科研之友
微信
新浪微博
Facebook
分享链接
摘要
The delivery of high quality video to multiple viewers accessing the content from various devices and networks with varying conditions is a challenging task. Transferring the same content (often with large volume) to multiple viewers drives unicast transmission inefficient. Furthermore, traditional multicasting is not an optimal solution for heterogeneous networks since it either overwhelms slow receivers or starves the fast ones. Hence, new delivery solutions with the goal of reducing the output rate of the source and tailored to heterogeneous networks is necessary. In this paper, we present a novel scheme for layered video multicasting to viewers with the flexibility to access the content from various devices (in a heterogeneous network) and loss/erasure prone networks. In this scheme, the source node uses Scalable Video Coding (SVC) to obtain scalability together with modern elastic Forward Error Correction (FEC) for loss protection. There exists a non-trivial advantage in using elastic FEC in conjunction with SVC which enables us to introduce a new architecture that brings considerable savings in required source output bandwidth. Our solution adapts to the specific needs of viewers, shows resilience against packet loss, e. g. less than 5dB drop in Peak Signal-to-Noise Ratio (PSNR) for up to 30% loss rate, and especially, can reduce the source output rate by half for typical 3-layer SVC.
- 出版日期2013-5
