自适应多速率语音编码技术应用于VOIP系统的研究
摘要
自适应多速率(AMR,Adaptive Multi Rate)语音编码是由3GPP(3rd Generation Partnership Proiect)制定的应用于第三代移动通信W—CDMA系统中的语音压缩编码。它以更加智能的方式解决信源和信道编码的速率分配问题,使得网络资源的配置和利用更加灵活和高效。它支持八种速率:12.2kb/s,10.2kb/s,7.95kb/s,7.4kb/s,6.70kb/s,5.90kb/s,5.15kb/s和4.75kb/s,此外,它还包括低速率的(1.80kb/s)背景噪声编码模式。
实际的语音编码的速率取决于信道的条件,与采用固定的编码速率的语音编码方式相比,AMR语音编码则可根据信道的传输状况来自适应地选择一种最佳编码模式(以比特率来区分)进行编码传输。
传统的VoIP系统都采用固定编码速率的语音编码器,这种固定速率的话音编码器不能够根据信道状况自适应的调整编码速率,因此在网络状况不好的条件下性能就会严重下降。通过分析VoIP系统和自适应多速率语音编码技术的特点,本论文提出了一种将自适应多速率语音编码技术应用于VoIP系统的方法,并通过仿真验证了在不同网络环境下的这种VoIP系统中的AMR技术的有效性。
在VoIP系统中实现自适应多速率语音编码后,本文进一步对AMR编码算法中舒适噪声生成、话音激活检测、丢帧隐藏机制三个模块分别进行了优化,系统仿真表明,这种优化后的AMR语音编码算法能够进一步提升VoIP系统的性能,将AMR编码器的编码时延降低20%。
最后我们得出结论,本文提出的AMR优化算法可以很好的与VoIP系统相结合,提供较好的话音质量。
Traditionally, fixed bit-rate speech coding is used in VoIP system. This technology can not adaptively switch its bit-rate due to network condition. So we select AMR speech coder.
The AMR speech coder consists of the multi-rate speech coder, a source controlled rate scheme including a voice activity detector and a comfort noise generation system, and an error concealment mechanism to combat the effects of transmission errors and lost packets. The multi-rate speech coder is a single integrated speech codec with eight source rates from 4.75 kbit/s to 12.2 kbit/s, and a low rate background noise encoding mode. The speech coder is capable of switching its bit-rate every 20 ms speech frame upon command. Compared with fixed bit-rate speech coder AMR can provide more flexibility.
This paper focuses on how to efficiently merge VoIP system and Adaptive Multi-Rate (AMR) technology. The general knowledge of VoIP system and AMR were introduced in the first chapter. Based on the analyses of these two technologies, we proposed a method to merge them into one system in order to overcome the drawback of fixed bit-rate voice coding in VoIP system. By simulation we proved that VoIP system based on AMR speech coder can achieve higher performance gain than that based on fixed bit-rate coder.
Besides, we do optimization on comfort noise generation system, voice active detector and error concealment mechanism to enhance the performance of the system. Simulation shows that the optimized AMR algorithm in VoIP system reduces the time latency to about 80% compared to standard AMR based VoIP system. Therefore, our proposed optimal AMR speech coder can work well in VoIP system and the performance gain is huge compared with fixed bit-rate speech coder.
实际的语音编码的速率取决于信道的条件,与采用固定的编码速率的语音编码方式相比,AMR语音编码则可根据信道的传输状况来自适应地选择一种最佳编码模式(以比特率来区分)进行编码传输。
传统的VoIP系统都采用固定编码速率的语音编码器,这种固定速率的话音编码器不能够根据信道状况自适应的调整编码速率,因此在网络状况不好的条件下性能就会严重下降。通过分析VoIP系统和自适应多速率语音编码技术的特点,本论文提出了一种将自适应多速率语音编码技术应用于VoIP系统的方法,并通过仿真验证了在不同网络环境下的这种VoIP系统中的AMR技术的有效性。
在VoIP系统中实现自适应多速率语音编码后,本文进一步对AMR编码算法中舒适噪声生成、话音激活检测、丢帧隐藏机制三个模块分别进行了优化,系统仿真表明,这种优化后的AMR语音编码算法能够进一步提升VoIP系统的性能,将AMR编码器的编码时延降低20%。
最后我们得出结论,本文提出的AMR优化算法可以很好的与VoIP系统相结合,提供较好的话音质量。
Traditionally, fixed bit-rate speech coding is used in VoIP system. This technology can not adaptively switch its bit-rate due to network condition. So we select AMR speech coder.
