Wireless Network Intrusion Detection System Based on Back Propagation Neural Network Improved by Genetic Algorithms

摘要

In order to improve the detection rate of intruders in wireless network, and to solve the problem that the back propagate neural network (BPNN) is invalid when these initial weight and threshold values of BPNN are chosen impertinently, Genetic Algorithms (GA)'s characteristic of getting whole optimization value was combined with BPNN's characteristic of getting local precision value with gradient method. After getting an approximation of whole optimization value of weight and threshold values of BPNN by GA, the approximation was used as first parameter of BPNN, to train (educate) the BPNN again (in other words, learning). The educated BPNN was used to recognize intruders in wireless network. Experiment results shown that this method was useful and applicable, and the detection right rate of intruders was above 95% for the KDD CUP 1999 data.