CNS Core: Small: Bridging the Silos of Edge Computing with Connected Namespaces

边缘计算提供了在数据网络路径中嵌入的计算元素上更接近其来源处理数据的机会。它可以为许多重要的延迟敏感和数据密集型应用程序带来变革，例如智能城市，自主系统和交通，增强和虚拟现实（AR/VR），视觉分析等。然而，实现边缘计算的潜在好处并不是微不足道的，因为它依赖于跨不同资源和服务做出及时决策的能力，跨许多孤立的提供商进行分解，并通过动态变化的操作条件进行管理。对于从边缘计算中寻求好处的应用程序，这在获取正确的可用资源集方面提出了挑战。对于向边缘计算提供资源的利益相关者来说，这在保持对其资源的期望控制程度方面提出了挑战。为了解决这个问题，该项目开发了一个新的去中心化边缘交换，它建立在一个新的系统级抽象（称为连接命名空间）上，以实现边缘计算中动态的、细粒度的和跨利益相关者的资源编排。重要的是，该交易所本地集成了定制加密技术的使用，即多方计算（MPC），以实现完全的隐私，同时满足整个系统的所有实时性能要求。本项目开发的技术允许边缘计算应用程序利用最佳可用边缘资源，即使这些资源分散在多个参与方之间并由多个参与方拥有，同时仍然为每个利益相关者提供有关其资源使用和共享策略的控制。在不牺牲机密性和控制的情况下提供性能的能力消除了对边缘计算的关键担忧，否则会阻碍其采用。因此，本研究为新边缘基础设施层的未来发展提供了一条道路，支持重要类别的新兴应用程序，包括边缘硬件和软件技术的新执行平台，并催化边缘计算作为收入流和经济增长的新来源的采用。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Edge computing provides opportunities to process data closer to its sources, on computing elements embedded in the data network path. It can be transformative for many important classes of latency-sensitive and data-intensive applications, such as Smart City, autonomous systems and transportation, augmented and virtual reality (AR/VR), visual analytics, etc. However, realizing the potential benefit of edge computing is not trivial, because it relies on the ability to make timely decisions across diverse resources and services, disaggregated across many siloed providers, and governed by dynamically changing operating conditions. For applications seeking benefits from edge computing, this presents challenges in obtaining the right set of available resources. For stakeholders contributing resources to edge computing, this presents challenges in maintaining desired degree of control over their resources. To address this problem, this project develops a new Decentralized Edge Exchange that builds on a new system-level abstraction, called connected namespaces, to enable dynamic, fine-grained, and cross-stakeholder resource orchestrations in edge computing. Importantly, the exchange natively integrates use of tailored cryptographic techniques, namely, multi-party computation (MPC), to achieve full privacy while meeting all the real-time performance requirements of the complete system. The technologies developed in this project permit edge computing applications to utilize the best available edge resources, even when those are fragmented across and owned by multiple parties, while still providing each stakeholder with controls regarding their resource use and sharing policies. The ability to deliver on performance without sacrificing confidentiality and controls removes critical concerns about edge computing, which otherwise hamper its adoption. As a result, this research presents a path for future developments in the new edge infrastructure tier, support for important classes of emerging applications, new execution platforms comprising novel hardware and software technologies for the edge, and catalyzes the adoption of edge computing as a new source of revenue streams and economic growth.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.