#USA #Beijing #Huawei #AI Agents #Core Network

Introduction

The landscape of mobile networks is on the cusp of a profound transformation, necessitated by the rise of artificial intelligence (AI) agents. Traditionally, mobile networks have been primarily focused on providing basic communication functions such as voice calls, text messaging, and internet access. However, the emergence of AI demands a paradigm shift: the architecture of core networks must evolve to manage a colossal influx of AI agents, expected to exponentially grow over the next few years.

The Future of Core Networks

Huawei strongly believes that AI-first core network architecture will be pivotal in this transformative journey. Incorporating AI into the network’s interfaces, functionalities, and procedures empowers operators to evolve from merely connecting devices to delivering intelligent network capabilities for next-generation digital services.
According to Huawei's "Intelligent World 2035" report, the number of AI agents globally could reach a staggering 900 billion by 2035, which is roughly 100 times the Earth's current population. With this anticipated surge, the core network architecture must not only accommodate new connections but also facilitate innovative services that leverage the power of AI.

Transitioning from Data Transmission to Task Processing

As the deployment of AI-enabled devices and applications gains traction, services powered by AI agents are becoming common use cases for mobile networks. This shift marks a movement from mere data transmission to proactive task processing. Huawei identifies three typical scenarios to enhance the performance of AI agent services:
Agent Over New Radio (AoNR):
To ensure minimal latency during task execution, the AoNR technology aims to facilitate high-quality user interactions while adhering to stringent low-latency requirements. Given that AI-enabled portable devices face limitations regarding processing power, battery capacity, physical size, and cost, achieving rapid response times becomes crucial.
Wide-Area Computing Network (WCN):
By aggregating computational resources across mobile networks, the WCN encourages efficient computation by distributing tasks and providing standardized computing services. This infrastructure allows AI agents to process data more rapidly and effectively.
* Agent Computing Network (ACN):
To guarantee the secure collaboration of various AI agents, the ACN will facilitate interactions among specialized agents operating in different domains. Complex tasks will often require coordination among these agents, which hinges upon a secure and dynamic communication framework that the ACN offers.

Key Characteristics for Next-Gen Networks

To bridge the gap between the capabilities of end devices and network infrastructures, next-generation core network architecture must exemplify three essential features:
1. Intent-based Interfaces:
These interfaces enable end devices to communicate personalized task execution requirements directly to the network, including specifications for resource allocation, network interaction, and quality of service.
2. Enhanced Functionality:
In addition to traditional control and user data planes, a Data AI Plane (DAP) will be introduced, playing an integral role in interpreting task-related intents and dynamically coordinating resource allocation in real-time.
3. Integrated AI Procedures:
Decision-making processes based on AI will be deeply integrated into the core network's operational procedures, ensuring coherent coordination of connectivity and computational resources.

Evolutionary Stages of AI-Enhanced Networks

Looking ahead, Huawei has outlined two stages in the evolution of AI-driven core networks. The first stage focuses on utilizing AI capabilities as supplementary enhancements to improve user service quality. The second stage envisions an AI-first core network that ensures what Huawei terms "enhanced connectivity"—whereby AI services tailored to specific tasks guarantee completion, and "beyond connectivity"—where the network extends its value proposition to include computational services and facilitate AI agent interactions. This will necessitate a comprehensive redesign of network architecture, particularly regarding interaction mechanisms, data architecture, and computational frameworks.

Conclusion: Collaboration in an AI-Driven Ecosystem

The transition to AI-integrated networks mandates robust collaboration within the ICT ecosystem, encompassing end devices, chipsets, and networks. By advancing the synergy between devices and networks, along with cutting-edge architecture, next-generation core networks will empower new connections and services, unlocking business opportunities in the age of intelligent technologies. This strategic alignment will pave the way for smarter, more responsive mobile networks capable of supporting an unprecedented scale of AI-driven interactions.