ultra-high-speed all-optical networking technologies for next generation networking mikio yagi,...
TRANSCRIPT
Ultra-high-speed all-optical networking technologies for next generation networking
Mikio Yagi, Shiro Ryu (1), and Shoichiro Asano (2)
1: Information and Communication Labs., Japan Telecom Co., Ltd.2: National Institute of Informatics
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks
September 29, 2004
2
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Agenda
• Future network features and applications
• Key technologies and issues for realization of all-optical network
• Our recent activities– Field trial 1 : Application of all-optical regeneration system – Field trial 2 : Application of automatic chromatic dispersion
compensator
• Conclusion
Future network features and applications
4
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Network applications for world-wide high-speed network
• GRID computing• Genome information analysis• High energy and nuclear fusion
research
• Space and astronomical science
• IT-Based Laboratory (ITBL)
• Storage area network (SAN)
5
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
What is needed for future network ?
Task
Result
• Communication style– Human to human– Human to computer– Computer to computer
• Bit-rate free• Protocol free• Network topology free
• High capacity• Short delay• High security• On demand
Global GRID computing
6
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Current network : IP-based Network “PRISM”
GSR
PRISM: Progressive Revolutionary Integration on Service Media
DWDM
IP based client
IP based client
10G POS ring
using MPLS
Customer router
Electrical processing equipments
7
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Future network: All-optical network
DWDMMesh
Network
IP routerPXCPXC
PXC
PXC
PXC
PXC
Photoniccrossconnect
Interwork
DWDM
Any client signal
Any client signal
GMPLS: Generalized Multi-Protocol Label Switching
All-optical processing equipments
8
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Features of all-optical network
High speed / High capacity
Short transmission delay time
High security
Protocol free
Bit-rate free
Topology free
On demand
These functions are essential for the future network applications.
Key technologies and issues for realization of the all-optical network
10
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Key technologies for the future network (1)
• Physical layer
• Control plane
• Others– Transport layer– Management– Service application
11
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Key technologies for the future network (2) : Physical layer
• Switching technologies on repeater node– Optical crossconnect (OXC)/Photonic crossconnect (PXC)
• High-speed Switching
– Link aggregation– Optical add/drop multiplexing (OADM)– Optical queuing
• All-optical signal processing technologies– All-optical regeneration
• 2R regeneration (regeneration and reshaping)• 3R regeneration (regeneration, reshaping, and retiming)
– Optical wavelength conversion– Compensation of fiber parameter effect (Chromatic dispersion /
Polarization-mode dispersion)
• Optical signal quality measurement technology
Fiber array
Micro mirror(MEMS mirror)
Fiber array
Optical Signal
Optical Signal
Micro mirror array
3 Dimension Micro Electro Mechanical Systems (3D MEMS)
1㎜
12
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Key technologies for the future network (3) : Physical layer
• Switching technologies on repeater node– Optical cross connect (OXC)/Photonic cross connect (PXC)
• High speed Switching
– Link aggregation– Optical add/drop multiplexing (OADM)– Optical queuing
• All-optical signal processing technologies– All-optical regeneration
• 2R regeneration (regeneration and reshaping)• 3R regeneration (regeneration, reshaping, and retiming)
– Optical wavelength conversion– Compensation of fiber parameter effect (Chromatic dispersion /
Polarization-mode dispersion)
• Optical signal quality measurement technology
13
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
What’s problem on physical layer ?
In the future all-optical network
– The route of the path changes dynamically
• Network protection/restoration
• Reconfiguration of peer-to-peer wavelength path service
Fiber parameters along the path are changed after reconfiguration.
14
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Fiber parameters cause signal degradation
Polarization-mode dispersion
(PMD)
40Gbit/s data signalreceiver
Com
pen
sation
of signal
distortion
Y
X
Y
X
XY
40Gbit/s data signaltransmitter
Chromatic dispersion(CD)
Signal-to-noise ratio (SNR)
15
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Compensation technologies on each effect
Polarization-mode dispersion
(PMD)
Y
X
Y
X
XY
Chromatic dispersion(CD)
Signal-to-noise ratio (SNR)
All-optical signal regeneration
• All-optical 2R regeneration
• All-optical 3R regeneration
All-optical signal regeneration
• All-optical 2R regeneration
• All-optical 3R regeneration
PMD compensatorPMD compensator CD compensatorCD compensator
16
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Key technologies for the future network (4) : Physical layer
• Switching technologies on repeater node– Optical crossconnect (OXC)/Photonic crossconnect (PXC)
• High-speed Switching– Link aggregation– Optical add/drop multiplexing (OADM)– Optical queuing
• All-optical signal processing technologies– All-optical regeneration
• 2R regeneration (regeneration and reshaping)• 3R regeneration (regeneration, reshaping, and retiming)
– Optical wavelength conversion– Compensation of fiber parameter effect (Chromatic dispersion /
Polarization-mode dispersion)
• Optical signal quality measurement technology
17
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Key technologies for the future network (5) : Control plane
MPLS
IP
+Label Label Path
GMPLS
λ（ Lambda） SONET/SDH
Etc.
