The Future of Internet
ETPs and the Future Internet
3.5 The European Technology Platforms and the Future Internet
The e-Mobility platformhas created a Post-IP Working Group and has delivered a Post-IP white paper, focusing on the future services and the evolution of their usage, the technical challenges, the roadmap and impact, discussing the role of Europe, the ways to have a major and timely impact, and the forums and standardization bodies that should be needed.
Post-IP is not seen as a new version of the IP protocol (say, IPvXx) but the next generation Internet protocols and architecture. Internet seems to undergo a major evolution every 20 years. The first generation Internet was a research Internet and the second generation was shaped by the industry’s adoption for commerce. The third generation Internet (or the Post-IP Internet) will be the wireless-mobile Internet. Such a claim is based on the following premises: there exist over 2.5 billion mobile phones worldwide - with only 500 million wired Internet terminals – with a significant fraction (~20%) of these phones having now Internet capability through the 2.5G and 3G cellular services; in 5 years, all mobile phones expected to be Internet-capable devices.
From a mobility perspective, a major problem with current Internet is that it is not power and spectrum efficient. According to ITRS (2004 revision), power consumption in nano-electronic devices must be reduced by 75% until 2010 and by 95% until 2016. Battery technologies will need to be able to cope with the increased power needs of new VvLSI systems. Currently, CMOS technology scaling, frequency increase and augmented capacity of interconnects move at a much faster pace than the growth in battery capacity. Cost of memory continuous a decrease trend. VvLSI advances will not solve the problem as major energy consumption is due to higher layer protocols excessive signalling.
Future mobile communications will have the following characteristics/needs: they will need to be power/bandwidth limited; dynamic in nature (intermittent and dynamic connectivity); have a variety of service delivery modes (uni-, multi- and broadcast); no more one person one device but more likely one person many devices and device to device communications; cross (layer, spectrum, network) operation; self-organisation, network management and QoS management; shift from networking of networks towards services networking; multi-modal services; more emphasis on security, privacy, trust and dependability; flexible billings.
Currently an evolutionary approach to the Internet has been followed, based on patch work on the current IP and Internet architecture. Post-IP Internet should follow a revolutionary approach to mitigate the major current Internet limitations, i.e. its architecture, the adopted end-to-end paradigm, and the Internet Protocols. The Post-IP approach should define a new architecture with management capability supporting multi-domain and be wireless-friendly (i.e. to be energy and spectral efficient and be capable of supporting a variety of wireless networks, from very low power sensor networks to wide area mobile networks).
The Post-IP environment should support services differently than today: the service provisioning itself is expected to be transformed; service delivery will be context-aware; services will be proactive, open and ubiquitous; service-networks will be formed.
From a services point of view, it is expected that complexity will grow with service plurality leading to new requirements on control and management of services provisioning. Conflicts are likely to emerge between new ways of service provisioning and the existing limitations of the current IP networking environment. Consequently, future research has to provide the means to overcome the IP limitations and allow for management of new levels of complexity. In addition, a wide spectrum of horizontal and vertical services offerings should be allowed as options within a general service framework, as well as allow for new business interfaces to freely embrace the current Internet and mobile world and future service solutions, without restricting the degree of freedom the Internet provides today. Finally, future service provisioning solution have to take into consideration human mentality and customer behaviour patterns that ask for carefree services that provide discovery mechanism, personalization, guidelines on “how to use” and a recognizable party that is responsible for overall customer care.
The future Internet or “Post-IP” should be an optimized network- and service- layer solution to the ubiquitous, mobile service-enabled communications between ~1012 mobile devices and billions of human users, with guarantees and built-in orchestrated security, reliability, robustness, mobility, context, access, service support and management of the communication and services. Major technical challenges are:
The design of an overall system with integrated features in the core functionality (i.e. mobility, QoS and security) supporting intermittent and dynamic connectivity with energy and frequency constraints in mind. The creation and evaluation of alternative, breakthrough and beyond-OSI layering architectures, to overcome static layering and its disadvantages, enabling services to exploit network context information and use self-organizing technology for optimisation.The decoupling of the ID’s of information objects from their addressing, including users, nodes, documents, interfaces, domains and other structures; a key component of Post-IP is should be that its naming and addressing structure must not be restricted or tied to the architecture. The specification of a migration strategy between current and Post-IP.
In order to maximise the European impact, the Post-IP Internet should be built on the European strength and competence in mobile networks, adopting a holistic approach to future architecture design. Effort should be coordinated across different projects with a well established support action. To this end: build on European pre-competitive research tradition, with large FP projects for consensus building and a good mechanism for preparation of future standardisation; Ensure collaboration between research projects and test-bed activities, as well as global collaboration supported by EUu and support actions in order to take initiative internationally. Finally, suitable standardisation process and forum - European driven - should be identified early enough.
According to the NEM (Networked and Electronic Media) platform , the future is basically shaped by the convergence of the future Internet and the fixed, mobile and broadcast networks evolution towards the delivery of digital content. It is about developing the future advanced access, core and delivery networks that brings the end-users the digital content anywhere, anyway, anyhow with the capacities needed and the (low) costs expected. This is already happening today at the network and the service layers.
We know enough to envisage a coherent converged future of universally available digital content and communications. We can start by designing a future where many things are possible and we can work on the enabling technologies now. First, we should discuss the structure of what we want (open the architectural forum) and keep in mind that effective and efficient networks that are flexible, controllable and manageable are going to become pervasive. We need a smaller number of larger players to achieve the operational efficiencies necessary for viability in the provision of future networks. State involvement is probably necessary to encourage and protect commercial investment, pay where there is no business case (social obligation driven) and lead with infrastructure rather than follow commercial development.
