The 4WARD Project
4WARD’s Strategic Objective
4WARD aims to increase the competitiveness of the European networking industry and to improve the quality of life for European citizens by creating a family of dependable and interoperable networks providing direct and ubiquitous access to information.
These future wireless and wireline networks will be designed to be readily adaptable to current and future needs, at acceptable cost. 4WARD’s goal is to make the development of networks and networked applications faster and easier, leading to both more advanced and more affordable communication services.
In our approach, we combine on one hand innovations needed to improve the operation of any single network architecture and on the other hand multiple different and specialised network architectures that are made to work together in an overall framework.
We will work
- on innovations overcoming the shortcomings of current communication networks like the Internet
- in a framework that allows the coexistence, inter-operability, and complementarity of several network architectures
- in an integrated fashion, avoiding pitfalls like the current Internet’s “patch on a patch” approach.
This work is structured into six work packages: three of them consider innovations for a single network architecture, i.e., Generic Path, In-Network Management and the Network of Information, one work package studies the use of Virtualisation to allow multiple networking architectures to co-exist on the same infrastructure, another work package looks at the design and development of Interoperable Architectures, and finally one work package that ensures that all envisaged developments take proper account of essential Non-Technical Issues.
FP7 Project Reference: 216041
Start Date: 2008-01-01
Duration: 24 months
Project Cost: 23.25 million euro
Contract Type: Collaborative project (generic)
End Date: 2010-06-30
Project Status: Execution
Project Funding: 14.45 million euro
WP1 - Business Innovation, Regulation, and Dissemination
Far-reaching technological innovations like those of 4WARD can only be leveraged if their non-technical context is taken into account.
Therefore research in 4WARD on networking technologies for the Future Internet is accompanied by research on the context of its intended usage in order to ensure maximum impact. The Business Innovation, Regulation, and Dissemination (BIRD) WP covers all of those aspects that are important in all WPs. It focuses especially on society, business, and governance issues and ensures that their respective requirements are properly addressed by all of these WPs’ technical solutions.
Various non-technical driving forces are analyzed to provide guidance to the technical research:
- Impact on existing and newly enabled business models
- Innovative and creative ways to generate value and employment opportunities
- Regulatory and public policy issues
- Application and business scenarios
4WARD technical results will be disseminated in the context of non-technical drivers, to bridge the gap between innovative research results and utilization for benefit of the economy and the society at large.
Publishing the project’s results in international journals and presenting it at conferences and public workshops as well as the publication of a 4WARD book will ensure that the project’s results are disseminated to the relevant players.
WP2 - New Architecture Principles and Concepts (NewAPC)
WP2 New Architecture Concepts and Principles (NewAPC) explores a new approach to allow for a plurality and multitude of network architectures: the best network for each task, each device, each customer, and each technology. Unlike the multitude we had in the past, where different incompatible technologies were competing with each other, NewAPC is working on developing an architecture framework that will allow networks to bloom as a family of interoperable networks. Networks will coexist and complement each other, each of them addressing individual requirements such as mobility, QoS, security, resilience, wireless transport and energy-awareness.
NewAPC’s mission and objective is to develop an architecture framework through which it will be possible to derive a family of interoperable network architectures and through which prototyping and rapid development of network architectures will be fostered. The architecture framework will comprise models, building blocks, and a design process which can be used by the network architect in order to derive suited network architectures in a smooth, efficient and simple way. Therefore, guidelines need to be developed that govern the composition of functionalities, such as mobility, QoS and security according to the desired properties and that lead to that proper building blocks can be applied by the network architect. Furthermore, NewAPC will investigate the use of invariants ito ensure interoperability, consistency and coherency for the communication system as a whole.
WP3 - Network Virtualisation (VNet)
One of the basic tenets of 4WARD is that the Future Internet shall allow multiple networking solutions to coexist, not only in the link and the application layer as in the Internet today, but also in the network and transport layers. Network Virtualisation is ideally suited to allow the coexistence of different network architectures, legacy systems included. Virtualisation is thus not only an enabler for the coexistence of multiple, possibly revolutionary, architectures, but also provides a smooth path for the migration towards more evolutionary approaches. This way, virtualisation can help to keep the Internet evolvable and innovation-friendly, particularly since it can mitigate the need to create broad consensus regarding the deployment of new technologies among the multitude of stake holders that make up today’s Internet. By decoupling the infrastructure from the services, virtualisation can provide the opportunity to roll out new architectures, protocols, and services without going through the slow and difficult process of creating such consensus.