Fiber
• Generalized MPLS (Multi-Protocol Label Switching)– Control and signaling mechanisms of MPLS label path have been extended in ord
er to apply those mechanisms to not only label paths, but also SONET/SDH paths, lambda paths and etc.
• MPLSMPLS is the set of extensions to OSPF, IS-IS, and RSVP to support the routing of paths (aka traffic engineering)
• MPMPSS is a concept that says the MPLS control plane can be leveraged to support routing of lambda paths
• GMPLSGMPLS is the realization of the MPS concept, created by extended MPLS to support non-packet non-packet paths (’s, time-slots, fibers)
18
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Key technologies for the future network (6)
• Physical layer
• Control plane
• Others– Transport layer– Management– Service application
There are a lot of issues to resolve for realization of the all-optical network.
Our recent activities
20
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Super SINET project
Super SINET is an ultrahigh-speed network intended to develop and promote Japanese academic researches by strengthening collaboration among leading academic research institutes.
http://www.sinet.ad.jp/english/super_sinet.html
•High energy and nuclear fusion
•Space and astronomical science
•Genome information analysis (bio-informatics)
• Supercomputer-interlocking distributed computing (GRID)
•Nanotechnology
21
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Our challenge
• For realization of future ultra-high-speed all-optical network – Physical layer
• Field trials
– Application of all-optical 2R regeneration system
– Application of automatic chromatic dispersion compensation system
– Control plane
– Service application
Field trial 1:
Application of all-optical 2R regeneration system
23
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
How can all-optical 2R regeneration be realized?
• 2R regeneration :– regeneration and reshaping
Amplified amplitude
Noise of level 1
Noise of level 0
Input signal
Input vs output characteristic of an optical device that has non-linear effect
24
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
How can all-optical 2R regeneration be realized? (Cont.)
• 2R regeneration :– regeneration and reshaping
Amplified amplitude
Noise suppression Noise of
level 1
Noise of level 0
IN
OUT
Input signal
Input vs output characteristic of an optical device that has non-linear effect
Output signal
An electro-absorption
modulator (EAM) has the effect
of noise suppression.
Opticaldevice
Input Output
Optical device
25
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Research background
• The optical signal quality is degraded by the loss of the OADM system.
• The OADM system causes the signal quality degradation for the through signal at the destination.
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
It is necessary to compensate for the degradation.
Optical add/drop multiplexing (OADM)
26
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Research background (Cont.)
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M2R regeneration
system
• The optical signal quality is degraded by the loss of the OADM system.
• The OADM system causes the signal quality degradation for the through signal at the destination.
It is necessary to compensate for the degradation.
27
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Research background (Cont.)
• This experiment– 40-Gbit/s 12-channel WDM field trial using an installed 320-
km-long fiber.– Applied OADM system with an all-optical 2R regeneration
system.
• This experiment– 40-Gbit/s 12-channel WDM field trial using an installed 320-
km-long fiber.– Applied OADM system with an all-optical 2R regeneration
system.
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
WD
M
2R regenerationsystem
28
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Location and cable for the 2R system field trial in Tokyo area
Fiber type : SMF
Shinjuku StationTokyo Station
Total fiber length : 80 km x 4 spans = 320 km
National Institute of Informatics (NII) Building
4.5-km-long installed fiber cable
11.7-km-long installed fiber cable
29
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Our transmission system
40-Gbit/s receiver Bit-error rate tester(Performance evaluation)
40-Gbit/s transmitter Repeater
30
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Performance evaluation : Q-factor
off
on
1 + 0
1 - 0Q =
1: ON level average value
1: ON level noise standard deviation
0: OFF level average value
0: OFF level noise standard deviation
Histogram
Q=20dB :: BER= 8×10-24
Q=17dB :: BER= 1×10-12
Q-factor ++ Transmission quality ++
31
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Performance evaluation cases
WD
M
WD
M1. Dropped at 160-km by the OADM ; “Dropped at 160km”
2. 320-km transmission without 2R; “320km w/o 2R”
3. 320-km transmission with 2R: “320km with 2R.”
