Channel Estimation Based on Extended Kalman Filtering for MIMO-OFDM Systems

Abstract

MIMO-OFDM systems are considered as the solution scheme for future wideband wireless communication systems. And channel state information play a crucial role in data detection module. But Kalman filtering (KF) is a linear processing method which is not fit for processing nonlinear problem. While wireless channels have some nonlinear characteristics. Owing to the above reasons, this paper describes a channel estimation method based on extend Kalman filtering (EKF) for MIMO-OFDM systems. The proposed method based on EKF can exploit pilot symbols and an extended Kalman filter to implement channel estimation without the aid of any prior knowledge of channel statistics. In comparison with the channel estimation methods based on the least square (LS) and the least mean square (LMS) algorithms, the method base on EKF has better performances theoretically. Computer simulations also demonstrate the channel estimation method based on EKF outperforms the channel estimation methods based on LS and the LMS. Hence, the channel estimation method based on EKF can offer a dramatic system performance improvement at the cost of modest computational complexity