This project studies how short-term wireless channel dynamics can be predicted accurately enough to support transmit power control in wireless body area networks. The goal is to reduce energy usage while preserving reliable communication around the human body, where channel behavior can change quickly with posture and motion.
The work combines lightweight deep learning models with system-level power control. Rather than treating prediction as a standalone benchmark, the project connects prediction accuracy to communication performance and resource allocation.
Related papers include the 2019 IEEE ICC paper on deep learning channel prediction and the 2020 IEEE Internet of Things Journal paper on accurate channel prediction by lightweight deep learning.