Channel estimation for OFDM systems with multiple transmit antennas by filtering in time and frequency

We address pilot-symbol aided channel estimation (PACE) for orthogonal frequency division multiplexing (OFDM) based systems, with multiple transmit antennas. An approach is presented which estimates and separates N_T superimposed signals, corresponding to N_T transmit antennas, in two dimensions. More specifically, we extend the concept of using two one dimensional (1D) estimators, instead of a truly two dimensional (2D) estimator for MIMO-OFDM systems, by dividing the estimation as well as the separation task into two stages. Each estimator only separates a subset of the N_T superimposed signals, together with smoothing and interpolation in the respective dimension. Design rules for the optimization of the proposed 2×1D estimator are established, dependent on the performance, computational complexity, induced delay, and overhead due to pilot symbols.