ML-Based Joint Estimation of Carrier Frequency Offset and Doubly Selective Channels for OFDM Transmissions

This paper proposes a pilot-aided joint channel estimation and synchronization scheme for burst-mode orthogonal frequency division multiplexing (OFDM) systems over time- and frequency-selective (doubly selective) channels. By exploiting the basis expansion model (BEM) for representing doubly-selective channels, a maximum likelihood (ML) cost function of carrier frequency offset (CFO) and BEM coefficients is formulated to develop a ML framework for the joint synchronization and channel estimation problem. Inheriting the properties of the ML estimation, the proposed estimator is unbiased and its mean-squared-error (MSE) performance achieves Cramer-Rao lower bounds (CRLB) asymptotically (for a large data records). Simulation results demonstrate that, over a wide range of Doppler spreads, the proposed estimation scheme offers a high robustness against the time variation of fast fading channels and outperforms the linear minimum mean square (LMMSE) algorithm. In addition, the ML-based algorithm achieves CRLB at very low signal-to-noise ratio (SNR) regimes.