5G-A to 6G Explained: From Origins to Now

Why it is suddenly everywhere

1. 01

   China’s September 2026 information-and-communications plan set a concrete bridge: 500,000 new 5G-A base stations by 2030, continued 6G trials, and commercial 6G only when conditions are ready.

2. 02

   ITU’s radio experts agreed in February 2026 on 20 draft performance requirements for IMT-2030, the formal name for 6G; final Study Group approval is expected in December.

3. 03

   3GPP has now fixed the schedule for Release 21, its first normative 6G release: package approval in March 2027 and the final protocol freeze in March 2029.

4. 04

   The trend is not evidence that a 6G phone is imminent. It marks the moment policy, live 5G-A deployment and global standard-writing moved onto one visible calendar.

The road from 5G to 6G

1. 20195G becomes commercial infrastructure

   Operators begin large-scale 5G rollouts. The public story centres on faster phones, while industrial networks, capacity and low latency develop more quietly.

2. 2023-12ITU publishes the 6G framework

   Recommendation M.2160 establishes IMT-2030 and six usage scenarios, adding AI-and-communication and integrated sensing-and-communication to familiar connectivity goals.

3. 20245G-Advanced begins

   3GPP Release 18 starts the 5G-A era: an upgrade path inside 5G rather than a new generation, designed to improve capacity, positioning, energy use and non-terrestrial links.

4. 2025The bridge reaches cities

   China reports 5G-A available in more than 300 cities. Operators test uplink-heavy applications, drone sensing and industrial automation on infrastructure that can also host 6G trials.

5. 2026-036G requirements become measurable

   ITU-R Working Party 5D agrees 20 draft technical performance requirements, seven of them new for 6G, turning a broad vision into criteria candidate radio systems can be evaluated against.

6. 2026-063GPP fixes the writing schedule

   Release 20 remains the final 5G-A-only release; Release 21 is designated the first normative 6G release, with work stretching into 2029.

7. 2026-09China publishes its build-out plan

   The new five-year plan targets 95 percent 5G/5G-A user penetration and 500,000 new 5G-A stations by 2030, while saying 6G commercialization should start at an appropriate time.

What “5G-A to 6G” actually means

The phrase sounds like a countdown to replacing every handset and tower. It is better understood as a bridge. 5G-Advanced, or 5G-A, is the second half of the 5G standards cycle, beginning with 3GPP Release 18 and continuing through Releases 19 and 20. It upgrades an installed network. 6G, called IMT-2030 at the ITU, is the next architecture and radio standard, whose first normative 3GPP specifications will be written in Release 21. One is being deployed; the other is still being defined.

That overlap is intentional. Telecom generations take most of a decade to negotiate, test, manufacture and deploy; even the annual Snapdragon first-launch race sits at the end of a long chain from standards and modems to certification and finished phones. In 2026, 3GPP lists 126 work items and 74 study items for Release 20 alone. They cover energy efficiency, AI and machine learning, satellite integration, mobility, coverage, public safety, and sensing plus communications. These are immediate improvements to operating 5G networks, but they also let engineers test ideas that may become foundational in 6G.

China’s plan makes the overlap unusually visible. At the end of July 2026, the country had 5.154 million 5G base stations and 1.303 billion 5G mobile users. The 2030 target is not “tear those out for 6G”; it is to add 500,000 5G-A stations, raise the density of 5G and 5G-A stations from 34 to 50 per 10,000 people, and continue 6G research on top of that foundation. Even the wording—commercialize 6G “at an appropriate time”—is a guardrail against reading a research timetable as a launch date.

Why “faster downloads” is the wrong headline

The skepticism is understandable. Many users experienced 5G as a new icon rather than a new life. In a lively Hacker News discussion, engineers and users repeatedly noted that ordinary browsing can feel little different from good 4G. The most persuasive defense was less glamorous: at a packed stadium or concert, the network may remain usable because newer radio systems handle far more simultaneous demand. Capacity, coverage and reliability are collective benefits; a speed test on one phone does not measure them well.

The same lesson matters for 5G-A. China’s plan mentions 10-gigabit-per-second downlink peaks in selected parks, transport hubs and venues, but its more revealing target is one-gigabit uplink in central urban areas. Downlink brings video to a phone. Uplink carries a factory camera, a drone feed or a robot’s sensor stream back to an edge computer. People’s Daily’s interviews identify weak uplink and incomplete coverage—not insufficient headline download speed—as current bottlenecks.

