Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

European Research on Future Internet Design

Peter Stuckmann, Rainer Zimmermann

European Commission, Directorate-General Information Society and Media,

{Peter.Stuckmann|Rainer.Zimmermann}@ec.europa.eu

Abstract. Future Internet has become the federating theme for European research on communication networks and
services. In the core lies research on communication networks towards an efficient, scalable and reliable Future Internet
coupled with research on the underlying technologies, in particular mobile and wireless access and optical networks. This
paper first presents the motivation for a bold initiative for Future Internet research in Europe. In this context a changing
business environment for telecommunications and the Internet and the opportunities to provide future Internet services
are discussed. From a technical perspective the limitations of the current Internet technology are outlined. The research
activities that address the challenge of Future Internet research are introduced under three main lines: Future Internet
Architecture and Network Technologies, Spectrum-efficient Access to Future Networks and Converged Infrastructures in
Support of Future Networks. Examples for first promising approaches to significantly change the principles of the
Internet architecture and protocols are presented.

INTRODUCTION
The European Union is presently funding collaborative research and development activities in the 7th
Framework Programme (FP 7), which is covering the period of 2007-2013 [1]. In the ICT Work Programme 2009-
2010 [2], Future Internet has become a federating theme for European research on communication networks and
services. This focus on Future Internet is motivated by the changing requirements to the current Internet that was
designed in the 1970s to support communication between computing systems for communities of expert users. It
was not designed to cope with the wide variety, and the ever growing number of networked and mobile users and
applications, business models, edge devices, networks and environments that it has now to support. Its structural
limitations are increasingly being recognized world-wide.
Clean slate or evolutionary approaches or a mix of those can be equally considered. The evolutionary approach
builds on the evolution of the current existing Internet to conceive pragmatic and viable solutions for commercial
rollout [3]. A clean-slate approach eliminates legacy Internet design constraints. Both approaches target the same
usage vision and will have to be synchronized on phased agendas. Once promising clean-slate architectural concepts
have been identified and evaluated as fulfilling the design goals, evolutionary paths for deployment have to be
identified. Since the current Internet has grown to become so large, it will be commercially and operationally very
challenging to introduce new architectural principles [4]. One deployment possibility is virtualization, which would
enable logically independent networks built on a common physical infrastructure. In this case new network
functionalities and protocols could be deployed, but also specialized networks could be provided by building overlay
or underlay techniques running new protocols on top of or below the network and transport protocols. In this
transition scenario a parallel network can be run for applications that truly need the improved functions. Users or
providers would migrate to the new system over time, similar to the way some are now abandoning the traditional
telephone system for Internet-based phones, even as the two networks run side by side.
A big challenge in clean-slate research on Future Internet is the need to evaluate the designed architecture.
Whereas analysis and simulation can allow a first performance evaluation, the use of prototypes is crucial, as a
system has to be built in order to evaluate it and to convince others that it is the appropriate solution. For global
systems as the Internet, it is almost impossible to get a new idea adopted that has not yet been tried at scale and
under realistic conditions. In addition, an experimental facility would be useful intellectually as it enables the
researchers to uncover things that would otherwise have been omitted [4].
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

Besides the necessities and opportunities that lie in network research, key innovations in underlying technologies
such as mobile and wireless communications and optical networks are expected to drive the development of the
Future Internet. Therefore these aspects have to be studied jointly with network architecture research as they build
together the future network infrastructure foundation. Modern communication infrastructures will be characterised
by mobile and wireless broadband access on the last mile or meter and the interconnection to ultra-high capacity all-
optical networks.

CHANGING APPLICATION REQUIREMENTS FOR THE FUTURE INTERNET

The Internet has emerged to a critical infrastructure for society and economy as a whole, similar to any other
utility, e.g., infrastructure for electricity and water supply. Society is undergoing a paradigm shift, the evolution of
the society and the Internet being now tightly interconnected. Daily life factors including health, transport,
knowledge, and culture rely increasingly on the Internet in the developed world and it is bringing economic
development of emerging economies.
In addition to 1 billion fixed Internet hosts expected to be connected by 2011 it is expected that mobile and other
type of handheld devices will be directly connected to the Internet leading to about 3 billion connected hosts by
2011 [3].
Since the Internet was designed for fixed terminals, it shows inefficient behavior for mobile and nomadic
terminals. Therefore the generic and efficient support of mobile terminals and mobile applications is one of the
major design goals of Future Internet architectures and technologies.

