Huawei launches ICNMaster MDAF to support core network high stability

Huawei launched ICNMaster MDAF, a system intended to support high-stability operation of core networks and to address interruptions that the company said could affect tens of millions of users and public “digital well-being” and “sense of social participation.”

Huawei positioned the launch against a backdrop of evolving 5G and cloud technologies and a rise in service diversity that, the company said, produced software-hardware decoupling, multi-generation coexistence, complex inter-generation APIs, signaling surges, transport network faults, and data center failures. GlobalData statistics in the release showed 42% of operators experienced core network service interruptions in the past three years, and TM Forum had issued a core network high-stability assessment standard with dozens of operators participating.

The release described ICNMaster MDAF as built around a Mixture of Models (MoM) architecture, multi-agent collaboration, and a network digital twin. According to the text, the MoM approach routed routine, high-frequency events to fast inference models and assigned complex anomalies to deep reasoning models; multi-agent orchestration and conflict resolution were combined with digital-twin validation to form a closed loop of perception, analysis, decision, and execution.

Huawei said the system built on earlier components named Fault Management Agent and Complaint Handling Agent and introduced an agent-network digital twin synergy. The company said the system enabled proactive risk prevention, reduced incident probability, rapid fault recovery, and service restoration within minutes, and that it aimed to support operators moving toward L4 high-stability.

Huawei said the “Agent + Digital Twin” collaboration paradigm validated by ICNMaster MDAF laid a technical and practical foundation for realizing fully autonomous driving networks in the future.