Multi-Connectivity Solutions for LTE and NR Networks
Dual connectivity, NR, Machine Learning.
Dual connectivity (DC) technology (or, more generally, multiple connectivity (MC)) is an important feature in the initial journey of New Radio (NR), the access network of 3GPP to 5G, being the basis for the first ways of implementing this generation of communication systems. With MC, users can be simultaneously connected to legacy technologies (4G-LTE and Wi-Fi) and the new 5G-NR technology. However, such technology poses additional challenges for the network, especially with regard to energy management and the numerous connectivity configuration options. This work aims to conceive strategies based on machine learning capable of efficiently exploring the performance of MC towards 5G eMBB (enhanced Mobile Broadband) and 5G URLLC (Ultra Reliable Low Latency Communications) in heterogeneous networks. The idea is to dynamically configure the best set of MC parameters to provide increased data throughput (eMBB target) and a more robust connection (URLLC target), being relevant in scenarios with the presence of small cells using mmWaves, for example, both considered essential to meet 5G specifications.