Joint Delay and Throughput Network Resource Orchestration in SAGIN Based on SDN

Due to the convergence of rapidly evolving terrestrial, airborne, and satellite communication networks, the space-air-ground integrated network (SAGIN) will be an important architecture to support 6G communications in the future. This architecture can coordinate different levels of networks in real time and dynamically provide efficient and personalized services to users according to different situations. However, due to the complexity and heterogeneity of its architecture, it is extremely difficult to realize resource scheduling and dispatching. In this article, A reconfigurable SAGIN architecture integrated with software defined networking (SDN) is proposed to address the above issues. To further satisfy the quality of service requirements (QoS) of users. The optimization problem of joint unmanned air vehicle(UAV) trajectory design and link selection is formulated for maximizing the ratio between downlink throughput and link delay. Then, the k-means-DDPG (K-DDPG) algorithm is applied to determine the optimal UAV trajectories and links. Finally, simulation results show that the adopted algorithm has good convergence. The performance of the adopted algorithm is significantly better than other algorithms.