基于频谱重整技术的认知蜂窝网络设计理论与资源优化研究

The explosive growth of data traffic and the spectrum scarcity have been great challenges of wireless communications. As the mobile communications evolve towards 4G and 5G networks, the licensed spectrum for 2G and 3G networks will be suffering from increasingly serious under-utilization. In this project, we study the design and optimization theory of cognitive cellular networks (CCN), and propose a novel broadband spectrum refarming (SR) technology to facilitate the spectrum sharing on licensed cellular spectrum between different generations of cellular networks. The research topics include SR system modelling, interference analysis and control, SR system resource allocation, and SR-based CCN (SR-CCN) network design and optimization. The key innovations of the project include: 1) Using cognitive radio techniques, the project proposes the broadband SR technique to support spectrum sharing between OFDMA and CDMA cellular networks; 2) By applying large random matrix theory, the method for quantifying system interference temperature is proposed; 3) Under the condition that the cross-channel state information is unavailable, the resource allocation schemes for SR systems are proposed by exploiting the inner power control mechanism of the primary systems; 4) Using stochastic geometry and convex optimization, we establish the theories of networks planning and resource optimization for cognitive cellular networks. The results of the project will help to establish the intellectual property base in novel spectrum sharing technology, and provide new theoretical foundations and methodological guidance for the development of future mobile communications.

无线通信面临着数据流量爆炸式增长和频谱资源严重短缺的重大挑战。随着通信网络向4G和未来5G演变，已授权蜂窝频谱的有效利用率却变得越来越低。本课题研究异代蜂窝网络频谱共享的认知网络设计理论，提出宽带频谱重整（Spectrum Refarming，SR）技术以实现已授权蜂窝频谱的高效利用。研究内容包括：SR系统建模、干扰分析与控制；SR系统资源分配；以及基于SR技术的认知蜂窝网络设计与优化。 创新性工作有：1）运用认知无线电原理，提出支持OFDMA网络与CDMA网络频谱共享的宽带频谱重整技术；2）利用高维随机矩阵理论构建干扰温度量化的系统性方法；3）利用主系统内在的功率控制机制，建立在交叉信道信息未知情况下的次系统资源分配方法；4）利用随机几何及凸优化方法建立认知蜂窝网络的网络规划和资源优化理论。研究成果将形成我国拥有独立知识产权的频谱共享技术，为未来移动通信的发展提供新的理论依据和方法指导。

无线通信面临着数据流量爆炸式增长和频谱资源严重短缺的重大挑战。随着通信网络向4G、5G和未来6G演变，已授权蜂窝频谱的有效利用率却变得越来越低。本项目研究异代蜂窝网络频谱共享的认知网络设计理论，提出宽带频谱重整技术以实现已授权蜂窝频谱的高效利用。主要研究内容包括：频谱重整蜂窝系统建模、干扰分析与控制、以及基于频谱重整的认知蜂窝网络设计与资源优化。重要结果包括：1）利用认知无线电技术和随机矩阵理论，建立了基于OFDMA与CDMA宽带频谱重整技术的认知蜂窝网络设计理论与系统资源优化技术体系；2）利用优化理论，提出了一系列认知蜂窝网络动态频谱共享关键技术，包括频谱共享干扰信道估计与干扰功率预测、协作式频谱共享系统性能分析、LTE与WiFi非授权频段频谱共享等；3）利用压缩感知理论和优化理论，提出了若干认知蜂窝网络关键技术，包括具有次奈奎斯特速率的宽带频谱感知方法、认知网络中协作频谱占用预测、认知网络中高能效传输机制等。研究成果有助于形成我国拥有独立知识产权的频谱共享理论与技术体系，为未来移动通信的发展提供新的理论依据和方法指导。本项目已发表国际期刊论文43篇（SCI 和EI 检索）和国际会议论文22篇（EI 检索），出版专著1部，申请中国发明专利8项，授权8项，超额完成既定任务。