1 Optical Transport Networks : motivation, challenges and possibilities Optical Transport Networks: motivation, challenges and possibilities Huaiyuan Ma PhD Trial Lecture Feb 26th, 2015 NTNU, Trondheim
1Optical Transport Networks : motivation, challenges and possibilities
Optical Transport Networks: motivation, challenges and possibilities
Huaiyuan MaPhD Trial Lecture
Feb 26th, 2015NTNU, Trondheim
2Optical Transport Networks : motivation, challenges and possibilities
Outline
Why OTN ?
Challenges
Potential solutions
3Optical Transport Networks : motivation, challenges and possibilities
Outline
Why OTN ?
Challenges
Potential solutions
4Optical Transport Networks : motivation, challenges and possibilities
Focus
We focus large scale OTN, nation wide, backbone
1st generation (SDH) → 2nd generation OTN
5Optical Transport Networks : motivation, challenges and possibilities
Driving force for OTN evolution
From Ref [7]
6Optical Transport Networks : motivation, challenges and possibilities
Driving force for OTN evolution
From Ref [3]
7Optical Transport Networks : motivation, challenges and possibilities
Why OTN ?
1st Generation: SDH • Based on TDM, 8000 frames per secondBased on TDM, 8000 frames per second
• Hierarchical Multiplexing structureHierarchical Multiplexing structure
– Multi-level multiplexing based transmission rateMulti-level multiplexing based transmission rate
From Ref [4]
8Optical Transport Networks : motivation, challenges and possibilities
Cons:Cons:– Cannot provide high and dynamic data transmission rateCannot provide high and dynamic data transmission rate
→ → Not statistic multiplexingNot statistic multiplexing
– Low channel utilizationLow channel utilization
– → → Fixed bandwidth allocationFixed bandwidth allocation
– Lack of end-to-end monitoringLack of end-to-end monitoring
SDH
Pros:Pros: – – Good OAMGood OAM
9Optical Transport Networks : motivation, challenges and possibilities
Huge bandwidth resource: DWDM
Large bandwidthLarge bandwidth 8 ~ 160 lambdas on a single fiber8 ~ 160 lambdas on a single fiber
10Optical Transport Networks : motivation, challenges and possibilities
DWDMPros:Pros:
• Large bandwidthLarge bandwidth
• Transparent data transmissionTransparent data transmission• Low energy consumptionLow energy consumption
– Green networkGreen network
Cons:Cons:• A point-to-point system, not a network A point-to-point system, not a network
– Lack of lambda switch capabilityLack of lambda switch capability
• Low lambda ( wavelength ) utilization, some unused lambdasLow lambda ( wavelength ) utilization, some unused lambdas
– Subjected to wavelength continuity constraint Subjected to wavelength continuity constraint
• Fiber non-linear effect Fiber non-linear effect
– Higher light powerHigher light power
11Optical Transport Networks : motivation, challenges and possibilities
The 2nd Generation: OTN
SDH + lambda management
+ = OTN
DWDM + OADM/OXC
SDHSDH : good OAM, network protection, … DWDMDWDM: large bandwidth ApproachApproach: improve the bandwidth utilization of DWDM
by introducing OADM/OXC to form optical network and enhance SDH by adding wavelength management (GMPLS)
12Optical Transport Networks : motivation, challenges and possibilities
OTN Hierarchy
From Ref [4]
13Optical Transport Networks : motivation, challenges and possibilities
OTN TCM (Tandem Connection Monitoring)
From Ref [4]
14Optical Transport Networks : motivation, challenges and possibilities
The 2nd Generation: OTN
Stronger Forward Error Correction (FEC) More Levels of Tandem Connection Monitoring (TCM) Transparent Transport of Client Signals
digital wrapper Switching scalability
15Optical Transport Networks : motivation, challenges and possibilities
Outline
Why OTN ?
Challenges
Potential solutions
16Optical Transport Networks : motivation, challenges and possibilities
Alarming
From Science 2010[6]
Capacity (bits/s) growth trend
17Optical Transport Networks : motivation, challenges and possibilities
Some observations: Amdahl's rule
Exponential growth (i) of data network traffic in the US (red circles) and (ii) of the (average) Exponential growth (i) of data network traffic in the US (red circles) and (ii) of the (average) processing power of the top 500 supercomputers. (Flop: floating point operation per second). processing power of the top 500 supercomputers. (Flop: floating point operation per second).
