Analyzing the time-limited behavior of mobile devices carriers

摘要

In general cases, wireless devices in the cyber world are representatives of their carriers in the physical world. The networking applications performance highly depends on the mobility characteristics of devices carriers. Recently, it has pointed out that the non-Poisson statistics with heavy tail existed in many studies of dynamics of devices carriers. But most of these studies claim that power-law characterizes diverse aspects of devices carriers' mobility patterns. In this paper, we analyze the behavior of library visitation which is obtained from Radio Frequency Identification tags and find that the time interval of entering to the library is well fitted with three sections of lines instead of the traditional distribution. At the same time, we divide the data into groups and extract period bursty events. Through careful analysis in different groups, we draw a conclusion that the aggregate behavior is composed of "heterogeneous groups".