Besides the increase in number of users and connected devices expected, new applications requirements for the
Future Internet are emerging.

High-quality and Shared Content Dissemination

As digitalization of data progresses, it is now expected that the majority of new media will arrive in digital form,
with the analogue form being the exception. For instance, digital videos will not only increase in number, but also in
size, due to increases in resolution and the ease of creation and manipulation. The increase of number of digital
videos and their distribution over an increasing number of locations creates the need for specific multimedia search
engines. Progress in network multimedia communication is also leading to 3D videos, virtual reality and gaming.
Digital TV channels are also progressively penetrating the Internet space.
Not only high-quality content is made available over the Internet by large content providers. Users are also
enabled to easily produce, offer, share and consume content on the Internet and are becoming "prosumers". Whereas
communication will remain an important service to be supported by the global network, dissemination of content,
either distributed by content providers or made available by prosumers, is expected to be one of the main functions
of the Future Internet.

Connecting Objects and Things

While the current Internet is a collection of rather uniform devices, it is expected that the "Internet of Things"
will be characterised by a much higher level of heterogeneity, as objects of totally different in terms of functionality,
technology and application fields are expected to belong to the same communication environment. The Internet of
Things can be defined as "a world-wide network of uniquely addressable and interconnected objects, based on
standard communication protocols". This enables applications involving real-world objects, but also business
applications based on network-assisted machine-to-machine interaction.

Service-oriented Internet
Whereas a lot of computing and storage applications are today still executed locally on end user devices such as
PCs, a service-oriented Internet would allow the access to complex physical computing resources, data or software
functionality in the form of services. One example is the "cloud-computing" approach to infrastructure services,
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

where large-scale data centers provide virtual execution and storage environments as Internet services with the same
functionalities to physical machines but far greater flexibility and scalability.

These sets of expected applications for the future Internet are only examples; the target architecture should
generically and flexibly support different requirements and traffic patterns.

FUTURE INTERNET SERVICE PROVISIONIONING

Today the telecom and the Internet industry are structured differently with regards to applications. On the one
hand in the telecom industry, applications are network infrastructure-centric including many standards. Applications
could look similar but are expected to be fully interoperable. On the other hand, in the Internet industry, service
providers are mostly answering a customer need without paying attention to the interoperability and relying on
lightweight infrastructure. Network infrastructure providers (ISPs) get their main revenue from end-users who pay
for network connectivity. Service providers, e. g., Google, get their main revenue from advertisers. In the past, user
micro-payments have proven to be too unpredictable and too much of a burden to be acceptable to users. History has
shown that users usually prefer flat rates or subscriptions.
The Future Internet shall provide a twofold path: enabling to focus on a specific user need and to develop a
solution without paying attention to the network infrastructure. Once a leader on a specific application emerges, the
Future Internet should also enable the growth of an eco system of players being able to build their own applications
on top of its system.
In the Future Internet, access to the network will be made available ubiquitously and connectivity becomes a
fundamental service that communities use and rely upon. The transmission technologies and access networks will
increasingly be designed and deployed for horizontal integration and service-agnostic platforms, as addressed under
the "technology neutrality" paradigm where services can be provided through different networks and different
technologies.

Bit Pipes and Advanced Network Services

The provision of "a bit pipe" is becoming a commodity business as end-user services are separated from the
network infrastructure. However, based on the pure data forwarding service provided by the communication
infrastructure, various kinds of advanced network services can be established. While the focus of today's Internet is
mainly on non-real time messaging, web browsing and multimedia, it is expected that new applications will demand
for new capabilities from the networks. Bulk data transfers, and real-time data transmission and information delivery
are examples of high interest especially for industrial applications. Interconnecting more than computing machines,
e.g. sensor networks will push connectivity needs that have nowadays reached the limit of the original design. The
network operator is also well positioned to provide context information to service providers for future innovative
applications, e.g. based on location information. For content dissemination, already today video download service
providers are buying content distribution network (CDN) services from infrastructure providers in order to be able to
provide their content in acceptable quality. As user numbers and the amount of content are expected to rapidly
increase, this kind of support service for efficient content dissemination is promising to have further potential.

