Energy Efficient Millimeter Wave Massive MIMO Wireless Communications

适用于大规模连接的低功耗高性能无线设备A。Lee Swinlehurst和Michael Green University of California Irvine所谓的“物联网”(IoT)指的是所有类型的物理设备之间的互联，以使它们能够收集数据并相互通信，或者与允许它们与互联网交换数据的接入点，从而与整个世界交换数据。有太多这样的设备可能有用的应用，包括医疗、交通、民用基础设施、环境监测以及个人通信。这些设备感知环境并相互通信和互联网的能力将带来一场革命，将改善我们的健康、我们的环境和我们的整体生活质量。为了使物联网革命成为现实，需要取得重大进展，以降低这些设备的成本(以便它们可以无处不在地部署)以及它们的能源消耗(以便它们可以在不更换的情况下长时间运行)。在过去的几十年里，数字硬件的进步导致了电路的小型化，这反过来又允许具有数十亿个晶体管的复杂微处理器占据不超过针头的空间。纳米制造的进步使得微型传感器和执行器的制造成为可能，它们只需要很少的能量或最小限度地进入环境就可以运行。然而，与模拟电子转换、放大和通信相关的物理限制要克服得更慢。该项目的目标是帮助弥补这一差距，并开发简单、低功耗、低复杂性的射频设备，这些设备可以与复杂的软件结合使用，使物联网愿景成为现实。需要新的技术和硬件实施来降低收发器的大小、成本和功耗，从而实现物联网设备的无处不在的部署以及高能效的千兆比特每秒无线网络。在这一愿景的推动下，我们的项目专注于几个新的研究举措，将推动无线系统朝着这个方向发展：(1)基于可重新配置的一位直接检测(DDA)天线的大规模MIMO射频硬件的新概念，这种天线潜在地不需要混频器或本地振荡器的产生和分配，并且能够实现更简单的天线馈电、高能量效率、更低的成本和超快的信令；(2)基于电感调谐的新的集成电路实现，以验证DDA概念，并说明如何使用无线LO分布的无混频解调来显著简化毫米波频段的射频前端；(3)基于角频域中的新推导，上行链路/下行链路信道容量由不依赖于白量化噪声假设的1比特ADC/DAC产生。这些结果将更适合于倾向于稀疏和频率选择性的毫米波信道。(4)具有1比特DAC的新型有限域下行链路预编码器，其不需要标准的乘法和加法算术运算(因此大大简化了必要的数字计算)，并且仍然显著优于传统方法，并且可以与信道编码技术相结合；(5)用于低分辨率ADC和DAC硬件的更现实的模型，其考虑了数据中的带外发射和互调产物，以便确定在什么条件下可以忽略它们以及如何以其他方式减轻它们。这项预期的研究解决了下一代无线系统的关键挑战，也与低功率物联网技术的成功具有很高的相关性。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Low-Power High-Performance Wireless Devices for Massive ConnectivityA. Lee Swindlehurst and Michael Green University of California IrvineThe so-called "Internet of Things'' (IoT) refers to the inter-networking of all types of physical devices in order to allow them to collect data and communicate with one another, or with an access point that allows them to exchange data with the internet and hence the entire world. There is a plethora of applications where such devices could be useful, including medicine, transportation, civil infrastructure, environmental monitoring, as well as personal communications. The ability of these devices to sense their environment and communicate with each other and the internet will lead to a revolution that will improve our health, our environment and our overall quality of life. In order for the IoT revolution to become a reality, significant advances are needed to reduce the cost of these devices (so that they can be deployed ubiquitously) as well as their energy consumption (so that they can operate for long periods of time without replacement). Advances in digital hardware over the past decades has resulted in circuit miniaturization that has in turn allowed for complicated microprocessors with billions of transistors to occupy a space no bigger than the head of a pin. Advances in nanofabrication have allowed for the creation of tiny sensors and actuators that require only very small amounts of energy or minimal access to their environment in order to operate. However, the physical constraints associated with analog electronic conversion, amplification and communication have been slower to be overcome. The goal of this project is to help remedy this gap, and develop simple, low-power, low-complexity radio-frequency devices that can be used in conjunction with sophisticated software to enable the IoT vision to become a reality. New techniques and hardware implementations are needed that reduce the size, cost and power consumption of transceivers that will enable ubiquitous deployment of IoT devices as well as energy efficient gigabit-per-second wireless networks. Motivated by this vision, our project focuses on several new research initiatives that will push wireless systems in this direction: (1) A new concept for massive MIMO RF hardware based on reconfigurable one-bit direct detection (DDA) antennas that potentially require neither mixers nor local oscillator generation and distribution, and enable simpler antenna feeding, high energy efficiency, lower cost and ultrafast signaling; (2) New integrated circuit implementations based on inductor tuning to validate the DDA concept and illustrate how mixer-less demodulation via wireless LO distribution can be used to dramatically simplify the RF front end for millimeter-wave frequency bands; (3) Uplink/downlink channel capacity results with one-bit ADCs/DACs that do not rely on the assumption of white quantization noise, based on a new derivation in the angular frequency domain. These results will be better suited for millimeter-wave channels that tend to be sparse and frequency-selective. (4) New types of finite field downlink precoders with one-bit DACs that do not require the standard arithmetic operations of multiplication and addition (and hence considerably simplify the necessary digital computations), and yet still substantially outperform conventional methods and can be combined with channel coding techniques; (5) More realistic models for low-resolution ADC and DAC hardware that account for out-of-band emissions and intermodulation products in the data, in order to determine under what conditions they can be ignored and how they can be mitigated otherwise. The anticipated research addresses critical challenges for next-generation wireless systems and is also of high relevance for the success of low-power IoT technology.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.