Maximum-likelihood carrier frequency offset estimation for OFDM systems over frequency-selective fading channels

This paper considers carrier frequency offset (CFO) estimation for OFDM systems over frequency-selective fading channels. We derive three new maximum-likelihood (ML) CFO estimators. The first estimator takes the pre-DFT signal at the receiver as Gaussian and averages the likelihood function over the Gaussian variable. The other two estimators first remove the channel impulse response (CIR) and average the resulting likelihood function over time-domain transmitted symbols. We demonstrate that the presence of virtual carriers is critically important for estimation. All our proposed estimators can achieve the asymptotic Cramer Rao bound (ACRB) and can also be enhanced using embedded pilots.