5G NR Spectral Efficiency Calculator

Defining Spectral Efficiency in Modern Cellular Networks

Radio frequency spectrum is one of the most commercially valuable and physically constrained resources on Earth. In telecommunications engineering, spectral efficiency (SE) quantifies how densely information can be packed into an allocated frequency band. Expressed in bits per second per Hertz (bps/Hz), spectral efficiency measures the net data transmission rate normalized against the consumed spectrum:

In 3GPP 5G New Radio (NR) systems, radio access network engineers distinguish between two fundamental definitions of spectral efficiency:

• Nominal Channel Spectral Efficiency ($\text{SE}_{\text{nominal}}$): The total user plane throughput divided by the nominal licensed carrier bandwidth ($BW_{\text{channel}}$, e.g., 20 MHz, 40 MHz, or 100 MHz). This metric reflects the operator's commercial return on spectrum investment, penalizing guard bands and duplex gaps.
• Effective Transmission Spectral Efficiency ($\text{SE}_{\text{effective}}$): The throughput divided strictly by the occupied active subcarrier transmission bandwidth ($BW_{\text{trans}} = N_{\text{PRB}} \times 12 \times \Delta f$). Because 5G NR leaves symmetric guard bands at the channel edges to comply with adjacent channel leakage ratio (ACLR) specifications, $BW_{\text{trans}}$ is typically 95% to 98% of $BW_{\text{channel}}$ (e.g., 98.28 MHz for a 100 MHz C-Band carrier with 273 PRBs at 30 kHz subcarrier spacing).

The Shannon-Hartley Theorem & Physical Layer Bounds

The absolute physical upper boundary for spectral efficiency across an additive white Gaussian noise (AWGN) channel is governed by the Shannon-Hartley theorem. Formulated by Claude Shannon in 1948, the channel capacity $C$ in bits per second is expressed as:

Here, $\text{SINR}$ is the linear Signal-to-Interference-plus-Noise Ratio ($\text{SINR}_{\text{linear}} = 10^{\text{SINR}_{\text{dB}} / 10}$). For example, an operating SINR of +24.0 dB corresponds to a linear ratio of 251.19, establishing an unconstrained theoretical SISO limit of:

In practical 3GPP NR deployments, actual spectral efficiency operates approximately 1.5 to 3.0 dB below the Shannon bound (known as the Shannon Gap). This implementation delta arises from:

1. Constellation Shaping & Quantization: Practical transceivers transmit discrete, non-Gaussian discrete constellation points (QPSK, 16-QAM, 64-QAM, 256-QAM, 1024-QAM). Higher modulation orders suffer from increasing constellation shaping loss.
2. Forward Error Correction (FEC) Blocklength Penalties: 3GPP NR employs Low-Density Parity-Check (LDPC) codes for the user plane (PDSCH/PUSCH). While LDPC approaches within fractions of a decibel of the turbo/polar limits, finite code block sizes incur residual parity overhead.
3. Air-Interface Overhead ($OH$): Physical channels require Demodulation Reference Signals (DMRS), Channel State Information Reference Signals (CSI-RS), Primary/Secondary Synchronization Signals (PSS/SSS), Physical Downlink Control Channel (PDCCH) allocations, and Cyclic Prefix (CP) guard durations. Per 3GPP TS 38.214 Section 4.1.2, this overhead typically consumes 14% of resources in Frequency Range 1 (FR1) Downlink and 18% in Frequency Range 2 (FR2).

Multi-Antenna Spatial Multiplexing Scaling (MIMO Mechanics)

To surpass the SISO Shannon boundary without increasing allocated RF bandwidth, modern cellular systems employ Multiple-Input Multiple-Output (MIMO) spatial multiplexing. By transmitting multiple independent data streams over rich multipath propagation environments, MIMO scales spectral efficiency quasi-linearly with the number of orthogonal spatial layers ($\nu$):

where $\lambda_i$ denotes the eigenvalues of the instantaneous spatial channel matrix $\mathbf{H}\mathbf{H}^H$. Under uncorrelated high-scattering conditions:

• $2 \times 2$ Dual-Stream MIMO: Doubles raw physical capacity ($\approx 2 \times \text{SE}_{\text{SISO}}$), standard in handheld terminals.
• $4 \times 4$ Quad-Stream MIMO: Quadruples throughput ($\approx 4 \times \text{SE}_{\text{SISO}}$), representing the commercial state of the art for premium 5G smartphones on C-Band (3.5 GHz).
• $8 \times 8$ MU-MIMO / Massive MIMO: Base stations equipped with 32T32R or 64T64R Active Antenna Units (AAUs) utilize beamforming to co-schedule 8 to 16 orthogonal spatial streams across separate UEs simultaneously on the same time-frequency PRBs, boosting aggregated sector spectral efficiency to $40\text{ to }80+\text{ bps/Hz}$.

ITU-R IMT-2020 vs. 3GPP Realistic Deployment Benchmarks

The International Telecommunication Union Radiocommunication Sector (ITU-R) defined the formal technical performance criteria for 5G in Recommendation ITU-R M.2410-0 ("Requirements related to technical performance for IMT-2020 radio interface(s)"). Under Section 4.4:

• IMT-2020 Minimum Peak Downlink Spectral Efficiency: $30.0\text{ bps/Hz}$.
• IMT-2020 Minimum Peak Uplink Spectral Efficiency: $15.0\text{ bps/Hz}$.

To fulfill the ITU-R 30 bps/Hz peak requirement, 3GPP designed the NR air-interface to support 8 spatial layers with 256-QAM modulation (code rate $R = 948/1024$ yields $7.41\text{ bps/Hz/layer} \times 8 \times (1 - 0.14) \times (273/273) \approx 51\text{ bps/Hz}$ raw continuous FDD).

However, in commercial field deployments, Time Division Duplexing (TDD) duty cycles profoundly alter the measured nominal spectral efficiency:

• Under a standard 2.5 ms DDDSU frame pattern (~74.29% Downlink duty cycle), the transmitter is silent during guard periods and uplink slots. Consequently, even if the active downlink slots instantaneously achieve $21.92\text{ bps/Hz}$ across 4 layers of 256-QAM, the net time-averaged nominal spectral efficiency across the carrier is $16.28\text{ bps/Hz}$.
• This explains why drive-test tools often report lower spectral efficiency figures than laboratory bench tests: duty cycle partitioning, control channel overhead, and channel fading all act on the raw theoretical capability.

3GPP TS 38.214 MCS-to-Spectral Efficiency Reference Table

The lookup table below outlines standardized 3GPP TS 38.214 modulation orders ($Q_m$), target code rates ($R$), raw SISO spectral efficiencies, and achievable multi-layer efficiencies across FDD and TDD allocations: