Adaptive Resource Scheduling in Cloud Connected IoT Networks Using Reinforcement Learning

The rapid expansion of Internet of Things (IoT) deployments has led to highly dynamic, heterogeneous, and resource constrained network environments. Cloud connected IoT systems rely on efficient resource scheduling to ensure low latency, high throughput, and energy efficiency while meeting diverse Quality of Service (QoS) requirements. Traditional static and heuristic based scheduling approaches struggle to adapt to fluctuating workloads, varying network conditions, and unpredictable device behaviors. This paper proposes an adaptive resource scheduling framework for cloud connected IoT networks using reinforcement learning (RL). The proposed approach enables intelligent decision making by continuously learning optimal scheduling policies based on real time system feedback. By modeling the scheduling problem as a Markov Decision Process (MDP), the RL agent dynamically allocates computation, communication, and storage resources across cloud and edge layers. Simulation based evaluation demonstrates that the proposed framework significantly improves system responsiveness, resource utilization, and energy efficiency compared to conventional scheduling methods. The results confirm the effectiveness of reinforcement learning in enabling scalable, autonomous, and adaptive resource management for next-generation cloud IoT ecosystems.