New method reveals faint patterns in radio observations
The DT-FISTA algorithm helps recover data lost to radio frequency interference.
Radio frequency interference (RFI) contaminates radio observations. This leaves masked regions in sampled images when affected samples are removed.
Replacing masked samples with zero or the mean causes spectral leakage. This leakage can obscure faint parabolic scintillation arcs in secondary spectra.
The DT-FISTA method uses a sparse Fourier representation. It estimates masked intensities from the strongest Fourier components and refines them with weaker components. Nesterov momentum extrapolation accelerates these updates.
The method uses 2D fast Fourier transforms (FFT) and pointwise operations. It requires time and memory. This is faster and uses less memory than the dense Wiener implementation.
Tests on pixel patches from Five-hundred-meter Aperture Spherical radio Telescope (FAST) observations show lower reconstruction errors. DT-FISTA suppresses mask-induced artifacts and reveals previously obscured faint arcs.
Key facts
- DT-FISTA reduces runtime and memory requirements by factors compared to the dense Wiener implementation.
- Faint arcs often have power of the dominant pulsar signal power.
- The method enables efficient interstellar medium (ISM) studies.