Gathering, processing and sharing data in a sensor-internet

Sensor networks have been proposed and deployed for study of scientific phenomena with levels of detail that was previously impossible. Recently urban sensing, participating sensing emerge as an effort to sensors in our daily life. To date, these deployments are isolated patches, each with different mechanisms to deliver data to their users, and often have no method to share data with others. We propose a Sensor-Internet architecture to interconnect isolated sensornets focusing on reprocessing existing data and sharing them conveniently among different organizations. ❧ The thesis of this dissertation is that simple design and easy-to-use of gathering, processing and sharing sensor data enables people to observe current status of physical phenomena and add value to the shared data. We substantiate this thesis through the following three approaches. First, we provide easy-to-use abstractions to enable a sensor data sharing system, so that many people can participate sharing sensor data. Second, we explore an efficient design of sensor data sharing system. Third, we present prototype systems and practical sensor applications to show our easy-to-use and efficiency in gathering, processing, and sharing sensor data.

传感器网络已被提出并部署，用于以此前无法实现的精细程度开展科学现象研究。近年来，城市感知（urban sensing）、参与式感知（participating sensing）作为融入日常生活的传感应用模式应运而生。截至目前，此类传感网络部署均为孤立的零散系统，各自采用不同机制向其用户传输数据，且通常不具备与其他系统共享数据的能力。为此，我们提出传感器互联网（Sensor-Internet）架构，用于互连各类孤立的传感网络，核心聚焦于对现有数据进行再处理，并实现不同组织间的便捷数据共享。
本学位论文的核心论点为：通过简洁的设计与便捷易用的传感器数据采集、处理及共享机制，可使人们实时观测物理现象的当前状态，并为共享数据赋予额外价值。我们通过以下三种路径验证这一论点：其一，构建具备易用抽象接口的传感器数据共享系统，以支持更多用户参与传感器数据的共享；其二，开展传感器数据共享系统的高效化设计研究；其三，搭建原型系统并开发实际传感应用案例，以此验证我们在传感器数据采集、处理与共享环节的易用性与高效性。