一种对丢包具有鲁棒性的语音编码算法—iLBC  

Speech Coding Algorithm with Robustness to Packet Losses-iLBC

作　　者：郑国宏[1] 晋军[2] 程剑[2] 陈亮[1] 

机构地区：[1]解放军理工大学通信工程学院电子信息工程系,江苏南京210007 [2]解放军理工大学通信工程学院卫星通信系,江苏南京210007

出　　处：《军事通信技术》2008年第4期60-65,共6页Journal of Military Communications Technology

摘　　要：文中介绍了一种适用于网络的,对丢包具有鲁棒性的语音编码算法——iLBC。iLBC在编码时先进行LPC分析,然后对LPC残差信号采用自适应码本编码,这个自适应码本与传统的CELP算法中的自适应码本不同,没有利用语音帧之间的相关性,使得iLBC算法对丢包具有鲁棒性;解码时如果有语音包丢失,则采用丢包隐藏(PLC)算法。介绍了iLBC编解码算法的基本原理;重点分析了iLBC的自适应码本编码方式以及丢包隐藏算法(PLC);比较了在丢包率相同的情况下,iLBC和ITU-T G.729A的MOS分,结果表明,iLBC更适合于有丢包的网络。A speech coding algorithm with robustness to packet losses and suitable for network named iLBC was introduced.iLBC codec performs LPC analysis first and then codes the LPC residual signal by using a special adaptive codebook.In contrast to traditional CELP,iLBC applies an adaptive codebook without introducing data dependency across frame boundaries which results in iLBC codec's robustness with packet losses.If packet losses occurs when decoding,packet loss concealment(PLC) is used.The principle of iLBC coding algorithm was introduced with an emphasis on the adaptive codebook coding and PLC.Finally,the mean opinion score of iLBC and ITU-T G.729A at the same packet loss rates were compared.The results show that iLBC is more suitable for the network with packet losses.

关 键 词：语音编码 自适应码本 丢包隐藏 

分 类 号：TN91[电子电信—通信与信息系统]