无线传感器网络时空信息集成框架研究

Providing a Spatio-Temporal Information Integration Framework for Wireless Sensor Networks

针对协作信号信息处理是构建无线传感器网络的关键技术之一。网络中有效信息分布时空变化的特点,提出了一种基于移动代理的时空信息集成框架。通过定义传感节点的目标检测信令和观测节点的时间延迟窗“节拍”,得到信息集成的时空坐标系;移动代理携带集成算法执行基于系统效能函数优化的访问路径,完成时域和空域的信息集成。节拍是空域信息集成的时间点,节拍序列构成时域信息集成的时间轴。仿真实验结果证明了框架的有效性。

CSIP (Collaborative Signal and Information Processing) is a key technology in constructing wireless sensor networks. Qi and Xu's paper[6] on CSIP did not provide spatio-temporal information integration framework, which is, in our opinion, a serious deficiency that we aim to remedy by providing such a framework. In the context of a wireless sensor network, scattered sensing nodes provide only locally spatial information that is time-varying and feasible over limited time periods. Furthermore, the data of nodes, which contribute to specific goals, are usually varying in the space and time domains. Like Ref.6, our CSIP is based on mobile agent. Unlike Ref.6, in order to provide a spatio-temporal information integration framework, we introduce the concept of rhythm; one rhythm is defined as a period of time triggered by an object detection signal (ODS) message, in which ODSs from nodes are collected and analyzed to program the visiting itineraries by the rule of priority of frequency and that of First In Last Out (FILO) for mobile agents. A rhythm acts as a label for spatial processing, and thus the series of rhythms become a time axis for temporal integration. Our integration framework has the following desirable features: energy efficiency, scalability, fault tolerance and progressive accuracy.