Abstract
Abstract
Differentiated service for packets entering the network is available through packet matching. Network security and differentiated services mean an inevitable choice for routers. Recursive data flow matching algorithm (RFC) is a high performance packet matching algorithm. However, with the increase of rule dimension and scale in the rule base, system memory consumption is unavoidable. This paper lowers memory consumption via improvement on RFC by dividing the rule base into several subsets and storing each rule in a separate subset. In addition, a variety of methods are used to streamline the RFC data structure for further improvement in algorithm speed and memory performance. The experimental results show that the improved algorithm of RFC greatly reduces the overall memory consumption of RFC, while greatly improving package matching performance.
Subject
General Physics and Astronomy
Reference8 articles.
1. Scalable Packet Classification Through Rulebase Partitioning Using the Maximum Entropy Hashing;Choi;IEEE/ACM Trans. Netw.,2009
2. HaRP: Rapid Packet Classification via Hashing Round-Down Prefixes;Fong;IEEE Transactions on Parallel and Distributed Systems,2011
3. Packet classification using hierarchical intelligent cuttings;Gupta,1999
4. Packet classification using multidimensional cutting;Singh,2003
5. EffiCuts: Optimizing Packet Classification for Memory and Throughput;Vamanan;SIGCOMM Comput. Commun. Rev.,2010