That is why the bridge connects naturally to systems such as Apollo Go’s robotaxi network. A self-driving fleet does not need every decision sent through a cell tower, but it does depend on maps, remote assistance, fleet coordination and graceful recovery when connectivity fails. The March 2026 Wuhan stoppage showed that a networked fleet can spread learning quickly and also share a correlated failure. Better communications infrastructure expands what is possible; it also makes resilience a design requirement. The wired side of that foundation is moving too: hollow-core fiber, which lets light travel through air instead of glass, is now being deployed to cut latency between exchanges and AI data centers.

The network starts to sense and think

ITU’s six IMT-2030 scenarios include four recognizable communications goals—immersive, hyper-reliable low-latency, massive, and ubiquitous connectivity—and two additions that explain the excitement: AI integrated with communications, and integrated sensing and communications (ISAC). The latter means radio infrastructure can use reflected signals to infer the presence, position or movement of objects while carrying data. In plain terms, parts of the network begin acting a little like distributed radar.

This is not only a 6G laboratory idea. China Mobile has already described 5G-A sensing trials for detecting low-altitude objects and supporting drone delivery in Shenzhen. That does not make today’s network a universal radar grid, nor does it prove a business model. It shows why 5G-A is a bridge: operators can learn about interference, positioning, privacy and operational cost before the feature is built more deeply into a 6G standard.

The other new axis is intelligence inside the network. The ambition is to coordinate processing across device, base station, core network and cloud, rather than merely transport requests to a distant data centre. That could support the kind of always-available services imagined by the AI agent phone, but the hard questions are not just latency. Who can inspect the context an assistant sends? Where is an inference performed? What happens when the network predicts intent incorrectly? A more intelligent network increases both capability and the amount of trust its operators must earn.

Ubiquitous connectivity adds a spatial change. Satellite and non-terrestrial networking already appear in 5G-Advanced work, while China’s plan calls for broader direct-to-phone satellite use and deeper integration between satellite and terrestrial mobile systems. The long-term vision is not one impossibly powerful tower. It is a layered network across ground, air and space, handing a connection between systems according to location and need.

Why now—and what remains unresolved

Three clocks aligned in 2026. The standards clock moved from aspirations to requirements: ITU defined measurable criteria and 3GPP dated the first specification freezes. The deployment clock moved from scattered demonstrations to a large 5G-A footprint. The policy clock attached 2030 numbers to that bridge. A People’s Daily explainer on September 22 translated those dry milestones into everyday examples, and “5G+6G” became a top search the next day.

But a calendar is not a use case. 5G promised industrial control, autonomous transport and immersive media; several arrived more slowly or narrowly than launch marketing implied. 6G will face the same test. High-frequency spectrum can offer capacity but struggles with range and obstacles. Sensing raises privacy and governance questions. Satellite links add coverage but also cost and coordination. Operators must justify another investment cycle while still completing 5G standalone and 5G-A networks.

The useful way to watch the next four years is therefore not to count laboratory gigabits. Watch whether 5G-A improves weak uplinks and crowded places; whether sensing produces tightly bounded, consent-aware services; whether satellite links close genuine coverage gaps; and whether Release 21 stays on schedule. If those pieces work, 6G will look less like a sudden sixth-generation switch and more like the name eventually given to a bridge already under construction.

Key facts

Current 5G stations in China

5.154 million (Jul 2026)

New 5G-A target

500,000 by 2030

ITU 6G scenarios

6

Draft requirements

20

First normative 6G release

3GPP Release 21

Final Rel-21 protocol freeze

Mar 2029

What people are saying

Common questions

Is 6G already available in China?

No. 5G-A is commercially deployed, while 6G remains in research, trials and standardization. China’s plan says commercialization should begin when appropriate; 3GPP’s first normative 6G release will not finish freezing until 2029.

Will a 5G phone become useless?

Not because 6G standards are being written. Mobile generations overlap for years, and the current plan expands 5G and 5G-A coverage through 2030. Future device compatibility will depend on spectrum, chipsets and operator deployment, none of which is settled by a research milestone.

What is the clearest difference between 5G-A and 6G?

5G-A evolves the deployed 5G system through Releases 18–20. Release 21 is the first normative 6G release. Features such as sensing, AI and satellite links appear on both sides of the boundary, which is exactly why 5G-A functions as a bridge.

When should ordinary users expect 6G?

Around 2030 is the industry target, not a guaranteed universal launch date. Standards must freeze, chips and equipment must be built, spectrum must be assigned, and operators must find viable services before coverage becomes widespread.