Enhancing IoT cyber security with hybrid deep learning: a novel approach for malware detection and source code piracy prevention using adaptive tensor flow and IPSO

The Internet of Things (IoT) has become a cornerstone of modern infrastructure, connecting devices, software, and applications. However, this connectivity brings significant cybersecurity challenges, such as malware threats and software piracy. This study proposes a novel hybrid deep learning approach to address these challenges, integrating Dynamic TensorFlow Deep Neural Network with Improved Particle Swarm Optimization. The proposed system excels in detecting malware and identifying source code duplication, utilizing advanced preprocessing techniques like tokenization and feature weighting to enhance detection accuracy. Furthermore, the Enhanced Long Short-Term Memory network is employed to analyze network traffic, identifying anomalies indicative of software piracy. Experimental results, using datasets from Maling (malware detection) and Googode Jam (source code piracy detection), demonstrate that the proposed hybrid method achieves superior performance, with 95% accuracy and a significant reduction in false alarms. This research highlights the merits of combining deep learning with optimization techniques to provide scalable, robust, and efficient cybersecurity solutions for IoT ecosystems. This research highlights the transformative potential of integrating hybrid deep learning and optimization techniques to tackle IoT cybersecurity challenges. The proposed approach not only achieves superior detection accuracy and reduced false alarms compared to existing methods but also addresses the dual threats of malware and source code piracy within a unified framework. These advancements mark a significant contribution to the development of scalable, efficient, and robust cybersecurity solutions, reinforcing the protection of IoT ecosystems and ensuring the resilience of critical digital infrastructures.