SEGURANÇA E CONFIABILIDADE EM IDS BASEADOS EM AGENTES.

Abstract

Lack of security is a constant concern in open distributed systems. Threats are present within environments insecure, uncertain and constantly changing. Due to this problem, many tools for evaluating vulnerabilities of the network as well as for their protection are being developed as techniques for encryption and software systems such as antivirus, firewall and IDS (Intrusion Detection System). Among these, there are IDS systems that are being conceived, designed and implemented, using techniques executed by agents. However, it is necessary to assure security and reliability of exchanged messages inside IDS. For this purpose, this paper proposes a security solution for IDS based on agents. The proposed solution provides a methodology and a secure mechanism for communication among agents, through information protection configuration mechanisms, authentication and authorization, key control and messages persistence using XML. The proposed solution is implemented as an extension to the IDS-NIDIA (Network Intrusion Detection System based on Intelligent Agents), whose architecture has an intelligent agent society that communicate in a cooperative way in a distributed environment. The implementation of the prototype and tests proposed in this work show the applicability of the proposed solution.