Internet traffic classification using feed-forward neural network

摘要: Many network activities can benefit from accurate traffic classification and categorization, such as QOS control, security monitoring, accounting. In this paper, a new approach based on feed-forward neural is proposed for classification, which eliminates the disadvantages of port-based or payload-based methods. Extensive experimentation comparison have been carried out to explore approach; it has found that, combined with fast correlation-based feature selection filter, better performance more results be obtained using method compared other techniques. For its good elimination accessing contents packets, technique expected promising application prospect in internet classification.