Integer frequency offset recovery in OFDM transmissions over selective channels

Carrier frequency offset (CFO) in OFDM systems is normally estimated in two steps. The fractional part of the CFO is recovered first and the remaining ambiguity is subsequently resolved by detecting the integer frequency offset (IFO). Conventional IFO recovery algorithms for OFDM signals are sensitive to multipath distortions as they are derived without explicitly taking into account the frequency selectivity of the transmission channel. In this paper, we propose a novel scheme in which the channel response and IFO are jointly estimated using a maximum likelihood (ML) approach. In doing so we exploit one or more pilot blocks placed at the beginning of the frame and carrying known symbols. Since the complexity of the resulting ML algorithm may be relatively large, we also suggest suboptimal solutions unifying various earlier proposals. Computer simulations are used to demonstrate the superiority of the proposed schemes over existing alternatives. It is shown that excellent performance can be achieved with affordable complexity even in the presence of highly dispersive channels.