DS-CDMA system with joint channel estimation and MAP detection intime-selective fading channels

摘要

In this paper, maximum a posteriori (MAP) detection is applied tona direct-sequence code-division multiple-access (DS-CDMA) system jointlynwith identification and estimation of time-selective fading channels. Bynsampling the outputs of the matched filter and combining antenna arraynelements, strong and time-varying multiple-access interference (MAI) isncharacterized and suppressed instantaneously. The decision statisticsnfor MAP detection are obtained from the conditional probability densitynfunction of the prediction error. The prediction is accomplished bynapproximating the fading channel with a constrained nonlinear statenmodel. Unknown parameters such as auto-regressive (AR) processncoefficients, noise covariance matrices, and the antenna array vectornare estimated based on received sample vectors only. Also, differentialnmodulation is applied to eliminate the need for pilot insertion. Throughncomputer simulations, near-optimum bit error rates (BERs) arenfound