Research of Hierarchical Intrusion Detection Model Based on Discrete Cellular Neural Networks
Kang Xie · Journal of Information and Computational Science · 2013
In this paper, a new method of hierarchical intrusion detection algorithm based on the Discrete Cellular Neural Networks (HDCNN) is put forward to solve the problems of low accuracy and slow speed in the existing intrusion detection algorithm. In order to obtain the template parameters for the HDCNN classifier, we use energy function constraint method to construct a new particle swarm optimization fitness function, jumping out the premature convergence. Emerging evidence has indicated that this new approach is affordable to parallelism and analog VLSI implementation. Experimental results and comparative studies based on the KDD cup 99 data sets are given, show that the proposed model exhibits an excellent performance owing to the higher attack detection rate and shorter processing time.