Mobile brand analysis based on WiFi hotspots on campus

With the rapid development of mobile Internet, the ways in which users access the network become diverse, which provides much convenience for us to collect huge amount of users' behavior information. In this paper, we combine data acquisition based on wireless access and brand analysis innovatively. This paper analyzes proportions of mobile terminal brands on campus using the data collected from WiCloud which is a WiFi-based data acquisition platform built by our lab, and classifies and analyzes the characteristics of mobile terminal brands. Finally, we put forward a method of predicting the proportions of mobile brands on campus based on ARIMA model, and the results prove that our brand prediction method has good fitting effect and good prediction ability.