I. Introduction
Over the past few years, with the development of mobile communication technology, especially the commercialization of 5G in many countries around the world, traditional terrestrial networks have experienced an explosive growth in terms of throughput and user access amount. Nevertheless, with the advance of the Internet of Things (IoT), relying solely on the ground network can no longer meet the exploding requirements on high-speed and reliable network access at anytime and anywhere on the earth, in consideration of its limited coverage and network capacity. This has aroused widespread concern in the academia and industry on the air-ground coordination and the space-air-ground integrated network (SAGIN) [1], [2]. In particular, many telecommunications companies around the world have released their plans for SAGIN development in the coming years. For example, in 2020, China Unicom launched their SAGIN construction plan and began to promote the maturity of the SAGIN industry chain. In academia, Liu et al. [1] reviewed recent works on SAGIN ranging from network design and resource allocation to performance analysis and optimization, and presented some challenges and future research directions. Kato et al. [3] focused on the utilization of artificial intelligence (AI) in optimizing the performance of SAGIN including intelligent traffic control, intelligent Resource Allocation, Smart Anomaly Detection, etc. After that, Cheng et al. [4] developed a SAGIN simulation platform, which adopts centralized and decentralized controllers to optimize network functions, such as access control and resource scheduling.