NeTS: Small: GATE: Greening At The Edges

该项目通过容量自适应网络设计、节能资源分配和流量调度等一系列方法来降低光纤和无线接入网络的能耗。当网络上有大量需要带宽的应用程序(如视频流)时，容量自适应接入网络保持在“高容量高功率”模式，而当网络上负载较少的应用程序(如语音和未来智能电网流量)时，容量自适应接入网络切换到“低容量低功率”模式。该项目将数据中心目前使用的方法推广到通信链路，并专注于将问题形式化并获得最优或接近最优的解决方案。容量自适应接入网的设计理念对接入网设计提出了容量自适应的新要求，而不是简单地为峰值利用率提供保障。该项目的能量感知网络控制、资源分配和流量调度方案除了考虑网络服务质量之外，还考虑额外的关键性能指标--能耗，而传统方案仅考虑延迟和损失等网络性能。通过理论分析、仿真和试验台实验，研究结果不仅将证明接入网络的能效提高，而且还将潜在地为核心网、数据中心网络和其他无线系统量身定做。广泛的影响：降低接入设备的能耗不仅在正常运行期间对环境产生积极影响，而且在紧急情况下将提供公共利益。远程访问设备目前由公共电力、备用电池和不同程度的本地发电相结合供电。该项目正在开发的技术，在紧急后备时延长运行时间，有助于保护公众健康和安全。该项目将通过本科生和研究生参与该项目，并将研究成果纳入本科生和研究生课程工作，将研究与教育结合起来。PI将利用新泽西州教育机会计划(EOP)，该计划为代表不足的人群提供教育机会，作为招收学生进入研究计划的过程的一部分。

This project addresses the reduction of energy consumption in optical and wireless access networks via a series of approaches such as capacity-adaptive network design, and energy-efficient resource allocation and traffic scheduling. The capacity-adaptive access network stays in the 'high-capacity high-power' mode when the network is heavily loaded with bandwidth-hungry applications such as video streaming, and switches into the 'low-capacity low-power' mode when the network is lightly loaded with less bandwidth demanding applications such as voice and future smart grid traffic. The project generalizes the approaches currently being used in data centers to communication links, and focuses on formalizing the problem and obtaining optimal or near-optimal solutions. It will not only improve energy efficiency of current access networks, but will also provide insights and guidelines on how to upgrade the capacity of access networks efficiently.The concept of designing capacity-adaptive access networks introduces new requirements of 'capacity adaptive' on the design of access networks as opposed to simply provisioning for peak utilization. This project's energy-aware network control, resource allocation, and traffic scheduling schemes consider an additional key performance metric, energy consumption, in addition to network quality of service, while conventional schemes merely consider network performance such as delay and loss. Research results, via theoretical analysis, simulations, and testbed experiments, will not only demonstrate improved energy efficiency of access networks, but will also be potentially tailored for core networks, datacenter networks, and other wireless systems.Broader Impact:Reducing the energy consumption of access equipment will not only have positive environment impacts during normal operation but it will provide public benefit during emergencies. Remote access equipment is currently powered by a combination of public power, battery backup, and varying degrees of local generation. Techniques such as those being developed by this project that extend operation when on emergency backup help protect the public health and safety. The project will integrate research and education through participation of both undergraduate and graduate students in the project and by incorporation of research outcomes into undergraduate and graduate course work. The PI will utilize the New Jersey Educational Opportunity Program (EOP), which provides educational opportunities to under-represented populations, as part of the process for recruiting students into the research program.