Controlling the Internet

Today's Internet control architecture implicitly restricts business to a set of primitive models such as settlement-
free peering or flat-rate subscription. Any more sophisticated approach has either to be implemented as a "walled-
garden" losing the benefits of global interconnectivity, or built by abusing the current architectural components,
using ad-hoc solutions for provider independence [5].
It is important that the Future Internet is designed to accommodate conflicting interests, so called "tussle
networking" [6] such as conflicting policies, traffic patterns and compensation modes. It is fundamental to recognize
the powerful capability of the current Internet to accommodate new applications developed by an increasing user
community. It is thus essential to keep the entry barrier as low as possible and design the Future Internet so as to
allow and steer open and innovative application development without impeding the Internet genericity, evolvability,
openness and accessibility. The Future Internet shall thus cultivate the opportunity for new players to take benefit of
the infrastructure foundation but also the pillars of the Future Internet without sacrificing on its global architecture
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

objectives and principles. A Future Internet should be able to adapt to the changes in society's demands on the
Internet as they occur without requiring permanent redesign. Additions and extensions to the network architecture
should be facilitated without rigorous standardization processes but also without replacement of the infrastructure
equipment.
This approach of "Design for Tussle" is also motivated by today's tussle over resource control by users and
network providers. Network providers offered several pricing schemes but flat-rate prevailed. Users of peer-to-peer
applications took advantage of that and by increasing the number of their parallel transport connections absorbed all
the bandwidth of the network making performance dismal for the traditional interactive web-browsing users. To
prevent congestion in their networks, operators introduced techniques such as port-selective packet dropping and
deep packet inspection, while users reacted by encrypting their traffic.
Besides filtering of traffic, operators are also starting with differentiated charging for transporting the content.
Operators want to charge higher fees for traffic belonging to premium customers. Advocates of net neutrality see
this to be against the principle that all Internet traffic should be treated equally. The principle is seen as the basis for
the preservation of current freedoms and for the Internet as an engine for innovation and creativity where business
services and user-generated content can be easily offered and accessed with global reachability.
This discussion is closely linked to the governance aspect that will become an important part of the Future
Internet, in particular if it comes to issues with a public policy component addressing aspects like privacy, security,
freedom of expression, intellectual property rights, and data protection. The term "Internet Governance" emerged in
the early 1990s to describe elements of the management of critical Internet resources, in particular the domain name
system (DNS), but also IP addresses, root servers, and Internet Protocols. It was based on a conceptual
understanding of self regulation and private sector leadership, of "Governance without Government". One concrete
result of this discussion was the establishment of the Internet Corporation for Assigned Names and Numbers
(ICANN) in 1998 and as the next step the United Nations Working Group on Internet Governance (WGIG) in 2003
and the Internet Governance Forum (IGF) in 2006 [7].

