Intent

Description

In the 3GPP context, an Intent is defined as expectations including requirements, goals, and constraints given to a 3GPP system, without specifying how to achieve them. This definition, introduced in Rel-19 and carried forward into Rel-20, captures the essence of intent-based management and networking: the expression of what is desired rather than how to accomplish it. The Intent concept is central to the evolution toward autonomous networks, where the system itself determines the optimal actions to fulfill the expressed expectations, leveraging AI/ML, policy frameworks, and closed-loop automation.

An Intent is typically expressed in a declarative form, often using a structured or natural-language-like syntax, and is consumed by an Intent-driven management system. The system translates the Intent into actionable policies, configurations, and workflows. For example, an operator might express an Intent such as 'ensure high-quality video streaming for premium users in the downtown area during peak hours.' The system would then interpret this Intent, decompose it into network-level requirements (e.g., bandwidth, latency, QoS parameters), and orchestrate the necessary resources across RAN, core, and transport domains to fulfill it. The Intent does not prescribe specific network functions, slices, or configurations; those decisions are left to the system's intelligence.

The Intent concept spans multiple domains within 3GPP. In TS 22.870, Intent is discussed in the context of service requirements and user expectations, particularly for 6G systems where users or verticals may express high-level goals that the network must satisfy. In TS 28.312, Intent is addressed from a management and orchestration perspective, defining how Intent-driven management can be realized in the 5G-Advanced and 6G management architecture. This dual treatment reflects the breadth of the Intent concept, which applies both to service-level expectations and to network management operations.

Technically, an Intent typically includes several components: the goal or expectation (what should be achieved), the context or scope (where and when it applies), the constraints (what limitations must be respected), and possibly the priority or criticality. The Intent lifecycle includes expression, translation, activation, assurance, and deactivation. During translation, the Intent is converted into a set of policies or intents at lower levels of abstraction. During assurance, the system continuously monitors whether the Intent is being satisfied and takes corrective action if deviations are detected. This closed-loop operation is a hallmark of intent-based systems and distinguishes them from traditional imperative management approaches.

The introduction of Intent in 3GPP standards is closely tied to the broader industry movement toward autonomous networks, as championed by organizations such as TM Forum and ETSI. 3GPP's adoption of Intent in Rel-19 and Rel-20 ensures that intent-based capabilities are integrated into the standardized network architecture, enabling multi-vendor interoperability and consistent implementation. As networks grow in complexity with network slicing, edge computing, and diverse service requirements, Intent provides a scalable and flexible way to manage that complexity by raising the level of abstraction at which operators and users interact with the network.

Purpose & Motivation

The Intent concept exists to simplify the management and operation of increasingly complex networks by allowing stakeholders to express what they want without needing to specify how to achieve it. Traditional network management relies on imperative instructions—explicit configurations, commands, and policies that dictate exactly how the network should behave. While effective for well-understood, static scenarios, this approach becomes unwieldy as networks grow in scale, heterogeneity, and dynamism. Operators managing 5G-Advanced and 6G networks must handle thousands of parameters, multiple domains, and rapidly changing conditions, making manual or scripted imperative management impractical.

Intent-based management addresses this by shifting the burden of decision-making from the human operator to the network system. The operator expresses high-level goals—such as 'optimize for energy efficiency' or 'guarantee ultra-reliable low-latency communication for critical services'—and the system determines the best way to achieve them. This reduces operational complexity, minimizes human error, and enables faster response to changing conditions. It also democratizes network management, allowing non-experts (e.g., vertical industry customers) to express their needs in business terms rather than technical parameters.

Prior to the standardization of Intent in 3GPP, intent-based networking was primarily a vendor-specific or open-source concept, lacking a common definition and framework. The introduction of Intent in TS 22.870 and TS 28.312 provides a standardized foundation that ensures interoperability, consistent terminology, and a clear architectural approach. This standardization is essential for multi-vendor deployments and for enabling intent-based services across different network domains and administrative boundaries. As 6G emerges, Intent is expected to become a fundamental interface between users, services, and the network, underpinning the vision of autonomous, self-managing networks.