低指纹密度下分段式高精确度室内定位方法

针对室内环境中传统定位方法在大定位区域、低指纹密度下定位精确度低、计算复杂度高的问题，提出了一种基于线性插值法和分布重叠的分段式定位方法。该方法采用传统的最近邻法进行粗定位，得到可信区域；利用线性插值法更新可信区域内指纹数据库，增加指纹密度；在可信区域内，采用基于分布重叠的指纹相似度匹配法实现精定位。实验结果表明，在低指纹密度下，该定位方法定位精确度较高，算法复杂度适中，具有一定的适用性。

Segmented high accuracy indoor positioning under low fingerprint density

Aiming at low positioning accuracy and high computational complexity of the traditional indoor positioning method under large area and low fingerprint density, a new segmented high accuracy positioning algorithm based on linear interpolation and distribution overlapping is proposed. The algorithm uses the traditional nearest neighbor method for coarse positioning, and then updates the fingerprint database in the trusted area with linear interpolation method to increase the fingerprint density. Finally, fine positioning is realized by fingerprint matching method that based on distribution overlapping. Experimental results show that, even under low fingerprint density, the proposed method can get good positioning performance with high accuracy and suitable complexity.