TECHNICAL LIMITATIONS OF THE CURRENT INTERNET

Internet technology can be characterised by its design principles, namely layering, packet switching, a network of
collaborating networks, intelligent end-systems as well as the end-to-end argument [8, 4].
It has been simple to link any new network to the Internet, providing instant benefits resulting from the
interconnectivity with a huge range of communicating peers. The transparency of the Internet has facilitated the
deployment of successively more complex network-agnostic applications and services. Together, these two
attributes characterise the "hourglass" approach to protocol architecture – the network layer, the waist of the
hourglass, is simple enough to operate on top of and integrate any link technology, and transparent enough that
anything can run over it (see Figure 1). The hourglass approach has led directly to a virtuous circle of increased
network reach enabling new styles of usage and vice versa.
However, with its increasing success, the Internet architecture is progressively losing its original simplicity and
transparency. One of the main causes are the raise of new classes of applications, additional operational and
management requirements, variety of business models, security mechanisms and scalability enablers that give rise to
ad-hoc solutions that extend the architecture without regard to the original key design principles.
Examples are firewalls to support end-user and site security and Network Address Translation (NAT) to cope
with the exhaustion of IPv4 address space. There is, however, a growing consensus among the scientific and
technical community that the methodology of continuously "patching" the Internet technology will not be able to
sustain its continuing growth at an acceptable cost and speed.
This loss of flexibility is already being felt as the number of Internet nodes grows another order of magnitude.
The size and scope of today's Internet make the deployment of new network technologies very difficult while
experiencing increasing demand in terms of connectivity and capacity. This situation where technological
innovation meets natural resistance is called "ossification". Examples are the slow deployment of technologies such
as multicast or Internet Protocol version 6 (IPv6). Innovation has happened mainly in the applications and in the
underlying transmission technologies, rather than in the core technology, the network and transport layers TCP/IP
(see Figure 1). The following technological limitations have to be overcome to meet the future challenges for the
global communication network.
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

Figure 1: Innovation and Ossification in the Internet

Limited Support of Mobile and Wireless Terminals

While the Internet was designed for stationary computers, today laptops and smart phones are constantly on the
move. With today's technology, a laptop changes its address and reconnects as it moves from one wireless network
or access point to another, disrupting the data flow. Alternatively, the Internet standard Mobile IP allows routing all
traffic back to the first access point as a laptop moves to a second or a third location, but delays and inefficiencies
could result. As a clean slate solution, the address system would have to be restructured so that addresses are based
more on the device and less on the location. This way, a laptop could retain its address even if it moves from one
wireless network or access point to the other.

Lack of Built-in Security

With the evolution towards Internet-based services, traditional telecom networks as backup for mission-critical
services are expected to gradually disappear. Therefore, built-in security mechanisms are one of the main goals of
Future Internet design.
The Internet was designed to be open and flexible, and all users were assumed to be trustworthy. Thus, the
Internet protocols were not designed to authenticate users and their data, allowing spammers and hackers to easily
cover their tracks by attaching fake return addresses onto data packets. Internet applications such as firewalls and
spam filters attempt to control security threats. But because such techniques don't penetrate deep into the network,
bad data still get passed along, clogging systems and possibly fooling the filtering technology.
The network would have to be redesigned to be sceptical of all users and data packets from the start. Data would
not be passed along unless the packets are authenticated. Faster computers today should be able to handle the
additional processing required within the network.

Scalability Issues
In addition, the vision is also under threat from basic engineering problems. The routing system, which is the
single most critical part of the Internet infrastructure, is facing significant scalability issues [4]. At just the time
when the Internet is becoming critical infrastructure, the core protocols may become increasingly fragile as more
manual configuration is needed to avoid cascading problems due to overload, accidental mis-configuration or attack.
The IPv6 standard allows expanding the address pool, but nearly a decade after most of the standard was completed,
the vast majority of software and hardware still use the older IPv4 technology. Even if more migrate to IPv6, not all
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

addressing issues would be solved. Researchers are questioning whether all devices truly need addresses. Sensors in
a home could communicate locally and relay the most important data through a gateway bearing an address. As
routing and addressing are becoming the main challenge of a global network with billions of users and objects,
scalability is one of the most critical design criteria for Future Internet architectures.

Performance and Quality of Service Challenges

While mechanisms for providing Quality of Service (QoS) within the Internet as well as Asynchronous Transfer
Mode (ATM) networks have been very well studied, the interaction problems between the network layers are still
unresolved and the management of such services, including configuration, policy setup, charging, inter-provider
setups, etc. is still a challenge [4].

RESEARCH PROGRAMME ON FUTURE NETWORKS

In the ICT Work Programme 2009-2010 [2], Future Internet has become the federating theme for European
research on communication networks and services. The planned activities under the heading "The Network of the
Future" can be seen as the basis for future network infrastructure foundations enabling future Internet services and
applications that are complementary funded under the headings "Internet of Services", the "Internet of Things", and
the "Media Internet". In addition horizontal activities like "Trustworthy ICT" and "Future Internet Experimental
Facility" are complementary parts of the programme (see Figure 2). Together with the ongoing activities under work
programme 2007-2008 the total funding for Future Internet research lies in the order of 800M€, roughly half of
which is allocated to research on future networks. The ongoing research activities and the target outcome of future
projects are structured under the following three major lines.

