An Approach for Could IDS Based on Deep Learning technique to Novel Optimized ANCNNED for Secure Intrusion Detection Method

The spread adoption of cloud computing resulted from recent improvements but security concerns persist because of how cloud environments operate across multiple nodes. The analysis of heterogeneous network data through IDS in cloud networks generates high false positive rates together with low accuracy and overfitting problems. This paper introduces an optimized security framework for cloud-based intrusion detection which uses an Ensemble Optimized deep learning method.The IDS framework operates through a comprehensive process starting from data collection and proceeding through preprocessing and feature extraction and selection and ending with detection and classification followed by performance evaluation. Analysis based on DC-Res2Net extracts features before GKSO optimizes the selection process. The Novel ANCNNED approach uses an ensemble of materialized by AlexNet CNN and EfficientDet to conduct classificatory operations. The detection system employs the Enhanced Zebra Optimization algorithm for optimization purposes.