Adaptive Resource Management in IoT-Fog-Cloud Networks via Hybrid Machine Learning Models

Abstract

The rapid proliferation of Internet of Things (IoT) devices has intensified the need for efficient resource management in IoT-Fog-Cloud networks to ensure seamless data processing, reduced latency, and energy-efficient operations. This study introduces a hybrid machine learning-based framework for adaptive resource management tailored to IoT-Fog-Cloud networks. The proposed model integrates supervised and unsupervised learning techniques to dynamically allocate computational, storage, and network resources based on real-time workload variations and application-specific requirements. A multi-layered architecture is employed, where fog nodes handle latency-sensitive tasks, and cloud resources manage high-computation processes, ensuring an optimal balance between performance and resource utilization.The framework leverages predictive analytics to forecast workload distribution, enabling proactive resource allocation and minimizing service disruptions. Furthermore, reinforcement learning algorithms are used to optimize link stability and routing efficiency, reducing communication delays and network congestion. Simulation results demonstrate the model’s effectiveness, achieving up to a 30% reduction in latency, 20% improvement in resource utilization, and 25% enhancement in energy efficiency compared to traditional approaches. This research highlights the adaptability and scalability of hybrid machine learning models in heterogeneous IoT-Fog-Cloud environments, addressing challenges such as dynamic workload fluctuations and limited fog resource capacities. The findings underscore the potential of intelligent resource management strategies to advance IoT applications in diverse domains, including smart cities, healthcare, and industrial automation. Future work will explore real-world deployments and integration with emerging technologies like 6G and edge AI to further enhance system robustness and efficiency