Abstract:

Conventional training based methods, comprising linear least square (LS) channel estimators are known to be optimal under the assumption of rich multipath channels. However, underwater acoustic (UWA) channels are naturally sparse, so the LS channel estimators can just get poor performance in UWA channels. A sparse UWA channel estimation algorithm based on compressive sensing (CS) for orthogonal frequency division multiplexing (OFDM) is proposed. Firstly, the training waveforms that consist of the elements of a complete orthogonal short time Fourier (STF) basis are used as the pilots of OFDM to “sense” the UWA channel, and then, a CS recovery algorithm is adopted to estimate the channel. Simulation results based on the real UWA channel demonstrate that the performance of the proposed schemes is better than the conventional LS estimators for it can lead to significant reductions in the training energy, loss in spectral efficiency and improvement in the robustness to noise. This sparse channel estimator is matched to the UWA channel very well and especially adapt for the moving platform OFDM UWA communication system.