A commensal Fast Radio Burst search pipeline for the Murchison Widefield Array

TL;DR

This paper demonstrates a real-time, commensal FRB search pipeline using the Murchison Widefield Array, capable of detecting bright nearby FRBs and verifying system performance with pulsar observations.

Contribution

It introduces a new FRB search pipeline integrated with the MWA's multibeam beamformer, enabling real-time detection and verification using known pulsars.

Findings

Pipeline successfully detects known bright pulsars.

Estimated sensitivity varies with bandwidth upgrades.

Projected FRB detection rate is a few to tens per year.

Abstract

We present a demonstration version of a commensal pipeline for Fast Radio Burst (FRB) searches using a real-time incoherent beam from the Murchison Widefield Array (MWA). The main science target of the pipeline are bright nearby FRBs from the local Universe which are the best candidates to probe FRB progenitors and understand physical mechanisms powering these extremely energetic events. The new MWA beamformer, known as the "MWAX multibeam beamformer", can form multiple incoherent and coherent beams commensally to any on-going MWA observations. One of the beams is currently used for FRB searches (tested in 10 kHz frequency resolution and time resolutions between 0.1 and 100 ms). A second beam is used for a Search for Extraterrestrial Intelligence (SETI). This paper focuses on the FRB search pipeline and its verification on selected known bright pulsars. The pipeline uses the FREDDA implementation of the Fast Dispersion Measure Transform algorithm (FDMT) for single pulse searches. Initially, it was tested during standard MWA observations, and more recently using dedicated observations of selected 11 bright pulsars. The pulsar PSR J0835-4510 (aka Vela) has been routinely used as the primary probe of the data quality because its folded profile was always detected in the frequency band 200 - 230 MHz with typical SNR >10. Similarly, the low DM pulsar PSR B0950+08 was always detected in folded profile in the frequency band 140 - 170 MHz, and so far has been the only object for which single pulses were detected. We present the estimated sensitivity of the search in the currently limited observing bandwidth of a single MWA coarse channel (1.28 MHz) and for the upgraded, future system with 12.8 MHz (10 channels) of bandwidth. Based on expected sensitivity and existing FRB rate measurements, we estimate an expected number of FRB detections to be between a few and a few tens per year.