CRII: CNS: A Hybrid Software Defined Networking-Information Centric Networking Framework for the Reuse of Computation at the Network Edge

CRII：CNS：一种混合软件定义网络-以信息为中心的网络框架，用于在网络边缘重用计算

• 批准号: 2306685
• 负责人: Spyridon Mastorakis
• 金额: $ 17.47万
• 依托单位: University of Notre Dame
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-01-01 至 2024-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2306685&HistoricalAwards=false
• 关键词: CRII; CNS; Hybrid; Software; Defined

The adoption of next-generation applications, such as Augmented Reality and Internet-of-Things (IoT), is expected to grow rapidly in the years to come. The number of IoT devices is expected to reach 75 billion by 2025, while the scale of potential use-cases (e.g., smart cities with hundreds of thousands or millions of residents) is also expected to grow. In such use-cases, users may request the processing of data with temporal, spatial, or semantic similarity by servers physically close to them. The repeated processing of considerably similar data may result in the same processing outputs and thus the execution of duplicate (redundant) computation. This project advances our understanding of the notion of “computation reuse”, where the processing outputs are reused/shared among processing requests for similar data, having the potential to: (i) reduce the time for the processing of similar data and thus the overall latency perceived by next-generation applications; and (ii) eliminate the execution of redundant computation, so that the available computing resources are effectively utilized. Our project will result in a novel network framework, so that requests for the processing of similar data are identified and forwarded to the same servers, facilitating the reuse of computation in realistic infrastructure deployments. The research outcomes of this project will be incorporated into university course curriculum, while this project will also offer opportunities that broaden participation and diversity in STEM.This project aims to advance the state-of-the art in edge computing architectures in order to facilitate the reuse of computation through a novel, hybrid network framework based on Information-Centric Networking (ICN)/Named-Data Networking (NDN) and Software-Defined Networking (SDN). The proposed framework capitalizes on the best of both ICN/NDN and SDN to realize a service-centric, computation reuse-aware, and “softwarized” architecture through: (i) the semantically meaningful ICN/NDN naming and stateful forwarding plane for the seamless and adaptive task forwarding towards edge servers that can reuse similar previously executed tasks; and (ii) the separation of forwarding and control planes and the programmable (logically) centralized SDN intelligence for the orchestration of the computation reuse semantics. This project will tackle fundamental challenges in the design and development of network systems for the realization of computation reuse in realistic edge computing environments, where multiple edge servers may be available for redundancy, fault tolerance, and load balancing purposes, facilitating the pervasive edge computing deployment.This project is jointly funded by Networking Technology and Systems(NeTS)program and the Established Program to Stimulate Competitive Research (EPSCoR).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.

预计在未来几年中，预计采用下一代应用程序，例如增强现实和贸易Internet（IoT）。到2025年，物联网设备的数量预计将达到750亿，而潜在用例（例如，具有数十万或数百万居民的智能城市的规模也将增长。在这样的用例中，用户可以通过物理接近的服务器要求使用临时，空间或语义相似性来处理数据。考虑相似数据的重复处理可能会导致相同的处理输出，从而执行重复（冗余）计算。该项目提高了我们对“计算重复使用”概念的理解，其中处理输出在处理请求中的相似数据的请求中被重复使用/共享，有可能：（i）减少处理相似数据的时间，从而减少下一代应用程序所感知的整体延迟； （ii）消除冗余计算的执行，以便有效地利用可用的计算资源。我们的项目将导致一个新颖的网络框架，因此可以将类似数据处理的请求确定并转发到同一服务器，从而支持现实基础架构部署中的计算重复使用。该项目的研究成果将纳入大学课程课程，而该项目还将提供扩大参与和多样性的机会。该项目旨在推进边缘计算体系结构的最先进的艺术，以通过基于信息的网络（ICN）网络（iCN）网络（new newdn）（ndn）（ndn）（ndn）（ndn）（ndndn）（ndndn）（ndndn）（ndndndn）（ndndn）（ndndn），以促进计算的重新使用。拟议的框架可以利用ICN/NDN和SDN的最佳能力实现以服务为中心的，计算重复使用的意识，并通过：（i）语义上有意义的ICN/NDN NAMING和Stane-NDN naming and State flectry Plaching and the eamemless and the eameless and to Edgaige tocriptive to Egge tocrize tocress fortherver to Enge tocress forthers the Edge Edge toce the Edge Edge the可以重复使用类似的执行任务； （ii）转发和控制平面的分离以及可编程的（逻辑上）集中式SDN智能，用于编排计算语义语义。该项目将解决网络系统设计和开发的基本挑战，以实现实际边缘计算环境中的计算重复使用，其中可能可用于多个边缘服务器可用于冗余，容忍度和负载平衡目的，并支持普遍的边缘计算部署。该项目由网络技术和系统（网络）计划（网络）计划（EPSIS）计划（EPSIS）启发（EPSIS）启发（EPSIS）。 NSF的法定使命，并使用基金会的知识分子优点和更广泛的审查标准来评估，被认为是宝贵的支持。

项目成果

Reservoir: Named Data for Pervasive Computation Reuse at the Network Edge

• DOI: 10.1109/percom53586.2022.9762397
• 发表时间: 2021-12
• 期刊: 2022 IEEE International Conference on Pervasive Computing and Communications (PerCom)
• 作者: Md Washik Al Azad;Spyridon Mastorakis

The Promise and Challenges of Computation Deduplication and Reuse at the Network Edge

• DOI: 10.1109/mwc.010.2100575
• 发表时间: 2021-09
• 期刊: IEEE Wireless Communications
• 影响因子: 12.9
• 作者: Md Washik Al Azad;Spyridon Mastorakis

An NDN-Enabled Fog Radio Access Network Architecture with Distributed In-Network Caching

• DOI: 10.1109/icc45041.2023.10279580
• 发表时间: 2023-01
• 期刊: ICC 2023 - IEEE International Conference on Communications
• 作者: Sifat Ut Taki;Spyridon Mastorakis

Harpocrates: Anonymous Data Publication in Named Data Networking

• DOI: 10.1145/3532105.3535025
• 发表时间: 2022-01
• 期刊: Proceedings of the 27th ACM on Symposium on Access Control Models and Technologies
• 作者: Md Washik Al Azad;R. Tourani;Abderrahmen Mtibaa;Spyridon Mastorakis