Description
Application specific Congestion control for Data Communication (ACDC) is a standardized 3GPP feature designed to provide granular, application-aware congestion management at the network access level. Unlike traditional congestion control methods that apply blanket restrictions to all traffic, ACDC operates by categorizing applications into distinct ACDC categories and applying selective access control barring (ACB) mechanisms based on these categories during periods of network congestion. The core principle is to allow the network to prioritize access attempts for critical or operator-designated applications while temporarily restricting or delaying access attempts from less critical applications, thereby alleviating signaling and data plane congestion and preserving service availability for high-priority use cases.
The architecture of ACDC involves coordination between the core network (CN) and the user equipment (UE). The network, specifically the Access and Mobility Management Function (AMF) in 5GC or the Mobility Management Entity (MME) in EPS, is responsible for defining and broadcasting ACDC-specific barring parameters via system information blocks (SIBs). These parameters, defined per ACDC category, include a barring factor and a barring time. The UE, which is provisioned with an ACDC category for each of its applications (typically via the USIM or device management), reads these broadcast parameters. When an application triggers a service request (e.g., for a mobile-originated data session or signaling connection establishment), the UE evaluates the request against the barring parameters for that application's assigned ACDC category. This evaluation involves drawing a random number and comparing it to the broadcast barring factor; if access is barred, the UE enters a timer-based wait state before retrying, effectively throttling access attempts from applications in congested categories.
Key components of the ACDC system include the ACDC category, a numeric identifier (e.g., 1-16) assigned to an application; the network-defined barring parameters for each category; and the UE's ACDC capability and rule enforcement logic. The UE's behavior is specified in protocols such as TS 24.301 (NAS) and TS 36.331 (RRC), ensuring standardized interaction with the radio access network (E-UTRAN, NG-RAN). ACDC does not directly manage data flow or quality of service (QoS) once a session is established; its role is exclusively focused on the initial access control phase. It acts as a preventive, signaling-layer filter to mitigate congestion storms that can be caused by a massive number of UEs simultaneously attempting to access the network for non-urgent services, such as social media updates or background syncs, during events that strain network capacity.
Purpose & Motivation
ACDC was created to address the critical problem of signaling and access congestion in cellular networks, particularly during mass events, emergencies, or network failures. Prior to ACDC, congestion control mechanisms like Access Class Barring (ACB) and Extended Access Barring (EAB) were relatively blunt instruments. ACB could bar entire classes of UEs (e.g., based on subscriber type), and EAB could target UEs configured for low access priority, but neither could distinguish between different *applications* running on the same device. This limitation became acute with the proliferation of smartphones and background applications, where a single UE could generate numerous access attempts for non-essential services, contributing to network overload and potentially blocking access for critical communications like emergency alerts or mission-critical operator services.
The historical context for ACDC's introduction in 3GPP Release 13 was the growing need for smarter, more application-aware network management. Operators required a tool to ensure network resilience and service continuity for high-value applications (e.g., voice, IoT alarm systems, public safety communications) even under extreme load. ACDC solves this by shifting the congestion management paradigm from a device-centric or subscriber-class-centric model to an application-centric model. It allows an operator to define policies that, for instance, permit a banking application or a smart meter reporting service to access the network while temporarily barring a video streaming app during a congestion event. This targeted approach improves overall network efficiency, enhances the user experience for priority services, and provides operators with a powerful policy tool for traffic steering and congestion mitigation, forming a foundational element for service differentiation and network slicing preparations.
Key Features
- Application-aware access control based on configured ACDC categories
- Network-configurable barring parameters (barring factor, barring time) broadcast per ACDC category
- UE-based enforcement where the device applies barring rules per application request
- Backward compatibility and coexistence with legacy barring mechanisms like ACB and EAB
- Support for both EPS and 5GS network architectures
- Provisioning of ACDC categories via USIM (EF_ACDC) or device management
Evolution Across Releases
Introduced the core ACDC framework for LTE (EPS). Defined the concept of ACDC categories, network broadcasting of ACDC barring parameters via SIBs, and UE behavior for applying category-specific barring during RRC connection establishment for mobile-originated calls and data sessions. Initial support was focused on mitigating signaling congestion.
Enhanced ACDC by extending its applicability to delay-tolerant access and introducing refinements for better integration with other core network functions. Clarified interactions with NAS signaling and improved the provisioning mechanisms for ACDC category assignment to applications.
Brought ACDC into the 5G System (5GS) architecture. Adapted the ACDC mechanisms to work with the 5G Core (5GC) and the NG-RAN, ensuring consistent application-specific congestion control across 4G and 5G. Defined support in 5G NAS (TS 24.501) and NR RRC (TS 38.331).
Introduced enhancements for ACDC in the context of network slicing and edge computing. Further optimized the barring parameters and UE procedures to support more dynamic and granular congestion control scenarios, aligning with 5G-Advanced service requirements.
Continued evolution within the 5G-Advanced framework, focusing on performance optimizations and potential integration with AI/ML-based network management for predictive congestion control. Refined signaling efficiency for ACDC parameter updates.
Further work on enhancing ADC for new 5G use cases, including tighter integration with service-based architecture (SBA) and policy control. Explored optimizations for reduced capability (RedCap) devices and non-terrestrial networks (NTN) scenarios.
Defining Specifications
| Specification | Title |
|---|---|
| TS 22.011 | 3GPP TS 22.011 |
| TS 22.818 | 3GPP TS 22.818 |
| TS 24.105 | 3GPP TS 24.105 |
| TS 24.301 | 3GPP TS 24.301 |
| TS 25.300 | 3GPP TS 25.300 |
| TS 25.304 | 3GPP TS 25.304 |
| TS 25.331 | 3GPP TS 25.331 |
| TS 26.938 | 3GPP TS 26.938 |
| TS 31.102 | 3GPP TR 31.102 |
| TS 36.304 | 3GPP TR 36.304 |
| TS 36.306 | 3GPP TR 36.306 |
| TS 36.331 | 3GPP TR 36.331 |