Resource optimization for small-cell wireless networks is more complicated than the traditional applications. The solution needs to be delivered promptly to respond the highly dynamic temporal and spatial variations. It seems that the machine learning strategy is more flexible and adaptive than the conventional optimization methods, since ML has the potential to find the implicit function relationship between arbitrary input data and output results. In this work, we focus on a generic D2D network and to show the effectiveness of ML apply to solve the power optimization problem with different optimization models. The research spans over all stages such as analysis, design, implementation, and validation. It is shown that the ML method has achieved several benchmarks in terms of QoS metrics for different optimization models.

Deep learning, Machine learning, Neural networks, Optimization, Wireless communications
doi.org/10.1109/GLOBECOM38437.2019.9013762
2019 IEEE Global Communications Conference, GLOBECOM 2019
Department of Systems and Computer Engineering

Zhou, J. (Junxiu), Liu, X. (Xian), & Huang, C. (2019). Machine learning for power allocation of a D2D network. In 2019 IEEE Global Communications Conference, GLOBECOM 2019 - Proceedings. doi:10.1109/GLOBECOM38437.2019.9013762