A Survey of Clustering Algorithms for Traffic Classification

Network traffic in the world wide is predicted to increase every year double the times. So, traffic occurs in the network. To manage a network traffic classification is necessary because of increasing number of users and Qos. Classification algorithm provides a major role in traffic classification (i.e. flow or packet classification .In this paper different classification algorithms used are discussed. Traffic classification algorithm divided into supervised and unsupervised algorithm. Unsupervised algorithm uses unlabelled data to process batches of flows. So it can identifies a new classes of traffic application. Supervised algorithm works well for known dataset (flow) .Here, different classification algorithms like k-means, Model based clustering, identity based clustering, and k medians are presented.