IRNC: AMI: Collaborative Research: Software-Defined and Privacy-Preserving Network Measurement Instrument and Services for Understanding Data-Driven Science Discovery

全球范围内的数据密集型科学发现对国际研究和教育网络的速度和管理提出了新的要求。在这些国际网络的连接点上，测量网络数据流以了解网络流量模式、识别网络异常并为网络控制和规划提供见解至关重要。然而，随着网络速度的不断提高、网络流量的庞大以及测量目标的不断变化，在高速网络上进行流量级测量具有极大的挑战性。先进测量仪器和服务（AMIS）项目利用多核、可编程网络处理器来原型化和部署先进的测量仪器，以实现准确的网络监控和深入的流量分析服务。该仪器支持线速率高达100Gbps的流量粒度测量和软件应用程序编程接口，以检查选定的流量，对用户流量的性能没有影响。通过可扩展的硬件和开源软件堆栈，测量服务为网络运营商提供了有效的工具来量化流量级网络性能，并通过保护隐私的计算分析来研究网络流量。该项目建立在学术界、工业合作伙伴、网络运营商和国际联盟的联盟基础上，他们带来了独特的专业知识和资源，以实现高性能、可编程流粒度网络测量的目标。该项目的成果将大大有利于数据驱动的科学发现，如天文学和空间天气研究，并将通过包括EPSCoR大学和西班牙裔服务机构在内的多所大学的参与，促进代表性不足的群体（如西班牙裔和女学生）的广泛参与。

Data intensive science discovery at a global scale has imposed new requirements on the speed and management of international research and education networks. At the connection points of these international networks, it is critical to measure the network data flows to understand network traffic patterns, identify network anomalies and provide insights to network control and planning. However, the ever-increasing network speed, the massive amount of network flows and the changing measurement objectives have made the flow-level measurement on very high-speed networks extremely challenging. The Advanced Measurement Instrument and Services (AMIS) project leverages many-core, programmable network processors to prototype and deploy an advanced measurement instrument to enable services for accurate network monitoring and in-depth traffic analysis. The instrument supports flow-granularity measurement at line rate up to 100Gbps and software application programming interfaces to examine selected flows, with no impact to the performance of user traffic. With scalable hardware and an open source software stack, the measurement services equip network operators with effective tools to quantify flow-level network performance and study network flows through privacy-preserving computational analytics. This project is built on a consortium of academia, industrial partners, network operators and international alliances, who bring unique expertise and resources to achieve the objectives of high performance, programmable flow-granularity network measurement. The outcomes from this project will significantly benefit data driven science discovery, such as astronomy and space weather studies, and will promote broadened participation of underrepresented groups (such as Hispanic and female students) through the involvement of multiple universities, including an EPSCoR university and a Hispanic Serving Institution.