01 · FEIT
Ericsson: AI RAN centralized + distributed
In het rapport How to deploy AI in mobile networks (2025) beschrijft Ericsson een AI RAN-strategie over gecentraliseerde rApps én gedistribueerde radio-site deployments voor non-real-time en real-time automation — vendorcopy, geen onafhankelijke adoptie-meting.
«Ericsson's AI RAN strategy spans both centralized (rApps) and distributed (radio site) deployments, enabling non-real-time and real-time automation and new AI use cases.»
02 · FEIT
AI-RAN: inferencing lokaal bij RAN-software
Ericsson’s Cloud RAN compute and acceleration paper stelt dat AI-inferencing in de meeste gevallen lokaal bij de RAN-software moet draaien vanwege stringente latency-eisen van RAN-protocollen — terwijl training centraal kan.
«While AI training for RAN can be centralized and executed in external data centers, AI inferencing needs in most cases to be executed locally together with the RAN software due to the stringent latency requirements imposed by the RAN protocols.»
03 · FEIT
Ericsson: purpose-built Silicon + Cloud RAN hardware
In How to deploy AI in mobile networks (2025, §5.2) beschrijft Ericsson twee hardware-architecturen voor AI-RAN: purpose-built RAN Compute with Ericsson Silicon én Cloud RAN compute op x86 van partners — vendor Silicon naast cloud-compute, geen EMF-dossier.
«Ericsson solutions support two hardware variants or architectures: purpose-built RAN Compute with Ericsson Silicon and Cloud RAN compute currently based on x86 from Ericsson ecosystem partners.»
04 · FEIT
Nokia: founding member AI-RAN Alliance
In de whitepaper AI for Radio Access Networks (2024) noemt Nokia zichzelf one of the founding members of the AI-RAN alliance met board-representatie — vendor-self-report over AI-RAN-alliantie, geen marktaandeelcijfer.
«Nokia was one of the founding members of the AI-RAN alliance and its commitment extends to representation on the board.»
05 · FEIT
3GPP TR 37.817: AI-enabled RAN intelligence
3GPP TR 37.817 v17.0.0 (2022-04) Scope definieert principles for RAN intelligence enabled by AI, een functional framework en use cases/solutions of AI enabled RAN — standaardstudie Rel-17, geen vendor-marketing.
«The present document provides descriptions of principles for RAN intelligence enabled by AI, the functional framework (e.g., the AI functionality and the input/output of the component for AI enabled optimization) and use cases and solutions of AI enabled RAN.»