Transcript Prof. Younghee Lee

Computer Networks

Lecture 13: Future Networks
Prof. Younghee Lee
Prof. Younghee Lee
Future Internet Services – Survey [1]

Findings: agree
– Some 66% of the experts agreed that at least one devastating
attack will occur in the next 10 years on the networked information
infrastructure or the country’s power grid
– 59% of these experts agreed with a prediction that more government
and business surveillance will occur as computing devices
proliferate and become embedded in appliances, cars, phones, and
even clothes.
– 57% of them agreed that virtual classes will become more
widespread in formal education and that students might at least
occasionally be grouped with others who share their interests and
skills, rather than by age.
– 56% of them agreed that as telecommuting and home-schooling
expand, the boundary between work and leisure will diminish and
family dynamics will change because of that.
– 50% of them believe that anonymous, free, music file-sharing on
peer-to-peer networks will still be easy to perform a decade from now.
[1] “In a survey, technology experts and scholars evaluate where the network is headed in the next ten years”, 202-4194500 http://www.pewinternet.org/ , January 9, 2005
Prof. Younghee Lee
Future Internet Services – Survey [1]
 Findings:
disagree
– Just 32% of these experts agreed that people would use the internet
to support their political biases and filter out information that
disagrees with their views. Half the respondents disagreed with or
disputed that prediction.
– Only 32% agreed with a prediction that online voting would be
secure and widespread by 2014. Half of the respondents disagreed or
disputed that idea.
Prof. Younghee Lee
Future Internet Services – Survey [1]

Findings: In the emerging era of the blog, experts believe the internet will
bring yet more dramatic change to the news and publishing worlds.
They predict the least amount of change to religion.
– Connections across media, entertainment, advertising, and commerce
» Google and Starbucks have a chance to build all-new new distribution models
– Health care is approximately 10 years behind
» boom in the next 10 years
– Government will be forced to become increasingly transparent,
– Digitization and the Internet make for a potent brew
» In the future, everyone will be famous for fifteen minutes in their own reality show
– Hyperlinks subvert hierarchy.
– The ‘always-on’ internet, combined with computers talking to computers
– The next decade should see the development of a more thoughtful internet.
We've had the blood rush to the head, we've had the hangover from that blood
rush; this next decade is the rethink.”
– The dissemination of information will increasingly become the dissemination
of drivel.
» As more and more ‘data’ is posted on the internet, there will be increasingly less ‘information
Prof. Younghee Lee
Future Internet Services – Survey [1]

Findings: Experts are both in awe and in frustration about the
state of the internet. They celebrate search technology,
peer-to-peer networks, and blogs; they bemoan institutions
that have been slow to change.
– What dimensions of online life in the past decade have caught them
by surprise?
» Pleasant surprises: These experts are in awe of the development of the
Web and the explosion of information sources on top of the basic
internet backbone. They also said they were amazed at the
improvements in online search technology, the spread of peer-to-peer
networks, and the rise of blogs.
» Unpleasant surprises: The experts are startled that educational
institutions have changed so little, despite widespread expectation a
decade ago that schools would be quick to embrace change. They are
unhappy that gaps exist in internet access for many groups – those
with low income, those with lower levels of educational attainment, and
those in rural areas. And they still think there is a long way to go before
political institutions will benefit from the internet.
Prof. Younghee Lee
Future Internet Services: summary

Revised summary in terms of technology
– Broadband networks: core (fiber) + access (wireless)
– Rethinking the Internet: improved or new one?
» Security
» Network management..
– Ubiquitous computing
» Smart device, smart space
– Dissemination networks
» Content based networks
– Web based services: Good surprise:
» Web + explosion of information sources
 Search technology, P2P, blog
» What about ad hoc environment or weakly connected environment?
– Applications: summary
» Media (news, publishing, IPTV..), entertainment, advertising, commerce,
telecommuting, home schooling, p2p
» Need more efforts: Healthcare, Education, information divide
» Less impact: politics, on-line voting, religions..
Prof. Younghee Lee
Services

