Anatomy of a Secure and Scalable Multiagent System for Event Capture and Correlation

Abstract

Event monitoring and correlation across a large network is inherently difficult given limitations in processing with regards to the huge quantity of generated data. Multiple agent systems allow local processing of events, with certain events or aggregate statistics being reported to centralized data stores for further processing and correlation by other agents. This paper presents a framework for a secure and scalable multiagent system for distributed event capture and correlation. We will look at what requirements are necessary to implement a generic multiagent system from the abstract view of the framework itself. We will propose an architecture that meets these requirements. Then, we provide some possible applications of the multiagent network within the described framework.