Optimal Training Sequences For Efficient MIMO Frequency-Selective Fading Channel Estimation

摘要

In this paper, novel channel estimation schemes using uncorrelated periodic complementary sets of unitary sequences are proposed for multiple-input multiple-output (MIMO) frequency-selective fading channels. When the additive noise is Gaussian, the proposed best linear unbiased estimator (BLUE) achieves the minimum possible classical Cramer-Rao lower bound (CRLB), if the channel coefficients are regarded as unknown deterministics. On the other hand, the proposed linear minimum mean square error (LMMSE) estimator attains the minimum possible Bayesian CRLB, when the underlying channel coefficients are Gaussian and independent of the additive Gaussian noise. The proposed channel estimators can be implemented with very low complexity via FFT, which makes them very suitable for practical systems such as, but not limited to, MIMO orthogonal frequency division multiplexing (MIMO-OFDM) systems.