Abnormal Traffic Detection Based on Generative Adversarial Network and Feature Optimization Selection

Abstract

Complex and multidimensional network traffic features have potential redundancy. When traditional detection methods are used for training samples, the detection accuracy of the supervised classification model is affected due to small data samples. Therefore, a method based on generative adversarial networks (GANs) and feature optimization is proposed. First, the feature correlation and redundancy are analyzed by the potential redundancy of network traffic. The feature optimization selection method of collaborative learning automata is proposed. Second, the confrontation interactive training principle of the generative confrontation network is adapted, in which a model of the generative confrontation network is proposed to solve the problem that small training label samples. Third, the interdomain distance is minimized by using GAN and the multiple kernel variant of maximum mean discrepancy (MK-MMD). The shared features between the source domain and target domain distribution are learned by applying the information between GAN confrontation training and classification network supervision training, improving the detection accuracy. Forth, random noise data and original training label samples are mixed to form a new training set. The accuracy is further improved by adopting generative models to continuously generate samples. The final classification results are output by the 16-dimensional Softmax classifier. The method has a small loss rate when the datasets are used to train by the experimental analysis of algorithm parameters and simulation data. The model optimized by MK-MMD has strong generalization ability. The average detection accuracy rates are 91.673% (two-classification) and 91.480% (multiclassification) by comparing machine learning and other shallow neural networks, and are the highest values among the compared methods. Moreover, the effectiveness and superiority of the proposed method are verified to be the best by comparing the recall rate, false positive rate (FPR), F-measure, AUC. When the interference of other samples are mixed, the proposed method is also robust.

Effective Internet of Things botnet classification by data upsampling using generative adversarial network and scale fused bidirectional long short term memory attention model

“…Shaikh et al 37 have used GAN based models to detect threats to IoT devices from within and outside the network of interest. Ma et al 38 have proposed multiple kernel variant of maximum mean discrepancy (MK‐MMD) in combination with GAN to reduce inter‐domain distance. The information among argument training and classification network supervision training of GAN is optimized.…”

COVID-19 and Global Increases in Cybersecurity Attacks: Review of Possible Adverse Artificial Intelligence Attacks

“…ML has been combined with other automation approaches, such as neural fuzzing in cybersecurity research for advertise attack findings [48]. According to the National Institute of Standards and Tech-nology, a neural network approach, known as generative adversarial networks (GAN), has recently been connected to deep fakes and false data duplication [49].Two neural networks (i.e. generative and discriminative networks) are used in GAN to replicate content features, analyse those features and improve the realism of how the machine represents those characteristics over time through a training process.…”

Small object intelligent detection method based on adaptive recursive feature pyramid

“…To solve these problems, the researchers optimize the small object detection method based on various optimization strategies, such as data enhancement [14] , [15] , [16] , [17] , [18] , multi-scale learning [19] , [20] , [21] , [22] , context learning [23] , [24] , [25] , [26] , [27] , and generative confrontation learning [28] , [29] , [30] , [31] , [32] , [33] , [34] , which are analyzed as follows:…”