基于协同过滤的无线传感器网络节点定位算法研究

发布时间：2019-01-27 12:46

【摘要】：无线传感器网络(Wireless Sensor Network,WSN)是由大量的传感器节点组成的自组织网络系统。随着信息技术的快速发展,WSN已经成为当今的新兴科技技术。近年来它的应用范围极其广泛,例如:医疗、军事、智能家居、环境观测等。这些应用中如果没有精确的节点位置信息,则毫无意义。针对WSN中节点定位精确度较低的问题,本文提出了基于协同过滤的无线传感器网络节点定位算法。本文主要内容如下:首先,本文详细叙述了 WSN的基本概念、主要特点及其应用,并介绍了WSN中节点定位的算法分类、算法原理、以及经典定位算法的详细定位过程。其次,在深入地研究WSN定位算法的基础上,针对定位算法中常出现的镜像误差,提出了基于协同过滤的镜像误差修正定位算法。该算法结合节点间的几何特征,引入定位模型,选出合理的参考节点,过滤出精确度高的候选点作为节点的位置信息,减少了节点间的误差传递,降低了整个网络的定位误差。再次,针对两个及两个以上相邻的节点同时发生镜像误差的情况,提出了WSN中基于协同过滤方法的改进定位算法。该算法利用节点间的相互协作和节点间的几何关系,引入定位模型,不仅迅速有效地对已经发生镜像误差的节点进行修正定位,还可以对该分布情况下的未知节点进行快速高效地定位,且避免了节点在定位过程中产生镜像误差的现象,从而降低了定位误差,减少了通信开销,提高了定位效率。最后,利用NS-2仿真软件对本文提出的基于协同过滤的无线传感器网络节点定位算法进行仿真实验。通过改变算法中的相关参数,对本文提出的定位算法与二阶段模拟退火定位算法、避免镜像误差的乐观定位算法、基于磁极思想的求精算法、无线传感器网络SL-n迭代定位算法的定位结果进行比较,得出了如下结论:在相同的网络环境下,本文提出的基于协同过滤的无线传感器网络节点定位算法,在定位精确度和能量消耗方面都优于另外两种算法。
[Abstract]:Wireless Sensor Network (Wireless Sensor Network,WSN) is an ad hoc network system composed of a large number of sensor nodes. With the rapid development of information technology, WSN has become a new technology. In recent years, it is widely used, such as medical, military, smart home, environmental observation and so on. There is no point in these applications without accurate node location information. Aiming at the problem of low accuracy of node location in WSN, a node localization algorithm based on cooperative filtering is proposed in this paper. The main contents of this paper are as follows: firstly, the basic concept, main characteristics and application of WSN are described in detail, and the classification and principle of node location algorithm in WSN are introduced, as well as the detailed localization process of classical localization algorithm. Secondly, based on the in-depth study of the WSN localization algorithm, a mirror error correction algorithm based on collaborative filtering is proposed for the image errors that often occur in the localization algorithm. Combined with the geometric characteristics of nodes, the algorithm introduces a location model, selects reasonable reference nodes, filters out candidate points with high accuracy as the location information of nodes, reduces the error transfer between nodes and reduces the positioning errors of the whole network. Thirdly, an improved localization algorithm based on cooperative filtering method in WSN is proposed for two or more adjacent nodes with image errors occurring at the same time. Based on the cooperation between nodes and the geometric relationship between nodes, the algorithm introduces a localization model, which not only modifies the nodes with mirror errors quickly and effectively, but also modifies them. The unknown nodes in this distribution can be located quickly and efficiently, and the mirror image error can be avoided in the localization process, thus reducing the location error, reducing the communication overhead and improving the localization efficiency. Finally, the simulation experiment of the node location algorithm based on collaborative filtering is carried out by using NS-2 simulation software. By changing the relevant parameters of the algorithm, the location algorithm proposed in this paper and the two-stage simulated annealing location algorithm, the optimistic location algorithm to avoid mirror error, the refinement algorithm based on magnetic pole thought, The results of SL-n iterative localization algorithm for wireless sensor networks are compared, and the following conclusions are obtained: under the same network environment, this paper proposes a node localization algorithm based on cooperative filtering for wireless sensor networks. It is superior to the other two algorithms in location accuracy and energy consumption.
【学位授予单位】：辽宁大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【参考文献】