CCSS: Collaborative Research: Learn-and-Adapt to Manage Dynamic Cyber-Physical Networks

该项目研究了针对云、交通和电力网络中出现的网络物理系统（CPS）机会的算法创新。设想的进展涉及管理CPS的学习和适应方法的模型和算法；可持续数据中心网络的在线数据驱动工作负载平衡智慧城市交通网络的实时交通管理；以及智能电网的学习辅助在线分散能源管理。拟议的研究具有巨大的社会、环境和经济效益潜力，因为它直接影响无线互联网、可扩展云网络、智能交通和支持智能家居和智能汽车的微电网。云用户的体验可以通过显著降低请求延迟而得到增强。由于有效的信号控制和车辆路线，智能汽车最终可以消耗更少的燃料，并且在十字路口等待的时间更短。最后，电网运营商可以从放宽的电池容量要求和变电站的快速电动汽车充电中受益。通过与中学生和社区接触，让本科生参与研究，增加代表性不足的少数民族的参与，以及促进科学素养的公众，将产生更广泛的变革性影响。

This project investigates algorithmic innovations targeting the cyber-physical system (CPS) opportunities emerging from cloud, transportation, and power networks. The envisioned advances pertain to models and algorithms for learn-and-adapt approaches to managing CPS; online data-driven workload balancing for sustainable data center networks; real-time traffic management for transportation networks in smart cities; and learning-aided online decentralized energy management for smart power networks.The proposed research has great potential for social, environmental, and economic benefits, as it directly impacts the wireless Internet, scalable cloud networks, intelligent transportation, and microgrids supporting smart homes and cars. Cloud users' experience can be enhanced with markedly low request delay. Smart cars can end up consuming less fuel, and experience less waiting time at road intersections due to efficient signal control and vehicle routing. Finally, power grid operators can benefit from relaxed battery capacity requirements and fast electric vehicle charging at substations. Broader transformative impact will be effected by reaching out to middle-school students and the community, involving undergraduates in research, increasing participation of under-represented minorities, and promoting a scientifically literate public.