Configured Grant Small Data Transmission

Description

Configured Grant Small Data Transmission (CG-SDT) is a feature introduced in 3GPP Release 17 for New Radio (NR) that enables User Equipment (UE) to transmit small data payloads in the uplink using pre-configured radio resources without requiring dynamic grant scheduling from the gNB. It operates within the RRC_INACTIVE state, allowing the UE to remain in this low-power state while performing data transmissions, thereby avoiding the transition to RRC_CONNECTED state and its associated signaling overhead. The gNB provides the UE with a configured grant configuration via RRC signaling, which includes parameters such as time-domain resources (periodicity, offset), frequency-domain resources (Resource Blocks), modulation and coding scheme (MCS), and power control settings. This configuration is stored by the UE and can be used repeatedly for multiple transmissions according to the defined periodicity.

Architecturally, CG-SDT integrates with the existing NR uplink framework and leverages the Configured Grant Type 1 mechanism, where resources are fully configured via RRC and activated without Layer 1 signaling. The UE autonomously selects an appropriate configured grant resource from its stored configuration based on data arrival and transmits using the Physical Uplink Shared Channel (PUSCH). Key components include the RRC protocol for configuration delivery, MAC layer for handling logical channel prioritization and multiplexing within the grant, and PHY layer for the actual transmission. The gNB monitors the pre-configured resources for UE transmissions and uses the configured grant's associated HARQ process for potential retransmissions, though retransmissions may require fallback to dynamic grant procedures in some scenarios.

CG-SDT's operation involves the UE evaluating its data buffer against the configured grant's parameters, such as the maximum transport block size, to determine suitability. If the data fits and timing aligns, the UE formats the MAC PDU, applies the configured MCS, and transmits. The gNB, aware of the configuration, decodes the transmission. This process eliminates the need for Scheduling Request (SR), Buffer Status Report (BSR), and UL grant dynamic scheduling, significantly reducing latency for small data bursts. It is particularly designed for sporadic, low-volume traffic patterns typical in IoT, wearables, and sensor applications, where connection establishment overhead would be disproportionate to payload size.

In the broader network context, CG-SDT enhances radio resource utilization by minimizing control plane signaling, which frees up resources for other users and reduces network congestion. It supports network efficiency by allowing more devices to be served with limited signaling capacity. The feature is part of 3GPP's broader Small Data Transmission (SDT) framework, which includes other methods like RA-SDT (Random Access based SDT), providing flexibility based on deployment scenarios and UE capabilities. CG-SDT's design ensures coexistence with other NR functionalities and maintains reliability through configured HARQ processes and potential fallback mechanisms.

Purpose & Motivation

CG-SDT was developed to address the inefficiencies of traditional connection-oriented data transmission for small, intermittent data packets in 5G networks. Prior to its introduction, UEs needing to send data, even minimal amounts, had to perform a full transition from RRC_INACTIVE to RRC_CONNECTED state, involving random access, RRC resume procedures, and dynamic scheduling requests. This process incurred significant signaling overhead, latency, and power consumption, which was disproportionate for applications sending only a few bytes periodically, such as IoT sensors, smart meters, or wearable health monitors. The limitations of previous approaches hindered scalability for massive IoT deployments and degraded user experience for latency-sensitive small data applications.

The motivation for CG-SDT stems from the growing demand for massive Machine-Type Communication (mMTC) and ultra-reliable low-latency communication (URLLC) use cases in 5G, where efficient handling of small data is critical. 3GPP identified that existing mechanisms were optimized for large, continuous data flows but not for sporadic small packets. CG-SDT solves this by enabling 'always-on' data transmission capability without the connection setup penalty, aligning with 5G goals of network efficiency, low latency, and energy savings. It reduces control plane load on the gNB, allowing more devices to be supported simultaneously, which is essential for dense IoT environments.

Historically, LTE introduced similar concepts like Semi-Persistent Scheduling (SPS) for voice over LTE, but CG-SDT extends this principle specifically for the RRC_INACTIVE state in NR, leveraging the enhanced inactive state introduced in 5G. It addresses the specific challenge of small data transmission in NR's more complex frame structure and wider bandwidths. By pre-configuring resources, CG-SDT minimizes air interface signaling, reduces device energy consumption by avoiding state transitions, and lowers latency to meet stringent requirements for industrial IoT and real-time monitoring applications, thereby fulfilling a key gap in 5G's capability portfolio.

Key Features

• Pre-configured uplink resources without dynamic scheduling
• Operation in RRC_INACTIVE state to avoid connection setup
• Reduced latency and signaling overhead for small packets
• Support for periodic and sporadic traffic patterns
• Autonomous UE transmission based on stored configuration
• Integration with HARQ processes for reliability

Evolution Across Releases

Defining Specifications