Machine Learning for the Knowledge Plane
Abstract
In 2003, Dave Clark (MIT) proposed a new abstract internet layer called the Knowledge Plane (KP) whose purpose is to support network applications such as general fault management, self configuration, and performance management that require distributed, self-aware cognitive processing. The heart of this idea is to combine techniques from machine learning with new architectural concepts in networking to make the internet self-aware and self-managing. This will require revolutionary advances in both statistical learning methods and network protocol design and implementation. We performed pilot work on the machine learning portion of the Knowledge Plane. This consisted of three components: (a) we wrote a document formulating the various learning tasks that will arise within the Knowledge Plane and within KP applications, (b) we worked with networking researchers to design several KP scenarios for purposes of developing a new DARPA-funded KP program, and (c) we participated in DARPA working meetings and planning workshops to help finalize the proposed KP program.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2006
- Accession Number
- ADA454287
Entities
Organizations
- Oregon State University