Service? [5]
– Telco view:
» end user pays for
» provided using communication infrastructure
– Internet view:
» An executing component
» is accessed using a communication infrastructure
» Composition forms a hierarchy of services and applications

Need to change the concept of communication infrastructure to enlarge
the communication market
– Sharable => can be composed with various service elements for application
– Service software infrastructure => sharable
– Network virtualization, Service virtualization

Service Market of enabled by SOA, web services and Web 2.0 [5]
– By 2011, 63% of products in the software infrastructure market and 56% in the software
application market will support web services and ‘Web 2.0’ technologies
– The worldwide software market for service-oriented architecture(SOA), web services and
Web 2.0 will grow from $41 billion to $142 billion from 2006 through 2011(CAGR 28%)
– The worldwide application software market enabled by SOA, web services and Web 2.0
will grow from $17.2 billion to $76.2 billion from 2006 through 2011.
Prof. Younghee Lee
Future Internet Services – wireless [4]

The Internet is becoming wireless
– Laptop sales exceeded desktop PC sales in July 2003
– By 2015, # wireless Internet terminals >> # wired!
– Laptops, cell-phones, PDA’s, Embedded devices (sensors, actuators,
RFID,…)
– Wireless access networks must scale and handle new types of
devices (sensors, etc.)

Internet Architecture
– IPv6, QoS, Mobile IP.. : deployment problems
– B-ISDN, 3G ?
– Need to encourage bottom-up transformation without loss of
investment in legacy system
» Evolutionary strategy
» New approaches to protocol standards
» Economic incentives for deployment
Prof. Younghee Lee
Future Internet Services – wireless [4]

Mobile data
– Terminal mobility (authentication, roaming and dynamic handoff), mobile
IPv6, Multicasting, IP multicast, Security, Wireless TCP: R&D during 90’s

Mobile P2P
– Router mobility, delayed delivery, caching and opportunistic data delivery (in
network storage), content-aware/location-aware data delivery


MANET
Sensors: Applications:
– Highway safety: accident prevention
» Opportunistic, attribute-based binding of sensors and cars, tight real-time and reliability
constraints
– Assisted Living
» Emergency event triggers interaction between object sensors and body sensors and initiate
external communication: Heterogeneous ad-hoc network, Sensors used to detect events and
specify location, Real-time communication with care provider
– Overlay services
» used for content distribution or dynamic binding between sensor devices and servers, agents,
end-users
 Use of XML or similar content descriptor
Prof. Younghee Lee
Future Internet Services

Characteristics
–
–
–
–

User Centric Service Network: Personalization
1 to n, n to n Data dissemination
Peer-to-Peer/Ad-hoc Information dissemination
Heterogeneous devices for various services and information
Constraints of today’s Internet services
– Not reflect user’s preference at network level
– Limitation of TCP/IP
– Need binding information prior to the communication
between the nodes
– IP address based routing
» Do not directly related with service provider/requester
» Service based routing between Service requester – service provider
Prof. Younghee Lee
Context aware service in other future
internet initiatives
 Projects
in NSF NeTS FIND
– “A Geometric Stack for Location-Aware Networking”
» New geometric stack for enabling general purpose spatial
communications
– “Service-Centric End-to-End Abstractions for Network
Architecture”
» Services can be specified by end-systems, mapped to resources inside
the network, and controlled during their operation.
– “An Internet Architecture for User-Controlled Routes”
» An Internet architecture that enables users or their end-systems to select
the paths their packets take through the network.
– “CABO: Concurrent Architectures are Better Than One”
» Allows a service provider to operate multiple virtual networks, each of
which is tailored to a specific application
Prof. Younghee Lee
Context aware service in other future
internet initiatives (cont.)
 EC
FP7 Work Programme 2007-08
– Target outcome of “Objective ICT-2007.1.1: The
Network of the Future”
» a) Ubiquitous network infrastructures and architectures supporting: …;
iv) context awareness;
– Target outcome of “Objective ICT-2007.1.2: Service and
Software Architectures, Infrastructures and
Engineering”
» a) Service architectures, platforms, technologies, methods and tools
that enable context awareness and discovery, advertising,
personalization and dynamic composition of services.
Prof. Younghee Lee
Pervasive Adaptation: FP7 EU [3]


