6G Network Evolution — Features Studied in 3GPP

Discover the features shaping 6G — what 3GPP has studied or is currently studying. Each feature below maps to a 3GPP study item (Technical Report) or to normative work (Technical Specification).

3GPP structures 6G across two releases:

• Release 20 — the bridge/pre-6G release. It completes 5G-Advanced and launches the first 6G study items (Technical Reports, TRs). Some adjacent features are already normative (Technical Specifications, TSs) in Rel-20.
• Release 21 — the first normative 6G release (first 6G TSs). Rel-21 scope is planned, not yet frozen; no Rel-21 6G TSs exist in the corpus at this snapshot date.

Each feature lists: what it is, why it matters, the technical approach, and the associated TR/TS or study/work item where known.

Status key: Study = Technical Report / feasibility study. Normative = Technical Specification with approved content. Planned = Rel-21 scope stated by 3GPP but not yet specified.

Part 1 — RAN (Radio Access Network)

1. 6G Radio Scenarios and Requirements (channel modelling)

• Spec / item: TR 38.914 · work item FS_6G_RAN · Study (in progress)
• What it is: The RAN-level study of propagation scenarios, channel models, and radio-layer KPIs needed to design the 6G air interface — the RAN counterpart to the SA1 requirements study (TR 22.870).
• Why it matters: Accurate channel models for the new bands are the prerequisite for all subsequent physical-layer and AI-native link design; without them nothing downstream can be calibrated.
• Technical approach: Defines deployment scenarios (dense indoor hotspot, outdoor macro, sidelink/D2D, NTN-integrated, sub-THz short range); extends channel models to sub-THz (100–300 GHz) and the upper mid-band (7–24 GHz); sets KPIs such as Tbps-class peak downlink, high edge rates, very low BLER targets, and ~1 cm positioning. Evaluates waveform candidates for high-mobility / high-Doppler conditions.

2. Sub-THz and Upper Mid-Band New Radio (spectrum)

• Spec / item: under TR 38.914 / 6G radio study · work item FS_6G_NR_spectrum · Study
• What it is: Study of new 6G candidate spectrum — sub-THz (100–300 GHz, the D-/G-band region) and the upper mid-band (7–24 GHz, "FR3").
• Why it matters: 5G mmWave stops around 52.6 GHz. Sub-THz unlocks Tbps peak rates over short ranges (indoor hotspots, backhaul); FR3 balances coverage and capacity for wide-area 6G.
• Technical approach: Sub-THz needs near-field beamforming, reconfigurable intelligent surfaces, and new antenna structures because path loss and Doppler behaviour differ from mmWave. FR3 is studied for >1 GHz carrier bandwidth with beam management traded between FR1 and FR2 characteristics. Both feed the 6G band definitions.

3. AI-Native Radio (AI for PHY and L2)

• Spec / item: part of FS_6G_RAN, building on Rel-19 AI/ML for NR (TR 38.843) · work item FS_6G_AI_NR · Study (transitioning from 5G add-on to 6G native)
• What it is: Where 5G treated AI/ML as an add-on for specific functions (CSI feedback, beam management, positioning), 6G studies making AI-native the default design principle across PHY (channel estimation, decoding), MAC (scheduling), and RRC (mobility).
• Why it matters: AI-native radio adapts to fast-changing environments without manual parameter tuning, approaching optimal performance across heterogeneous scenarios.
• Technical approach: Neural-network-based channel coding candidates (e.g. autoencoders for specific channels), learned waveforms, and over-the-air training signalling; a framework for the network to provision/version AI models to UEs with trust handling; and an "AI split" placing functions at the UE, gNB-DU, gNB-CU, or core per latency/compute trade-off.

4. Integrated Sensing and Communication (ISAC) — radio aspects

• Spec / item: radio design under the 6G RAN study; requirements TS 22.137, architecture TS 23.137 · work item ISAC_REQ / Sensing · Requirements normative, radio under study
• What it is: The same network serves double duty — carrying communication traffic while also performing radar-like measurement, reusing its own transmissions to locate, follow, and identify objects nearby.
• Why it matters: A defining 6G capability — automotive perception, object/UAV tracking, gesture recognition, cm-level positioning, and environmental monitoring, all without dedicated sensor infrastructure.
• Technical approach: Sensing KPIs (range, velocity, angular resolution), multi-target tracking, and co-existence with communication QoS; monostatic/bistatic sensing geometries; interference management between sensing and data transmission.

