China Mobile and Huawei Complete Field Trials of User-level Dynamic Network Slicing in Shenzhen
China Mobile and Huawei have wrapped up comprehensive verification trials for user-level dynamic network slicing technology in Shenzhen. The innovation enables dedicated resource reservation for individual subscribers to deliver exclusive SVIP network channels, with integrated intelligent hardware modules delivering end-to-end service quality guarantees. The trials mark a definitive shift for 5G-Advanced networks from best-effort universal connectivity to a new operational phase defined by on-demand guaranteed target data rates, delivering a replicable, deployable benchmark for global telecom operators transitioning from data volume-centric operations to experience-driven service models.
Legacy industry slicing architectures rely on static priority resource allocation to safeguard critical services. This framework creates uneven resource distribution under fluctuating wireless channel conditions, with surplus spectrum allocated during robust coverage and insufficient bandwidth reserved when signal quality deteriorates. The newly validated dynamic slicing architecture abandons rigid fixed-resource prioritisation, built around two foundational core capabilities: intelligent spectral efficiency prediction and elastic dynamic resource reservation.
Rather than locking static bandwidth allocations permanently, the system continuously monitors real-time network load metrics and subscriber geographic positions to calculate live spectrum utilisation efficiency. Protected bandwidth resources are allocated precisely to match active user demand, eliminating both bandwidth shortages and idle spectrum waste. Reserved capacity is automatically released during off-peak service periods to facilitate resource sharing across all connected subscribers.

High-density university campuses, urban metro networks and high-speed intercity railways present three notoriously challenging operating environments for wireless connectivity, characterised by extreme population density, sharp service demand surges and highly variable user mobility. Network resource constraints within these zones routinely degrade user experience for latency-sensitive applications including live streaming, voice over instant messaging, video conferencing and audio playback, which demand unobstructed dedicated transmission pipelines. Full-scale overload testing covered all three high-pressure scenarios to validate the universal adaptability of dynamic slicing.
Trials conducted at the Yuehai Campus of Shenzhen University delivered consistent target guaranteed throughput for all connected users, regardless of proximity to signal transmission nodes or peripheral coverage fringes, with no sharp deterioration in service quality triggered by positional shifts. Short-form video browsing during class breaks and high-definition live broadcasts for campus club events maintained seamless, stable connectivity throughout peak demand windows. Tests on Shenzhen Metro Line 4 and the Xiamen-Shenzhen high-speed railway deployed real-time channel forecasting and dynamic bandwidth orchestration to cut the proportion of low-throughput connections within mobile overload zones. Measured peak stalling rates for short-form video content fell by sixty per cent during metro rush hours. Subscribers travelling at full line speed within rolling stock maintained uninterrupted high-definition video streaming and multi-party video conferencing sessions.
China Mobile will continue advancing frontline 5G-Advanced technical innovation alongside cross-industry ecosystem partners to build robust digital communications infrastructure. Iterative network architecture upgrades, refined user experience frameworks and expanded vertical application ecosystems will unlock the full spectrum of digital use cases across all operating environments, responding to sustained consumer demand for premium-grade communications services and supplying robust technical foundations and innovative momentum for broader digital transformation.
