Recursive MMSE Channel Estimation for MIMO-OFDM Systems

Recursive minimum mean-square-error estimation (MMSE) with low-complexity computation is considered for online implementation of channel acquisition in multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) systems. The channel correlation matrix is estimated by three approaches respectively, i.e., averaging measurements, direct tracking signal subspace of measurements and iterative calculation from measurements. Simulations demonstrate that the MMSE-based algorithms will obtain higher accuracy than least square estimation and its recursive pattern. In comparison of the three approaches, it is shown that the method of tracking signal subspace of measurements, which can separate the measurement space into signal subspace and its orthogonal noise subspace, is comparable to optimal MMSE estimation.