Analyzing traffic characteristics between backbone networks based on Hadoop

Nowadays, along with the development of mobile network, the network traffic is growing rapidly, including traffic between backbone networks. ISPs are seriously puzzled by the explosive growth of traffic between backbone networks. Therefore, the research and analysis of traffic characteristics between backbone networks becomes essential. In this paper, based on the Hadoop platform, we analyze the monthly traffic features of two provinces by using the real data collected by our self-developed traffic monitoring system. Besides, we use SARIMA model to select a daily trace that can represent the monthly trace best. Then, through analyzing the distribution of packet counts, flow duration, and user online time and so on, we examine the basic traffic characteristics and user behavior characteristics from the perspective of flows.