The AMR speech coder consists of the multi-rate speech coder, a source controlled rate scheme including a voice activity detector and a comfort noise generation system, and an error concealment mechanism to combat the effects of transmission errors and lost packets. The multi-rate speech coder is a single integrated speech codec with eight source rates from 4.75 kbit/s to 12.2 kbit/s, and a low rate background noise encoding mode. The speech coder is capable of switching its bit-rate every 20 ms speech frame upon command. Compared with fixed bit-rate speech coder AMR can provide more flexibility.
This paper focuses on how to efficiently merge VoIP system and Adaptive Multi-Rate (AMR) technology. The general knowledge of VoIP system and AMR were introduced in the first chapter. Based on the analyses of these two technologies, we proposed a method to merge them into one system in order to overcome the drawback of fixed bit-rate voice coding in VoIP system. By simulation we proved that VoIP system based on AMR speech coder can achieve higher performance gain than that based on fixed bit-rate coder.
Besides, we do optimization on comfort noise generation system, voice active detector and error concealment mechanism to enhance the performance of the system. Simulation shows that the optimized AMR algorithm in VoIP system reduces the time latency to about 80% compared to standard AMR based VoIP system. Therefore, our proposed optimal AMR speech coder can work well in VoIP system and the performance gain is huge compared with fixed bit-rate speech coder.
引文
[1]3GPP TS 26.071 V5.0.0,"AMR Speech Code General Description," June 2002.
[2]3GPP TS 26.073 V5.0.0,"ANSI-C Code for the Adaptive Multi Rate(AMR)Speech Code," June 2002.
[3]3GPP TS 26.915 V3.0.0(2000-03):"Echo Control For Speech and MultiMedia Services"
[4]3GPP,3G TR 26.975 V1.1.0(2000-01):"Codec:Performance Characterization of the AMR Speech Codec"
[5]3GPP TS 26.093 V4.0.0,Source Controlled Rate Operation(Rel.4),June 2002.
[6]John G.Proakis,Digital Communications.Third Edition.McGraw-Hill.Inc,1995
[7]3GPP TS 26.093 V4.0.0,"AMR Speech Code about Source Controlled Rate Operation," June 2002.
[8]Mullner R.,Ball C.F.,Ivanov K.,Hartmann D.,Winkler H.,"AMR-wideband:Enjoying superior voice quality at full coverage and competitive capacity in GERAN networks," IEEE Vehicular Technology Conference,2005,p 2320-2324
[9]Wei Bo,Dong Hui,Gibson,Jerry D.,"Application of NB/WB AMR speech codecs in the 30-kHz TDMA system," IEEE Transactions on Wireless Communications,November,2004,p 1897-1901
[10]Guo Yi-Feng,Kuo Geng-Sheng,"A new defined lower bit rate AMR mode in GSM and WCDMA networks," IEEE Vehicular Technology Conference,2007,p 725-729
[11]Barreto,Andre Noll,Pirhonen,Riku,"Capacity increase in GSM networks using source-adaptive AMR,"IEEE Vehicular Technology Conference,2006,p 553-557
[12]Dong,H.,Gersho,A.;Gibson,J.D.;Cuperman,V.,"A multiple description speech coder based on AMR-WB for mobile ad hoc networks," ICASSP,IEEE International Conference on Acoustics,2004,p Ⅰ277-I280
[13]Werner,Marc,Junge,Thomas,Vary,Peter,"Quality control for AMR speech channels in GSM networks," ICASSP,IEEE International Conference on Acoustics,2004,p Ⅲ1076-Ⅲ1079
[14]Harri Holma,Juan Melero,Janne Vainio,Timo Halonen and Jail Makinen,"Performance of Adaptive Multirate(AMR)Voice in GSM and WCDMA,"P.0.Box 301,FIN-00045 Nokia Group,Finland
[15]R.Mullner,C.F.Ball,K.Ivanov,H.Winkler,R.Perl,and K.Kremnitzer,"Exploiting AMR-WB Audio Bandwidth Extension for Quality and Capacity Increase"
[16]Jirvinm.K."Standardization of the Adaptive Mulcirate Codec," Euripco.
[17]Haloncn,T.,Melero,1.and Romero,J."GSM,GPRS and EDGE Performance:Evolution Toward 3GNMTS," Wiley & Sons,2002.
[18]TS 26.101.Mandatory speech codec speech processing functions;AMR speech frame structure,3rd Generation Partnership Project(3GPP);TSG2SA Codec working group[S].
[19]GSM Recommendation 05.03.Digital cellular telecommunications system;channel coding,ETSI[S].
