Cognitive Radio for Satellite TT & C System: A General Dataset Using Software-defined Radio

Area of Science:

• Satellite Communications
• Cognitive Radio
• Deep Learning

Background:

• Satellite communication systems require adaptive technologies like cognitive radio.
• Existing datasets lack space-environment signal features, hindering research.
• There is a critical need for specialized datasets for satellite signal processing.

Purpose of the Study:

• Introduce RML24, the first open-source benchmark dataset for deep learning in satellite Telemetry, Tracking, and Command (TT&C) signal processing.
• Provide a realistic dataset that bridges the gap between terrestrial and space communication signal characteristics.
• Facilitate advancements in intelligent, adaptive satellite systems through accessible data and code.

Main Methods:

• Generated over 1.3 million signal samples using Software Defined Radio (SDR) and Radio Frequency (RF) platforms.
• Included 22 modulation schemes common in TT&C systems.
• Incorporated simulated satellite channel models, RF link effects, and propagation phenomena for realism.

Main Results:

• RML24 is the first open-source dataset tailored for deep learning in satellite TT&C signal processing.
• The dataset features realistic signal characteristics, including channel models and RF effects.
• Comprehensive annotations support deep learning tasks like parameter estimation and demodulation.

Conclusions:

• RML24 provides a high-quality, specialized resource to accelerate deep learning research in satellite communications.
• The dataset enables the development of more adaptive and intelligent cognitive radio systems for space.
• Public availability of RML24 and associated code promotes broader research and innovation in the field.