The building blocks for this are a Uuniversal Network Protocol that is capable of being the best at everything (i.e., availability, reliability, security, flexibility, variable bandwidth, manageable, QoS, virtual circuits, variable overhead), media integration (why can’t I click on my TVv screen? How can I get the news I want?), intelligence in the networks (where, how much, what for, etc.), powerful devices (with comfortable user interfaces?) and many more.
The services are now driving the networks to the limits. Thus, there is the need and opportunity to enhance them, ultimately to the point of being capable of handling all society’s networked service needs. We should start now following the foundations outlined earlier and preparing the building blocks, aiming to achieve convergence, integration, interoperability in technical and non-technical domains to make things seamless. Move on from the existing business roles and commercial domains to unlimited exciting models for the future and push the network(s) beyond all known demands. Think pan-European! Think Global!
NEM is working in this direction, working towards: the network of the Future (Future of Internet), new applications mixing media and communication, electronic content from all to all, interoperable digital rights management systems, convergence of various media and content formats, development of new middleware for media applications, the home environment, the house environment, the building environment, electronic content from all to all, terminals, devices for fixed, portable, nomadic use, security and trust.
Networked European Software and Services Initiative (NESSI), aims to provide a unified view for European research in Service Architectures and Software Infrastructures, with 9 working groups focusing on: business process management; semantic technologies; service oriented infrastructure; services engineering; services sciences; software engineering; trust, security and dependability; user service interaction; e-health.
Ubiquitous wide area networking infrastructure is at the heart of everyday life. People now expect to be connected wherever they are, whatever they are doing, and they care little about the technology involved in achieving this. IT must evolve in the same direction and provide access to rich application functionality with security, reliability, predictable performance, etc, with complexity of technology hidden. There is need for an integrated view of infrastructure involving networks, computing, storage, data, information, etc.
Service Oriented Architecture (SOA) is a paradigm for organizing and utilizing distributed capabilities that may be under the control of different ownership domains. NESSI envisages a world of services delivered via a distributed system including fixed and wireless networks, powerful computational infrastructure, diverse edge devices – computers, PDA’s, phones, sensors, etc. There is a need for a new software infrastructure for composition of services, which is flexible, predictable, cost-effective and enabling a rich ecosystem of services. Finally, intelligent composition and efficient delivery of services is at the core of the NESSI vision.
Today’s Internet service model is based on simple packet transport and functionality is pushed to the edge. Simplicity is a key to its success. It is essentially best-effort.
In the future Internet, the infrastructure service model should support manageability. NESSI core services supplement with richer, more diverse functionality. Simplicity should be retained for users/applications. It should have predictable non-functional properties.
NESSI needs future Internet but will not build it. Requirements in order for the future Internet to support the NESSI vision are: converged resources architecture (networks, middle-boxes, computing, storage, edge devices…); virtualisation and appropriate levels of abstraction (well-defined interfaces (functional and non-functional properties) and manageability (e.g. support for failure detection and mitigation); flexibility; mobility; appropriate security options. NESSI will support new applications but will not build them.
There is a need for broad collaboration. Broad interoperability is needed between products, technologies, service providers, across international boundaries and for a controlled sharing of data and information. End-to-end properties of composite applications should be assured, should be delivered by multiple stakeholders and should be predictable and manageable. Open standards for interfaces, protocols, and data are essential but there is also a need for a coordinated approach to adoption, as cultural, policy and regulatory issues are also important.
Concluding, the convergence of IT and networking is at the heart of the knowledge economy. NESSI sees an approach based on services as fundamental and welcomes the broad consensus developing around this view. NESSI is working on an open framework enhancing flexible infrastructure with higher-level services. Coherent approach to services requires a cooperative approach and NESSI is showing that a multi-stakeholder approach works. NESSI contribution to the future Internet is primarily related to dynamic service composition platforms.
For the Integral Satcom Initiative (ISI)the future Internet design should address the wide network heterogeneity, the ubiquity requirement (global connectivity anywhere and anytime), pervasiveness, dependability, security and trust, as well as for end-user devices supporting a large variety of different access technologies via a common set of standardised network protocols, capable of optimally managing multiple connections in parallel.
The role of satellite and space segments of the future network will be to provide truly ubiquitous and broadband connectivity. This is done today for fixed terminals located in remote areas and portable stations and is soon expected for mobile collective platforms such as aircraft, vessels and trains and emergency and disaster scenarios. (In the future, it is expected to provide complementary direct access for 4G end-user devices, exploiting satellite inherent broadcast and multicast capabilities and for interplanetary networks.
Delay Tolerant Networks (DTN’s) is usually a network of regional networks, i.e. an overlay on top of regional networks, including the internet. Its basic features are: intermittent connectivity, long delay with very high jitter, large asymmetry and high error rate links. For such networks major challenges are to improve efficiency of TCP convergence layer (restart on disconnect, slow start mechanism), design appropriate security mechanisms, research solution to routing problem.
Challenges addressed by ISI are: design and deploy hybrid terrestrial/satellite systems to reduce transmission costs, increase efficiency and flexibility, provide dependability; increase the spectrum and power efficiency of satellite systems by at least an order of magnitude with respect to the state-of-the-art; enable the exploitation of higher frequency bands (Kka and above); align the form factor and power consumption of user terminals to those of terrestrial systems, in particular for mobile services; introduce software defined radio and cognitive radio concepts and architectures; improve network integration and quality of service support; lower the cost of terminals, networks, space segments, services; achieve the harmonization of regulatory frameworks and fair regulation for ground components and spectrum usage; strongly pursue open standardization of technology.