Stochastic Diffusion, Adaptive Estimation, and Prediction Models for Wireless Networked Systems

本研究的目的是开发真实的时间的无线网络的建模，估计和预测的方法。该建模方法可推广到包括“城市峡谷”在内的各种传播环境，并直接解决发射机数量引起的随机和不断变化的干扰效应和噪声源等问题。 该研究将产生通用方法，可提高通信信道利用率，从而提高各种无线网络系统应用程序的性能。可扩展性包括基本算法、建模和分析工具、通过仿真和硬件实现的演示和验证、研究结果的文档记录以及工程专业学生教育。如果成功，这项研究的特别好处将产生于网络控制系统，如移动的传感器网络、无人驾驶车辆、远程手术、触觉协作、健康监测、农业地区/农场、家庭网络、以及无线网络多媒体应用。 例如，高度可靠的电力的经济有效的输送越来越依赖于联网的SCADA（监督控制和数据采集系统）和分布式控制系统，这些系统通过商业上可获得的并且经常是基于互联网协议（IP）的通信技术来操作。研究结果将通过不同的平台进行传播，从而能够开发出性能和可靠性更高、布线成本更低的商用软件和硬件（收发器）。 教育计划特别强调培养多样化的学生群体，并通过利用学生和导师团队，吸收研究生和本科生，并注重面向项目的综合系统设计，将代表性不足的少数群体纳入其中。与橡树岭国家实验室的伙伴关系将提供现实的数据集和应用，并有机会从从业人员直接反馈给学生。

The objective of this research is to develop methods for modeling, estimation, and prediction of wireless networks in real time. The modeling approach generalizes to diverse propagation environments including "urban canyons", and directly addresses issues such as the random and ever changing interference effects and noise sources caused by the transmitter population. The research will result in general methods that will improve communication channel utilization and thereby performance of various wireless networked system applications. Deliverable includes fundamental algorithms, modeling and analysis tools, demonstration and validation via simulation and hardware implementation, documentation of research results, and engineering student education.If successful, particular benefits of this research will accrue to networked control systems, such as mobile sensor networks, unmanned vehicles, remote surgery, haptics collaboration, health monitoring, agricultural areas/farms, home networking, and wireless networks multimedia applications. For example, the economically efficient delivery of highly reliable electric power is increasingly dependent on networked SCADA (Supervisory Control and Data Acquisition System) and distributed control systems operating over commercially available, and frequently internet protocol (IP) based communication technologies. The results will be disseminated using different platforms to allow the creation of commercial software and hardware (transceivers) with increased performance, reliability, and reduced wiring costs. The educational plan places a particular emphasis upon the cultivation of a diverse student population and inclusion of under-represented minorities, by utilizing student and mentor teams, incorporating graduate and undergraduate students, and focusing upon project-oriented integrated systems design. A partnership with Oak Ridge National Laboratory will provide realistic data sets and applications, and opportunities for direct feedback from practitioners to the students.