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。关键词: 报文分类; 前缀匹配; 规则缩减; 规则膨胀Abstract: 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 in thezrule 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.Key words :Packet Classification; Range Matching; Rule Space; Rule Reduction; rule expansion
猜您喜欢
评论