Currently, the Internet Protocol, or IP, is build upon a static, layered architecture as designed in the early 80s. With the exponential growth of the internet this one-size-fits-all design starts to fail us in certain aspects. As a result, the current trent in network research is to navigate away from static, layered architecture to networks that feature functional atomization, diffusion and sedimentation.
As part of my BSc degree at Lancaster University, I have been working on a current research project, Autonomic Network Architecture, or ANA for short. The project aim is to develop an autonomic network architecture, featuring flexible, dynamic and fully autonomous formation of network nodes as well as whole networks.
The overall target of the project is to identify fundamental autonomic networking principles, which would enable a network to scale not only in size, but also in functionality. These requirements lead to properties such as self-monitoring, self-repair, self-management, self-optimization and self-protection which through research are proven to be the building blocks of networks that are richer in functionality and scalability.
A new Autonomic Network Architecture will emerge as a result of this research. This architecture will provide a framework for network function re-composition. The design that enable flexible, dynamic and fully autonomic formation of large-scale networks in which the functionalities of each constituent network node are also composed in an autonomic fashion. This architecture will also allow dynamic adaptation and re-organisation of the network according to the working, economical and social needs of the users. Moreover, it must support mobile nodes and multiple administrative domains.
My involvement focused more on the Functional composition framework, the mechanism that would enable nodes in this architecture to communicate with-in a compartment, based on the functionality they required. Currently, in the Internet, this is done statically, by looking the destination address in the IP packet and routing it accordingly, without taking in consideration other parameters, such as the load of the destination.
Functional composition will enable the dynamic routing of messages, using the help of a classifier. A node that would request a service, it would query a classifier, in order to give it the destination where it should route that kind of messages at the present time. The classifier after taking in consideration various user defined variables, as well as dynamic conditions, such as the load of a node, the type of application or who is requesting the service, will redirect the node to next hop, and so on. This builds a dynamic path with modules, which are put together using the functional composition framework, that compromise this modular autonomic network.
This is just a small introduction of the ANA framework and functional composition, since its a rather large project. I will post a more detailed post on Functional Composition soon. For more information regarding the ANA project please visit the project homepage
