Transfer Learning for Tilt-Dependent Radio Map Prediction

Fifth generation wireless networks (5G) will face key challenges caused by diverse patterns of traffic demands and massive deployment of heterogeneous access points. In order to handle this complexity, machine learning techniques are expected to play a major role. However, due to the large space of parameters related to network optimization, collecting labeled data to train models for all possible network configurations can be prohibitive. In this paper, we analyze the possibility of transferring knowledge, in which a machine learning model trained on a particular network configuration is used to predict a quantity of interest in a new, unknown setting. We focus on the tilt-dependent received signal strength maps as quantities of interest and we analyze the case where the knowledge acquired for a particular antenna tilt setting is transferred to a different tilt configuration of the same antenna. We test our algorithm in two different scenarios (i) where training data is obtained from real measurements and (ii) where training data is artificially generated. Promising results showing the suitability of our transfer learning approach are obtained through extensive experiments on a real dataset.