CNS Core: Small: Towards a Knowledge Plane for Coordinating Network Policies

在计算机网络研究中应用知识管理的概念和方法的兴趣越来越大。虽然现有的工作主要围绕网络数据（即网络状态）展开，但该项目询问知识管理系统是否以及如何处理网络语义（即网络的含义）。为了了解语义网络知识管理的好处，本项目研究了网络策略协调。这些解决方案将大大提高网络策略的可预测性、灵活性、可管理性和安全性，为推进现代通信网络的策略基础设施提供了新的视角。该项目的贡献是开发网络知识平面，这是一个编排基础设施，用于不同的、可能冲突的网络策略，这些策略可以以复杂的方式进行交互：在第一个推动力中，项目引入数据完整性约束（ic）作为统一的抽象，将策略从网络中提出来，为策略推理创建一个通用框架；接下来，通过利用与IC维护的连接，该项目将为软件定义网络（sdn）开发编排服务；最后，通过利用与语义查询转换的连接，该项目将开发一个协商工具，以协调全球互联网独立利益相关者的政策制定。该项目的成果将为政策联网面临最严峻挑战的领域带来重大收益。它将使网络策略更容易理解、预测、重用、管理和发展。此外，该项目将在两个方面产生广泛影响：(1)编排框架将广泛适用，促进sdn超越其初始部署，并加速部署新的策略扩展，以发展互联网；(2)网络中知识表示和推理技术的新颖使用将开辟一个新的研究空间，围绕逻辑属性安排策略，创造一个重新思考通信网络基本策略问题的机会。项目网站（https://ravel-net.org）将继续运作，以发布项目进展情况。具体来说，支持工具将是开源的。源代码也将在github服务器（https://github.com/ravel-net）上提供，并附有在线文档，以方便独立验证。网络痕迹和真实数据集将被收集并公开。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。

There has been an increased interest in applying knowledge management concepts and methods in computer networking research. While existing work centers around network data (i.e., the state of the network), this project asks if and how knowledge management systems can handle network semantics (i.e., the meaning of the network). To understand the benefits of semantic network knowledge management, this project investigates network policy coordination. The solutions will substantially improve the predictability, flexibility, manageability, and security of network policies, offering a fresh perspective to advance the policy infrastructure for modern communication networks.The contribution of the project is to develop network knowledge plane, an orchestration infrastructure for disparate, potentially conflicting network policies that can interact in complex ways: in the first thrust, the project introduces data integrity constraints (ICs) as a unifying abstraction to lift the policies out of the network, creating a common framework for policy reasoning; next, by leveraging connections to IC maintenance, the project will develop an orchestration service for software defined networks (SDNs); finally, by leveraging connections with semantic query transformation, the project will develop a negotiation facility to coordinate policy makings by independent stakeholders of the global Internet.The results of the project will offer significant gains to areas that are the most critical challenges to policy-based networking. It will make network policies easier to understand, predicate, reuse, manage, and evolve. Moreover, the project will have broad impact for two ways: (1) the orchestration framework will be broadly applicable, facilitating the SDNs to move beyond their initial deployment, and accelerating the deployment of new policy extensions to evolve the Internet, (2) the novel use of knowledge representation and reasoning techniques in networking will open up a new research space that arranges policies around logical properties, creating an opportunity to rethink the fundamental policy issues of communication networks.The project website (https://ravel-net.org) will be maintained to disseminate the progress of the project. Specifically, supporting tools will be made open-source. Source code will also be available on the github server (https://github.com/ravel-net) with online documentation to facilitate independent validation. Networking traces and real-world dataset will be collected and made public.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.