Random Access Sliced 5G-Advanced Network for Machine Type Communication

Massive machine-type communication (mMTC) has been highlighted as one of the major services that enable the development of the Internet of Things (IoT) paradigm with varying Quality of Service (QoS) requirements for 5G and beyond networks. However, it is extremely challenging to use a monolithic physical network to handle several mMTC applications with varying QoS needs due to the limitations experienced during the random access (RA) procedure, which leads to collisions and network overload. While 5G-Advanced promises better network performance and more connections through the introduction of machine learning, it also requires more access resources for the RA process. The introduction of network slicing to the access network can ameliorate these issues, leading to improvements in collision resolution and network congestion management. Conse-quently, to tackle the afore-mentioned challenges, we propose a network-slicing random access scheme (NSRAS) that combines network slicing and machine learning algorithms to dynamically vary the physical random access resources into multiple virtual resources to reduce collisions during the RA procedure while meeting the QoS requirements of the different types of machine-type communication devices (MTCDs). Our proposed scheme shows improvements in the average access delay and the outage proportion for MTCDs.