aspg-3-2758-1791417849 20261008000409 American Scientific Publishing Group admin@americaspg.com American Scientific Publishing Group Fusion: Practice and Applications FPA 2770-0070 2692-4048 2024 16 1 Hitesh Kumar Sharma Research Scholar, Amity University, Gwalior, M.P, India Samta Jain Goyal Associate Professor, Amity University, Gwalior, M.P, India Sumit Kumar Assistant Professor, G.N.S University, Sasaram, Bihar, India Abhishek Kumar Research Scholar, Amity University, Gwalior, M.P, India As IoT devices increase, accuracy and data security become increasingly crucial. This research recommends a powerful threat detection system that accelerates message responses to improve IoT security. The recommended strategy finds dangers in using many data sources. Our deep learning system is DenseNet. It groups photographs nicely. We show how the approach works using real-world experiments. It has few false positives and negatives and is effective at recognizing items. Through ablation research, we examine how design and component selections impact technique performance. This clarifies the method's fundamentals. The research reveals that feature selection, fusion, and DenseNet design improve the technique. We discuss the need for fine-tuning hyperparameters to improve approaches and monitor more individuals. The strategy makes IoT communities safer and more robust by laying the groundwork for threat detection and response. This approach solves message transmission delay concerns, making the IoT safer. These discoveries may benefit hacking specialists. They improve and speed up IoT security. 2024 101 117 2758 https://creativecommons.org/licenses/by/4.0/ 10.54216/FPA.160108 https://www.americaspg.com/journal/3/article/2758 https://www.americaspg.com/storage/61714834987.pdf