摘要
在高速分组分类中,基于TCAM的硬件实现方法是一种重要的解决方案。但TCAM存储空间价格昂贵,并且要求区间匹配转换成前缀匹配,大区间报文匹配的直接转化通常会引起规则空间的膨胀。本文提出一种规则的优化表示方案,提高了区间规则表示的效率,使区间匹配转化的最坏情况由2w-2减少到w,有效节省了空间。
TCAM based hardware implementation method is the cornerstone in high speed packet classification. But the memory in TCAM is very expensive and this method has a prerequisite that range match must be resolute into prefix rule match. Always such conversion will result inthe rule space exploding. This paper presents a new method to reduce the rule space by inserting specific new rules, such method conflicts free with the direct change ,so they can be applied together. Using this way to process the rule set ,The results present that the space is effectively reduced.
出处
《微计算机信息》
北大核心
2007年第21期1-2,10,共3页
Control & Automation
关键词
报文分类
前缀匹配
规则缩减
规则膨胀
Packet Classification, Range Matching, Rule Space, Rule Reduction, rule expansion