PräKIpio

5G campus networks have high reliability and latency requirements, especially in industry 4.0 environments. Therefore, planning is a major challenge, especially when it comes to accurate predictions of expected service levels. Especially interference needs to be taken into account to guarantee highly reliable network operation.

The projects goal is to support and simplify the planning, operation and optimization of 5G campus networks with NR MIMO. For this purpose, methods of machine learning (ML), in particular artificial intelligence (AI), are used for radio propagation prediction and error analysis. The project has two research goals:

• The prediction of radio coverage for the planning and analysis of 5G campus networks using AI/ML-based prediction models, which can make predictions on essential characteristics of radio coverage in planning tools through additional information.
• The automated analysis of failures in 5G campus networks by blending measurements and predictions with signalization protocols.

The HHI is in charge of the design and implementation of AI/ML methods. These are used to support the prediction of expected radio channels and interferences. Radio channel modeling, the simulation of 4G/5G radio systems on link and system level and the physically precise replication of 3GPP-compliant radio systems in software are special skills at HHI that are built on years of experience. For the training of the AI/ML methods, a large amount of data is required, which can be provided by software simulations. In addition, the HHI will operate and further expand the infrastructure on the 5G-Berlin campus.