A Fast Implementation of the Extended Kalman Filter on FPGA for UAV Applications

Extended Kalman Filter algorithm is widely used in UAV applications. Fast operation of the algorithm is important for precisely calculating the attitude of the UAV. In this paper, we present a very fast FPGA implementation of the Extended Kalman Filter algorithm. A dedicated processor is designed using verilog-HDL and logical blocks are used to accelerate the algorithm. Maximum operating frequency of the FPGA is 130MHz and the propagation of the state, which is the critical path of the algorithm, is completed at 1050 clock cycles (8.074 mu s).