A Novel Federated Prefilter Design for Ultra-Tightly Coupled GPS/INS Integration

D-J. Jwo, C-F. Yang

Abstract:	The tracking performance is a challenge to the receiver designer in complex environments such as high dynamic scenarios, indoor environments, and urban areas, etc., where the GPS tracking loops may lose lock due to the signals being weak, subjected to excessive dynamics or completely blocked. Traditional GPS tracking loop can be replaced by the vector tracking loop (VTC) for solving some of the problems in complex/harsh environments. As a specific configuration of the VTC, The ultra-tightly coupled (UTC), sometimes referred to as the deeply coupled, GPS/INS navigation system integrates the signals from the tracking loops of the GPS receiver with the position and velocity obtained from the INS. In such configuration, the Doppler signal derived from INS aids the tracking loop to remove the dynamics from the GPS signals, which reveals many benefits, e.g., disturbance and multipath rejection capability, improved tracking capability for weak signals, and reduction of acquisition time. Due to the large computational load of the centralized filter configuration, the federated filter-based UTC integration structure has been developed to distribute the computing tasks between different filters, leading to overall filter complexity reduction and computational requirements, which is attractive for efficient real-time implementation. Since the measurement (I and Q accumulator outputs) for the prefilters and the system states are highly nonlinearly related, the nonlinear filters are potentially useful. In addition to the extended Kalman filter (EKF) and the unscented Kalman filter (UKF) approaches, one of the alternative methods is to use the Bayesian filter/particle filter (PF) to improve the mitigation/interference suppression capability. For certain problems, a Gaussian assumption cannot be applied with confidence. To overcome these problems, the probability-based estimator PF was proposed to approximate the posterior distribution for highly nonlinear systems. The PF possesses superior performance as compared to EKF and UKF as an alternative estimator for dealing with the nonlinear, non-Gaussian system. Incorporation of the adaptive algorithm into the nonlinear filters will be performed for dealing with the low quality signal and high dynamic environment. Exploration of the determination of the process and measurement noise covariance matrices needs to be involved for dealing with the varying dynamic scenario. Performance assessment will involves the improvement of tracking capability for weak signals in high dynamic environments and rejection capability for non-Gaussian disturbance. Performance evaluation for the proposed design as compared to the others will be carried out.
Published in:	Proceedings of the ION 2013 Pacific PNT Meeting April 23 - 25, 2013 Marriott Waikiki Beach Resort & Spa Honolulu, Hawaii
Pages:	1010 - 1022
Cite this article:	Jwo, D-J., Yang, C-F., "A Novel Federated Prefilter Design for Ultra-Tightly Coupled GPS/INS Integration," Proceedings of the ION 2013 Pacific PNT Meeting, Honolulu, Hawaii, April 2013, pp. 1010-1022.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In