1
2
32R regeneration
system
32
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Result of 320-km transmission with OADM / 2R system
WD
M
WD
M
1
2
32R regeneration
system
Dropped at 160-km
Q-factor: 18.8dB
320-km w/o 2R
Q-factor: 16.9dB
320-km with 2R
Q-factor: 17.7dB
1.9dB degradation
0.8dBimprovement
33
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Discussion
• OADM system with/without 2R regeneration system
– 0.8-dB improvement over 320-km transmission with 2R– Nearly the same as the quality of the signal dropped at 160-km.
• From a point of view of the system design,
– It is preferable that transmission characteristics of the express channel and the dropped channel are equal.
We have confirmed that the all-optical 2R system has the
possibility to realize such a condition in an OADM system.
Field trial 2:
Application of automatic chromatic dispersion compensator
35
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Influence of chromatic dispersion
40Gbit/s data signalReceiver
Com
pen
sation
of signal
qu
ality
40Gbit/s data signalTransmitter
• When the wavelength path is dynamically reconfigured, accumulated physical parameters including chromatic dispersion (CD) are changed.
– CD is one of the most important parameters for the system over 40 Gbit/s.
36
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Influence of chromatic dispersion (Cont.)
40Gbit/s data signalreceiver
Com
pen
sation
of signal
qu
ality
40Gbit/s data signaltransmitter
short Group delay largelong Group delay small Index of reflection
z
longInput
Output
z
Reflect point of input signal depends on wavelength. It causes the difference of group delay.
short
Tunable chromatic dispersion compensator - Chirped fiber Bragg grating (CFBG) -
37
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Automatic chromatic dispersion compensator
40Gbit/s data signalreceiver
Automatic chromatic dispersion
compensator
40Gbit/s data signaltransmitter
Signal quality monitoring(Q-factor, Bit-error rate)
Hill-climbing method
Tunable chromatic dispersion
compensator
Device controller
Input Output
38
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Performance evaluation
• Rerouting operation
− GMPLS signaling
− Operation of automatic chromatic dispersion compensator
39
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Sequence of path setup and operation of automatic chromatic dispersion compensator
GMPLS control plane
GMPLS Control plane
Data plane
40-Gbit/s receiver
-DEM
UX
Automaticchromatic dispersion
compensator-MU
X
PXC PXC
1. Path Setup Request
2. RSVP - PATH ( Path setup )3. RSVP - RESV
Service plane1. Service request
1. Service request
2. Path setup
4. Service in
40
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Sequence of path setup and operation of automatic chromatic dispersion compensator
GMPLS control plane
GMPLS Control plane
Data plane
40-Gbit/s receiver
-DEM
UX
-MU
X
PXC PXC
1. Path Setup Request
2. RSVP - PATH ( Path setup )3. RSVP - RESV
Service plane1. Service request
1. Service request
2. Path setup
4. Service in
Automaticchromatic dispersion
compensator
Automaticchromatic dispersion
compensator
41
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Experimental result
Variation of Q-factor in case of network protection operation
• Network protection operation by switching a line between Line1 and Line 2 at second span in every 10 minute.
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
20.0
0:00 0:10 0:20 0:30 0:40 0:50 1:00 1:10 1:20 1:30
Time [hour:min.]
Q-f
acto
r [d
B]
Line2 Line2 Line2 Line2
Line1 Line1 Line1 Line1
Recovery time : 40 sec
Efforts to improve the time for CD compensation- GMPLS multilayer integration –
43
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Make the CD compensation operation faster
40Gbit/s data signalreceiver
Automatic chromatic dispersion
compensator
40Gbit/s data signaltransmitter
Signal quality monitoring(Q-factor, Bit-error rate)
Hill-climbing method
Tunable chromatic dispersion
compensator
Device controller
Input OutputQuality measurement (Q-factor, BER) takes long time ( ~ 40 sec).
Measurement of CD makes path setup faster
Measurement of CD makes path setup faster
The multilayer integration among a GMPLS control plane, a measurement plane, and a data plane is essential.
