NeTS: Medium: Connecting the Next Billion: Rethinking Wireless Network Design Principles for the Internet-of-Everything

在未来十年内，预计将有数十亿台设备无线连接到互联网，以实现智能家居、身体/健康监测和环境监测等应用。这个庞大的物联网（IoT）预计将在一个小环境中的数百个设备中活跃，并发送间歇性但及时的数据。现有的无线技术不适合管理物联网应用，因为它们的设计是为了处理空间不是很密集的智能手机等设备，并且在访问无线媒体时交换大量数据。本提案的研究目标是满足对包含从物理和介质访问控制到网络的所有层的新物联网架构的需求。提出的设计利用物联网应用的独特特性（如间歇性、高密度和用户动态），实现“干扰平均”现象，为更可靠、更快速的服务保障提供基础。这种方法非常适合物联网应用，因为它的传输是即时的，可靠的，并且在密集部署下有效地工作。该项目有望成为低成本公共接入环境监测、医疗保健和智能生活领域关键无线服务的关键推动因素。到2020年，预计将有250亿至750亿台设备连接到互联网，构成所谓的庞大物联网（IoT）。许多物联网应用将基于大量低成本设备动态地与接入点或邻近设备建立连接，以通信小束延迟敏感数据。这与典型的无线局域网（WLAN）设置形成鲜明对比，后者由位置相对稀疏的站点产生密集的流量。因此，现有的无线资源分配技术不适合服务于即将到来的物联网网络。因此，迫切需要开发高效实用的通信策略，以支持大量密集的移动设备产生间歇性和延迟敏感的流量——这在新兴的物联网设备网络中越来越重要。本研究解决了这一需求，为“轻量级”通信和网络战略的原则开发奠定了基础和必要的手段，为新兴物联网网络的有效运行提供了低复杂性和低开销的解决方案。为此，本研究提出了一种包容干扰的范式，以适应即将到来的物联网应用的上述非传统动态。所提出的策略允许许多用户同时共享资源，从而绕过了现有无线解决方案的沉重成本。该方法的新颖之处在于：(i)物理层和网络层操作考虑了快速用户和间歇性流量动态以及不同的服务质量要求，以及它们对典型用户观察到的干扰的影响；（ii）多用户访问设计结合了避免冲突范式和拥抱冲突范式的优势；（iii）通过减少或消除实际实施的信号要求，将间接负荷降至最低；（iv）它由强大的实验组件支持，其最终目标是在实际物联网应用中实例化理论上新颖的最佳技术。这些目标的实现需要在代码设计、资源分配、分布式算法设计和实现等不同领域开发新工具。本项目提出的跨学科研究将有助于填补我们对大规模网络如何有效管理新兴设备和流量动态的理解空白。新的方法和技术反过来又期望对个别学科本身作出根本性的贡献。此外，该项目将帮助参与的学生均衡地接触各种理论和应用技术，涵盖不同领域，包括无线网络，优化，算法设计和无线系统实现，从而有助于培养合格的劳动力。

Within the next decade, billions of devices are expected to be connected to the Internet wirelessly to enable applications like smart homes, body/health monitoring, and environmental monitoring, among many others. This vast Internet of Things (IoT) is expected to be active in the hundreds of devices within a small environment and that send intermittent, but timely data. Existing wireless technologies are unsuitable for managing IoT application since they are designed to handle devices such as smartphones that are not very dense in space and exchange huge amounts of data when they access wireless medium. The research objective of this proposal is to answer the need for a new IoT architecture encompassing all layers from Physical and Medium Access Control to Networking. Proposed design utilizes the unique characteristics (e.g., intermittence, high density, and user dynamics) of IoT applications to achieve an "interference-averaging" phenomenon that provides the foundation for more reliable and rapid service guarantees. This approach is well-suited to IoT applications since its transmissions are immediate, reliable, and works efficiently under dense deployments. This project is expected to be a key enabler for low-cost public access to critical wireless services in environmental monitoring, healthcare, and smart-living.By 2020, there will be anywhere between an estimated 25-75 billion devices that will connect to the Internet, making up the so-called vast Internet-of-Things (IoT). Many of the IoT applications will be based on a large population of low-cost devices dynamically making connections with access points or neighboring devices to communicate small bundles of delay-sensitive data. This contrasts sharply with the typical wireless local area network (WLAN) setup with intense traffic generated by relatively sparsely positioned stations. Accordingly, the existing wireless resource allocation technologies are not well-suited to serve the upcoming IoT network. Therefore, there is a pressing need for the development of efficient and practical communication strategies to support a large number of densely-packed mobile devices generating intermittent and delay-sensitive traffic - a scenario of increasing significance in the emerging IoT-device networks.This research addresses this need by undertaking the well-founded development of a framework for establishing the foundations and the means necessary for the principled development of "light-weight" communication and networking strategies that provide low-complexity and low-overhead solutions for the provably efficient operation of emerging IoT networks. To that end, this research proposes an interference-embracing paradigm to accommodate the above non-traditional dynamics of upcoming IoT applications. Proposed strategy allows many users to share the resources simultaneously, thereby bypassing the heavy costs of existing wireless solutions. The approach is novel in that: (i) the physical and network layer operation accounts for the fast user and intermittent traffic dynamics and the diverse quality-of-service requirements, as well as their impact on the observed interference by a typical user; (ii) the multi-user access design combines the strengths of collision-avoiding and collision-embracing paradigms; (iii) the overhead load is minimized by reducing or eliminating the signaling requirements for practical real-world implementation; and (iv) it is supported by a strong experimentation component with the ultimate objective of instantiating novel theoretically optimal techniques in practical IoT applications.The attainment of these objectives requires the development of new tools in such diverse areas as code design, resource allocation, distributed algorithm design, and implementation. The cross-disciplinary research proposed in this project will help fill the gap in our understanding on how large scale networks can effectively manage the emerging device and traffic dynamics. The new approaches and techniques are in turn expected to make fundamental contributions to the individual disciplines themselves. Also, the project will help provide participating students a balanced exposure to a wide variety of theoretical and applied techniques spanning different fields including wireless networking, optimization, algorithm design, and wireless systems implementation, and thereby contribute to the education of a competent workforce.