面向5G超密集蜂窝网的毫米波混合式Mesh组网研究

So far there still lacks an effective methodology to leverage mmWave to accomplish flexible and efficient ultra-dense networking (UDN) for 5G networks. Thus, research will be conducted in this project to investigate networking mechanisms, algorithms, and protocols for mmWave-based UDNs, and ultimately develop a complete framework for mmWave-based UDNs. By considering the characteristics of mmWave communications and ultra-dense cellular networks, this project starts with a unique design of heterogeneous networking architecture, which is featured with hybrid mesh networking for integrating backhaul and access networks, control and management of mmWave networks via macro cellular networks, and separation between user plane and control plane. Based on this architecture, a complete methodology for mmWave-based UDNs is established as follows. For backhaul networks, key physical and medium access control (MAC) mechanisms are developed first. These mechanisms include beam switching and configuration for dynamic and static beams, fast detection of link degradation, and quick response to link failure. Besides, a novel resource management algorithm is derived to consolidate centralized optimization and distributed scheduling, with its long-term performance proved to be optimal. Based on these mechanisms and the resource management algorithm, networking protocols are designed to achieve flexible mesh networking in mmWave backhaul networks. For access networks, key mechanisms are developed to handle mmWave mobile communications. Such mechanisms include fast cell discovery, quick beam alignment, and efficient mobility management. They work cooperatively with virtual cells that are created from multiple small cells. Based on these mechanisms and virtual cells, algorithms and protocols are designed to deliver stable services to users moving across mmWave access networks. Finally, a simulation platform will be developed, through which the research results are validated. The effectiveness, advantages, and performance gains of the entire methodology developed in this project will also be demonstrated by running typical networking scenarios on the simulation platform.

如何利用毫米波实现5G超密集蜂窝网的灵活高效组网至今没有系统有效的方法。为此，本项目深入研究基于毫米波通信实现超密集蜂窝网的组网机制、算法和协议，并创立一套面向超密集蜂窝的毫米波组网方法。首先根据毫米波和超密集蜂窝网的特点构建了异构网架构，其特征包括以混合式Mesh网络融合接入和回传网、以低频蜂窝网辅助管控高频蜂窝网、及控制面与用户面分离。然后在此架构基础上研究组网方案：针对回传网络，首先研究毫米波的动静态波束切换与配置、故障链路快速检测与应对等机制，同时推导集中式优化和分布式调度相融合的组网算法，最后基于这些机制和算法设计协议实现回传网的灵活组网；针对接入网，首先设计支持毫米波移动通信的快速小区发现、波束对齐和移动性管理等机制，并通过建立和维护虚拟小区的算法和协议来实现接入网的可靠组网。本项目的组网架构、机制、算法和协议将通过仿真来验证，并在仿真平台上运行典型组网场景来展示研究成果。

利用毫米波实现5G超密集蜂窝网的灵活高效组网至今没有系统有效的方法。为此，本项目完成了基于毫米波通信实现超密集蜂窝网的组网机制、算法和协议的深入研究，创立了一套面向超密集蜂窝的毫米波Mesh组网方法。首先根据毫米波和超密集蜂窝网的特点构建了异构网架构，其特征主要是以混合式Mesh网络实现5G接入回传一体化网络。然后在此架构基础上研究组网方案：针对回传网络，首先研究实现了毫米波的动静态波束切换与配置、故障链路快速检测与应对等机制，然后研究实现了集中式优化和分布式调度相融合的组网机制与算法，最后基于这些机制和算法设计协议实现回传网的灵活组网；针对接入网，首先设计了支持毫米波移动通信的快速小区发现、波束对齐和移动性管理等机制，然后针对超高频实现了新型多址接入协议，并通过建立和维护虚拟小区的算法和协议来实现接入网的可靠组网。本项目还开发了基于5G毫米波Mesh组网的仿真平台，上述研究的组网架构、机制、算法和协议等研究成果均在仿真平台上得到验证。