Security Model Of Mobile Agent System Based On Middleware

Mobile Agent technology is a promising technology.It is a newly distributed computing technology,which has incomparable advantage to the traditional distributed computing system,the security problem of Mobile agent system is a big obstacle on the way of Agent technology developing and implements in commercial domain,especially that of Mobile Agent.First of all,we study security protection aspect of some current Agent system.Its surprise that few of them aware of Agent security,and most of them integrated security technology are based on the concern of traditional attack.We also give a completely and presentation to current study on protecting mobile agent.Nearly everyone is developed in the case of some special hypothesis,and aim at particular attack behavior.Such partly protections obviously have no practical value.After comprehensively study the structure of mobile agent system,partition of the mobile agent system was reallocated,furthermore a security model of mobile agent system was proposed based on middleware.This paper study two main problems of the security model:middleware platform itself security and security technology deployment to mobile agent system.Chapter 4 was focused on former,to make sure of security,middleware platform was discussed through out all its booting, runtime and management stages.Chapter 5 was dedicated to deployment,security mechanism was embedded in three primary service of agent platform,directory service,management service,communication service.The proposed security model of mobile agent system is a general model,which core are close to mid-ware methodology.It is not only good for agent security configuration,but also good for whole system development.