부산대학교 도서관

Wireless resource allocation is one of the difficulties of 6G integrated satellite-terrestrial network (ISTN), which will become one of the important feature of 6G wireless networks. Different from the traditional two-stage resource allocation method of selecting the access network type and allocating resources, this paper proposes a single-stage user-centric wireless resource allocation method for ISTN, which comprehensively gathers the user service demand and the available wireless resources per area, and executes the system-optimal integrated scheduling and allocation based on reinforcement learning. We define the important network functions and basic procedure of ISTN. Using the classical mathematic models of the components involved in wireless resource allocation, a DQN mathematic model of our proposed user-centric network resource allocation method is designed