EM/SAGE Based ML Channel Estimation for Uplink DS-CDMA Systems over Time-Varying Fading Channels

Abstract

The matrix inversion for the maximum likelihood (ML) channel estimation requires high complexity for the direct-sequence code-division multiple-access (DS-CDMA) systems. The prime motivation of the paper is to propose channel estimators that achieve mean square error (MSE) performance of ML channel estimator in an iterative manner without any matrix inversion. Therefore, two computationally efficient solutions to the problem of ML channel estimation are proposed. We compare the both algorithms in terms of the number of used iteration and show that the proposed algorithms converge the same MSE performance of the ML estimator as the increasing number of iterations.