Sparse channel estimation in OFDM systems using compressed sensing techniques in a Bayesian framework

Compressed Sensing (CS)-based sparse channel estimation in a Bayesian framework for Orthogonal Frequency Division Multiplexing (OFDM)-based communication systems is presented in this paper. An OFDM signal model having interference free region of the received training sequence is developed. The significance of Bayesian approach in the formulation of an estimator is shown by Bayesian Bound Analysis. Based on the developed signal model, the interference-free region of the received OFDM signal is used for sparse channel estimation, utilizing CS reconstruction algorithms and prior statistical knowledge of channel. The proposed CS-based channel estimation method in the statistical framework results in a low complexity estimator, where the received samples used for estimation are less than that required for conventional techniques using Maximum Likelihood and Maximum a posteriori methods. The estimation methods are analyzed by numerical simulations and are found to have better performance when compared with previous algorithms.