Programmable Named Data-based Next Generation Software-Defined WANs

基于可编程命名数据的下一代软件定义广域网

• 批准号: RGPIN-2021-02556
• 负责人: Karmouch, Ahmed
• 金额: $ 2.99万
• 依托单位: University of Ottawa
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=761631
• 关键词: Programmable; Named; Data; based; Next

Software Defined Networking is an emerging paradigm that provides networks the ability of having extensive control over traffic flows to meet application needs. Research and practical architecture designs efforts are driving today's technology to embrace the concept of programmability of data plane devices along with an abstract layer for the control plane. Furthermore, Named Data Networking is a promising content-centric architecture that uses routable content names instead of IP addresses to achieve location-independent forwarding. With Named Data Networking, Software Defined Networking will transform the network Industry into an open ecosystem by enabling the development of new softwarized network architectures, such as a softwarized wide-area named-data-network architecture, that can serve resource-constrained devices interacting with modern dispersed and content-centric applications. We believe that the stars are now aligned to develop softwarized wide-area named-data-networks thanks to (i) advances in network softwarization and virtualization;(ii) the growing trend of adopting programmable switches; (iii) the emergence of content centric networking concepts and (iv) the maturation of artificial intelligence techniques. However, a number of fundamental challenges must be addressed in order to create the synergy among these technologies and realize a softwarized network. The long-term objective of this research program is to address key research challenges for realizing the vision of a novel softwarized wide-area named-data-network. To achieve this goal, we will focus on the following three primary short-term objectives: (1) Design a novel virtualized and programmable data plane that can serve both IP and named-data packets. The data plane must be able to parse/de-parse and process packets at line speed, and offer a wide range of packet processing capabilities. (2) Develop a consistent and scalable control plane orchestration with the goal of minimizing the controller and data plane interactions while achieving scalability and consistency to ensure the efficient coexistence of network functions for both IP and named-data flows. (3) Extend the data plane with intelligence to adapt to the needs of evolving applications (e.g., the ability to create preconfigured traffic highways with no store-and-forward processing) and to bring new customized services to content-based applications (e.g., line-speed in-network packet load processing and caching). Realizing the vision of softwarized wide-area named-data-network will be essential in future next generation networks and services. It will contribute to rapid deployment of innovative network services while reducing network operator's capital and operational expenditures. The proposed research program will produce the trainees needed to drive next wave of network innovation. The outcome of this research will serve as the seed for Canada's national leadership in the fast-paced digital advances.

软件定义网络是一种新兴的模式，它为网络提供了对流量进行广泛控制的能力，以满足应用需求。研究和实际的架构设计工作正在推动当今的技术将数据平面设备的可编程性的概念沿着控制平面的抽象层。此外，命名数据网络是一种很有前途的以内容为中心的架构，它使用可路由的内容名称而不是IP地址来实现位置无关的转发。借助命名数据网络，软件定义网络将通过开发新的软件化网络架构（例如软件化广域命名数据网络架构）将网络行业转变为开放的生态系统，该架构可以为与现代分散和以内容为中心的应用程序交互的资源受限设备提供服务。我们相信，由于（i）网络软件化和虚拟化的进步;（ii）采用可编程交换机的增长趋势;（iii）以内容为中心的网络概念的出现;以及（iv）人工智能技术的成熟，现在的星星已经对准开发软件化广域命名数据网络。然而，为了在这些技术之间创造协同效应并实现软件化网络，必须解决一些根本性的挑战。该研究计划的长期目标是解决实现新型软件化广域命名数据网络愿景的关键研究挑战。为了实现这一目标，我们将集中在以下三个主要的短期目标：（1）设计一个新的虚拟化和可编程的数据平面，可以服务于IP和命名数据包。数据平面必须能够以线速解析/解解析和处理数据包，并提供广泛的数据包处理功能。(2)开发一致且可扩展的控制平面编排，目标是最大限度地减少控制器和数据平面交互，同时实现可扩展性和一致性，以确保IP和命名数据流的网络功能高效共存。(3)通过智能扩展数据平面，以适应不断发展的应用程序的需求（例如，创建没有存储转发处理的预配置的业务高速公路的能力）以及将新的定制服务带到基于内容的应用（例如，线速网内分组加载处理和高速缓存）。实现软件化的广域命名数据网络的愿景将是未来下一代网络和业务的关键。它将有助于创新网络服务的快速部署，同时减少网络运营商的资本和运营支出。拟议的研究计划将培养推动下一波网络创新所需的培训生。这项研究的结果将成为加拿大在快速数字化发展中的国家领导地位的种子。