An intelligent packet forwarding in router using self learning classifier
J Nithisha, J. Vinisha · 2017
Traffic measurements provides important role in variety of network applications like traffic monitoring, accounting and security analysis. In our day to day activities in the network lots of packets transmitted from one place to another. Router plays major role in data forwarding. Traffic may occur in router since millions of packet transmitted through router. So that, we need to monitor the traffic flow in Router. In existing system packets are forwarded as flows, needs larger counter to measure and hold per flow traffic details. In our proposed work, flows are classified based on application wise and clustering of flows used to monitor the traffic details. Deep packet inspection (DPI), and port based approaches were used for classification in last decades. In Our system, unsupervised machine learning algorithm used for classification. The incoming packets are clustered into flow group and forwarded to the router. It gives advantages and provide efficient memory utilization. Then priority is assigned to each flow set and flows are forwarded to the destination based on priority. Because of clustering of flows it provides intelligent packet delivery in router.