Milestone-Proposal:First Real-Time Speech Communication on Packet Networks: Difference between revisions

From IEEE Milestones Wiki
(Article updated via HTTP request)
(Article updated via HTTP request)
Line 1: Line 1:
{{ProposalEdit|a1=First Real-Time Speech Communication on Packet Networks|a2a=MIT Lincoln Laboratory, Lexington, MA|a2b=Boston Section|a3=* 1974: First real-time two-way packet speech, on ARPAnet – MITLL (Lexington, MA) and USC/ISI (Marina del Rey, CA) in August with 9.6  kbps CVSD (Continuously Variable Slope Delta Modulation); MITLL and Culler-Harrison (Goleta, CA) in December with 3.5 kbps LPC (Linear Predictive Coding) * 1976: First packet speech conferencing with 3.5 kbps LPC on ARPAnet among MITLL, CHI, USC/ISI * 1979: Real-time 2.4 kbps LPC conferencing using ARPAnet and Atlantic SATNET, among MITLL, USC/ISI, UCL (London), NDRE (Norway), BBN (Cambridge, MA) * 1982: Internet packet speech and conferencing including Wideband SATNET, Packet Radio Net. Local Area Ethernets, and telephone network interfaces – participants were MITLL, USC/ISI, and SRI|a4=This pioneering work on speech in packet networks developed and demonstrated systems which were forerunners of the voice-over internet protocol (VoIP) systems that are now so widely in use. The real-time voice work included development of a new Network Voice Protocol (NVP), because the packet and reliability constraints of the available Transmission Control Protocol (TCP) implementation made it unsuitable for real-time communication. This protocol development was an immediate forerunner of the separation of TCP and IP, so that the real-time packet speech work played a large role in the development of the protocols, which are still in wide use today. The technology and protocols for real-time speech over packet networks also enabled other real-time internet applications such as packet video, so that now systems like Skype enable real-time voice and video at home and in offices for extremely large number of people; this is a capability of significant social impact.
{{ProposalEdit|a1=First Real-Time Speech Communication on Packet Networks|a2a=MIT Lincoln Laboratory, Lexington, MA|a2b=Boston Section|a3=* 1974: First real-time two-way packet speech, on ARPAnet – MITLL (Lexington, MA) and USC/ISI (Marina del Rey, CA) in August with 9.6  kbps CVSD (Continuously Variable Slope Delta Modulation); MITLL and Culler-Harrison (Goleta, CA) in December with 3.5 kbps LPC (Linear Predictive Coding) * 1976: First packet speech conferencing with 3.5 kbps LPC on ARPAnet among MITLL, CHI, USC/ISI * 1979: Real-time 2.4 kbps LPC conferencing using ARPAnet and Atlantic SATNET, among MITLL, USC/ISI, UCL (London), NDRE (Norway), BBN (Cambridge, MA) * 1982: Internet packet speech and conferencing including Wideband SATNET, Packet Radio Net. Local Area Ethernets, and telephone network interfaces – participants were MITLL, USC/ISI, and SRI|a4=This pioneering work on speech in packet networks developed and demonstrated systems which were forerunners of the voice-over internet protocol (VoIP) systems that are now so widely in use. The real-time voice work included development of a new Network Voice Protocol (NVP), because the packet and reliability constraints of the available Transmission Control Protocol (TCP) implementation made it unsuitable for real-time communication. This protocol development was an immediate forerunner of the separation of TCP and IP, so that the real-time packet speech work played a large role in the development of the protocols, which are still in wide use today. The technology and protocols for real-time speech over packet networks also enabled other real-time internet applications such as packet video, so that now systems like Skype enable real-time voice and video at home and in offices for extremely large number of people; this is a capability of significant social impact.
References
References
Clifford J. Weinstein and James W. Forgie, “Experience with speech communication in packet networks,” IEEE Journal on Selected Areas in Communications, Vol. 1, No. 6, December 1983.
Clifford J. Weinstein and James W. Forgie, “Experience with speech communication in packet networks,” IEEE Journal on Selected Areas in Communications, Vol. 1, No. 6, December 1983.
Line 8: Line 9:
The narrowband (2 – 5 kbps) digital vocoder algorithms were needed due to the limited ARPAnet link rates;  these algorithms did not run in real time, so they had to be modified and adapted to run on the new DSP systems which were just emerging.
The narrowband (2 – 5 kbps) digital vocoder algorithms were needed due to the limited ARPAnet link rates;  these algorithms did not run in real time, so they had to be modified and adapted to run on the new DSP systems which were just emerging.


