With the growing interest in satellite networks, satellite-terrestrial integrated networks (STINs) have gained significant attention because of their potential benefits. However, due to the lack of a tractable network model for the STIN architecture, analytical studies allowing one to investigate the performance of such networks are not yet available. In this work, we propose a unified network model that jointly captures satellite and terrestrial networks into one analytical framework. Our key idea is based on Poisson point processes distributed on concentric spheres, assigning a random height to each point as a mark. This allows one to consider each point as a source of desired signal or a source of interference while ensuring visibility to the typical user. Thanks to this model, we derive the probability of coverage of STINs as a function of major system parameters, chiefly path-loss exponent, satellites and terrestrial base stations' height distributions and density, transmit power and biasing factors. Leveraging the analysis, we concretely explore two benefits that STINs provide: i) coverage extension in remote rural areas and ii) data offloading in dense urban areas.

翻译：随着卫星网络日益受到关注，卫星-地面集成网络（STINs）因其潜在优势而备受瞩目。然而，由于缺乏适用于STIN架构的可处理网络模型，目前尚无法开展相关分析研究来评估这类网络的性能。本文中，我们提出了一种统一网络模型，将卫星网络和地面网络联合纳入同一分析框架。我们的核心思想基于分布在同心球面上的泊松点过程，并为每个点赋予随机高度作为标记。这使得每个点既可作为期望信号源也可作为干扰源，同时确保对典型用户的可见性。借助该模型，我们推导出STIN覆盖概率作为主要系统参数（包括路径损耗指数、卫星与地面基站的高度分布及密度、发射功率和偏置因子）的函数。基于该分析，我们具体探索了STIN提供的两大优势：i) 偏远农村地区的覆盖扩展，以及ii) 密集城区的数据卸载。