Impact of QoS and Task Processing on IoT Edge Energy Consumption

As the utilization of the Internet of Things (IoT) continues to grow, there is a persistent demand for new applications, primarily driven by considerations such as battery energy and heat management. Consequently, the development of energy-efficient solutions becomes imperative to prolong system operating times. This research proposes a solution based on optimizing the task processing location of IoT edge devices, a factor directly impacting network energy consumption. The challenges associated with task processing are intricately linked to energy consumption. As processing demands escalate and computing power and communication capabilities increase, network energy consumption rises accordingly. The proposed solution involves offloading tasks from edge end devices to a smart gateway using various methods. This approach leverages the resources available in edge devices, which are closer in proximity and possess more abundant resources compared to end devices. By enhancing computing capacity, this strategy significantly impacts energy consumption, leading to notable improvements in energy efficiency during task processing. In this study, the integration of a Markov decision model and reinforcement learning techniques is employed to optimize the selection of the processing location. The results demonstrate an average energy consumption improvement of at least $10 \%$ when choosing the appropriate processing mode. Simulation results have been thoroughly evaluated and discussed, showcasing the efficacy of the proposed approach in enhancing energy efficiency within IoT networks.