Underwater Acoustic Channel Tracking with Cluster Variation Learning for Acoustic Mobile OFDM Communication.

In this paper, we propose a channel tracking method with cluster variation learning in multicarrier communications, by exploiting the cluster-specific channel coherence for time-varying underwater acoustic (UWA) channels. We develop a cluster variation model due to the cluster-specific channel coherence. In the cluster variation model, the path delay and the Doppler scale assemble the range and the velocity in the target movement. Inspired by the target tracking for its movement, we propose a channel tracking method which can incorporate the cluster variation model with the measurement model efficiently through Kalman filtering. The cluster variation model can learn its parameters block by block through pilots, therefore reduce its model mismatch adaptively. The proposed method has a moderate computational complexity, since it eliminates the explicit Doppler scale estimation and Doppler variation model comparing to the conventional channel estimation. Using simulated data as well as experimental data in mobile UWA communications with one phone element, we study the system performance in term of mean square error (MSE) and symbol error rate (SER) performance, and show that the proposed channel tracking method improves the performance significantly. [ABSTRACT FROM AUTHOR]