Deep Learning For Robust Power Control For Wireless Networks

Robust optimization is an important task in wireless communications, because due to fading and feedback delay there is inherent uncertainty in channel state information in a wireless environment. This paper aims to show that a deep learning approach for network utility maximization can produce more robust solutions than the traditional model-based approach. We focus on the classic power control problem for sum-rate maximization in a wireless network with multiple interfering links. By injecting samples of random channel realizations into the unsupervised training process, the neural network is able to learn to adapt to the uncertain channel environment.