基于WiFi信号强度定位算法模拟.pdf

摘 要 随着移动互联网的发展，利用移动设备进行导航定位以及使用基于位置的服 务已经成为日常生活中必不可少的事情。然而在室内环境中，由于建筑物的遮挡 使得基于卫星系统的导航难以在此类环境中使用，故需要有效的室内定位解决方 案。好在公共场合的无线网络的普及，使得利用无线网络进行室内定位成为了一 个成本较低而且容易实现的室内定位方案。本课题研究的是基于多 AP 的 Wi-Fi 网络信号强度的室内定位模拟。首先论述室内定位技术以及基于 Wi-Fi 的定位技 术，并对 Wi-Fi 位置指纹定位算法进行研究，包括最近邻法、K 近邻法、加权 K 近邻法以及朴素贝叶斯法。然后设计一个系统，运用 MATLAB 平台实现对定位算 法进行模拟，使用 MATLAB 中的图形界面(GUI)将定位结果直观地展现出来。最后 研究针对 AP 的分布、指纹密度以及 K 值对定位结果的影响。 关键词：室内定位，位置指纹，最近邻，定位误差，MATLAB 3 Abstract With the development of the mobile internet, it’s essential to use navigation or location-based services by mobiles in the daily life. But with the architecture of the building, the signal of satellites was covered up, that makes the difficulty of satellite navigation indoors. So, the rapid way to solve indoor-locate is what we need. Good to the widespread of WLAN in public, using wireless network signal to locate indoors becomes a low-cost and easy-to-implement scheme. The topic is to simulate the indoor-locate based on signal intensity of Wi-Fi wireless network with several APs. Firstly, introduce the technology of indoor signal positioning system, and Wi-Fi- based positioning. Study the aching algorithm of Wi-Fi-based fingerprints positioning, including NN, KNN, WKNN, and Bayes algorithm. Secondly, design the system of simulating the positioning algorithm by MATLAB platform, and use GUI tool to show the outcome of the position intuitively and in convenience. At last, analyze the influence of positioning with distribution of APs, the density of fingerprints and the choice of K in KNN or WKNN. KeyWords: Indoor positioning ； Location fingerprint ； Nearest neighbor ； Positioning error； MATLAB 4