基于相似性的无线传感网定位算法研究

发布时间：2018-03-12 14:38

本文选题：无线传感器网络　切入点：节点定位算法　出处：《中国矿业大学》2017年硕士论文　论文类型：学位论文

【摘要】：无线传感器网络(Wireless Sensor Network,WSN)是一门集中了嵌入式技术、微电子技术、传感器技术、分布式信息处理技术和通信网络等技术的交叉学科,在环境监测、医疗卫生、国防军事、交通管理等领域具有广阔的应用前景。现有应用大都需要预先获知传感器节点的位置,没有位置信息的监测数据往往毫无意义。因此,对无线传感器网络节点定位的研究具有重要意义。首先,本文从信号相似性出发,提出一种基于信号相似性的无线传感器网络节点定位算法。该方法通过比较两个邻居节点接收到其它节点发送来信号的相似性,计算这两个节点间的相对距离。使用最短路径算法计算非邻居节点间的相对距离后得到整个网络的相对距离图。在已知至少3个锚节点位置的前提下,利用MDS算法计算出节点的物理位置。其次,在信号相似性基础上,结合结构相似性,提出一种融合结构相似性的无线传感器网络节点定位算法。首先,根据传感器节点是否能够通信生成一个网络拓扑图。然后计算任意两个节点之间信号相似性值作为拓扑图边的权值。使用基于加权SimRank算法计算两个节点之间的结构相似度并作为相对距离。最后,在已知至少3个锚节点位置的前提下,使用MDS算法完成对未知节点的定位。最后,分别在规则的和非规则的无线传感器网络中对本文所提出的两种算法进行了实验验证,结果表明,本文提出的两种基于相似性的无线传感器网络节点定位算法具有较好的定位精度。
[Abstract]:Wireless Sensor Network (WSNs) is an interdisciplinary subject that focuses on embedded technology, microelectronics, sensor technology, distributed information processing technology and communication network, in environmental monitoring, health care, national defense, military affairs, etc. Most of the existing applications need to know the location of sensor nodes in advance, and monitoring data without location information are often meaningless. The research on node localization in wireless sensor networks is of great significance. Firstly, this paper starts from the similarity of signals. A node localization algorithm based on signal similarity in wireless sensor networks is proposed, which compares the similarity of signals sent by two neighbor nodes to other nodes. The relative distance between the two nodes is calculated. The relative distance graph of the whole network is obtained by using the shortest path algorithm to calculate the relative distance between the non-neighbor nodes. The physical location of nodes is calculated by using MDS algorithm. Secondly, based on the similarity of signals and structural similarity, a node location algorithm based on structural similarity in wireless sensor networks is proposed. A network topology graph is generated according to whether the sensor node can communicate. Then the similarity value of the signal between any two nodes is calculated as the weight of the edge of the topology graph. Based on the weighted SimRank algorithm, the junction between the two nodes is calculated. Construct the similarity and act as the relative distance. Finally, Under the condition that at least three anchor nodes are known, MDS algorithm is used to locate the unknown nodes. Finally, the two algorithms proposed in this paper are experimentally verified in both regular and irregular wireless sensor networks. The results show that the proposed two similarity based node localization algorithms in wireless sensor networks have good localization accuracy.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【参考文献】