Improved carrier frequency offset estimator for common referenced sampling and RF oscillators in OFDM transceivers

Orthogonal frequency division multiplexing (OFDM) has been proven to be very sensitive to carrier drifts due to fact that both carrier frequency offset (CFO) and sampling frequency offset (SFO) destroy the orthogonality cross multiple subcarriers. The sampling clock and the local oscillator, from which the carrier frequency is derived, sharing a common reference clock is becoming a configuration extensively used in wireless transceivers. In this paper, we propose a maximum likelihood (ML) estimation scheme with the common reference clock assumption. We further propose two CFO estimators with low computational complexity - a near ML estimator and a method of moments estimator. With the common reference clock assumption, the CFO estimation range is widened to many multiples of subcarrier spacing without any additional training symbols or constructing the training symbol by multiple identical parts. The performance of the proposed algorithms in comparison with existing estimators is studied through simulations. The proposed algorithms are validated on a real-time hardware testbed.