MULTIPLE CFO ESTIMATION IN OFDM UNDERWATER COMMUNICATION SYSTEMS

Abstract

When channel is severely frequency selective, Orthogonal Frequency DivisionMultiplexing (OFDM) is a suitable candidate for high rate communications. In multipath fadingchannels like harsh underwater channels, due to Inter Carrier Interference (ICI) and CarrierFrequency Offset (CFO) the performance of this technique is dramatically decreased. Therefore,these effects must be alleviated to overcome the performance degradation. To do that, first theCFO's must be estimated. In this paper, we consider an underwater channel with three sets ofpaths and we assume that the paths in each set have similar Doppler scale factor. For minimumtransmission power, Zero Padded OFDM (ZP-OFDM) is considered. We propose a new methodbased on MUltiple SIgnal Classification (MUSIC) technique that facilitates the estimation ofmultiple CFO's in Under Water Acoustic (UWA) channel and compare it with the derivedMaximum Likelihood (ML) estimator algorithm. Simulation results show that the proposedmethod is robust against the received Signal to Noise Ratio (SNR) variations.