Adaptive MMSE-SVD for OFDM downlink MU-MIMO in a high mobility environment

Abstract

In this paper, we propose an adaptive minimum mean square error filtering combined with singular value decomposition (MMSE-SVD) for OFDM downlink multi-user multi-input multi-output (MU-MIMO) transmission. During data transmission, base station (BS) updates the multi-user MMSE transmit filter by using linear channel prediction, while user equipments (UEs) update their eigenmode receive filters, constructed by SVD, by using decision feedback adaptive channel estimation. The uncoded BER performance achievable by adaptive MMSE-SVD is evaluated by computer simulation. It is shown that proposed adaptive MMSE-SVD can increase an allowable maximum Doppler frequency (f_DT) for keeping BER < 10⁻² by about 4 times.