#Huawei #MoM #Network Stability

Introduction

In today's digital era, our daily lives and economic activities are heavily reliant on network infrastructure. The stability of these networks is directly linked to society's sense of 'digital welfare' and 'social engagement.' As service disruptions can affect millions, maintaining the core network's stability is crucial as it serves as the brain of the network.

Enhancing Network Stability with Huawei

On December 23, 2025, Huawei introduced the ICNMaster MDAF solution, a state-of-the-art system built on the principles of Models of Models (MoM), multi-agent collaboration, and digital twin technology. This intelligent, high-stability system empowers global operators to enhance their transition toward L4 high stability while ensuring service continuity through proactive risk prevention, rapid fault recovery, and service reinstatement within minutes.

With the advancement of 5G and cloud technologies, network services are diversifying. Challenges arise in the form of hardware-software decoupling, coexistence of multiple generations, complex inter-generational APIs, signal fluctuations, transport network failures, and data center outages. These issues exert more pressure on operators to maintain carrier-class reliability in core networks. According to GlobalData, 42% of operators have experienced core network service disruptions in the last three years, with increasing frequency raising industry-wide concerns.

The TM Forum has introduced a core network reliability assessment standard to provide guidance to global operators on improving core network reliability progressively. Many operators are already participating in these assessments, establishing a consensus within the industry to prioritize enhancing core network reliability.

Huawei's Solve: ICNMaster MDAF

To address demands for high stability in core networks, Huawei has released the ICNMaster MDAF intelligent high stability system. Building upon previous innovations such as the “Fault Management Agent” and “Complaint Processing Agent,” this product introduces groundbreaking technologies within the MoM architecture, multi-agent collaboration, and agent-network digital twin synergy. Key features include automated issue closure, reduced event likelihood, and service recovery within minutes, reinforcing L4 high stability of the core network.

Transitioning to Multi-Model Architecture

Unlike traditional single-model architectures, the MoM architecture executes scenario leaps by fully integrating the benefits of fast inference models with deep reasoning models like DeepSeek. It offers a smart model traffic routing framework where tasks are dynamically allocated. High-frequency, routine events are promptly addressed by accurate models, while complex anomalies necessitating deep reasoning are assigned to specialized models, ensuring optimal efficiency and precision.

From Single-Agent to Multi-Agent Collaboration

Traditional single agents only automate isolated individual scenarios. In contrast, multi-agent collaboration technology organizes multiple agents systematically, enabling agent orchestration, conflict resolution, and more. Improvement actions are verified through the network digital twin system, ensuring corrective measures are accurate and effective before implementation in the live network. This establishes a self-terminating loop of perception, analysis, decision-making, and execution.

Conclusion

Huawei's ICNMaster MDAF MoM-based multi-agent collaborative high stability solution represents a significant step towards fully autonomous networks, allowing operators to provide not just connectivity but a robust, flexible, and intelligent digital foundation for the future. It addresses the long-standing conflict between network stability and operational efficiency, signifying a strategic shift in the telecommunications industry from a network-centric to a user-centric operational model. The