Optinen verkkotekniikka

1
Optinen verkkotekniikka
TKK, 9. 4. 2002 S-38.164 Laajakaistainen
välitystekniikka kevät 2002

• Jaakko Aarnio
• Nokia Research Center
• jaakko.aarnio_at_nokia.com

2
Yhteenveto

• Internet- ja multimedialiikenteen huima kasvu on
  johtaa siihen, että dataliikenne syrjäyttää
  perinteisen puhelinliikenteen hallitsevan roolin
  tietoliikenneverkojen liketoiminnassa.
  Kaistanleveysvaatimukset tulevat sen myötä
  kasvamaan suuresti lähitulevaisuudessa - syntyy
  akuutti tarve paljon suurempiin siirto ja
  kytkentänopeuksiin kuin mitä nykyiset ATM- ja
  SDH-verkot pystyvät tarjoamaan. Se edellyttää mm.
  uusien optisten järjestelmien käyttöönottoa
  verkkojen infrastruktuurissa.
• Yksimuotokuidun kapasiteetin voidaan arvioida
  olevan noin 50 terabittiä sekunnissa (Tbps), joka
  on noin neljä kertaluokkaa suurempi kuin
  elektroniikan nopeus. Kuidun kapasiteettia
  pyritään hyödyntämään yhä paremmilla
  transmissiotekniikoilla, optisella
  kytkentäisyydellä ja kuidun alkuinvestointien
  jakamisella monipalvelu-tilaajaverkoissa.
• Aallonpituusmultipleksointi (wavelength division
  multiplexing, WDM) on yksi keino hyödyntää kuidun
  kapasiteettia, missä loppukäyttäjän laitteet
  toimivat vain sähköisellä nopeudella, mutta useat
  eri loppukäyttäjien WDM-kanavat voidaan
  multipleksoida samaan kuituun. Lähitulevaisuudessa
  odotetaan asteittaista siirtymistä optisten
  ratkaisujen käyttöön myös tilaajaverkoissa
  (kiinteät ja langattomat verkkoratkaisut).
• Tämä esitys on katsaus optisten ratkaisujen
  kehityksestä data- ja tietoliikenneverkoissa.
  "IST-road map for optics communications, 2001" on
  liitetty myös mukaan.

3
Content

• Need for optical networking
• Basic technologies
• Optical networking in
• core/metro solutions
• access for fixed and wireless solutions
• Technology roadmap
• Standardization

4
Need for optical networking

• User and service requirements
• Increasing traffic volumes
• see also internet indicators
• http//www.internetindicators.com/

5
Market evolution for Fiber Networks
CumulativeMarketsize
Access wave
Metro wave
Core wave
2000
2005
2010
1990
1995
6
Internet economy and carriers' earnings

• Global internet commerce revenue since 1998
• gt216 Billion US Dollars (March 26, 2001)
• ref http//www.internetindicators.com/facts.html
• The Internet Infrastructure Indicator
• The Internet Applications Infrastructure
  Indicator
• The Internet Intermediary Indicator
• The Internet Commerce Indicator

Carriers earn best on voice
Revenues in Billions of US dollars
Source Datamonitor 1999
7
User and Service Requirements
64kb/s
100kb/s
1Mb/s
10 Mb/s
100Mb/s
10000
1Gb/s
1000
D
F
G
E
Response time (ms)
100
C
A. POTS B. Videoconferencing (low quality) C.
Videoconferencing (high quality) D.
Teleworking E. Telelearning F. Information
exchange and retrieval G. Entertainment
A
B
10
including Internet
1
102
103
104
106
105
107
108
109
Information content (bits)
Source Heinrich Hertz Institute
8
Evolution of Transport Technology
Ref D.T. Neilson al. 35,8Tbps (320 Gbps,
112112 ports) MOEMS based OXC demonstrated at
OFC 2000, PD12
WDM
SDH
switch granularity
link capacity
PDH
Teleph
Mbit/s
1
10
0.1
100
0.01
1000
10000
100000
Historical traffic increase by a factor of 30-60
at each step. (source ACTS/Horizon)
9
Some Basics (Fiber attenuation)
By elimination of water peak attenuation, total
usable bandwidth in fiber has been increased to
some 45 THz (50), including L-band (1560-1600
nm) as 4th window (not shown in Fig.).
RefLucent Tech., Allwave fiber
10
Basic technologies

