Integrated , Distributed Fault Management for Communication Networks

This report describes an integrated, distributed fault management (IDFM) system for communication networks. The architecture is based on a distributed intelligent agent paradigm, with probabilistic networks as the framework for knowledge representation and evidence inferencing. A static strategy for generating the suggestive test sequence is proposed, based on which a heuristic dynamic strategy is initiated. Another dynamic strategy, formulated as a Markov decision problem, is also provided. To solve this problem, reinforcement learning techniques are investigated. Research supported by the U.S. Army Research Laboratory under the Federated Laboratory Program, Cooperative Agreement DAAL01-96-2-0002, by NASA under NASA Cooperative Agreement NCC3-528, and by a grant from Lockheed Martin Telecommunications.