Performance Evaluation of Full-Duplex IAB Multi-Cell and Multi-User Network for FR2 Band

One of the approaches to support high data rates in beyond 5G cellular networks is the dense deployment of the small cell millimeter-wave (frequency range 2 (FR2)) (>24.5 GHz) base stations. However, this dense deployment of base stations leads to high cost, because it requires fiber backhaul connection with the core network. One prominent solution is the integrated access and backhaul (IAB) network proposed in 3GPP Release 16, where some portion of wireless spectrum is used for backhaul to serve base stations instead of fiber, and the remaining portion of the spectrum is used for user equipment for communication. The partition of the spectrum into access and backhaul degrades the spectral efficiency of the IAB network. Thus, in this paper, we propose a full duplex (FD) enabled IAB network with large-scale array systems to enhance the spectral efficiency of the IAB network. In the FD IAB network, the IAB node is FD which transmits and receives at the same frequency/time resources. We consider a multi-cell multi-user IAB network with all the interferences due to FD transmission and evaluate the performance in terms of bit-error-rate and spectral efficiency. Moreover, an algorithm for choosing different users is proposed which is based on the cross-correlation of RF precoder weights. Further, identification of the optimal beam's index is also proposed which is based on the index of synchronization signal block and half-frame bit of physical broadcast channel payload of 5G NR. Finally, successive interference cancellation is proposed for self-interference mitigation in the digital domain.

Automated summary

This paper proposes a full duplex IAB network with large-scale array systems to enhance spectral efficiency, using part of the spectrum for backhaul to serve base stations and the remaining portion for user equipment communication. The algorithm for user selection is based on the cross-correlation of RF precoder weights, and an optimal beam's index identification method is proposed using 5G NR signals. Additionally, successive interference cancellation is suggested for self-interference mitigation in the digital domain.