Cloud-based deployments are increasingly vulnerable to many kinds of assaults; hence powerful anomaly detection systems are needed to prevent security breaches. While effective to some degree, current methods like RSSI, GTM, and APG have drawbacks in terms of scalability, precision, and accuracy. In order to improve the forecast accuracy of cloud assaults and optimize blockchain-based cloud installations, this study suggests a novel anomaly detection system that combines multimodal feature analysis, deep learning models, and QoS-aware sidechains. The suggested strategy considerably surpasses traditional approaches in terms of precision, accuracy, recall, and AUC performance by optimizing feature variance across various sample types and utilizing cutting-edge deep learning algorithms. Additionally, the framework shows increased efficiency in terms of throughput, energy usage, and block mining latency, makingcombining Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN), the system analyses system logs to find anomalous patterns of activity linked to various threats. The architecture guarantees the transparency and integrity of system records through the use of Blockchain technology, while Deep Learning models offer accurate and quick anomaly detection. The advantages of both Deep Learning and Blockchain technology support the choice to merge them. Blockchain technology ensures the accuracy of anomaly detection and prevents tampering with system records by providing a distributed, immutable ledger. On the other hand, deep learning models produce high precision, accuracy, recall, and AUC measures because of their remarkable pattern recognition skills and ability to adjust to shifting attack strategies. Experimental data analyses show the effectiveness of the suggested framework.
Read full abstract