Channel Estimation for OFDM Systems Based on RLS and Superimposed Training Sequences

The recursive least squares (RLS) algorithm and the superimposed training sequences are applied to the orthogonal frequency division multiplexing (OFDM) systems to estimate channel state information (CSI). In order to reduce the interference caused by the unknown information data that is added with the superimposed training sequences, we present the following method: first, the information data are detected using the CSI of the previous one block, and then subtracted them from the current received signals. Second, the remainder signals mainly including the training sequence are used to estimate CSI. For decreasing the computation complexity, we use the same training sequence in all OFDM blocks. The computation complexity and the mean square error (MSE) performance of our proposed RLS method are compared with the original RLS method.