Fig. 2

BC-HyIDS phase 1: signature based detection architecture. Phase 1 architecture from proposed system is elaborated in figure. It shows input, output and processing steps for signature based attack detection. Classifiers used where artificial neural network, XGBoost and isolation random forest. All classifier are trained using CIC-IDS 2017 dataset. This is a standard intrusion detection dataset used for modern attack detection. Ensemble approach using majority voting algorithm is used to get final prediction as attack or normal packet