Collaborative Research: CNS Core: Medium: IoCT: System Mechanisms for Enabling an Internet of Collaborative Things

协作研究：CNS 核心：媒介：IoCT：实现协作物联网的系统机制

• 批准号: 2211302
• 负责人: Prashant Shenoy
• 金额: $ 39.87万
• 依托单位: University of Massachusetts Amherst
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-10-01 至 2025-09-30
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2211302&HistoricalAwards=false
• 关键词: Collaborative; Research; CNS; Core; Medium

The rapid growth of Internet-of-Things (IoT) devices points to a future with trillions of devices deployed at a planetary scale for use in diverse application domains, ranging from healthcare, to transportation, to energy, to manufacturing, to environmental sensing, and more. Such growth poses significant challenges. First, due to the relatively slower growth of cloud infrastructure, the current practice of offloading data from IoT devices for computing at cloud servers will not be sustainable due to scaling challenges. Second, future IoT devices will operate in environments where they will host multiple applications simultaneously and move across physical spaces with different owners, thus requiring flexible sharing across trust domains of sensing, processing, communication, and action resources present in the devices and the physical spaces. Addressing these emerging challenges requires a radical rethinking of the current device-edge-cloud architecture of IoT systems. In this project, the researchers are exploring a new architecture approach called the “Internet of Collaborating Things” (IoCT). In IoCT, instead of offloading device data to edge and cloud servers, the application computation is distributed across an amorphous collection of interconnected IoT devices that produce and consume data, while harnessing emerging on-device compute accelerators. The project will address three key challenges associated with the IoCT vision: (i) secure multitenancy on resource-constrained IoT devices; (ii) secure collaboration among IoT devices; and (iii) resilient management of resources distributed across IoT devices and the spaces they are operating in. The societal benefits of the project are twofold. First, it will provide the foundations for a massive-scale IoT infrastructure, without the large cost and energy footprint associated with using traditional server infrastructure. Second, it will allow IoT applications involving rich high-dimensional sensors and sophisticated physical control to be run anywhere by securely harnessing resources present in nearby devices and physical spaces. This includes autonomous urban robots interacting in complex manners with social spaces and built environments. Additionally, the project team will undertake specific broader impact activities to advance education, transition technology, and broaden participation. In particular, for creating and nurturing an IoCT community of researchers and practitioners, the project team will release open-source software, create experimental testbeds, organize workshops, and contribute to broadening participation in computing through outreach to local high schools using two pre-college programs: the Los Angeles Computing Circle at UCLA and the Summer Turing Institute at UMass Amherst.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.

物联网 (IoT) 设备的快速增长预示着未来将在全球范围内部署数万亿设备，用于不同的应用领域，从医疗保健、交通、能源、制造到环境传感、这种增长带来了重大挑战。首先，由于云基础设施的增长相对较慢，由于扩展挑战，当前从物联网设备卸载数据以进行云服务器计算的做法将无法持续。在同时托管多个应用程序的环境中运行并与不同的所有者跨物理空间移动，因此需要跨设备和物理空间中存在的传感、处理、通信和操作资源的信任域进行灵活共享，解决这些新出现的挑战需要对当前的设备边缘进行彻底的重新思考。在该项目中，研究人员正在探索一种称为“协作物联网”(IoCT) 的新架构方法。在 IoT 中，应用程序计算不是将设备数据卸载到边缘和云服务器，而是分布在各处。无定形的集合该项目将解决与物联网愿景相关的三个关键挑战：(i) 资源有限的物联网设备之间的安全多租户；物联网设备；(iii) 对分布在物联网设备及其运行空间的资源进行弹性管理。该项目的社会效益是双重的，首先，它将为大规模物联网基础设施提供基础，而无需大规模建设。相关成本和能源足迹其次，通过安全地利用附近设备和物理空间中的资源，它将允许物联网应用程序丰富的高维传感器和复杂的物理控制在任何地方运行，这包括以复杂的方式与社交互动的自主城市机器人。空间和建筑环境。项目团队将开展具体的更广泛影响的活动，以促进教育、过渡技术和扩大参与，特别是为了创建和培育研究人员和从业者的物联网社区，项目团队将发布开源软件。 ，创建实验测试平台，组织研讨会，并通过推广影响到当地高中，使用两个大学预科项目：加州大学洛杉矶分校的洛杉矶计算圈和麻省大学阿默斯特分校的夏季图灵研究所，为扩大计算参与做出贡献。该奖项是 NSF 的法定使命，并被认为是值得的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来提供支持。

项目成果

Dělen: Enabling Flexible and Adaptive Model-serving for Multi-tenant Edge AI

Dälen：为多租户边缘 AI 提供灵活、自适应的模型服务

• DOI: 10.1145/3576842.3582375
• 发表时间: 2023-05-09
• 期刊: Proceedings of the 8th ACM/IEEE Conference on Internet of Things Design and Implementation
• 作者: Qianlin Liang;Walid A. Hanafy;Noman Bashir;Ahmed Ali;David E. Irwin;Prashant Shenoy
• 通讯作者: Prashant Shenoy

Understanding the Benefits of Hardware-Accelerated Communication in Model-Serving Applications

了解模型服务应用程序中硬件加速通信的好处

• DOI: 10.1109/iwqos57198.2023.10188785
• 发表时间: 2023-06
• 期刊: Proceedings of IEEE/ACM 31st International Symposium on Quality of Service (IWQoS
• 作者: Hanafy, Walid A.;Wang, Limin;Chang, Hyunseok;Mukherjee, Sarit;Lakshman, T. V.;Shenoy, Prashant
• 通讯作者: Shenoy, Prashant

Model-driven Cluster Resource Management for AI Workloads in Edge Clouds

边缘云中 AI 工作负载的模型驱动集群资源管理

• DOI: 10.1145/3582080
• 发表时间: 2022-01-18
• 期刊: ACM Transactions on Autonomous and Adaptive Systems
• 影响因子: 2.7
• 作者: Qianlin Liang;Walid A. Hanafy;Ahmed Ali;Prashant Shenoy
• 通讯作者: Prashant Shenoy