CRII: CNS: Towards an Efficient Serverless Mobile Edge Computing Network

该奖项全部或部分由2021年美国救援计划法案（公法117-2）资助。无服务器计算已成为一种事件驱动的云计算范式，与传统云计算相比具有许多优势。这包括自动化服务器管理、按使用付费定价和大规模并行性，所有这些都使其成为移动的边缘计算（MEC）的一种有吸引力的方法。然而，边缘节点的资源受限性质和MEC中突发工作负载的延迟敏感性质两者导致从服务提供商和应用程序开发人员的角度来看在MEC网络中实现无服务器计算的承诺的根本挑战。为了释放无服务器移动的边缘计算的全部潜力，该项目围绕三个相互关联的任务进行组织：（i）为保证最佳响应时间的无服务器请求开发具有本地意识的负载平衡方案，（ii）开发具有成本效益的、延迟驱动的资源配置方案，以及（iii）开发允许应用程序开发人员利用无服务器计算提供的并行性的巨大潜力的原理和技术。该项目将产生重要的社会和教育影响。MEC已经在许多领域实现了一系列物联网（IoT）应用，包括安全和监控、医疗保健、工业制造、农业等。一个高效的无服务器移动的边缘计算网络将直接影响这些应用，并最终影响整个社会。此外，该项目中开发的算法将直接添加到现有的无服务器计算平台上，我们提出的并行辅助在线学习方法将成为社区在处理基于并行计算的机器学习边缘应用程序时的基准。从教育的角度来看，该项目将开发一个广泛共享的课程，并培训本科生和研究生，同时扩大参与计算和工程。该奖项反映了NSF的法定使命，并已被认为是值得通过评估使用基金会的智力价值和更广泛的影响审查标准的支持。

This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2).Serverless computing has emerged as an event-driven cloud computing paradigm that enjoys many advantages over traditional cloud computing. This includes automated server management, pay-as-you-use pricing, and massive parallelism, all of which make it an attractive approach for mobile edge computing (MEC). However, both the resource-constrained nature of edge nodes and the latency-sensitive nature of bursty workloads in MEC result in fundamental challenges to fulfilling the promise of serverless computing in MEC networks from the perspectives of both service providers and application developers. To unleash the full potential of serverless mobile edge computing, this project is organized around three interrelated tasks: (i) the development of locality-aware load balancing schemes for serverless requests that guarantee an optimal response time, (ii) the development of cost-effective, latency-driven resource configuration schemes, and (iii) the development of principles and techniques that allow application developers to utilize the great potential of parallelism offered by serverless computing. This project will lead to important societal and educational impacts. MEC has enabled a range of Internet of Things (IoT) applications across many areas, including security and surveillance, healthcare, industrial manufacturing, agriculture, and others. An efficient serverless mobile edge computing network will directly impact these applications, and ultimately, society at large. Moreover, the developed algorithms in this project will be a direct add-on for existing serverless computing platforms, and our proposed parallelism-assisted online learning approach will be a benchmark for the community when dealing with parallel computing-based machine learning edge applications. From an education perspective, this project will develop a curriculum that will be broadly shared, as well as train undergraduate and graduate students, while broadening participation in computing and engineering.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.