Performance Evaluation of LQG based Optimal GNSS Tracking Loop using Non-linear Measurements

Abstract

This paper proposes a new GNSS signal tracking loop based on linear quadratic Gaussian regulator (LQG) algorithm and evaluates its signal tracking performance. The proposed LQG based tracking loop consists of the extended Kalman filter (EKF) and linear quadratic regulator (LQR). The EKF uses I, Q correlation sum as measurements, and estimates code, carrier phase, the Doppler frequency and the Doppler rate as states. While conventional loop filter uses simple integrator for numerically controlled oscillator (NCO) input, the proposed loop filter adopts optimal control gain using LQR controller. To assess the performance of the proposed tracking loop, this paper performs Monte-Carlo software simulation and IF data processing from hardware GPS signal simulator using GNSS software receiver. The simulation results show that the performance of the proposed LQG based tracking loop ranges in error from 0.1 to 1 Hz DLL and from 20 to 40 Hz PLL for static user and shows equal or better performance than the those of the conventional PLL for high dynamic user.