On optimal resampling for OFDM signaling in doubly-selective underwater acoustic channels

Orthogonal frequency division multiplexing (OFDM) is under strong consideration as a modulation for underwater acoustic channels. Underwater acoustic channels are wideband in nature due to the fact that the signal transmission bandwidth is comparable to the carrier frequency. Furthermore, these channels are characterized by long delay spreads, sparsity and time variation. The time-variation challenges the use of OFDM - destroying carrier orthogonality and introducing significant inter carrier interference. Much of the current work on time varying channels assumes a single common Doppler shift for each path. As such, resampling is optimal preprocessing in such cases. The current work examines the choice of resampling factor for the case of distinct Doppler shift on each path. The optimal resampling factor is selected to maximize the Fisher information for the transmitted data; Fisher information is also used to evaluate the information loss due to resampling. The findings suggest that the optimal resampling factor is near that associated with the Doppler shift of the strongest path. It is also observed that the information loss is modest in many cases using this practical preprocessing.