NeTS: Small: Toward Optimal, Efficient, and Holistic Networking Design for Massive-MIMO Wireless Networks

近年来移动互联网设备(例如智能手机、平板电脑等)的激增不断地将当前的无线网络推向其容量极限。提高频谱利用率是扩大网络容量的一个关键途径。为了使未来的无线网络能够支持每秒几千兆位的数据速率，近年来进行了大量的研究工作，被称为“大规模多输入多输出”系统(M-MIMO)，其中天线的数量被扩展到数百甚至数千。然而，到目前为止，这些努力大多局限于物理层或信号处理方面的问题。关于如何最佳地利用M-MIMO技术以及开发控制和优化算法以从组网角度改善网络容量、延迟和能量消耗性能的知识非常有限。拟议的研究将通过促进基于M-MIMO的无线网络的开发来使网络社区能够在吞吐量、延迟和能量消耗方面显著提高网络性能。通过这个项目，研究者还旨在促进M-MIMO在正规课堂教学中的研究，并教育学生对M-MIMO有很强的理解和了解。研究人员将主持一次研讨会，向无线网络领域的公司展示研究成果。这项研究不仅将促进M-MIMO无线网络设计方面的知识，而且将通过一个统一的研究计划来探索对M-MIMO网络的网络级理解，其中包括开发易处理的跨层理论模型、探索理论性能界限和分布式算法。本项目致力于回答以下首要问题：M-MIMO物理层的线性容量扩展规律能否转化为上层在吞吐量、延迟和能量消耗方面的性能收益？具体地说，该项目将围绕以下三个相互关联的主要领域展开研究工作：1)M-MIMO多跳回程网络的最优路由和拥塞控制；2)M-MIMO蜂窝网络的高效调度设计；3)M-MIMO无线网络的能量分析。该项目旨在弥合物理层在M-MIMO方面的进步和无线网络研究之间的差距。

The proliferation of mobile-Internet devices in recent years (e.g., smartphones, tablets, etc.) has continuously pushed the current wireless networks to their capacity limits. A key approach to enlarge network capacity is to increase spectrum efficiency. To allow future wireless networks to support multi-Gigabit per second data rates, there have been significant research efforts in recent years termed 'massive multiple-input multiple-output' systems (M-MIMO), where the numbers of antennas are scaled up to hundreds or even thousands. However, so far, these efforts are mostly limited to problems at the physical layer or signal processing aspects. There is very limited knowledge on how to optimally employ M-MIMO technologies and develop control and optimization algorithms to improve network capacity, delay, and energy expenditure performances from networking perspectives. The proposed research will enable the networking community by facilitating the development of M-MIMO-based wireless networks with substantially increased network performances in terms of throughput, delay, energy expenditure. Through this project, the investigator also aims to advance M-MIMO research in formal classroom instruction and educate students to inculcate a strong comprehension and knowledge of M-MIMO. The investigator will host a workshop to present research findings to companies in the areas of wireless networking. This research will not only advance the knowledge in the design of M-MIMO wireless networks but will also serve a critical need in the general networking research community by exploring a network-level understanding of M-MIMO networks through a unified research program, which consists of the development of tractable cross-layer theoretical models, exploration of theoretical performance bounds and limits, and the development of distributed algorithms.This project strives to answer the following overarching questions: Could the linear capacity scaling law at the M-MIMO physical layer be translated into upper layers' performance gains in throughput, delay, and energy expenditure? Specifically, the project will focus the research efforts along the following three interrelated thrust areas: 1) optimal routing and congestion control for M-MIMO multi-hop backhaul networks; 2) efficient scheduling design for M-MIMO cellular networks; and 3) energy analytics for M-MIMO wireless networks. The project aims to bridge the gaps between physical layer advances in M-MIMO and wireless networking research.