Inferensys

Glossary

Resource Block Grid

The two-dimensional time-frequency lattice structure in OFDM systems, consisting of resource elements defined by subcarriers in the frequency domain and OFDM symbols in the time domain.
Developer building agentic RAG system, retrieval pipeline diagram on laptop, technical workspace with notes.
TIME-FREQUENCY LATTICE

What is Resource Block Grid?

The resource block grid is the fundamental two-dimensional time-frequency lattice structure in OFDM systems, defining the allocation of physical resources for data transmission.

A resource block grid is the two-dimensional time-frequency lattice that constitutes the physical transmission structure in OFDM-based systems such as LTE and 5G NR. It is formed by resource elements (REs) — the smallest allocable unit defined by a single OFDM symbol in the time domain and a single subcarrier in the frequency domain. The grid organizes these elements into resource blocks (RBs), which serve as the minimum scheduling unit for user data allocation.

In 5G NR, the grid architecture is parameterized by the OFDM numerology, where subcarrier spacing and cyclic prefix length scale the time-frequency dimensions. A standard resource block spans 12 consecutive subcarriers in frequency and one slot in time, with the number of OFDM symbols per slot varying by configuration. This structured lattice enables orthogonal multiple access, where different users are assigned distinct, non-interfering regions of the grid, and provides the reference framework for embedding demodulation reference signals (DMRS) essential for coherent channel estimation.

TIME-FREQUENCY STRUCTURE

Key Characteristics of the Resource Block Grid

The resource block grid is the fundamental two-dimensional lattice that organizes OFDM transmissions. It defines the atomic unit of allocation—the resource element—and the scheduling granularity—the resource block—across both time and frequency domains.

01

Resource Element: The Atomic Unit

A Resource Element (RE) is the smallest physical resource unit in the OFDM grid, occupying one subcarrier in the frequency domain and one OFDM symbol in the time domain. Each RE carries a single complex-valued modulation symbol (QPSK, 16QAM, 64QAM, or 256QAM).

  • In LTE, a single RE with 15 kHz subcarrier spacing and normal CP occupies approximately 66.7 µs in time
  • The total number of REs in a 20 MHz LTE carrier with normal CP is 84,000 per subframe
  • REs are grouped into larger structures for scheduling: resource blocks, resource block groups, and control channel elements
1 subcarrier × 1 symbol
RE Dimensions
84,000
REs per LTE Subframe (20 MHz)
02

Resource Block: The Scheduling Granularity

A Resource Block (RB) is the minimum unit of resource allocation that the MAC scheduler can assign to a user. In LTE, an RB spans 12 consecutive subcarriers (180 kHz) in frequency and one slot (0.5 ms, 7 OFDM symbols with normal CP) in time, containing 84 REs.

  • 5G NR introduces flexible RB definitions that scale with numerology, but the 12-subcarrier frequency span remains constant
  • The total number of RBs depends on the channel bandwidth: 6 RBs for 1.4 MHz, 100 RBs for 20 MHz LTE
  • RBs can be allocated as contiguous blocks or distributed across the bandwidth using resource allocation types 0, 1, or 2
12 subcarriers × 1 slot
RB Dimensions (LTE)
180 kHz
RB Bandwidth
03

Time-Frequency Lattice Structure

The resource block grid forms a two-dimensional lattice where the horizontal axis represents OFDM symbols (time) and the vertical axis represents subcarriers (frequency). This orthogonal structure ensures that subcarriers do not interfere with each other despite overlapping in frequency.

  • The grid dimensions for a 10 ms LTE radio frame with 20 MHz bandwidth: 100 RBs × 140 OFDM symbols (normal CP)
  • Reference signals, synchronization signals, and control channels occupy specific, predefined positions within this lattice
  • The lattice structure enables localized and distributed resource allocation schemes for frequency diversity
100 × 140
Grid Dimensions (20 MHz LTE Frame)
10 ms
Frame Duration
04

Reference Signal Mapping

Specific REs within the resource block grid are reserved for reference signals (pilots) that enable channel estimation and coherent demodulation. These REs are transmitted at known power and phase, allowing the receiver to estimate the channel response.

