A survey of classification algorithms for network traffic

Abstract

Network traffic in the world wide is calculated to rise every year twice the times. To keep pace and profit from this increased amount of flows efficiently. And offer new services. Some efficient techniques needed. Day by day new applications are invented and they have heterogeneous nature in network environment and communication between these new devices also a critical part. improving the network performance, establish proper service policies in router, handling network security risks, management of network operations and provide Qos services to users in internet. To solve these issues classification techniques are used. In this survey different classification algorithms are discussed. K-means algorithm, classification using clustering algorithm, Classification based on Fuzzy Kernel K-means Clustering, Support vector machine algorithm, and self-learning classifier Bayesian classification, C5.0 and traffic classification using correlation information and robust network traffic algorithms are presented.