[20]Wilson W S Wong,Elvino S Sousa."Power assignment strategies and traffic control for wireless multmedia DS2CDMA systems[A]," Proc.Of IEEE International Symposium on Personal,Indoor and Mobile Radio Communications,PIMRC[C].Taipei:IEEE,1996.
[21]Q.Shen,A.W.Krzymien."Power Assignment in CDMA personal communication systems with integrated voice/data traffic[A]," conference Record/ IEEE Global Telecommunications Conference[C].London:IEEE,1996.
[22]K.K Paliwal and B.S.Atal."Efficient vector quantization of L PC parameters at 24 bitsPframe[J]," IEEE Trans.Speech and Audio Processing,1993,1(1):3-14.
[23]K.J3/4r vinen,J.Vainio,P.Kapanen,T.Honkanen,P.Haavisto,R.Salami,C.Laflamme,and J.- P.Adoul."GSM enhanced full rate speech codec [A]," in Proc.IC2 ASSP' 97[C],1997:771-774.
[24]T.Honkanen,J.Vainio,K.J3/4rvinen,P.Haavisto,R.Salami,C.Laflamme,and J.- P.Adoul."Enhanced full rate speech codec for IS-136 digital cellular system[A]," in Proc.ICASSP' 97[C],1997:731-734.
[25]R.Hagen,E.Ekudden,B.Johansson,and W.B.Klei2 jn.."Removal of sparse - excitation artifacts in CEL P[A]," in Proc.ICASSP' 98[C],1998:Ⅰ- 145-Ⅰ-148.
[26]M.R.Schroeder,B.S.Atal."Code - Excited Linear Pre2 diction(CEL P):High quality speech at very low bit rates[A]," in Proc.ICASSP' 85[C],1985.pp.937-940.
[27]F.Itakura."Line spectral representation of linear predic2 tire coefficients of speech signals[J]," J.Acoust.Soc.Amer.,vol.57,Supplement no.1,S35,1975.
[28]B.S.Atal,R.V.Cox,P.Kroon."Spectral Quantization and Interpolation for CEL P Coders[A]," ICASSP[C],1989,pp.69-721
[29]3GPP TS 26.090 V5.0.0,"AMR Speech Codec;Transcoding functions,"Jun.2002.
[30]C.Olivier,A.Magnus and S.Krister,"Capacity and Speech Quality Aspects Using Adaptive Multi-Rate(AMR)," Proc.of IEEE PIMRC 1998,vol.3,pp.1535 - 1539,Sept.1998.
[31]G.Helmer,A.Garavaglia,G.Knopp and W.Koch "Capacity Optimization for Adaptive Multi-rate in UMTS Systems," Proc.of IEEE VTC 2004 Fall,vol.7,pp.26- 29,Setp.2004.
[32]H.Holma,J.Melero,J.Vainio,T.Halonen and J.Makinen,"Performance of Adaptive Multirate(AMR)Voice in GSM and WCDMA," Proc.of IEEE VTC 2003 Spring,vol.5,pp.22-25,April 2003.
[2]3GPP TS 26.073 V5.0.0,"ANSI-C Code for the Adaptive Multi Rate(AMR)Speech Code," June 2002.
[3]3GPP TS 26.915 V3.0.0(2000-03):"Echo Control For Speech and MultiMedia Services"
[4]3GPP,3G TR 26.975 V1.1.0(2000-01):"Codec:Performance Characterization of the AMR Speech Codec"
[5]3GPP TS 26.093 V4.0.0,Source Controlled Rate Operation(Rel.4),June 2002.
[6]John G.Proakis,Digital Communications.Third Edition.McGraw-Hill.Inc,1995
[7]3GPP TS 26.093 V4.0.0,"AMR Speech Code about Source Controlled Rate Operation," June 2002.
