Subspace-Based Super-Resolution Sparse Channel Estimation in MIMO-OFDM Systems

This chapter introduces a parametric sparse MIMO-OFDM CE scheme that can enable super-resolution estimation of path delays with arbitrary values based on the finite rate of innovation (FRI) theory. Since the wireless MIMO channels exhibit sparsity, the compressive sensing methods can be employed to achieve effective channel estimation. Furthermore, the spatial and temporal correlations of MIMO channels can be exploited to enhance the estimation performance. Specifically, the spatial channel correlation leads to the common sparse pattern on the path delays of different antennas, while the temporal channel correlation leads to the relatively unchanged sparsity pattern during several OFDM symbols. Taking advantage of these characteristics, the considered scheme outperforms existing state-of-the-art methods, and reduces the pilot overhead through joint signal processing across different antennas.