CAREER: Protocols for Low-Power Wide-Area Networks in White Spaces

该项目将整合电视空白区域低功耗广域网（LPWAN）的设计和实施方面的研究和教育。lpwan使低功耗设备能够为各种物联网（IoT）应用进行长距离传输。今天，lpwan有许多主要的限制，使其采用具有挑战性。首先，它们依靠有线基础设施来集成多个网络以覆盖非常大的区域（例如，智能城市）。缺乏适当的基础设施阻碍了它们适用于农村/偏远广域应用，如农业（如智能农业）和工业物联网（如油田监测）。其次，有限频谱下lpwan的快速增长提出了共存的挑战。第三，当前的（非蜂窝）lpwan不能有效地支持移动节点，例如智能农业中的拖拉机和无人机。最后，lpwan还没有设计成支持实时通信，这阻碍了它们在许多应用（例如，过程控制）中的采用。该项目将通过开发SNOW（白色空间传感器网络）的理论基础和系统来解决这些挑战，这是一种利用未使用电视频谱的LPWAN，称为空白空间。结果可扩展到许多其他lpwan。由于有大量的空白空间，该项目将为许多农村应用提供连接。它将通过课程开发、学生研究以及向少数民族/代表性不足的学生和K-12学生提供服务来整合教育。该项目将设计并实现基于SNOW的LPWAN架构和完整协议栈，以支持以下可扩展集成、共存、移动性和实时通信。(1)通过最小化集成网络中的时延，实现多个snow的可扩展无缝集成。考虑到质量和执行时间之间的权衡，求解方法包括全局优化和快速启发式。(2)提出了一种基于强化学习的新方法来处理多个独立网络共存的问题。这是通过开发一个在低功耗节点上实用的高效q学习框架来实现的。这将是首个用于LPWAN和处理低功耗网络共存的q -学习方法。(3)提出轻量级跨层方法，通过处理多普勒效应和白空间地理空间变化的影响，实现SNOW节点的移动性。(4)提出了SNOW的实时通信框架，这将是LPWAN实时调度的第一个成果。(5)它将在TI CC1310和通用软件无线电外设上实现所提出的协议。这些协议将通过在两个不同的无线电环境中进行实验来进行评估——一个是城市试验台，另一个是智能农业试点农田。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project will integrate research and education in the design and implementation of a Low-Power Wide-Area Network (LPWAN) in the TV White Spaces. LPWANs enable low-power devices to transmit over long distances for various Internet-of-Things (IoT) applications. Today, LPWANs have a number of major limitations making their adoption challenging. First, they rely on wired infrastructure for integrating multiple networks to cover very large areas (e.g., smart city). Lack of proper infrastructure hinders their applicability to rural/remote wide-area applications such as agricultural (e.g., smart farming) and industrial IoT (e.g., oil-field monitoring). Second, rapid growth of LPWANs in the limited spectrum raises the challenge of coexistence. Third, current (non-cellular) LPWANs are not designed to effectively support mobile nodes, e.g., tractors and drones in smart farming. Finally, LPWANs are not yet designed to support real-time communication hindering their adoption for many applications (e.g., process control). This project will address these challenges by developing theoretical foundations and systems for SNOW (Sensor Network Over White Spaces), an LPWAN that exploits unused TV spectrum, called white spaces. The results are extendable to many other LPWANs. Due to abundant white spaces, this project will enable connectivity for many rural applications. It will integrate education through course development, student research, and outreach to minority/underrepresented and K-12 students. This project will design and implement an LPWAN architecture and complete protocol stack based on SNOW to support scalable integration, coexistence, mobility, and real-time communication as follows. (1) It proposes a scalable seamless integration of multiple SNOWs by minimizing latency in the integrated network. Considering tradeoffs between quality and execution time, the solution approach includes both global optimization and fast heuristics. (2) It proposes a novel approach based on Reinforcement Learning to handle coexistence with many independent networks. This is done by developing an efficient Q-learning framework that is practical at low-power nodes. This will be the first Q-learning approach for LPWAN and for handling coexistence in a low-power network. (3) It proposes lightweight cross-layer approaches to enable mobility of SNOW nodes by handling Doppler effect as well as the effects of geospatial variation of white spaces. (4) It proposes a real-time communication framework for SNOW which will be the first result on real-time scheduling for LPWAN. (5) It will implement the proposed protocols on TI CC1310 and also on universal software radio peripheral devices. The protocols will be evaluated through experiments in two different radio environments -- an urban test-bed and an agricultural field piloting smart farming.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.