Collaborative Research: IMR: MM-1A: Scalable Statistical Methodology for Performance Monitoring, Anomaly Identification, and Mapping Network Accessibility from Active Measurements

合作研究：IMR：MM-1A：用于性能监控、异常识别和主动测量映射网络可访问性的可扩展统计方法

• 批准号: 2319593
• 负责人: George Michailidis
• 金额: $ 20万
• 依托单位: University of California-Los Angeles
• 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: 2023-10-01 至 2026-09-30
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2319593&HistoricalAwards=false
• 关键词: Collaborative; Research; IMR; MM; 1A

This collaborative project aims to develop and implement new statistical methodology for the detection and mapping of various aspects of delays and bandwidth quality in computer networks. Part I of the project will develop the so-called extreme delay tomography, where network links experiencing extreme delays or congestion will be identified. The goal is to do so by using only end-to-end measurements from the perimeter of the network without observing all links directly. Part II of the project will focus on combining multiple data sets and measurements to build a state-wide and potentially nation-wide map(s) of broadband accessibility at the resolution of a US-census block. The project brings together researchers from the University of Michigan, the University of California, Los Angeles, and Merit Network, with expertise in network tomography, statistics of extremes, and network measurement research. One core goal is to utilize and develop new theory on statistics of extremes that can be used to detect and identify anomalous delays in the network from end-to-end delay measurements. This will expand the fields of statistics, network tomography, and provide novel tools for situational awareness and quality of service. The second core activity of the project will contribute new statistical methodology for integrating multiple observational data sets with existing demographic covariates for the purpose of mapping broadband access. This research activity will contribute to the fields of data integration and transfer learning.The broader impacts of the project will be manifold. The direct impact of the research will be on: (i) Developing novel practical tools for network situational awareness and security; (ii) Mapping the broadband availability that will aid in understanding and identifying priority goals for public policy and also in addressing social disparities due to the broadband gap; (iii) The project will engage graduate and undergraduate students and thus contribute to training of the next generation of network measurement researchers and statisticians. The outcomes of the project will have a broader impact on the fields of statistics and network measurement research through novel theoretical, methodological, and algorithmic developments, data products and software. The latter will be disseminated to the broader research communities through journal publications, conference presentations and open-source software.An account of the research goals, achievements, and landmarks will be made publicly available on https://www.merit.edu/initiatives/#activeresearchprojects. This address will maintain pointers to data products, research reports, presentations, and events related to the sponsored research.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.

该合作项目旨在开发和实施新的统计方法，用于检测和映射计算机网络中延迟和带宽质量的各个方面。 该项目的第一部分将开发所谓的极端延迟断层扫描，其中将识别经历极端延迟或拥塞的网络链路。目标是通过仅使用网络周边的端到端测量来实现此目的，而不直接观察所有链路。该项目的第二部分将侧重于结合多个数据集和测量结果，以美国人口普查区块的分辨率构建全州范围甚至全国范围的宽带接入地图。该项目汇集了来自密歇根大学、加州大学洛杉矶分校和 Merit Network 的研究人员，他们在网络断层扫描、极端统计和网络测量研究方面拥有专业知识。 一个核心目标是利用和开发关于极端统计的新理论，该理论可用于通过端到端延迟测量来检测和识别网络中的异常延迟。这将扩展统计、网络断层扫描领域，并为态势感知和服务质量提供新颖的工具。该项目的第二个核心活动将贡献新的统计方法，将多个观测数据集与现有人口协变量相整合，以绘制宽带接入图。这项研究活动将为数据集成和迁移学习领域做出贡献。该项目的更广泛影响将是多方面的。该研究的直接影响将在于：（i）开发用于网络态势感知和安全的新型实用工具； (ii) 绘制宽带可用性图，这将有助于理解和确定公共政策的优先目标，并有助于解决宽带差距造成的社会差距； (iii) 该项目将吸引研究生和本科生，从而有助于培训下一代网络测量研究人员和统计学家。该项目的成果将通过新颖的理论、方法和算法开发、数据产品和软件对统计和网络测量研究领域产生更广泛的影响。 后者将通过期刊出版物、会议演示和开源软件向更广泛的研究社区传播。研究目标、成就和里程碑的描述将在 https://www.merit.edu/initiatives 上公开发布/#activeresearchprojects.该地址将保留与赞助研究相关的数据产品、研究报告、演示文稿和活动的指针。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。