Integrating System Dynamics into Predictive Analytics for Dynamic Mobile Network Capacity Planning

Abstract

This work assesses how System Dynamics (SD) complements Predictive analytics for dynamic mobile network capacity planning. Including SD within predictive analytics models improves responsiveness to the challenges posed by dynamic environments by including stocks and flows, feedback and time delays, and therefore model accuracy. Even with advances in machine learning, traditional predictive analytics still struggle with the dynamic feedback mechanisms as much. This research seeks to enhance mobile network forecasting and capacity planning using a hybrid approach employing real-time data. While computational complexity is a challenge, this integrated approach considerably improves network performance and planning accuracy.