Preamble-based Estimation of Highly Frequency Selective Channels in MIMO-FBMC/OQAM Systems

Filter bank-based multicarrier systems using offset quadrature amplitude modulation (FBMC/OQAM) are among the top candidates for replacing orthogonal frequency division multiplexing (OFDM) in future wireless networks. However, despite their many proven advantages over both OFDM and other FBMC schemes, they are known to suffer from a selfinterference effect, inherent to these systems. Together with the nonnegligible frequency selectivity of their subchannels in realistic propagation conditions, this constitutes a fundamental difference to OFDM, especially in receiver signal processing tasks such as channel estimation. The challenge is intensified in multiple-input multiple-output (MIMO) systems, where the multi-antenna interference has also to be taken into account. Recently, there has been a significant progress in the study of channel estimation for FBMC/OQAM systems, mainly for the single-antenna case. Notably, a method that provides a closedform estimate of the channel, while making no assumption on its frequency selectivity and using a very short preamble for training, was reported and shown to significantly outperform classical methods. Moreover, the preamble was optimized with respect to the channel estimation mean squared error. In this paper, these results are extended to an analogous MIMO setup. The MIMO-OFDM case is also revisited as a special case. Simulation results are presented that verify the good performance of the proposed method for both moderately and highly frequency selective channels.