MIMO Precoding with ESPRIT-Like Blind Channel Estimation

This paper considers the Tomlinson-Harashima Precoding (THP) when using over multiple antenna system. In spite of its good performance, THP is very sensitive to erroneous Channel State Information (CSI) i.e., its performance largely depend on channel estimation accuracy. Most algorithms that employed in channel estimators are training based. In this paper, a correlation-based algorithm is proposed in order to estimate channel in blind method, where it does not require any training sequence for channel estimation. The CSI can be identified by an Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT)-like method and Singular Value Decomposition (SVD). In particular, we focus on the THP performance with respect to the channel estimation errors by employing spatial power loading which provides significant improvement with respect to conventional methods. We develop this loading for the THP in minimum average Symbol Error Rate (SER) sense. Simulation results are used to show the performance advantage attained by the power loading in blind channel estimation scenario.