| Summary: | 碩士 === 國立臺灣大學 === 資訊工程學研究所 === 105 === In recent years, many video service providers start to provide 3D video contents,moreover, there is a new service which is called multi-view 3D video arise. The multi-view 3D video provides multiple view point to be chosen, also, it offer better immersive experience than traditional single view 3D video to users. But, the transmission of all views in multi-view 3D video would require significantly bandwidth consumption. Another interesting and well developed technology is the Wireless Local Area Networks (WLANs) which could support efficient multicast service on streaming with the limited wireless bandwidth. However, multicast to a set of heterogeneous user over multiple wireless Access Points (APs) also been a complicated problem.
In this thesis, we wont to solve the multicast multi-view 3D video in WLANs problem. We exploit the Depth Image Based Rendering (DIBR) technology to synthesize the user subscribed view from nearby left and right views. Therefore, in this problem we need to decide the user to AP association and the view session to be multicast from each AP. And the aim of our problem is maximize the number of satisfied users with bandwidth constraint. We propose an efficiently algorithm to solve this problem, and the simulation results show that our algorithm could further consider the view thesis and the inter APs coordination, and effectively satisfy most user demands.
|
|---|
