Iterative Decoding, Offset and Channel Estimation for OFDM using the Unscented Kalman Filter

An iterative decoding and estimation algorithm for OFDM is presented for time-varying fading channels with unknown carrier frequency offset (CFO). The unscented Kalman filter (UKF) is applied to obtain accurate joint estimates of CFO and multipath channel coefficients. A new unscented Inverse covariance filter is also developed, which reduces the computational complexity of the filter by appropriate approximations to the covariance matrices. The UKF and LDPC code decoder are then combined in an iterative structure. Simulation results are provided for high speed mobile environments under the IEEE 802.16 OFDM standard.