Semi-Blind Channel Estimation for OFDM using Least Squares

Subspace blind channel estimation which takes advantage of induced cyclostationarity in orthogonal frequency division multiplexing (OFDM) has been explored by many researchers. This method however, requires the knowledge of the probabilistic model of the channel input which is not always available. In this paper, a semi-blind channel estimation method is proposed which considers the OFDM channel input as an unknown deterministic signal. This method is semi-blind because pilot tones provided by the IEEE 802.11a std. are used to better track channel variations. Simulation results show the proposed approach provides an equally efficient bit error rate performance