Coordinated Machine Learning for Energy Efficient D2D Communication

We address the problem of a coordination among machine learning tools solving different problems of radio resource management. We focus on energy efficient device-to-device (D2D) communication in a scenario with many devices communicating adhoc directly with each other. In such scenario, deep neural network (DNN) is a convenient tool to predict the channel quality among devices and to control the transmission power. However, addressing both problems by a single DNN is not suitable due to a dependency of the power control on the predicted channel quality. Similarly, a simple concatenation of two DNNs leads to a high cumulative learning error and an inevitable performance degradation. Hence, we propose a mutual coordination of the DNNs for channel quality prediction and for power control via a feedback and a knowledge transfer to mitigate the accumulation of errors in individual learned models. The proposed coordination improves the energy efficiency by 10-69% compared to state-of-the-art works and reduces the training time of DNNs more than 3.5-times compared to DNNs without coordination.