超密集网络中基于服务器虚拟机的多点协作技术研究

To cater for the ever-increasing demand of traffic and variety of applications in ultra-dense networks, Servers can be virtualized as multiple storage-intense virtual machines and computation-intense virtual machines. Virtual machines in different points can be cooperated and concentrated to zones lacking of resources based on server virtual machines. From the viewpoint of wireless transmission, storage-intense virtual machines between multiple points can cooperate to achieve joint cooperative transmission and base station ON-OFF switching between base stations, which can improve the energy efficiency. From the viewpoint of wired routing, storage-intense virtual machines between multiple points in networks can work in cooperative manner to store traffic of remote service providers to reduce the cost involved in wired routing. Regarding the computation-intense virtual machine, multiple points can cooperate as computing resource pool to execute fast networks optimization. The project will study impacts brought about by server virtual machine to networks capacity, energy efficiency and cost in ultra-dense networks, considering renewable energy, we will propose joint base station cooperative transmission and ON-OFF switching mechanism. Moreover, taking the different storage capacity of points into account, we will propose joint wired routing and cooperative traffic caching of remote service providers. Finally, the project will propose a framework of joint multi-point cooperation in networks to trade off the allocation between virtual machines with the constraint of overall storage capacity of servers to improve the performance of ultra-dense networks.

为了满足超密集网络中日益增长的业务及应用需求，网络各端点配置的服务器可以虚拟化为以数据存储为主的虚拟机和以优化计算为主的虚拟机。多个端点之间的虚拟机可以进行联合协作，将网络资源集中到资源需求密集的区域。多点之间的存储型虚拟机在无线方面可以协作实现基站之间的联合协作发送及开关切换，以提升无线能效。在有线传输方面, 多点之间的存储型虚拟机可以进行协作对远端业务提供者的业务进行备份，以减少有线传输成本及时延。对于计算型虚拟机，多点之间可以协作形成计算资源池以支持网络整体的快速优化。本项目将研究服务器虚拟机对超密集网络容量、能效、成本等各部分性能影响及其机理，基于可再生能量等提出无线部分的联合基站协作发送及开关机制。接着，考虑端点存储容量差异性提出异步有线传输及多点协作备份机制。最后，项目将在服务器容量限制下提出网络整体的联合多点协作机制对虚拟机分配进行折衷优化，从而提升超密集网络性能表现。

为了满足超密集网络中日益增长的业务及应用需求，网络各端点配置的服务器可以虚拟化为以数据存储为主的虚拟机和以优化计算为主的虚拟机。多个端点之间的虚拟机可以进行联合协作，将网络资源集中到资源需求密集的区域。多点之间的存储型虚拟机之间，在无线方面可以协作实现基站之间的联合协作发送及开关切换，以提升无线能效。在有线传输方面, 多点之间的存储型虚拟机可以进行协作对远端业务提供者的业务进行备份，以减少有线传输成本及时延。对于计算型虚拟机，多点之间可以协作形成计算资源池以支持网络整体的快速优化。本项目将研究服务器虚拟机对超密集网络容量、能效、成本等各部分性能影响及其机理，基于可再生能量等提出无线部分的联合基站协作发送及开关机制。接着，考虑端点存储容量差异性提出异步有线传输及多点协作备份机制。最后，项目将在服务器容量限制下提出网络整体的联合多点协作机制对虚拟机分配进行折衷优化，从而提升超密集网络性能表现。. 通过三年的项目周期，本项目成功对3C融合的理论进行深入研究。申请人通过虚拟化技术，结合云计算、区块链技术等实现物联网中数据的安全存储；进一步的，结合纳什议价博弈将3C融合的思想引入星地传输，实现兼顾用户公平性与可靠性的星地传输机制，该机制为面向6G的物联网通信提供了一种重要思路。此外，申请人进一步通过无人机技术将3C思想移植到移动式基站上，通过无人机移动性，结合指针网络深度学习机制等，在时延与能耗限制下对物联网中的移动式边缘服务进行了研究。本项目发表的工作均在国内外本专业顶刊上，且通过项目研究与国内外各顶尖团队实现了研究合作，加强了申请人研究的影响力。