Globally optimum ML estimation of timing and frequency offset in OFDM systems

We present a globally optimum maximum likelihood estimator for carrier frequency and symbol timing synchronization of orthogonal frequency division modulation (OFDM) systems. The approach proposed here is an extension of the previously proposed estimation method given by Van de Beek, Sandel and Borjesson (see IEEE Trans. on Signal Processings, vol.45, no.3, p. 1800-05, 1997). Our research indicates that the previously proposed likelihood metric does not globally characterize the estimation problem. Based on this observation, a new likelihood function for the joint timing and frequency offset estimation is derived. The resulting probabilistic measure is used to develop an unbiased ML estimator. Furthermore, a moment estimator for carrier frequency offset estimation under uncertain symbol timing is proposed and the statistical property of this estimator is analyzed. In comparison to the previously proposed methods, the proposed estimator is computationally and statistically efficient which makes the estimator more attractive for real time applications. The performance of the proposed estimator is assessed through computer simulations.