On the basis of analysis on structure model of wireless sensor networks（WSNs）

a novel 3D localization algorithm is proposed based on support vector machine（SVM）.Taking advantage of SVM can classify on nonlinear sample in high-dimensional space

the feature vector is estimated by distance between each node and anchor node.Finally

the cubic space of unknown nodes is classified to realize location of unknown nodes.The experimental results indicate that the algorithm achieves average localization correct rate 88 % and is insusceptible to anchor nodes density and range error

so it approves that this algorithm is practical.