[8]Mullner R.,Ball C.F.,Ivanov K.,Hartmann D.,Winkler H.,"AMR-wideband:Enjoying superior voice quality at full coverage and competitive capacity in GERAN networks," IEEE Vehicular Technology Conference,2005,p 2320-2324
[9]Wei Bo,Dong Hui,Gibson,Jerry D.,"Application of NB/WB AMR speech codecs in the 30-kHz TDMA system," IEEE Transactions on Wireless Communications,November,2004,p 1897-1901
[10]Guo Yi-Feng,Kuo Geng-Sheng,"A new defined lower bit rate AMR mode in GSM and WCDMA networks," IEEE Vehicular Technology Conference,2007,p 725-729
[11]Barreto,Andre Noll,Pirhonen,Riku,"Capacity increase in GSM networks using source-adaptive AMR,"IEEE Vehicular Technology Conference,2006,p 553-557
[12]Dong,H.,Gersho,A.;Gibson,J.D.;Cuperman,V.,"A multiple description speech coder based on AMR-WB for mobile ad hoc networks," ICASSP,IEEE International Conference on Acoustics,2004,p Ⅰ277-I280
[13]Werner,Marc,Junge,Thomas,Vary,Peter,"Quality control for AMR speech channels in GSM networks," ICASSP,IEEE International Conference on Acoustics,2004,p Ⅲ1076-Ⅲ1079
[14]Harri Holma,Juan Melero,Janne Vainio,Timo Halonen and Jail Makinen,"Performance of Adaptive Multirate(AMR)Voice in GSM and WCDMA,"P.0.Box 301,FIN-00045 Nokia Group,Finland
[15]R.Mullner,C.F.Ball,K.Ivanov,H.Winkler,R.Perl,and K.Kremnitzer,"Exploiting AMR-WB Audio Bandwidth Extension for Quality and Capacity Increase"
[16]Jirvinm.K."Standardization of the Adaptive Mulcirate Codec," Euripco.
[17]Haloncn,T.,Melero,1.and Romero,J."GSM,GPRS and EDGE Performance:Evolution Toward 3GNMTS," Wiley & Sons,2002.
[18]TS 26.101.Mandatory speech codec speech processing functions;AMR speech frame structure,3rd Generation Partnership Project(3GPP);TSG2SA Codec working group[S].
[19]GSM Recommendation 05.03.Digital cellular telecommunications system;channel coding,ETSI[S].
[20]Wilson W S Wong,Elvino S Sousa."Power assignment strategies and traffic control for wireless multmedia DS2CDMA systems[A]," Proc.Of IEEE International Symposium on Personal,Indoor and Mobile Radio Communications,PIMRC[C].Taipei:IEEE,1996.
[21]Q.Shen,A.W.Krzymien."Power Assignment in CDMA personal communication systems with integrated voice/data traffic[A]," conference Record/ IEEE Global Telecommunications Conference[C].London:IEEE,1996.
[22]K.K Paliwal and B.S.Atal."Efficient vector quantization of L PC parameters at 24 bitsPframe[J]," IEEE Trans.Speech and Audio Processing,1993,1(1):3-14.
[23]K.J3/4r vinen,J.Vainio,P.Kapanen,T.Honkanen,P.Haavisto,R.Salami,C.Laflamme,and J.- P.Adoul."GSM enhanced full rate speech codec [A]," in Proc.IC2 ASSP' 97[C],1997:771-774.
[24]T.Honkanen,J.Vainio,K.J3/4rvinen,P.Haavisto,R.Salami,C.Laflamme,and J.- P.Adoul."Enhanced full rate speech codec for IS-136 digital cellular system[A]," in Proc.ICASSP' 97[C],1997:731-734.
[25]R.Hagen,E.Ekudden,B.Johansson,and W.B.Klei2 jn.."Removal of sparse - excitation artifacts in CEL P[A]," in Proc.ICASSP' 98[C],1998:Ⅰ- 145-Ⅰ-148.
[26]M.R.Schroeder,B.S.Atal."Code - Excited Linear Pre2 diction(CEL P):High quality speech at very low bit rates[A]," in Proc.ICASSP' 85[C],1985.pp.937-940.
[27]F.Itakura."Line spectral representation of linear predic2 tire coefficients of speech signals[J]," J.Acoust.Soc.Amer.,vol.57,Supplement no.1,S35,1975.
[28]B.S.Atal,R.V.Cox,P.Kroon."Spectral Quantization and Interpolation for CEL P Coders[A]," ICASSP[C],1989,pp.69-721
[29]3GPP TS 26.090 V5.0.0,"AMR Speech Codec;Transcoding functions,"Jun.2002.
[30]C.Olivier,A.Magnus and S.Krister,"Capacity and Speech Quality Aspects Using Adaptive Multi-Rate(AMR)," Proc.of IEEE PIMRC 1998,vol.3,pp.1535 - 1539,Sept.1998.
[31]G.Helmer,A.Garavaglia,G.Knopp and W.Koch "Capacity Optimization for Adaptive Multi-rate in UMTS Systems," Proc.of IEEE VTC 2004 Fall,vol.7,pp.26- 29,Setp.2004.
[32]H.Holma,J.Melero,J.Vainio,T.Halonen and J.Makinen,"Performance of Adaptive Multirate(AMR)Voice in GSM and WCDMA," Proc.of IEEE VTC 2003 Spring,vol.5,pp.22-25,April 2003.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |