5G NR & LTE MIMO Throughput & Spatial Multiplexing Calculator

Dimension multi-antenna spatial multiplexing gains, channel Rank Indicators (RI), spatial correlation cross-talk, array beamforming directivity, and Massive MIMO MU-MIMO capacity scaling across 3GPP Rel 15–18.

Quick Commercial Topology Presets
Section A: Architecture & Transmission Mode
Section B: Carrier & Spatial Channel
ρ = 0.00
Tip: Standard 4T4R SU-MIMO is 0.0 dB. Macro 32T32R/64T64R beamforming arrays provide +6 to +12 dB.
Aggregate Multiplied MIMO Throughput
1,628.4 Mbps
Effective Spatial Gain: 4.00× vs. SISO (407.1 Mbps)
High-Rank 4x4 SU-MIMO / Full Spatial Multiplexing Peak
Spatial Layer Allocation & Decorrelation Efficiency
Full Rank (Rank 4)
Aggregate Net Bit Rate
1.628 Gbps
1,628.40 Mbps total
Multi-Stream Gain Factor
4.00×
Ideal theoretical: 4.00×
Multiplexing Stream ηmimo
100.0%
0.0% correlation loss
Active Spatial Streams
ν = 4 Layers
Max topology rank: 4
Effective Spectral Efficiency
16.28 bps/Hz
Across 100 MHz channel
Beamforming Array Boost
+0.0 dB (1.00×)
SNR capacity multiplier
SISO Baseline Throughput
407.1 Mbps
Single spatial layer
Channel Matrix Condition
Optimal (κ ≈ 1.0)
Orthogonal eigenvectors
Step-by-Step Mathematical Substitution Audit
SISO Base = 407.10 Mbps | Configuration: 4x4 SU-MIMO (Rank ν = 4) | Spatial Correlation ρ = 0.00 → ηmimo = 1 - 0.00 = 1.000 (100%) | Rmimo = 4 · 407.10 · 1.000 = 1,628.40 Mbps | Spatial Multiplexing Multiplier = 1,628.40 / 407.10 = 4.00×

Multi-Antenna Physics & Spatial Multiplexing Architecture

Multiple-Input Multiple-Output (MIMO) technology represents the most fundamental spectral leap in modern mobile telecommunications. Defined extensively within 3GPP TS 38.211, TS 38.214 (5G NR), and 3GPP TS 36.211 (LTE-Advanced), multi-antenna techniques exploit the spatial dimension of the wireless propagation channel to either transmit multiple independent information streams simultaneously (spatial multiplexing) or provide robust diversity protection against destructive multipath fading (transmit diversity).

1. Spatial Multiplexing vs. Transmit Diversity (SM vs. TxDiv)

The fundamental trade-off in multi-antenna link design is governed by the Diversity-Multiplexing Trade-off (DMT) established by Zheng and Tse. Operators must dynamically balance raw bit-rate multiplication against link robustness:

Spatial Multiplexing Theoretical Upper Bound vs. Diversity Capacity
C_{\text{SM}} = \sum_{i=1}^{\nu} \log_2\left(1 + \frac{P_i \sigma_i^2}{N_0}\right) \quad \text{vs.} \quad C_{\text{TxDiv}} = \log_2\left(1 + \frac{\sum_{i=1}^{N_{\text{Tx}}} |h_i|^2 P_0}{N_{\text{Tx}} N_0}\right)

2. Channel Matrix Singular Value Decomposition (SVD) and Rank Indicator (RI)

In a physical transmission environment with $N_{\text{Tx}}$ transmit antennas and $N_{\text{Rx}}$ receive antennas, the wireless channel is characterized by an $N_{\text{Rx}} \times N_{\text{Tx}}$ complex channel impulse matrix $\mathbf{H}$. Using Singular Value Decomposition (SVD), any arbitrary channel matrix $\mathbf{H}$ can be decomposed into orthogonal spatial eigen-modes:

Singular Value Decomposition (SVD) of the MIMO Channel Matrix
\mathbf{H} = \mathbf{U} \mathbf{\Sigma} \mathbf{V}^H = \sum_{i=1}^{R} \sigma_i \mathbf{u}_i \mathbf{v}_i^H

where $\mathbf{U}$ is an $N_{\text{Rx}} \times N_{\text{Rx}}$ unitary matrix of receive spatial eigenvectors, $\mathbf{V}$ is an $N_{\text{Tx}} \times N_{\text{Tx}}$ unitary matrix of transmit beamforming precoders, and $\mathbf{\Sigma} = \text{diag}(\sigma_1, \sigma_2, \dots, \sigma_R)$ contains the non-negative singular values ordered such that $\sigma_1 \ge \sigma_2 \ge \dots \ge \sigma_R \ge 0$.

The algebraic rank $R \le \min(N_{\text{Tx}}, N_{\text{Rx}})$ determines the maximum number of decoupled, interference-free parallel channels (eigen-pipes) supported by the propagation environment. The ratio of the extreme singular values defines the channel condition number ($\kappa$):

Condition Number and Stream Decorrelation Loss
\kappa = \frac{\sigma_{\text{max}}}{\sigma_{\text{min}}} = \frac{\sigma_1}{\sigma_R}

When the environment provides rich, isotropic multipath scattering (e.g., dense urban microcells with multiple non-line-of-sight reflections), the condition number is close to 1.0 ($\kappa \approx 1$), meaning all singular values are of comparable magnitude. The User Equipment (UE) reports a Rank Indicator (RI = 4), enabling the gNodeB scheduler to allocate four full data layers.

