Towards smart phone traffic classification

摘要: The appearance of smart phones and their continuing rapid uptake has large affects on our society in as much they represent a paradigm shift the traditional industrial structure. Telecom market is changing day by day, networks with complicated varied traffic have almost reached capacity because increase user service releases phones. Therefore, necessity for phone monitoring analysis increased. Traffic an essential element efficient reliable networks. In this paper, we propose new classification application method. proposed method composed several consecutive steps: grouping HTTP User-Agent field, extracting common strings LCS algorithm finally classifying traffic. addition, to classify unknown from previous methods, process that extracts header signatures grouped information improve completeness. We achieved about 90% accuracy rate target campus network.