SpecEES: Collaborative Research: Spatially Oversampled Dense Multi-Beam Millimeter-Wave Communications for Exponentially Increased Energy-Efficiency

毫米波（mmW）频段中可用的大量频谱为无线通信网络的数据速率呈指数级增长提供了途径。在第五代（5G）网络等新兴系统中，毫米波频率的使用可能会使网络容量、移动性和频谱效率得到前所未有的改善。然而，毫米波频段的开发需要解决许多技术挑战。特别是，当今实现中存在的技术限制需要在算法、信号处理方法、电路架构和集成方法方面采用新的范例，以便使5G无线成为现实。例如，需要先进的信道模型，使设计人员能够实现未来的无线网络基础设施。此外，还需要新的算法、软件、硬件和电子电路来实现高效的毫米波天线阵列处理。该项目将利用爱因斯坦狭义相对论中产生的著名物理学，即因果光锥，来显著提高毫米波无线基站中关键阵列信号处理组件的性能。具体来说，电磁波的时空特性，如狭义相对论所描述的，在新的架构中被利用来提高能量效率，降低噪声，并改善阵列接收器的线性度。将对毫米波信道时空特性的系统范围研究与这些体系结构相结合，以设计新型毫米波阵列接收器和最佳波束形成算法。狭义相对论描述了多维时空连续体中的一个区域，由于光速恒定和波动方程的性质，该区域不被传播的波所占据。因此，与无线传播信道相对应的所有传播波的支持区（ROS）都被限制在一个“光锥”内。这个锥体之外的时空区域（被称为“别处”）是一个空洞，在那里无线通信信号无法传播。虽然没有波，但别处被电子噪声和由现实世界放大器和数据转换器引起的非线性失真所占据。该项目探索了对毫米波天线阵列进行空间过采样的可能性，然后在离散空间和连续时间维度上应用众所周知的sigma-delta调制技术的多维扩展，以实现噪声和失真整形，从而有效地将不需要的接收分量转移到其他地方。尽管sigma-delta算法已被用于模数转换器（adc），但本文提出这些算法的多维扩展不仅限于adc；相反，有可能将这些算法应用于阵列中使用的低噪声放大器，adc和其他电路元件，这反过来又导致阵列处理多维电路理论中新概念的创建。该技术有望改善放大器的噪声系数和线性度，并以线性成本降低天线和接收器的数量，指数提高阵列数字化的ADC性能。由此产生的毫米波阵列处理器应用于无线通信、相控阵雷达和射电望远镜天线孔径。该项目是俄亥俄州和纽约四所大学之间的多机构合作，并将通过一年一度的布鲁克林5G峰会实施多种教育和社区外展活动。该项目包括为女性工程师和学生提供指导，开发新的教育材料，以及让代表性不足的群体参与无线通信主题。推广活动将通过社区活动、研讨会、布鲁克林5G峰会（包括5G中的女性活动）以及IEEE会议组织的科学推广和学术活动来实现。毫米波电路研究和教育计划将理论与实践系统原型相结合。行业参与对新兴无线技术至关重要，在整个项目中都有规划，并通过一年一度的布鲁克林5G峰会进行促进。开源模型、设计和原型芯片将提供给公众和无线行业。

The vast amount of spectrum available in the millimeter-wave (mmW) bands offer a path for exponential growth in data rates for wireless communications networks. In emerging systems such as fifth-generation (5G) networks, the use of mmW frequencies will potentially enable unprecedented improvements in network capacity, mobility, and spectral efficiency. However, the exploitation of mmW bands requires solutions to many technical challenges. In particular, the technology limitations present in today's implementations require new paradigms in algorithms, signal processing methods, circuit architectures, and integration methods in order for 5G wireless to become a reality. For example, there is a need for advanced channel models that let designers implement the wireless network infrastructure of the future. There is also a need for new algorithms, software, hardware, and electronic circuits for efficient mmW antenna array processing. This project will exploit well-known physics arising from Einstein's Special Theory of Relativity, namely the causality light-cone, to significantly improve the performance of key array signal processing components in mmW wireless basestations. Specifically, the spatio-temporal properties of electromagnetic waves, as described by Special Relativity, are exploited in novel architectures to improve the energy efficiency, reduce the noise, and improve the linearity of array receivers. A system-wide study of spatio-temporal properties of mmW channels is combined with these architectures to design new types of mmW array receivers and optimum beam forming algorithms. The Special Theory of Relativity describes a region in the multidimensional spacetime continuum that is not occupied by propagating waves due to the constant speed of light and the nature of the wave equation. As a result, the region of support (ROS) of all propagating waves, which correspond to wireless propagation channels, are confined inside a ``Light Cone''. The region of spacetime outside this cone (known as ``Elsewhere'') is a void within which wireless communications signals cannot propagate. Although devoid of waves, the Elsewhere is occupied by both electronic noise and nonlinear distortion arising from real-world amplifiers and data converters. The project explores the possibility of spatially over-sampling the mmW antenna arrays and thereafter applying multidimensional extensions of well-known sigma-delta modulation techniques across both discrete space and continuous-time dimensions to achieve noise and distortion shaping, which effectively move the unwanted received components into Elsewhere. Although sigma-delta algorithms have been employed in analog-to-digital converters (ADCs), it is here proposed that multidimensional extensions of these algorithms are not limited to just ADCs; rather, it is possible to apply these algorithms to low-noise amplifiers, ADCs and other circuit components used in arrays, which in turn leads to the creation of new concepts in multi-dimensional circuit theory for array processing. The technique is expected to lead to improved amplifier noise figure and linearity and exponentially improved ADC figure-of-merit for array digitization at a linear cost in the number of antennas and receivers. The resulting mmW array processors have applications in wireless communications, phased-array radar, and radio telescope antenna apertures. The project is a multi-institutional collaboration between four universities in Ohio and New York, and has multiple education and community outreach activities, which will be implemented via the annual Brooklyn 5G Summit. The project includes mentoring for female engineers and students, development of new educational material, and engagement of underrepresented groups in wireless communications topics. Outreach will be achieved through community activities, workshops, the Brooklyn 5G Summit including events for women in 5G, and scientific outreach and academic events organized within IEEE conferences. The mmW circuits research and education program combines theory with hands-on system prototyping. Industry engagement, which is critically important for emerging wireless technologies, is planned throughout the project, and facilitated via the annual Brooklyn 5G Summit. Open source models, designs and prototype chips will be offered to the public and wireless industry.