CAREER: Self-Organization Services and Application Support Constructs for Sensor Networks

自主自配置是部署可扩展无线传感器网络的重要组成部分。尽管网络服务取得了重大进展，但即使是中等规模的试验床的经验也表明，每个节点的配置是大规模部署的重大障碍。为了解决这个问题，本研究将考虑分布式传感器网络中自配置的三个主要组成部分：位置校准、分组和移动性。重点是从工程和计算机科学的多个学科中仔细设计分布式自组织算法，以及它们与无线通信、处理和传感方面的最新技术的集成。结果将被应用于一组应用程序构造的开发，这些应用程序构造将作为理解移动模式和使用无线微型传感器协调行为的“字母表”。建议的应用程序构造将通过设计和实现一个游戏应用程序来实现，该应用程序将解释其用户的行为并生成旨在改变玩家行为的响应。这项工作的影响将来自于在游戏背景下严格练习开发的概念，以及在日常生活环境中实际部署游戏变体，在日常生活环境中，智能传感器将与其用户无缝交互。这项工作的结果将在会议上公布，在传感器网络的新课程中讲授，并将发布在网络上，供有兴趣采用它的其他机构使用。在该项目下开发的硬件和软件将提供给社区使用，以便它们可以应用于其他传感器网络应用。

Autonomous self-configuration is an essential component for the deployment of scalable wireless sensor networks. Despite the significant progress of network services, experience with even modest sized testbeds has shown that the per-node configuration is a significant barrier for large scale deployments. To address this issue this research will consider three main components of self-configuration in distributed sensor networks: location calibration, grouping and mobility. The emphasis will be on the careful design of distributed self-organization algorithms drawing from multiple disciplines in engineering and computer science and their integration with the state-of-the-art technologies in wireless communication, processing and sensing. The results will be applied to the development of a set of application constructs that will serve as an "alphabet" for understanding mobility patterns and orchestrating behaviors using wireless microsensors. The proposed application constructs will be pursued through the design and implementation of a game application that will interpret the behavior of its users and generate responses aiming to change the player's behavior. The impact of this work will arise from the rigorous exercise of the developed concepts in the context of the game and the actual deployment of the game variants in everyday life environments, where intelligent sensors will seamlessly interact with their users. The results of this work will be published in conferences, taught in a new course on sensor networks and will be posted on the web for other institutions interested in adopting it. The hardware and software developed under this project will be made available to the community so that they can be applied in other sensor network applications.