A novel structure for MMSE transceivers over slowly time-varying channels

This paper addresses the design problem of decision feedback (DF) transceiver without zero-forcing constraint over slowly time-varying narrowband multi-input multi-output (MIMO) channels. The space-time generalized triangular decomposition (ST-GTD) is applied for the design of minimum mean square error (MMSE) DF transceiver. With space-time powerloading, the proposed space-time geometric mean decomposition (ST-GMD) MMSE transceiver maximizes Gaussian mutual information over the equivalent channel seen by each space-time block. For practical applications, the causal ST-GTD MMSE transceiver which does not require channel prediction but shares the same asymptotic bit error rate (BER) performance with the ST-GMD MMSE system is also developed. In high signal to interference plus noise ratio (SINR) region, our results show that the proposed MMSE transceivers have better BER performance than the conventional GMD-based MMSE transceiver; the average BERs of the proposed systems are a non-increasing function of the ST-block size.