Abstract: Network traffic characteristics and traffic models are the basis for network planning and design and performance evaluation. This paper devoted to the research on the service analysis and traffic modeling of the integrated ground-air-space network. Based on the analysis of the high-speed railway private network service of the integrated space-to-ground network, the paper abstracted the characteristics of the high-speed rail private network such as traffic type, service transmission content, packet arrival interval, and packet length, and constructed the simulated traffic data set of high-speed railway private network and analyzed its traffic characteristics. Aiming at the self-similarity, burstiness and periodicity of the high-speed railway private network traffic, the paper used the alpha-stable self-similar model to model the high-speed rail traffic, and verified the reliability of the model by the Q-Q (Quantile- Quantile) diagram. The research results showed that the model can effectively describe the traffic characteristics of the high-speed railway private network. The research results of this paper provided an important basis for the design and simulation of traffic forecasting models and the rational allocation of network resources.