Improvement of maximum likelihood time delay estimation for GPS positioning in multipath, jamming and low SNR environments.
In this paper, we propose a robust maximum likelihood (RML) approach for delay and amplitude estimation of weak GPS signals in multipath and interference environments such as that encountered in indoor applications. First, to formulate the problem we exploit the fact that during the acquisition period the propagation delay causes only a circular shift to the spreading C/A code vector. This allows the decomposition of the received data vector into a constant component plus an undesired signals component. Then, a sample mean model is developed using a moving average over long time durations to reduce the effect of undesired zero-mean signals to one of a colored Gaussian noise. The developed model has an antijam structure which relies on the exploitation of a temporal structure property of the GPS signal, namely the replication of the C/A-code. An efficient temporal whitening technique is derived from the sample covariance matrix and applied to suppress the colored noise effect. The ML time-delay estimation of the superimposed multipath parameters becomes tractable and results in an efficient algorithm by adopting a sequential procedure.
|Main Author:||Sahmoudi, M.|
|Other Authors:||Amin, M. G.|