WSN定位误差分析及精度控制方法
发布时间:2018-08-17 17:16
【摘要】:近年来,对于小型无线移动设备的定位得到了越来越多的关注。对WSN(Wireless Sensor Networks,无线传感器网络)的定位是其中一个分支。在WSN中,通常都存在一些为位置已知的节点,被称为锚节点,其他节点为位置未知的普通节点。未知节点通过测量自身与通信范围内锚节点的距离或到达角等信息来得到自身位置。事实上,利用信号强度传播模型,可以通过RSS(Received Signal Strength,接收信号强度)来估计出收发端的距离。基于RSS的测距方法得到了广泛的关注和研究,因为它不需要额外的硬件支持,且算法很简单。但该系统存在两个很大的问题。1)由于无线信道快衰落和慢衰落的影响,基于RSS的测距精度很低。慢衰落即阴影衰落,它对于测距精度的影响已经得到了广泛的研究。而快衰落由多径效应引起,一般认为可以通过多次测量取平均的方法来予以消除,所以很少有相应的研究。但是,当所要求的精度给定时,所需的测量次数依旧悬而未决。2)该定位系统的定位精度不止受到测距精度的影响,还与锚节点相对于未知节点的分布状况有关。这就使得不同的节点的定位精度不同,有时甚至相差很大。一些节点的定位精度很高,而另外一些可能很低。这就使得整个系统可靠性很差、很不稳定。针对以上问题,本文主要完成了以下方面的工作:1)对于测距问题,很重要的一个问题就是如何得到无线信道的参数。通过测量少数几个点的RSS值,再利用最小二乘法来估计所需参数。本文在实际环境下进行了实验,证明了该方法的有效性和准确性。2)从理论上详细地分析了快衰落对测距精度的影响。首先假设快衰落服从对数正态分布,推导出了原有的测距方法的误差分布函数。理论分析表明该方法存在固有偏差,但通过一个很简单的修正即可以变为无偏估计。为了证明该方法的性能,进一步推导了基于RSS测距的CRLB(Cram′er-Rao Lower Bound,克拉美罗下限),即测距误差的方差所能达到的最低程度。各个方面的论证都表明,测距误差的方差与收发端距离的平方成正比。3)测距后即得到了一系列的方程组,本文通过线性迭代的方法进行求解。通常使用最小二乘法,但在基于RSS测距的背景下,该方法可以被证明并非最优。本文进一步给出了最优的迭代矩阵,以达到最小化定位误差的目的。4)为了解决系统的不稳定性问题,本文定义了GDOP(Geographical Dilution of Precision,几何精度因子),并基于此概念提出了APC(Automatic Precision Control,自动精度控制)算法。该算法通过迭代动态地调节测量次数,从而使定位精度稳定在给定的限制附近。
[Abstract]:In recent years, more and more attention has been paid to the positioning of small wireless mobile devices. Location of WSN (Wireless Sensor Networks, wireless sensor networks is one of the branches. In WSN, there are usually some known nodes, called anchor nodes, and other nodes are common nodes with unknown location. Unknown nodes obtain their position by measuring the distance or angle of arrival between themselves and the anchor nodes in the communication range. In fact, using the signal intensity propagation model, the distance between the transceiver and receiver can be estimated by RSS (Received Signal Strength,. The ranging method based on RSS has been widely studied because it does not require additional hardware support and the algorithm is very simple. However, there are two big problems in this system. 1) because of the influence of fast fading and slow fading in wireless channel, the ranging accuracy based on RSS is very low. Slow fading is shadow fading, and its influence on ranging accuracy has been widely studied. However, fast fading is caused by multipath effect, which is generally considered to be eliminated by multiple measurements, so there is little research on it. However, when the required precision is timed, the required number of measurements is still unresolved (.2). The positioning accuracy of the positioning system is affected not only by the ranging accuracy, but also by the distribution of the anchor nodes relative to the unknown nodes. This makes the positioning accuracy of different nodes different, sometimes even very different. Some nodes have high positioning accuracy, while others may be very low. This makes the whole system very poor reliability, very unstable. Aiming at the above problems, this paper mainly completes the following work: 1) for the ranging problem, one of the most important problems is how to get the parameters of the wireless channel. By measuring the RSS values of a few points, the least square method is used to estimate the required parameters. In this paper, experiments are carried out in a practical environment, and the effectiveness and accuracy of this method are proved. 2) the influence of fast fading on ranging accuracy is analyzed in detail theoretically. First, the error distribution function of the original ranging method is derived from the logarithmic normal distribution of fast fading suit. Theoretical analysis shows that the method has inherent deviation, but it can be changed into unbiased estimation by a simple correction. In order to prove the performance of the method, the CRLB (Cram'er-Rao Lower Boundary) based on RSS ranging is further derived, that is, the minimum degree of variance of ranging error. It is shown from all aspects that the variance of ranging error is proportional to the square of the distance between the transceiver and the receiver) and a series of equations are obtained after ranging. The method of linear iteration is used to solve the problem in this paper. The least square method is usually used, but in the context of RSS ranging, the method can be proved not to be optimal. In order to solve the instability problem of the system, the optimal iterative matrix is given to minimize the positioning error. In this paper, GDOP (Geographical Dilution of Precision, geometric precision factor is defined, and based on this concept, APC (Automatic Precision Control, automatic precision control algorithm is proposed. The algorithm dynamically adjusts the number of times of measurement by iterating so that the positioning accuracy is stable near the given limit.
