Diapositive 1

Report
Optimization of WDM system architecture
using 100G technology
Guillaume Crenn, Product Line Manager
Presentation Agenda
 100G Market & Technology overview
 100G benefit in WDM networks
 Potential issue and Workaround
 Next step to increase system capacity
Page 2
100G Market status
Explosive Growth in
Ethernet/IP Services/Traffic
Ethernet Services Revenue
Source: Ovum
CAGR: 2009-2016: 13.2%
Page 4
Explosive Growth in
Ethernet/IP Services/Traffic
IP Traffic Growing at a much faster rate than
Ethernet Services Revenue
Source: Cisco
CAGR: 2010-2015: 32%
Page 5
Explosive Growth in
Ethernet/IP Services/Traffic
Services at rates greater than 1G growing at an alarming rate
Market is driving demand for higher rate services at a lower cost
Services Greater than 1G
900000
800000
700000
600000
500000
Services Greater than 1G
400000
300000
200000
100000
0
2008 2009 2010 2011 2012 2013 2014 2015 2016
Source: Ovum
CAGR: 2008-2016: 35.8% 10G services Grew > 60% CAGR
Page 6
How are Service Providers
Reacting? (1/2)
Ethernet Services and IP Traffic Growth Result in Huge Growth in Global
Page 7
Bandwidth Demand
How are Service Providers
Reacting? (2/2)
40G & 100G Volumes
250,000
200,000
150,000
40G & 100G Volumes
100,000
50,000
0
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Source: Ovum
Service Providers are putting in bigger Pipes Today
to better utilize expensive fiber infrastructure
Page 8
100G technology overview
100G transmission technology
 10G transmission mainly uses NRZ modulation and Direct detection
 OOK-NRZ (On Off Keying – None Return to Zero) modulation format
 40G transmission: different type of modulation used by different
vendors
 No real 40G standardization
 Direct or coherent detection technology available
 100G: Standardization of the line interface
 This standardization has caused the technology to mature faster than 40G and
therefore will result in a higher adoption rate
 Same technology used by all players on the terrestrial 100G market
today
 DP or PM-QPSK modulation Format and Coherent receiver
 Optical design similar for all competitors
 Performance differentiation at DSP level
Page 10
100G transmission technology
 Routers are rapidly moving toward 100G interconnect to
accommodate the increase in IP traffic
 Native 100G client interfaces standardized and available
Physical Layer
100 Gigabit Ethernet
7m over Copper Cable
100GBase-CR10
100m over OM3 MMF
100GBASE-SR10
125m over OM4 MMF
100GBASE-SR10
10km over SMF
100GBASE-LR4
10km over SMF
100GBASE-LR10
40km over SMF
100GBASE-ER4
 Numerous component vendors have introduced 100G technology to
the market in the last 3-6 months causing the economics to improve
quickly
Page 11
100G Competition Status
 Field trial over Long haul routes achieved on operators live networks
since 2 to 3 years
 Few 100G channels added on live 10G and 40G networks
 First network deployment announced beginning of last year (2011 H1)
 Few Metro Networks already running at 100G
 1rst commercial route for 100G traffic: Paris-Frankfurt
 Massive 100G network deployment since 2011 H2 and 2012 H1
 Transponders (native 100G client ports) and Muxponder (aggregation of 10G
client ports in a 100G line available in all 100G players portfolio
Page 12
100G benefits in WDM networks
100G Network design rules
 Utilizes the same hardware as 10G and 40G (filters, amps, etc.)
