Conference Review

1
Conference Review

• ICIN'2006 TPC

2
Some facts and figures

• 214 delegates
• lt100 companies
• 31 countries
• 3 Tutorials
• 1 workshop
• 3 key note presentations
• 17 sessions
• 57 papers
• 6 poster session
• 2 exhibitions

• 10 networking sessions
• 2 cocktail receptions
• 1 gala dinner
• 400 bottles of wine
• 1600 meals

3
TPC chairmans observations

• The subject of convergence was very timely
• The competing business models of Qos, payment and
  customer service vs no-Qos, free and self service
  seem to be critical to the way the industry will
  develop.
• The telco model for supporting convergence
  demands flexibility which drives complexity
• Complexity will lead to problems in network,
  service and customer management. No one is
  addressing this issue
• We need self managing solution that incorporates
  automatic load control, data management/cleansing,
  rigorous interworking rules
• No account of user demand/actions such as P2P

4
Hot topics for the conference

• Context based services
• Identity management
• TV services
• Qos and overload control
• IMS Architecture
• Service delivery platforms/methods
• Business models

5
The Future

• Topics for next conference
• Community services
• Managing the complexity
• Delivering Interoperability
• Delivering the promised services with Qos
• IMS and SIP
• Date of next Conference
• 8th 11th October 2007
• Bordeaux
• TPC Chairman Ulrich Reber (Siemens)
• IAB Chairman Stuart Sharrock

6
8th 11th October 2007
7
Backup
8
Session 1A Challenges to the IMS

• IMS is bringing together two worlds (IETF and
  TLC), there is the need to take full advantage of
  it
• IMS can start with TELCO Services, but it has
  soon the need to take advantage e show the
  benefit of the convergence of IP and TLC
• IMS has still to prove its validity in terms of
  Revenue generating Services, consolidation of the
  Architecture and timing implementation
• Skype has demonstrated that there can be almost
  zero-cost telephony
• Are There Alternative to the IMS ? It seems there
  are Skype is one. Its approach is oriented
  towards the view of VOICE as a Software
  Application. Telco have to change attitude in
  developing new services as fast as Web 2.0
  companies. IMS can be used as a means.

9
1B QoS

• Over-engineering cannot economically meet the
  diverse performance and BW needs of multi-service
  NGNs.
• Policy driven, dynamic, inter-domain,
  application-driven resource management is
  therefore necessary.
• ETSI TISPAN RACS and ITU-T RACF offer two
  important standards approaches.
• A stand alone Bandwidth Manager is one QoS
  option.
• Separating the Session Border Conroller functions
  into Border Controller and Access Border Gateway
  elements and distributing the ABG to the access
  node may offer advantages for guaranteeing QoS.
• E2E QoS is still a research topic, but crucial as
  the demand for bandwitdh increases
• Overlap between QoS signalling and Session
  Control Signalling. How to avoid it?

10
2A TV services

• Basic Triple Play offering is not enough to
  attract/retain customers. A new TV experience is
  required. Personalized interactive TV services
  promise additional value and revenues for
  operators and content providers be it in a
  fixed or mobile environment
• Ease of use is key for success for IP TV as well
  as for mobile TV
• DRM / key management are a pre-condition for
  content-providers to allow operators the
  broadcast of their content

11
2B User profile and context

• Context based service seem to be the future
• Platforms for the aggregation of context
  information come into place
• Focus point will be to consider an appropriate
  management of user permissions of his own context
• How to handle the huge amount of subscriber data
  to be stored in the network seems to be of
  interest to many market players

12
3A Convergent Services Delivery / SDP

• The Session addressed different IMS and SDP
  issues. The first presentation proposed solutions
  for the implementation of enriched voice services
  in the combined CS and PS environment. Very
  detailed presentation with almost no response on
  the part of the audience.
• The second presentation analyzed two approaches
  for the implementation of instant group
  communication within an IMS environment which
  were successfully validated in the HPs IMS
  Experience Center and was, judging from the
  discussion, very well accepted.
• The third presentation, in my opinion the best of
  all three, was also well accepted by the
  audience. It defined the SDP as a collection of
  generic SOAs. The generic SOA is obtained by
  extracting technology independent concepts from
  the current SOA in such a way that implementation
  technologies and distribution mechanisms may be
  chosen and not imposed. Each generic SOA
  visualises a service-oriented reference point.

