Characterizing smartphone traffic with MapReduce

With the demand of increasing revenue by attracting customers consuming mobile internet services, traffic monitoring and analysis become more critical on network management and service provisioning for operators and ISPs. At the same time, the rapid growing number of smartphone clients led to more complexity and explosive mobile Internet traffic. It makes traditional traffic analyzing methods face challenge of dealing with huge amount of data. In this paper, we propose and implement a distributed computing system which aims to perform high-speed data-intensive network traffic analyses by leveraging MapReduce programming model. Running on this efficient system, we summarize characteristics of smartphone mobile internet traffic by analyzing massive real data which is captured from a living UMTS network of a major service provider of China. Our analysis is crucial for cellular providers to anticipate usage patterns and future traffic growths.