【摘要】 针对WKNN算法中未知节点的定位中邻近参考点参数固定、定位不灵活且误差较大的问题，提出了基于RSSI的加权近邻改进算法。首先对RSSI值进行高斯滤波处理，通过FCM聚类确定未知节点所属类别，采用隶属度阈值对聚类结果进行修正。然后根据FCM的聚类子样本数设定WKNN算法的近邻值，实现了WKNN算法的自适应计算。实验结果表明，该方法的定位精度明显好于KNN和WKNN的定位算法，其平均误差不超过0.36 m。更多还原

【Abstract】 To solve problems including fixed K adjacent reference points, inflexibility and large errors while locating unknown nodes in the weighted k-nearest neighbor(WKNN) algorithm, a method based on weighted nearest neighbor improvement algorithm of received signal strength indication(RSSI) was proposed. First, the RSSI value was proceeded by Gaussian filtering. The category of unknown nodes was identified by FCM clustering. The clustering results were corrected by membership threshold.Then, the nearest neighbor value of WKNN algorithm was set according to the number of clustering subsamples of FCM, which realized the adaptive calculation of WKNN. Experimental results show that the positioning accuracy of this method is significantly better than KNN and WKNN positioning algorithm,and the average error does not exceed 0.36 m.更多还原