Performance Evaluation of Extended Access Barring for Narrowband IoT

Abstract

Narrowband Internet of Things (NB-IoT) has been introduced as a new narrowband radio technology which provides low-power wide-area IoT in release 13 of 3rd Generation Partnership Project (3GPP). It is expected that NB-IoT supports more than 50,000 devices in a cell. As the number of machine devices including those in the extreme coverage increases continuously, network congestion among the machine devices seems inevitable in future. In release 11 of 3GPP, Extended Access Barring (EAB) was introduced to resolve the congestion in the random access procedure. EAB controls the access behavior of devices by dividing them with different classes. As a consequence, the performance of EAB for NB-IoT becomes the interests of this thesis. In this thesis, we first survey the key features of NB-IoT to support machine type communications. Then, through extensive simulations on the random access procedure of NB-IoT, we investigate the system performance in terms of resource utilization, access delay, and throughput. As one of the key observation, it is shown that current NB-IoT system can support up to 120,000 devices if 12 preambles are assigned for random access procedure, and up to 250,000 devices if 48 preambles are assigned.