摘要
在分布式模拟训练系统中,语音系统的设计通常采用专用硬件的手段实现,该方法不仅增加了系统设计的费用和复杂度,也增加了系统硬件组成和布局的难度.设计了一种分布式环境下的基于软件化的实时语音通信系统,利用DirectX对语音数据进行采集和回放,并以网络通信的方式对数据进行打包、传输和解包.同时,针对噪声等周围环境和无人说话时产生的冗余数据造成网络通信量增加的问题,采用离散傅里叶变换对语音信号在频域进行过滤,有效降低了系统的网络负担.该系统在某型模拟器中得到了较好的应用.
Audio communication using special hardware in distributed simulation not only increases the expense and complexity of system design,but also increases the difficulties of hardware composing and layout.Real-time audio communication was designed based on software in distributed environment.Audio data were collected and replayed by adopting DirectX and packed,transmitted and unpacked on the network.Aiming at redundant data produced in nobody speaking and noisy environment increasing the burden,audio signals were filtered in the frequency domain based on discrete Fourier transform（DFT）,the burden of the network was reduced effectively.The system has good effects in a certain project.
出处
《武汉工程大学学报》
CAS
2012年第5期60-63,共4页
Journal of Wuhan Institute of Technology
关键词
分布式仿真
语音通讯
网络
离散傅里叶变换
distributed simulation
audio communication
network
discrete Fourier transform（DFT）
