Building an AI-Native Core Network for Next-Gen AI Agent Services
In an era dominated by advancements in machine intelligence, the telecommunications landscape is transitioning dramatically from traditional communication methods to innovative networks that function natively with artificial intelligence (AI). The evolution of a dedicated AI-native core network promises to reshape the delivery and performance of digital services in ways previously unimaginable. This network transformation is epitomized by Huawei’s vision for what they call an AI-native core network, aimed at catering to a new generation of AI agents.
Traditionally, mobile networks started by facilitating communication primarily through voice calls and text messages, followed by mobile internet access. As we progress toward a future filled with intelligent digital systems that rely heavily on AI agents, the core network is required to adapt significantly. One of the fundamental changes is that the core network must now be equipped to understand intents and make intelligent decisions, ensuring seamless task completions driven by AI. According to Huawei’s report, "Intelligent World 2035," the number of AI agents could balloon to 900 billion by 2035, a stark increase from the current global population. This exponential growth underscores the urgent need for a core network architecture capable of handling a vast influx of AI-driven connections and associated services.
As we embrace AI in our daily lives, we’re witnessing the dawn of a new era where AI agent services will become mainstream, ushering the telecommunication sector from a focus on mere data transfer to a sophisticated ecosystem that efficiently supports specialized tasks. Huawei proposes three scenarios that exemplify the integration of AI services into core network functionalities:
1. Agent in New Radio Networks (AoNR) - This concept revolves around optimizing task execution times. Mobile AI devices, characterized by limited computational power and battery capacity, require low latency and high reliability in executing their designated tasks. The AoNR aims to guarantee that task completion times remain consistently below predetermined thresholds, ensuring a smooth user experience.
2. Wide Computing Networks (WCN) - This scenario focuses on enhancing computational efficiency by interlinking computational resources across mobile networks. By allocating computing resources based on task demands, WCN can offer standardized computational services, operating fluently over diverse network conditions.
3. Agent Computing Network (ACN) - In complex scenarios, tasks often necessitate collaboration among multiple AI agents across various domains and ecosystems for successful execution. The ACN framework introduces trust through digital identity across domains, facilitating agent discovery and task delegation based on specific skills. It promotes seamless interactions among AI agents, allowing them to work together dynamically to achieve complex objectives.
To bridge the increasing disparity between endpoint capabilities and core network services, the architecture of next-gen core networks needs to embody three pivotal features:
Looking further into the future, Huawei envisions two stages pivotal for transforming core networks into AI-native ecosystems. The preliminary phase involves implementing additional layers of AI functionalities aimed at enhancing customer experiences. The latter stage foresees the emergence of a next-gen AI-native core network that encompasses what Huawei describes as "expanded connectivity" – where AI-driven services are empowered to ensure task completion, and "beyond connectivity" – advancing past mere communication to deliver computing services and facilitate agent collaboration. Such transformative shifts necessitate a complete overhaul of network architecture, focusing on interaction mechanisms, data architecture, and computation frameworks.
Moreover, the transition to AI-native networks demands extensive collaboration across the entire ICT ecosystem, including endpoint devices, integrated circuits, and network infrastructure. By fostering cohesive interactions between endpoint devices and networks, alongside an advanced architectural framework, AI-native core networks will not only handle novel types of connections and services, but will also unlock new business opportunities in the age of intelligent technology.
Traditionally, mobile networks started by facilitating communication primarily through voice calls and text messages, followed by mobile internet access. As we progress toward a future filled with intelligent digital systems that rely heavily on AI agents, the core network is required to adapt significantly. One of the fundamental changes is that the core network must now be equipped to understand intents and make intelligent decisions, ensuring seamless task completions driven by AI. According to Huawei’s report, "Intelligent World 2035," the number of AI agents could balloon to 900 billion by 2035, a stark increase from the current global population. This exponential growth underscores the urgent need for a core network architecture capable of handling a vast influx of AI-driven connections and associated services.
As we embrace AI in our daily lives, we’re witnessing the dawn of a new era where AI agent services will become mainstream, ushering the telecommunication sector from a focus on mere data transfer to a sophisticated ecosystem that efficiently supports specialized tasks. Huawei proposes three scenarios that exemplify the integration of AI services into core network functionalities:
1. Agent in New Radio Networks (AoNR) - This concept revolves around optimizing task execution times. Mobile AI devices, characterized by limited computational power and battery capacity, require low latency and high reliability in executing their designated tasks. The AoNR aims to guarantee that task completion times remain consistently below predetermined thresholds, ensuring a smooth user experience.
2. Wide Computing Networks (WCN) - This scenario focuses on enhancing computational efficiency by interlinking computational resources across mobile networks. By allocating computing resources based on task demands, WCN can offer standardized computational services, operating fluently over diverse network conditions.
3. Agent Computing Network (ACN) - In complex scenarios, tasks often necessitate collaboration among multiple AI agents across various domains and ecosystems for successful execution. The ACN framework introduces trust through digital identity across domains, facilitating agent discovery and task delegation based on specific skills. It promotes seamless interactions among AI agents, allowing them to work together dynamically to achieve complex objectives.
To bridge the increasing disparity between endpoint capabilities and core network services, the architecture of next-gen core networks needs to embody three pivotal features:
- - Intent-based Interfaces - These interfaces enable endpoint devices to communicate specific task requirements directly to the network, encompassing connectivity, computational resources, and performance quality.
- - Enhanced Functional Layers - Beyond the conventional control and user planes, a Data and AI Plane (DAP) must be integrated. This layer will interpret task-related intents and dynamically allocate resources in real-time, ensuring tasks are fulfilled swiftly and efficiently.
- - AI-Integrated Procedures - Decision-making processes driven by AI will be deeply embedded within the core network procedures, facilitating rapid coordination between communication requirements and computational resources.
Looking further into the future, Huawei envisions two stages pivotal for transforming core networks into AI-native ecosystems. The preliminary phase involves implementing additional layers of AI functionalities aimed at enhancing customer experiences. The latter stage foresees the emergence of a next-gen AI-native core network that encompasses what Huawei describes as "expanded connectivity" – where AI-driven services are empowered to ensure task completion, and "beyond connectivity" – advancing past mere communication to deliver computing services and facilitate agent collaboration. Such transformative shifts necessitate a complete overhaul of network architecture, focusing on interaction mechanisms, data architecture, and computation frameworks.
Moreover, the transition to AI-native networks demands extensive collaboration across the entire ICT ecosystem, including endpoint devices, integrated circuits, and network infrastructure. By fostering cohesive interactions between endpoint devices and networks, alongside an advanced architectural framework, AI-native core networks will not only handle novel types of connections and services, but will also unlock new business opportunities in the age of intelligent technology.
Topics Telecommunications）
【About Using Articles】
You can freely use the title and article content by linking to the page where the article is posted.
※ Images cannot be used.
【About Links】
Links are free to use.