CRII: CSR: Opportunistic Computation Offloading for Battery-Free Computing Devices

物联网（IoT）的指数级增长提出了一个问题，即如何在不牺牲其功能的情况下可持续地为数十亿新IoT设备提供动力。目前的无电池物联网设备是通过收集环境能源来供电的，但存在一些缺点，包括不可预测的能源可用性、有限的处理能力和严重受限的内存资源。该项目旨在通过利用数十亿持续供电（即，电池和墙壁供电）物联网设备已经在现场通过卸载额外的计算，他们的需求。为了实现这一目标，该项目将（1）研究具有异构无线电支持的集成框架，以在无电池系统上执行部分计算，同时在需要时利用墙壁或电池供电的边缘设备进行额外计算;（2）设计一个调度器，该调度器考虑设备上的计算不确定性，可收获的能量特性和时序约束，以确定最佳通信和计算策略;以及（3）研究低开销技术以经由中间结果分析来估计设备上计算的不确定性（例如，深度神经网络（Deep Neural Networks）这项研究的成功完成可能会通过增加无电池计算设备的功能来影响电子废物。增强的无电池计算能力可能会影响许多领域，包括医疗保健（例如，连续健康监测器），农业（例如，土壤湿度监测），空间探索（例如，超小型卫星）以及其他需要纵向、可持续、零维护数据收集的科学领域。这项研究的成果将使多个科学领域的研究人员和软件开发人员能够利用无电池、可持续的计算来进行应用和研究。除了为本科生和研究生提供研究机会外，研究成果还将被纳入大学课程。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

The exponential growth of the Internet of Things (IoT) raises the question of how to power billions of new IoT devices sustainably without sacrificing their capabilities. Current battery-free IoT devices are powered by harvesting ambient energy, but with several shortcomings, including unpredictable energy availability, limited processing capability, and severely constrained memory resources. This project aims to address these challenges by leveraging the billions of persistently powered (i.e., battery- and wall-powered) IoT devices already in the field by offloading additional computation to them on demand. To realize this goal, this project will (1) study an integrated framework with heterogeneous radio support to perform partial computation on battery-free systems while also utilizing wall- or battery-powered edge devices for additional computation when needed; (2) design a scheduler that considers on-device computational uncertainty, harvestable energy characteristics, and timing constraints to determine optimal communication and computation strategies; and (3) investigate low overhead techniques to estimate the uncertainty of on-device computation via intermediate result analysis (e.g., from deep neural networks). The successful completion of this research could impact electronic waste (e-waste) by increasing the capabilities of battery-free computing devices. Enhanced battery-free computing capabilities could impact many fields, including healthcare (e.g., continuous health monitors), agriculture (e.g., soil moisture monitoring), space exploration (e.g., nanosatellites), and other scientific fields that require longitudinal, sustainable, zero-maintenance data gathering. The outcomes of this research will allow researchers and software developers across several scientific domains to utilize battery-free, sustainable computing for their applications and research. In addition to providing research opportunities to undergraduate and graduate students, the research outcomes will be integrated into university course curriculum.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.