项目名称: 基于部分网络状态信息的无线视频调度研究
项目编号: No.61201165
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 电子学与信息系统
项目作者: 周亮
作者单位: 南京邮电大学
项目金额: 25万元
中文摘要: 资源调度是多媒体通信中的关键技术之一。然而,现有的资源调度研究均建立在网络状态信息已知或可以精确获取这一强假设条件上。显而易见,这一假设往往在实际的应用场景中难以成立。由于忽略了网络状态信息的不确定性对多媒体资源调度的设计和实现性能的影响,现有调度算法并不适用于直接扩展至部分网络状态信息的情况。本项目以实际的无线多媒体通信场景为研究对象,着力解决基于部分网络状态信息的无线视频传输中的资源调度问题,以减少系统调度性能对网络信息透明性或动态变化性的依赖。具体而言,对控制信道受限、用户QoE模型信息部分未知、网络拓扑结构动态变化等三个不同层面进行分析和建模,分别利用马尔科夫决策理论、在线测试-优化理论、图染色理论设计出适合于无线视频通信特性的资源调度机制,提供具有强鲁棒性和普适性的解决方案,并通过仿真实验和理论分析的手段评价和比较所设计机制的性能。
中文关键词: 视频传输;无线通信;资源调度;码率控制;路由设计
英文摘要: Resource scheduling is one of the key techniques for multimedia communications. However, most existing resource scheduling algorithms for multimedia communications assume that perfect network state information is available or known to the scheduler. Clearly, this assumption may be invalid in many practical multimedia communication scenarios. It is not appropriate to directly extend existing algorithms to the case of partial network state information since they completely ignore the impact of network state information uncertainty on the performance of the resource scheduling. To this end, this program aims at proposing a series of resource scheduling strategies for wireless video transmission, in the context of practical communication scenarios, based on partial network state information to reduce the dependence of the network transparency and dynamic on the scheduling performance. Specifically, we first analyze and model the uncertainty of the network state information from three perspectives: constrained control channel, partial user QoE function, and dynamic network topology. Next, we design efficient, robust and generic resource scheduling strategies by employing Markov decision theory, online testing-optimization theory, and graph coloring theory, respectively. Finally, the performance of the proposed scheme
英文关键词: Video Transmission;Wireless Communications;Resource Allocation;Rate Control;Routing Design