Abstract

Aiming at the characteristics of large passenger flow and the complex and variable indoor environment in the airport terminal, an indoor localization algorithm based on Received Signal Strength feature extension and Spectral Regression Kernel Discriminant Analysis is proposed. In the offline phase, the Least Square-Support Vector Machine regression model is used to estimate the distance between the terminal and the Access Point, and the Received Signal Strength features are extended based on this. The Spectral Regression framework is introduced on the basis of Kernel Discriminant Analysis. The non-linear features of the Original Location Fingerprint were extracted by this algorithm to generate a new feature fingerprint dataset. During the online stage, Spectral Regression Kernel Discriminant Analysis was firstly used to process the extended Spectral Regression Kernel feature of the point to be positioned, and then use the weighted K nearest neighbor algorithm for position estimation. Experimental results show that the algorithm in this paper can effectively reduce the average error and improve the indoor localization accuracy in the complex terminal environment with large passenger flow and non-line-of-sight environment.

中文翻译：

---

基于RSS特征扩展和谱回归核判别分析的终端室内定位算法

摘要

针对机场航站楼客流量大、室内环境复杂多变的特点，提出了一种基于接收信号强度特征扩展和频谱回归核判别分析的室内定位算法。在离线阶段，使用最小二乘支持向量机回归模型来估计终端与接入点之间的距离，并在此基础上扩展接收信号强度特征。Spectral Regression 框架是在核判别分析的基础上引入的。该算法提取原始位置指纹的非线性特征，生成新的特征指纹数据集。在线上阶段，谱回归核判别分析首先对待定位点的扩展谱回归核特征进行处理，然后采用加权K近邻算法进行位置估计。实验结果表明，本文算法在客流量大、非视距环境的复杂终端环境中，能有效降低平均误差，提高室内定位精度。