Research On Channel Response Estimation Of Underwater Acoustic Channel

Shallow water is a significantly strategic region in both military and civilianaspects. As a carrier of energy with farthest propagation and best performance on thesea, the sound wave is widely used in underwater information transmission over along distance. A method of channel response estimation is given to obtain theunderwater acoustic channel transfer function. It provides vitally technical guaranteefor the understanding of the sound propagation characteristics in ocean, the reliableunderwater acoustic communication, the use of underwater acoustic equipments andthe information exchange between ships.This thesis analyzes the main characteristics of shallow water acoustic channeland the difficulties of realizing the channel response estimation, and makes themodeling for shallow water acoustic channels. Combined with the results of soundpropagation experiment on the sea, an algorithm for the case of a known source isdeveloped and then extended to the case in which the source signal is not knownprecisely.The main research contents in this dissertation are as follows:1、Summarizes the research situation of the underwater acoustic channelresponse estimation roundly, and briefly describes the structure and assumptions ofthe shallow water multipath channel model. Then the effect of marine environment onthe underwater acoustic channel is discussed from four aspects: velocity profile,sound transmission loss, types of seafloor sediment and ocean ambient noise.2、An underwater acoustic channel response estimation method based on thede-dispersion transform(DDT) is presented here. The channel response function canbe estimated accurately for known source signals without knowledge of any oceanenvironmental information a priori. Compared with traditional methods, this approach can effectively eliminate distortion of received signals due to dispersion byDDT and therefore improve the efficiency of channel response estimation.3、As for the case in which the source signal is not known, by using the modefilter blind deconvolution algorithm, the source signal is extracted from the datareceived by a vertical array based on the normal mode theory and inherent propertiesof underwater acoustic channel. The modal dispersion is removed by employing DDT,thus improve the accuracy of the algorithm.4、Numerical simulations are performed in two typical shallow water waveguides—ideal and summer waveguide, respectively. The basic theoretical derivation andinitial value estimation process are given. Then, by comparing the performance ofunderwater acoustic channel response estimation (arrival time and amplitudeestimation of different frequency components of normal modes) before and after theDDT, the advantage of the proposed DDT-based method is verified. Finally, themethod is proved to be robust and has fast convergence performance by analyzing theLaoShan bay sound propagation data.