Analysis of capacity of picocell with dominating video streaming traffic

This paper focuses on the research of centralized wireless networks where main load is formed with on-line video streaming services. In case of congestion channel resources are not enough for satisfying all the users requirements. This brings to degradation of video playback (appearance of rebufferings, forced transcoding to lower bit rate etc). Congestion has probabilistic nature, since it depends not only on number of users but also on there location relatively to a base station and chosen bit rates of videos. For considered networks we propose a model and two techniques of congestion probability estimation. Basing on these estimations network capacity is analyzed.