A novel method of adaptive weighted K-nearest neighbor fingerprint indoor positioning considering user's orientation
A novel method of adaptive weighted K-nearest neighbor fingerprint indoor positioning considering user's orientation
复制标题
考虑用户方位的自适应加权K近邻指纹室内定位新方法
DOI：
10.1177/1550147718785885
复制
发表时间：
2018-06-28
影响因子：
2.3
通讯作者：
Wang, Yongkang
中科院分区：
文献类型：
--
作者：
Bi, Jingxue;Wang, Yunjia;Wang, Yongkang
室内环境中影响Wi-Fi信号的因素很多，其中人体有着重要的影响。而且，它的特性与用户的定位有关。为了消除用户人体因素引起的定位误差，提高定位精度，提出了一种考虑用户方位的自适应加权K近邻指纹定位方法。首先，提出了方向指纹数据库模型，该模型包括位置、方向和每个参考点处的平均接收信号强度指示符的序列。其次，以信号域和位置域的混合距离作为聚类特征，采用模糊c均值算法对方向指纹数据库进行聚类。最后，提出了一种自适应算法，通过匹配操作选择K个参考点，去除信号域距离、最小和最大坐标值较大的参考点，并计算剩余参考点的加权平均坐标作为定位结果。实验结果表明，该方法的平均误差减小了0.7m，均方根误差减小到1.3m左右。实验结果表明，本文提出的自适应加权K近邻指纹定位方法能够提高定位精度。
There are many factors affecting Wi-Fi signal in indoor environment, among which the human body has an important impact. And, its characteristic is related to the user's orientation. To eliminate positioning errors caused by user's human body and improve positioning accuracy, this study puts forward an adaptive weighted K-nearest neighbor fingerprint positioning method considering the user's orientation. First, the orientation fingerprint database model is proposed, which includes the position, orientation, and the sequence of mean received signal strength indicator at each reference point. Second, the fuzzy c-means algorithm is used to cluster orientation fingerprint database taking the hybrid distance of the signal domain and position domain as the clustering feature. Finally, the proposed adaptive algorithm is developed to select K-reference points by matching operation, to remove the reference points with larger signal-domain distances, minimum and maximum coordinate values, and calculate the weighted mean coordinates of the remaining reference points for positioning results. The experimental results show that the average error decreases by 0.7m, and the root mean square error decreases to about 1.3m by the proposed technique. And, we conclude that the proposed adaptive weighted K-nearest neighbor fingerprint positioning method can improve positioning accuracy.