5. Reconfigurable Intelligent Surfaces (RIS)

• Spec / item: TR 38.858 (Rel-18/19 foundation), extended for 6G · work item FS_RIS · Rel-18 study complete; 6G normative work planned
• What it is: Engineered surfaces whose electromagnetic response can be tuned so that incident radio waves are steered toward a chosen direction, redirecting signals without an active transmit chain.
• Why it matters: Useful for closing coverage gaps at sub-THz frequencies, where propagation loss is high, and for broad indoor/outdoor coverage at lower energy cost than deploying more active base stations.
• Technical approach: A control interface on the network side to set the surface's phase-shift pattern, channel-estimation methods that take the surface into account, and treatment of the RIS as a new network entity with management and orchestration hooks.

6. Network Energy Efficiency / Green RAN

• Spec / item: TS 28.310 (energy-efficiency management), scenarios in TR 38.914 · work item FS_6G_NR + energy management · 5G normative; 6G extensions under study
• What it is: Study of radio-node energy consumption at architectural and physical-layer levels — base-station sleep modes, dynamic spectrum sharing, and inter-node coordination to reduce carbon footprint.
• Why it matters: 6G treats a large energy-efficiency-per-bit improvement as a design constraint as fundamental as peak data rate.
• Technical approach: Extended DTX/DRX including base-station deep sleep; AI-driven traffic prediction with pre-emptive sleep scheduling; energy-proportional hardware in the TR 38.914 scenarios; OAM energy KPIs and reporting (TS 28.310, TS 28.552).

Part 2 — SA (Service & System Aspects / Architecture)

7. 6G Use Cases and Service Requirements

• Spec / item: TR 22.870 · work item FS_6G_Requirements (SA1) · Study (TR reported complete)
• What it is: The foundational SA1 requirements study. It collects use cases from verticals (industry, public safety, health, immersive media, space), derives KPIs, and identifies gaps versus 5G service requirements (TS 22.261).
• Why it matters: The later architecture and radio work is shaped by what this study concludes. It sets the headline targets — Tbps-class peak rates, sub-millisecond latency, non-terrestrial coverage, and very high device density (order 10⁷ devices/km²).
• Technical approach: Use-case workshops → candidate KPI tables → gap analysis against TS 22.261 → new normative requirements, covering holographic communication, tactile/haptic services, extreme-reliability industrial control, and space-segment connectivity.

8. Mobile Metaverse Services

• Spec / item: TS 22.156 · work item Mobile Metaverse (SA1) · Normative (active in Rel-20)
• What it is: Service requirements for immersive, multi-user 3D shared environments delivered over the mobile network — avatars, shared spatial experiences, haptics, and holographic rendering.
• Why it matters: A key 6G use-case driver, requiring order-of-magnitude improvements in latency, bandwidth, and multi-user synchronisation over 5G.
• Technical approach: Stage 1 requirements for very low end-to-end latency (haptics and visual rendering), high per-user data rates, synchronisation across UEs in a shared scene, split rendering (compute in the network), and privacy for avatar/identity data.

9. Integrated Sensing and Communication (ISAC) — architecture

• Spec / item: TS 22.137 (Stage 1), TS 23.137 (Stage 2), SA6 enablement TS 23.255 · Requirements normative; architecture active
• What it is: The system-level framework that turns sensing into a network service.
• Why it matters: Supplies the architecture and APIs through which applications can request sensing and receive the results, exposing sensing as a network service.
• Technical approach: A Sensing Service Function interfacing with the core, a sensor-data management framework, and procedures for network-controlled versus UE-controlled sensing; SA6 adds application-layer sensing enablement via SEAL (TS 23.255).

10. Ambient IoT / Zero-Energy Devices (Phase 2)

• Spec / item: TS 22.369, TS 23.369, TS 23.370, radio TS 38.191 / TS 38.291 · work item AmbientIoT_Ph2 · Normative (multiple Rel-20 CRs approved)
• What it is: Battery-free devices powered by harvested energy (from radio signals, light, or motion) that talk over NR, with the network serving as both their energy source and their interrogator.
• Why it matters: Enables extreme device density by removing the battery barrier for billions of sensors, tags, and actuators (logistics, healthcare, smart environments).
• Technical approach: Device classes DO (Duty-Cycled Only) and DO-A (DO with Active transmission) at sub-mW power; a gNB-Reader/interrogator architecture (TS 23.369); AIoT NAS (TS 24.369) and radio PHY (TS 38.191 / TS 38.291); temporary identifiers (TS 23.003) and OAM support (TS 28.541).

