The goal of this project is to design methods based on machine learning (ML) and artificial intelligence (AI) for identifying and eliminating radio frequency interference (RFI) from the raw data collected by radio telescopes. RFI are typically human-made radio signals from terrestrial communication and broadcasts. They appear mostly as high-intensity signals at specific frequency channels, but broadband RFI are also found.
The project began in 2024 when a group of undergraduate and postbac students led by Adrita Khan planned to write a review paper on the use of ML in dealing with RFI. The review paper is currently in preparation and it will be submitted to New Astronomy Reviews once completed, inspired by Ndung’u et al. (2023). The review paper might include benchmarking of the most widely used ML models on some representative datasets.
In future we plan to develop our own model for the identification and elimination of RFI.
People
- Khan Asad, PhD, supervisor.
- Adrita Khan, Research Intern, CASSA & CCDS, AMCS MSc Student, NSU.
- Fariha Tanjim Shifat, R&D Engineer, Penta Global Ltd.
- MD Sajin, CSE Graduate, BRAC University.
- Prantik Paul, CSE Lecturer, BRAC University.
- Hasin Rayhan, BSc student in EEE, Dhaka University.
- Fatema Akter, MSc student, North Dakota State University, USA.
