Price Learning in Joint Resource Allocation and Power Control Game in Heterogeneous Cellular Networks

In the next generation heterogeneous cellular networks, small cells, such as D2D and Relay, are commonly deployed in a macro cell. Interference coordination between the small cells and the macro cell are usually tackled by effective resource allocation and power control. Stackelberg game model can be used to formulate the resource allocation and power control problem into two-step games considering the interference between different cell tiers. Within the leader and follower games, deciding an appropriate price value is difficult. Thus, a price learning method based on reinforcement learning is proposed for the Stackelberg game model. Then, a resource allocation and power control algorithm is derived from this Stackelberg game model. The simulation results can demonstrate the performance advantages of the proposed algorithm compared with two other benchmarks.