Precoding Resource Block Group

Description

The Precoding Resource Block Group (PRG) is a fundamental concept in the 5G New Radio (NR) physical layer, specifically within the context of channel state information (CSI) reporting and precoding for multiple-input multiple-output (MIMO) transmissions. In the downlink, the gNB (base station) uses precoding to shape and direct transmission beams towards user equipment (UE), optimizing signal quality and throughput. The PRG defines the granularity at which the same precoding matrix can be applied across a set of contiguous physical resource blocks (PRBs) in the frequency domain. This grouping is crucial because wireless channel conditions can vary across frequency; applying a single precoding matrix over too wide a bandwidth might lead to suboptimal performance, while per-PRB precoding would incur excessive signaling overhead.

The size of a PRG is configurable by the network and can be adapted based on channel conditions, UE capability, and system bandwidth. Specifications such as 3GPP TS 38.213 and 38.214 define how PRG sizes are signaled to the UE, typically through higher-layer RRC configuration or dynamic DCI indications. The UE performs channel measurements and reports recommended precoding matrices (Precoding Matrix Indicators, PMI) for these PRG-sized chunks. The gNB then uses these reports to select and apply the appropriate precoding across the designated PRGs. This process enables efficient closed-loop MIMO operation, where the feedback overhead is managed by grouping PRBs expected to have similar channel characteristics.

Architecturally, PRG interacts with other physical layer components like the CSI-RS (Channel State Information Reference Signals) for channel estimation and the PDSCH (Physical Downlink Shared Channel) for data transmission. The choice of PRG size represents a trade-off: larger PRGs reduce CSI feedback overhead and precoding complexity but may not capture fine-grained frequency selectivity, potentially degrading performance in highly frequency-selective channels. Smaller PRGs offer more precise precoding at the cost of higher signaling. In massive MIMO (mMIMO) systems, efficient PRG configuration is vital for leveraging beamforming gains across wide bandwidths, such as in FR2 (mmWave) bands, where beam management is critical.

Purpose & Motivation

PRG was introduced in 5G NR (Release 15) to address the limitations of LTE's precoding granularity and to support the wider bandwidths and advanced MIMO schemes of 5G. In LTE, precoding was typically applied with a granularity tied to subbands or the entire system bandwidth, which could be inefficient for the larger bandwidths (up to 400 MHz in NR) and more diverse deployment scenarios of 5G. The motivation was to create a flexible mechanism that balances the trade-off between precoding accuracy and signaling overhead, enabling efficient beamforming and spatial multiplexing.

Without PRG, the network would face a dilemma: either use a single precoding matrix for the entire bandwidth, leading to performance loss in frequency-selective fading channels, or signal precoding information per PRB, resulting in prohibitive control channel overhead. PRG provides an intermediate, configurable granularity that allows the gNB to adapt to channel coherence bandwidth. This is particularly important for 5G's support of diverse use cases, from enhanced mobile broadband (eMBB) requiring high throughput to ultra-reliable low-latency communication (URLLC) needing robust link adaptation.

The creation of PRG was driven by the need to optimize massive MIMO operations, where precise beamforming is key to capacity and coverage. By grouping PRBs, the system reduces the payload size for CSI feedback (like PMI and CQI) and downlink control signaling, conserving uplink and downlink resources. This efficiency enables scalable MIMO deployments across FR1 (sub-6 GHz) and FR2 (mmWave) spectra, supporting features like multi-user MIMO (MU-MIMO) and coherent joint transmission, which are foundational to 5G's performance targets.

Key Features

• Configurable granularity for precoding application across frequency
• Reduces CSI feedback overhead by grouping PRBs with similar channel conditions
• Supports flexible adaptation to channel coherence bandwidth
• Enables efficient beamforming for wideband 5G NR transmissions
• Integrates with CSI reporting mechanisms (e.g., PMI, CQI)
• Facilitates scalable massive MIMO and MU-MIMO operations

Evolution Across Releases

Defining Specifications