Network Virtulization allows virtual networks, running on a shared physcial network, to efficently use network resource,extend network functions and provide diversity services and applications. Network Virtulization has been considered as one of the future Internet trends. As the physical networks continue to converge and grow in size and complexity, the potential for large-scale failures also increases. Such large-scale failures can be caused by software bugs, system configuration errors, natural disasters or malicious attacks. Accordingly, network recovery from massive failures is critical and has attracted great attention in the research community. In the literature, most research focused on fast recovery right after small-scale failures happen, under the implicit assumption that all failed components can be repaired instantly and simultaneously. However, after large-scale failures, the physical network can only be progressively repaired due to the limited repair resources. Accordingly, virtual networks running on the physical network can only be gradually restored. In this project, we explore new research on theory and key technologies of progressive repair and progressive joint restoration under the network virulization environments after large-scale failures. We aim to design efficient progressive repair and restoration algorithms for various network technologies, from wireless access network to core network, with the consideration of not only multiple immediate-term objectives, but also short-term as well as medium to long term objectives. Key techniques to be investigated include graph theory, network optimization, probability analysis, and multiple criteria decision making.
网络虚拟化技术使构建在物理网络之上的虚拟网络能够高效利用网络资源并提供多样化的服务和应用需求,被认为是未来互联网的发展趋势。随着物理网络规模的不断增加,自然灾害的频繁发生以及各种人为因素都使网络发生大范围失效(比如网络里有多条链路及节点失效)的可能性日益增加。当失效范围大、修复资源有限时,物理网络必然要分步骤地进行渐进修复。同样,运行在物理网络之上的虚拟网络也将渐进恢复。本项目拟研究网络大范围失效下渐进的物理网络修复与虚拟网络协同恢复的理论和关键技术,尤其着重于物理网络包含移动通讯的情况。我们将针对不同场景设计多种高效渐进修复和恢复策略,综合考虑修复或恢复过程中不同阶段网络性能收益、成本等多目标间的平衡。本项目研究涉及多个相关领域。研究内容不仅限于计算机科学和网络通讯,还需要图论、运筹学、 网络优化、博弈论等理论和技术,预期将产生一批交叉学科的成果。
网络虚拟化技术使构建在物理网络之上的虚拟网络能够高效利用网络资源并提供多样化的服务和应用需求,被认为是未来互联网的发展趋势。随着物理网络规模的不断增加,自然灾害的频繁发生以及各种人为因素都使网络发生大范围失效(比如网络里有多条链路及节点失效)的可能性日益增加。当失效范围大、修复资源有限时,物理网络必然要分步骤地进行渐进修复。同样,运行在物理网络之上的虚拟网络也将渐进恢复。本项目研究网络大范围失效下渐进的物理网络修复与虚拟网络协同恢复的理论和关键技术,尤其着重于物理网络包含移动通讯的情况,如车载网。我们针对不同场景 (如车载网,数据中心,无线网)设计多种高效渐进修复和恢复策略,综合考虑修复或恢复过程中不同阶段网络性能收益、成本等多目标间的平衡,以及协议设计。
{{i.achievement_title}}
数据更新时间:2023-05-31
跨社交网络用户对齐技术综述
面向云工作流安全的任务调度方法
城市轨道交通车站火灾情况下客流疏散能力评价
基于FTA-BN模型的页岩气井口装置失效概率分析
TGF-β1-Smad2/3信号转导通路在百草枯中毒致肺纤维化中的作用
面向瘦虚拟化的系统资源映像与恢复技术研究
面向无线认知网络的网络编码理论与关键技术研究
面向农田复杂时变环境的无线传感器网络高效传输与虚拟化关键技术研究
网络虚拟化中虚拟网映射问题的关键技术研究