Abstract:
This project aims to provide an infrastructure for distributed reasoning over multiagent systems.In particular, the project has focused on fault diagnosis in telecommunication networks. Giventhe uncertainty over diagnosis and the distribution of data, this project has continued an earlierproject, which proposed the use of Bayesian networks to manage uncertainty.In this project, we deal with one of its limitations: the inability to manage a single Bayesiannetwork in large communication networks. For this purpose, we have worked on the study ofMultiply Sectioned Bayesian Networks, which facilitate their distribution in different nodes.The project has proposed a communication mechanism between the nodes to maintainconsistency, which has been implemented using a JGroups middleware multicasting. To validatethis model, it has been applied to diagnosis in FTTH scenario, integrating the reasoning system ina multi-agent platform deployed over a simulated FTTH scenario.