Network Intrusion Detection Algorithm Based on Multi-level and Parameters of the Fusion Feature
LI Qingnian · Bulletin of Science and Technology · 2012
The traditional network intrusion detection algorithm,the use of all are the same or similar characteristics as the intrusion detection criterion,but in a multi-layered network,the different levels of the characteristics and differences between the parameters,the test result error is bigger.Therefore put forward based on the parameters of features of the integration of network intrusion detection algorithm.Through the extraction multilayer network operating difference characteristics and parameters by using nonlinear regression way for each layer,characteristic parameters of difference compensation,according to the adaptive way fusion of multilayer difference characteristics the fusion processing,as the basis of a test.Experiments show that the algorithm enhances the accuracy of inspection,to make the ideal effect.