An apparatus, a method and a computer program for predicting a future quality of service of a wireless communication link based on a predicted future environmental model that is predicted using a time-series projection. The method includes determining environmental models of one or more active transceivers in the environment of the mobile transceiver over points in time, determining a predicted future environmental model of the one or more active transceivers at a point in time of the future using a time-series projection of environmental models, predicting the future quality of service of the wireless communication link for point in time of the future using a machine-learning model. The machine-learning model is trained to provide information on a predicted quality of service for a given environmental model, and the predicted future environmental model is used as input to the machine-learning model.