Resource-Constrained Wireless Video Communication

该提案旨在提高无线网络上功率受限设备的视频通信质量，特别是实时视频。本论文的研究目标是研究资源受限的无线视频通信，考虑的资源包括带宽、功率和时延。该方法是一种新的延迟功率率失真理论，它提供了一种手段来量化延迟，功耗，比特率和视频失真之间的关系。在这个项目中，PI将提出的延迟-功率-速率-失真理论应用于跨层设计，在总资源约束下联合优化视频编码器和无线通信组件;并开发一个通用的测试平台，实现所提出的算法。在这个建议中概述的想法和方法，预计将填补一些重要的空白，在资源受限的无线视频通信的基本理解，并将提供系统设计和算法实现的理论基础。时延-功率-速率-失真理论不仅可以为资源受限的无线视频通信系统的设计提供重要的理论依据，而且可以为实际的无线视频通信系统提供性能界限。提出的视频编码和跨层设计的设计方法将填补一个关键的空白-现有的机制缺乏在总资源约束下实现最佳性能的能力。初步的研究已经引起了业界的关注，这表明在应用所提出的技术，以支持实时视频通过无线网络的极大兴趣。综合研究和教育计划将培养无线，多媒体和嵌入式系统工程师，能够应对世纪的通信挑战。此外，PI希望利用这个项目来激发公众和所有年龄段学生的学习过程和发现兴趣。

The proposal aims to improve the quality of video communications, especially real-time video, over wireless networks for power limited devices. It is comprehensive in that it includes theoretical work, algorithm design, and testbed implementation.The objective of this research is to study resource-constrained wireless video communication; the resources under consideration include bandwidth, power, and delay. The proposed approach is a novel delay-Power-Rate-Distortion theory, which provides a means to quantify the relationship among delay, power consumption, bit-rate, and video distortion. In this project, the PI will apply the proposed delay-Power-Rate-Distortion theory to cross-layer design that jointly optimizes the video encoder and the wireless communication component under total resource constraints; and develop a general-purpose testbed with implementation of the proposed algorithms. The ideas and approaches outlined in this proposal are expected to fill a number of important gaps in fundamental understanding of resource-constrained wireless video communication and will provide the theoretical underpinning for system design and algorithm implementation. The delay-Power-Rate-Distortion theory is expected to not only yield important principles in design methodologies for resource-constrained wireless video communication, but also to provide performance bounds for practical wireless video communication systems. The proposed design methodologies for video coding and cross-layer design will fill a critical void -- the existing mechanisms lack the capability of achieving optimal performance under total resource constraints. The preliminary research has drawn attention from industry, which showed great interest in applying the proposed techniques to supporting real-time video over wireless networks. The integrated research and education program will train wireless, multimedia, and embedded system engineers, capable of coping with challenges of communication in the 21st century. Furthermore, the PI expects to use this project to inspire the learning processes and the discovery interests of the general public and students of all ages.