Optimizing Edge Server Deployment in 5G Networks: A Forecasting-Based Approach with AI-Driven Strategies
Shyam Sunder Gupta, Pushpendra Singh Danghi, Mayur Bhoyar, S Inayath Ahamed, Paresh Tanna, Nishma Bhaskarani
Abstract
The fast development of 5G networks requires utilizing edge servers to ensure further demands for ultra-low latency and high bandwidth. This paper introduces a forecasting-based strategy with AI adopted to enhance overall edge server deployment in 5G networks. Analyzing historical network traffic data and data of the users’ activity at the moment, complex statistical algorithms make exact forecasts of future network load and consequently, perform preparations for it beforehand. The strategy uses reinforcement learning and optimisation algorithms to identify the ideal server placement in concerning latency minimisation, cost, and load distribution. Consequently, this research proposes solutions to some of the challenges including; users’ mobility which is dynamic, the services offered and the users’ power needs. Simulation has revealed that the formulated framework enhances the following performance parameters of the network, indicating enhanced quality of service, reduced latency, and high reliability. This work continues the exploration of intelligent and scalable 5G edge networks, and facilitates potential application like autonomous car, smart glasses, and IoT.