基于区域网格划分的SVM室内定位算法

SVM INDOOR POSITIONING ALGORITHM BASED ON REGIONAL MESHING

随着城市化的快速发展,人们越来越多的时间处于室内,因此室内定位方法和系统的研究越来越得到关注。针对室内环境相对复杂造成的RSSI值波动大、定位响应不及时的问题,采用有效的预处理,兼顾定位速度和定位精度,提出一种基于区域网格划分的SVM定位算法。离线采集阶段进行SVM"位置指纹"模型训练。在线定位阶段,进行数据筛选和数据修正等预处理,再进行区域网格匹配。在K个区域网格内使用SVM分类器进行二分类,通过投票法确定移动终端的位置。实验结果表明,该定位算法在保证定位速度的前提下能够有效地提高定位精度。

With the rapid development of urbanization, people spend more time indoors, so the study of indoor positioning methods and systems is getting more attention. Aiming at the problem that RSSI value fluctuates greatly and positioning response was not timely due to the relatively complex indoor environment, an SVM positioning algorithm based on regional meshing was proposed by using effective pre-processing and considering the positioning speed and accuracy. In the offline acquisition stage, the algorithm performed SVM position fingerprint model training. In the online positioning stage, the algorithm performed pre-processing, such as data filtering and data modification, and matched region grids. SVM classifiers were used for two classifications in K region grids, and the position of mobile terminals was determined by voting method. Experimental results show that the proposed algorithm can effectively improve the positioning accuracy while ensuring the positioning speed.