Energy-Efficient Deep Learning Models for Cyber Threat Prediction in Constrained IoT Infrastructure

As the Internet of Things (IoT) continues to grow, the need for efficient and effective cybersecurity solutions becomes increasingly critical. This paper presents a novel approach to cyber threat prediction in constrained IoT environments through the application of energy-efficient deep learning models. Given the limited computational resources and energy constraints of many IoT devices, traditional deep learning models often fail to deliver the required performance while maintaining energy efficiency. To address this, we propose a hybrid deep learning architecture that balances accuracy with energy consumption by leveraging lightweight neural network models and energy-aware training techniques. Our model demonstrates promising results in terms of threat detection accuracy while significantly reducing the computational burden on resource-constrained IoT devices. Through extensive experiments, we show that the proposed solution outperforms existing models in both prediction accuracy and energy efficiency, offering a viable solution for secure IoT networks without compromising system performance. This research contributes to the growing field of sustainable AI, aiming to enhance cybersecurity in IoT ecosystems while minimizing the environmental and operational impact of deep learning technologies.