Analysis of Physical Connectivity and Cross-Layer Service Matching in User-Service-Oriented ISTN

Aiming at service-oriented design requirements for the integrated satellite–terrestrial network (ISTN), this article proposes a cross-layer analysis framework and a distributed cognitive space service network architecture. These address challenges in traditional single-layer research, including unquantified cross-layer deviations, incomplete link analysis, and a lack of multidimensional evaluation. A distributed on-orbit architecture for LEO satellites is constructed to enable cross-layer cooperation across physical-layer access, network-layer routing, and application-layer service matching. Beyond channel-fading-based analysis, a multilink model incorporating node-induced interference is established. It derives uplink access success rate expressions, quantifies impacts of user density, link distance, and carrier bands on connectivity, and verifies interference-attenuation coupling via simulations. By integrating mutual information and entropy theory, a cross-layer deviation framework is built, using confluent hypergeometric distribution to model service matching uncertainty. This achieves quantitative modeling of “physical-network layer” cooperation gains and “network-application layer” adaptation deviations. The results provide theoretical tools for optimizing space-based intelligent networks. The architecture and methods directly support enhancing large-scale satellite network quality and constructing objective functions, providing a key technical path for service-oriented future space–ground integration systems.