中文摘要：

基于WLAN的定位服务现今已成为智慧城节中一个很有吸引力的研究领域。在各种定位算法中，经典欧氏距离法的度量方式只考虑各实际位置点RSS向量之间的绝对距离，往往忽视各实际位置点RSS向量之间的相对距离；并且只能给各AP赋予相同的权重。为克服欧氏距离湛-的不足，提出了基于卡方距离及灵敏度法的WLAN室内定位方法（CSKNN）。该方法利用位置指纹信息建立参考点的指纹信息和测试点的指纹信息，然后利用更能反映特征量之间相对距离的卡方距离并结合灵敏度法对各AP权重进行修正，得出在当前定位环境中各AP在定位系统中的贡献，用加权后的卡方距离依据各参考点的指纹信息计算待定位点的位置。结果表明，该方法比传统的欧氏距离法精度高。

英文摘要：

WLAN-based localization service has become a hotspot for smarter city nowadays. Among the localization algorithms, the clas- sical Euclidean distance solely keeps count of the absolute distance between the RSSI vector and overlooks the relative distance between the RSSI vector. And it can only give the same weight to every AP. In order to overcome the defects of Euclidean distance, a new algo- rithm based on Chi-square distance and sensitivity method for WLAN indoor localization is proposed. The algorithm uses fingerprinting technique to make training dataset and testing dataset,tht;n uses Chi-square distance and sensitivity method to correct the training dataset which will be used in the online localization phase and get the weight of every AP in the algorithm in order to improve positioning accura- cy. The results show that the proposed algorithm has better accuracy compared with the classical Euclidean distance.