ICE-T: RC: Horizontal Resource Management in Distributed Edge Clouds

下一代物联网(IoT)、智能可穿戴设备和增强现实应用和设备的延迟、带宽和移动性需求使传统云平台无法满足其需求。边缘云计算将计算和存储资源部署在网络的边缘，并通过云模式提供，已成为解决这些新兴应用需求的一种有前途的替代方案。本项目研究边缘云中网络感知的横向资源管理，并将重点关注三个协同主题。首先，该团队将为边缘云设计网络感知协调算法，包括结合多路径传输和虚拟机迁移的新边缘云迁移技术，以在用户和设备移动性的情况下提供网络透明的低延迟访问。其次，该团队将研究分布式边缘云中的网络感知位置，在了解分布式边缘和虚拟网络功能交互的同时，整合应用的延迟和带宽需求。第三，项目研究将解决大规模分布边缘云的控制面架构的基本问题，并将研究未来边缘云的分散式和层次式控制面架构。项目团队将通过协调算法(如地球弹性缩放算法)以及边缘迁移和网络感知布局的视角来研究这两种架构的权衡和优势。我们的研究还将涉及原型实施和在多个实验研究试验台上的大规模实验。该项目将扩展马萨诸塞大学阿默斯特分校和瑞典Umea大学之间的合作。该奖项反映了NSF的法定使命，并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，被认为值得支持。

The latency, bandwidth, and mobility needs of the next-generation Internet of Things (IoT), smart wearables, and augmented reality applications and devices makes traditional cloud platforms ill-suited to their needs. Edge cloud computing, where computational and storage resources are deployed at the edge of the network and offered through a cloud model, has emerged as a promising alternative to address the needs of these emerging applications.This project examines network-aware horizontal resource management in edge clouds and will focus on three synergistic topics. First, the team will design network-aware orchestration algorithms for edge clouds, including new edge cloud migration technique that combines multi-path transport and virtual machine migration to provide network-transparent low-latency access in the face of user and device mobility. Second, the team will study network-aware placement in distributed edge clouds that incorporates the application's latency and bandwidth needs while being aware of interacting distributed edge and virtualized network functions. Third, the project research will address the fundamental issue of the control plane architecture for massively distributed edge clouds and will study both decentralized and hierarchical control plane architectures for future edge clouds. The project team will study the tradeoffs and benefits of both architectures through the lens of orchestration algorithms such as geo-elastic scaling algorithms as well as edge migration and network-aware placement. Our research will also involve prototype implementation and large-scale experimentation on multiple experimental research testbeds.This project will extend a collaboration between the University of Massachusetts, Amherst, and Umea University, Sweden.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.