A parallel and distributed simulation method for ultra-dense satellite internet protocols

With the rapid advancement of satellite technology, ultra-dense satellite Internet networks are progressively emerging as a pivotal component of non-terrestrial networks, leveraging their potential for wide-area coverage and high-throughput transmission. However, such networks are characterized by considerable node scale, highly dynamic topology, and complex protocol interaction, posing substantial challenges to achieving efficient and accurate network simulation and protocol performance evaluation. To address these issues, this paper presents a parallel and distributed simulation method based on the lightweight simulation platform UltraStar. Designed for large-scale dynamic topologies and complex protocol interactions, the method employs parallel and distributed computing to significantly enhance the simulation efficiency of satellite Internet networks. Furthermore, we systematically evaluate the performance of different transport layer protocols and congestion control algorithm combinations in dynamic satellite network environments, analyzing their adaptability and bottlenecks across various scenarios. Simulation results demonstrate that the proposed parallel and distributed simulation method effectively supports the simulation of satellite Internet networks with tens of thousands of nodes. Compared to existing algorithms, this method significantly enhances simulation efficiency and demonstrates superior speedup and scalability. Moreover, the distinct performance of various transport layer and congestion control combinations across different scenarios uncovers key performance trends, offering a valuable reference for protocol selection and systematic optimization.