Towards Intelligent Multi-Access Edge Computing Using Machine Learning

Abstract

Multi-Access Edge Computing (MEC) has been introduced as a part of the 5G architecture to reduce network latency and traffic compared to cloud computing. Applications are running in MEC centers close to the end users improving user experience. Due to limited resources of a MEC center, only selected applications can be placed there. In this paper we introduce a machine learning-based approach to optimize network traffic by intelligent selection of applications running in a MEC center. Our approach is based on current and historical network traffic data to predict the potential for network traffic savings. We propose a smart city architecture for integrating this approach and discuss its benefits and challenges. Using smart waste management as a particular use case, we illustrate the application of our solutions.