Dynamic Resource Allocation Framework for Resilient and Secure IoT Communication Using Federated Learning and Quantum Cryptography

Abstract

 In the growing Internet of Things (IoT) landscape, secure and efficient resource allocation remains a critical challenge, especially with increasing data sensitivity and device connectivity. This study introduces a novel Dynamic Resource Allocation Framework that integrates Federated Learning (FL) and Quantum Cryptography (QC) to optimize resource distribution while bolstering security across IoT devices. Leveraging FL allows for model training directly on edge devices, ensuring data privacy with no data transfer to central servers, which resulted in a 30% decrease in latency and a 40% improvement in processing speed compared to traditional centralized methods. Simultaneously, QC secures communication channels through quantum key distribution, achieving an 85% reduction in data breach incidents over six months in real-world testing. The framework’s resource allocation algorithm adapts dynamically based on device load and data sensitivity, demonstrating a 20% improvement in resource utilization and a 15% increase in network efficiency. Experiments show that the framework also reduces power consumption by 25% on average, extending device battery life by up to 40% in low-power IoT devices. This approach not only enhances the resilience and security of IoT communications but also optimizes resource management, making it suitable for diverse applications in smart cities, industrial automation, and healthcare IoT networks.