C114 News March 13 Special Report (Jiang Junmu) From the &#34;clear listening&#34; of the voice era to the &#34;fast downloading and smooth viewing&#34; of the data era, regardless of how the connectivity capability has improved, mobile communication networks have always played the role of a &#34;pipeline&#34; over the past thirty years. However, with the wave of mobile AI sweeping the world, various AI applications and intelligent agents are beginning to enter daily life, business operations, industrial control, urban governance, and other aspects. Can the network still operate under the old logic?
At the 2026 Mobile World Congress (MWC 2026), a barometer-like event, Huawei's Vice President and President of the Wireless Network Product Line, Direction, provided an answer with the Agentic MBB full series solution.
As a combination of Huawei's trend insights, innovative strength, and engineering capabilities, the Agentic MBB full series solution does not bring about certain technical breakthroughs or linear hardware performance enhancements, but rather a business-centric system-level network intelligence, meaning that the network evolves from a &#34;pipeline&#34; to an actively perceiving, even thinking and decision-making &#34;intelligent organism.&#34; This may provide a glimpse into the leap in industrial thinking towards the mobile AI era and the path of network reconstruction.
<img src="https://imageproxy.pbkrs.com/https://image.c114.com.cn/20260313/17733802352920.jpg?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://image.c114.com.cn/20260313/17733802352920.jpg"/>
Diverse Demands of Mobile AI Call for Network Reconstruction
The explosive development and accelerated popularization of AI agents are reshaping the demand landscape for mobile communication networks.
Direction pointed out at the press conference that multi-modal interaction requires a fivefold increase in uplink capacity, real-time collaborative intelligence needs to compress end-to-end latency to within 400 milliseconds, and ubiquitous intelligent agent services are forcing network coverage to extend from urban areas to &#34;villages, roads, and blind spots.&#34; Extending the timeline to 2035, the objects of network services will expand from nearly 10 billion people to nearly 1 trillion intelligent agents, most of which will operate on mobile devices.
Faced with these complex and diverse demands, the traditional &#34;one-size-fits-all&#34; resource supply model has become unsustainable, and multi-dimensional network capabilities are more important than ever.
Huawei's breakthrough approach is to reconstruct the network on three levels: first, the reconstruction of network capabilities, shifting from a past focus on downlink to an &#34;all-round player&#34; with significant uplink and low latency characteristics; second, the reconstruction of resource management, replacing traditional human-centric configurations with intelligent dynamic management; and finally, the reconstruction of interaction modes, evolving from API-based interactions to business-intent-based interactions, allowing the network to truly understand service demands.
The purpose of these three reconstructions is to achieve precise matching between network supply and business demand, thereby fully unleashing network value, which has also become the underlying logic of Agentic MBB.
RAN Agent: Equipping the Network with a Brain
As the core lever of the aforementioned reconstruction, Huawei has introduced two core innovations: RAN Agent and Adaptive Air. Among them, RAN Agent is not only the industry's first wireless network intelligent agent but also represents a qualitative change in network intelligence from &#34;functional addition&#34; to &#34;system native&#34;—relying on a unique communication large model and wireless digital twin system (RDTS)
It has taken the lead in building network-level, system-level, and verifiable global intelligent capabilities, opening up a new dimension of network intelligence for wireless networks.
<img src="https://imageproxy.pbkrs.com/https://image.c114.com.cn/20260313/17733802359492.jpg?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://image.c114.com.cn/20260313/17733802359492.jpg"/>
As a wireless solution deeply integrated with communication large models, this intelligent agent possesses intent interaction, global network perception, and intelligent decision-making capabilities, achieving precise network resource scheduling around users and services, systematically outputting the best strategies from a holistic network perspective, and comprehensively enhancing user experience, operational efficiency, and network energy efficiency.
In terms of architectural design, the RAN Agent plays a core hub role that connects the upper and lower layers. Northbound, it relies on the innovative A2A-T interface to connect with the operators' business intelligent agents, undertaking business intents and accurately meeting diverse scenario needs such as VIP assurance, regional user experience enhancement, and network-level energy-saving and consumption reduction; southbound, it bridges the new generation Adaptive Air base station equipment, collaborating with network element intelligence and network intelligence to achieve precise resource scheduling. Through end-to-end efficient collaboration, it forms a multi-objective, cross-scenario minute-level &#34;prediction - analysis - decision - execution&#34; full closed loop, enabling daily network optimization, extreme energy savings, and business revenue growth across all scenarios of single-domain autonomy.
The role of RDTS is to digitize the physical network, building a data foundation for the RAN Agent: through site-level digital twins, making sites visible, manageable, and controllable; through industry-leading multi-dimensional modeling, completely linking the physical world encompassing devices, networks, and environments with the digital world. The combination of the two achieves capabilities such as proactive operation and maintenance, virtual iteration, and strategy pre-validation. For example, strategies can be rehearsed in a virtual space, and validated solutions can then be deployed to the physical network, significantly reducing trial and error costs.
