Efficient Time of Arrival Estimation Algorithm Achieving Maximum Likelihood Performance in Dense Multipath

Abstract

Robust and accurate time-of-arrival (TOA) estimation in dense multipath channels such as those encountered in ultra-wideband (UWB) systems is a considerable challenge especially when the signal-to-noise ratio (SNR) is low. The exact maximum likelihood (EML) TOA estimator in dense multipath conditions has the potential to attain accurate TOA estimation, however, it is too complex for practical implementation. There is a substantial performance gap between the known practical algorithms for TOA estimation and the EML estimator. In this paper, a novel practical TOA estimation algorithm is developed that attains the EML performance when the multipath arrivals are dense. When the multipath arrivals density is low the estimator does not attain the maximum likelihood performance but still outperforms other known practical estimators. The estimator does not need to know the channel characteristics accurately, thus, it is robust to various multipath channels. The approach taken is to approximate the received multipath signal as a Gaussian process and derive the maximum likelihood estimator. In order to further decrease the computational load of the new algorithm, we develop a low complexity approximation with negligible performance degradation. The algorithm is useful for either single channel realization or multiple channel realizations using diversity either in time, frequency or space. When applying diversity technique a substantial performance gain is attained due to the optimal combining of the channel realizations and thus reliable TOA estimation is attainable even at low SNR. The estimator's performance can be closely predicted by a closed-form analytical error expression.