Figure 2: Future Internet as a Federating Theme

Future Internet Architectures and Network Technologies

The main goal of this activity is overcoming structural limitations of the current Internet architecture resulting
from an increasingly larger set of applications, of devices and edge networks to be supported. The first target
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

outcome are novel Internet architectures and technologies enabling dynamic, efficient and scalable support of a
multiplicity of user requirements and of applications with various traffic patterns, variable end-to-end quality of
service, point-to-point or point-to-multipoint distribution modes, and supporting legacy and future service
architectures. The target architecture should support personalised rich media networking, machine-to-machine
communication, wireless sensor networks, ad-hoc connectivity networks as well as personal and body area
networks. It should also be wireless-friendly, natively support mobility, be spectrum- and energy-efficient, support
future very-high-data-rate all-optical connections as well as heterogeneous wired/wireless access domains. Routing
and location-independent addressing or naming, dynamic peering, signalling, resource virtualisation, and end-to-end
content delivery techniques are related research issues.
Besides the architectural concept, flexible and cognitive network management and operation frameworks are
needed. They are expected to enable dynamic, ad-hoc and optimised resource allocation, control and deployment,
administration with accounting that ensures both a fair return-on-investment and expansion of usage, differentiated
performance levels that can be accurately monitored, fault-tolerance and robustness associated with real-time trouble
shooting capabilities. The management architecture should target self-organised and self healing operations,
cooperative network composition, service support and seamless portability across multiple operator and business
domains.
Migration paths and coexistence through overlay, federation, virtualisation and other techniques should be
investigated to support several network and management architectures including legacy systems. Benchmarking
capability of the proposed architectures is to be considered from the onset. Clean slate or evolutionary approaches or
a mix of those can be equally considered.

Spectrum-efficient radio access to Future Networks

One of the important infrastructure foundations of the Future Internet will be next-generation mobile radio
technologies. They should be designed to be cost-, spectrum- and energy-efficient and adapted for implementation in
future high-capacity mobile radio systems. Key technology building blocks expected to be addressed are adaptive
modulation and coding schemes, multiple antenna and user detection schemes, cross-layer design and low–latency
transmission schemes. They are expected to be complemented by co-operative technologies at base station and/or
terminal level, novel network topologies and related dynamic channel modelling and estimation. Integrated projects
are expected to take a comprehensive approach to the key technology building blocks and develop system evolution
paths by jointly designing radio transmission techniques and radio interface protocol stacks and considering
spectrum co-existence and sharing.
Cognitive radio and network technologies are to be developed in order to reduce the management complexity and
enabling seamless service provision in a radio environment with a large number of heterogeneous radio access
technologies. These should support environment-aware, self-reasoning- and learning-capable mobile devices that
can change any parameter or protocol based on interaction with the environment with or without network assistance.
As a complement to cellular systems, novel radio network architectures enabling the innovative usage of
licensed, unlicensed or unused radio spectrum with the aim of radical cost- and energy-reduction are to be
researched. Target environments range from short to medium distance including systems based on femto-cells, ad-
hoc networks and vehicular networks, up to wide-area terrestrial and satellite-based radio access networks.

Converged infrastructures in support of Future Networks

The second important infrastructure foundation that comes before the last mile or meter, are ultra-high-capacity
optical transport and access networks. They are expected to be based on state-of-the-art photonics with transparent
core-access integration, optical flow and packet transport, dynamic wavelength allocation and end-to-end service
delivery capability. They should overcome the limitations of segmentation between access, metro and core networks
and domains, enable lower cost optical access and address the need for energy efficiency. Integrated projects are
expected to address also a network control plane supporting flexible management capability of multi-domain and
multi-operator contexts with end-to-end carrier grade performance.
Another objective is converged service capability across heterogeneous access. Concepts should be developed
that go beyond incremental steps in service platforms. What is needed are breakthrough technologies and
architectures for seamless ubiquitous broadband services, integrating wired and wireless, fixed and mobile
technologies in hybrid access networks, including hybrid-satellite networks. These enable generic support for
service portability and continuity across composite networks through the service-network interface, with ubiquitous
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

access from any network, from any technological or administrative domain, from any location and with a variety of
access devices.

