HYBRIDNET-GUARD: An Advanced Deep Learning Framework for IoT Gateway Protocol Attack Detection with MTSS-FE and Wombel Optimization Algorithm (WBOA)

The Internet of Things (IoT) is a crucial aspect of the digital age, though at the same time, it is subject to critical issues concerning cybersecurity and data privacy. With the increase in smart infrastructure and device connectivity, it is becoming more significant to protect IoT gateways against protocol-based attacks. HybridNet-Guard is an advanced deep learning framework presented in this paper to identify such attacks with high accuracy and efficiency. The framework has a feature extraction strategy that captures temporal and spatial patterns in the data and a new hybrid optimization approach to the selection of the most relevant features. As experimental tests demonstrate, HybridNet-Guard is more effective than other models, with a high accuracy of 99.71%, a recall of 99.61% and a Matthews Correlation Coefficient (MCC) of 99.68%, a low Hamming Loss and Jaccard Scores of 0.29% and 86.18%, respectively. These findings demonstrate that HybridNet-Guard not only improves detection accuracy but also ensures scalability and adaptability for real-world IoT environments. By bridging the gap between accuracy and computational efficiency, the framework provides a practical solution for protecting IoT gateways against evolving protocol-based attacks, advancing the state-of-the-art in IoT security research.