Virtualisation further provides a general approach for network service providers to share a common physical infrastructure. This is particularly beneficial in network domains where the deployment costs per user are predominant and an encumbrance for frequent technology replacement as is the case for instance in access networks.
The goal of VNet is to develop a systematic and general approach to network virtualisation. The problem space is divided into three main areas:
1) Virtualisation of Network Resources: While the virtualisation of some types of resources, such as servers and links, is well known and already widely used today, we aim for a generalised approach that allows the use of a broad variety of resources as part of a unified virtualisation framework. Virtualisation of both wireless and wireline resources will be studied. The performance of shared resources and the secure separation of virtual networks sharing a resource will be important issues. The secure, flexible, and efficient exploitation of wireless spectrum and access infrastructure is expected to significantly improve cost-efficiency and utilisation of expensive wireless infrastructures. To allow for the integration of a variety of resources into the common framework, VNet aims to develop standardised interfaces for management and control of the virtualised resources.
2) Provisioning of Virtual Networks: Based on a substrate of virtualised network resources, and using their control and management interfaces, VNet aims to develop a systematic approach to instantiating complete virtual networks using the virtual resources, allowing the on-demand deployment of new virtual networks on a potentially large scale. The virtualisation framework includes the discovery of available physical and virtual resources, as well as the scalable provisioning, control, and aggregation of resources to form complete networks.
3) Virtualisation Management: Once a virtual network has been instantiated, management mechanisms are required to deal with the virtual resources it is based upon. These mechanisms should support the deployment, control, and dynamic re-allocation of resources on demand during the lifetime of the virtual network. A particular challenge is the dynamic management of volatile and mobile resources that may enter or leave the virtual network at any time.
WP4 - In-Network Management
In-Network Management is a novel paradigm to manage networks in the Future Internet. It will be based on a lean architecture to operate new services in the Future Internet. Discovery of network capabilities and adaptation of management operation to current working conditions are key elements in the novel management paradigm.
Traditional networks: management as external processes
In traditional Internet management, the management functionality resides outside the network, in dedicated management stations and servers. In commercial networks, interactions between these elements often occur out-of-band, through special communication networks. For emerging large-scale, dynamic network environments however, the approach turned out to be inadequate and alternative approaches must be developed.
A new Management Paradigm
In-Network Management is a new paradigm for network management, where management functions come as embedded capabilities of the devices.
With this approach, network elements have embedded “default-on” management capabilities, consisting of several autonomous components which interact with each other in the same device and with components in neighbouring devices. Glued together with a set of discovery and self-organizing algorithms, the network elements form a thin “management plane” embedded in the network itself.
The In-Network Management paradigm can be interpreted as pushing management intelligence into the network, and, as a consequence, making the network more intelligent: as a consequence, objectives and costs of management operations can be adapted according to local working conditions. The network, which now includes the management plane as a part, can execute end-to-end management functions on its own and perform, for instance, reconfigurations in an autonomous fashion. It reports results of management actions to an external management system, and it triggers alarms if intervention from outside is needed.
The architecture of In-Network management first of all models how management capabilities are embedded inside the services of a node. On this basis, it is then possible to compose them, in such a way that the embedded functions are coordinated with each other. Out of smaller autonomous components, more complex management functions can be constructed in the management plane.
The goal of In-Network Management is to achieve scalable, robust management systems with low complexity for large-scale, dynamic network environments. The guiding principles to achieve this goal are decentralization and self-organization.
WP5 - A new path abstraction
The main objective of Formux is to develop an approach and the necessary mechanisms for a generic path abstraction that can incorporate recent advances on techniques to enhance the efficiency of data transport, including use of multiple routes, exploiting the benefit of network coding, or realising diversity gains in a wireless network by cooperative coding. The generic path abstraction is also intended to provide an easy-to-use and efficient operation for both user and network, adapting transport procedures to the capabilities of the underlying network. Expected benefits are improved performance and resilience, as well as a stronger semantics such as connecting areas or information objects with each other or making concepts like anycasting transparently available.
