ZTE cultivates intelligence where networks and AI converge

An interview with Li Xiaotong, Vice President of ZTE Corporation

In the rapidly evolving landscape of telecommunications, we stand at the threshold of a transformative era where artificial intelligence and 5G-Advanced technologies are converging to reshape the very foundation of network infrastructure. As the industry transitions from the traditional mobile internet era to what many are calling the Mobile AI era, we had the opportunity to engage with Li Xiaotong, Vice President of ZTE Corporation and General Manager of RAN Products, to explore how these revolutionary forces are driving innovation and redefining the role of communication networks.

From connection pipes to intelligent engines

“We are witnessing a fundamental shift in how we conceptualize network infrastructure,” Li begins, setting the stage for our discussion. “Just as the Renaissance period in Italy redefined art and science, today’s Mobile AI era is redefining the role of communication networks. Networks are no longer merely pipes for data transmission—they are becoming intelligent engines capable of understanding, predicting, and adapting to user needs in real-time.”

This transformation is driven by profound changes in traffic patterns that few in the industry fully grasp yet. “While generative AI currently represents only a fraction of total network traffic, its impact on network architecture is disproportionately significant,” Li observes. “AI interactions fundamentally alter the traditional 90-10 downlink-uplink ratio. We’re seeing AI applications generate 26% uplink traffic compared to traditional apps’ 10%—a shift that requires completely rethinking network design principles.”

Li anticipates the explosive growth in AI Agent applications, creating what he terms “a new traffic taxonomy”. Traditional architectures should evolve to accommodate AI agents generating vastly increased interaction volumes, with unprecedented demands for both uplink capacity and deterministic latency.

AIR RAN: Intent-driven intelligence at scale

When asked about ZTE’s response to these emerging challenges, Li introduces us to AIR (AI Reshaped RAN) RAN—ZTE’s groundbreaking solution for the Mobile AI era. “The industry talks about ‘AI-assisted’ networks, but we realized this approach was fundamentally insufficient,” Li explains. “True transformation requires ‘AI-native’ architecture where intelligence is embedded in the network’s DNA, not merely layered on top.”

AIR RAN represents a paradigm shift to what Li calls “intent-driven intelligent service models.” Unlike rule-based systems that react to predefined conditions, AIR RAN interprets high-level intentions and dynamically orchestrates resources across time, frequency, spatial, and power domains. “We focus AI deployment precisely where non-linear complexity demands it most—in signal variations, traffic fluctuations, and dynamic radio environment changes,” Li explains.

The technical foundation lies in ZTE’s SoC accelerator and heterogeneous computing architecture, which enables the industry’s first native-AI BBU. The commercial deployments of native-AI BBU validate this approach: video service quality issues reduced by 80%, live streaming uplink rates enhanced by over 20%, gaming latency optimized by 18%, and 5G base station daily energy consumption reduced by more than 12%.

Transforming business models: From traffic to experience

The conversation naturally turns to the commercial implications of these technological advances. Li emphasizes how AIR RAN is enabling a fundamental transformation in operator business models—from traffic monetization to experience monetization.

“The traditional telecom revenue model is hitting a wall,” Li states candidly. “Operators today face a critical challenge: growing traffic without matching revenue gains. To overcome this, they must shift to experience-driven value propositions, prioritizing user satisfaction and innovative service models.”

ZTE’s approach centers on three strategic pillars that Li believes will define the next decade of telecom economics. First, experience assurance through what he terms “precision networking”—precise allocation, perception, and evaluation creating closed-loop optimization. Second, multi-dimensional efficiency improvements spanning spectral, energy, and operational domains. Third, value-based user segmentation enabling differentiated service delivery.

Source: RCRWirelessNews