The paper presents a scalable RF data generation framework which aims to address the challenges of limited data generation/testing platforms for spectrum sensing systems. The proposed framework combines simulation and real-world data generation methods to enable large and diverse data sets for training and testing RF ML models and spectrum sensing systems. The framework includes modules for metadata generation which allows for easy experimentation. The effectiveness of the proposed framework is demonstrated through experiments including signal detection and modulation classification. This paper contributes to the development of a comprehensive framework for generating RF IQ data with ease that can significantly reduce the development and deployment time of complex wireless systems.
Citation and Bibtex
Sankar, H. P., Subbaraman, R., Hu, T., & Bharadia, D. (2024). RFSynth: Data generation and testing platform for spectrum information systems. In 21st USENIX Symposium on Networked Systems Design and Implementation (IEEE DySPAN 24).
@inproceedings{sankar2024rfsynth, title={RFSynth: Data generation and testing platform for spectrum information systems}, author={Sankar, Hari Prasad and Subbaraman, Raghav and Hu, Tianyi and Bharadia, Dinesh}, booktitle={IEEE International Symposium on Dynamic Spectrum Access Networks (IEEE DySPAN 24)}, year={2024} }
