Semiblind Iterative Receiver for Coded MIMO-OFDM Systems

In this paper, a semiblind iterative receiver is proposed for coded multiple-input-multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) systems. A novel iterative extended soft-recursive least square (IES-RLS) estimator for joint channel and frequency offset estimation is introduced. Extrinsic bit information obtained from the channel decoder and expected symbol decisions obtained from the demodulator are simultaneously exploited by the IES-RLS. The proposed receiver combines the MIMO data demodulator, the proposed channel estimator, and the channel decoder in an iterative manner to yield an improved bit error rate (BER) performance. To arrive at a feasible algorithm, the first-order linearization of the received vector signal with respect to the frequency offset is used in the IES-RLS channel estimator. The BER performance, a constellation-constrained mutual information analysis, and an EXIT chart analysis are used to verify the effectiveness of the proposed receiver. Simulation results show the superiority of the proposed semiblind receiver, compared with conventional semiblind receivers.