Testbed Design and Performance Emulation for Satellite-Terrestrial Integrated Networks

Area of Science:

• Network Engineering
• Telecommunications
• Computer Science

Background:

• Satellite-terrestrial integrated networks (STINs) offer extended coverage for remote IoT and sensor telemetry.
• Evaluating STINs is challenging due to complex dynamics in mobility, channels, resources, and control planes.

Purpose of the Study:

• To present a software-based modular testbed and performance-emulation framework for STINs.
• To facilitate traceable evaluation of STIN strategies for various applications.

Main Methods:

• Developed a framework integrating scenario generation, model-driven processing, replaceable algorithm engines, and scheduler-based control.
• Modeled terrestrial, UAV, and LEO satellite entities with abstracted link budgets and various network functions.
• Employed a lightweight emulation model parameterized by NTN and link-budget assumptions for strategy evaluation.

Main Results:

• Revealed tradeoffs between access, handover, routing, and scheduling strategies in STINs.
• Demonstrated sensitivity of performance to workload, mobility, outage, demand, and model parameters.
• Validated the framework's utility for analyzing mixed sensing and communication traffic.

Conclusions:

• The proposed framework supports traceable STIN strategy evaluation.
• It is suitable for remote sensor networks, sensing-data backhaul, and remote-IoT services.
• The framework aids in understanding performance under explicit emulation assumptions.