II-NEW: Advancing Access and Home Networking Research in NEULab

有了这笔仪器拨款，首席研究人员将在普林斯顿大学网络最终用户实验室(NEULab)建立一个现实的接入和家庭网络环境，包括一个DSL接入系统、一个电缆调制解调器接入系统和一个具有三个屏幕的家庭网络：带机顶盒的高清电视、计算机和各种移动设备。这个测试床将通过GigE链路和一个可配置的路由器连接到国家主干测试床Vini。PI将在理论基础上进行各种实验，并反过来塑造理论的发展，以支持受限最后一跳和异类家庭网络上的富媒体三网融合应用的网络结构、协议和算法。通过这项建议，NEULab的几个关键部分将得到建立，特别是在DSL接入网、电缆调制解调器接入网和三屏家庭网络上。关键设备包括DSLAM、CMTS、Spirent信道模拟器、铜场、家庭网关、两级交换机和具有WLAN功能的消费电子产品。智能价值：将验证或证伪以下领域开发的现有数学理论和算法，以及受拟议试验床启发的未来数学理论和算法：宽带接入光纤/DSL网络、P2P内容分发、内容感知网络、分布式调度算法、联合功率控制和调度、随机网络效用最大化和绿色信息技术。还将讨论涉及难以分析表征的指标的开放问题，例如延迟、抖动和收敛速度。在这7个研究领域下，列出了9个可以在试验台上回答的具有挑战性的问题。更广泛的影响：这项建议代表了一种建立在皮？S行业合作和技术转让之上的战略，通过普林斯顿大学NEULab保护伞下的一套实验平台，加速将理论转化为实践。PI还将通过本科生的独立工作和本科生论文，积极参与实验设施的建设和运行，并每年向普林斯顿以外的10名本科生开放暑期本科课程。

With this instrumentation grant, the principal investigator will build a realistic access and home networking environment in Princeton's Networked End User Lab (NEULab) including a DSL access system, a cable modem access system, and a home network with triple screens: HDTV with set-top box, computer, and a variety of mobile devices. This testbed will be connected to national backbone testbed VINI via GigE links and a configurable router. The PI will run a variety of experiments, based on theoretical foundation and in turn shaping the development of theory, on network architectures, protocols, algorithms supporting media-rich triple-play applications over constrained last hop and heterogeneous home networks. Through this proposal, several key parts of NEULab will be established, especially on DSL access network, cable modem access network, and triple-screen home network. The key equipments include DSLAM, CMTS, Spirent channel emulators, copper-farm, home gateways, two stages of switches, and consumer electronics with WLAN capabilities.Intellectual Merit: Existing mathematical theories and algorithms developed in the following areas will be validated or falsified, and future ones inspired by the proposed testbeds: Broadband access fiber/DSL networks, P2P content distribution, content-aware networking, distributed scheduling algorithms, joint power control and scheduling, stochastic network utility maximization, and green information technology. Open issues involving metrics that are difficult to be analytically characterized, such as delay, jitter, and rate of convergence, will also be addressed. A list of 9 challenging questions that can be answered in the testbed is provided under these 7 research areas. Broader Impacts: This proposal represents a strategy built on top of the PI?s industry collabora- tions and technology transfers to accelerate the transfer of theory into practice, through a set of experimental platforms under the umbrella of NEULab within Princeton University. The PI will also actively involve undergrad students in the construction and running of the experimental facilities through their independent work and undergrad theses, and hold summer undergraduate program open to 10 undergrads outside Princeton every year.