【学位授予单位】:哈尔滨工业大学
【学位级别】:硕士
【学位授予年份】:2015
【分类号】:TN929.5;TP212.9
本文编号:2188340
[Abstract]:In recent years, more and more attention has been paid to the positioning of small wireless mobile devices. Location of WSN (Wireless Sensor Networks, wireless sensor networks is one of the branches. In WSN, there are usually some known nodes, called anchor nodes, and other nodes are common nodes with unknown location. Unknown nodes obtain their position by measuring the distance or angle of arrival between themselves and the anchor nodes in the communication range. In fact, using the signal intensity propagation model, the distance between the transceiver and receiver can be estimated by RSS (Received Signal Strength,. The ranging method based on RSS has been widely studied because it does not require additional hardware support and the algorithm is very simple. However, there are two big problems in this system. 1) because of the influence of fast fading and slow fading in wireless channel, the ranging accuracy based on RSS is very low. Slow fading is shadow fading, and its influence on ranging accuracy has been widely studied. However, fast fading is caused by multipath effect, which is generally considered to be eliminated by multiple measurements, so there is little research on it. However, when the required precision is timed, the required number of measurements is still unresolved (.2). The positioning accuracy of the positioning system is affected not only by the ranging accuracy, but also by the distribution of the anchor nodes relative to the unknown nodes. This makes the positioning accuracy of different nodes different, sometimes even very different. Some nodes have high positioning accuracy, while others may be very low. This makes the whole system very poor reliability, very unstable. Aiming at the above problems, this paper mainly completes the following work: 1) for the ranging problem, one of the most important problems is how to get the parameters of the wireless channel. By measuring the RSS values of a few points, the least square method is used to estimate the required parameters. In this paper, experiments are carried out in a practical environment, and the effectiveness and accuracy of this method are proved. 2) the influence of fast fading on ranging accuracy is analyzed in detail theoretically. First, the error distribution function of the original ranging method is derived from the logarithmic normal distribution of fast fading suit. Theoretical analysis shows that the method has inherent deviation, but it can be changed into unbiased estimation by a simple correction. In order to prove the performance of the method, the CRLB (Cram'er-Rao Lower Boundary) based on RSS ranging is further derived, that is, the minimum degree of variance of ranging error. It is shown from all aspects that the variance of ranging error is proportional to the square of the distance between the transceiver and the receiver) and a series of equations are obtained after ranging. The method of linear iteration is used to solve the problem in this paper. The least square method is usually used, but in the context of RSS ranging, the method can be proved not to be optimal. In order to solve the instability problem of the system, the optimal iterative matrix is given to minimize the positioning error. In this paper, GDOP (Geographical Dilution of Precision, geometric precision factor is defined, and based on this concept, APC (Automatic Precision Control, automatic precision control algorithm is proposed. The algorithm dynamically adjusts the number of times of measurement by iterating so that the positioning accuracy is stable near the given limit.
【学位授予单位】:哈尔滨工业大学
【学位级别】:硕士
【学位授予年份】:2015
【分类号】:TN929.5;TP212.9
【参考文献】
相关硕士学位论文 前1条
1 刘文娟;基于RSSI测距的WSN定位系统设计[D];大连理工大学;2010年
,本文编号:2188340
本文链接:https://www.wllwen.com/kejilunwen/wltx/2188340.html