International Conference on Advances in Engineering Research on packet classification cased on improved cross- product method

Abstract

Packet filtering and forwarding is the core problem of gigabit network intrusion forensics system, which requires network forensics machine to classify IP packet, and packet is processed differently according to classification results. This paper presents an improved packet classification method based on cross-product. It classifies the addresses into several classes, and then cross-products these classes. To compress the large of cross-product table, this paper uses no-overlapping partition method to combine repeated rows or columns. Simulation experiments have verified the validity of our method.