Anomaly Detection and RFI Classification with Unsupervised Learning in Narrowband Radio Technosignature Searches

ABSTRACT The search for radio technosignatures is an anomaly detection problem: candidate signals represent needles of interest in the proverbial haystack of radio-frequency interference (RFI). Current search frameworks find an enormity of false-positive signals, especially in large surveys, requiring manual follow-up to a sometimes prohibitive degree. Unsupervised learning provides an algorithmic way to winnow the most anomalous signals from the chaff, as well as group together RFI signals that bear morphological similarities. We present GLOBULAR (Grouping Low-frequency Observations By Unsupervised Learning After Reduction) clustering, a signal processing method that uses HDBSCAN to reduce the false-positive rate and isolate outlier signals for further analysis. When combined with a standard narrowband signal detection and spatial filtering pipeline, such as turboSETI, GLOBULAR clustering offers significant improvements in the false-positive rate over the standard pipeline alone, suggesting dramatic potential for the amelioration of manual follow-up requirements for future large surveys. By removing RFI signals in regions of high spectral occupancy, GLOBULAR clustering may also enable the detection of signals missed by the standard pipeline. INTRODUCTION Listen (BL) Initiative (Worden et al. 2017) has used various facilities, including the Robert C. Byrd Green Bank Since the work of Cocconi & Morrison (1959) and Telescope (GBT), to conduct radio-frequency searches Drake (1961), radio frequencies have comprised the most of numerous targets, ranging in scale from the planetary widely explored domain in the search for extraterrestrial (e.g., Traas et al. 2021; Franz et al. 2022) to the intelligence (SETI) due to their favorably low extinction galactic (e.g., Gajjar et al. 2021; Choza et al. 2024), for across cosmic distances and the known efficacy of unambiguously artificial signals, or "technosignatures." Choza et al. and novel, is hindered by the enormous amount of radiofrequency (2024) also employed unsupervised learning in their use interference (RFI) present at all observing of DBSCAN to cluster their false-positive event set in a bands. Since most RFI signals are themselves technosignatures, 2-dimensional feature space. A robust filtration framework is therefore Rejection) clustering, a method that leverages unsupervised critical to distinguish desired signals from RFI. learning to reduce false positives in technosignature Past GBT searches (e.g., Enriquez et al. 2017; Price searches before the spatial filter step by identifying et al. 2020) have primarily used two criteria to reject common types of RFI at high spectral resolution and removing RFI.