NeTS-NOSS: SENSIMIDE: Integrated Software Development and Multi-Mode Simulation for Large-Scale Sensor Networks

提案编号：0627183PI：Rich Wolski，Co-Pi：Chandra Krintz Institution：加州大学圣巴巴拉分校标题：Nets-NOSS：SENSIMIDE：大规模传感器网络的集成软件开发和多模式模拟摘要尽管传感器网络似乎有望带来变革性的科学甚至社会变革，但目前它们的发展仍处于萌芽阶段。一个关键的障碍是缺乏开发工具来支持传感器网络系统的连贯设计、高效工程和有效部署。该项目培育了一个新的软件开发平台，该平台结合了多模式仿真、功率和电池建模、无线电行为的统计表征以及高级编译和程序准备策略，以提供统一的编程和调试基础设施。SENSIMIDE(Sensornet Simulation IDE)将对异类设备的全系统、周期关闭的仿真与来自活动物理设备的实时测量相结合，以提高可扩展集成仿真的逼真度。基于集群的高性能仿真环境无需修改即可实时执行SensorNet操作系统和应用程序代码，使用户能够透明地在物理设备上执行和在仿真中移动。此外，SENSIMIDE在多个抽象级别上检测和启动仿真，从而为用户提供重要且新颖的调试和性能检查能力。SENSIMIDE将这种能力与新的编译策略和调试支持相结合，为用户提供了两个新的灵活性维度。首先，用户可以对仿真系统中同时执行的传感器集合进行编程、测量和控制。其次，仿真系统接受外部开发的电源、电池和无线电通信模型，作为在这些重要研究领域促进新系统级研究的一种方式。通过为传感器网络软件系统的开发、调试和全系统分析提供全面的基础设施，SENSIMIDE降低了传感器网络开发和部署的门槛，同时促进了对关键系统级问题的新研究。PIS项目将通过出版物和原型软件向更广泛的传感器网络研究和教育界传播SENSIMIDE。

Proposal Number: 0627183PI: Rich Wolski, Co-PI: Chandra Krintz Institution: University of California, Santa BarbaraTitle: NeTS-NOSS: SENSIMIDE: Integrated Software Development and Multi-Mode Simulation for Large-Scale Sensor NetworksAbstractDespite the potential for transformative scientific and even social change that sensor networks seem to promise, their development is, at present, still nascent. A key impediment is the lack of development tools that enable the coherent design, efficient engineering, and effective deployment of sensor network systems.This project fosters a new software development platform that combines multi-mode simulation, power and battery modeling, statistical characterization of radio behavior, and advanced compilation and program preparation strategies to provide a unified programming and debugging infrastructure. SENSIMIDE (Sensornet Simulation IDE) couples full-system, cycle-close, emulation of heterogenous devices with real-time measurements from active physical devices, to improve the fidelity of scalable ensemble simulations. A high-performance, cluster-based emulation environment executes sensornet operating systems and application code in real-time, without modification, making it possible for the user to move between execution on physical devices and in emulation transparently. In addition, SENSIMIDE instruments and actuates emulations at multiple levels of abstraction thereby affording users vital and novel debugging and performance inspection capabilities.SENSIMIDE interfaces this capability with new compilation strategies and debugging support that provide the user with two new dimensions of flexibility. First users can program, instrument, and control ensembles of simultaneously executing sensors within the emulation system. Secondly, the emulation system accepts externally developed power, battery, and radio communication models as a way of promoting new systems level research in these important areas of study.By delivering a comprehensive infrastructure for the development, debugging, and full-system analysis of sensornet software systems, SENSIMIDE lowers the barrier-to-entry for sensor network development and deployment while promoting new research into critical systems-level problems. The project PIs will disseminate SENSIMIDE to the wider sensor network research and educational communities via publication and prototype software.