Currently, many features of the RAN Agent have been validated in the live network. China Mobile, in collaboration with Huawei, has globally deployed the first large-scale wireless network intelligent agent and established eight major high-level autonomous intelligent network demonstration zones in Shanghai, Anhui, Guangdong, and other regions, fully integrating AI+ networks into mobile production processes. So far, 400,000 5G sites have completed intelligent upgrades, empowering over 2,000 network engineers. The practical data is quite convincing: in Henan, precise energy savings based on the intelligent agent achieved an additional energy gain of 5.5% per 5G base station; in Fujian, AI's precise prediction and intelligent identification increased the accurate issuance rate of traffic overlay packages to 90%; in Anhui, the introduction of natural language interaction shortened fault repair time by 20%; in Shanghai, the full automation upgrade of operational processes improved the self-closure rate of wireless network optimization by 15%.
Adaptive Air: Enhancing Multi-Dimensional Capabilities with Network Element Intelligence
The &#34;intelligent awakening&#34; of the network relies on the support of underlying network elements. Huawei's new generation Adaptive Air solution deeply integrates industry-leading radio frequency and baseband with intelligent algorithms, providing diversified network capabilities for business and meeting the rapid development needs of AI
In terms of RF upgrades, Huawei has launched the GigaGreen Plus series, achieving significant breakthroughs in new materials, new processes, and new antenna feeder architectures, resulting in a 30% reduction in hardware size and weight while comprehensively enhancing performance and energy efficiency.
The world's first 256T U6GHz AAU and the tri-band UWB MetaAAU, which deeply integrate exclusive technologies such as ELAA, new filters, and amplifiers, will provide a multi-dimensional ultimate experience of &#34;10Gbps downlink and 1Gbps uplink&#34; after deployment on the 5G-A network. The strategic significance of the U6GHz AAU lies in its forward-looking layout of the 6GHz frequency band—U6GHz is the last 700MHz contiguous bandwidth resource in the Sub-6GHz mid-band, combining the dual advantages of wide coverage and large capacity, which not only meets the diverse business needs of today but also enables a smooth evolution from 5G-A to 6G. After multiple discussions at the World Radiocommunication Conference (WRC), this frequency band has been established as one of the mainstream frequency bands for future mobile communications. Meanwhile, the new generation EasyAAU and the industry's first tri-band 8T8R RRU, relying on disruptive innovations such as sapphire amplifiers, achieve significant improvements in coverage and capacity, while also being extremely lightweight and energy-efficient. For example, the new generation EasyAAU reduces power consumption by 40%, increases coverage by 30%, and weighs only 12 kilograms, making it easy to install.
In terms of baseband upgrades, Huawei has released the new generation UBBPi series, which achieves double the capacity and energy efficiency through innovative technologies such as system-level Chiplet, ultra-heterogeneous parallel processing architecture, and near-storage computing.
With the industry's leading all-dimensional intelligent collaborative algorithm, this series of products achieves &#34;time, frequency, space, and power&#34; global intelligent optimization under 20 times the computational complexity, doubling the edge experience of cells and improving the average experience by 40%, truly enabling every user and every business to have a full-scenario Cell-Free experience—whether at the center or edge of the cell, users can receive consistent service quality.
&#34;Though the old country is old, its fate is renewed.&#34; The mobile communication industry has come a long way, but under the wave of mobile AI, its mission is being redefined—In the past, the value of the network was measured in &#34;bits,&#34; resembling a race for speed; in the future, the value of the network will be measured in &#34;intelligence,&#34; which includes the ability to understand business intentions, adapt to diverse scenarios, create differentiated experiences, and foresee failures in advance. Huawei's Agentic MBB translates the new value standard and the concept of &#34;network and AI moving towards each other&#34; into executable solutions, from the release of RAN Agent to the upgrade of Adaptive Air, from the layout of U6GHz to the evolution of L4 intelligent networks, helping operators reconstruct networks, solidify their foundations, and turn the era of mobile AI into tangible commercial success.
<img src="https://imageproxy.pbkrs.com/https://www.c114.com.cn/news/images/t21.gif?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://www.c114.com.cn/news/images/t21.gif"/>

159998

512480

588780

Huawei launched the Agentic MBB full series solutions at the 2026 World Mobile Communications Conference, marking the transition of mobile communication networks from the traditional "pipeline" role to an intelligent "smart organism" with perception and decision-making capabilities. This solution addresses the diverse needs of the mobile AI era, emphasizing the reconstruction of network capabilities, intelligent resource management, and the expansion of service targets. It is expected that by 2035, network services will cover nearly one trillion intelligent entities

C114 通信网

Tianhong CSI Computer Theme ETF

GTJA Allianz CSI All Share Semiconductors & Semiconductor Equipment ETF

GTJA Allianz SSE STAR Chip Design Thematic ETF

ChinaAMC SSE STAR Semiconductor Material & Equipment Thematic ETF

SPDR S&P Semicon

Huawei

Direxion Semicon Bull 3X

Guotai CES Semiconductor Chip Industry ETF

Guotai CSI Computer Theme ETF

ChinaAMC Guozheng Semiconductor Chip ETF

VanEck Semiconductor ETF

China Southern CSI Semiconductor Industry Custom ETF

iShares Semiconductor ETF

From Pipeline to Intelligent Lifeform: Huawei Agentic MBB Helps Operators Solidify Network Foundation in the Mobile AI Era