<p>Despite significant improvements in cloud security, existing solutions are insufficient to protect resources against malicious threats, which often suffer from high false positive rates. Thus, the given paper introduces a novel intrusion detection and prevention system (IDPS) that combines signature-based and anomaly-based techniques. The proposed approach uses a newly generated dataset from a real-world network environment using Modern Honey Network (MHN) and Snort, which can be preprocessed using a dimensionality reduction feature. Thus, to determine and improve the overall performance of the proposed framework, the pre-processed data is subsequently fed into an integrated approach with the help of WOA-XGBoost (whale optimization-Extreme Gradient boost) algorithm. By integrating real-time traffic analysis with hyperparameter optimization, our approach achieves superior accuracy compared to conventional techniques with high sensitivity and reduces false positive rates. This research addresses critical gaps in intrusion detection and prevention systems by effectively mitigating evolving threats in modern cloud environments.</p>