Search for: All records

Abstract We investigate the effectiveness of the statistical radio frequency interference (RFI) mitigation technique spectral kurtosis ( $\widehat{\mathit{SK}}$) in the face of simulated realistic RFI signals. $\widehat{\mathit{SK}}$estimates the kurtosis of a collection ofMpower values in a single channel and provides a detection metric that is able to discern between human-made RFI and incoherent astronomical signals of interest. We test the ability of $\widehat{\mathit{SK}}$to flag signals with various representative modulation types, data rates, duty cycles, and carrier frequencies. We flag with various accumulation lengthsMand implement multiscale $\widehat{\mathit{SK}}$, which combines information from adjacent time-frequency bins to mitigate weaknesses in single-scale $\widehat{\mathit{SK}}$. We find that signals with significant sidelobe emission from high data rates are harder to flag, as well as signals with a 50% effective duty cycle and weak signal-to-noise ratios. Multiscale $\widehat{\mathit{SK}}$with at least one extra channel can detect both the center channel and sideband interference, flagging greater than 90% as long as the bin channel width is wider in frequency than the RFI. 

We explore the statistical radio frequency interference (RFI) mitigation technique spectral kurtosis (SK) in the context of simulated realistic RFI signals. SK is a per-channel RFI detection metric that estimates the kurtosis of a collection of M power values in a single channel to discern between human-made RFI and incoherent astronomical signals of interest. We briefly test the ability of SK to flag signals with various representative modulation types, data rates, and duty cycles, as well as accumulation lengths M and multi-scale SK bin shapes. Multi-scale SK uses a rolling window to combine information from adjacent time-frequency pixels to mitigate weaknesses in single-scale SK. High data rate RFI signals with significant sidelobe emission are harder to flag, as well as signals with a 50% effective duty cycle. Multi-scale SK using at least one extra channel can detect both the center channel and side-band interference, flagging most of the signal at the expense of larger false positive rates. 

A cyclostationary process is one whose autocorrelation function is periodic or nearly periodic. The modulation schemes used to encode information give rise to cyclostationarity in many human-generated sources of interference. In contrast, nearly all astrophysical signals are expected to be wide-sense stationary on timescales of interest, making cyclostationarity a potentially robust way of discriminating between interference and astronomical sources. We are developing an algorithm that employs a well-known method of detecting cyclostationary signals and testing its efficacy against a suite of simulated interference covering a wide range of modulation schemes. We present receiver operating characteristic curves and binary classification scores for different types of interfering signals. Our algorithm performs well for many modulation schemes, with F1 and φ coefficient scores in excess of 0.9 in some cases, though it shows weaknesses in the case of frequency modulation. We also apply our algorithm to archived Robert C. Byrd Green Bank Telescope observations of a bright millisecond pulsar. We use standard pipelines for blindly detecting and timing pulsars and preliminarily find improvement in data quality according to several metrics, though some undesirable effects are still present. We also show that our algorithm has no negative impact when detecting Galactic HI emission. We thus believe that cyclostationary signal processing shows promise as a means of interference mitigation and discuss opportunities and challenges for employing it more widely.