Securing the Cloud : Leveraging Deep Learning and ORCA Optimization for Enhanced IDS

Although cloud computing has completely changed how data is stored and accessed, there are still ongoing security risks. Due to their limited scalability and adaptability in dynamic cloud environments, traditional intrusion detection systems (IDS) and machine learning-based techniques frequently fail. This study suggests a Network Intrusion Detection System (NIDS) based on a Region-based Convolutional Neural Network (RCNN) optimized with the ORCA metaheuristic algorithm in order to address these problems. The RCNN is appropriate for detecting subtle intrusion patterns because it is made to model intricate spatial relationships in structured data. To improve detection performance, ORCA is used to adjust hyperparameters like learning rate, batch size, and dropout rate. Metrics like accuracy, precision, recall, and F1-score are used to assess the suggested model on a cloud-based IoT network dataset. The results show that the RCNN-ORCA model achieves lower false alarm rates and higher detection accuracy than conventional classifiers like SVM and CNN. For cloud-based intrusion detection, this method addresses issues of imbalance, scalability, and dynamic threat adaptation while demonstrating the efficacy of fusing deep learning with nature-inspired optimization.