基于LAA的认知蜂窝网络设计理论与优化研究

With the rapid development of information technologies, the next generation wireless cellular networks need to address technical scenarios including seamless wide-area coverage and hot-spot high-capacity. Licensed spectrum becomes inadequate to support the explosive growth of wireless data traffic, impeding the development of wireless cellular networks. In this project, we study the fundamental theory and key technologies of licensed-assisted-access (LAA) using cognitive cellular networks, where spectrum sharing technique is adopted to extend radio spectrum resource. By investigating the mechanism of deep spectrum sharing between the LAA cellular networks and the legacy networks in 5GHz unlicensed bands, we propose efficient spectrum sharing theory for heterogeneous networks by introducing active spectrum grabbing, access and vacating based on spectrum sensing. According to the protocol design of spectrum sharing in unlicensed bands, we establish the theory of mathematical modelling and performance analysis for the protocol. Considering the service requirement and legacy networks offloading, the multi-dimensional joint optimization theory for spectrum sharing protocol is proposed, fulfilling a high utilization of unlicensed bands. This project aims to establish the fundamental theory of LAA using cognitive cellular networks and propose the novel spectrum sharing techniques that can profoundly improve the capacity of wireless cellular networks, realizing a full access of licensed and unlicensed bands for wireless cellular networks. The achievements of this project will provide theoretical reference and technique support for sustainable development of next generation wireless cellular networks.

随着信息技术的发展，连续广域覆盖和热点高容量等场景对下一代无线蜂窝网络容量提出了极高的要求。仅依靠授权频谱将不足以支持快速增长的数据流量，成为无线蜂窝网络发展的重大瓶颈。本项目利用频谱共享技术拓展蜂窝网络频谱资源，研究基于授权协助接入（LAA）的认知蜂窝网络基础理论和关键技术。通过研究LAA蜂窝网络与5GHz非授权频段上免授权网络的深度频谱共享机理，提出基于频谱感知进行主动频谱捕捉、接入和释放的异构网络间高效频谱共享理论；基于频谱共享协议设计，建立协议的数学建模和性能分析理论；构建面向用户需求及负载可切换的多维联合优化理论，实现无线蜂窝网络对非授权频谱的高效利用。通过本项目的研究将建立基于LAA的认知蜂窝网络设计基础理论，研究并突破能够大幅度提升无线蜂窝网络容量的频谱共享技术，实现无线蜂窝网络对授权频谱和非授权频谱的全面接入，为我国未来无线蜂窝网络的可持续发展提供重要的理论基础及技术支撑。

本项目旨在利用频谱共享技术拓展无线蜂窝网络可利用的频谱资源，研究基于授权协助接入（LAA）的认知蜂窝网络基本理论和关键技术。在机会型LAA频谱共享方面，对LAA-LTE系统与5GHz免授权频段上现行的WiFi系统间深度共享机理进行了研究，提出了基于频谱感知进行主动频谱捕捉、接入和释放的MAC层接入协议，实现了信道容量和时延性能的量化和近似；建立了协议参数和系统资源配置优化模型和优化算法，以保证LAA与WiFi系统间公平、高效的频谱共享；利用机器学习技术提出了LAA异构网络资源配置和分布式动态网络接入联合优化，实现了LAA用户网络接入和资源配置对网络环境变化的快速响应；面向LAA用户的QoS需求，建立了用户网络接入和物理资源配置的联合优化，实现了可支持QoS用户数的最大化。在共存型LAA与WiFi系统间的频谱共享方面，通过量化PHY-MAC跨层性能，揭示了共存型LAA的反射传输对WiFi系统性能的影响；为共存型LAA设备设计了基于随机序列的反射传输机制，实现了免用户间协作、低主次系统间信令传递需求和用户规模可扩展的多用户接入技术。