Research On Key Technologies Of The Water Pipe Underwater Acoustic Communication System

The water pipe is the basic unit that composes the urban water supply system and risers.It can realize underwater acoustic communication,and support applications such as Internet of Things in underwater pipeline network.The geometric underwater acoustic channel model based on the shallow sea can effectively simulate the propagation characteristics of acoustic waves on the foundation of low complexity.However,the water pipe is different in water quality,noise,and channel physical structure from the ocean.The existing geometric underwater acoustic channel model cannot truly simulate the water pipe acoustic characteristics.At the same time,the commonly used channel estimation methods have weak ability to eliminate multipath effects,resulting in poor performance of the receiver in terms of accuracy and effect.Therefore,this paper proposes a channel modeling and estimation algorithm suitable for the water pipe and underwater acoustic environment to solves the problem of modeling research on underwater acoustic channel.And it realizes the purpose of improving the quality of underwater acoustic communication.This paper conducts research on the channel modeling and channel estimation of the water pipes,the main work is as follows:(1)Introduces the acoustic propagation characteristics of the water pipe,investigates the principles of the underwater acoustic Orthogonal Frequency Division Multiplexing(OFDM)communication system and channel estimation,and focus on the transmission loss and multipath effects of acoustic waves in the water pipe.(2)Considering that shallow sea and water pipe are similar and there is no sound velocity jump layer,we adjust and optimize the parameters of the shallow sea geometric underwater acoustic model(such as sound velocity,time delay and propagation loss)to obtain a geometric underwater acoustic channel suitable for the water pipe model.(3)A complex residual network channel estimation algorithm based on Least Square(LS)is proposed.Facing the serious multipath effects of the water pipe underwater channel,this paper combines traditional estimation algorithms and deep learning to perform effective channel estimation at the receiver to reduce the interference of the water pipe to the signal.(4)Apply the proposed water pipe underwater acoustic channel model and estimation algorithm to the actual water pipe scene.Through the water pipe experiment and semi-physical simulation,the effectiveness and reliability of the algorithm proposed in this paper are verified in the actual water pipe environment.And an effective solution is provided for the underwater networking application of urban water supply system.