A peer to peer traffic identification method based on support vector machine and artificial bee colony algorithm

Today, Peer-to-peer (P2P) traffic is the most important network flow on the Internet; meanwhile it gives rise to many security problems for the network management. Therefore P2P traffic identification is the hottest topic of P2P traffic management. Support vector machine (SVM) has advantages with resolving small samples for P2P classification problems. However, the performance of SVM is primarily dependent on its parameters. At present, genetic algorithm and particle swarm optimization algorithm are commonly used to learn best parameters for SVM. However, these methods are easy to run into local optimal solution. Thus, artificial bee colony algorithm is proposed to optimize the parameters of SVM and has been applied to P2P traffic identification. And the real campus traffic data is used as the research object to do simulating experiments. The experimental results show that the parameters of support vector machine optimized by artificial bee colony have high P2P classification accuracy.