MRI: CRON: Development of a Cyberinfrastructure Reconfigurable Optical Network for Large-Scale Scientific Discovery

在本提案中，我们建议为可重构光网络（CRON）开发一个网络基础设施。随着超级计算设施和存储资源分布在美国和全球各地，科学应用和合作的成功越来越依赖于高速光网络。国家和国际光网络和区域网络的发展和部署使更多的科学家、研究和教育机构能够在以前无法实现的水平上进行联系和合作。然而，研究项目与已部署的高速网络之间存在着严重的差距，主要原因在于：(1)物理资源有限；例如，LONI网络的昂贵资源限制了同时用户的数量，并且路易斯安那州有许多研究和/或教育机构由于最后一英里问题而无法访问网络。(2)不可修改的环境——由于目前部署的网络既用于生产也用于研究，个人用户不能通过修改环境参数（如网络协议栈和操作系统）来强加给其他用户；(3)局部环境？为一个特定的高速网络开发的应用程序或协议应该在各种不同的网络上正确工作，而不考虑不同的网络带宽和延迟。但是，每个已部署的物理网络都有一个物理特性，例如带宽和延迟。该项目的目标是通过开发虚拟网络和计算网络基础设施（CRON）来弥合物理网络和研究之间的差距，该基础设施可提供对各种网络环境的集成和自动化访问。为了提供虚拟环境，CRON由两个组件组成：硬件（包括交换机、光纤、网络模拟器）和软件（根据用户的需求自动重新配置硬件）。的要求。CRON通过实现三个关键目标来解决关键问题：(i)无限虚拟网络资源CRON将使研究人员能够探索新的网络技术，并在不受物理限制的情况下快速评估其对应用程序的影响，并允许教育工作者向无法访问物理网络的学生介绍最先进的网络环境。（ii）可重构性CRON将允许研究人员和教育工作者修改虚拟环境，而不会干扰共享CRON的其他人。（iii）完整的环境CRON将提供各种网络环境，包括区域网络和全球跨洋光网络。

In this proposal, we propose to develop a Cyberinfrastructure for Reconfigurable Optical Networks (CRON). With supercomputing facilities and storage resources geographically dispersed across the US and around the globe, the success of scientific applications and collaborations relies increasingly on high-speed optical networks. The development and deployment of national and international optical networks and regional networks make it possible for more scientists, research and educational institutes to connect and collaborate at previously unachievable levels.However, there is a critical gap between the research projects and the deployed high speed networks, caused by the following problems: (1) Limited physical resource ? e.g., the expensive resource of the LONI network has limitations on the number of simultaneous users, and there are many research and/or educational institutions in Louisiana that cannot access the networks due to the last mile problem.; (2) Unmodifiable environment - because the currently deployed networks are used for production as well as research, individual users cannot impose on other users by modifying environmental parameters, such as network protocol stacks and operating systems; and (3) Partial environment ? applications or protocols developed for one specific high speed network should work correctly over all kinds of different networks, irrespective of different network bandwidths and delays. However, each deployed physical network has one physical characteristic, such as bandwidth and delay.The goal of this project is to bridge the gap between physical networks and research by developing a virtual networking and computing cyberinfrastructure, CRON, that provides integrated and automated access to diverse networking environments. To provide the virtual environments, CRON consists of two components: hardwares including a switch, optical fibers, network emulators and softwares automatically reconfiguring the hardwares based on users? demands. CRON resolves the critical issues by achieving three key objectives: (i) Unlimited virtual networking resource CRON will enable researchers to explore new network technologies and rapidly assess their impact on applications irrespective of physical limitation and allow educators to introduce the state-of-art networking environment to students who cannot access those physical networks. (ii) Reconfigurability CRON will allow researchers and educators to modify the virtual environments without interfering with others who share CRON. (iii) Complete environment CRON will provide all kinds of networking environments, including regional networks and global transoceanic optical networks.