ABSTRACT. propose a new objective of network research: to build a fundamentally different sort of network that can assemble itself given high level instructions, reassemble itself as requirements change, automatically discover when something goes wrong, and automatically fix a detected problem or explain why it cannot do so. … to achieve this goal … not sufficient to improve incrementally on the techniques and algorithms we know today.
Instead, we propose a new construct, the Knowledge Plane, a pervasive system within the network that builds and maintains high level models of what the network is supposed to do, in order to provide service and advice to other elements of the network.
the knowledge plane is novel in its reliance on the tools of AI and cognitive system … best suited to meeting the uncertainties and complexity of our objective.
Internet … emphasis on generality and heterogeneity, the “narrow-class” combination of simple, transparent network with rich end-system functionality, and the deeply embedded assumption of a decentralized, multi-administrative structure are critical strengths, but lead to frustrated users when something fails, and high management overhead with much manual configuration, diagnosis, and design.
… the network … to take care of itself rather than depending on people to attend to it. … a high-level view of its design goals and the constraints on acceptable configurations, then it could make many low-level decisions on its own. …
the ability to abstract and isolate high-level goals from low level actions, to integrate and act on imperfect and conflicting information, and to learn from past actions to improve future performance.
previous attempts to do “high-level network management” … fail
… one possible reason is that …
… its ability to host a broad and changing array of applications … devise separate construct that creates, reconciles, and maintains the many aspects of high-level views, and then provides services and advices as needed to other elements of the network.