METHODS AND TOOLS FOR NETWORK TRAFFIC CLASSIFICATION

Since the first releases of intercommunication models between end-devices, like ring topology, the whole picture of now days network infrastructure was changed unrecognizably. Modern systems of networking consist of many complex intermediate modules like switches, routers, firewalls, hubs etc. and the main goal of these inventions was to provide more reliable and scalable ground for communication (Quality of Service). Meanwhile, rapid growth of traffic on the Internet forced network engineers and software reliability engineers to pay more attention on the optimization of data flow from both sides, developing network-oriented software and application-oriented network. To apply effective solutions on these tasks, engineers need to research specifics of the current network state. The more whole system evolves, more data about network traffic we gain, and now it helps us to make optimization and tuning of intermediate devices, rather than just scaling it up with more bare hardware. Which protocols are used the most? What types of applications loads the network bandwidth the most? and etc. Classification of packets can help resolve the answers, and there are different approaches to achieve this. In this study, I tried to explore already known tools and methods that can be applied to solve such tasks.