Conversely, in environments with strong Line-of-Sight (LoS) conditions, closely spaced antennas, or wave propagation through narrow physical conduits (the notorious "Keyhole" or "Pinhole" channel), the spatial correlation factor $\rho \to 1.0$. The singular values $\sigma_2, \sigma_3, \sigma_4 \to 0$, causing $\kappa \to \infty$. The spatial matrix collapses to rank deficiency ($R = 1$), forcing the gNodeB to fall back to Rank 1 single-layer transmission despite having four physical transceivers installed.

3. Massive MIMO & Multi-User MIMO (MU-MIMO) Mechanics

In commercial Single-User MIMO (SU-MIMO), total user throughput is strictly constrained by the physical size and battery consumption of the mobile handset. Commercial smartphones are practically limited to 4 receiver antennas (4 Rx) at sub-6 GHz frequencies due to chassis clearance, hand-blocking losses, and thermal limits. Consequently, a single smartphone cannot exceed Rank 4.

5G NR macro base stations resolve this bottleneck by deploying Active Antenna Units (AAUs) equipped with 32 or 64 digital transceiver chains (32T32R or 64T64R) arranged in dense planar antenna panels (e.g., an $8 \times 8$ cross-polarized array containing 128 or 192 physical patch elements). Through Multi-User MIMO (MU-MIMO), the gNodeB breaks the 4-layer terminal ceiling by co-scheduling multiple spatially separated devices on the exact same physical time-frequency resource blocks:

4. Channel Reciprocity and Sounding Reference Signals (SRS) in TDD

In Frequency Division Duplex (FDD) networks, uplink and downlink operate on paired frequencies separated by a duplex spacing (e.g., 45 MHz or 100 MHz). Because the duplex distance exceeds the channel coherence bandwidth, fast-fading multipath characteristics are statistically independent. The gNodeB cannot infer downlink channel properties from uplink signals; instead, it relies on the UE transmitting CSI codebook feedback (RI, PMI, CQI). In a 64T64R system, quantizing a $4 \times 64$ channel matrix requires massive signaling overhead that would consume the entire uplink control bandwidth.

Massive MIMO achieves its full potential on Time Division Duplex (TDD) bands (such as 3.5 GHz n77/n78 and 2.6 GHz n41) by exploiting RF Channel Reciprocity. Because uplink and downlink share the exact same radio frequency over interleaved time slots:

TDD Radio Channel Reciprocity Theorem
\mathbf{H}_{\text{DL}}(t) = \mathbf{H}_{\text{UL}}^T(t + \Delta t) \quad \text{for } \Delta t \ll T_{\text{coherence}}

Within the channel coherence time $T_{\text{coherence}}$ (typically 5 to 20 ms depending on UE velocity), the physical propagation path is symmetric. The UE transmits periodic Sounding Reference Signals (SRS) on the uplink across the entire carrier bandwidth. The gNodeB's digital baseband receiver directly estimates the complete $64 \times 4$ channel matrix from the incoming SRS waveforms, applies internal hardware calibration coefficients to equalize transceiver filter mismatches, and computes optimal downlink precoding weights without requiring explicit CSI feedback from the smartphone.

5. Worked Engineering Case Study: C-Band 100 MHz 4x4 SU-MIMO vs. Keyhole Correlation

Consider a Tier-1 commercial 5G NR deployment operating on Band n78 (3500 MHz) with a $100\text{ MHz}$ carrier bandwidth, $30\text{ kHz}$ subcarrier spacing ($\mu = 1$, 273 PRBs), 256-QAM modulation (MCS 27, code rate 948/1024), and a standard $2.5\text{ ms}$ DDDSU TDD frame pattern (yielding an effective downlink user-plane duty cycle of $74.29\%$ and $14\%$ control overhead).

3GPP MIMO Topology & Performance Benchmark Matrix

Reference operational parameters, spatial ranks, theoretical limits, and realistic macrocell capacity multipliers defined across 3GPP LTE-Advanced and 5G NR releases:

Antenna Topology Operating Mode Max Rank (ν) Theoretical Multiplier Practical Macro Multiplier Primary Limiting Factor
1T1R (SISO) Single Spatial Stream Rank 1 1.00× 1.00× Bandwidth & Modulation Ceiling
2T2R (2x2 MIMO) SU-MIMO Spatial Multiplexing Rank 2 2.00× 1.85× – 1.95× Cross-Polarization Isolation
2T2R (TxDiv / SFBC) Space-Frequency Block Coding Rank 1 1.00× (SNR boost) 1.15× (Effective) Diversity Gain Saturation
4T4R (4x4 MIMO) SU-MIMO Spatial Multiplexing Rank 4 4.00× 3.40× – 3.80× UE Antenna Physical Separation
8T8R (8x8 MIMO) SU-MIMO (mmWave / FWA CPE) Rank 8 8.00× 6.20× – 7.10× Angular Spread & Terminal Dimensions
32T32R Massive MIMO MU-MIMO Multi-User Scheduling 8 Streams 8.00× (Cell Total) 6.50× – 7.20× Inter-User Interference Leakage
64T64R Massive MIMO MU-MIMO High-Density Grid 16 Streams 16.00× (Cell Total) 12.00× – 13.80× Pilot Contamination & Channel Inversion