Sensor networks with limited nodes are deployed randomly, and thus there exist a number of monitoring holes which lead to incomplete sensor data sets. Aiming at the problem, this paper prposes a Kriging interpolation algorithm for sensor data based on irregular grid. An irregular grid meshing algorithm is designed to adapt to the random nodes deployment. A fast search algorithm is proposed to sift the neighboring nodes of the interpolation points. Kriging matrix is solved and fast Kriging data interpolation is obtained. Experimental results based on Intel-Berkeley data set show that this algorithm has higher accuracy.%传感器网络节点数量的有限性和部署的随机性使其监测区域存在测量空洞,导致传感数据集不完整.为此,提出一种基于不规则网格的传感数据Kriging插值算法,对监测区域做不规则划分,以适应节点随机部署的特性.通过近点邻域搜索算法,确定待插值点的邻居节点,并据此求解Kriging矩阵,实现快速插值.基于英特尔-伯克利传感数据集的实验结果表明,该算法具有较高的插值精度.
展开▼