基于动态分簇的矿井WSNs目标跟踪定位技术研究

Research on WSNs target tracking techniques in coal mine based on dynamic clustering

无线传感器网络(WSNs)定位技术是传感器网络的重要支撑技术。现有的目标定位技术如RS-SI、三边定位以及加权质心算法等,较多考虑的是平面网络或是如何提高定位精度问题,带来了较大的能耗。充分考虑了WSNs能耗有限的特点,提出了一种基于动态分簇的目标定位技术。该技术算法给出了基于簇结构的网络形成的算法思想,簇首以及簇成员节点的选取原则。基于动态簇的传感器网络的定位技术分为网络初始化、动态簇跟踪目标和簇的重组3个阶段。通过实验仿真,证明了簇组建成功概率与权值关系,以及不同运动场景和节点布置密度情况下与标跟踪的精度关系。该算法降低了网络能量开销,提高了定位精度,适合矿井传感器网络应用。

Localization is an important support technology of Wireless Sensor Networks. About the existing protocols such as RSSI, three border localization and weighted centric algorithm, more consideration are taken to the precision and planar network, which bring larger energy consumption. A low overhead target localization protocol of WSNs based on dynamic clustering is implemented, fully taking characteristics of WSNs such as network topology and limited power into account. The algorithm of target tracking based on dynamic clusters in coal mine is divided into three parts, that is network initialization, target tracking of dynamic clusters and cluster regrouping. Both the performance analysis and experimental results show that cluster formation of a successful relationship between probability and weight and the relation of target tracking precision in different scene. This algorithm reduces synchronization energy consumption and improves the precision, so it can be used for WSNs in coal mine.