Optimum Power Allocation based on Traffic Matching Service for Multi-beam Satellite System

Like IoT or 5G networks, future multi-beam satellite (MBS) network becomes denser causing the number of beams increases. In addition, the characteristics of traffic requests in the MBS network is asymmetric distribution and time-varying. In view of above condition, a significant challenge power allocation facing is matching the traffic requests while with fairness into consideration as the network scales up. In this paper, we consider optimum power allocation based on traffic demands service for multi-beam satellite system, which has been formulated as an optimization problem in pursuit of traffic matching degree and fairness maximization. The optimal solution to this optimization problem can be obtained by the conjecture-based multi-agent Q-learning power allocation (MAQ-PA) algorithm. Simulated results show the effectiveness of the proposed power allocation algorithm.