论文部分内容阅读
目前浮动关键词模式匹配算法的性能是IP包内容检测及过滤系统的瓶颈。而现有的浮动关键词匹配模式算法存在吞吐率低或支持的关键词数量少的问题。为了解决这些问题,文中提出了一种基于Bloom filter的改进算法。对某些短模式,采用部分无状态过滤思想,使该算法具有快速、可升级、大规模、易实现的特点。仿真试验表明:只要模式出现概率小于0.01,系统吞吐率至少可以实现9Gbps的匹配速率。
At present, the performance of floating keyword pattern matching algorithm is the bottleneck of IP packet content detection and filtering system. However, the existing floating keyword matching mode algorithm has a problem of low throughput or limited number of keywords supported. In order to solve these problems, an improved algorithm based on Bloom filter is proposed in this paper. For some short modes, the idea of partial stateless filtering is adopted, which makes the algorithm fast, scalable, large-scale and easy to implement. The simulation results show that as long as the probability of mode occurrence is less than 0.01, the system throughput rate can achieve at least 9Gbps matching rate.