Huawei Urges Operators to Rebuild Networks Around AI Data Centers

Why AI changes the network, according to Huawei

The World Broadband Association (WBBA) held the Broadband Development Congress (BDC) as part of Network X 2026 in Vienna. Bob Chen gave a keynote titled "Reshaping Network Value for New Growth in the AI Era."

Chen argued that over the past three decades every boom in services, from voice to the internet to video, depended on network support. AI is now developing quickly, he said. According to the speech, large AI models now have close to 10 trillion parameters, and token consumption rose by almost 50 times over the past year. In his view, the objects, frequency and quality of network connections have changed dramatically, and the industry has to rethink how network value is created.

The decisive factor, Chen said, is a change in philosophy for network construction: planning and build-out should follow the needs of AI services and AIDC layouts, while AI technology should be used to improve operations and maintenance.

Rebuilding the architecture around AIDCs

Networks have traditionally been organized by administrative region, split into national backbones, regional backbones and metropolitan networks. Chen said the number of AIDCs has grown almost tenfold with the rise of AI, and traffic volume is now driven by computing power rather than population.

He used Zhongwei in China as an example. The city once accounted for only about 0.1% of China's total traffic. It has since become one of the country's eight national computing hubs, and its intelligent traffic now makes up more than 10% of the country's total bandwidth. Chen said similar cases are appearing worldwide.

His recommendation is an AIDC-centered architecture, with networks organized in layers around core, regional and edge data centers. Direct one-hop mesh links between data centers would enable deterministic latency circles of 1-5-20 ms and fast time to market.

New business models for enterprise and home broadband

For enterprise customers, Chen said operators can move from selling bandwidth alone to selling "bandwidth + latency + reliability". He cited an AI-focused industrial design company in China that uses AI rendering to shorten production cycles sharply, and needs latency below 2 milliseconds for real-time rendering. According to Chen, OXC, ASON and quantum encryption allow networks to offer deterministic latency, high reliability and strong security, which raises the value of leased lines. Operators could also use computing platforms to sell "network + computing power + model/application" as integrated services.

For home broadband, he proposed a strategy of "one fiber, one network, one home". One Network means high-quality whole-home Wi-Fi using FTTR or Wi-Fi mesh. One Home means smart applications supported by AI-capable home terminals and AI agents. Home AI agents also change how users interact with broadband, Chen said, letting them control home networks with simple voice commands.

Redefining operations and maintenance

Chen called for end-to-end intelligence across planning, construction, maintenance, optimization and operations, built on intelligence at the level of network elements, the network and services. He gave three examples:

• Wi-Fi optimization: Most home network complaints stem from Wi-Fi problems. AI-based optimization identifies sources and types of interference and adjusts transmit power and other parameters in real time, raising Wi-Fi speeds in interference-heavy environments by about 20%.
• Energy saving: Network equipment used to run at full power regardless of traffic. Intelligent traffic analysis now allows dynamic power adjustment for OLT chassis and line cards, cutting total equipment energy consumption by more than 20%.
• Fault handling: Home broadband troubleshooting is split into three levels. O&M AI agents automatically fix air-interface problems such as roaming failures (self-healing). AI agents in the NOC handle configuration parameter issues remotely. On-site visits use O&M Copilot to pinpoint physical fiber problems.

What it means for operators

The speech is Huawei's pitch to network operators to shift their planning from geography to computing power, and to broaden their products beyond raw connectivity. The performance figures, such as the 20% gains in Wi-Fi speed and energy use, come from the company's own presentation. Chen closed by saying that the three "R" can reshape the value of the UBB network and help the industry pursue new growth in the AI era.