Research On Energy Efficiency Of Cellular Networks Based On Game Theory

With the increasing of 5G operating frequency and the application of millimeter wave,the number of base stations will be improving in a rapid speed.Meanwhile,the speed requirement of transmission for 5G is higher than 4G.Thus,the full duplex and energy harvesting technology have become potential techniques to be applied in 5G base stations.In this trend,it's necessary for the 5G base station to apply both full duplex and energy harvesting technology.The increasing of base stations combing the upgration of wireless technique provides new directions for energy efficiency optimization strategy analysis.The single cell is composed of macro base stations and micro base stations.The optimization strategy of a micro base station is influenced by the interference between base stations.In this paper,the game for micro base stations' switch-off strategy in the single cell is established.After the existence of mixed strategy Nash equilibrium is proved,the real scenario base station switch-off strategy based on Q Learning is proposed.Simulation results indicate that the proposed switch-off strategy is capable of improving the energy efficiency compard to tradtional switch-off strategy.In the circumstance that the frequency resource is limited,the inter cell interference can not be easily canceled.The number of reciprocally influencing base stations is growing exponentially.The wireless network is representing the characteristic of large scale networks.In this paper,the base station transmit power control strategy between two energy harvesting time points is proposed.And the approximation based on mean field game theory is conducted on the differential game when the number of players is large.Meanwhile,a numerical algorithm sovling the mean field game Nash equilibrium is proposed.Numerical simulation results indicate that the proposed algorithm can improve the coverage performence when the network is in a low energy state and improve the energy efficiency performance of the whole network compared to traditional strategy.