ICT-FET: ICT- Future Emerging Technologies
Pervasive Adaptation: IST-2007.8.2: 17 M€
– Key features:
»
»
»
»
Self-adapting software, hardware, protocols, architectures, …
Massively scalable, Capable of adapting to highly dynamic contexts
Autonomous adaptation strategies (bio-inspired, stochastic, ...)
Multidisciplinary, human-centric research
– Evolve-able pervasive systems
» From short term adaptation to long term Evolution
» To robustly respond to dynamically changing environments, operating conditions..
– Networked societies of artefacts
» From local autonomy to collaborative systems
» Cooperatively attempting goals with society-like behaviour
» Beyond localising and recognising other artefacts: adapting to each others and to
changing needs helping others to adapt to yourself
» For context-sensitive service delivery in rapidly changing environments
– Adaptive security and dependability, Dynamicity of trust, Security for
tiny and massively networked devices
Prof. Younghee Lee
Future Internet Services
 Characteristics
[1], [5], [6]
– Nomadic access, always on experience
– Real time Internet services: end to end QoS
– From free-of-charge access to value-based transactions
» collaborative services emerge and the distinction between
producers and consumers becomes blurred.
» some business models will be based on payment for services,
others on payment by third parties, e.g. advertisers.
» Service brokers will form supermarket, where consumers can pick
and choose and compose their personalized service
Prof. Younghee Lee
Future Internet Services
 Characteristics
[1], [5], [6]
– Creating trust: trusted and secure system
– Simplification of the user interface
» An intuitive, easy-to-access user interface across a wide range of
devices.
» The open system that users can combine services offered from
several domains to compose new personalised ad-hoc services
– User empowerment: applications built on-demand by nonIT experts. (Demand rather than supply driven)
» transition to a world of user generated content and services
» User: Consumers as suppliers
» The creation and governance of applications of service-oriented
infrastructures
» New language and design paradigms, tools, and interfaces for
empowering unskilled people
Prof. Younghee Lee
Future Internet Services
 Characteristics
[1], [5], [6] (continued)
– Increased machine-to-machine (M2M) interaction:
» Significant part of the growth of internet usage
» Need for standard low-cost, always-on ubiquitous network
access and secure, efficient (i.e. low-bandwidth) and timesensitive protocols
– Evolvability and Adaptability
» dynamically assembled and characterized by the ability to reconfigure themselves to accommodate unforeseen events
– Innovation
» emerging business models and discover new possibilities
Prof. Younghee Lee
Challenges [5]

High-level Service Description Languages
– powerful abstraction constructs to enable stakeholders to describe
service-based applications in terms of what they require.
– allowing all kinds of parties to develop
– Process execution language, e.g., BPEL.
» hard procedure that requires formally trained software developer
– High-level process fragments to facilitate reuse
– For complex composite service-based applications
– High-level language support for adaptation

Service Design Theories
– To deal with large scale, dynamic, global reach and open serviceenabled applications and systems: Modelling Languages,
– To provide suitable interfaces to allow runtime testability and
debugability.: Extended service interfaces
Prof. Younghee Lee
Challenges[5]

Service Domains
– To understand dynamic problem domains, and to incorporate domain
knowledge
» Models of dynamic problem domains
– Problem domain aware reasoning and search techniques
» Future service discovery mechanisms must take knowledge about the
problem domain into account
» Now: precise queries + domain knowledge from search engines such as
Google.
» In the future: automated discovery => tightly coupled with problem domain
knowledge: large scale context aware semantic service discovery
» problem domain knowledge could be retrieved from diverse sources (from
Wikipedia pages to domain-specific reference models)
– Domain specific composition and monitoring technologies
» Natural language parsing techniques to be applied to match formal and
informal service queries to informal problem domain descriptions such as
Wikipedia.
Prof. Younghee Lee
Challenges[5]

