Approximate Maximum Likelihood Sequence Detection

The problem of multiuser detection in asynchronous Direct-Sequence Code-Division Multiple-Access systems with residual errors in timing and amplitude estimation is addressed. Prior work has shown that the maximum-likelihood (ML) solution is sensitive to timing errors as small as a fraction of a chip period. In this paper, residual errors in both synchronization and fading coeecient estimation are assumed. A modiied ML metric is derived with the knowledge of the assumed statistics of the residual errors. As a closed-form solution does not exist, an approximate ML metric is derived based on perturbation theory. A recursive implementation scheme for the approximate formulation is proposed with the aim of reducing the complexity and facilitating dynamic updates of the receiver parameters. Performance is investigated via simulation and analytical bounds for both Gaussian and at Rayleigh fading channels.