Detecting radio frequency interference in radio‐antenna arrays with the recurrent neural network algorithm

• Published in: Astronomische Nachrichten Vol. 339; no. 5; pp. 358 - 362
• Main Authors: Burd, P. R., Mannheim, K., März, T., Ringholz, J., Kappes, A., Kadler, M.
• Format: Journal Article
• Language: English
• Published: Weinheim WILEY‐VCH Verlag GmbH & Co. KGaA 01.06.2018; Wiley Subscription Services, Inc
• Subjects: Algorithms; Antenna arrays; Astronomy; Correlation coefficients; Filtration; Interferometry; methods: Data analysis; methods: Numerical; Neural networks; Radio astronomy; radio continuum: General; Radio frequency; Radio frequency interference; radio lines: General; Radio signals; Radio telescopes; Recurrent neural networks; techniques: Interferometric; Training
• ISSN: 0004-6337; 1521-3994
• DOI: 10.1002/asna.201813505

Signal artifacts due to radio frequency interference (RFI) are a common nuisance in radio astronomy. Conventionally, the RFI‐affected data are tagged by an expert data analyst in order to warrant data quality. In view of the increasing data rates obtained with interferometric radio telescope arrays, automatic data filtering procedures are mandatory. Here, we present the results from the implementation of a RFI‐detecting recurrent neural network (RNN) employing long short‐term memory (LSTM) cells. For the training of the algorithm, a discrete model was used that distinguishes RFI and non‐RFI data, based on the amplitude information from radio interferometric observations with the Giant Metre‐Wave Radio Telescope (GMRT) at 610 MHz. The performance of the RNN is evaluated by analyzing a confusion matrix. The true positive and true negative rates of the network are ∼99.9 and ∼97.9%, respectively. However, the overall efficiency of the network is ∼30% because of the fact that a large amount non‐RFI data is classified as being contaminated by RFI. Matthews correlation coefficient is 0.42, suggesting that a still more refined training model is required.