We group the challenges around a set of claims against which the work in PURSUIT can be evaluated on a project level. With that in mind, we formulate our objective and claims as designing, developing and evaluating a novel information-centric pub/sub-based internetworking architecture that
- Provides an improved impedance match towards application-level concepts: One of the major functions of an internetworking architecture is to provide an appropriate interface to network functions that be effectively utilised by application developers. We call this to match the impedance between application concepts and networking concepts. Such impedance match can be ‘measured’ in the required middleware approaches that map application-layer concepts, such as ontologies, agents or web services, onto the network services being exposed at the internetworking level. We assert that the PURSUIT design framework provides an improved impedance match compared to the current IP due to the used information-centric model that is exposed at its API.
- Provides tussle delineation of crucial functions: Building a distributed system requires the mediation of interests stemming from (distributed) stakeholders. The occurring tussles (i.e., the conflicts of such interests) need to be accommodated by the provided architecture in order to provide a suitable ground for innovation and overall societal viability. We assert that the PURSUIT design provides the necessary delineation of functions so as to cope with the various tussles within the network architecture and adapt to varying (in particular future) business and investment models.
- Enables optimisation of sub-architectures: The internetworking architecture provides a unifying approach for internetworking disparate systems on global scale. We believe that the current heterogeneity of technological approaches to networking in particular domains will prevail, expressed in a variety of wireless and wireline technologies. It is the clear goal of each infrastructure provider to optimally utilise the resources its infrastructure provides. Hence, optimisation of individual sub-architectures is an important design goal in order to make the internetworking architecture viable in large scale. We assert that the function separation postulated by PURSUIT lends itself to an optimisation of underlying sub-architecture.
- Provides high performance: The internetworking function is crucial to make distributed applications work in high performance. Hence, developing design choices that follow suit with the expected evolution of technologies such as optical or new wireless is essential for any adoption of the network architecture. We assert that the design choices in PURSUIT are well suited for the performance needs of the Future Internet.
- Scales to the needs of the Future Internet: Scale is essential for the internetworking function in a global Internet. The proliferation of digital technologies drives the need for scale in the future, driven by increasing digitisation of material that is being made available in the network. An example for illustrating the required scale is that of online social networks like Facebook and Twitter, where relationships (through, e.g., following tweets) reach hundreds of thousands and more per user, the user base still increasing for both platforms. We assert that the PURSUIT solutions are well suited to the scaling needs of the Future Internet from the perspective of information items as well as network systems providing the delivery of these items.
Objective: Provide a sound architectural framework for information-centric networking
- Define architectural invariants and their specific or general viability
- Translate invariants into coherent set of concepts
- Provide a set of coherent architectural arguments for their viability
- Include particular the proper (socio-)economic arguments
Design Choice Challenges
Objective: Develop a set of design choices to support our architectural claims
- Rendezvous throughout all (recursive) levels of the architecture
- Realisation of a scalable inter-domain topology formation
- Approaches to topology management (with focus on optical and wireless)
- Scalable and high-performing forwarding
- Information-centric middlewares
Objective: Provide the required proofs for our architectural claims
- Implementation (prove that it runs – and performs)
- Simulation (prove that it scales – and performs)
- Socio-economics (prove that its design is viable)
- Economics (prove that it is economically sensible)
Objective: Provide the required tools for disseminating our results
- Implementation (a tool to create a community)
- Test bed (a place to meet and try out)
- Website (a place to exchange)
- Course material (a tool to educate the new generation)
- Exploitation strategies (a tool to convince the stakeholders)
Publications and presentations are means to an end for all the above