A Convex Optimization-Based Beamforming Framework for Spectral Efficiency Enhancement in 5G Heterogeneous Massive MIMO Systems

Abstract

This paper presents a convex optimization-based beamforming framework for enhancing spectral efficiency (SE) in heterogeneous massive multiple-input multiple-output (mMIMO) 5G networks. Although universal frequency reuse improves spectral utilization, it introduces significant multi-cell interference that degrades system performance. To address this challenge, coordinated uplink (UL) and downlink (DL) beamforming schemes are developed to maximize the weighted sum SE under practical constraints. The resulting optimization problem is inherently non-convex and NP-hard; however, by reformulating the signal-to-interference-plus-noise ratio (SINR) and co-channel interference constraints, a tractable convex formulation is obtained. The proposed framework is efficiently solved using CVX, ensuring global optimality. Simulation results demonstrate that the proposed method achieves a minimum SE of 30 bit/s/Hz in the downlink and significantly outperforms benchmark techniques under practical conditions (Kr ≥ 4, N ≥ 12, SNR = 30 dB). Furthermore, increasing the number of users and base station antennas yields substantial additional SE gains, highlighting the scalability of the proposed approach for dense 5G deployments. Simulation results demonstrate that the proposed method achieves a minimum spectral efficiency of 30 bit/s/Hz in the downlink, outperforming existing benchmark schemes under practical conditions (Kr ≥ 4, N ≥ 12, SNR = 30 dB). Furthermore, increasing the number of users and base station antennas leads to significant additional SE gains, highlighting the scalability of the proposed approach for dense 5G deployments.