Summary from listing
The employer is seeking a Master's student to develop an action-conditioned world model for downlink link adaptation in AI-native 5G/6G radio access networks. The thesis will characterize and preprocess radio and baseband trace data, train a latent world model to predict throughput, block error rate, channel quality, and spectral efficiency, evaluate its prediction accuracy, integrate it into an offline reinforcement-learning pipeline, compare different link-adaptation approaches, and document the results. The role requires a relevant Master's program, machine learning and data analysis skills, Python and PyTorch experience, wireless communications knowledge, and the ability to work with sequential data and reproducible experiments.
