Study on the impact of clustering for non-orthogonal multiple access based on multilevel code for radio-over-fiber fronthaul application

With the stringent requirements of fifth-generation and beyond wireless communications technology coupled with the potential bottleneck in the optical fronthaul, radio-over-fiber (RoF) technology offers an alternative solution to distribute the wireless signals while meeting these requirements. The inherent centralized architecture of RoF enables advanced coordination functionality to be performed within the fronthaul network. Advanced coordination functionality such as non-orthogonal multiple access (NOMA) has gained traction as a means to improve spectral efficiency for larger user-base networks. Among various NOMA implementations, the recently proposed multilevel code (MLC) was found to be superior to the commonly investigated superposition coding successive interference cancellation scheme in a two-user scenario. In this paper, we generalize the definition of the MLC to $n$n user cases and discuss two power allocation schemes for the generalized MLC definition, which are based on the spectral efficiency (SE) criterion and symbol error rate (SER) criterion, respectively. Simulation-based performance analysis on the MLC-SE and the MLC-SER schemes was conducted over different cluster sizes and randomly generated user distributions. The results show, to the best of our knowledge, that the best performance in terms of user fairness and bit error rate can be obtained in a two-user MLC-SER scheme. (c) 2021 Optical Society of America

Automated summary

This research generalizes the definition of multilevel code (MLC) for $n$n user cases and discusses two power allocation schemes based on the spectral efficiency (SE) criterion and symbol error rate (SER) criterion. Simulation-based performance analysis shows that the best user fairness and bit error rate performance can be obtained in a two-user MLC-SER scheme.