Large-Scale User Behavior Characterization of Online Video Service in Cellular Network

Abstract

Online video service has become prevalent in recent years. A better understanding of the user behavior of such service is crucial for allocating the network resources and adjusting the service design. While there are some measurement studies on the non-mobile video services in fixed network, the usage of video service in mobile network is yet to be explored. In this paper, we present a detailed analysis of the user behavior characteristics of a leading comprehensive online video service, namely Youku, in cellular network. This paper is based on a large-scale data set containing over 17 billion traffic traces, collected from a major cellular network in Northeastern China. We analyze the user behavior from three key aspects: data consumption, service usage, and mobility. We provide an insight into how the mobile video service is utilized by users (especially the heavy users), by measuring the user intensities and various representative behavior features in each analysis aspect, such as active time, replay rate, video category, access location, and residence time. We reveal the patterns of different user behaviors, and discuss the implications for practical application. The findings of this paper can provide direct help for network operators and service providers to improve the network performance and user experience.