基于模糊聚类的位置指纹室内定位优化技术研究
发布时间：2019-02-17 15:48
【摘要】：随着无线通信和智能移动终端技术的飞速发展,人们对位置感知服务的要求越来越高,而高精度的室内定位技术是实现位置感知服务的核心。在当前室内环境WAN分布广泛和运用普遍化的今天,使得基于WLAN定位系统成为研究的热点课题。基于位置指纹的WLAN室内定位具有实现简单、成本较低,无需知道AP的位置和发射功率即可实现定位,对额外的硬件设备需求量少等优点,使其在学术界和工业界得到广泛的关注和运用。位置指纹定位技术主要分为两个阶段：采样阶段和定位阶段,当前位置指纹定位技术在实现定位时,在兼顾位置指纹定位算法的精度和效率上还没有相对完善的机制,因此本文针对这个问题对位置指纹定位技术实施优化改进。 本文通过分析比较几种典型的位置指纹定位算法,KNN定位算法在时间复杂度和定位精度上都有一定的优势,但是KNN算法在定位时需要耗费巨大的时间与指纹数据库进行比对从而确定K个指纹数据的选取,且由于K的选择是固定的而影响了某些位置处的定位精度。 针对KNN定位算法存在的不足,本文提出了一种基于模糊聚类改进的KNN定位算法,对定位环节的庞大指纹数据库运用聚类分析方法实现模糊聚类,继而采用KNN算法实现移动终端的定位。对改进后的位置指纹定位技术性能利用MATLAB仿真实验进行测试,并在Android平台上通过原型系统进行实验验证。 最后通过仿真实验分别对传统的KNN算法和改进后的KNN算法进行分析比较,在不影响定位系统其他性能的机制下,最终实验结果表明改进后的位置指纹定位在时间性能和匹配效率上都有极大的提高。
[Abstract]:With the rapid development of wireless communication and intelligent mobile terminal technology, people demand more and more location sensing services, and high precision indoor positioning technology is the core of location sensing services. With the wide distribution and generalization of WAN in indoor environment, WLAN based positioning system has become a hot topic. WLAN indoor location based on position fingerprint has the advantages of simple realization, low cost, no need to know the location and transmitting power of AP, and less demand for additional hardware equipment. It has been widely concerned and applied in academia and industry. The location fingerprint location technology is mainly divided into two stages: sampling stage and location stage. When the current location fingerprint localization technology realizes the location, there is not a relatively perfect mechanism in taking into account the accuracy and efficiency of the location fingerprint location algorithm. Therefore, this paper optimizes and improves the location fingerprint location technology. Through the analysis and comparison of several typical location fingerprint location algorithms, the KNN localization algorithm has some advantages in time complexity and location accuracy. However, the KNN algorithm needs to spend a great deal of time comparing with the fingerprint database to determine the selection of K fingerprint data, and because the selection of K is fixed, the location accuracy of some locations is affected. In view of the shortcomings of KNN localization algorithm, this paper proposes an improved KNN location algorithm based on fuzzy clustering, which uses clustering analysis method to realize fuzzy clustering for large fingerprint database of location link. Then the location of mobile terminal is realized by KNN algorithm. The performance of the improved location fingerprint location technology is tested by MATLAB simulation experiment and verified by the prototype system on the Android platform. Finally, the traditional KNN algorithm and the improved KNN algorithm are analyzed and compared by simulation experiments. The experimental results show that the time performance and matching efficiency of the improved location fingerprint location are greatly improved.
【学位授予单位】：云南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925.93
本文编号：2425328
[Abstract]:With the rapid development of wireless communication and intelligent mobile terminal technology, people demand more and more location sensing services, and high precision indoor positioning technology is the core of location sensing services. With the wide distribution and generalization of WAN in indoor environment, WLAN based positioning system has become a hot topic. WLAN indoor location based on position fingerprint has the advantages of simple realization, low cost, no need to know the location and transmitting power of AP, and less demand for additional hardware equipment. It has been widely concerned and applied in academia and industry. The location fingerprint location technology is mainly divided into two stages: sampling stage and location stage. When the current location fingerprint localization technology realizes the location, there is not a relatively perfect mechanism in taking into account the accuracy and efficiency of the location fingerprint location algorithm. Therefore, this paper optimizes and improves the location fingerprint location technology. Through the analysis and comparison of several typical location fingerprint location algorithms, the KNN localization algorithm has some advantages in time complexity and location accuracy. However, the KNN algorithm needs to spend a great deal of time comparing with the fingerprint database to determine the selection of K fingerprint data, and because the selection of K is fixed, the location accuracy of some locations is affected. In view of the shortcomings of KNN localization algorithm, this paper proposes an improved KNN location algorithm based on fuzzy clustering, which uses clustering analysis method to realize fuzzy clustering for large fingerprint database of location link. Then the location of mobile terminal is realized by KNN algorithm. The performance of the improved location fingerprint location technology is tested by MATLAB simulation experiment and verified by the prototype system on the Android platform. Finally, the traditional KNN algorithm and the improved KNN algorithm are analyzed and compared by simulation experiments. The experimental results show that the time performance and matching efficiency of the improved location fingerprint location are greatly improved.
【学位授予单位】：云南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN925.93
【参考文献】
相关期刊论文 前6条
1 房秉毅;;基于超宽带技术的室内定位系统[J];电子技术应用;2006年07期
2 杨帆;赵东东;;基于Android平台的WiFi定位[J];电子测量技术;2012年09期
3 冯浩然;袁睿翕;慕春棣;;无线网络中基于信号强度的定位及算法比较[J];计算机工程与应用;2006年32期
4 周傲英;杨彬;金澈清;马强;;基于位置的服务:架构与进展[J];计算机学报;2011年07期
5 罗军舟;吴文甲;杨明;;移动互联网:终端、网络与服务[J];计算机学报;2011年11期
6 余小鹏;周德翼;;一种自适应k-最近邻算法的研究[J];计算机应用研究;2006年02期
，本文编号：2425328
本文链接：https://www.wllwen.com/kejilunwen/wltx/2425328.html
