Abstract: For preventing URL （Universal Resource Locator） attacks, this paper proposed a model based on artificial immunity. Applying the self-protection mechanism of biological immune system to network security, through deploying resource safely, avoiding guessing primary key, requesting valid check and business exception check, this model prevented URL attack in different levels while responding to valid request normally and made the applications acquire the secondary immunity. This paper also proposed and analyzed the key technology of implementing the model: the algorithm of avoiding guessing primary key and an enhanced static password algorithm based on random number. This model has been applied to two practical applications. The test results and the practical applications confirm the effectiveness of the model.

Key words: artificial immunity, Universal Resource Locator （URL） attack, black/white list, random number

摘要： 针对统一资源定位符（Universal Resource Locator，URL）攻击提出了一个基于人工免疫的URL攻击防范模型。该模型将生物免疫系统自我保护机制应用于网络安全，通过Web资源安全部署、数据主键防猜测处理、请求合法性检测和业务层异常检测，从系统的不同层面对URL攻击进行防范，同时保证合法的URL请求得到响应，并且在系统受到一次攻击后就会具有二次免疫能力。提出并分析了实现该模型的关键技术：数据主键防猜测处理算法和基于随机数改进的静态密码算法。该模型被成功地应用于两个实用系统中，测试结果和实际应用证明了该模型的有效性。

关键词: 人工免疫, 统一资源定位符攻击, 黑/白名单, 随机数