 Compatible with all MUX/DMUXs, ROADMs, OADMs, already installed in your
network
 Works with both 50GHz and 100GHz channel plans
 Utilizes standard ITU Grid
 Required OSNR around 14 dB
 Compatible with the large majority of 10G and 40G installed systems
 Transmission over 3000 to 4000 kilometers
 ±50 000 ps/nm of Chromatic dispersion tolerance
 Much more tolerant than the 10G or 40G optical interfaces
 10G tolerance around 2000 ps/nm max
 PMD tolerance > 30ps (100ps of DGD)
 Much more tolerant than the 10G or 40G interfaces
 10G tolerance around 10 to 12 ps
Page 14
100G deployment: green field
 Green field deployment: new WDM network based on 100G only
 High capacity backbone network
 Capacity per fiber pair increases by a factor 10 compared to 10G design
 Granularity to lower bit rate client ports using Muxponder
 Client ports can be 100G, using transponders
 Client can also be 40G, 10G or lower by using Muxponder
 Infrastructure Cost reduction on network infrastructure due to dispersion
compensation removal
Page 15
Real Life Customer Scenario
ODC
ODC
ODC
ODC
ODC
3200 Km Network – 38 spans ranging from 66km to 150km
100G does not require dispersion compensation like 10G.
Eliminating Dispersion Compensation Saves 37%
(On commons: chassis, amps, dispersion, management cards)
This can be used to pay for sellable customer services
Deploying with 100G instead of 10G paid for
the first 100G Channel!
Page 16
100G deployment: Upgrade and
optimization of existing systems
100G TRP or MXP
100Gb/s
Additional wavelength
100G TRP or MXP
100Gb/s
Additional wavelength
10G/20Gb/s
Additional wavelength
Ekinops PM C1002/C2002
 Upgrading Existing Systems with Additional 100G capacity
 100G FEC allows good performance over existing line systems
 Same OSNR and power per channel range for existing 10G/40G
channels and new additional 100G channels
Page 17
100G Option for Expanding
Capacity
 100G performance: = to or better than most 10G
systems
 100G Alien waves can use existing 10G design rules with
potential bypassing of regens
Typical 10G
10-12 ps
~ 300-700 ps/nm
~ 2,500 km
16-17 dB
PMD
Dispersion
Distance
OSNR
Ekinops 100G
30 ps
50,000 ps/nm
> 3,000 km
14 dB w/DynaFEC
32 Channel System, 30 Channels Lit
POP-7
POP-1
12.7 km 109.4 km
3.2 dB
23.1 dB
Hut-1
105.1 km
21.4 dB
102km
48km
21.1 dB 10.1 dB
Hut-2
58.6 km
12.3 dB
Hut-5
Hut-3
POP-2
105.5km
21.5 dB
POP-3
100G
100.2km
20.4 dB
Hut-6
109.6km
23.9 dB
120.6km
20.6 dB
Hut-7
107.1km
21.5 dB
Hut-8
POP-4
ReGen, Add/Drop
107.5km
21.3 dB
105.3km
20.6dB
104.5km
20.3dB
Hut-9
Hut-10
POP-5
100.2km
19.5dB
100.4km
19.8dB
Hut-11
105.4km
20.5dB
Hut-12
94.5km
18.9dB
Hut-13
POP-6
Field Trial
102.8km
20.3dB
Hut-15
Hut-14
(10G ReGen)
98.4km
19.3dB
Hut-16
90.2km
17.7dB
Hut-17
85.3km
17.1dB
Hut-18
100G
100G
9-Spans, 751 km
64.6km
12.8dB
100G Express
14-Spans, 1,387 km
Extending to Full Route
Page 18
100G Option for Expanding
Capacity (Continued)
32 Channel System, 30 Channels Lit
POP-7
POP-1
12.7 km 109.4 km
3.2 dB
23.1 dB
Hut-1
105.1 km
21.4 dB
102km
48km
21.1 dB 10.1 dB
Hut-2
58.6 km
12.3 dB
Hut-5
Hut-3
POP-2
105.5km
21.5 dB
POP-3
100G
100.2km
20.4 dB
Hut-6
109.6km
23.9 dB
120.6km
20.6 dB
Hut-7
107.1km
21.5 dB
Hut-8
POP-4
ReGen, Add/Drop
107.5km
21.3 dB
105.3km
20.6dB
104.5km
20.3dB
Hut-9
Hut-10
POP-5
100.2km
19.5dB
100.4km
19.8dB
Hut-11
105.4km
20.5dB
Hut-12
94.5km
18.9dB
Hut-13
POP-6
102.8km
20.3dB
Hut-15
Hut-14
(10G ReGen)
98.4km
19.3dB
Hut-16
90.2km
17.7dB
Hut-17
85.3km
17.1dB
Hut-18
100G
100G Express
100G
9-Spans, 751 km
64.6km
12.8dB
14-Spans, 1,387 km
 Eliminate the need for equipment upgrades to existing 10G systems
Avoid:
 Cost of new filters, amps, DCM, etc.
