基于免疫计算的抗蠕虫病毒Web系统

为了高精度地检测、识别和消除Web系统中的蠕虫病毒，为了实现Web系统的故障自修复，提出了Web系统的正常模型和免疫计算方法。Web系统的正常模型是由各个组件的时空属性表示的，存储在自体数据库中，系统中所有组件的时空属性唯一确定了该软件系统的正常状态。免疫计算方法包括自体/异体的检测算法、已知蠕虫病毒的识别算法、未知蠕虫病毒的识别算法、蠕虫病毒的消除算法和受损系统的修复算法。通过数学逻辑论证，正常模型能在理论上实现对自体和软件故障的100%检测率。“人工智能”网络课程Web原型上的抗蠕虫病毒实验表明，正常模型和免疫计算对实现Web系统的抗蠕虫病毒功能是有效的和必要的，能提高蠕虫病毒检测的精度和系统修复的效率。

Anti-worm Web system based on immune computation

To detect recognize and eliminate worms in the Web system accurately,and realize self repairing of the Web system,a normal model and immune computation are proposed for the Web system.The normal model of the Web system is represented with the space-time properties of all the components in the system,and stored in the self database.The space-time properties uniquely identify the normal state of the software system.The approaches of immune computation include self/non-self detection algorithm,recognition algorithm for known worms and unknown worms,elimination algorithm for worms,and failover algorithm for damaged systems.According to mathematic logic proof,the normal model is proved necessary and important for 100% detection of selfs and non-selfs.The anti-worm experiments on the Web-based A! course system show that,the normal model and immune computation are both effective and necessary for the anti-worm function of the Web system,and they can increase the detection accuracy and failover efficiency.