Research On Security Situation Of DDoS Attack Assessment Method Based On Machine Learning

In the environment of big data,the scale of Distributed Denial of Service(DDoS)attack continues to expand rapidly,and its harm to network security is becoming more and more serious.In order to conduct a deep research on security situation assessment method,this paper introduces the basic principles and characteristics often different types of DDoS attack,and illustrates the importance and urgency of security situation assessment for DDoS through comprehensive analysis of various DDoS attacks.The traditional network situation assessment methods cannot effectively evaluate the DDoS attack security situation in the new network environment.In this paper,a DDoS attack security situation assessment model based on improved Fuzzy C-means(FCM)clustering has been proposed.This model generates a fusion feature according to network flow changes in IP address of old and new users and unilateral and bilateral network flow,and calculates the risk index of each network node on the basis of fusion feature and obtains the security situation information of the whole network by fusing the risk indexes of all network nodes,and clusters the fusion situation information with improved FCM into five security levels,so as to quantitatively evaluate the DDoS attack security situation of the whole network through the proposed situation risk degree recognition model.Experiments on real DDoS data show that the proposed model can assess the DDoS attack security situation reasonably and effectively and be more flexible and accurate than existing methods.In this paper,eight deep learning methods are analyzed and studied.The security situation assessment method based on deep learning has been proposed and the security situation assessment model based on neural network has been built.This model uses Convolution Neural Network(CNN),Recurrent Neural Network(RNN),Back Propagation algorithm(BP)and Long Short-Term Memory(LSTM)to train each network security index in depth,so as to realize the comprehensive evaluation of network security situation.Experiment results show that the proposed model can evaluate the network security situation more intuitively and effectively.