Joint channel tracking and symbol detection for OFDM systems with Kalman filtering

This paper proposes a new joint channel tracking and symbol detection scheme for pilot symbol-assisted OFDM systems in multipath fading. The proposed scheme uses Kalman filters for both channel tracking and subsequent equalization which are combined in the coupled estimator structure. Modelling the multipath fading channel as random processes to describe channel variations in a general AR framework lends itself to a state-space representation that enables the application of Kalman filtering for the tracking of channel variations. However, the proposed tracking algorithm requires knowledge of the transmitted symbols. This implies that an iterative method should be sought to obtain alternatively either channel or transmitted symbols. To compose the coupled estimator structure, a linear Kalman filter equalizer with the corresponding state-space model is therefore proposed for the detection of transmitted symbols. With the proposed Kalman filters, iterative structure is utilized to decode transmitted symbols and subsequently to track channel parameters. Finally, the performance of the proposed method is investigated through the experimental results