• Fiber capacity
• Optical multiplexing
• Wavelength Division Multiplexing
• Evolution of capacity cost

11
Dense vs. Coarse WDM transmission
Ref J. Campbell, Coarse WDM makes waves in
Metro/access markets, Laser Focus World, Nov 2000
12
Dense vs. Coarse WDM transmission

• CWDM devices available
• light sources not yet available
• due to lack of standard

13
Wavelength Allocation

• ITU-T Standards
• 192.100 THz (1560.61 nm) as
• reference wavelength, 100 GHz
• grid.
• Wavelength Located in flat gain
• region of Erbium-Doped Fiber
• Amplifier (EDFA), the flat gain
• region depended on the saturation
• level of EDFA, etc.
• New developments in S-band (Thullium Doped
  FAs, 1450-1500 nm) to broaden the wavelength
  region at short wavelength region.
• Wavelength Spacing depend on
• the techonologies of laser, optical
• multiplexer and demultiplexer, etc.
• L-band doubles the gain bandwidth
• of an optical amplifier.

14
Cost Capacity evolutionSubmarine Networks
Cost/circuit (US/km)
Nb of tel. circuits
From 1956 to 1998 Capacity x 7200 Cost/circuit
10000
10 000 000

1 000 000

Cost percircuit
100 000
10000
TAT-2
TAT-4
TAT-1
10 000
1000
TAT-3
TAT-7
TAT-5
1000
100
TAT-6
TAT-8
TAT-9
100
10
TAT-12/13
10
1
Gemini
1
1860
1950
1970
2000
1960
1980
1990
2010
15
Cost evolution of 600 km 2.5 Gb/s circuits
Source Alcatel
16
Key technology trends

• Multimedia and other services requiring more
  bandwidth lead to a strong increase in traffic
  (about 35 annual growth for data and 10 for
  voice in core transport).
• Optical communication technologies will have
  direct impact on future datacommunication network
  infrastructure to efficiently and economically
  support the wide diversity of present and future
  applications and services.
• Optical networks have huge bandwidth 10 THz
  available with many beneficial features (e.g.
  cost-efficiency, survivability, scaleability,
  reconfigurability, transparency).
• New networking solutions are enabled by new
  developments in component technologies (e.g.
  wavelength conversion, optical switching
  components, technical solutions for signal
  monitoring in optical layers).
• Terabit switching platform development may
  combine IP, ATM, Gig-E, SDH/SONET, and wavelength
  switching.

17
Optical networking

• WDM network evolution
• Evolution in access and core
• Optical networking as part of communication
  networks
• Simplification of protocol stacks - Optical
  Internetworking

18
Scenario for the WDM deployment in the transport
network
Ref ACTS/Horizon
19
Optical cross connect (OXC) - principle of
operation

• Optoelectronic crossconnects
• available at 512x512 port counts
• Transparent design based on
• MOEMS technology

Optical space switch
O/E/O conversion
WDM
Opaque
f conversion
optical bypassing
Transparent
OXC
Note OAM signaling is considered separately.
Ref Agilent technology, OFC 2000
20
Is telecom still going to rule the backbone and
regional environment ?
Mainstream market demand
IP with de-facto interfaces over optical network
SDH, ATM over optical network
PRODUCT PERFORMANCE
disruption ?
data 35/yr in US
voice 10/yr
TIME
Sustaining trajectories
21
Network Evolution Options
(10 Mb/s)
(100 Mb/s)
(50 Mb/s)
Radio/ Mobile
VDSL
FTTH
Copper ADSL
ACCESS
PON (ATM) (Super)
SDH ADM
WDM
32 x 10 Gb/s
TRANSPORT
SDH DXC
OADM
WDM
Photonic Transport Layer
OXC
1 Tb/s WDM
22
Evolution in Layering
Optical layer (e.g.WDM)
(source ACTS/Horizon)
23
Issues of optical networking

