Spectrum sharing for LTE and WiFi coexistence using decision tree and game theory

In this paper, a spectrum sharing scheme adapted for the operation of Long Term Evolution (LTE) in unlicensed band (LTE-U) and WiFi system is proposed. This scheme optimizes the resource in unlicensed band based on decision tree learning and repeated game. We consider a flexible architecture decoupling control plane from data plane in the coexistence system. In control plane, strong ability of data processing is provided. Controllers learn the latest data set in the pool to build decision trees. According to the decision tree, controllers deduce the network status of the opponent. Repeated game is played to share spectrum resources for maximizing the resource utilization in the coexistence system. At the beginning of each game stage, controllers use decision tree to make informed decisions quickly. In repeated game, an encouragement mechanism is proposed to stimulate operators to share spectrum resources. Simulation results show that dynamic spectrum sharing is able to improve throughput and spectrum efficiency in unlicensed band. At the same time, this scheme also balances the traffic load and guarantees fairness among multiple systems.