CSR:Small:System Support for Edge Computing Applications

CSR：小型：边缘计算应用的系统支持

• 批准号: 1816399
• 负责人: Qun Li
• 金额: $ 50万
• 依托单位: College of William and Mary
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-10-01 至 2023-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1816399&HistoricalAwards=false
• 关键词: CSR; Small; System; Support; Edge

Edge computing promises to enable many applications not readily supported by the cloud computing paradigm. By placing resource-rich nodes in close proximity to mobile or Internet of Things (IoT) devices, edge computing offers better responsive services, along with higher scalability and availability than traditional cloud platforms. Edge computing, as a new computing paradigm, also poses vastly different research challenges due to its distinct infrastructure from other systems. In this project, a number of novel technologies will be developed to run various applications on an edge computing system with less deployment effort and higher efficiency. Specifically, the project proposes a comprehensive architecture for edge computing application development and identifies task scheduling, service migration, data management as enabling techniques for edge computing to prevail. The project will focus on new algorithms and system design methodologies in (1) a task queue management and inter-edge collaboration system, (2) a highly efficient and robust service migration support based on Docker/Virtual Machine, and (3) a combined edge/cloud computing architecture for distributed data management. With respect to broader impacts, this project will advance the state of the art in emerging edge computing systems by designing system support to facilitate the development of a wide variety of edge computing applications, thereby having a positive impact on our daily lives. The new paradigm derived from the research will lead to a number of interesting problems that are not explored in traditional cloud or mobile systems. The findings of this research will be incorporated into new and existing undergraduate and graduate courses.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.

边缘计算有望实现许多云计算范式不容易支持的应用程序。通过将资源丰富的节点放置在靠近移动的或物联网（IoT）设备的位置，边缘计算提供了更好的响应服务，沿着具有比传统云平台更高的可扩展性和可用性。边缘计算作为一种新的计算范式，由于其与其他系统不同的基础设施，也带来了截然不同的研究挑战。在该项目中，将开发一些新技术，以在边缘计算系统上运行各种应用程序，从而减少部署工作量并提高效率。具体来说，该项目提出了一个用于边缘计算应用程序开发的综合架构，并将任务调度、服务迁移、数据管理确定为边缘计算的支持技术。该项目将重点关注以下方面的新算法和系统设计方法：（1）任务队列管理和边缘间协作系统，（2）基于Docker/虚拟机的高效和强大的服务迁移支持，以及（3）用于分布式数据管理的边缘/云计算组合架构。在更广泛的影响方面，该项目将通过设计系统支持来促进各种边缘计算应用程序的开发，从而推动新兴边缘计算系统的发展，从而对我们的日常生活产生积极的影响。从研究中得出的新范式将导致许多在传统的云或移动的系统中没有探索的有趣问题。该奖项反映了NSF的法定使命，并被认为值得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来支持。

项目成果

Efficient Privacy-Preserving Federated Learning for Resource-Constrained Edge Devices

• DOI: 10.1109/msn53354.2021.00041
• 发表时间: 2021-12
• 期刊: 2021 17th International Conference on Mobility, Sensing and Networking (MSN)
• 作者: Jindi Wu;Qi Xia;Qun Li

Neuron Manifold Distillation for Edge Deep Learning

• DOI: 10.1109/iwqos52092.2021.9521267
• 发表时间: 2021-06
• 期刊: 2021 IEEE/ACM 29th International Symposium on Quality of Service (IWQOS)
• 作者: Zeyi Tao;Qi Xia;Qun Li

Efficient Live Migration of Edge Services Leveraging Container Layered Storage

• DOI: 10.1109/tmc.2018.2871842
• 发表时间: 2019-09
• 期刊: IEEE Transactions on Mobile Computing
• 影响因子: 7.9
• 作者: Lele Ma;Shanhe Yi;Nancy Carter;Qun A. Li

vTrust: Remotely Executing Mobile Apps Transparently With Local Untrusted OS

vTrust：使用本地不受信任的操作系统透明地远程执行移动应用程序

• DOI: 10.1109/tc.2022.3152074
• 发表时间: 2022
• 期刊: IEEE Transactions on Computers
• 影响因子: 3.7
• 作者: Tang, Yutao;Qin, Zhengrui;Lin, Zhiqiang;Li, Yue;Yi, Shanhe;Xu, Fengyuan;Li, Qun
• 通讯作者: Li, Qun

FABA: An Algorithm for Fast Aggregation against Byzantine Attacks in Distributed Neural Networks

• DOI: 10.24963/ijcai.2019/670
• 发表时间: 2019-08
• 作者: Qi Xia;Zeyi Tao;Zijiang Hao;Qun A. Li