Optimisation of Dynamic Request Scheduling in Mobile Edge Computing

Abstract

  The goal of this research project is to explore possible uses of deep learning algorithms for predictive task scheduling in the context of Mobile Edge Computing (MEC) to maximize dynamic request scheduling. Because there is a growing demand for low-latency applications, it is crucial to plan activities efficiently to maximize resource efficiency in MEC scenarios and improve overall system performance. We employ deep learning models to predict task arrival and user mobility patterns, respectively, to achieve the purpose of this study. This gives us the ability to plan events more precisely and adaptable than we could in the past. We will be able to significantly reduce the amount of time spent waiting for tasks and improve the system's throughput by integrating these predictive skills into the scheduling process. CloudSim is utilized in the design of the simulation environment. This enables the analysis and simulation of dynamic scheduling problems in MEC. The results show that the deep learning-based scheduling technique outperforms conventional scheduling strategies, leading to notable improvements in task latency and resource allocation efficiency. The use of this research represents a progression in the creation of scalable and intelligent scheduling techniques for MEC systems.