基于KWAP-KNN算法的无线传感覆盖室内定位优化

Indoor Location Optimization of Wireless Sensor Coverage Based on KWAP-KNN Partition Clustering Algorithm

分区聚类算法用于室内定位,具有较好的定位结果。为了提高无线传感覆盖室内定位精度,在分析基于Wi-Fi的室内定位信号的基础上,提出一种基于仿射传播算法(WAP)优化K最近邻算法(KNN)的WAP-KNN聚类算法,并开展实验测试分析。在分区时选择K-means,K个聚类中心作为信号发射装置所在位置,在划分交叉区域及未覆盖区域时采取聚类方式,在小区域内采用KNN聚类完成。实验结果证明,WAP聚类算法均比AP聚类算法性能更高,聚类效果更好。该研究对提高无线传感覆盖的室内定位精度具有很好的实际应用意义。

The partition clustering algorithm is used in indoor localization and has good localization results.In order to improve the indoor positioning accuracy of wireless sensing coverage,a clustering algorithm based on KWAP-KNN is proposed based on the analysis of indoor positioning signals based on Wi-Fi,and the experimental test analysis is carried out.When partitioning,k-means is selected,and K clustering centers are used as the location of the signal transmitter.When dividing cross areas and uncovered areas,clustering is adopted,and KNN clustering is adopted in small areas.The experimental results show that WAP clustering algorithm is lower than AP clustering algorithm in terms of similarity between sample points and clustering time,and the clustering effect is better.The research has a good practical significance for improving the indoor positioning accuracy of wireless sensor coverage.