CNS Core: Small:On Parallelizing Optical Network Design Problems:Towards Network Optimization as a Service

规划、部署和设计构成Internet基础设施的网络涉及复杂的问题，通常称为“网络设计”问题。网络设计问题的有效和高效解决方案对于互联网的运营和经济及其支持关键和可靠的教育、政府、商业和与科学相关的通信服务的能力至关重要。随着网络规模的扩大，网络设计变得越来越困难。本研究项目旨在通过研究以可扩展的方式解决大型网络的网络设计问题的新方法来应对这一挑战。该项目旨在为网络设计问题带来高效、可扩展的解决方案的范式转变。项目办法将是利用所有人的共同特点，即优化过程既包括路线选择方面，又包括资源分配方面，开发可适用于各种问题的并行解决办法。研究计划是将布线维度上的解空间划分为一组相互排斥的集体穷举的子空间，这些子空间可以作为独立子问题并行探索。具体地，每个子空间对应于每个节点对的固定路由路径，并且相应的子问题仅涉及沿着给定路径搜索最优资源分配。因此，子问题是相互独立的，可以通过利用可用的集群/云计算资源来加速整体执行时间来并行解决。该方法将着眼于智能方法，以修剪实际上需要探索的一组子空间。这个项目的算法结果将被集成到一个开放源码工具中，供学生、研究人员和从业者使用。这项研究将为未来的网络优化即服务(NOAAS)模型奠定基础，该模型将远程提供网络优化，并以云服务的形式按需交付。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Planning, deploying, and engineering the networks that make up the Internet infrastructure involve complex problems generically referred to as "network design" problems. Effective and efficient solutions to network design problems are crucial to the operation and economics of the Internet and its ability to support critical and reliable education, government, commercial and science-related communication services. Network design gets increasingly difficult as networks grow larger. This research project aims to address this challenge by investigating new approaches for tackling network design problems for large networks in a scalable manner. This project aims to make contributions that will lead to a paradigm shift in efficient, scalable solutions to network design problems. The project approach will be to develop parallel solutions that are applicable to a wide range of problems by exploiting a feature common to all, namely, that the optimization process incorporates both a routing aspect and a resource allocation aspect. The research plan is to partition the solution space across the routing dimension into a set of mutually exclusive and collectively exhaustive subspaces, which may be explored in parallel as independent subproblems. Specifically, each subspace corresponds to a fixed routing path for each node pair, and the corresponding subproblem involves searching only for an optimal resource assignment along the given paths. Consequently, subproblems are independent of each other and may be solved in parallel by leveraging available cluster/cloud computing resources to speed up the overall execution time. The approach will look at intelligent ways to prune the set of subspaces that practically need to be explored. The algorithmic results of this project will be integrated into an open-source tool that will be made available to students, researchers, and practitioners. The research will lay the foundation for a future "Network Optimization as a Service (NOaaS)" model for network optimization to be delivered remotely and on demand as a cloud service.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.