A Novel Framework for Network Traffic Classification Using Unknown Flow Detection

The overall goal of any organization or enterprise is to ensure optimum use of the available network resources in supporting the attainment of the operations and hence its strategic objectives. The services offered by any organization or enterprise mainly include user support services. Network Traffic Analysis (NTA) is the inference of information from observation of unknown traffic flows, for example analysis of the presence, absence, amount, direction and frequency of traffic. Traffic flow is a sequence of packets sent from a particular source to a particular unicast, any cast or multicast destination that the sources desire to label as a flow. A flow could consist of all packets in a specific transport connection or a media stream. So NTA can be useful in taking various network based decisions such as traffic routing, bandwidth allocation, etc. This paper presents a novel framework for network traffic classification which can be extended for network traffic analysis based decision making.