Information transfer is currently realized by establishing TCP connections or just by sending datagrams with UDP. Paths exist implicitly through the choices made by IP routing protocols but these are largely unknown to end users and can be difficult to control for the network operator. The quality of a path is largely unpredictable due, for instance, to congestion or poor transmission quality, especially when wireless links are used. The generic path will retain the facility a simple socket-like interface as in IP while providing enhanced control facilities, as necessary, for both user and operator. The generic path will be defined in the context of the Information Network paradigm developed in 4WARD, enabling connections to information objects rather than just to specific hosts.
The generic path will provide necessary awareness of constraints at lower layers for efficient wireless communication and for optimal use of new dynamic optical switching technology, for instance. New coding techniques such as network coding and rateless source coding can significantly improve transfer efficiency and the work package will investigate how they can be exploited within the generic path abstraction. Routing is clearly an essential consideration for the quality of a generic path. Concurrent use of multiple paths for both point-to-point and swarm-like communications will be evaluated. Path selection will be aware of lower layer characteristics to ensure necessary quality and reliability for given applications.
It is clearly necessary to revisit the issue of network QoS which, despite efforts over many years, has still not received a completely satisfactory solution. Realised generic path performance depends on mechanisms such as end-to-end congestion control and router queue management that determine how they share network resources. Economic issues such as the creation of a true provider market by empowering user controlled path choice must be taken into account.
Lastly, it is essential that the generic path enable user mobility. It is envisaged that users will have an increasing range of access facilities with the ensuing requirement that the generic path should facilitate mobility management for multi-homed users over all kinds of wired and wireless interface.
WP6 - Networking of Information (NetInf)
The traditional role of networking has been to interconnect remotely located devices like computers or telephones. This function is increasingly recognised to be ill-adapted and inadequate for the information-centric applications that currently generate the vast majority of Internet traffic.
In 4WARD we take a different approach, instead of the node-centric paradigm, we adopt an information-centric paradigm. In this paradigm, the communication abstraction presented to applications is based on transfer of application data objects instead of the end-to-end reliable byte-stream used by the majority of applications today.
The current semantic overload of the IP-address as both node identifier and locator, indicting the current point of attachment in the network topology, is replaced by a clear separation of information self-certifying object identifiers and locators. Several models for abstracting the location and focusing on networking between (mobile) hosts have been proposed, e.g., the Host Identity Protocol (HIP), the Internet Indirection Infrastructure (I3), the Layered Naming Architecture and the NodeID proposal. 4WARD will build on this prior work and and by taking it one step further, we will be able to design a networking architecture where mobility, multihoming and security is an intrinsic part of the network architecture rather than add-on solutions. It will also allow users to gain increased control over incoming traffic, this enables new possibilities for defending against denial of service attacks. The self-securing property also intrinsically facilitates possibilities for effective content protection and access rights management.
The need for information-centric networking is manifested by the increasing number of overlays that are created for the purpose of information dissemination (e.g., Akamai CDN, BitTorrent, Skype, and Joost). Their objective is to distribute information by relying on users to exchange pieces of data between themselves, massively distributing the load away from any central server, and scaling automatically to any group size. 4WARD will integrate much of the functionality of these overlays, including caching functions where the ‘copies’ are as the originals. in a common and open information networking service that can generally be used by applications.
4WARD extends the networking of information concept beyond “traditional” information objects (e.g., web pages, music/movie files, streaming media) to conversational services like telephony, and store-and-forward services like email. Special attention will be paid to how this affects wireless communication and to how services can be made to work in an environment with a heterogeneous and disruptive communication infrastructure.. Also how networking of information can extend to include real world objects, and by this enabling new types of services, will be investigated.
Strategic Advisory Group, SAG
The Strategic Adrivsory Group has been set up to provide the project with the long term vision and strategic advice it needs to be successful. It is staffed by members from the partner organisations who have expertise in the subject matter of the project but are not directly involved in its work.
Strategic Advisory Group responsibilities:
It provides guidelines for the long term technical direction of the project
It reports to the Board, and the recommendations it makes are taken into account by the PMT.
Strategic Advisory Group members:
Olle Viktorsson, Ericsson
Adam Wolisz, TUB.
Staffan Truvé, SICS.
Heiner Stüttgen, NEC.
Werner Mohr, NSN.
Kurt Lösch, Alcatel-Lucent.
Roberto Kung, Orange France Telecom.
Dipankar Raychaudhuri, Rutgers University.