基于单元格的无线传感器网络节点定位算法

Reliable and Precise Distributed Algorithms for Node Localization in Wireless Sensor Networks Based on Cells

在一种由基本单元格构成的离散化无线传感器网络模型中,针对典型的Bounding Box定位方法存在的问题,利用所有位置已知的锚节点对于未知节点位置形成的约束,提出了2种分布式的节点定位算法LIE和LIE-R,并对新算法的性能进行了详细的评估。理论分析及仿真实验结果均表明,文中所提出的算法具有良好的定位性能。

In order to improve the localization accuracy, coverage and stability of Bounding Box method, we propose two new methods based on discrete communication model. In the full paper, we explain our new methods in detail. In this abstract, we just give only a briefing and add some pertinent remarks to listing the two topics of explanation. The first topic is： the discrete network model. The network is subdivided into many congruent square cells, and all the positions and distances are expressed as the number of cells, the communication range of each node is defined as a rectangle centered at the node. The second topic is： intersection and exclusion. Sensor nodes use geometric constraints induced by both neighboring and non-neighboring anchors to decrease the uncertainty of their position estimate. A sensor node can position itself anywhere within the intersection of the communication ranges of its neighboring anchors, also within the exclusion of those other anchors that are out of range of it. Our new method＇s important feature is providing detailed analytic density of known anchor nodes and complexity expressions for the basic probabilistic error, the minimal estimates of the new method. Simulation results in more realistic scenarios show preliminarily that LIE（Localization with Intersection and Exclusion） and LIE-R （Localization with Intersection and Exclusion-Reduction） can significantly achieve higher accuracy with less anchor nodes than Bounding Box method. Furthermore, LIE achieves the same expected accuracy as LIE-R with almost only 1/2 of the anchors.