Real time LMMSE channel estimation for wireless OFDM systems with transmitter diversity

We propose an enhanced low-complexity channel estimation technique for wireless orthogonal frequency-division multiplexing (OFDM) systems with transmitter diversity in the presence of multipath propagation. It shows that real time knowledge of the system's signal-to-noise ratio (SNR) and channel's RMS delay spread, derived from a channel frequency response (CFR), can further improve the linear minimum mean-squared error (LMMSE) estimation, especially if the OFDM training symbols used require interpolation at the receiver end. With additional preprocessing on the CFR, applied before the LMMSE estimation, and with knowledge of the SNR, computer simulations show that the proposed channel estimation technique gives up to 2 dB improved performance gain for high SNRs over a single fixed set case scenario LMMSE estimator in terms of the mean-squared error (MSE).