Ill condition identification for linear prediction-based MIMO-OFDM channel estimation

Based on our previously developed signal perturbation free (SPF) transmit scheme, a new linear prediction (LP)-based channel estimation algorithm with ill condition identification is proposed for MIMO-OFDM systems. Based on the estimated correlation matrix of the received signal without signal perturbation error by employing the SPF transmit scheme, a SPF LP-based MIMO-OFDM channel estimation algorithm is developed. In this algorithm, the common ill-conditioning problem existing in most LP-based blind methods is properly solved based on two ill-condition test rules. Computer simulations show that the proposed SPF LP-based channel estimation algorithm significantly outperforms the same algorithm without ill condition identification, the nulling-based semi-blind method as well as the LS method.