Explicit Time Delay Estimation Algorithm Based on Maximum Correntropy Criterion and Approximate Prolate Series

Abstract

Time delay estimation (TDE) algorithms have been widely used in many applications, such as digital communications systems, source location, wireless sensor networks. Explicit TDE (ETDE)-based algorithms are very popular because of low computation complexity and sufficient accuracy in real-time application. However, conventional ETDE-based algorithms use the sinc function as filter coefficients, leading to big truncation error with few filter taps. Furthermore, these algorithms have poor performance in Non-Gaussian environment. In this paper, we use approximate prolate function rather than sinc function as filter coefficients to improve the TDE accuracy with small filter taps. And we use correntropy as cost function to restrict Non-Gaussian noise. A novel ETDE algorithm using maximum correntropy criterion (MCC) and approximate prolate series (APS) is proposed in this paper. Simulation results show that the proposed algorithm has better convergence performance and smaller steady-state error in non-Gaussian noise environment.