Regulation Aware Spectrum Sharing in 5G and B5G Wireless Networks

该提案探讨了频谱共享在5G及Beyond-5G(B5G)系统中的作用。人们普遍认为，移动通信服务的需求增长将超过可用的频谱。在这方面，在移动网络运营商之间共享频谱，或从控股公司租用频谱，以使频谱与需求最佳匹配，是一个越来越有吸引力的解决方案。虽然这一概念已经在认知无线电的背景下进行了探索，但在本提案中，我们将探索频谱共享与新兴技术趋势(如大规模多输入多输出(MIMO)系统和网络切片)之间的相互作用，从而允许应用程序和频谱使用的虚拟化。 这一伙伴关系的长期目标是为TELUS可能不得不在完全不同的频谱许可模式下运营的未来做好准备。其目的是了解这样的许可证制度将施加什么限制，以及TELUS的业务将不得不如何改变以应对。在这里，我们提出了实现这一目标的初步步骤，分析了新兴技术和合作模式之间的相互作用，以了解在5G(和Beyond-5G-B5G)网络的背景下可能发生的事情、监管限制和新的分析工具以及利用共享频谱的技术。 在追求这些目标的过程中，我们提出了三个目标：使用频谱共享透镜分析大规模MIMO系统，提出鼓励频谱共享的信号处理和资源分配方案；提供频谱共享的经济分析；最后，使用机器学习技术，特别是强化学习，在有限的系统状态信息下评估系统性能和经济价值。

This proposal explores the role of spectrum sharing in 5G and beyond-5G (B5G) systems. It is well-accepted that growth in demand for mobile communication services will outstrip the availability of spectrum. In this regard, sharing spectrum amongst mobile network operators, or leasing spectrum from a holding company, to best match spectrum to demand is an increasingly appealing solution. While this notion has been explored in the context of cognitive radio, in this proposal we explore the interplay of spectrum sharing and emerging technology trends such as massive multiple-input, multiple-output (MIMO) systems and network slicing, allowing for virtualization of applications and spectrum usage. The long-term goal of this partnership is to prepare for a future where TELUS may have to operate under a completely different spectrum licensing paradigm. The aim is to understand what constraints such a licensing regime would impose and how TELUS operations would have to change in response. Here we propose initial steps toward that goal, analyzing the interplay between emerging technologies and cooperation modes to understand what is possible in the context of 5G (and beyond-5G -B5G) networks, the regulatory constraints and new tools to analyze and techniques to exploit shared spectrum. In pursuing these goals, we propose three thrusts: analyzing massive MIMO systems using a spectrum sharing lens and proposing signal processing and resource allocation schemes to encourage spectrum sharing, providing economic analyses of spectrum sharing and, finally, using machine learning techniques, especially reinforcement learning, to evaluate system performance and economic value with limited information about the state of the system.