Online Energy-Efficient Power Control in Wireless Networks by Deep Neural Networks

Alessio Zappone (1) , Merouane Debbah (2) , Zwi Altman (3)

1 Institute for Cartography - Instituts für Kartographie [Dresden]
2 L2S - Laboratoire des signaux et systèmes
3 Orange Labs [Issy les Moulineaux]

The work describes how deep learning by artificial neural networks (ANNs) enables online power allocation for energy efficiency maximization in wireless interference networks. A deep ANN architecture is proposed and trained to take as input the network communication channels and to output suitable power allocations. It is shown that this approach requires a much lower computational complexity compared to traditional optimization-oriented approaches, dispensing with the need of solving the optimization problem anew in each channel coherence time. Despite the lower complexity, numerical results show that a properly trained ANN achieves similar performance as more traditional optimization-oriented methods.

Domains: Computer Science [cs]

Alessio Zappone, Merouane Debbah, Zwi Altman. Online Energy-Efficient Power Control in Wireless Networks by Deep Neural Networks. 19th International Workshop on Signal Processing Advances in Wireless Communications (SPAWC 2018), Jun 2018, Kalamata, Greece. ⟨10.1109/SPAWC.2018.8445857⟩. ⟨hal-01962086⟩