基于WLAN的室内定位指纹算法研究及应用

发布时间：2018-04-14 06:10

本文选题：WLAN室内定位 + 位置指纹　；参考：《杭州电子科技大学》2017年硕士论文

【摘要】：随着无线通讯技术和智能终端的广泛应用,使基于位置服务(Localtion Based Service,LBS)的应用逐渐从室外延伸到室内。由于室内WLAN部署广泛且移动终端可以容易获取接收信号强度,使基于位置指纹的WLAN室内定位方法受到了国内外学者的重视。针对室内环境中RSS信号波动性大以及不同终端信号接收能力差异性的问题,本文通过对RSS特性进行分析,提出一种新颖的位置指纹定位算法,并进一步对指纹算法进行了优化研究,最后设计开发了室内定位应用系统。本文主要工作和创新点如下:(1)针对RSS信号时变性以及不同终端信号接收能力差异性,导致WLAN位置指纹定位不稳定的问题,提出基于RSS空间位置线性相关的定位算法。该算法以离线采集的多组RSS样本形成的特征矩阵构建离线指纹数据库,定位时,通过计算实时RSS矩阵与指纹库参考点相关性,得到最相关的k个参考点,并利用二次加权质心算法计算用户的最终位置。为了有效降低信号时变性的影响,RSS采样时进行了滤波、排序等处理,构建离线指纹数据库时尽量增加采样次数,但需要对样本进行聚合处理以适应定位相关性计算。(2)为了降低定位匹配计算过程中计算开销,研究提出适用于RSS空间线性相关定位算法的位置指纹聚类方法。以RSS特征矩阵作为聚类样本,利用相关系数作为相似性度量标准,通过K-means聚类方法将指纹数据库分割成相对较小的指纹子库,匹配计算时采用动态筛选策略对指纹子库进行判断,通过匹配被选中的指纹子库即可估算最终位置,这样可以缩小匹配计算过程中指纹搜索空间,提升定位系统的效率。(3)不同AP布局会对定位性能造成不同影响,研究以参考点位置指纹区分度最大化为目的,提出一种AP布局规划参考方案。以参考点与相邻参考点欧氏距离之和表示为参考点指纹区分度,将所有参考点指纹区分度之和定义为当前AP布局区分度SD,以SD最大化为条件来布置AP的位置,这样可以有效改进系统的定位精度。
[Abstract]:With the wide application of wireless communication technology and intelligent terminal, the application of Localtion Based Service (LBS) is gradually extended from outdoor to indoor.Because the indoor WLAN is widely deployed and the mobile terminal can easily obtain the received signal strength, the WLAN indoor location method based on position fingerprint has been paid more attention by domestic and foreign scholars.In order to solve the problem of high volatility of RSS signal in indoor environment and the difference of receiving ability of different terminal signals, this paper proposes a novel location fingerprint location algorithm by analyzing the characteristics of RSS.Furthermore, the fingerprint algorithm is optimized and the indoor location application system is designed and developed.The main work and innovation of this paper are as follows: (1) aiming at the problem of RSS signal time-varying and different terminal signal receiving ability, which leads to the instability of WLAN location fingerprint location, a location algorithm based on RSS spatial position linear correlation is proposed.In this algorithm, the off-line fingerprint database is constructed from the characteristic matrix of multi-groups of RSS samples collected offline. When locating, the correlation between the real-time RSS matrix and the reference points of the fingerprint database is calculated, and the most relevant k reference points are obtained.The final position of the user is calculated by the quadratic weighted centroid algorithm.In order to effectively reduce the influence of time-varying signal on RSS sampling, filtering and sorting are carried out, and the sampling times are increased when constructing off-line fingerprint database.In order to reduce the computation cost of location matching, a location fingerprint clustering method suitable for RSS spatial linear correlation localization algorithm is proposed.RSS feature matrix is used as clustering sample, correlation coefficient is used as similarity measure, fingerprint database is divided into relatively small fingerprint subdatabase by K-means clustering method.The dynamic screening strategy is used to judge the fingerprint subdatabase, and the final position can be estimated by matching the selected fingerprint subdatabase, which can reduce the fingerprint search space in the course of matching calculation.To improve the efficiency of the positioning system, different AP layouts will have different effects on the location performance. In order to maximize the fingerprint differentiation of reference points, a reference scheme for AP layout planning is proposed in this paper.The sum of Euclidean distance between reference points and adjacent reference points is taken as the fingerprint differentiation degree of reference points. The sum of fingerprint differentiation degrees of all reference points is defined as the current AP layout differentiation degree SD.The position of AP is arranged under the condition of maximum SD.This can effectively improve the positioning accuracy of the system.
【学位授予单位】：杭州电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN925.93

【参考文献】