【作者】 梁甲金；

【导师】 邓平；

【作者基本信息】 西南交通大学 ， 通信与信息系统， 2010， 硕士

【摘要】 无线传感器网络(Wireless Sensor Networks, WSN)通过引入网络技术使具有数据采集能力、信息处理能力和无线通信能力的传感器节点不再是单个的感知单元,而是能够交换信息、协调控制的有机结合体。无线传感器网络已成为一种全新的信息获取、处理、传输和控制系统,并在军事、工业、商业、医疗、环境监测等领域有着广阔应用前景。无线传感器网络广泛应用在环境恶劣、不可到达区域,实现监测与跟踪任务。在这些应用中,节点的位置信息起到很重要的作用,因此,对节点自身定位方法的研究具有重要的理论意义和极强的实用价值。本文在详细分析现有定位算法的基础上,重点对基于移动锚节点的无线传感器网络节点定位算法和移动锚节点路径规划方法进行了研究。首先,在查阅大量相关文献的基础上,本文综述了无线传感器网络技术在国内外的研究现状,介绍了无线传感器网络定位计算方法和定位算法分类方法,并详细分析了近年来该领域具有代表性的算法和研究成果。其次,考虑到无线传感器网络中的节点是随机部署在二维空间内的特定应用环境中,移动锚节点利用周围节点信息进行动态路径规划是一个重要研究方向。本文针对静态路径规划缺少灵活性的缺点,提出了一种禁忌搜索动态路径规划方法,通过仿真分析验证了该方法在节点非均匀分布情况下的有效性。然后,研究分析了基于半定规划的定位模型,仿真分析了该研究领域内主要算法的定位性能。针对集中式半定规划定位算法存在计算复杂度大的问题,本文重点探讨了一种基于多功率移动锚节点的分布式半定规划定位算法,仿真结果表明该算法可以大大减小算法的计算复杂度。最后,通过分析研究近年来移动锚节点辅助定位领域内的研究成果,本文提出了一种基于移动锚节点结合正三角形路径的多坐标系定位算法,结合正三角形路径,该算法可以取得很好的定位性能。在研究一种分布式移动锚节点辅助定位算法的基础上,针对该算法存在依赖直线移动路径的缺点,结合动态路径规划,提出了一种改进的分布式多功率移动锚节点辅助定位算法。仿真结果表明,改进算法在网络节点非均匀分布和网络区域存在障碍物的情况下可以取得比原算法更好的定位性能。更多还原

【Abstract】 The application of network technology has made wireless sensor networks (WSN) node which is consisted of sensing, data processing, and communicaiting components not only to be a sensor unit, but also to be a combination unit which can exchange information, collaborate and control. As a new information acquisition and process platform, Wireless sensor networks possess wide application prospects in military, industry, medical treatment, environment monitoring and other fields. Wireless sensor networks have widely used in the areas of poor conditions, to achieve the monitoring and tracking tasks. In these applications, the location of nodes play an important role in wireless sensor networks. Therefore, the study of node location method has important theoretical and practical value. This thesis focuses on the study of localization algorithm and path planning of the mobile anchor for wireless sensor networks based on the analysis of related literatures.First of all, the research status of wireless location techniques for wireless sensor networks at home and abroad are summarized based on the analysis of a lot of related literatures. The classification methods of localization systems and algorithms for wireless sensor networks are introduced and the principles and characteristics of recent representative localization algorithms are analyzed and discussed.Secondly, as sensor nodes are randomly deployed in two-dimensional applications, in the localization algorithm with a mobile beacon, how to choose the appropriate beacon trajectory is very important. At present relatively little research has been made. In this paper, a new dynamic path planning method is proposed. When mobile anchor nodes receive messages sent from neighbor nodes, the path that the mobile anchor should travel along is determined according to the information of neighbor nodes based on this search algorithm.Then, several semi-definite programming relaxation models are analyzed, the advantages and the disadvantages of several location algorithms in these fields are researched based on the simulation. Focus on the computation costs of semi-definite programming, in this thesis a new convex position estimation algorithm is studied. In this algorithm, a mobile anchor broadcasts packets with multi-power level on its trajectory. Simulation results show that this algorithm can greatly reduce the computational complexity and also perform well.Finally, a new distributed multi-coordinate localization algorithm using multi-power mobile anchors is proposed based on the study of a lot of recent representative localization algorithms. The algorithm can achieve very good positioning performance when it using triangle path. Based on the study of a distributed mobile anchor localization algorithm, aim at the defect of node beeline movement and using dynamic path planning method, a improved localization algorithm is proposed for the case of nodes non-uniform distribution. The simulation results show that this improved algorithm can obtain higher localization precision compared with the original algorithm in the network environments where the sensor nodes are non-uniform deployed and obstacles still occur in the area.更多还原