Advanced wireless communications and signal processing techniques for 6G wireless networks.

自2019年以来，第五代（5G）商用移动网络已在全球范围内推出，并在一些国家达到了巨大规模。例如，到2020年底，中国已部署了50多万个5G基站，为1亿多5G用户提供服务。自从最初推出第一代（1G）以来，每十年就会出现新一代。因此，十年后，5G的继承者将是6G无线。为什么需要6G ？目前，全球无线流量呈指数级增长，到2030年将达到每月5016 EB（艾字节），而2020年为每月62 EB。因此，5G可能无法适应2030年及以后的巨大移动流量。因此，2030年及以后，6G系统将提供极高的容量、可靠性和效率等。然而，许多经典无线技术的性能和效率指标（例如频谱效率（SE）和能源效率（EE））已经趋于稳定。这就需要新颖的系统设计来解决这些瓶颈。为此，需要对能源使用、频谱共享、多址接入、参数估计、资源分配等复杂的通信任务进行优化。因此，可以实现6G对高可靠性、低延迟和大量连接的要求。尽管无线技术呈指数级增长，但可用频谱有限，无法满足高数据需求的用户和设备的1000倍扩展。因此，研究人员必须提高SE（每单位带宽实现的数据传输速率）。增强SE的主要方法是使用认知无线电，它提供灵活的机会性频谱接入。另一种方法是非正交多址，即多个用户同时重用同一频谱。因此，这些解决方案提供比正交方法更高的SE，但具有明显更高的信号处理复杂性。在6G环境中，新频谱的未授权部分可能会受到其他系统和环境噪声的干扰。因此，6G设备必须能够动态地、认知地选择最佳的工作频段。然而，使用这些技术的代价是适应和估计的高度复杂性。因此，该项目将利用机器学习技术（深度学习、强化学习等）中的概念/技术。因此，本项目旨在开发整体、集成的解决方案，涉及多天线无线等经典技术和智能面板、无人机、能量收集等新兴技术。通过整合这些，研究可以找到协同效应和效率。该项目将为超5g和6G无线开发新的分析和设计方法、算法和应用程序，以实现这些收益。这是一个完美的时机，因为全球对这些系统的研究正在扩大。

Since 2019, commercial fifth-generation (5G) mobile networks have been rolled out worldwide and have already reached an enormous scale in some countries. For example, China has deployed over 500 000 5G base stations at the end of 2020, serving more than 100 million 5G subscribers. Since the original rollout of the first generation (1G), a new generation has appeared every one decade. Thus, in a decade, the successor of 5G is 6G wireless. Why is 6G needed? Currently, global wireless traffic is escalating exponentially, up to 5016 EB (exabyte) per month in 2030 compared with 62 EB per month in 2020. Thus, 5G may not fit the tremendous volume of mobile traffic in 2030 and beyond. Thus 2030 and beyond, the 6G system will provide extreme capacity, reliability, efficiency, etc. However, many classical wireless technologies have plateaued their performance and efficiency metrics (e.g., spectral efficiency (SE) and energy efficiency (EE)). This calls for novel system designs to attack these bottlenecks. To this end, it is necessary to optimize complex communication tasks such as energy use, spectrum sharing, multiple access, parameter estimation, resource allocation, and others. Thus, 6G requirements of high reliability, lower latency, and massive connectivity are realizable. Despite the exponential growth of wireless, the available spectrum is limited and cannot sustain the 1000 times expansion in users and devices with high data requirements. Therefore, researchers must improve the SE (the data transfer rate achieved per unit bandwidth). The primary way of enhancing the SE is with cognitive radio, which offers flexible, opportunistic spectrum access. Another way is non-orthogonal multiple access, where multiple users reuse the same spectrum simultaneously. These solutions thus provide higher SE than orthogonal approaches but have a significantly higher signal processing complexity. In the 6G landscape, the unlicensed parts of the new spectrum may suffer from interference by other systems and environmental noises. 6G devices, therefore, must be able to dynamically and cognitively select the best operating band. However, the cost of using these techniques is the high complexity of adaptation and estimation. Thus, this project will exploit concepts/techniques from machine learning techniques (deep learning, reinforcement learning, and others). Therefore, this project aims to develop holistic, integrated solutions involving classical techniques such as multiple-antenna wireless and emerging techniques such as intelligent panels, drones, energy harvesting, etc. By integrating these, research can find synergies and efficiencies. The project will develop new analysis and design methods, algorithms, and applications for beyond-5G and 6G wireless to achieve those gains. The timing is perfect, as the global research efforts into these systems are expanding right now.