13
3B Identity management in NGN

• An Identity Centric Architecture plus peer to
  peer technology and a REST based approach can
  allow for the creation of disruptive, simple and
  totally decentralized communication architecture
• Identity Management Identity and Role Management
  (IRM) extension approach defines roles as being
  assumed by a person (or principal) extending his
  identity (ID) according to the current context.
• A solution for handling identity of end-users
  accessing Internet applications through different
  access networks is discussed. It based on a
  mechanism translating to the application level
  the authentication information provided by the
  access network.
• New business models related identity management
  is presented. It will bring new business
  opportunities to telecom operators.

14
4A Next Generation Services

• Summary
• Usibility is the key success factor for next-gen
  services
• FMC services based in UMA/IMR are already in
  place
• Enterprise servivces can be mobilized by Web
  Services
• Issues
• Terminals supporting FMC are still not fully
  available
• Users may be confuzed by too many 3rd party
  soft-clients on their next-genration mobile
  terminals

15
4B Programming models for IMS

• SIP servlet technology
• IMS platforms implemented (playground, proof of
  concept)
• Service enablers (well defined QoS, mobility
  management, presence ..)
• Triggering of AS
• Comparison of technologies for AS JAIN SIP, SIP
  Servlets, Parlay, Parlay X
• Triggering of AS in IMS is still inmature
• Service (feature) indication remaining problem
  little specifics on SCIM
• 3rd party service development Parlay, Parlay X
• SIP control SIP servlet is JAIN SIP X (easier
  to use but less complete)

16
5A Business Aspects of service convergence

• The Business Drivers behind the major decision to
  switch from Stove-Pipe Architecture to a
  Converged service layer could probably be
  addressed in a separate conference as the scope
  is very wide.
• Should Business drivers be addressed in a
  Technical conference like ICIN ? Feedback from
  the delegates are welcome !
• Should Business Drivers be a separate track in
  ICIN with papers evaluated on different on
  different criteria ?

17
5B Service Management and Delivery

• Management of the many intelligent elements in an
  NGN architecture is a key challenge
• The Sip event distribution can provide a
  framework for distribution and collection of
  management information
• Add-hoc networks demand service delivery
  solutions without central elements
• A service broker based on a fully distributed
  processing model can help
• Complexity in the end-user environment (multiple
  devices, identities, and networks) effects many
  service delivery problems
• A home gateway is a useful approach in
  simplifying service delivery

18
6A Service Composition in NGN

• Main point necessity to solve the problem of
  easy integration of telco services or telco
  enablers (call control, presence, localization,
  user context) inside Web Services.
• Extension of SOA (Web Services) to ubiquitous
  communication environment, enabling customization
  based on user context, including his physical
  environment service composition is managed by
  fixed proxy for mobile elements, following a
  meta-programming schema that composes Services
  Elements in accordance with a service template.
• Need for an architecture for a Telco convergent
  SOA platform service blending seamlessly both
  Telco services and classical IT services
• Telco services have real time constraints ,
  beneficiates from asynchronous events .
• Proposal for an event oriented Application Server
  for web services supporting stateful Web
  Services, asynchronous interactions and
  notifications , the engine may be based on
  asynchronous engines, such as JAIN SLEE.
• Service Creation Composition revisited with
  combination of both Software Development Kit
  (SDK) and classical fast Visual SCE, instance of
  environment applied to Parlay X API.

19
6B Network Infrastructure and Protocols

• Overload control in NGN
• Is a non-functional requirement in NGN/IMS to be
  addressed
• GOCAP control protocol being worked out for
  implementation in SIP Nodes
• Multimedia content adaptation in IMS
• multiplicity of devices and access types require
  solutions for content adaptation
• Service specific overlay network and associated
  MM routing could help addressing dynamic content
  adaptation
• Protocol extensions could open door for security
  risks
• But extension are required and inhibi them is
  pointless
• Change need be managed in an open manner
• Shall always promote interoperability

20
7A User profile and context

• There is a definite role for the operator to help
  manage the vast amount of user profile and
  context information
• Question is how do we get hold of all the
  information in a controlled way, that will ensure
  that the context information is always current
• Properly implemented it will certainly help
  service interaction and invocation

