基于Speex信构分离和置信传播的运动水声通信  

Moving Underwater Acoustic Communication Based On Speex Framework Separation and Confidence Propagation

作　　者：冯向超 倪秀辉 巩小东 孟雍祥 梁俊燕 杜洋 FENG Xiangchao;NI Xiuhui;GONG Xiaodong;MENG Yongxiang;LIANG Junyan;DU Yang(Institute of Oceanographic Instrumentation,Qilu University of Technology(Shandong Academy of Sciences),Qingdao 266318,China;National Technical University of Ukraine“Igor Sikorsky Kyiv Polytechnic Institute”,Kyiv 03056,Ukraine;Shandong Technological Center of Oceanographic Instumentt Co.,Ltd,Qingdao 266005,China;School of Information and Control Engineering,Qingdao University of Technology,Qingdao 266525,China)

机构地区：[1]齐鲁工业大学(山东省科学院)海洋仪器仪表研究所,山东青岛266318 [2]乌克兰国立技术大学(基辅工学院),乌克兰基辅03056 [3]山东省海洋仪器仪表科技中心有限公司,山东青岛266005 [4]青岛理工大学信息与控制工程学院,山东青岛266525

出　　处：《探测与控制学报》2024年第3期127-135,共9页Journal of Detection & Control

基　　金：青岛市关键技术攻关及产业化示范类项目(22-3-3-hygg-8-hy);山东省重点研发计划项目(2022CXGC020409);山东省科学院海洋仪器仪表研究所培养基金项目(HYPY202110)。

摘　　要：针对水声语音信号受快变信道影响引起的传输可靠性差的问题,提出基于Speex信构分离和置信传播的运动水声通信技术,该技术由语音信号处理和数字通信两部分构成。在语音信号处理部分,基于Speex算法,将整帧分成多个子帧,利用合成分析技术对子帧进行编码,并将语音信号的结构信息与内容信息进行分离,只传输语音内容信息,使得语音信息的压缩比由7.60提升至8.42,提高了传输效率;利用接收参数计算激励信号,并通过合成滤波器重构语音信号。在数字通信部分,基于叠加训练(ST)方案,提高对时变水声信道的跟踪能力;基于置信传播,采用双向信道估计(BCE)算法,实现对每个短块的精准信道估计。将Speex算法、ST方案、BCE算法和Turbo均衡相结合,构建水声语音通信系统,提高水声语音信号传输的可靠性。通过仿真和海上试验验证了该技术的有效性。Aiming at the problem of poor transmission reliability of underwater acoustic voice signal,a moving underwater acoustic communication technology based on the Speex algorithm and confidence propagation was proposed.The technology consisted of two parts,compression and reconstruction of the voice signal and digital communication.In voice signal processing,the structure of the speech signal was separated from the information.The Speex algorithm was used to divide the whole frame into multiple subframes,and the subframes were encoded by the synthesis analysis technique.The structure of the speech signal was separated from the information,and only the voice information was transmitted.The compression ratio of the voice information was increased from 7.60 to 8.42,which improved the transmission efficiency.The excitation signal was calculated by the receiving parameters,and the speech signal was reconstructed by the synthetic filter.In digital communication,the superimposed training(ST)scheme was adopted to improve the tracking ability of time-varying underwater acoustic channels.Based on confidence propagation,the bidirectional channel estimation(BCE)algorithm was used to achieve accurate channel estimation for each segment.Moreover,the Speex algorithm,ST scheme,BCE algorithm,and Turbo equalization were combined to build an underwater acoustic voice communication system,which improved the reliability of underwater acoustic voice signal transmission.Finally,the effectiveness of the proposed technique was verified by computer simulation and experimental results with horizontal distance of 5.5 km between the transmitter and receiver in Jiaozhou Bay.

关 键 词：Speex算法 叠加训练(ST)方案 置信传播 双向信道估计(BCE)算法 TURBO均衡 

分 类 号：TN929.3[电子电信—通信与信息系统]