Intelligent Subcarrier Allocation in Hybrid Beamforming Multi-User mMIMO-OFDM Systems

This paper proposes a genetic-algorithm (GA)-based subcarrier allocation in orthogonal frequency division multiplexing (OFDM)-based hybrid beamforming multi-user massive multiple-input multiple-output (MU-mMIMO) systems. Our goal is to maximize the system sum-rate capacity under the total transmit power constraint through optimally selecting K
max
users out of K available users to be served over each sub-carrier. Considering the energy-efficient hybrid beamforming architecture deployed at the base station (BS), the non-convex optimization problem is solved in four steps: (i) designing a radio frequency (RF) beamformer using slow time-varying angle-of-departure (AoD) information of users to generate the beams for all subcarriers, (ii) designing a baseband (BB) precoder for each subcarrier using the corresponding low-dimensional effective channel state information (CSI) seen from the BB stage based on regularized zero-forcing (RZF) technique, (iii) optimizing subcarrier allocation using GA with equal power allocation (EQ-PA) among users (iv) performing GA-based power allocation over each subcarrier to further improve the system sum-rate. Illustrative results indicate that the proposed algorithm performs significantly better than the random and greedy subcarrier allocation schemes in terms of the achieved sum-rate.