#USA #None #Huawei #AI Agents #Core Network

The Emergence of AI-Native Core Networks for Next-Gen Services

In the age of rapid technological advancements, traditional mobile networks have primarily facilitated communication through voice calls, SMS, and mobile internet access. However, with the rise of artificial intelligence (AI), the necessity for a robust infrastructure to manage an overwhelming number of AI agents is paramount. This shift is prompting a rethinking of core network architecture which must now operate with intention, make intelligent decisions, and ensure effective end-to-end task execution driven by AI.

Huawei has declared that developing an AI-native core network is vital for this transformation. By integrating AI across interfaces, functions, and procedures, operators can transition from merely providing connectivity to deploying intelligent network capabilities that underpin the next generation of digital services.

According to Huawei's Intelligent World 2035 report, it is anticipated that the number of AI agents globally could reach a staggering 900 billion by the year 2035. This prediction highlights the need for core network architecture capable of supporting long-term growth of new connections and services. As AI-enabled devices and applications gain traction, agent services are expected to become standard scenarios within mobile networks, shifting the focus from mere data transport to task-oriented functionality.

To ensure the optimal performance of AI agents, Huawei has identified three critical service scenarios:

1. Agent Over New Radio (AoNR): This scenario addresses the strict latency requirements faced by AI wearable devices, which often have limitations in processing power, battery capacity, size, and cost. AoNR aims to consistently provide an optimal user experience while ensuring end-to-end latency remains within stringent thresholds.

2. Wide Computing Network (WCN): A WCN interconnects computing resources through mobile networks, thereby enabling task-based computing planning and standardized IT services. This approach is designed to enhance computational efficiency across the board.

3. Agent Computing Network (ACN): In complex scenarios that require the collaboration of multiple AI agents across different domains and ecosystems, an ACN facilitates secure and dynamic cooperation. It offers features such as inter-domain trusted digital identity, agent discovery and routing based on skills, and dynamic task-based networking.

For the next-gen core network architecture to effectively bridge the gap between devices and networks, it must exhibit three essential characteristics:

• - Interfaces (from signaling to intention): Intent-based interfaces allow terminals to directly communicate their customized task requirements to the network. This includes specifications related to networking, computing, and user experience.
• - Functions (C/U Plans to C/U+DAP): The architecture introduces a dedicated data and AI plan (DAP) alongside traditional control and user plans, functioning as an intelligence layer that interprets task intentions and dynamically aligns resources in real-time.
• - Procedures (from AI overlay to AI-native): Deep integration of AI-driven decision-making into fundamental network procedures is necessary for seamless coordination of connectivity and computing services.

Looking to the future, Huawei has laid out a roadmap for the evolution of its AI-driven core network in two distinct stages. The first stage revolves around enhancing customer experience through the incorporation of AI functionalities within the network. The subsequent stage aims at establishing an AI-native core network, facilitating what Huawei terms as 'enhanced connectivity.' This model is not only centered on task-driven AI capabilities that guarantee task fulfillment, but also extends beyond connectivity to incorporate functions such as providing IT services and fostering agent collaboration.

To accommodate these advanced services, the network architecture will have to undergo significant restructuring, particularly in interaction mechanisms, data architecture, and computing frameworks. Moreover, this transition to AI-native networks necessitates a synergistic collaboration across the entire telecommunications ecosystem, including terminals, chips, and networks. With the synergistic mechanism between devices and networks, along with an advanced core network architecture, the next-gen core network will unlock new connections and services, paving the way for fresh business opportunities in the era of intelligence.