21
7B Business Drivers for NGN

• Paper 1 Which are the technical issues, most of
  the times underestimated, behind the MNVO
• Paper 1 How these technical issues can force to
  a concrete type of business model
• Paper 2 How a concrete service can boost the
  revenues of a company (we all hope!!)
• Paper 2 Which are the technical challenges for
  an effective delivery of mobile TV
• Paper 3 Which are the economics of the movement
  towards IMS from PSTN.
• Paper 3/4 How important is to correctly
  estimate the service traffic, to adequate
  resources
• Paper 4 How difficult is to get accurate
  estimations, due to big amount of parameters
  involved

22
8A Charging key points

• Network
• Variety of services, content, terminals and
  accesses
• Multiple service providers (Operator, ASP, ISP,
  ITSP, Internet Community providers)
• End-user
• Value of bit is variable. 100 byte SMS is
  valuable than 100 B (milliseconds) streaming
  video
• Business
• Revenue share models
• Charging as part of Customer Care
• Conclusion
• Networks need to have differentiated charging.
• At the end end-user is the king, he needs to
  have a predictable and understandable charging
  scheme from his service providers.

23
8B Authentication security

• Non-IMS and IMS approaches
• Much emphasis on home devices and
  telemetricsopen issue on whether to use SIP for
  device control
• Multitude of authentication methods
  password-based, ISIM (soft- and smart card),
  embedded URLs
• Interesting, refreshing research resultsmuch is
  left for further study
• Need for standards solutions beyond IMS
• We will hear much more at ICIN 2007!

24
9 P2P Summary

• P2P is here now and it is well tested
• Kazaa, Skype, Groove,
• Amount of applications are increasing
• As seen in these 3 presentations
• IP Multimedia Control System using IMS and P2P
• Mobile Peer to Peer architecture
• A combined SIP and P2P for scalability, cost etc.
• I believe that both P2P and the Classical
  Centralized Service Architecture will be used
• Powerful Clients with Amazing Servers!!!

25
9 P2P Chair Summary

• P2P one of the hot topics in the conference
• Session with 3 different views on usage of P2P in
  Telecom
• In network architectures
• Mobile P2P file sharing
• All three speakers made great presentations and
  on time, allowing time to QA and Summary
• Rounded up the conference well based on the
  Keynotes and other speeches ?

26
Summary of the Workshop (1/2)

• More than 45 questions around NGN and IMS have
  been collected from the participants in a
  brainstorming session
• Three clusters have been identified
• Business Models / Charging
• Technology
• Services and Applications

27
Summary of the Workshop (2/2)

• The discussion lead us to some answers / topics
• Flexible charging is a must (who is charged,
  community charging, event based charging, on
  transport- and on application level, transparent
  tariff models, bonus programs, different tariff
  models in different markets, support of
  whole-sale business models)
• There is a risk of increased complexity
  (distributed architectures need integration,
  multi-vendor problem, performance issues, service
  interaction, global environment / roaming, keep
  lower levels service agnostic)
• Co-existence of IN and NGN/IMS is a given (no
  big-bang to NGN/IMS, migration of successful IN
  services like VPN needed)

28
Various messages

• There are similarities between the IMS hype
  what was the IN hype in the 90's. IMS can't solve
  all aspects of service creation.
• Signalling And Managemen Control is not enough.
  We need content control.
• Ad's insertion in content might be the way to
  bring money back in the telco world

29
Managing Control Plane Overload in Next
Generation Networks

• Will the NGN architectures for the control planes
  and application environments support todays
  services and tomorrows user behaviours?
• What thought is being given to managing peaks in
  processing resource demand?
• How can service levels be protected under
  overload demand in the multi-protocol,
  multiple-connected meshes of functions in NGNs?
• How can these issues be addressed in a future
  proof, implement once way that copes with
  multi-vendor inter-operability.

30
A Service Specific Overlay Networkfor Adapting
SIP Multimedia Calls

• How can IMS-based networks accommodate multiple
  access technologies and heterogeneity of handsets
  for delivery of content?
• Are SIP/IMS current adaptation mechanisms
  flexible enough for delivery of multimedia
  content when roaming between several access
  technologies?
• Can media adaptation be addressed in the network
  by dynamic nodes?
• How SMART (Smart MultimediA Routing and
  Transport) and SSON (Service Specific Overlay
  Network) concepts help better responding dynamic
  content adaptation?

31
Network Protocols extensions Are They Bad?

• Interoperability of the standard-compliant
  implementations becomes ambiguous when the
  network protocols are extended or selectively
  implemented
• Protocol options and features can be detrimental
  to the given protocol security
• There are definite dangers of unwarranted
  protocol extensions
• What would be a prudent approach to extension of
  network protocols?