| dc.description.abstract | BACKGROUND: As the second phase of 5G standardization efforts encapsulated in Release 16 comes to its freeze and completion date in June 2020, aspects of some promised features and services started to crystallize. Among of which, New Radio (NR)-based access to unlicensed spectrum, commonly known as 5G NR-U. Current technical reports have identified Listen-Before-Talk (LBT) as a working assumption in the process of standardizing NR-U channel access scheme. LBT was originally developed for Licensed-Assisted Access (LAA) in release 13 of the 3GPP specifications, which was based on ETSI regulations. This research examines how next-generation wireless systems using LBT perform under vastly presumed 5G NR dense deployments, and how the coexistence landscape manifests in the homogeneous prospect rather than the widely investigated heterogeneous counterpart, e.g. with Wi-Fi.
METHODS: In this work, a simulator was developed in C++ to help analyze different intra-network NR-U co-channel scenarios under saturated traffic. The simulator was validated with Markov Chain analytical model to confirm the procedures and algorithms conform to the standard delineated by the 3GPP specifications.
RESULTS: Simulation results indicated inefficiency in channel utilization of homogeneous dense deployments with high priority traffic classes. For instance, the effective channel utilization drops to less than 10% when only 20 devices share the channel with traffic tagged as priority 4, e.g., voice calls. Moreover, mean delay between successful packet transmissions in aforesaid scenario turned out to be around 1 second and exponentially increasing with the number of devices sharing the channel. We demonstrated through simulations how LBT devices can be unfair when sharing the channel with others exhibiting different traffic priority classes. A video streaming device – i.e. class 3 – for example, takes away 42% of the channel when sharing it with other 7 devices browsing the internet – i.e. class 2 – leaving them with 34% of useful channel time to split. The remaining 24% of the time packets collide with each other, rendering the channel futile and reducing the overall throughput.
CONCLUSION: Literature is inundated with research on cross-technology coexistence analysis. This work aims to study same-technology wireless coexistence performance and underlines the importance of improving channel access mechanisms in next-generation wireless communication. | en_US |
