An MMSE Approach to Channel Shorting for Underwater Acoustic FH-FSK Communication

Underwater acoustic channel is the biggest obstacle which causes the underwater acoustic communication falling behind the wireless communication. Generally speaking, underwater acoustic channel will represent doubly spread, i.e. multipath spread and Doppler spread. Also the multipath spread will be quite long in the underwater acoustic channel. For example, the multipath spread will be several milliseconds in the shallow sea and several seconds in the deep sea. Multipath spread will lead to Inter-Symbol Interference (ISI), and Doppler spread will cause the channel to be rapidly time-varying. Although ISI can be suppressed by channel equalizer, but because of the long multipath spread, it increases the complexity about the design of the equalizer. In this paper, channel shortening will be used to reduce the ISI. We present two methods of channel shorting, which are Time Reversal (TR) filter and the Minimum Mean Squared Error (MMSE) filter. When the two filters are used in FH-FSK, we found that the performance of MMSE filter is better than TR filter. Taking into account the characteristics of sparse underwater acoustic channel, Compressive Sensing (CS) algorithm is used for channel estimation. Finally, We perform a series of numerical simulations and the experiments on the lake to show that MMSE filter outperformances the TR filter under the FH-FSK system using the bit error rate (BER) as an evaluation criterion.

A New Blind Equalization Algorithm Suitable for Sparse Underwater Acoustic Channel

The Constant Modulus Algorithms is a mature algorithm for a long time. The researchers are fond of its solidity. But in the area of shallow sea or blue sea, the CMA has poor constringency. In this passage, based on a improved blind equalization algorithm, we proposed a new algorithm which suitable for sparse underwater acoustic channel. The computer simulation results demonstrate fast convergence and fewer accounts.