Extended Kalman Filter based loosely coupled INS/GPS integration scheme using FPGA and DSP

Abstract

GPS provides an accurate position and velocity information but with a slower update rate, whereas INS provides position, velocity and attitude information with a higher update rate. However, INS alone cannot provide proper solutions due to the increasing error characteristics of the accelerometer and gyros of the inertial sensor. To get an accurate navigation solution, it is necessary to integrate INS with GPS. This paper discusses integration of INS/GPS systems using Extended Kalman Filter (EKF). TMS320vc33 floating point DSP is used for INS and EKF computations. A FPGA (Field Programmable Gate Array) based GPS data acquisition system and Dual Port RAM (DPRAM) have been designed and used for the presented architecture.