Comparison of Radio-Frequency Interference Mitigation Strategies for Dispersed Pulse Detection

Hogden, John; Wiel, Scott Vander; Bower, Geoffrey C.; Michalak, Sarah; Siemion, Andrew; Werthimer, D.

doi:10.1088/0004-637x/747/2/141

Cited by 11 publications

(12 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This imposes the requirement to clean the data of RFI before dedispersion. Examples of RFI excision techniques for FRB searches have been described in Hogden et al (2012). For our survey, removing RFI in real-time imposes specific constraints both on the algorithm (only information from the past data is available) and on the processing.…”

Section: Spectral Processing and Rfi Excisionmentioning

confidence: 99%

Limits on fast radio bursts at 145 MHz with artemis, a real-time software backend

Karastergiou¹,

Chennamangalam²,

Armour³

et al. 2015

Mon. Not. R. Astron. Soc.

101

104

View full text Add to dashboard Cite

Fast Radio Bursts (FRBs), are millisecond radio signals that exhibit dispersion larger than what the Galactic electron density can account for. We have conducted a 1446 hour survey for Fast Radio Bursts (FRBs) at 145 MHz, covering a total of 4193 sq. deg on the sky. We used the UK station of the LOFAR radio telescope -the Rawlings Array -, accompanied for a majority of the time by the LOFAR station at Nançay, observing the same fields at the same frequency. Our real-time search backend, ARTEMIS, utilizes graphics processing units to search for pulses with dispersion measures up to 320 cm −3 pc. Previous derived FRB rates from surveys around 1.4 GHz, and favoured FRB interpretations, motivated this survey, despite all previous detections occurring at higher dispersion measures. We detected no new FRBs above a signal-to-noise threshold of 10, leading to the most stringent upper limit yet on the FRB event rate at these frequencies: 29 sky −1 day −1 for 5 ms-duration pulses above 62 Jy. The non-detection could be due to scatter-broadening, limitations on the volume and time searched, or the shape of FRB flux density spectra. Assuming the latter and that FRBs are standard candles, the non-detection is compatible with the published FRB sky rate, if their spectra follow a power law with frequency (∝ ν α ), with α +0.1, demonstrating a marked difference from pulsar spectra. Our results suggest that surveys at higher frequencies, including the low frequency component of the Square Kilometre Array, will have better chances to detect, estimate rates and understand the origin and properties of FRBs.

show abstract

Section: Spectral Processing and Rfi Excisionmentioning

confidence: 99%

Limits on fast radio bursts at 145 MHz with artemis, a real-time software backend

Karastergiou¹,

Chennamangalam²,

Armour³

et al. 2015

Mon. Not. R. Astron. Soc.

101

104

View full text Add to dashboard Cite

show abstract

“…More details on each of these steps can be found in Lorimer and Kramer (2004) and Lyon et al (2016) and references therein. The first major step in the pulsar search pipeline after data digitization, channelization, RFI excision and 'clipping' (Hogden et al, 2012), involves de-dispersion, which corrects for the frequency dependent delay the incoming signals experience due to the presence of free electrons in the interstellar medium. The amount of dispersion is governed by the integrated electron density along the line of sight, also known as the dispersion measure (DM).…”

Section: Pulsar Detectionmentioning

confidence: 99%

Separation of pulsar signals from noise using supervised machine learning algorithms

Bethapudi

Desai

2018

Astronomy and Computing

View full text Add to dashboard Cite

We evaluate the performance of four different machine learning (ML) algorithms: an Artificial Neural Network Multi-Layer Perceptron (ANN MLP), Adaboost, Gradient Boosting Classifier (GBC), and XGBoost, for the separation of pulsars from radio frequency interference (RFI) and other sources of noise, using a dataset obtained from the post-processing of a pulsar search pipeline. This dataset was previously used for the cross-validation of the SPINN-based machine learning engine, obtained from the reprocessing of the HTRU-S survey data (Morello et al., 2014). We have used the Synthetic Minority Over-sampling Technique (SMOTE) to deal with high-class imbalance in the dataset. We report a variety of quality scores from all four of these algorithms on both the non-SMOTE and SMOTE datasets. For all the above ML methods, we report high accuracy and G-mean for both the non-SMOTE and SMOTE cases. We study the feature importances using Adaboost, GBC, and XGBoost and also from the minimum Redundancy Maximum Relevance approach to report algorithm-agnostic feature ranking. From these methods, we find that the signal to noise of the folded profile to be the best feature. We find that all the ML algorithms report FPRs about an order of magnitude lower than the corresponding FPRs obtained in Morello et al. (2014), for the same recall value.

show abstract

“…Several methods have been investigated to cater for this (for example, (Hogden et al, 2012)), the major challenge being to try and remove as much RFI as possible without affecting any true astrophysical dispersed pulses which might be present in the data. We implement a simple RFI-thresholding process which limits the amount of strong, bursty RFI, whose effectiveness can be tweaked by applying different thresholding factors.…”

Section: Rfi Thresholdingmentioning

confidence: 99%

Multibeam Gpu Transient Pipeline for the Medicina Best-2 Array

Magro

Hickish

Adami

2013

J. Astron. Instrum.

View full text Add to dashboard Cite

Radio transient discovery using next generation radio telescopes will pose several digital signal processing and data transfer challenges, requiring specialized high-performance backends. Several accelerator technologies are being considered as prototyping platforms, including Graphics Processing Units (GPUs). In this paper we present a real-time pipeline prototype capable of processing multiple beams concurrently, performing Radio Frequency Interference (RFI) rejection through thresholding, correcting for the delay in signal arrival times across the frequency band using brute-force dedispersion, event detection and clustering, and finally candidate filtering, with the capability of persisting data buffers containing interesting signals to disk. This setup was deployed at the BEST-2 SKA pathfinder in Medicina, Italy, where several benchmarks and test observations of astrophysical transients were conducted. These tests show that on the deployed hardware eight 20 MHz beams can be processed simultaneously for ∼640 Dispersion Measure (DM) values. Furthermore, the clustering and candidate filtering algorithms employed prove to be good candidates for online event detection techniques. The number of beams which can be processed increases proportionally to the number of servers deployed and number of GPUs, making it a viable architecture for current and future radio telescopes.

show abstract

Comparison of Radio-Frequency Interference Mitigation Strategies for Dispersed Pulse Detection

Cited by 11 publications

References 30 publications

Limits on fast radio bursts at 145 MHz with artemis, a real-time software backend

Limits on fast radio bursts at 145 MHz with artemis, a real-time software backend

Separation of pulsar signals from noise using supervised machine learning algorithms

Multibeam Gpu Transient Pipeline for the Medicina Best-2 Array

Contact Info

Product

Resources

About