Research On Doppler Shifts Estimation Based On Lfm Signal In Underwater Acoustic Communication

As people accelerate the development and utilization of ocean, underwater communication in the ocean are more and more attractive, whether in the field of military or the field of commerce. However, due to the electromagnetic wave which is the medium used in air-ground communication environment are faster than the acoustic wave which is the medium used in underwater acoustic environment of a rate of about five orders of magnitude, the Doppler effect has a serious impact on the quality of underwater acoustic communication and has greatly restricted the development of underwater communication. Therefore, in order to ensure the accuracy and reliability of underwater communication and prevent the signal distortion, estimation and compensation of the Doppler shift factor is necessary.This paper analyzes the characteristics of underwater acoustic channel, including the theory of acoustic speed in sea, propagation loss, multipath effects, ambient noise, and the Doppler effect, and the effect of these characteristics on the propagation of acoustic waves. Two main methods to estimate the Doppler shift factor are introduced, including ambiguity function method and block Doppler estimation method.Because the accuracy of Doppler estimation methods for existing is not enough and the computational complexity of those methods are large, this paper proposed a Doppler shift factor estimation method based on LFM(Linear Frequency- modulated) signal. Firstly construct the theory of noiseless received signal by known LFM signa l, then using least squares method minimize the error of the signal to obtain the objective function, at last to find the optimum of the objective function to obtain Doppler shift. The innovation is dividing the interval by the extremes of the derivative o f the objective function to narrow the search, and then using particle swarm algorithm to search the global optimum, it improves estimate accuracy. Finally, compare the performance of several algorithms show that the new proposed method is simple to achieve with a small amount of computation, and can guarantee the accuracy and reliability of the estimates.