A novel particle filtering approach to blind symbol detection and timing estimation

This paper addresses the problem of joint estimation of the timing epoch and the sequence of transmitted symbols in a digital transmission system. Since optimal timing recovery is analytically intractable in general, most existing synchronization methods are either heuristic or based on approximate maximum likelihood (ML) arguments. We introduce herein an alternative approach which is derived within a Bayesian estimation framework and is implemented via the sequential Monte Carlo (SMC) methodology. Assuming the relative delay between the received signal and the local clock reference varies according to an autoregressive model, the proposed SMC technique guarantees asymptotically optimal timing recovery and, when adequately combined with a closed-loop receiver architecture, minimal symbol error rate (SER).