CNS Core: Medium: Collaborative: Exploring and Exploiting Learning for Efficient Network Control: Non-Stationarity, Inter-Dependence, and Domain-Knowledge

CNS 核心：中：协作：探索和利用学习实现高效网络控制：非平稳性、相互依赖和领域知识

• 批准号: 1901103
• 负责人: Zhi-Li Zhang
• 金额: $ 33.13万
• 依托单位: University of Minnesota-Twin Cities
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-10-01 至 2024-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1901103&HistoricalAwards=false
• 关键词: CNS; Core; Medium; Collaborative; Exploring

Traditionally, computer network protocols and control mechanisms are designed and engineered in accordance with certain theoretical models or design principles, under (often simplifying) assumptions about the network environment in which they operate. Network operations are mostly performed by operators through manual configurations of control parameters and resources, sometimes guided by measurement analysis and performance optimization. With the increasingly wide range of applications and complex network scenarios, traditional methods do not always perform well. To address this challenge, machine learning (ML) techniques have been applied to a wide range of networking and distributed systems problems, from reducing data center cooling costs to traffic optimization and application management. While preliminary results are promising, applying machine learning techniques to networking pose many important research questions that must be explored systematically and in depth. The proposed research constitutes an important first step toward providing a principled understanding of the fundamental limitations and promising new opportunities in learning-based network control from both theoretical and practical perspectives. It will help advance the emerging visions of self-driving networks and AIOps (Artificial Intelligence for IT Operations), and bring benefits to network operators, users, and the society at large. This project also integrates research with education and broadens participation in computing, especially with recruitment and training of female and under-represented students and outreach activities to K-12. Networks are a collection of control and (distributed) data plane elements that operate at different time scales on diverse types of data, respond and adapt to changes in traffic demands and the network state to achieve disparate objectives. The networking environments are highly dynamic and uncertain, with non-stationarity caused by surges and time-of-day changes in traffic demands, and unpredictable network failures; they are also inherently correlated, inter-dependent and constrained, in part due to complex interactions of various network entities. Moreover, networks are engineered systems -- there are basic principles that govern their designs and operations, with constraints that cannot be violated and inherent relations that could yield substantial performance gains. The proposed research focuses on learning-based network control problems to address these challenges along the following inter-related research thrusts. In Thrust 1, Network-Centric Learning Techniques, this project will explore the fundamental limits (from a theoretical perspective) and advance new network-centric ML techniques for non-stationary, correlated and constrained environments. In Thrust 2, Network-wide Learning-based Control and Horizontal/Vertical Interactions, this project will study and develop innovative learning-based network control algorithms in a network-wide framework by exploiting the (horizontal and vertical) interactions and leveraging shared learning. Last but not the least, in the Evaluation Thrust, this project will evaluate the proposed learning-based network control algorithms and compare them with conventional optimization and other ML based approaches.The project information such as publications, algorithms developed, data collected and personnel, will be made publicly available at https://web.cs.ucdavis.edu/~liu/Research/Holistic.htm during the entire project duration and for five years after the completion of this project.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.

传统上，计算机网络协议和控制机制是根据某些理论模型或设计原则，在（通常简化）关于它们操作的网络环境的假设下设计和工程化的。网络操作主要由操作员通过手动配置控制参数和资源来执行，有时由测量分析和性能优化来指导。随着应用范围的日益广泛和网络场景的日益复杂，传统的方法并不总是表现良好。为了应对这一挑战，机器学习（ML）技术已被应用于各种网络和分布式系统问题，从降低数据中心冷却成本到流量优化和应用程序管理。虽然初步结果令人鼓舞，但将机器学习技术应用于网络提出了许多重要的研究问题，必须系统而深入地探索。拟议的研究构成了重要的第一步，从理论和实践的角度提供了一个基本的限制和有前途的新的机会，在基于学习的网络控制的原则性理解。它将有助于推进自动驾驶网络和AIOps（用于IT运营的人工智能）的新兴愿景，并为网络运营商、用户和整个社会带来好处。该项目还将研究与教育结合起来，扩大了对计算的参与，特别是招聘和培训女学生和代表人数不足的学生，并开展了从幼儿园到12年级的外联活动。网络是控制和（分布式）数据平面元素的集合，这些元素在不同的时间尺度上对不同类型的数据进行操作，响应并适应流量需求和网络状态的变化，以实现不同的目标。网络环境是高度动态和不确定的，具有由流量需求的激增和时间变化以及不可预测的网络故障引起的非平稳性;它们也固有地相关，相互依赖和受约束，部分原因是各种网络实体的复杂交互。此外，网络是工程系统-有一些基本原则支配着它们的设计和运作，有一些不能违反的限制，有一些内在的关系可以产生很大的性能增益。建议的研究重点是基于学习的网络控制问题，以解决这些挑战沿着以下相互关联的研究推力。在Thrust 1，网络为中心的学习技术中，该项目将探索基本限制（从理论角度），并为非平稳，相关和受约束的环境推进新的网络为中心的ML技术。在Thrust 2，基于网络学习的控制和水平/垂直交互中，该项目将通过利用（水平和垂直）交互和利用共享学习，在网络范围内的框架中研究和开发创新的基于学习的网络控制算法。最后，在评估重点中，该项目将评估所提出的基于学习的网络控制算法，并将其与传统优化和其他基于ML的方法进行比较。项目信息，如出版物，开发的算法，收集的数据和人员，将https://web.cs.ucdavis.edu/~liu/Research/Holistic.htm在整个项目期间和项目完成后的五年内在www.example.com上公开。该奖项反映了NSF的法定使命，并已被认为是值得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的支持。

项目成果

Kaala: scalable, end-to-end, IoT system simulator

Kaala：可扩展、端到端的物联网系统模拟器

• DOI: 10.1145/3538393.3544937
• 发表时间: 2022
• 期刊: SIGCOMM workshop on networked sensing systems for a sustainable society
• 作者: Dayalan, Udhaya Kumar;Fezeu, Rostand A.;Salo, Timothy J.;Zhang, Zhi-Li
• 通讯作者: Zhang, Zhi-Li

Domain Disentangled Meta-Learning

• DOI: 10.1137/1.9781611977653.ch61
• 发表时间: 2023
• 作者: Xin Zhang;Yanhua Li;Ziming Zhang;Zhi-Li Zhang

Making content caching policies 'smart' using the deepcache framework

• DOI: 10.1145/3310165.3310174
• 发表时间: 2019-01
• 期刊: Comput. Commun. Rev.
• 作者: A. Narayanan;Saurabh Verma;Eman Ramadan;Pariya Babaie;Zhi-Li Zhang

Raven: belady-guided, predictive (deep) learning for in-memory and content caching

• DOI: 10.1145/3555050.3569134
• 发表时间: 2022-11
• 期刊: Proceedings of the 18th International Conference on emerging Networking EXperiments and Technologies
• 作者: Xinyue Hu;Eman Ramadan;Wei Ye;Feng Tian;Zhi-Li Zhang

5GNN: extrapolating 5G measurements through GNNs

• DOI: 10.1145/3565473.3569186
• 发表时间: 2022-12
• 期刊: Proceedings of the 1st International Workshop on Graph Neural Networking
• 作者: Wei Ye;Xinyue Hu;Tian Liu;Ruoyu Sun;Yanhua Li;Zhi-Li Zhang