The Report Minimal Technical Efficiency Necessities for IMT-2030 (6G) was formally finalized on the 51st assembly of ITU-R Working Social gathering 5D (WP 5D), which concluded on 12 February 2026. It marks the official transition of 6G from imaginative and prescient and idea to the implementation section, laying a strong basis for the following submission and analysis of candidate know-how proposals, in addition to the event of worldwide unified requirements.
This report systematically defines a number of key efficiency indicators that 6G radio interface applied sciences shall meet. For instance, peak community capability in large-scale networking, ubiquitous determinism and composite necessities, amongst different components, are posing new challenges for the event of 6G networks.
6G: The infrastructure for Human-Agent Synergy
Past the evolution of requirements, we’re witnessing the fast development of AI. ZTE argues that the business is shifting from the cell web period to an agent-centric web period, a shift that may rework service necessities throughout three dimensions: linked objects, interactive content material and repair paradigms. Connectivity, the corporate says, will increase from people + issues to people + issues + clever brokers, with always-on brokers demanding ubiquitous and deterministic connections. On the similar time, interactive experiences are anticipated to evolve from visual-audio content material to AI-driven multi-sensory immersion, rising necessities for uplink and downlink efficiency, latency and jitter. Service paradigms may also change, as conventional application-centric networks evolve into agent networks that help service brokers. These networks might want to ship built-in data providers that mix sensing, knowledge processing and intelligence, with generative methods to help numerous and ubiquitous AI providers.
6G is the infrastructure for a Human-Agent Synergy world. In response, ZTE says it has constructed a full-chain, systematic 6G strategic know-how structure throughout requirements, chips, structure and algorithms, encompassing two transformative know-how instructions and 4 evolutionary know-how instructions.
AI and space-air-ground integration as core pillars of ZTE 6G technical technique
Towards this backdrop, the corporate has outlined a strategic 6G know-how roadmap constructed round two main transformative instructions: the deep integration of synthetic intelligence and communications, and the convergence of house, air and floor networks.
The primary main route is the deep integration of AI and communications. ZTE’s imaginative and prescient of “Native AI” in 6G means that synthetic intelligence will permeate not solely the community layer, but in addition the radio entry community (RAN) and the air interface itself. The corporate says AI has already demonstrated worth in bettering metrics resembling spectral effectivity, however a number of key questions stay, together with how one can ship extra environment friendly and inexperienced infrastructure and how one can coordinate the evolution of agentic networks and repair brokers.
The second route is space-air-ground integration. As worldwide analysis and standardization efforts progress, ZTE sees the convergence of satellite tv for pc techniques and mobile networks as an inevitable pattern. By leveraging the large-scale industrial chain of terrestrial mobile networks, satellite tv for pc payloads primarily based on industrial-grade chips might obtain higher efficiency and value effectivity than conventional satellite tv for pc techniques. Nonetheless, the corporate notes that true integration would require unified end-to-end design throughout structure, air interface, spectrum and operational administration.
In the meantime, final spectral effectivity, modern AI IoT, purposeful enlargement and trusted safety stand because the 4 enduring evolutionary know-how instructions.
Effectivity is the important thing driver in direction of a well-balanced means of AI RAN
ZTE positions effectivity as a central design precept for 6G techniques, particularly as AI turns into extra deeply embedded in community infrastructure. The corporate separates this integration into two distinct dimensions: “AI for RAN” and “RAN for AI”.
AI for RAN refers to using synthetic intelligence to enhance community efficiency, cut back vitality consumption and decrease operational bills. Based on ZTE, this space presents large-scale and deterministic demand, and devoted chips provide the optimum cost-performance ratio. The corporate notes that solely a small variety of physical-layer AI modules have been standardized thus far, and for features with clear mathematical fashions, devoted chips paired with conventional algorithms are extra appropriate than GPUs, which may devour extra energy and incur larger prices with out delivering important efficiency features.
Against this, RAN for AI explores the evolution of the radio entry community into edge AI infrastructure. Right here, ZTE sees a number of challenges. The distributed nature of RAN means its computing scale is much smaller than that of centralized cloud AI infrastructure. Whereas edge AI provides latency benefits for particular use circumstances, the general market stays unsure, and operators haven’t but seen returns that justify large-scale GPU investments. Because of this, the corporate believes RAN for AI ought to comply with a demand-driven deployment method.
The convergence of AI and communications is an inevitable pattern, and ZTE has been embracing AI proactively and driving steady innovation alongside this path. The corporate launched the NodeEngine computing base station in 2020, adopted by the UniEngine hyper-converged machine and, in 2024, the AIREngine clever computing engine for base stations. Moreover, from a scientific end-to-end perspective, in 2025, we delivered the AIR resolution; right this moment, by shifting the main target from community efficiency to enterprise worth beneath the technical philosophy of serving AI with AI, AIR MAX goals to maximise effectivity in three instructions: vitality effectivity, labor effectivity and funding effectivity by the improve of technical structure, manufacturing effectivity and enterprise mannequin.
GigaMIMO leads AI-native wi-fi, ushering in a brand new section of 6G
ZTE’s GigaMIMO addresses key challenges round capability, protection and deterministic expertise for large-scale 6G community deployments.
Within the race to innovate core 6G applied sciences, ZTE has made a number of breakthroughs and developed a set of corresponding prototypes—with its GigaMIMO structure standing out as a flagship development for 6G tech innovation and the deep integration of AI. Leveraging its early work on large MIMO courting again to the pre-5G period, the seller has engineered each centralized and distributed GigaMIMO architectures, constructed to spice up community capability, increase protection and elevate end-user expertise.
• Powered by multi-dimensional technical innovation and refined AI integration—together with AI-driven hyperlink adaptation, beam prediction and clever scheduling—ZTE’s centralized GigaMIMO prototype, which options 2048 antenna parts within the U6G band (6425–7125 MHz), has delivered important efficiency breakthroughs in subject checks, delivering a 10x enhance in cell capability and enabling co-site protection with mid-band 5G.
• For its distributed GigaMIMO resolution, ZTE has innovatively deployed a portfolio of AI collaborative scheduling applied sciences, resembling high-precision inter-cell and inter-site synchronization, channel state data compression and suggestions, and low-complexity joint beamforming enhancement. Validated in a industrial community in partnership with China Cellular, these applied sciences have boosted edge person expertise by sixfold.
Moreover, by the co-design of large MIMO and AI algorithms, spectral effectivity is enhanced by greater than 30%, whereas making certain constant community expertise and differentiated service capabilities, laying a strong basis for the large-scale deployment of 6G.
The corporate can also be working with ecosystem companions. Along with China Cellular, the corporate accomplished the business’s first IoDT (Interoperability Growth Testing) verification through a Qualcomm prototype primarily based on GigaMIMO know-how with 400 MHz in a single service.
Business collaboration and immersive use circumstances
ZTE emphasizes that collaboration throughout the business chain can be important for 6G improvement.
As AI evolves from conversational interplay to closed-loop execution and bodily AI, gadgets will face challenges associated to computing energy, kind issue, battery life and value. The corporate says this makes end-edge-cloud collaboration essential for enabling a broad vary of AI providers.
One of many early use circumstances the corporate is exploring is AI-powered immersive communication. ZTE expects applied sciences resembling AI glasses and spatial computing to allow multi-sensory experiences that transcend sight and sound to incorporate contact and notion.
ZTE plans to showcase a spread of 6G-related reveals on the upcoming Cellular World Congress in Barcelona, Spain, highlighting each its long-term imaginative and prescient and ongoing technical work throughout AI integration, space-air-ground convergence and next-generation radio applied sciences.