Service Platforms
– virtualize vast numbers of heterogeneous computing systems and
resources: to share
– flexibility to adapt to application-specific and contextual needs:
preserving QoS and SLAs.
– Customization support:
» provide an application-specific response by discovering, coordinating and
composing independent matching services in a dynamic manner
» optimize itself in accordance with the specific application requirements

Service Quality
– to guarantee that appropriate end-to-end quality of service
» Dynamic quality assurance
» Run-time assurance of service quality and service-level-agreements
Prof. Younghee Lee
Challenges[5]

Software Service Life Cycle for
– open service networks that is spontaneous ad-hoc procedures
– Change-oriented life cycle (from development => deployment)
» barriers between development (with testing) and deployment will blur:
quantifiable real-time performance, and maintaining the temporal validity
of data
– Software resource discovery, self-deployability and self-adaptability
»
»
»
»
dynamic discovery
dynamic near-optimal allocation of heterogeneous resources
end-to-end real-time QoS and SLA
“intelligent” matchmakers
Prof. Younghee Lee
Challenges

Service architecture for highly reliable and secure service
–
–
–
–

Dynamic description technique based on pi-Calculus
Quality control
User centric context aware seamless service architecture
Media service and content oriented Internet service
Dynamic service creation technology based on Product
line
– User centric service creation
– Service discovery for user’s context or user’s preference
– Description and inference of user’s preference

Service based on Ambient Intelligence
– Self adaptive service infra
– Distributed system architecture of intelligence

Testbed system for development of future Internet
services
Prof. Younghee Lee
Internet requirements [2]





Openness
Scalability: use a platform which integrates billions or
trillions of entities
Dynamicity and proactiveness: adapt to changing
scenarios and contexts
No central control and governance:
Partial predictability
Prof. Younghee Lee
IT – Telco – Media convergence [2]



Telco: getting increasingly open to new services
ISP, P2P
WEB 2.0 (3.0) :
– individuals take an active ‘producer’ role
– services become community-centric.

Telcos are migrating to IP-based system (IMS)
– to increase average revenue per user
– However, the IMS is too complex to permit the rapid and low cost
application development => need simpler and powerful solution

Telco 2.0
– software development kits (SDKs) to open their networks to other
service providers to create federations
Prof. Younghee Lee
IT – Telco – Media convergence [2]

Difference between telco and web models
– Telco: traditional platforms (Service Delivery Platforms) and business
models (e.g. customers pay for services)
» looking for “killer” distributed service platforms to allow them to provide
services more flexibly
» Platforms for integrating the concept of web-as-a-platform
» Providing and operating service platforms rather than just infrastructure.
– Web model: based on web-as-a-platform and alternative business
models (e.g. advertisers pay)


Telcos are investing in service-oriented architectures (SOAs)
Convergence between IT and media
– Web 2.0 and IPTV
Prof. Younghee Lee
IT – Telco – Media convergence [2]




Key application domain: digital home (using IMS), urban ?
(public space)
software-as-a-service (SaaS)
The next step: applications as services: natural one for telco
Telco: Instead of viewing service platforms as add-ons, on
top of network => standardised platform for service provision
Prof. Younghee Lee
IT – Telco – Media convergence [2]
 Key
to success? To survive or IT booming
– Virtualisation of resources: Key cornerstone for
convergence
» Network virtualisation: a solution for the telco carriers to open their
networks
» Virtualisation of service platforms: same service to be developed
once and executed on top of different platforms.
» Virtualisation of real objects : such as sensor networks, to be
located, identified and, when applicable, operated over the
internet.
» And virtualisation of end-users
– Need to expand the market: many new service paradigm
Prof. Younghee Lee
Service Scenarios [7]

