A Polyclonal Selection Clustering for Packet Classification

Abstract

AbstractThe process of categorizing packets into “flows” in an Internet router is called packet classification. Packet classification is one of the most difficult problems in the Internet routers. Traditional packet classification algorithms focus on the time complexity and storage complexity of the classification and the rules used for classification are fixed and can not meet the increasing network requirement. In this paper, a polyclonal selection clustering algorithm for packet classification (PSC-PC) is proposed, which can produce the rules for classification automatically. Experimental results show that the rules obtained by PSC-PC are feasible for the packet classification, and the proposed algorithm is self-adaptive and self-learning, which makes it more applicable to the network whose types of application are changeable.KeywordsDefault RateTuple SpaceAntibody PopulationHead FieldInternet RouterThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.