About RASPNet

News and Updates

• July 19, 2024: Read the RASPNet press release here.
• June 12, 2024: The RASPNet website is live.

What is RASPNet

RASPNet is a large-scale dataset of airborne radar scenarios from across the contiguous United States. It was created to support the development and benchmarking of data-driven models within the radar community. There are 100 airborne radar scenarios in RASPNet, each consisting of 10,000 clutter realizations, which can be used for radar algorithm development and evaluation.

The motivation behind RASPNet stems from the absence of a publicly available dataset encompassing the diversity of scenarios and conditions that radar systems encounter. Many existing datasets are either too narrow in scope, or are encumbered by restrictions that limit their accessibility and utility for the broader research community in adaptive radar processing. RASPNet overcomes these challenges by providing a concrete benchmark on which practitioners can evaluate their algorithms, while providing a means to train data-driven methods for effective generalization to real-world radar scenarios.

Research Team

• Shyam Venkatasubramanian, Duke University
• Bosung Kang, University of Dayton
• Ali Pezeshki, Colorado State University
• Muralidhar Rangaswamy, Air Force Research Laboratory
• Vahid Tarokh, Duke University

Citations and Publications

• S. Venkatasubramanian, B. Kang, A. Pezeshki, M. Rangaswamy, and V. Tarokh. RASPNet: A Benchmark Dataset for Radar Adaptive Signal Processing Applications. arXiv preprint arXiv:2406.09638. pdf | bibtex | arXiv

Support

This work was supported in part by the Air Force Office of Scientific Research (AFOSR) under award FA9550-21-1-0235. Dr. Muralidhar Rangaswamy and Dr. Bosung Kang were supported by the AFOSR under project 20RYCORO51. The authors are grateful to Dr. Erik Blasch at AFOSR for his continued support, which has made this effort possible.