Signal detection and timing estimation via summation likelihood ratio test

Abstract

This paper presents a novel scheme for improving signal detection and timing estimation in communication systems that employ unique words (reference symbols) to aid signal acquisition. The newly proposed algorithm which is termed summation likelihood ratio test (SLRT) performs a summation of likelihood values over a suitable search range. Computer simulations of the SLRT and the generalized likelihood ratio test (GLRT) have been performed for a particular classification scenario of the traffic channel (TCH) and the fast associated control channel (FACCH) traditionally used in terrestrial wireless communications systems such as GSM and mobile satellite communication systems. The simulation results demonstrate that the SLRT outperforms the GLRT significantly.