本文提出了一种基于稀疏化最小二乘支持向量机的WSN节点区域定位算法。算法将WSN室内定位问题看作以传感器节点RSSI值为特征量的多分类问题,将节点RSSI值转化为节点位置,利用SVC良好的泛化能力和稀疏化的特点,实现区域定位,以达到较高的定位精度。实验结果表明,与SVC的区域定位比较,该算法有较好的定位精度与时效性。 In this paper, we propose a WSN node locating algorithm based on thinning least squares support vector machines. The algorithm considers the WSN indoor positioning problem as a multi-classification problem characterized by RSSI value of the sensor nodes, and transforms the node RSSI values ​​into node locations. By using the SVC’s good generalization ability and thinning characteristics, the algorithm realizes regional positioning to achieve better performance High positioning accuracy. Experimental results show that this algorithm has better positioning accuracy and timeliness than SVC.