基于多Agent的可控网络安全系统研究  

Research on controllable network security system based on multi-agent

作　　者：周剑岚[1] 刘先荣[2] 宋四新[2] 

机构地区：[1]华中科技大学控制科学与工程系,湖北武汉430074 [2]中国长江三峡工程开发总公司博士后工作站,湖北宜昌443002

出　　处：《系统工程与电子技术》2008年第6期1163-1167,共5页Systems Engineering and Electronics

摘　　要：移动安全Agent扫描各客户主机的漏洞,采集记录异常活动的审计日志,实现事前和事后的安全保障,但移动Agent自身的通信和迁移的安全性同样重要。首先结合硬件特征属性密钥和用户信息,实现基于Agent技术的多因素认证系统,在认证基础上,利用非对称加密技术和密钥管理,保障Agent通信和迁移的安全性。Agent作为软件,容易受到外部破坏,采用检测代理,通过Agent的协作,利用地址解析协议对网内节点的扫描,将广域网扫描机制转化为简单易行的内网扫描,从而保障客户主机中认证Agent的部署可靠性。实验结果表明,该系统效率高,可扩展性、通用性好。Mobile security agents scan the vulnerabilities, gather the auditing logs, and implement the preactive and postactive security tasks, but it＇s also necessary to guarantee the communication and mobility security for agent it-self. With the hardware characteristics, asymmetric encryption algorithm and user privacy, the multi-factor authentication is implemented based on agent, which guarantees the communication security. Agent is easily hit from the outside, so the scanning agent detects the network inner nodes using ARP protocol, which induces an applicable scanning method, and provides a feasible approach to guarantee the deployment of the multi-agent platform. The results obtained from experiments show that the system is high efficiency, extensible and open.

关 键 词：网络安全 安全服务 多智能体系统 移动安全智能体 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]