5G Network Slicing using Machine Learning Techniques

Abstract

Communication systems to be delivered with the Fifth Generation (Fifth Generation, 5G) are expected to meet the requirements of high reliability, low delay, high security, high capacity, and high-speed. Mobile providers are looking for programmable solutions to provide numerous different services, and the 5G network structure provides a solution to this need using Network Slicing. In this study, artificial intelligence-based machine learning algorithms and methods of placing users in communication slices were examined by creating realistic user and base station data. Considered communication slices were selected as advanced mobile network (enhanced Mobile Broadband, eMBB), large-scale machine-type communication (mMTC), and ultra-low-latency data communication (Ultra-Reliable Low Latency Communications, URLLC). Two different machine learning models were created and tested in the proposed simulation environment, and their performances were compared.