| IPTV services have been widely deployed in many countries in recent years. Theseservices predominantly include both Live TV and Video-on-Demand (VoD), offered by thesame service provider to the same pool of users over the same managed network.Understanding user behaviors in such setting serves as an important step to model andoptimize such a system.Previous studies on user behavior were on either Live TV or VoD. For the first time,we conduct an in-depth behavior study for IPTV users offering both Live TV and VoDservices to understand how users interact with the services. Our data is from the largestIPTV service provider in China, offering hundreds of live channels and hundreds ofthousands of VoD files, with traces covering more than1.9millions users over a period of5months. This large set of data provides us valuable insights on user behavior of suchsimultaneous system, and allows us exploring in details some new behavior issues whichhave not been studied before.Our results shed some new understanding on IPTV user behaviors which have strongimplications on system design. For example, we find that the average holding time for VoDis significantly longer than Live TV. However, if channel surfing is discounted for Live TV,the holding times of both services are not much different. Given similar video length, usersin VoD tend to view HD longer. On the other hand, for Live TV, channel popularity ismuch less dependent on its video quality. In contrast to the popular assumptions on userinteractivity, the transitions among Live TV, VoD and Offline modes are far from a Markovmodel. |
PDF Full Text Request