Switching Algorithm Based On Monte Carlo-Markov Decision Under Space-Air-Ground Integrated Network

As one of the next-generation mobile network visions, the Space-Air-Ground integrated network is an inevitable trend of the future network, and research on heterogeneous network switching algorithms under the Space-Air-Ground integrated network becomes more important. Existing heterogeneous network switching algorithms usually use fixed weights of attribute to make decisions, but when the single or multiple attributes of multiple networks are too different, users will be connected to the same network. In the Space-Air-Ground integrated network, large differences in networks between heterogeneous networks, such as delay, will result in excessive load on a single network. In this paper, we proposed the Monte Carlo-Markov decision process (MC-MDP) algorithm to balance the network load of multiple networks. It can dynamically adjust the access networks of users in the system while considering the user's service requirements and network differences. Monte Carlo method is used to improve the convergence speed of the Markov decision process (MDP) algorithm. Numerical results confirm the MC-MDP can improve the bandwidth resource utilization efficiency of the heterogeneous network and the convergence speed of the MDP algorithm.