Addressing the CQI feedback delay in 5G/6G networks via machine learning and evolutionary computing

5G networks apply adaptive modulation and coding according to the channel condition reported by the user in order to keep the mobile communication quality. However, the delay incurred by the feedback may make the channel quality indicator (CQI) obsolete. This paper addresses this issue by proposing two approaches, one based on machine learning and another on evolutionary computing, which considers the user context and signal-to-interference-plus-noise ratio (SINR) besides the delay length to estimate the updated SINR to be mapped into a CQI value. Our proposals are designed to run at the user equipment (UE) side, neither requiring any change in the signalling between the base station (gNB) and UE nor overloading the gNB. They are evaluated in terms of mean squared error by adopting 5G network simulation data and the results show their high accuracy and feasibility to be employed in 5G/6G systems.