Network coding based SVC multicast over broadband wireless networks
Network coding based SVC multicast over broadband wireless networks
复制标题
DOI：
10.1109/lcn.2012.6423657
复制
发表时间：
2012-10
期刊：
影响因子：
--
通讯作者：
Hao Zhou;Yusheng Ji;Yu Gu;Bao-hua Zhao
中科院分区：
文献类型：
--
作者：
Hao Zhou;Yusheng Ji;Yu Gu;Bao-hua Zhao
无线网络上的视频多播在面对网络和终端用户能力的异构性以及沿着的分组丢失时具有其自身的挑战。使用应用于不同视频层的不同调制和编码方案（MCS）的可缩放视频编码可以根据不同用户的信道条件向不同用户提供不同的视频质量。本文提出了一种分层的混合NC/ARQ方案来处理分组丢失，并结合结构化网络编码（SNC）技术对SVC流进行编码。重要的是为每层选择合适的MCS，决定在一个层中应该发送多少SNC分组，以及确定分配给多个视频会话的资源，以将可伸缩视频编码应用于无线多播流。我们证明了这样的资源分配问题是NP-困难的，并提出了一个最优算法与伪多项式运行时间下的合理假设。我们还讨论了意外的数据包丢失的现象造成的分层混合NC/ARQ方案在高丢包率的环境中的退化，并提出了一个解决方案，以克服这样的问题。我们的算法可以获得最佳的传输配置，最大化所有接收机的期望效用。仿真结果表明，我们的算法提供了显着的改善，视频质量的最佳算法，而不需要从接收机的反馈和算法使用分层的混合FEC/ARQ方案，它具有出色的能力，以打击意外的数据包丢失。
Video multicast over wireless networks has its own challenges when facing the heterogeneity of networks and end-user capabilities, along with packet losses. Scalable video coding using different modulation and coding schemes (MCSs) applied to different video layers can provide different video qualities to different users according to their channel conditions. A layered hybrid NC/ARQ scheme is proposed in this paper to handle packet losses together with a structure network coding (SNC) technique to encode the SVC stream. It is important to choose an appropriate MCS for each layer, decide how many SNC packets in one layer should be transmitted, and determine the resources allocated to multiple video sessions to apply scalable video coding to wireless multicast streaming. We prove that such a resource allocation problem is NP-hard and propose an optimal algorithm with a pseudo-polynomial run time under a reasonable assumption. We also discuss the phenomenon of unexpected packet losses caused degradation of the layered hybrid NC/ARQ schemes in a high packet loss ratio environment, and propose a solution for overcoming such a problem. Our algorithm can attain the optimal transmission configuration for maximizing the expected utility for all receivers. The results from simulations revealed that our algorithm offers significant improvements to the video quality over an optimal algorithm without needing feedback from the receivers and an algorithm using a layered hybrid FEC/ARQ scheme, and it has the outstanding ability to combat unexpected packet losses.