基于哈希树和有限状态机的XSS检测模型

XSS Detection Model Based on Hash Tree and Finite State Machine

目前对于跨站脚本(XSS)攻击缺乏有效的防御措施。针对该问题,提出一种基于哈希树的多模式匹配模型,以快速检测XSS攻击。将质数分辨定理作为哈希散列函数的基础,对特征值进行编码,使其成为特征节点进行动态增减,以构建哈希树,并结合改进的确定性有限状态机算法提取多模式相似度攻击向量,实现多模式匹配的快速检测。实验结果表明,该模型的检测准确率较高,漏报率和误报率较低,能满足大范围XSS脚本攻击的防范要求。

Aiming at the lack of effective prevention measures against Cross-site Scripting（XSS） attacks, this paper proposes multi-pattern matching model for rapidly detection based on the Hash tree. The contributions include： Using prime resolution theorem as the basis of hash function; Adopting a series of encoded methods to the feature values; Dynamically adding nodes to build a Hash tree; Combining an improved deterministic finite state machine algorithm to extract multiple pattern similarity attack vectors for rapid detection of XSS. Experimental results show that this model can effectively prevent reflective XSS.