Service Function Chain Anomaly Detection Based on Distributed Generative Adversarial Network in Network Slicing Scenario br

For the problem of Service Function Chain (SFC) anomalies due to hardware and software anomaliesin network slicing scenarios, a Distributed Generative Adversarial Network (GAN)-based Time Series anomalydetection model (DTSGAN) is proposed. First, to learn the characteristics of normal data in SFC, a distributedGAN architecture is proposed for anomaly detection of multiple Virtual Network Functions (VNFs) containedin SFC. Then, a feature extractor based on sliding window data is constructed for time series data, and thefeature sequence is obtained by extracting two derived characteristics and eight statistical features of the datato mine the deep-level features. Finally, in order to learn and reconstruct data characteristics, a three-layercodec constructed by Time Convolutional Network (TCN) and Auto-Encoder (AE) is proposed as a distributedgenerator, which measures the difference between reconstructed data and input data by anomaly score functionto detect the state of VNF, and then completes the anomaly detection of SFC. The effectiveness and stability ofthe proposed model are verified on the dataset Clearwater using four evaluation metrics: accuracy, precision,recall and F1 score