GPS Navigation using Neural Networks

Mangesh Chansarkar

Abstract:	Neural Networks have been proposed as nonlinear filters in a variety of applications such as Blind Signal Separation, Image Registration, Blind Deconvolution and many others. This paper presents a new approach to solving the GPS pseudorange equations using 3 layer Neural Networks. A three layer Radial Basis Function (RBF) Neural Network is designed which solves the nonlinear GPS pseudorange equations directly as opposed to the Linear Least Squares or Extended Kalman Filter approaches in traditional GPS receivers. A carefully selected cost function is minimized using a novel variation of the classical conjugate gradient algorithm such that training time for the Neural Network is minimized. Simulations have been performed at SiRF Technology Inc which show stable behavior even under bad geometry conditions where the traditional Recursive Least Squares and Extended Kalman Filter approaches show high sensitivity to measurement errors. Under good geometry conditions the neural network solution shows slightly improved noise performance compared to the expected performance of traditional Least Squares solution.
Published in:	Proceedings of the 12th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1999) September 14 - 17, 1999 Nashville, TN
Pages:	1235 - 1240
Cite this article:	Chansarkar, Mangesh, "GPS Navigation using Neural Networks," Proceedings of the 12th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1999), Nashville, TN, September 1999, pp. 1235-1240.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In