论文部分内容阅读
将相干水声通信技术和混合码激励线形预测语音编码技术相结合进行了研究,设计并实现了一个可应用于浅海水声信道的实时数字语音通信系统。系统针对语音编码数据的特点,根据语音数据对恢复话音重要程度的不同,采用分层的信道编码以减少总体冗余度提高系统传输效率,采用鲁棒的语音帧同步机制、引入缓存机制保证语音数据流的连续性以提高系统整体性能;采用基于信道估计的内嵌二阶锁相环分数间隔判决反馈均衡器以对抗水声信道多径的影响。对该系统进行的海试在6km和11km距离上成功地实现了较高质量(可分辨出不同说话人)的实时水下语音通信,海试结果表明该系统可以应用于半双工实时水下数字语音通信。
The coherent underwater acoustic communication technology and the hybrid code excitation linear predictive speech coding technology are combined to study and design and implement a real-time digital voice communication system which can be applied to shallow sea acoustic channel. According to the characteristics of voice encoded data, the system adopts layered channel coding to reduce the overall redundancy and improve the system transmission efficiency. Based on the different importance of voice data to the recovered voice, the system uses robust voice frame synchronization mechanism and introduces a caching mechanism to ensure voice Data stream continuity to improve the overall system performance; based on channel estimation embedded second-order PLL fractional interval decision feedback equalizer to counter the impact of underwater acoustic channel multipath. Sea trials on the system successfully achieved real-time underwater voice communications of higher quality (different speakers can be discerned) over distances of 6 km and 11 km, and sea trials showed that the system could be applied to half-duplex real-time underwater Digital voice communications.