Sparse Equalization Based On Minimum Symbol Error Rate In Underwater Acoustic Communications

The strategic position of the ocean in the country is constantly highlighted,and the underwater acoustic communication technology closely related to it has received more and more attention from various countries.It has become an urgent need to develop a high-speed and reliable underwater communication system.However,the complex and changeable underwater environment,as well as the high time delay caused by the use of sound waves as the transmission medium,leads to many difficulties in underwater acoustic communication.In recent years,sparse signals have become a hot research field,and many scholars have carried out research on sparse channel estimation and sparse channel equalization.The underwater acoustic channel has natural sparsity,and since the equalizer can be regarded as the inverse of the channel in the frequency domain,the equalizer is correspondingly sparse in this approximation.In order to improve the convergence characteristics of the system,it is of great significance to study the adaptive sparse equalizer that can be used underwater.At the same time,underwater acoustic communication has lower requirements for real-time performance due to slower transmission.But it requires the Symbol Error Rate(SER)is low enough,the traditional Minimum Minimum-Mean-Square-Error(MMSE)can't achieve desired SER,and based on Minimum-bit-Error-Rate(MSER)criterion of adaptive equalization algorithm can reduce system SER.It can improve system reliability and has advantages in underwater acoustic communication.The work of this thesis focuses on the derivation,simulation and practical application of sparse adaptive MSER equalization algorithm.1.Firstly,the existing adaptive sparse filtering algorithm based on MMSE criterion is introduced,especially the coefficient proportional adaptive filtering algorithm.The filtering algorithm of this type mainly distributes the independent step size proportional to the tap value of the filter,so as to improve the convergence of the algorithm.2.Inspired by the existing sparse adaptive filtering algorithm and under the study of sparse characteristics of underwater acoustic channel and Zero Force(ZF)equalizer,sub-gradient projection method based on MSER criterion is used in two modulation modes,and the adaptive proportional minimum error rate(PMSER)equalization algorithm for linear structure and decision feedback structure is obtained.Intuitively,the proposed algorithm allocates independent step sizes to different filter taps by using sparse matrix.Finally,in order to determine the sparse matrix,the rules for sparse matrix element selection are derived,and two sparse selection schemes are given,which are called SC-PMSER and Z-PMSER respectively.The performance of the proposed algorithm is compared and simulated in a static sparse simulation channel.3.In order to cope with the complex and changeable actual underwater acoustic channel,this paper combines PMSER algorithm with Turbo equalization structure,and designs the receiver with digital phase-locked loop and time reversal technology.The validity of the proposed structure is verified by simulation in an actual static channel.Finally,the designed receiver structure is used to decode the data received in the indoor pool and Qiandao lake experiment.The experimental results verify that the algorithm has the characteristics of low bit error rate and fast convergence.