大规模无线传感器网络定位误差分析及优化方法研究

受成本因素的限制，大规模无线传感器网络通常由大量廉价节点和少量位置已知的锚点组成，并采用基于RSS（Received Signal Strength）测距方法，严重制约了定位算法的性能，已成为该研究领域亟待解决的关键问题。本项目将通过分析大规模无线传感器网络的定位机制，建立多跳定位误差的标度率及误差传播公式，从而揭示多跳定位误差的形成机理和传播规律；探索锚点位置对定位精度的影响规律，揭示锚点最优部署位置与节点密度的依赖关系，进而研究基于随机网络模型的锚点部署优化算法；深入分析分布式定位涉及的RSS测距、单个节点定位和洪泛广播三个关键环节，采用信息融合和图论等方法研究面向大规模无线传感器网络的多层次定位性能优化方法，形成具有节能和误差传播抑制功能的分布式定位算法。本项目的研究成果有利于完善多跳定位的误差理论，并为开发和部署大规模无线传感器网络提供理论指导和实用方法。

Due to cost constraints, a large-scale wireless sensor network (WSN) normally consists of a mass of cheap sensor nodes and a small portion of anchor nodes with known locations, and obtain inter-node distance estimates based on RSS (Received Signal Strength) measurements, which significantly restricts the performance of localization algorithms. Hence, localization in large-scale WSNs has become an attractive research topic in the field of WSNs. This project will deal with the mechanism of sensor localization in large-scale WSNs, establish the scaling laws of and the propagation rule of multihop sensor localization errors, and then reveal intrinsic mechanism of as well as the propagation rule of multihop localization errors. And on this basis, by investigating the influence of anchor placement on localization accuracy, the dependency between the optimal anchor placement and the density of sensor nodes is demonstrated, and further, the optimal algorithm for placing anchors is presented. After carefully studying the key parts of distributed localization algorithms, including estimating inter-node distances based on RSS, localizing a single sensor node and flooding messages from anchor nodes, a systematic optimization framework is taken into consideration based on information fusion and graph theory, with the result that an energy-efficient distributed localization algorithm is proposed to restrict the propagation of localization errors. As a result, the outcomes of this project are helpful to the error theory of multihop localization in large-scale WSNs and provide theoretical guidance and practical methods for developing and deploying large-scale WSNs.