CSR: Small: Collaborative Research: Bridging Reliability Analysis and Reality in Sensor Systems: Theories and Applications

从研究人员、开发人员甚至消费者的角度来看，可靠性分析被认为是传感器网络系统被广泛部署到关键任务应用之前不可或缺的一步。然而，传统的网络可靠性分析理论与现实的传感器网络系统环境之间存在着很大的差距。本研究的主要目的是通过新的度量，分析模型和方法来缩小这种差距，用于大规模传感器网络的可靠性分析。具体而言，本研究包括三个主要组成部分：1)描述传感器网络可靠性行为的概念、度量和模型；2)有效评估传感器网络可靠性的新算法和方法；3)展示所提出理论方法的实际意义的应用/案例研究。通过该项目开发的新的可靠性理论和方法是对网络系统可靠性知识体系以及可靠传感器系统设计的基础贡献。此外，将建立一个用于铁路监测的软件工具和原型传感器系统，以验证所提出的研究，并使理论研究具有可操作性和适用性。该项目通过研究生和本科教育，以及通过综合推广和小组指导计划，促进工程学科女生的招聘和保留，产生了更广泛的影响。

From the viewpoint of researchers, developers, and even consumers, reliability analysis is recognized as an indispensable step before sensor network systems can be widely deployed for mission-critical applications. However, there is a big gap between traditional network reliability analysis theories and realistic sensor network system environments. The primary objective of this research is to close such gap through new metrics, analytical models, and approaches for the reliability analysis of large-scale sensor networks. Specifically, the research has three major components: 1) concepts, metrics, and models for describing the reliability behavior of sensor networks, 2) new algorithms and methodologies for efficient reliability evaluation of sensor networks, and 3) applications/case studies for demonstrating the practical significance of the proposed theoretical approaches. The new reliability theories and methodologies developed through this project are fundamental contributions to the body of knowledge on the network system reliability as well as to the design of reliable sensor systems. Additionally, a software tool and a prototype sensor system for railway monitoring will be built to validate the proposed research, and render the theoretical research operational and applicable. This project has broader impacts through graduate and undergraduate education, and through an integrated outreach and group mentoring program that will promote recruitment and retention of female students in engineering disciplines.