Deep reinforcement learning for QoS-driven cloud healthcare services selection: A framework and performance evaluation

Abstract

In recent years, the smart city concept has become popular due to its ability to improve the quality of life for urban residents. Smart community, smart transportation, and smart healthcare are among the several fields the idea covers. Integrating cloud computing technology into the healthcare industry has revolutionized healthcare delivery, enabling efficient data storage, analysis, and remote access to critical medical resources. However, choosing high-quality healthcare services from many cloud service providers remains challenging. This study presents the Quality of Service-driven Cloud Healthcare Services Selection (QCHSS) framework, underpinned by deep reinforcement learning, to tackle the intricate challenge of optimizing cloud-based healthcare services. QCHSS prioritizes Quality of Service (QoS) criteria, elevating patient experiences and outcomes. Leveraging Deep Reinforcement Learning (DRL), particularly the Deep Q-network (DQN) technique, we intelligently select cloud healthcare services, resulting in substantial improvements in availability, reliability, energy efficiency, and throughput. This research not only advances cloud-based healthcare service selection but also underscores the transformative potential of DRL in complex decision-making processes, offering a significant contribution to the field and enhancing healthcare service quality.