Optimization of the Load Balancing in the Edge Servers for Mobile Edge Computing using Deep Learning Algorithms

Mobile Edge Computing (MEC) addresses limitations in cloud-based services for mobile networks by utilizing Edge servers that offload tasks from mobile devices to nearby servers. However, inefficient utilization of resources in these servers can lead to performance issues. To address this, a Task Offloading based Load Balancer employing Reinforcement Learning (RL) algorithms is proposed. This project introduces modules for network generation, RL algorithm initialization, task offloading simulation, and performance analysis. Through simulations and graphical representations, the effectiveness of load balancing techniques is demonstrated. The RL-based load balancing system optimizes server selection based on parameters such as queue size, processing capacity, and distance, achieving improved task offloading and reduced delays. The results showcase the efficacy of employing RL algorithms for load balancing in MEC, enabling equitable distribution of tasks among Edge servers, enhancing system performance, and ensuring efficient utilization of resources.

 Keywords: Mobile Edge Computing (MEC), Load Balancing Techniques,  Resource Efficiency, Reinforcement Learning (RL).