The enormous demand for mobile video streaming in the future necessitates an accurate estimation of video quality of experience. In this paper, we give insights on how buffered video streaming performance metrics can be modeled in a multi-cellular context by a joint treatment of data flow dynamics seen by a streaming flow in its own cell and dynamic inter-cell interference dynamics seen in neighboring cells. In particular, we show that dynamic inter-cell interference strongly affects video streaming startup delay distributions, not only for users at the cell edge, but also in the cell center. The model provided can help designing and optimizing cellular networks in the future.
展开▼