Formula: 10*log(1.6)=2dB for 60% growth rate. Formula: 10*log(1.6)=2dB for 60% growth rate.
From Ref [5]
18Optical Transport Networks : motivation, challenges and possibilities
New Coding Techniques On/Off Keying (OOK) + direction detectionOn/Off Keying (OOK) + direction detection Quadrature Phase Shift Keying (QPSK) + coherent detectionQuadrature Phase Shift Keying (QPSK) + coherent detection 16-ary Quadrature Amplitude Modulation (16-QAM) + coherent detection16-ary Quadrature Amplitude Modulation (16-QAM) + coherent detection Polarization-Division Multiplexing (PDM): double channel capacityPolarization-Division Multiplexing (PDM): double channel capacity
From Ref [1]
19Optical Transport Networks : motivation, challenges and possibilities
Some warnings
(a) Spectral efficiency evolution of research experiments.(b) Trade-off between (single-(a) Spectral efficiency evolution of research experiments.(b) Trade-off between (single-polarization) spectral efficiency and SNR per bit in the linear regime for QAM (Quadrature polarization) spectral efficiency and SNR per bit in the linear regime for QAM (Quadrature
Amplitude Modulation) formats, assuming hard-decision forward error correction (FEC) with Amplitude Modulation) formats, assuming hard-decision forward error correction (FEC) with 7% coding overhead. 7% coding overhead.
From Ref [5]
32%/year << 86%/year, new coding techniques already push the spectral efficiency to the 32%/year << 86%/year, new coding techniques already push the spectral efficiency to the limit limit
20Optical Transport Networks : motivation, challenges and possibilities
Challenges Channel capacity growth rate contributed from higher-order coding techniques almost Channel capacity growth rate contributed from higher-order coding techniques almost approaches the Shannon limit.approaches the Shannon limit. Higher spectral efficiency requires higher SNR ratio.Higher spectral efficiency requires higher SNR ratio.
Higher light power causes more fiber non-linear effects.Higher light power causes more fiber non-linear effects. Limits the signal transmission distances.Limits the signal transmission distances.
We need look for other ways to address the network capacity issue caused by data We need look for other ways to address the network capacity issue caused by data traffic growthtraffic growth
21Optical Transport Networks : motivation, challenges and possibilities
Outline
Why OTN ?
Challenges
Potential solutions
22Optical Transport Networks : motivation, challenges and possibilities
Potential Solutions
Spatial Multiplexing ( SM ) Fiber bundles Multi-core fibers
Mode Division Multiplexing (MDM) Low-loss and low nonlinear fiber
From Ref [2]
Physical layer : Physical layer :
23Optical Transport Networks : motivation, challenges and possibilities
Potential Solutions
Trend: More meshed network topology Improve bandwidth utilization by ASON/GMPLS Potentially achieve good network protection
Network layer :Network layer :
24Optical Transport Networks : motivation, challenges and possibilities
Thanks !Thanks !
25Optical Transport Networks : motivation, challenges and possibilities
References1. Why complex modulated optical signals?, Lightwave,
http://www.lightwaveonline.com/articles/print/volume-30/issue-4/feature/why-complex-modulated-optical-signals.html
2. MDM, http://epic.semi.cas.cn/yjfx/201412/t20141210_271696.html
3. Cisco: Video, Internet-of-Things, mobile are prime drivers of Internet use, http://www.techtimes.com/articles/8271/20140611/cisco-video-mobile-big-internet-use.html
4. Optical Transport Network (OTN):A comprehensive study,http://www.mapyourtech.com/entries/general/optical-transport-network-otn-a-comprehensive-study
5. Winzer, P.J., "Challenges and evolution of optical transport networks," Optical Communication (ECOC), 2010 36th European Conference and Exhibition on , vol., no., pp.1,3, 19-23 Sept. 2010
6. Filling the Light Pipe, www.sciencemag.org
7. IP video to be 79% of all IP traffic by 2018, up from 66% in 2013,http://www.digitaltvnews.net/?p=24292