基于IP包拆分重组技术的混合语音压缩编码算法研究  

Research on hybird speech compression coding algorithm based on IP packet splitting and reassembling technology

作  者:李凌云 李肖克 陈奕钊 王国法[1] 王辉[1] Li Lingyun;Li Xiaoke;Chen Yizhao;Wang Guofa;Wang Hui(The 34th Research Institute of CETC,Guilin 541004,China)

机构地区:[1]中国电子科技集团公司第三十四研究所,广西桂林541004

出  处:《电子技术应用》2025年第2期70-74,共5页Application of Electronic Technique

摘  要:针对某特殊通信网业务系统中,在10 kb/s的窄带信道上传输1路标准G.729编码格式的VoIP语音数据的特殊通信场景,提出一种基于IP包拆分重组技术的混合语音压缩编码算法,将G.729压缩后的语音数据进行解压缩,再通过AMBE进行二次压缩,结合IP包拆分重组技术,保留语音数据中有效载荷,剔除多余开销数据,减小语音数据传输所需带宽。仿真实验验证了该方法的有效性,当G.729和AMBE的语音压缩编码速率分别为8 kb/s、2.4 kb/s,载荷长度为20 ms,IP包打包周期为8包时,实验表明无论在何种光路状态下,平均句子可懂度达85%以上,话音信号等级达3级以上,满足话音传输系统要求。Aiming at the special communication network service system,in order to transmit 1 channel of standard G.729 Voice over Internet Protocol(VoIP)voice data over 10 kb/s narrowband channel in the special communication scenario,a hybrid speech compression coding algorithm based on IP packet splitting and reassembling technology is proposed.The algorithm decomposes the voice data after G.729 compression,and then performs secondary compression through Advanced Multi-Band Excitation(AMBE).Combined with IP packet splitting and reassembly technology,the payload in the voice data is retained,the redundant overhead data is eliminated,and the bandwidth required for voice data transmission is reduced.The effectiveness of the method is verified by simulation experiment.The experiments show that when the speech compression coding rate of G.729 and AMBE is 8 kb/s and 2.4 kb/s respectively,the load length is 20 ms,and the IP packet packaging cycle is 8 packets,the average sentence intelligibility is above 85%and the voice signal level is above level 3 under any optical path state,which meets voice transmission system requirements.

关 键 词:语音压缩编码 G.729 AMBE IP包拆分重组 窄带通信 

分 类 号:TN912.3[电子电信—通信与信息系统]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象