future retail :
– to run a new retail store. Evaluating a suitable location for the store
running a simulation with different providers
– Census data from local market data from a research institute paying
for the simulation only based on the time that you use it

future health-care
– Enjoying Paris as a tourist in your car, sudden accident occurs.
– The car automatically alerts police and ambulance and reports
the location of the accident.
– The closest available ambulance picks you up. Your medical records
are being downloaded, analysed, translated into French

future service creation
– do not have to run your own small sized data-center, basic services..
– not worrying about scalability or seasonal peaks
– the execution of your services will happen somewhere in the Internet
and you are billed only on actual usage.
Prof. Younghee Lee
Service Architecture [8]

Service oriented architecture
– Current services emphasis assumes Internet cannot be changed.
– What if we can do services and Internet together in a clean slate
way?
– Internet does more than packet delivery
» Information dissemination
» Integration of sensor networks: Data aggregation, data-oriented
connectivity and search
» Virtualization: service customization, resource allocation & isolation
» SILO

Technology driven cross layer architecture
– Technologies offer new capabilities: how to exploit them?
» Optical networks: dynamic circuits and topologies
» Cognitive radios: flexible use of spectrum, dynamic topologies
» Location awareness
– Challenge and opportunity: how to create a workable architecture
Prof. Younghee Lee
Service and software: FP7 EU [9]
Projects & Service Framework
Prof. Younghee Lee
Service and software: FP7 EU [2]

Service platforms:
– automatic service discovery, description, composition and
negotiation
– SLA management and QoS, access rights and customer charging.
– Service aggregation platforms

Service engineering
– Methods and tools to enable faster development and support the
evolution of higher-quality lower-cost services.

Service Front Ends
– High-level, functionality-aware, network agnostic infrastructure
– Exposure of user session state information (including the provision
of session management API)
– Service brokering functions for the seamless blending of services.
Prof. Younghee Lee
Service and software: FP7 EU [2]

Virtualised service delivery platforms
– Virtualisation of service platforms
– same service to be developed once and executed on top of
different platforms



Service platforms to manage opportunities for real-time
multimedia.
SOA-Grid coupling to extend SOA for enterprise,
embedded, pervasive and real-time systems and provide
technology-independent standards for interoperability.
End-to-end solutions for instant, context-aware and
personalised service creation
Prof. Younghee Lee
Unifying Reference Architectures [9]





Extending SOA for embedded, pervasive and real-time
systems
Technology-independent standards for interoperability
(e.g., Service Component Architecture (SCA); and a
standardised enterprise service bus)
Advanced middleware (Grids, P2P, virtualisation)
Dependability, security and trust;
Means for mapping business processes to IT processes.
Prof. Younghee Lee
Service Oriented Architecture




The concept of objects => replaced by the concept of
services
S/W system that provides services to other applications
through publish and discoverable interface
Composition of services
Use of web services
Prof. Younghee Lee
Open service platform [10]
Prof. Younghee Lee
Service Description, Discovery and
Composition [5]


User specification of required services
Service creation by users
– Potential for intelligent agents in helping to compose new services
based on user profile
– Base services available in different platforms
– Mechanisms through which to match, negotiate and come to
acceptable solutions

Requirements for semantic technology
– Semantic interoperability is necessary so that service data and
information can not only be exchanged but also understood
– How to provide intelligence to the network
– Distributed ontologies (or ‘folksonomies’: user generated taxonomy)
Prof. Younghee Lee
A context aware middleware
Service provider/resources
Service user/applications
User Query/Response
Service/Info.
Security
Semantic service/resources discovery/composition
Security
Service interaction protocol
Service Context
Requirements of service interaction
Generic Routing Protocol
Location,
Service/Info.
Dynamic Group
… network context
Dissemination Routing
Interaction Routing
Unicast/Anycast/Multicast/Broadcast functions
Future Network
Prof. Younghee Lee
Privacy Protection
Composed service
User Preference,
User Context
Context
manager
Context
info
location
context
User
context
...
New Networks for future services





