Hybrid implementation of Extended Kalman Filter on an FPGA

Radar Tracker is a functional block of almost every radar system. It is used to smooth the radar measurements while estimating the closest path of the target. A typical implementation can be done using an Extended Kalman Filter (EKF). EKF is a recursive complex algorithm. Due to irregularity of the algorithm it is typically implemented on microprocessors using software. However, due to high computational cost, the performance of such system is limited to given processor capabilities. It´s possible to partition the algorithm in such a way that part of the algorithm would be offloaded to a co-processor. Such architecture would enable more capable systems. In particular, it allows increased Radar Tracker performance. The following paper demonstrates such a concept when EKF is implemented on a Field Programmable Gate Array (FPGA).