MRI: Acquisition of Instrumentation for Sensor, Cluster, and Network-Based Distributed Computing

MRI：采购用于传感器、集群和基于网络的分布式计算的仪器

• 批准号: 0521189
• 负责人: Raj Bhatnagar
• 金额: --
• 依托单位: University of Cincinnati Main Campus
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2005
• 资助国家: 美国
• 起止时间: 2005-09-15 至 2009-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0521189&HistoricalAwards=false
• 关键词: MRI; Acquisition; Instrumentation; Sensor; Cluster

This project, creating a distributed computing and decision making environment in which data gathered from wireless networks will be processed utilizing distributed algorithms implemented on AES (Advanced Execution System), addresses the following issues: How to efficiently-Gather massive but distributed data using a large number of sensors equipped with small wireless radios,-Aggregate data from a group of sensors to AES for real time processing, and-Make decisions based on distributed algorithms specifically designed for this purpose.AES envisions two components, both based on utilizing the concept of volunteer computational resources: An environment with-A component that allows loosely coupled scientific and decision making applications to be run using stand-alone volunteer computers (Galaxy AES) such as those participating in the PlanetLab project and-Multiple Beowulf clusters connected for both loosely coupled and more tightly coupled computations.Wireless sensor networks offer unprecedented opportunities for a broad spectrum of applications, including monitoring the environment in order to take corrective action if a hazardous situation arises. Tiny sensor devices sense environmental parameters and transmit data over a wireless medium. This work develops new algorithms/methods for mapping parallel applications onto highly dynamic volunteer resources that contribute in making decisions in real time based on sensor data gathered. Moreover, this instrumentation enables the construction of two sensor networks to continuously monitor outdoor and indoor air quality. The sensor networks are expected to integrate local computations and inter-sensor communication to perform complex global computations, inferencing, and decision making using the distributed algorithms developed for this purpose. The project, expected to provide a testbed capable of handling thousands of volunteer nodes, enables members of the PlanetLab community to use the Galaxy AES testbed for research in areas of P2P computing and volunteer computing. Benefits on the wireless sensor network side include management and control of air pollutants, informed formulation of public policy related to the environment, and educational training.

该项目创建了一个分布式计算和决策环境，在该环境中，从无线网络收集的数据将利用在AES（高级执行系统）上实现的分布式算法进行处理，解决以下问题：如何高效地-使用配备小型无线无线电的大量传感器收集大量分布式数据，-将来自一组传感器的数据聚合到AES进行实时处理，并基于为此目的专门设计的分布式算法做出决策。AES设想了两个组件，它们都基于利用志愿计算资源的概念：一个包含组件的环境，该组件允许使用独立的志愿计算机（Galaxy AES）（如参与PlanetLab项目的计算机）运行松散耦合的科学和决策应用程序；多个Beowulf集群连接在一起，用于松散耦合和更紧密耦合的计算。无线传感器网络为广泛的应用提供了前所未有的机会，包括监测环境，以便在出现危险情况时采取纠正措施。微型传感器设备感知环境参数并通过无线介质传输数据。这项工作开发了新的算法/方法，将并行应用程序映射到高度动态的志愿者资源上，这些资源有助于根据收集到的传感器数据实时做出决策。此外，该仪器使两个传感器网络的建设能够持续监测室外和室内空气质量。传感器网络有望集成本地计算和传感器间通信，以使用为此目的开发的分布式算法执行复杂的全局计算、推理和决策。该项目预计将提供一个能够处理数千个志愿者节点的测试平台，使PlanetLab社区的成员能够使用Galaxy AES测试平台进行P2P计算和志愿者计算领域的研究。无线传感器网络方面的好处包括管理和控制空气污染物，知情制定与环境有关的公共政策，以及教育培训。