Future Internet Research Coordination

While in Future Internet research a bottom-up approach is needed to generate ambitious and long-term ideas,
concepts and solutions, coherence is needed to generate critical mass. The coordination structure for European
Future Internet research is characterised by 3 pillars (see Figure 3). Academia is expected to play an important role
in clean-slate thinking and major contributions are expected towards new architectural concepts. After an initial
expert consultation in March 2006 [9], the EIFFEL think tank was established in October 2006 for discussing and
exchanging ideas and research trajectories on the future of the Internet architecture and governance building as a
foundation of the future networked society [10]. To have a significant impact, industry leadership is needed and the
programme was already successful in the first call in 2007 in attracting the major industry players to take up the
challenge of Future Internet research. The European technology platforms in the area of networks, satellite
communications, electronic media, software and services and smart systems have joined forces and have drawn up a
common vision towards a public-private partnership on Future Internet technologies [3]. To provide a forum for the
ongoing projects to exchange results and develop synergies, the Future Internet Assembly (FIA) was established that
is especially focused on cross-domain discussions to bridge the gap between the different research communities of
networks, services, content, and network security. The Bled conference in March 2008 was the kick-off of the FIA,
where the Bled declaration [11] was signed by 63 European Research Projects and European Technology Platforms
to express their commitment and call for European action towards the Future Internet. A first milestone has been the
publication of the book "Towards the Future Internet – A European Research Perspective" presenting first project
results [12].

Figure 3: Structure of Future Internet Research Coordination

 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

FIRST APPROACHES TO FUTURE INTERNET DESIGN

Research on Future Internet architectures and technologies has already started. Both industry-driven integrated
projects generating critical mass and focused projects concentrating on specific ideas have been launched in the first
phase of FP7. The project portfolio is structured in the three clusters Future Internet, Radio Access and Spectrum
and Converged and Optical Networks (see Figure 4). In the following, examples for first promising approaches to
significantly change the principles of the Internet architecture and protocols are presented.

Figure 4: Future Networks Project Portfolio and Clustering

Information-centric Paradigm and New Addressing Concepts

While today's Internet is based on a node-centric paradigm, several projects are adopting 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.

Networking across multiple paths and layers with the Generic Path

The project 4WARD [13] is integrating much of the functionality of peer-to-peer overlays, by including caching
functions where the "copies" are treated as the originals. This is done in a common and open information networking
service generalised for use by applications.
The current semantic overload of the IP address as both node identifier and locator, indicating the current point
of attachment in the network topology, is replaced by a clear separation of information self-certifying object
identifiers and locators. While previously proposed models for abstracting the location and focusing on networking
between (mobile) hosts are considered, the project is designing a network architecture where mobility, multi-homing
and security are an intrinsic part rather than add-on solutions.
The notion of a Generic Path (GP) is defined as a framework being able to efficiently realize "networking of
information" by exploiting cross-layer optimization and multiple network paths. A GP is defined as means to
organize the accessibility of a sufficient number of parts or copies of information objects stored in a group of hosts.
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

Incorporating the paradigm of information-centric networks means that a GP is actually hiding the physical location
of information objects. It does not matter where chunks or copies of information are stored, but the GP takes care of
delivering it to the destination. Because cross-layer information is available, new transmission techniques can be
used inside a GP. This is especially interesting for the introduction of network coding into fixed and wireless
networks. Here, multipath routing needs to be combined with specific capabilities of nodes. An important advantage
of this concept is that mobility of information objects and hosts becomes conceptually equivalent and is dealt with
by the GP internally.