  • Cell-specific Reference Signals (CRS) in LTE occupy REs at fixed subcarrier and symbol offsets, with density depending on the number of antenna ports
  • Demodulation Reference Signals (DMRS) in 5G NR are UE-specific and mapped to the scheduled RBs, supporting beamforming and MU-MIMO
  • Reference signal overhead typically consumes 4.76% to 14.29% of REs per RB, depending on antenna configuration
4.76% – 14.29%
Reference Signal Overhead per RB
Up to 8
Antenna Ports (LTE CRS)
05

Resource Element Groups and Control Mapping

REs are aggregated into Resource Element Groups (REGs) for control channel mapping. A REG consists of 4 consecutive REs within the same OFDM symbol that are not occupied by reference signals.

  • Control Channel Elements (CCEs) are formed from 9 REGs (36 REs) and are the basic unit for PDCCH transmission
  • PDCCH aggregation levels (1, 2, 4, or 8 CCEs) determine the coding rate and coverage of downlink control information
  • The control region occupies the first 1–3 OFDM symbols of each subframe in LTE, while 5G NR uses the flexible CORESET structure
4 REs
REG Size
9 REGs (36 REs)
CCE Size
06

Numerology-Dependent Grid Scaling

In 5G NR, the resource block grid scales with the numerology (µ), which defines the subcarrier spacing as 15 × 2^µ kHz. Larger subcarrier spacings compress the time-domain symbol duration while maintaining the 12-subcarrier RB width.

  • µ=0 (15 kHz): 1 ms slot, 14 symbols — baseline for sub-6 GHz
  • µ=1 (30 kHz): 0.5 ms slot — common for mid-band deployments
  • µ=2 (60 kHz): 0.25 ms slot — enables low-latency URLLC
  • µ=3 (120 kHz): 0.125 ms slot — used in mmWave bands above 24 GHz
  • The frequency-domain RB count scales inversely with subcarrier spacing for a given bandwidth
15 × 2^µ kHz
Subcarrier Spacing Formula
µ = 0, 1, 2, 3, 4
5G NR Numerologies
RESOURCE BLOCK GRID

Frequently Asked Questions

Essential questions about the two-dimensional time-frequency lattice structure that forms the fundamental scheduling unit in OFDM-based systems like LTE and 5G NR.

A Resource Block Grid is the two-dimensional time-frequency lattice structure that organizes physical resources in OFDM-based systems such as LTE and 5G NR. It consists of Resource Elements (REs) —the smallest physical resource unit, representing one subcarrier in the frequency domain and one OFDM symbol in the time domain. Resource Elements are grouped into Resource Blocks (RBs) , which form the minimum scheduling granularity. In LTE, one Resource Block spans 12 subcarriers (180 kHz) in frequency and 7 OFDM symbols (one 0.5 ms slot) in time for normal cyclic prefix, containing 84 Resource Elements. In 5G NR, the Resource Block definition remains 12 subcarriers in frequency but the time-domain extent varies with the configured numerology, allowing flexible slot durations. The grid provides a structured coordinate system where the scheduler allocates specific Resource Blocks to user equipment for data and control transmission.

Key structural elements:

  • Resource Element (k, l): Indexed by subcarrier k and OFDM symbol l, carrying one complex-valued modulation symbol
  • Resource Block: 12 consecutive subcarriers × 1 slot, the atomic scheduling unit
  • Bandwidth Part (BWP): A contiguous subset of Resource Blocks configured per UE in 5G NR
  • Control Resource Set (CORESET): Specific Resource Block regions reserved for downlink control channels

The grid enables orthogonal multiple access by assigning non-overlapping Resource Block allocations to different users, eliminating intra-cell interference while maximizing spectral efficiency through dynamic resource partitioning.

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.