Managing the complex feature set of wireless communication systems in dynamic contexts while meeting diverse service demands, including critical ones, is a challenge yet to be fully addressed.
At IDLab, we tackle this challenge by exploring innovative strategies for end-to-end wireless network management, leveraging trusted Machine Learning (ML) techniques and Digital Twin technology. Our approach involves tailoring networks to application needs, adjusting end-device behaviour based on detailed network feedback, and exploring advancements like in-band telemetry, programmable protocol stacks, and richer application-network interfaces. Additionally, we're developing intuitive network programming methods, such as AI-guided derivation of application requirements and natural language-based programming of networks using LLMs. Lastly, applications and stacks with reduced state space are taken into consideration to achieve better predictability and bring down management complexity.
Our focus includes public networks but also extends to private professional networks, where introducing and validating highly-reliable communication paradigms is crucial.
