AI as a Service for 6G–Satellite Network Architectures: Enabling Next-Generation E-healthcare care and Beyond

The integration of sixth-generation (6G) terrestrial networks with non-terrestrial satellite systems promises ubiquitous, high-capacity, and low-latency connectivity on a global scale. Artificial Intelligence as a Service (AIaaS) emerges as a pivotal enabler, offering distributed intelligence that can adapt to heterogeneous and dynamic space–ground environments. This paper proposes a multi-layer AIaaS framework for AI-native integrated 6G–satellite networks that unifies spaceborne, airborne, and terrestrial tiers via coordinated orchestration, AI-oriented network slicing, and federated learning. Beyond conventional communication services, the proposed architecture is positioned to transform the E-healthcare sector by enabling applications such as telemedicine, remote robotic surgery, real-time emergency response, and continuous monitoring through wearable devices. Simulation results under realistic LEO mobility and heterogeneous traffic conditions demonstrate up to 33.8% lower end-to-end latency, a +12.7 percentage-point increase in service continuity, routing stretch near unity, and 23.8% energy savings per inference compared to baselines. These gains underscore AIaaS as a practical enabler for SLA-driven, resilient, and health-aware AI service delivery in heterogeneous TN–NTN environments, supporting the digital transformation of the E-healthcare sector alongside other critical domains.