44
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Sequence of the multilayer integration
GMPLS control plane
GMPLS Control plane
Data plane
40-Gbit/s receiver
Chromatic dispersion analyzer (receiver)
-DEM
UX
Chromatic dispersion
compensator
-MU
X
PXC
PXC
PXC
Chromatic dispersion analyzer
(transmitter)
Measurement plane
1. Path Setup Request
Data plane
2. RSVP - PATH ( Path setup )3. RSVP - RESV
Service plane1. Service request
1. Service request
2. Path setup
45
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
GMPLS Control plane
Data plane
40-Gbit/s receiver
-DEM
UX
Chromatic dispersion
compensator
-MU
X
PXC
PXC
PXC
Sequence of the multilayer integration
GMPLS control plane
Measurement plane
1. Path Setup Request
Data plane
2. RSVP - PATH ( Path setup )3. RSVP - RESV
4. Data plane setup5. CD measurement
Service plane1. Service request
Chromatic dispersion analyzer (receiver)
Chromatic dispersion analyzer
(transmitter)
Chromatic dispersion analyzer (receiver)
Chromatic dispersion analyzer
(transmitter)
46
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
GMPLS Control plane
Data plane
40-Gbit/s receiver
-DEM
UX
Chromatic dispersion
compensator
-MU
X
PXC
PXC
PXC
Sequence of the multilayer integration
GMPLS control plane
Measurement plane
1. Path Setup Request
Data plane
2. RSVP - PATH ( Path setup )3. RSVP - RESV
4. Data plane setup5. CD measurement
Service plane1. Service request
7. Set the value to CD compensator
8. Service in
6. Receive the measured CD value
Chromatic dispersion analyzer (receiver)
Chromatic dispersion analyzer
(transmitter)
Chromatic dispersion analyzer (receiver)
Chromatic dispersion
compensator
CD value
47
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Location and experimental setup of field trial in Kyusyu area
GMPLS control plane
Data plane
Yame Station
Fukuoka Station
Tosu Station
Route 3 (66-km NZDSF)
Route 2 (35-km NZDSF)
Route 1 (35-km SMF)
First span (66-km NZDSF) Second span
Third span(31-km NZDSF)
Route 3 (31-km NZDSF)
Kyushu University
40-Gbit/s receiver
Chromatic dispersion analyzer (receiver)
-DEM
UX
Chromatic dispersion
compensator
Chromatic dispersion analyzer
(transmitter)
-MU
X
PXC
PXC
PXC
Kyushu University
GMPLScontroller GMPLS
controllerGMPLScontroller
GMPLScontroller
Optical amplifier
SC-DCF
Optical amplifier
SC-DCF
SC-DCF
SC-DCF
SC-DCF
VOA Optical amplifier
GMPLScontroller
40-Gbit/s, 24-channel WDM transmitters
SC-DCF: Slope-compensating dispersion compensation fiber , PXC: Photonic cross-connect
48
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Error count v.s. time in rerouting operation
0
2
4
6
8
10
00:00 10:00 20:00 30:00Time [min : sec]
Err
or c
ount
R3 R1 R2 R3 R1 R2 R3 R1 R2 R3
R1 : Route 1R2 : Route 2R3 : Route 3
0
2
4
6
8
10
28:00 28:10 28:20 28:30Time [ min : sec]
Err
or c
ount
8.4 sec
Fig.a : Error count v.s. time in rerouting operation.
Fig.b : Error count v.s. time in rerouting operation (details).
49
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Discussion
• A field trial of GMPLS multilayer integration among a GMPLS control plane, a measurement plane, and a data plane for ensuring the quality of a 40-Gbit/s wavelength path is effective for the GMPLS all-optical rerouting
• The time for the start of the service after a fault was measured to be about 8.4 seconds.
• A field trial of GMPLS multilayer integration among a GMPLS control plane, a measurement plane, and a data plane for ensuring the quality of a 40-Gbit/s wavelength path is effective for the GMPLS all-optical rerouting
• The time for the start of the service after a fault was measured to be about 8.4 seconds.
Conclusion
51
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
What to do for the future network further ?
• Higher capacity switching
• Routing processing for large scale network
• Higher speed transmission system technologies (160 Gbit/s …..)
• Signal quality monitoring method based on all-optical processing
• Network security (data plane, control plane)
There are a lot of issues to resolve for realization of the all-optical network.
There are a lot of issues to resolve for realization of the all-optical network.
52
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Future network applications
• GRID computing• Genome information analysis• High energy and nuclear fusion
research
• Space and astronomical science
• IT-Based Laboratory (ITBL)
• Storage area network (SAN)
53
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
Thank you for your kind attention !
54
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.
55
Internet2 Fall 2004 Member Meeting Extreme Networking: Experimental Ultra-High-Speed Networks 29/Sep./2004
Proprietary of Japan Telecom Co., Ltd, and and National Institute of Informatics.