基于凸规划的无线传感器网络定位算法研究

发布时间：2019-03-29 21:05

【摘要】：无线传感器网络(Wireless Sensor Networks,WSN)是一种集数据收集、融合、处理、传输等功能于一体的网络系统。作为物联网的基础和重要组成部分,他触及到了软硬件技术、计算机网络及通信技术等多门学科领域。在军事、医疗、环境监测、森林火灾、空间探测等无线传感器的网络应用中,不仅需要采集数据,往往还需要精确确定数据的位置信息。因此,节点定位技术在无线传感器中占据着十分重要、不可或缺的地位。在实际的环境中,常常会遇到不规则的网络拓扑、复杂的传播环境、不均匀的节点部署等问题,这对定位算法的性能提出了更高的要求。本文对已有定位算法进行了研究分析,甄别已有算法的不足,对研究的凸规划定位算法做了详细的分析,并提出了改进算法。本文的主要工作如下:(1)分析了目前国内外凸规划定位算法的研究现状,指出了凸规划定位算法中研究的不足。针对经典的RSSI-Convex算法,分析了该算法的基本原理、算法过程、以及存在的问题。(2)在对RSSI-Convex算法的分析中发现其对RSSI的条件利用不够充分,经过论证后,提出了基于RSSI所成圆的改进算法。该算法对原始算法的改进如下:针对未知节点通信半径内仅有一个信标节点的情况,不在一味地用信标节点坐标代替未知节点坐标。改进算法利用了物理中转变参考系的思想,达到了缩减未知节点未知区域的目的;针对未知节点通信半径内有一个以上信标节点的情况,利用RSSI所成的圆来进一步缩小由RSSI-Convex所确定的未知节点的未知区域,达到了提高定位精度的目的。实验结果表明,在对RSSI充分利用的条件下就可以提高节点的定位精度。(3)在RSSI-Convex凸规划的算法基础上,结合查阅的大量文献,提出了改进算法。算法通过引入变角的概念,对RSSI-Convex算法进行了大胆的改进:用扇形取代圆,用变角来进一步缩小未知节点的未知区域,达到提高定位精度的目的;通过概率均分的思想将未知节点合理的限定在角平分线上,使得未知节点的求解过程进一步简化。实验结果表明,改进算法比原算法具有更高的定位精度,并可以较好的改善所存在的问题。对比两个改进算法可以看出,基于动态角度划分的改进算法,其算法曲线较为平缓,可以满足信标节点密度低而对精度要求高的领域,如军事、医疗等。而基于接收信号强度的改进算法可以在不增加成本的前提下,提供更高的节点精度,使得在环境监测、森林火灾等领域得到的信息来源的准确性更高。
[Abstract]:Wireless sensor network (Wireless Sensor Networks,WSN) is a data collection, fusion, processing, transmission and other functions in one network system. As the foundation and important part of the Internet of things, he touched on many subject fields such as hardware and software technology, computer network and communication technology. In the application of wireless sensors, such as military, medical, environmental monitoring, forest fire, space detection and other wireless sensor applications, it is necessary not only to collect data, but also to accurately determine the location information of the data. Therefore, node positioning technology plays a very important and indispensable role in wireless sensors. In the actual environment, irregular network topology, complex propagation environment, uneven node deployment and other problems are often encountered, which put forward higher requirements for the performance of the localization algorithm. In this paper, the existing localization algorithms are studied and analyzed, the shortcomings of the existing algorithms are identified, and the convex programming localization algorithm is analyzed in detail, and an improved algorithm is proposed. The main work of this paper is as follows: (1) the present research status of convex programming localization algorithm at home and abroad is analyzed, and the deficiency of the research on convex programming localization algorithm is pointed out. In view of the classical RSSI-Convex algorithm, the basic principle, algorithm process and existing problems of the algorithm are analyzed. (2) in the analysis of the RSSI-Convex algorithm, it is found that the condition of RSSI is not sufficiently utilized. An improved algorithm for circle formation based on RSSI is proposed. The improvement of the original algorithm is as follows: for the case that there is only one beacon node in the communication radius of the unknown node, the beacon node coordinates are not always used to replace the unknown node coordinates. The improved algorithm makes use of the idea of changing reference system in physics and achieves the purpose of reducing the unknown region of unknown nodes. Aiming at the case that there is more than one beacon node in the communication radius of the unknown node, the unknown area of the unknown node determined by the RSSI-Convex is further reduced by using the circle formed by the RSSI, and the accuracy of the location is improved. The experimental results show that the localization accuracy of nodes can be improved by fully utilizing RSSI. (3) based on the algorithm of RSSI-Convex convex programming, an improved algorithm is proposed based on a large number of references. By introducing the concept of variable angle, the RSSI-Convex algorithm is improved boldly: the sector is used to replace the circle, and the variable angle is used to further reduce the unknown region of the unknown node so as to improve the positioning accuracy; The unknown node is reasonably limited to the angular bisection line by the idea of equal probability, which further simplifies the solving process of the unknown node. The experimental results show that the improved algorithm has higher positioning accuracy than the original algorithm and can improve the existing problems. Compared with the two improved algorithms, it can be seen that the improved algorithm based on dynamic angle partition has a smooth curve, which can meet the requirements of low density of beacon nodes and high precision, such as military and medical treatment. The improved algorithm based on the received signal strength can provide higher node accuracy without increasing the cost, which makes the accuracy of information sources in environmental monitoring, forest fire and other fields more accurate.
【学位授予单位】：广西师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP212.9;TN929.5

【相似文献】