Building the 6G standard: Narrowing the path to Release 21

What you should know:

• 6G's first release is becoming more concrete. September built on the 5G-based foundation already established for 6G, with progress on device requirements and physical-layer capabilities.
• 3GPP established an initial reference-device framework for 6G. Massive IoT, eMBB and FWA will have defined baseline characteristics, recognizing that different 6G products can have materially different technical requirements.
• Modulation produced two different outcomes. 1024-QAM uplink advanced for FWA, while constellation shaping will not be included in Release 21 after companies reported different evaluation results and implementation views.
• Some consequential choices remain open. Additional 5G-to-6G migration options move to December, while work also continues on 6G aggregation, the Core and 5G-Advanced satellite connectivity.

As discussed in our previous update, 3GPP has already established 5G technologies as the foundation for the 6G air interface. With that baseline increasingly clear, attention is turning to the specific enhancements, device requirements and deployment choices that will further define Release 21. At the September plenaries in Madrid, that work advanced across the first 6G reference-device types, modulation, migration from 5G to 6G, aggregation and the 6G Core, while 5G-Advanced satellite connectivity continued to evolve.

An initial set of 6G reference device types

LTE used user equipment (UE) categories to indicate different levels of device capability. 5G moved away from this approach because performance increasingly depended on an operator's carrier and bandwidth configuration. Yet practical differences in radio requirements still emerged for products such as automotive and XR devices.

Against that backdrop, 3GPP began an initial framework for three reference device types, covering massive Internet of Things (IoT), enhanced mobile broadband (eMBB) and fixed wireless access (FWA), as the initial use cases. 3GPP also agreed that each UE in 6G should be associated with a single reference device type. Each reference device type is associated with at least the following four baseline characteristics:

• the minimum number of transmit antennas or chains.
• the minimum number of receive antennas or chains.
• the smallest maximum UE bandwidth for the downlink.
• the smallest maximum UE bandwidth for the uplink.

These reference device types give the ecosystem use-case-oriented baselines for characterizing 6G products while allowing optional capabilities within each type. Future meetings may consider additional types for products such as automotive devices and wearables, as well as characteristics such as higher-order modulation support, transmit-power class or further capability segmentation.

Higher uplink modulation for fixed wireless access: raising the spectral-efficiency ceiling

Following guidance from the RAN Plenary in June and Q3 2026 working-group discussions, the September plenary considered two modulation questions for Release 21: how far to extend the highest supported modulation orders and whether to adopt modulation shaping.

In an OFDM system, information is carried across many subcarriers that act as parallel channels. Each subcarrier carries as many bits as are needed to represent a position in the underlying constellation. A larger constellation therefore allows each symbol to carry more bits under favorable channel conditions, increasing spectral efficiency and supporting higher data rates.

5G already supports up to 1024-QAM for the downlink and up to 256-QAM for the uplink. For Release 21, 3GPP RAN Plenary agreed to increase the uplink modulation order to 1024-QAM for FWA devices. The motivation is capacity.

The 5G-to-6G migration baseline, and the options still under study

Early in the 6G study, 3GPP identified 6G standalone (SA) with multi-RAT spectrum sharing (MRSS) as the default migration path under consideration from 5G to 6G. MRSS enables simultaneous operation of 5G and 6G within the same channel, so operators can serve devices from both generations within the same carrier.

Beyond that baseline, 3GPP has studied three additional 5G-to-6G migration options:

• Option 1: 6G-5G aggregation with 6G as the anchor.
• Option 2: 5G-6G aggregation with 5G as the anchor.
• Option 3: a dual stack, with the 5G radio access network (RAN) connected to the 5G Core (5GC) and the 6G RAN connected to the 6G Core (6GC).

This meeting was expected to decide whether to support any additional 5G-to-6G migration option on top of SA with MRSS. A decision was not possible because consensus was not reached, so the question has been postponed to December. The eventual choice will shape how operators reuse existing spectrum and infrastructure and how tightly early 6G deployments remain coupled to 5G.

6G-to-6G aggregation: toward a more unified approach

Migration concerns how two generations coexist. A separate question is how 6G combines its own carriers. Carrier aggregation is considered one of the most important features of 4G and 5G. It has broadly delivered the means to aggregate operators' spectrum assets, and it is considered the default aggregation technology for 6G. Carrier aggregation relies on tight synchronization and coordination across carriers, and it is typically implemented by a single vendor in collocated scenarios.

For deployments requiring non-collocated aggregation, such as combining a macro cell and a small cell, with possibly non-ideal backhaul, 5G resorted to dual connectivity (DC) techniques. The 6G study is examining 6G-to-6G aggregation across all of these scenarios, both collocated and non-collocated and with ideal or non-ideal backhaul, with the goal of supporting them through a more unified approach.

The RAN Plenary tasked RAN2, the air-interface protocol working group, to further study 6G-to-6G spectrum aggregation for non-collocated deployments with non-ideal backhaul latency of up to 10 ms, independent scheduling and multi-vendor interoperability as a design goal. A unified approach could support more flexible multi-vendor deployment across heterogeneous 6G network configurations.

Operator principles as the 6G Core Network decisions approach

The study of the 6G Core Network is expected to progress and reach some high-level decisions in Q4 2026 at SA2, the 3GPP architecture working group.

Ahead of those decisions, a multi-operator contribution in SP-260924 was presented at the meeting, laying out some key operators' considerations. This is a contribution presented by operators, not an adopted 3GPP architectural decision. It set out three operator considerations:

1. Evolution should be the starting point. Evolution serves as the baseline, with architectural change introduced where there is clear value.
2. Essential connectivity should remain interoperable and testable. AI can enhance connectivity, but connectivity should not depend on AI.
3. New 6G capabilities should be supported where justified. The aim is to support innovation without creating unnecessary complexity.

Together, these principles signal that operators want 6G innovation to remain deployable, interoperable and economically justified rather than pursuing architectural change for its own sake. SA2 will continue this work toward the high-level decisions expected in Q4 2026.

A standardized evolution path for satellite connectivity

Beyond the Release 21 work, the September plenaries also devoted considerable time to a related 5G-Advanced satellite effort, reflecting the growing importance of satellite connectivity within 3GPP. Although this is part of 5G Release 20 work rather than 6G, it demonstrates how important 5G capabilities continue to evolve even as the next generation takes shape.

3GPP has been developing standardized NR non-terrestrial networks (NR-NTN) since Release 17. At this meeting, 3GPP approved a new Release 20 work item, beginning with a study phase, focused on how standardized NR-NTN can coexist in the same cell or carrier with satellite systems that serve existing, unmodified devices. The work will explore how to continue supporting those devices while providing an evolution path toward devices designed with standardized NR-NTN capabilities.

As direct-to-device satellite deployments expand, this work could help support continuity between existing services and an evolving standardized satellite ecosystem.

Bottom line

The September plenaries show Release 21 developing as a selective evolution of proven foundations, with changes advancing where the use case, supporting evidence and implementation path are sufficiently well established. Several elements of 6G moved from broad study toward more defined technical direction, while important choices remain open where the industry has not yet reached consensus.

The December plenaries are expected to revisit several questions left open in September, including 5G-to-6G migration and the further definition of the new reference device types. More to come after the December plenaries in Boston.