Interoperability needed to be established among separate implementations of speech algorithms and protocols running at different sites, on different host computers and DSPs, over different types of networks, and with geographic separation across the US and between US and Europe. |a7=MIT Lincoln Laboratory conducted  pioneering R&D and experiments in packet speech, collaborating with the other organizations noted above, including USC/ISI, SRI, CHI, BBN, and others. Lincoln was a direct participant in all the milestone packet speech experiments cited above, spanning 1974-1982. In addition, in 1971, Lincoln conducted the earliest packet speech related experiments on the ARPAnet; those experiments showed the feasibility of packet speech transmission; real-time 2-way communication over the ARPAnet between Lincoln and CHI and between Lincoln and USC/ISI followed in 1974.|a8=Yes|a9=MIT Lincoln Laboratory is a secure facility. The main entrance and lobby area is open to the public, and the Laboratory often opens its auditorium to outside events including Boston Section IEEE meetings, IEEE Life Fellow meetings, Science on Saturdays for K-12, and concerts.|a10=Massachusetts Institute of Technology|a11=Yes|a12=IEEE Signal Processing Society Speech Technical Committee
Interoperability needed to be established among separate implementations of speech algorithms and protocols running at different sites, on different host computers and DSPs, over different types of networks, and with geographic separation across the US and between US and Europe.|a7=MIT Lincoln Laboratory conducted  pioneering R&D and experiments in packet speech, collaborating with the other organizations noted above, including USC/ISI, SRI, CHI, BBN, and others. Lincoln was a direct participant in all the milestone packet speech experiments cited above, spanning 1974-1982. In addition, in 1971, Lincoln conducted the earliest packet speech related experiments on the ARPAnet; those experiments showed the feasibility of packet speech transmission; real-time 2-way communication over the ARPAnet between Lincoln and CHI and between Lincoln and USC/ISI followed in 1974.|a8=Yes|a9=MIT Lincoln Laboratory is a secure facility. The main entrance and lobby area is open to the public, and the Laboratory often opens its auditorium to outside events including Boston Section IEEE meetings, IEEE Life Fellow meetings, Science on Saturdays for K-12, and concerts.|a10=Massachusetts Institute of Technology|a11=Yes|a12=IEEE Signal Processing Society Speech Technical Committee


Dr. Steve Young, Chairman, sjy_at_eng.cam.ac.uk|a13name=|a13section=|a13position=|a13email=|a14name=|a14ou=|a14position=|a14email=|a15Aname=|a15Aemail=|a15Aname2=|a15Aemail2=|a15Bname=|a15Bemail=|a15Bname2=|a15Bemail2=|a15Cname=|a15Ctitle=|a15Corg=|a15Caddress=|a15Cphone=|a15Cemail=}}
Dr. Steve Young, Chairman, sjy_at_eng.cam.ac.uk|a13name=Bruce Hecht|a13section=Boston|a13position=Chair & Chapter Coordinator|a13email=bruce.hecht@analog.com|a14name=|a14ou=|a14position=|a14email=|a15Aname=|a15Aemail=|a15Aname2=|a15Aemail2=|a15Bname=|a15Bemail=|a15Bname2=|a15Bemail2=|a15Cname=|a15Ctitle=|a15Corg=|a15Caddress=|a15Cphone=|a15Cemail=}}

Revision as of 13:49, 1 November 2010