虚拟AP融合CNN模型的自适应RSSI指纹定位方法  

ADAPTIVE RSSI FINGERPRINT POSITIONING METHOD BASED ON VIRTUAL AP FUSED WITH CNN MODEL

作　　者：吴仕勋[1] 黄文鲜 李敏 徐凯[1] Wu Shixun;Huang Wenxian;Li Min;Xu Kai(School of Information Science and Engineering,Chongqing Jiaotong University,Chongqing 400074,China)

机构地区：[1]重庆交通大学信息科学与工程学院,重庆400074

出　　处：《计算机应用与软件》2025年第1期72-81,共10页Computer Applications and Software

基　　金：重庆市教委科技研究计划青年项目(KJQN202000703);四川省川渝合作重点研发项目(2020YFQ0057);2018年重庆市技术创新与应用示范重大课题专项(cstc2018jszx-cyztzxX0034)。

摘　　要：基于RSSI(Received Signal Strength Indication)位置指纹的Wi-Fi室内定位现已被大量应用于各类基于位置信息的服务中。但指纹定位的精度受到RSSI信号的剧烈波动影响,难以满足高精度位置信息服务的需求。为克服该困难,提出一种结合虚拟AP技术与高精度CNN(Convolutional Neural Network)判别模型的定位方法。该方法通过距离比定位得到虚拟AP的位置,并将该信息与RSSI融合作为数据增强CNN模型的输入,确定样本的位置。设计实验方案采集实际的用户终端RSSI数据,构建指纹定位的数据集,验证所提出的指纹定位方案的有效性。实验结果表明,在该数据集上,所提出的方法在确定区域时的准确度达到91%,并将95%的定位误差控制在2 m以内。对比现有的定位方案,所提出的方案在定位精度上有显著提升。The indoor Wi-Fi positioning system based on RSSI positioning fingerprint has been widely used in all kinds of location-based services.However,the accuracy of fingerprint positioning is susceptible to the sharp fluctuation of RSSI which is difficult to meet the demand for precise indoor positioning.To overcome these difficulties,this paper proposes a new positioning method combining virtual AP technology with a high-precision CNN classification model.In the proposed method,the position of the virtual AP was obtained by distance ratio localization,and this information was fused with RSSI as the input of the data augmented CNN model to determine the position of the sample.Through designing the experimental scheme,we collected the actual user terminal RSSI data,built the dataset of fingerprint positioning,and verified the effectiveness of the positioning fingerprint method.The experiment results on this dataset show that the proposed method's accuracy of area determination is up to 91%,and 95%of the positioning error is controlled within 2 meters.Compared with the existing positioning methods,the proposed method has a significant improvement in terms of positioning accuracy.

关 键 词：接收信号强度指示 指纹定位 卷积神经网络 虚拟AP 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]