arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1808.06722cs.NI
收藏

弹性视频传输机制

Mechanisms for Resilient Video Transmission

Roger Immich, Eduardo Cerqueira, Marilia curado

摘要:

无线网络被设想为提供具有成本效益的内容递送(包括用于视频应用)的最重要技术之一。它们将允许成千上万的固定和移动用户以无处不在的方式访问、制作、共享和消费视频内容。这些网络上的实时视频服务正在成为日常生活的一部分,并已被用于传播从教育到娱乐内容的信息。然而,处理这些网络的波动带宽、稀缺资源和时变出错率的挑战突出了对错误恢复视频传输的需要。在此上下文中,已知前向纠错(FEC)和不等错误保护(UEP)方法的组合为无线用户提供具有体验质量(QoE)保证的视频应用的分发。本文提出了一个评估视频特征及其对最终用户感知质量的影响的过程。本文提出了一系列基于FEC的跨层视频感知机制和UEP,以增强几种类型的无线网络中的视频传输。在这些机制中使用了许多方法来设置自适应的冗余量,例如启发式技术、随机神经网络、蚁群优化和模糊逻辑。在第一个,启发式技术,机制依赖于人类的经验来定义最佳策略。通过真实视频序列和实际网络轨迹的仿真实验,验证了所提出机制的优点和缺点。使用众所周知的QoE指标进行评估。结果表明,所有提出的机制都能够超越竞争对手的感知视频质量和网络足迹。

英文摘要:

Wireless networks are envisaged to be one of the most important technologies to provide cost-efficient content delivery, including for video applications. They will allow thousands of thousands of fixed and mobile users to access, produce, share, and consume video content in a ubiquitous way. Real-time video services over these networks are becoming a part of everyday life and have been used to spread information ranging from education to entertainment content. However, the challenge of dealing with the fluctuating bandwidth, scarce resources, and time-varying error rate of these networks, highlights the need for error-resilient video transmission. In this context, the combination of Forward Error Correction (FEC) and Unequal Error Protection (UEP) approaches is known to provide the distribution of video applications for wireless users with Quality of Experience (QoE) assurance. This thesis proposed a procedure to assess the video characteristics and their related impact on the perceived quality to end-users. This thesis proposes a series of cross-layer video-aware and FEC-based mechanisms with UEP to enhance video transmission in several types of wireless networks. A number of methods to set an adaptive amount of redundancy were used in these mechanisms, such as heuristic techniques, random neural networks, ant colony optimisation, and fuzzy logic. In the first one, heuristic techniques, the mechanisms rely on human experience to define the best strategy. The advantages and drawbacks of the proposed mechanisms were demonstrated in realistic simulations using real video sequences and actual network traces. The assessments were conducted with well-known QoE metrics. The results show that all the proposed mechanisms were able to outperform the competitors on both perceived video quality and network footprint.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