Speech Coders – a VoIP perspective Roar Hagen CTO SIP/email: [email protected].
-
Upload
deon-meals -
Category
Documents
-
view
221 -
download
4
Transcript of Speech Coders – a VoIP perspective Roar Hagen CTO SIP/email: [email protected].
Agenda
• Speech Coders – a VoIP perspective• Demo• Q&A
QoS – (endpoints) status”A lot of talk, ... but not much work”
• Year after year the same story
• More then 3000 papers since 1984
• Limited ToS support at the end points
QoS – statusIndustry’s perspective
quality concerns
unproven technology
PSTN works fine
too busy to switch
not compelling economics
0 10 20 30 40 50 60
percentage of respondents
source: Forrester Research/AT&T (2000)
Background - Diverse Environment
PSTN Managednetwork
Wireless
PublicInternet
Managednetwork
Next generation codecs should Next generation codecs should address the needs of all applicationsaddress the needs of all applications[ ]
Packet Loss San Francisco – Hong Kong
0
10
20
30
40
50
60
70
80
90
100
6.00Tue22/1
18 6.00Wed23/1
18 6.00Thu24/1
18 6.00Fri
25/1
18 6.00Sat26/1
18 6.00Sun27/1
18 6.00Mon28/1
18 6.00Tue29/1
18 6.00Wed30/1
18 6.00Thu31/1
18
Time
Avg G.711
Worst 10s G.711
Avg G.729
Worst 10s G.729
Jitter San Francisco – Hong Kong
Homg Kong to China VoIP Call
Wireless VoIP – The Big Unknown ?
LAN
2G
/2.5
G
WLAN
Bluetooth
3GVehic
leW
alk
Fixed
Mbps1001010.1
Mobility
Approach
We need holistic view/approach for both
• Horizontal (end-to-end) perspective
• Vertical (top-down) perspective
Vertical (Top Down) Perspective
PhysicalPhysical
LinkLink MLPPP/FR/ATM AAL1MLPPP/FR/ATM AAL1
NetworkNetwork IP/WFQ/IP-precIP/WFQ/IP-prec
TransportTransport RTP/UDP/RSVPRTP/UDP/RSVP
SessionSession SIP/H.323SIP/H.323
PresentationPresentation Speech Codecs/…Speech Codecs/…
VoIP Aspirations
• IP innovation rather than PSTN replication
• New features and services through voice and data convergence
• End-to-end IP
• Better than PSTN sound quality
Current speech processing technology not designed for packet switched environments
* MEAN OPINION SCORECONGESTED NETWORKOVERPROVISIONED
NETWORK
MO
S
= U
SER E
XPER
IEN
CE
“FALL OFF A CLIFF” shape of curve forces
over provisioning
“FALL OFF A CLIFF” shape of curve forces
over provisioning
…congestion related VoIP QoS problems can be solved without over provisioning…
* MEAN OPINION SCORE
CONGESTED NETWORKOVERPROVISIONED NETWORK
MO
S
= U
SER E
XPER
IEN
CE
Operate AT and ABOVE congestion point without
customer knowing
Operate AT and ABOVE congestion point without
customer knowing
GIPS Ehanced G.711+GIPS NetEQ™
G.711+GIPS NetEQ™
G.711+ITU PLC
G.729A
G.711+No PLC
Matching PSTN QualityTelephony bandwidth speech test result
SOURCE LOCKHEED MARTIN GLOBAL TELECOMMUNICATION (COMSAT)
5
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
NETWORK CONDITION (% PACKET LOSS)
0% 5% 30%10% 15% 25%20%
MO
S
GIPS iPCM™-wb+GIPS NetEQ™-wb
G.722+GIPS NetEQ™-wbG.722.1
Source + no PLC
Better Than PSTN QualityWideband speech
0% 5% 10% 15% 20% 25%
5
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
MO
S
NETWORK CONDITION (% PACKET LOSS)
narrow band sound quality equal to
PSTN
wide band sound quality
Jitter Buffer/PLC Enhancements
Source: Lockheed Martin Global Telecommunications (COMSAT)
Delay gain with NetEQ™approx. 30-60ms compared to traditional jitter buffers
Adaptive jitter bufferJitter Fixed jitter buffer NetEQ™140
120
20
80
100
40
60
00 200 400 600 800 1000 1200 1400 1600 1800 2000
Packet number
Dela
y (
ms)
The NextGen Speech Codec Ideal
• Need one concept that will work for a long time– footprint importance
• Need to handle large diversity of transport network– low rate– high quality, high rate– packet loss– jitter– low delay
• Manageable IPR situation• Signal Robustness
– speech– music
• Suitable for variety of applications, e.g. IP video-conferencing
iLBC (internet Low Bitrate Codec)
• Speech sampled at 8 kHZ,• using a block-independent linear-predictive coding (LPC)
algorithm.• Bandwidth 13.867 kbps (52 bytes per 30 ms)• Frame size 30 ms (support for 20 ms in the next revision)• Complexity and memory requirements are similar to ITU
G.729A• Basic Quality is equal to or better than G.729. Packet loss
robustness is significantly better than G.729. • Packet loss concealment - Integrated example solution
MOS Results
1.5
2.0
2.5
3.0
3.5
4.0
0 5 10 15
Packet Loss [%]
MO
S
G.729A G.723.1 iLBC
Source: Dynastat Inc.
iLBC - IETF work
• IETF deliverables, submitted during February ‘02:– iLBC codec specification draft - experimental
standards track– iLBC RTP Payload Profile - regular
standards track (AVT)– Statement about IPRs in iLBC and its “freeware nature”
• MOS results submission to the AVT mailing list during March ‘02
Why iLBC !?
• Current low bit rate codecs: ITU G.729, G.723.1, GSM-EFR, and 3GPP-AMR were developed for circuit switched & wireless telephony and are all based on the CELP (Code Excited Linear Prediction) paradigm.
• CELP coders are stateful, they have memory, error propagation results from lost or delayed packets.
• iLBC treats every packet individually, making it suitable for packet communications.
More information
• Coming Soon - web site www.ilbcfreeware.org with:– Info about initiative– Info about codec– Latest iLBC IETF drafts (spec and payload format)– Latest iLBC float point Source code– FAQ list
• IETF drafts:– draft-andersen-ilbc-00.txt - codec spec (exper. stds track)– draft-duric-rtp-ilbc-00.txt - RTP payload profile (AVT group)
• Web site www.globalipsound.com • Free demo SIP client available, please request at:
SIP/email: [email protected]
Summary
• Current speech coding technology not suited for VoIP• VoIP opens possibilities
– Move quality exprience to the next level with wideband coders
• NGN will not be NGN unless we move step forward on all of its fields
• iLBC – internet Low Bit Rate Codec– Provide an open standard ”the Internet way” for coder
Demo