compared to conventional MCO methods. Furthermore, it delivers a 5.75× decrease in power consumption relative to GPU-based MCO implementations.

키워드