 Cost of field tech’s deployment, planning, etc.
 Downtime, customer satisfaction / SLA issues
 And with a smaller form factor… Eliminate space & power issues
 1 RU: Rack chassis even in very tight spaces
 230W (4.8A): Run power cables to existing fuse panels
 No new power cables to BDP
Page 19
100G deployment: Upgrade and
optimization of existing systems
• No need to deploy a new WDM system if you are missing capacity
• Even If 10G is your primary service rate, you can get 10x 10G into
the same spectrum as 1X 10G previously
10 X 10G channels
1 X 100G channel
100G Muxponder
9 free ports available
for additional
deployment
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10 X 10G channels
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
10G
Page 20
100G Issues and workaround
100G potential issue
 Issue with Most 100G solutions
 Only available in large form factors not suitable for applications
with limited space
 Footprint can also be an issue in co-location environments
But: Some low footprint solution are available for limited space
 Cost: 100G Muxponder designed for Long haul today remain
more expensive the 10 X 10G TRP
But:
a)The cost savings on Infrastructure deployment can reduce the
Price Gap for green field deployment
b) 100G standardization with all component suppliers working on
the same technology, will lead to efficient price reduction on
100G market
Page 22
Ekinops Proprietary information
Next steps for WDM systems
Optical Evolution
Flex Coherent
 Flex Coherent (100G)
 Modulation format evolution for distance versus capacity
optimization
 DP BPSK, DP QPSK, DP 8QAM, DP 16QAM
 Trade-off of Spectral efficiency versus OSNR
Modulation
Distance
(km)
Total capacity
Application
DP BPSK
6000
4 Tb/s
Very Long Haul & Submarine
DP QPSK
3000
8 Tb/s
Long Haul
DP 8QAM
1500
12 Tb/s
High capacity route
DP 16QAM
750
16 Tb/s
High capacity route
Page 24
100G Optical Evolution
Wave Bounding /SuperChannel
 400Gb/s -1Tb/s
 100G use 50Gb/s DSP and 25GHz Optical receiver (available
today)
 1000G (1T) will require 500Gb/s DSP and 250GHz optical
receiver
 Such type of DSP will not be available before the next decade
 Wavebonding: the information is distributed over several
subcarriers spaced as closely to form a SuperChannel
(example 10 x 100Gb/s @50GHz :500GHz)
10 x 100G = 1Tb/s @ 500GHz
Page 25
100G Optical Evolution
Flex Grid
 Spectrum efficiency optimization
 No need to keep 50GHz between sub carriers 37,5GHz is enough
for DP-DPSK
10 x 100G = 1Tb/s @ 500GHz
10 x 100G = 1Tb/s @ 375GHz
Page 26
100G Optical Evolution
Flex Grid
 Each Sub-Carrier transporting a lower Bit Rate,
compatible with current 100G components
10 x 100G = 1Tb/s @
375GHz
SuperChannel #1
10 x 100G = 1Tb/s @
375GHz
SuperChannel #2
10 x 100G = 1Tb/s
@ 375GHz
SuperChannel #3
 Flex-Spectrum DWDM filtering is adopted enabling Multicarrier add and drop.
 Current ROADMs are already compatible with flexgrid
 Capacity to Add & drop full or sub-part of each super channelPage 27
100G Optical Evolution
Optical amplifier schemes
 Optimized Hybrid Amplifier (Erbium/Raman)
 For 100G /400G/1T transport.
 Optimization of the Bandwidth,
 Optimization of the noise flatness
 Optimization of the noise figure at nominal point
Page 28
Thank You

similar documents