MRC-Based Decision-Directed LMMSE Channel Estimation for Efficient OFDM Transmit Diversity

This paper proposes an linear minimum mean square error (LMMSE) channel estimator with maximum ratio combining-based decision directed (MRC-DD) technique to increase spectral efficiency in orthogonal frequency division multiplexing (OFDM) transmit diversity. For Alamouti's transmit diversity, pilot symbols are exclusively allocated in two OFDM symbols on different antennas. This results in serious degradation of spectral efficiency. The proposed subcarrier allocation scheme uses repetition-coded data symbols to reduce the number of pilot symbols. The proposed channel estimator consists of an MRC-DD channel estimator and an LMMSE channel estimator. The MRC-DD channel estimator performs a two-step estimation procedure to obtain more accurate CFR estimates at the repetition-coded data locations: the first coarse LMMSE step and the second fine DD step. Then, the final LMMSE channel estimator gives the CFR estimates at the Alamouti-coded data locations. Simulation results show that the proposed channel estimator provides as good bit error rate performance as the conventional LMMSE channel estimator using more pilot symbols, whereas it improves spectral efficiency more than the conventional one.