From the course: AI for Telecom: Network Optimization and Security in 5G/Edge Systems

Unlock this course with a free trial

Join today to access over 24,600 courses taught by industry experts.

AI architecture for telecom

AI architecture for telecom

- [Narrator] AI isn't just a tool. It's becoming part of the telecom backbone. In this session, we'll break down how to design an AI that fits, scales, and delivers real impact across your network. Let's take a look at the reference architecture first. There are three layers of it. The application layer sits on top. Then we have the platform layer, and then the host layer where everything sits. You can have different apps in the application layer. If it's on the RAN side, you can have xApps, rApps. You can also use CORE and BSS apps, GenAI Bots, AIOPs, and SON extensions, which are then connected to the platform layer through APIs. That platform layer comprise of the NWDAF, non realtime RIC, machine learning pipelines, and data lakes, mainly. The host layer is the infrastructure layer. You can either have it on the cloud or you can have it on your own data center. If we go deep dive on the application layer, these are the components which actually runs the application logic, such as…

Contents