Along with the fast development of Internet technology, attack technology of networkhackers is changing, and there are lots of new attacks every day. The Network IntrusionDetection System is a new type of network security system, which has a good detection effecton Network protection. However, most detect systems base on misuse detection other thananomaly detection because of real-time ability.In this paper, an intrusion detection method based on fuzzy synthetic discriminationalgorithm is proposed, a new method of calculating degree of membership is introduced, andthe system model with implementation is supplied. The particle swarm optimization (PSO)algorithm and K-means clustering to fuzz element are employed to ensure justifiability ofmembership functions. The Maximum Eigenvalue is used to calculate subjective weight, whichdetermined the weight of factors with objective weight together to reduce the effect of single weight torecognition ability. Results show this scheme have a good effect on improving detectionperformance and real-time ability. |