Search of RRATs on declinations from $+42^{\circ}$ to $+55^{\circ}$ with a neural network

TL;DR

This study employed a neural network to search for rotating radio transients in a specific sky region, successfully identifying new RRATs and known pulsars, and developed a detection scheme for low S/N pulses.

Contribution

It introduces a neural network-based method for RRAT detection over a large sky area and presents a new scheme for identifying low S/N pulses.

Findings

Detected 15 known pulsars and 3 new RRATs.

Characterized the new RRATs' dispersion measures and pulse widths.

Developed a scheme for detecting pulses below traditional S/N thresholds.

Abstract

In an area of 3,300 square degrees, a search for pulsed dispersed signals using a neural network has been carried out. During the six-month observation period, pulses were detected from fifteen known pulsars as well as three new rotating radio transients (RRATs). The main characteristics of these new sources were provided. The dispersion measures of the transients and the half-widths of the pulses ranged from 7.2 to 59.9 pc/cm$^3$ and from 20 to 300 ms, respectively. A RRAT search scheme has been developed that allows for the detection of pulses with a signal-to-noise ratio (S/N) below the threshold required for reliable detection.