Simulation Of Malicious Nodes Detection Based On Machine Learing For Wsn

A Machine Leafing based malicious nodes detection scheme was designed and simulated in this paper, which has modelded the characteristic behavior of Selective Forwarding, Hello Flooding and abstracted four properties of these two types of malicious nodes closely related. The nodes can be classified by using Machine Learning algorithms. The experimental results show that this method can correctly classify a certain number of sensor nodes with a lower error rate.