Initial Access Codebook Design and CSI Type-II Feedback for Sub-6GHz 5G NR

Beam codebooks are a recent feature to enable high dimension multiple-input multiple-output (MIMO) in 5G new radio (NR). Codebooks comprised of customizable beamforming weights can be used to transmit reference signals and aid the channel state information (CSI) acquisition process. In this paper, we characterize the role of each codebook used during the beam management process and design a neural network to find codebooks that improve overall system performance. Evaluating a codebook is not purely about maximizing signal power, but instead, a holistic, system-level view of effective spectral efficiency is necessary to capture the relationships between codebooks, feedback, and spectral efficiency. The proposed algorithm is built on translating codebook and feedback knowledge into a consistent beamspace basis similar to a virtual channel model to generate initial access codebooks and select the subsequent refined beam training. This beamspace codebook algorithm is designed to directly integrate with current 5G beam management standards. Simulation results show that the neural network codebooks improve over traditional codebooks in received signal power, even in dispersive sub-6GHz environments. We further utilize our simulation framework to evaluate type-II CSI feedback formats with regard to effective multi-user spectral efficiency. Our results suggest that optimizing codebook performance can provide valuable spectral efficiency improvements, but 5G feedback quantization resolution limits multi-user performance in sub-6GHz bands due to the rich scattering environment.