摘要
基于TCAM的报文分类算法的关键问题在于如何高效地存储规则,而TCAM对范围形式的规则存储效率不高。文章提出了一种基于TCAM的报文分类算法——GD-TCAM算法,该算法基于格雷编码的纵向压缩,再利用TCAM的剩余位宽进行横向扩展,通过纵向压缩和横向扩展实现降低扩展系数的目的。通过利用预留表项的顺序移动法,改进TCAM的储存方式,保证分类的正确性、利于规则更新。经过理论证明和实验验证,GD-TCAM算法可以有效地降低扩展系数、降低能耗、便于规则更新。
The packet classification algorithms based on TCAM focus on efficient storing rules. However, TCAM can not store rules of range form in high efficiency. This paper proposes GD-TCAM packet classification algorithm based on TCAM. GD-TCAM conducts vertical compression based on gray code and horizontal scaling-up using the left bits in TCAM, thus to reduce the coefficient of scaling-up. By moving the reserved table entries serially, the storing method of TCAM is improved, the rules updating is convenient, and the classifying accuracy is guaranteed as well. Both theoretical proof and experimental verification show that GD-TCAM can reduce the coefficient of scaling-up and energy consumption as well as update rules conveniently.
出处
《信息工程大学学报》
2016年第6期724-729,共6页
Journal of Information Engineering University
基金
国家863计划资助项目(2009AA012200)
上海市科研计划资助项目(08dz1501600
13dz1108800)
