复杂条件下的无线传感器网络定位技术研究

发布时间：2018-01-25 23:01

本文关键词： 无线传感器网络 RSSI 卡尔曼滤波线性回归移动节点　出处：《南京航空航天大学》2014年硕士论文　论文类型：学位论文

【摘要】：无线传感器网络(WSN，Wireless Sensor Network)作为获取信息的重要手段，在军事战场、环境监测、医疗护理等诸多领域有广泛的应用。而对于诸多应用如目标跟踪、森林火灾监控、远程数据采集等都需要确定的位置信息，因此节点定位技术对WSN的应用至关重要，是WSN核心技术之一。本文主要研究了复杂条件下的节点定位技术。针对RSSI值常常受到各种复杂条件的干扰，提出了基于卡尔曼滤波的RSSI值筛选策略。另外一方面，在RSSI值与实际距离的信号衰减模型中，提出了信号衰减方程与线性回归理论结合的方法获取准确的环境参数，修正RSSI测距模型。实验证明该方法能够有效的抑制干扰对测距结果的影响。一些定位场合中，锚节点的个数超过三个，其性能会有优劣之分，应该尽量选取性能较好的锚节点进行定位，，否则将性能较差的锚节点加入定位会造成定位误差增大。因此本文提出了一种基于线性回归分析的定位策略。用相关系数和剩余标准差对锚节点的测距模型进行性能评估，制定了一套适合实际情况的定位策略。实验表明，采用修正的测距模型和新的定位策略，使得节点定位精度明显提高。为了解决锚节点稀疏的条件下，传统的静态节点定位方法往往无法实现移动节点定位的问题，提出了一种基于RSSI测距和蒙特卡洛预测相结合的移动节点定位算法。当移动节点进入锚节点稀疏区域时，通过RSSI测距和最小二乘曲线拟合预测当前时刻各自的采样区域，将最终的采样区域缩小在它们的交集以内，缩减采样区域的同时提升了样本采样成功率和定位的精度。最后利用RSSI线性回归分析技术和基于预测的移动节点定位方法相结合，实现了锚节点不同密度情况下的移动节点定位。实验表明，该定位方法效果良好。
[Abstract]:Wireless sensor network (WSNN Wireless Sensor Network) as an important means to obtain information, in the military battlefield, environmental monitoring. Medical care and many other fields have a wide range of applications, but for many applications such as target tracking, forest fire monitoring, remote data collection and other needs to determine the location information. Therefore, node location technology is very important to the application of WSN, and is one of the core technologies of WSN. In this paper, we mainly study the node location technology under complex conditions. Aiming at the RSSI value is often disturbed by various complex conditions, a RSSI value filtering strategy based on Kalman filter is proposed. In the signal attenuation model between the RSSI value and the actual distance, the method of combining the signal attenuation equation with the linear regression theory is proposed to obtain the accurate environmental parameters. The RSSI rangefinder model is modified. Experiments show that the proposed method can effectively suppress the influence of interference on ranging results. In some positioning situations, the number of anchor nodes is more than three, its performance will have advantages and disadvantages, we should try to select better anchor nodes to locate. Otherwise, the location error will increase if the anchor node with poor performance is added. Therefore, a localization strategy based on linear regression analysis is proposed in this paper. Correlation coefficient and residual standard deviation are used to carry out the location model of anchor node. Performance evaluation. A set of localization strategies suitable for the actual situation is established. The experimental results show that the modified ranging model and the new location strategy can obviously improve the accuracy of node localization. In order to solve the problem that the traditional static node localization method can not realize the mobile node location under the condition of sparse anchor nodes. A mobile node location algorithm based on the combination of RSSI ranging and Monte Carlo prediction is proposed, when the mobile node enters the sparse area of the anchor node. The RSSI ranging and least square curve fitting are used to predict the respective sampling regions at the current time, and the final sampling regions are reduced to their intersection. At the same time, the sample sampling success rate and localization accuracy are improved by reducing the sampling area. Finally, the RSSI linear regression analysis technique and the mobile node location method based on prediction are combined. The location of mobile nodes with different density of anchor nodes is realized, and the experimental results show that the proposed method is effective.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP212.9;TN929.5

【参考文献】