Active Networks, programmable routers
Content Routers for data dissemination
Data oriented networks
Networks with SoA concepts: SILO
Role based architecture
– a non-layered protocol
paradigm
Prof. Younghee Lee
SILO [12]

Services are defined generically at a typically fine level of
granularity
– E.g. in-order delivery, or CRC or flow control






Created dynamically by the SILO control agent per
application request out of methods, on-demand
Some common silo recipes can exist to replicate e.g. TCP/IP
Explicit cross-layer control and optimization by the control
agent
Ontology
Construction Agent, Management Agent, Tuning Agent
Universe of Services Storage
Prof. Younghee Lee
SILO [12]
File Xfer
VoIP (secure)
Switch
File Xfer
VoIP (secure)
Ordered delivery
Encryption
Reliable delivery
Error detection
Other service ...
Expedited forwarding
Forwarding
Forwarding
PHY
Prof. Younghee Lee
Forwarding
Conclusion
 Expand
Internet Service Market
– New Service paradigm on resources, Prosumer,
virtualization
 SOA,
Web services, GRID, P2P
 Context aware services
 New network Architecture
– Programmable router, SOA type network services,
Content based networks, Data oriented networks
Prof. Younghee Lee
Future Network?
Future of the Internet?

The Internet
We’re here
– 1st Jan 1983. === Flag day
The Internet
– ARPAnet switched from NCP to TCP/IP paradigm
– About 400 machines need to switch.
– As the Internet get bigger, it get much harder to change : Flag day?
– Maybe we reach to almost the summit of the networking system
based on the paradigm of the Internet which are “distributed control,
stateless,….”

Evolution?
– New requirements: mobility, security, NM, content dissemination…
– Small improvement, higher cost, complexity…

Future of the Internet?: No revolutionary technologies [5]
– Optimistic view: many new challenges for disruptive technologies
– Pessimistic view: networking research is at the end
– Realistic? View: continuous evolution and change
Prof. Younghee Lee
Future of the Internet?

Vision
–
–
–
–
–
–
Virtual world, virtualization of network
Connectivity oriented network -> content oriented network
Service creation by consumer : producer
Context aware networking: user context, environment context
Mobile Internet, multi-sensory information
Virtual real world with emotion [6]
[5]
Prof. Younghee Lee
References
[1] Mike Papazoglou, Klaus Pohl “Longer term research challenges in Software & Services”, Jan2008
[2] “The Future Internet: A Services and Software Perspective”, Draft to be reviewed by a working group in
Bled Feb 2008 FP7 EU
[3] “ICT-FET FP7 Work Programme: Extract from Work Programme 2007 for ICT – Information and
Communication Technologies under FP7”, 2007 EU
[4] Mario Gerla, “New Service and Architecture Requirements for the Future Internet: The Wireless, Mobile
and Sensor Network Perspective” based on the NSF WMPG report, Rutgers, Oct 2005, CCW OCT 2005
[5] “Expert Group on Services in the Future Internet, Consultation Meeting”, Brussels 12th November 2007
Directorate-General Information Society & Media, EC
[6] 김도현, 이영희, “미래 인터넷 서비스 기술,” 한국통신학회지 (정보와통신) 제25권 제3호, 2008. 3.
[7] NESSI Strategic Research Agenda, Vol. 3.FP7-2.exec, NESSI Roadmap. Public -Version 29, February
2008
[8]: GENI: Global Environment for Networking Innovations, “To Reinvent the Internet”, Guru Parulkar,
CISE/NSF, Nov. 2006
[9] http://cordis.europa.eu/fp7/ict/ssai/internet-services_en.html
[10] NEXOF-RA Project 216446, FP7 EU May, 2008
[11] Chris Brealey, “SOA Research Challenges:Present Day and the Future Service Component
Architecture”, CASCON Oct. 2008
[12] Ilia Baldine, “Future Internet Architecture: Issues and non-Issues The SILO perspective” NC Univ.
Prof. Younghee Lee