S^3: Packet Classification Using Simple Split Header Structure

摘要

Due to the complexity of the packet information search at router, packet classification is often a performance bottleneck in network infrastructure; therefore, it has received much attention in the research community. In general, there have been two major threads of research addressing this problem: algorithmic and architecture. In this paper we introduce new algorithmic direction to packet classification scheme. This novel algorithm splits the incoming rule(s) and is stored in a static data structure. A binary search tree is generated if the algorithm encounters the same rule. It is important to note that the uniqueness of the rule is judged by the source and destination address rule. We have considered five fields of IP packet header to carry on the further discussion i.e. source and destination addresses, source and destination port and protocol field.