Rank Selection for 5G-NR MIMO Downlink: A Realistic SINR and Multipath-Aware Approach

In 5G New Radio (5G-NR) Multi-Input MultiOutput (MIMO) systems, selecting the optimal Rank Indicator (RI) for downlink transmission is critical to maximizing throughput and spectral efficiency while minimizing packet error rate (PER). We propose the Realistic SINR and Transport-Aware Rank Selection (RSTAR) method, which optimizes RI by integrating Transport Block Size (TBS), Inter-Layer Interference (ILI), and codebook-based precoding, unlike traditional methods relying on Shannon capacity. RSTAR employs a two-stage process, considering ILI in both stages and using realistic Signal-to-Interference-plus-Noise Ratio (SINR) derived from 5G-NR codebooks. Simulations in Indoor Hotspot (InH) and Rural Macro (RMa) scenarios show that RSTAR achieves up to 21 % spectral efficiency gain and 9% PER reduction in InH environments with rich multipath (15-19 clusters), where ILI dominates (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8. 3} \times$</tex> InterCell Interference, ICI). However, in RMa scenarios (10-11 clusters), where ICI is more significant (ILI <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$3.5 \times$</tex> ICI), RSTAR's performance degrades due to its focus on ILI. Our contributions include: (1) introducing RSTAR, a novel RI selection method, and (2) analyzing its performance across diverse multipath environments, highlighting the trade-offs of ILI-focused strategies.