The Pubish/Subscibe Routing Paradigm

The PSIRP project is redesigning the entire Internet architecture based on the publish/subscribe routing paradigm
[14]. In this already existing concept, senders “publish” what they want to send and receivers “subscribe” to the
publications that they want to receive. In principle, no one receives any material to which they have not explicitly
expressed an interest by way of subscription. By explicitly subscribing to information, the subscribers offer their
resources to be "used by the sender" which means that resources are added to the communication in a scalable
fashion as the number of subscribers grows. In such a new Internet, multicast and caching will be the norm and
security and mobility will be designed directly into the architecture, rather than added as after-thoughts.
PSIRP’s work is focusing on the intersection of security, routing, wireless access, architecture design, and
network economics, in order to design and develop efficient and effective solutions. The architecture also addresses
multicast and caching functionalities to achieve optimal operational efficiency, as well as embodying security and
mobility directly into the foundations of its design. Moreover, a system of identifiers and labels, combined with
methods for scoping information, allows information to be addressed as opposed to end-hosts, effectively providing
a locator-identifier split. The result is a powerful yet flexible infrastructure with a high degree of resilience. The new
pub/sub-based internetworking architecture will restore the balance of network economics incentives between the
sender and the receiver and is well suited to meet the challenges of future information-centric applications and use
modes.
The conceptual architecture is based on a modular and extensible core, called the PSIRP component wheel. The
architecture does not have the traditional stack or layering of telecommunications systems, but rather components
that may be decoupled in space, time, and context.

Architecture based on Reachability and Forwarding Planes

The Trilogy project is developing a baseline architecture [6], which is comparable in scope to the current Internet
network and transport layers, but with a subtly different internal structure. A fundamental assumption of the
architecture is that it is based on a minimal packet delivery service. The ideal case is that packets are entirely self-
describing, meaning that other concepts such as connections, flows or sessions are higher level constructs, invisible
to the delivery service, and the network delivers each packet independently of every other. In particular, the
functions involving the networking infrastructure are divided into two planes, for reachability and forwarding, and
are distinguished from the transport services to which the network is totally transparent.
The reachability plane is responsible for hop-by-hop outgoing link selection and hence enabling network-wide
reachability. The forwarding plane is responsible for deciding how the transmission resource on each link is
apportioned between packets.
Distinct from the packet delivery service, the functions that are implemented are identified in a pure end-to-end
fashion. Transport services include functions, such as reliability, flow control or message framing that are totally
invisible to the packet delivery service. The reachability and forwarding planes are separate and together achieve the
packet delivery service. The key identifier space is the destination locator, which is handled in detail only by the
reachability plane. The forwarding plane treats locators as opaque tags which only have to be tested for equality, in
order to test for path consistency. The transport services use endpoint identifiers that label the communicating
parties, but these are totally independent of the locators used in the reachability plane.
For resource control, an accountability framework is proposed that is based on congestion volume.
Accountability enables a rational basis for sharing resources amongst users on a Future Internet that is a playground
for competing users, applications and businesses. In addition, the concept of resource pooling is introduced, which
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

enables separate network resources to behave like a single large pooled resource. Resource pooling enables better
resilience and efficiency.

CONCLUSIONS
Future Internet has emerged as the federating scheme for European Research on Communication Networks and
Services. Besides the focus on network science, underlying technologies such as mobile and wireless access and
optical networks remain the key drivers for Future Internet developments. The main challenges for Future Internet
design are support of mobile broadband applications, manageability and scalability including quality-of-service
support, security and trustworthiness as well as support of advanced high-quality content including 3D applications.
A first set of projects has been launched and first promising concepts are under development. Coordination is
ensured by providing platforms for inter-project cooperation to ensure coherence and the generation of critical mass.
Increased coordination effort is needed to further build the Future Internet Assembly. Co-operation with EU national
initiatives should be reinforced and international cooperation with regions having Future Internet initiatives is
desired.
In the next phase, projects should continue developing the technological and architectural foundations of the
Future Internet. In addition, an integrated approach to the individual innovations is needed. Conceptual results
should be evaluated and implemented in prototypes. The evaluation should show to what degree design criteria and
design goals are met. Benchmarking of architectures is needed to compare different approaches. Once architectural
concepts are evaluated as promising, evolution paths need to be studied including the assessment of their risks.

