Software-Radio Strategies for Heterogeneous Communication Networks

在过去的二十年中，由于在发送和接收侧使用多个天线可实现的速率和分集增益，开发多输入多输出（MIMO）无线通信的研究经历了巨大的增长。当前标准仅捕获MIMO技术的有效频谱益处的一小部分，因为它们仅使用少量天线并且仅获得约10 bps/Hz的频谱效率。传统MIMO系统的这种改进促使研究人员研究在通信端使用大量天线，以获得更高的频谱效率以及无线传输中的更高分集阶数。机器学习（ML）算法最近已经显示出其有前途的能力。在过去的几年里，用于通信系统的ML方法的工作量显著增长。本研究试图了解ML技术与传统无线通信理论相结合的可行性，以解决无线通信中当前和未来的挑战，例如MIMO均衡，估计和检测技术，配置上行链路和下行链路信道以及大规模MIMO系统中的调度波束成形。波束成形是一种允许基站向用户发送目标信息波束的技术，从而减少干扰并更有效地使用无线频谱。混合波束形成是大规模MIMO系统中最有前途的降低硬件成本和训练开销的方法之一，特别是在毫米波段。在这项研究计划中，我们的目标是开发可用于降低无线系统的能耗和硬件成本，同时提高频谱效率的算法。多用户MIMO系统的无线资源分配的一个共同特征是有限数量的无线资源应该在许多用户之间分配。在本研究中将在任何网络中研究的无线电资源分配的思想是有效地分配有限的资源，即，用户之间的总发射能量和可用频谱，以满足用户的服务需求。*最后，尽管大规模MIMO和毫米波技术最近在研究界引起了相当大的兴趣，但其RF前端实现、其对系统性能的影响以及缓解技术的有效性尚未得到彻底研究。因此，本研究计划使用新型/改进的数字信号处理算法研究RF非理想性的分析，建模和低复杂性抑制。在这项研究中，我们主要关注多天线信令策略对系统容量的影响，以及在容量最大化和误差最小化的意义上，空间自由度分配到分集、复用和干扰消除的最佳选择。

Research developing Multiple-input-Multiple-Output (MIMO) wireless communications has experienced tremendous growth over the past two decades due to the rate and diversity gains achievable using multiple antennas at both transmit and receive sides. The current standards capture only a fraction of the efficient spectral benefits of MIMO techniques, as they use only a small number of antennas and attain spectral efficiencies of only around ten bps/Hz. Such improvement of traditional MIMO systems triggers researchers to examine the use of a massive number of antennas at the communication ends to obtain higher spectral efficiencies in addition to higher diversity orders in wireless transmissions.***Machine Learning (ML) algorithms have shown their promising capacities recently. In previous years, the body of work on ML methods for communication systems has conspicuously grown. This research attempts to find an understanding of the feasibility of combining ML technologies with traditional wireless communication theories to address the current and future challenges in wireless communication, such as MIMO equalization, estimation and detection techniques, configuring uplink and downlink channels and scheduling beamforming in massive MIMO systems.***Beamforming is a technique that allows the base stations to transmit targeted beams of information to the users, decreasing the interference and using the wireless spectrum more efficiently. Hybrid beamforming is one of the most promising approaches for reducing the hardware cost and training overhead in massive MIMO, especially for mm-wave bands. In this research program, we aim to develop algorithms that can be used to reduce the energy consumption and hardware cost of the wireless systems while enhancing the spectral efficiency.***A common feature of the radio resource allocations for multi-user-MIMO systems is that the limited number of radio resources should be distributed among many users. The idea of the radio resource allocation that will be investigated in this research at any network is to distribute the limited resources effectively, i.e., total transmit energy and available spectrum, among users to meet users' service requirements.***Finally, although massive MIMO and mm-wave techniques have recently received a considerable amount of interest in the research community, their RF front-end implementation, their implication on the system performance as well as the effectiveness of the mitigation techniques have not yet been thoroughly investigated. Therefore, this research program investigates analysis, modelling and low complexity suppression of RF non-idealities using novel/modified digital signal processing algorithms.***In this research, we focus mainly on the impact of multi-antennas signalling strategies on system capacity and the optimum, in the sense of capacity maximization and error minimization, allocation of the spatial degree of freedom to diversity, multiplexing, and interference cancellation.**