An agent-based framework for distributed intrusion detections

摘要: Network application has become a part of our everyday life. With the increasing convenience and popularity network, more malicious users utilize network to obtain their vicious intentions. In order protect users’ information security privacy, various intrusion detection systems were proposed developed in last decade. Intrusion as an emerging technology made great achievements theory practice, whose aim is confidentiality, integrity or availability system resource. As complex system, development includes many aspects, such architecture design, design implementation components, test real cases, so on. Though have been presented, most them mainly focus on one two aspects systems. This thesis aims at providing rudimentary solution for agent-based Peer-to-Peer distributed framework. The major contributions this include following five aspects. 1. Introducing novel framework which involve different agents peers; 2. Designing functionalities each agent by using JACK/UML approach; 3. Representing knowledge about according employing ontology; 4. Developing efficient task allocation protocol used coordinate hosts collaboratively detect attacks; 5. Implementing testing reasonable manner utilizing environment, i.e. JACK . summary, integrates technology, architecture, ontology technique protocol. Implementation experiments