DISCLAIMER
The views expressed in this article are those of the authors and do not necessarily reflect the official view of the
European Commission on the subject

REFERENCES
[1] Peter Stuckmann et al., "Toward Ubiquitous and Unlimited-Capacity Communication Networks: European Research in
Framework Programme 7", IEEE Communications Magazine, May 2007, pp. 148-157
[2] European Commission, "ICT Work Programme 2009-2010", version agreed by programme committee, Nov. 2008,
http://cordis.europa.eu/fp7/ict
[3] Dimitri Papadimitriou, "Future Internet - the Cross-ETP Vision Document", version 1.0, Jan. 2009, available on
http://www.future-internet.eu
[4] Anja Feldmann, "Internet Clean-Slate Design: What and Why?", editorial note submitted to CCR, June 2007,
http://ccr.sigcomm.org/online/?q=node/229
[5] Louise Burness et al., "The Trilogy Architecture for the Future Internet", in: "Towards the Future Internet – a European
Research Perspective", IOS Press, 2009
[6] David D. Clark et al., "Tussle in Cyberspace: Defining Tomorrow’s Internet", Proceedings of SIGCOMM 2002, ACM Press
[7] David Hausheer et al., "Future Internet Socio-Economics - Challenges and Perspectives", in: "Towards the Future Internet – a
European Research Perspective", IOS Press, 2009
[8] David D. Clark, “The Design Philosophy of the DARPA Internet Protocols", Proceedings of SIGCOMM 1998, ACM Press
[9] Peter Stuckmann et al., "Communication Networks of The Future", consultation meeting report, CORDIS, Mar. 2006,
ftp://ftp.cordis.europa.eu/pub/ist/docs/ct/report-communication-networks_en.pdf
[10] Petri Mähönen et al., " EIFFEL: Evolved Internet Future for European Leadership", white paper from the EIFFEL Think-
Tank, Dec 2006, http://www.fp7-eiffel.eu/fileadmin/docs/EIFFEL-FINAL.pdf
[11] Bled Declaration, http://www.fi-bled.eu/Bled_declaration.pdf
[12] Georgios Tselentis et al., "Towards the Future Internet – a European Research Perspective", IOS Press, 2009, ISBN 978-1-
60750-007-0
[13] Henrik Abramowicz et al., "A Future Internet Embracing the Wireless World", in: "Towards the Future Internet – a
European Research Perspective", IOS Press
[14] Sasu Tarkoma et al., "The Publish/Subscribe Internet Routing Paradigm (PSIRP): Designing the Future Internet
Architecture", in: "Towards the Future Internet – a European Research Perspective", IOS Press, 2009
 Pre-print version of article published in IEEE Wireless Communications Magazine, October 2009

BIOGRAPHIES
PETER STUCKMANN [M] ([email protected]) received his engineering degree in 1999 and his doctor's degree in
2003 both from the Faculty of Electrical Engineering and Information Technology of Aachen University, Germany.
He is currently Project Officer with the European Commission, Directorate-General Information Society and Media, where he is
managing the research cluster on mobile and wireless communications as part of the European research programme FP7. Before
joining the European Commission in 2004 he has occupied several engineering and managerial positions both in industry and
academia in the area of radio interface engineering.
Dr. Stuckmann is the author of the text book "The GSM Evolution", Wiley & Sons, 2002, 2 book contributions, more than 20
journal and conference publications and one patent. He is member of the IEEE and the German VDE/ITG.

RAINER ZIMMERMANN ([email protected]) graduated in engineering from the Technical University of

Berlin in 1977 and obtained his Dr.-Ing. degree in 1985 from the same university. He worked from 1977 to 1985 as a researcher
for the Technical University Berlin and the Fraunhofer Society. He then joined the European Commission as a Project Officer for
projects in the fields of Production Engineering, High Performance Computing (1990) and Software (1992). Since 1995 he held
positions as Head of Unit for Telematics between Administrations-IDA, Software Systems and Services, Nanoelectronics and
Photonics. He is currently Head of Unit for Future Networks of the Directorate-General for Information Society and Media.