Game theoretical method for sum-rate maximization in full-duplex Massive MIMO Heterogeneous Networks

• We modeled a two-layer HetNets with Massive MIMO applied. The system works in full-duplex mode. Even full-duplex leads a huge interference, it is a more efficient system.
• The game theoretical method is chosen to decline the effect of interference. A game was built base on this scenario.
• By applying the game theoretical method, the maximum system sum rate is reached. In the experiment game theoretical method indeed perform better (higher sum rate) than random access method. It performed a better access scheduling in a dynamic system.
• It turns out that users׳ power constraint (UL transmission and DL receiving constraints) is the major factor which limits system׳s performance. And compared to non-optimized model, game theoretical method has a better antiinterference ability.

In this paper, we consider Massive MIMO in two-layer Heterogeneous Networks. The system has a large self-interference and co-channel interference due to operation in full-duplex mode. By using Game Theory, an optimized sum-rate is achieved. We investigate the potential sum-rate before and after the optimization under the power constraints. It is shown that the game theoretical method performed a very good access scheduling. Compared to non-optimized model, game theoretical method can achieve higher sum-rate.