11. Energy Services Phase 2 (network as an energy actor)

• Spec / item: TR 22.883 (study) → TS 22.261 (normative output) · work item FS_EnergyServ_Ph2 / EnergyServ_Ph2-REQ · Study complete → normative requirements
• What it is: The network as an active participant in energy grids — reporting energy consumption of UEs and nodes, responding to grid demand signals, enabling demand-response.
• Why it matters: Aligns the network with sustainability goals and supports grid flexibility services, complementing the per-bit energy-efficiency targets.
• Technical approach: New energy-related KPIs in TS 22.261; requirements to expose energy metrics to authorised third parties; per-session/application energy-footprint reporting.

12. Next-Generation Core Architecture (6G System)

• Spec / item: 6G Core architecture study (TR to be assigned) · work item FS_6G_Arch (SA2) · Study (starting in the Rel-20 timeframe)
• What it is: The SA2 study of the architecture for the 6G System — what evolves from or replaces the 5G Core's Service-Based Architecture (SBA).
• Why it matters: The 6G core must handle Tbps flows, sub-ms services, AI-native network functions, and seamless integration of terrestrial, non-terrestrial, and private networks.
• Technical approach: Disaggregated core NFs with finer-grained splitting; native compute integration (the network scheduling/migrating compute alongside data sessions); evolved network slicing from soft to hard resource partitions; and study of semantic communication (compressing at the level of meaning, not just bits).

13. Network Digital Twin (NDT)

• Spec / item: TS 28.561 (management of network digital twins) · work item NDT-OAM (SA5) · Normative (active in Rel-20)
• What it is: A live digital copy of the physical network, kept current from ongoing measurements and used to optimise, anticipate faults, and test changes before they are applied.
• Why it matters: 6G networks are too complex for human-in-the-loop management; NDTs enable autonomous, zero-touch operation and underpin intent-driven management.
• Technical approach: A management data model for the twin (data sources, synchronisation interfaces, lifecycle); integration with the closed-control-loop framework; and analytics fed from the core's data analytics function.

14. AI/ML Network Architecture (NWDAF evolution)

• Spec / item: TS 23.288 (NWDAF), TS 23.482 (AIML enablement), SA6 SEAL AIML TS 24.560 · work item AIML_Ph2-APP / eMDAS_Ph4 · Normative (Phase 2 CRs approved in Rel-20)
• What it is: Evolution of the 5G data-analytics function (NWDAF) toward an AI-native core, where AI/ML models are first-class network functions rather than analytical overlays.
• Why it matters: 6G defines AI as native infrastructure; the network must train, deploy, update, and retire models at scale for radio resource management, traffic prediction, and security.
• Technical approach: Distributed model training across multiple NWDAFs and a model repository/marketplace role; standardised AIML enablement APIs for vertical applications (TS 23.482); application-layer AIML service protocols (TS 24.560); and planned model lifecycle management (data provenance, versioning, explainability hooks).

Part 3 — CT (Core Network Protocols)

15. Ambient IoT CT Protocols

• Spec / item: TS 24.369 (AIoT NAS), TS 29.369, TS 29.506 · work item AmbientIoT_Ph2-CT · Normative (Rel-20 CRs approved)
• What it is: CT1/CT4 define a new, stripped-down NAS variant (AIoT NAS) and service-layer APIs for ambient IoT devices that cannot perform normal 5G registration.
• Why it matters: Zero-energy devices need a radically lightweight protocol stack; standard 5G NAS is too heavyweight for battery-less operation.
• Technical approach: A minimal, event-driven "check-in" NAS state machine (no persistent registration); AIoT temporary identifiers (TS 23.003); SBI data-management services (TS 29.369); and UDR extensions for AIoT device data (TS 29.506).

16. Combined QoS Configuration and Monitoring (cross-layer end-to-end QoS)

• Spec / item: TS 22.261 (requirements) with SA2/CT follow-on · work item CombQoS · Requirements normative; architecture/CT work follows
• What it is: A bidirectional QoS API letting authorised third parties provide end-to-end QoS parameters to the network and receive QoS monitoring reports back.
• Why it matters: 6G services (remote surgery, industrial control, holographic streaming) need the application to co-manage network QoS dynamically, not rely on static SLAs.
• Technical approach: New TS 22.261 clauses allowing a third party to inject end-to-end QoS parameters and to receive QoS reports; CT follow-on via Network Exposure Function (NEF) API extensions conveying third-party QoS intent to the policy (PCF) and session-management (SMF) functions.

Reference Table — 3GPP Specs / Studies