• Transparency in optical networks
• Interoperability and Interworking
• Network and equipment architectures
• Network management, monitoring and protection
• Passive and active components
• Network planning, design and performance
  evaluation
• Different levels of optical regeneration and
  wavelength conversion

24
State of the art Commercial
2000
2005
2010
1995
Gb routers
25
Research Work Needed

• Multiservice access networks (fixed and wireless)
• to share the cost of fiber plant while the
  demand of bandwidth increases
• who, what, where, when?
• Network Architectures
• Optimum wavelength routing for the client traffic
  (ATM. FR, Fast IP, Gigabit Ethernet)
• Dynamic wavelength assignment and routing,
  bandwidth on demand
• Network Protection Schemes
• Network control and management, and their
  compatibility with higher layer client (SDH, ATM,
  etc.)
• IP optics mapping
• Network Element
• Optical Cross-Connect (OXC) and Optical Add
  Drop Multiplexer (OADM) functionality,
  capabilities, structures
• Optical Amplifier with wider bandwidth and high
  gain
• Components are 1/4 or 1/100 cheaper
• Network Control and Management is the key

26
Networking
27
Standardization
ANSI T1X1
ITU-T
Optical InternetworkingForum
ATM Forum
IETF
IEEE
ITU-T
28
Standardization of optical networking and
internetworking

• Optical Domain Service Interface (ODSI), kick-off
  meeting, January 18, 2000
• Optical internetworking Forum - OIF
  http//www.oiforum.com
• Technical Subcommittee - T1A1 Performance and
  Signal Processing,
• http//www.t1.org/t1a1/t1a1-rms.htm
• Technical Subcommittee T1E1 - Interfaces, Power
  Protection of Networks,
• http//www.t1.org/t1e1/_e1-rms.htm
• Technical Subcommittee T1M1 - Internetwork
  Operations, Administration, Maintenance and
  Provisioning http//www.t1.org/t1m1/_m1-rms.ht
  m
• T1S1 Technical Subcommittee - Services,
  Architectures and Signaling
• http//www.t1.org/t1s1/t1s1-rms.htm
• T1X1 Technical Subcommittee - Digital Hierarchy
  and Synchronization
• http//www.t1.org/t1x1/_x1-rms.htm
• TIA FO-2 Committee on Optical Communication
  Systems
• http//www.tiaonline.org/standards/sfg/scope.html
  FO-2
• SONET Interoperability Forum (SIF)
  http//www.atis.org/atis/sif/sifinfo.htm
• ATM Forum http//www.atmforum.com/atmforum/abo
  ut/intro.html
• Internet Engineering Task Force (IETF)
  http//www.ietf.org/tao.html
• ITU-T Study Group 13 General network aspects

29
Provisional conclusions (1/3)

• Factual
• Explosive bandwidth demand in fixed transport
  networks
• circuit switched optical networking solutions are
  penetrating the core and metropolitan networks
  with protocol, bitrate or optical transparencies
• Strong trend towards packet orientation (e.g. IP)

Source Lightwave May-97 (assuming data was 20
of voice 1990)
30
Provisional conclusions (2/3)

• Issues to be resolved
• optimization of multiservice access networks to
  allow convergence to take place in the physical
  layer (fiber) simplification of solutions, e.g.
  QoS?
• Ethernet in the First Mile in 3-5 years very cost
  effective for residential access, strong
  competitor to xDSL and CATV
• Cost effective microwave photonic solutions for
  radio front-end?
• Evolution of enabling technologies to allow the
  full functionality of fault tolerant, flexible
  optical connection oriented or connectionless
  networks ?
• Role of higher network layers, e.g. IP, Gigabit
  Ethernet, 10 Gig-E, ATM, SDH over WDM/Optical
  networking ?
• degree of transparency ?
• Effect of fast evolving de-facto standards ?
• Optical signal processing (e.g. OCDMA) in access
  networks, Lan, MAN environments ?
• Role of OTDM in transport networks ?

31
Provisional conclusions (3/3)

• Speculative
• optical burst or label switching?
• optical packet switching?