Afternoon session with TDC - Corporate Business Development

1
Technology and Trends inResidential Broadband
Networks
Afternoon session TDC Corporate Business
Development
Copenhagen, December 6, 2004 Halldór Matthías
Sigurðsson (halldor_at_cti.dtu.dk)
2
Agenda

• Personal Profile
• Technology Trends
• Next Generation Networks
• Residential Broadband Networks
• Service Differentiation and Competition in
  Residential Broadband Networks
• Finding the Optimal Migration Strategy

3
Session 1
Personal Profile
4
Personal Profile

• Education
• B.Sc. University of Iceland
• M.Sc. Danish Technical University
• Ph.D. Center for Tele-Information
• Iceland Telecom RD
• MPEG-2 over VDSL
• Pace DSL set-top boxes
• Implementing NGN in Iceland
• Cooperation
• TDC
• Iceland Telecom RD
• Microsoft Research Laboratories Beijing
• KTH, UiO, USC, Norfa

Publications Projects Literature www.cti.dtu.dk/
halldor/phd
5
Competence Framework
Ph.D. Multimedia Services in Residential
Broadband Networks
6
Implementing NGN in Iceland

• 17 reduction in Total Cost of Ownership in
  deploying NGN
• Article in IEEE Communications Magazine

7
Session 2
Technology Trends in Residential Broadband
Networks
8
Motivation
FUTURE
CURRENT

• Moving from several dedicated networks to one
  converged multipurpose networks

9
Moores Law for Communication
10000000
Photonic
1000000
Mb/s
100000
Telephone lines
10000
(including mobiles)
Millions
1000
Processor
performances MIPS
100
10
Internet hosts Millions
1
0.1
Sources ITU, Intel, ECOC, OFC
0.01
1970
1975
1980
1985
1990
1995
2000
2005

• Moores law
• Technology development follows exponential curves
• Doubles every X months

10
Moores Law for Residential Networks
Based on Eldering, 1999

• Exponential growth with a period of 23 months
• ADSL is currently surpassing this growth
• Other sources, e.g. BT Technology Journal foresee
  a demand of 100 Mbps in access networks in 2010

11
Broadband Evolution
Alcatel, ref. 1, 2004

• The path of broadband infrastructure evolution
  Maxwell, 1999
• Phase 1
• High speed Internet Access
• Speed lt 1,5 Mbps
• Phase 2
• Broadband Services (Voice, Video and Data)
• Speed lt 10 Mbps
• Phase 3
• User-Centric Universal Broadband Services
  Alcatel, 2004
• Speed gt 20 Mbps

12
Current Trend / Intermediate Scenario
Alcatel, ref. 2, 2004

• Towards Services
• High Speed Internet
• Video on Demand
• Voice over IP
• Broadband Television

• Moving From Transmission
• Data centric
• Oversubscription
• Lack of Resource Management

13
Sustainable Broadband Services
Alcatel, ref. 2, 2004

• Services
• High Speed Internet
• Video on Demand
• Voice over IP
• Broadband Television

What is needed and How do we get there
14
Session 3
Next Generation Networks
15
Next Generation Networks

• Vendors
• Ericsson
• Alcatel
• Cisco

• Standardisation Bodies
• ITU-T
• IETF
• ETSI
• 3GPP

• Vendors started selling NGN solutions
• Standardisation did not keep up
• Today NGN is an umbrella term for future network
  infrastructure

16
Next Generation Networks

• Formal separation between several planes through
  standardized protocols or APIs
• All IP QoS enabled core network

17
Session 4
Phase 2 Near future Residential Broadband Networks
18
Motivation

• Real-time communication (voice and video) pose
  timing constraints and end-to-end QoS Requirements

19
Resource Management

• Over provisioning
• Provide more than you need
• Loose Control
• Prioritise certain traffic
• Strict Admission
• Resource Reservation

• Near future solution will be based on a
  combination of over provisioning and loose control

