Blind CFO tracking in MIMO-OFDM systems at low SNR

A novel blind carrier frequency offset (CFO) estimation and tracking algorithm at low SNR is proposed for multi-input multi-output orthogonal frequency division multiplexing (MIMO-OFDM) system. The CFO is estimated initially by cyclic correlation in nonzero cyclic frequency and estimating signal parameter via rotational invariance techniques (ESPRIT), and then CFO is tracked based on least mean square (LMS) criterion. Theory analysis shows that the algorithm doesn’t require pilot symbols or training sequences, adapts to the synchronization under any number of receive and transmit antennas and any distributed stationary noise. Simulation results verify the algorithm can track abruptly changed CFO and linearly changed CFO, and keep stable at low SNR.