ML estimation of timing, integer frequency and primary sequence index in LTE systems

This paper addresses the problem of maximum likelihood (ML) estimation of slot timing, integer carrier frequency offset and primary sequence index for the downlink of Long Term Evolution (LTE) systems. The proposed algorithm is designed to exploit the knowledge of the pilot Zadoff-Chu sequence embedded in the primary synchronization signal (PSS). The estimation process is affected by the presence of a large set of nuisance parameters, which need to be estimated jointly with the parameters of interest. As a consequence, the exact ML solution is extremely complex and we have developed a suboptimal algorithm designed to provide a good balance between estimation accuracy and complexity. In particular, a key finding is a reduced-rank representation for the frequency response of the channel, which is required by the ML estimator but is not available at receiver prior to having acquired synchronization. Compared to existing alternatives, the resulting scheme exhibits improved accuracy in the estimation of all three parameters of interest.