Font Size: a A A

Research On Network Attack Detection Technology Based On Abnormal Traffic Analysis

Posted on:2019-01-03Degree:MasterType:Thesis
Country:ChinaCandidate:B LiFull Text:PDF
GTID:2348330569988274Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the rapid development of internet technology,the harm caused by cyber attacks is becoming more and more serious.Before the outbreak of attacks,how to detect and identify attack quickly,accurately,and comprehensively has important practical significance for ensuring the safe operation of information systems and reducing economic losses.Based on the discovery and analysis of abnormal traffic,the technology of network attack detection is researched in order to further improve the real-time and accuracy of detection.The specific work is as follows:(1)Anomaly traffic detection model based on network traffic prediction is proposed.Understanding network traffic trends and abnormal network traffic in advance is a key step in reducing the probability of network attacks.Firstly,the phase space of network traffic time series is reconstructed,and the reconstructed traffic sequence is input into the model.Then,the Particle Swarm Optimization(PSO)algorithm is used to optimize the initial parameters of Elman neural network,and the trained Elman neural network is considered to predict network traffic.Finally,an abnormal traffic detection method based on K Nearest Neighbor(KNN)cumulative distance is used to detect the abnormal flow of the predicted traffic.The improvement of the prediction accuracy makes the detection of abnormal traffic timely.(2)A network attack detection model based on abnormal traffic is proposed.The abnormal flow detected by this method is used as an input of the model.Firstly,the Extreme Learning Machine(ELM)algorithm is used to project the linearly inseparable samples in the low-dimensional input space into the high-dimensional feature space in order to make it linearly separable.Then the KNN algorithm is used to classify the samples projected into the high-dimensional feature space,and the intrusion detection classifier is established to identify the type of network attack.Experimental results show that compared with other attack detection methods,the network attack detection model based on abnormal traffic analysis proposed in this paper improves the accuracy of intrusion detection,can effectively identify network attacks type and ensure the security of information systems.
Keywords/Search Tags:network attack detection, abnormal traffic detection, time series, network traffic prediction, classification problem
PDF Full Text Request
Related items