20
Session 5
Service Differentiation and Competition
in Residential Broadband Networks
21
Levels of Competition
ISP
BRAS
ISP
DSLAM
CPE
INTERNET
ISP
22
Competition among Service Providers
InternetVoice
Inter DomainVoice
Intra DomainVoice
ISP
BRAS
ISP
DSLAM
CPE
INTERNET
ISP
InternetVideo
Inter Domain Video
Intra Domain Video
23
Scenario 1Best Effort
InternetVoice
Inter DomainVoice
Intra DomainVoice
ISP
BRAS
ISP
DSLAM
CPE
INTERNET
ISP
InternetVoice
Inter Domain Video
Intra Domain Video

• Current Situation
• Internet / Public Garden Model
• Network Access Provider must rely on income from
  transmission
• No QoS - Voice, video and data compete for
  resources
• Transmission characterised by bottlenecks and TCP

24
Scenario 2One Way Service Differentiation
InternetVoice
Inter DomainVoice
Intra DomainVoice
ISP
VoIP
BRAS
ISP
DSLAM
CPE
HAN
INTERNET
STB
ISP
InternetVoice
Inter Domain Video
Intra Domain Video

• Upgrade in BRAS (Same DSLAM and CPE)
• Downstream Prioritising Upstream Competition
• Supports VoIP and VoD Not television
  broadcasting
• Competing Service Providers are worse off

25
Scenario 3Port Based Service Differentiation
InternetVoice
Inter DomainVoice
Intra DomainVoice
ISP
VoIP
BRAS
CPE
ISP
DSLAM
HAN
INTERNET
STB
ISP
InternetVoice
Inter Domain Video
Intra Domain Video

• New CPE, DSLAM upgrade
• Port based prioritisation
• Supports VoIP, VoD and broadcast television
• Intra Domain / Walled Garden Model
• Discriminates Inter and Internet Service Providers

26
Scenario 4Content Based Service Differentiation
InternetVoice
Intra DomainVoice
Inter DomainVoice
ISP
VoIP
BRAS
CPE
ISP
DSLAM
HAN
INTERNET
STB
ISP
InternetVoice
Intra Domain Video
Inter Domain Video

• New CPE, DSLAM upgrade
• Two way content based service differentiation
• Supports VoIP, VoD and broadcast television
• Inter Domain / Gated Garden Model
• Operators can choose which third party operators
  to allow access

27
Session 6
Finding the Optimal Path
28
Case Study

• Techno-Economic Study of the implementation of
  triple-play services
• Infrastructure
• Functional alternatives
• Components
• Start with DSL based
• Services
• High Speed Internet
• Voice over IP
• Video on Demand
• Broadband television broadcasts
• Simulate
• Investment Cost
• Operational Cost

29
Simulation Model

• Excel based model
• Graphical User Interface
• Visual Basic for Application

30
Conclusion

• Residential Broadband Networks are moving towards
  converged voice, video and data transmission
• Successful shift will require service
  differentiation/resource management in some form
• Most likely path requires new equipment (both at
  customer and operator)
• Operators will choose a migration strategy that
  grants the market control
• Operators will have to open their networks to
  third-party service providers

31
Questions
32
References

• Sigurdsson H.M.,Thorsteinsson S.E., Stidsen T.,
  2004. Cost Optimization in the Design of Next
  Generation Networks. IEEE Communications
  Magazine, September (Available online at
  http//www.cti.dtu.dk/halldor/publications/sigurd
  sson_ieee.pdf)
• Maxwell, 1999. Residential Broadband An
  Insiders Guide to the Battle for the Last Mile.
  Wiley, New York.
• Alcatel ref. 1, 2004. The User-Centric Broadband
  World. Strategic White Paper (Available online
  at http//www.alcatel.com/com/en/appcontent/apl/U
  serCentric_BB_swp_tcm172-74441635.pdf)
• Alcatel ref. 2, 2004. Broadband Network
  Architecture. Strategic White Paper (Available
  online at http//www.alcatel.com/com/en/appconten
  t/apl/UserCentric_BB_swp_tcm172-74441635.pdf)
• Alcatel, 2003. Evolution of the First and Second
  Miles for Pervasive Service Delivery, Technical
  Paper
• Odlyzko, A., 2001. Content is Not King. First
  Monday, Vol. 6, No. 2 (Available online at
  http//www.firstmonday.org/issues/issue6_2/odlyzko
  /index.html)
• Eldering, C., Eisenach J.A., 1999. Is There a
  Mooress Law for Bandwidth? , IEEE Communications
  Magazine, October 1999