AI-native RAN：营运商和供应商的框架

AI-native RAN有可能改变行动网路的经济性，但现在必须做出艰难的决定。

虽然 AI 开始被引进 RAN 以提高自动化和智慧化程度，但业界已经制定了一个雄心勃勃的愿景，即实现AI-native RAN，其中 AI 可以嵌入到行动网路的每个元素中。这有可能改变 5G 的经济效益、提供新服务并改善客户体验。然而，解决方案和生态系统仍在发展，对营运商制定策略提出了挑战。

基于新营运商研究，本报告概述了人工智慧原生平台的关键驱动因素和挑战。它还开发了AI-native RAN 的分类，并将其映射到主要活跃参与者，包括供应商和营运商。该框架帮助利害关係人了解他们在新兴市场中的地位并发现合作机会。

目录

• 什么是AI-native RAN，推动其采用的因素有哪些？
• AI-native RAN 的关键要素
  • 谁在领导这些元素的发展以及何时可用它们？
• 哪些供应商和其他利害关係人形成有助于加速部署的平台和生态系统？
• 嵌入式 AI 可以且应该支援的RAN 功能
• AI 处理能力将在网路中配置在哪里，以及关键的架构决策是什么？

"The AI-native RAN could transform the economics of mobile networks but challenging decisions must be made now."

AI is beginning to be introduced to the RAN to increase automation and intelligence, but the industry has set out an ambitious vision of an AI-native RAN in which AI can be embedded into every element of the mobile network. This has the potential to transform the challenging economics of 5G, enable new services and improve customer experiences. However, the solutions and ecosystem are nascent, which makes it challenging for telecoms operators to plan their strategies.

This report sets out the main drivers and challenges in the AI-native platform, based on new operator surveys. It creates a taxonomy of the AI-native RAN and maps this against the main active players, including vendors and operators. The framework enables stakeholders to understand their place in the emerging market and identify alliances.

Questions answered:

• What is the AI-native RAN and what are the drivers for its adoption?
• What are the main elements of an AI-native RAN? Who is leading the development of these elements and when will they be commercially available?
• Which vendors and other stakeholders are forming platforms or ecosystems and will these help to accelerate deployability?
• Which RAN functions can or should be supported with embedded AI?
• Where would the AI processing capability be located in the network and what are the main architectural decisions?

Who should read this report:

• Heads of strategy and technology within vendor companies in the RAN equipment, RAN software, AI platforms, AI models and data, and semiconductors sectors.
• CTO office and heads of network or data strategy within operators, especially those that aim to establish a roadmap for RAN AI and for virtualised RAN within the next few years.
• CEOs and CTOs within start-up companies that are focused on RAN AI.
• Leaders of standards groups or industry alliances that are working on RAN AI.