Radio Frequency Interference Detection Algorithms Research Articles

Lightweight deep neural network for radio frequency interference detection and segmentation in synthetic aperture radar

Radio frequency interference (RFI) poses challenges in the analysis of synthetic aperture radar (SAR) images. Existing RFI suppression systems rely on prior knowledge of the presence of RFI. This paper proposes a lightweight neural network-based algorithm for detecting and segmenting RFI (LDNet) in the time-frequency domain. The network accurately delineates RFI pixel regions in time-frequency spectrograms. To mitigate the impact on the operational speed of the entire RFI suppression system, lightweight modules and pruning operations are introduced. Compared to threshold-based RFI detection algorithms, deep learning-based segmentation networks, and AC-UNet specifically designed for RFI detection, LDNet achieves improvements in mean intersection over union (MIoU) by 24.56%, 13.29%, and 7.54%, respectively.Furthermore, LDNet reduces model size by 99.03% and inference latency by 24.53% compared to AC-UNet.

Development of a Dynamically Re-Configurable Radio-Frequency Interference Detection System for L-Band Microwave Radiometers.

Real-Time RFI Detection and Flagging (RT-RDF) for microwave radiometers is a versatile new FPGA algorithm designed to detect and flag Radio-Frequency Interference (RFI) in microwave radiometers. This block utilizes computationally-efficient techniques to identify and analyze RF signals, allowing the system to take appropriate measures to mitigate interference and maintain reliable performance. With L-Band microwave radiometry as the main application, this RFI detection algorithm focuses on the Kurtogram and Spectrogram to detect non-Gaussian behavior. To gain further modularity, an FFT-based filter bank is used to divide the receiver's bandwidth into several sub-bands within the band of interest of the instrument, depending on the application. Multiple blanking strategies can then be applied in each band using the provided detection flags. The algorithm can be re-configured in the field, for example with dynamic integration times to support operation in different environments, or configurable thresholds to account for variable RFI environments. A validation and testing campaign has been performed on multiple scenarios with the ARIEL commercial microwave radiometer, and the results confirm the excellent performance of the system.

A comparison framework for deep learning RFI detection algorithms

ABSTRACT In the realm of radio astronomy, the detection of radio frequency interference (RFI) stands as a pivotal pursuit. This study performs a novel comparative analysis of deep learning methodologies and introduces a novel transfer learning method, called fine-tuning. We compare various aspects and factors relating to this problem, including supervised fully convolutional network (FCN) architectures used within the literature, loss functions, regularization techniques, and training methodologies, to establish the most effective strategies for RFI detection. Moreover, the relationship between parameters, FLOPS, and inference times are examined. Fine-tuning involves pre-training models with low-quality aoflagger reference outputs, a very popular and accessible RFI flagging software package, and thereafter re-training the models with high-quality reference outputs. We utilize two data sets: real observations from LOFAR and simulated data from HERA. The mean squared error (MSE) loss function emerges as a robust performer if a high recall is desired. In contrast, the binary cross-entropy (BCE) loss function excels in generalization but falls short in classification performance. The Dice loss function emerges as the top performer, maximizing the F1 score and thereby serving as the choice for our further investigations. Notably, we highlight the important role of data quality and model capacity. In particular, we find that low-capacity models exhibit resilience when trained with low-quality flags from aoflagger, showcasing their ability to mitigate overfitting and overflagging tendencies. In contrast, high-capacity models excel when trained with high-quality flags. Fine-tuning proved to be an effective method to unlearn the overflagging tendencies of aoflagger, whilst requiring very little data.

Evaluation of Application Effectiveness on Ocean Salinity Satellite RFI Detection Algorithms

In order to alleviate the impact of radio frequency interference (RFI) on the accuracy of ocean salinity satellite remote sensing, scholars have proposed various detection and labeling algorithms for RFI based on remote sensing data from the SMOS satellite. However, the signals that generate RFI are diverse, and the factors that influence remote sensing observation data are complex. Existing algorithms often target specific hypothetical conditions, lacking general applicability, which frequently leads to an important gap between the nominal performance of the literature and practical applications, posing great challenges to data labeling work. To address this problem, this study conducted a comprehensive and systematic analysis of RFI simulation based on scene modeling, algorithm modeling, and RFI energy modeling. Three typical RFI detection algorithms were selected, and the simulation scene was divided into 3 typical scenes: ocean, land, and sea–land scenes, and RFI was analyzed in terms of weak, moderate, strong, and extremely strong based on energy. Through simulation analysis and evaluation of RFI detection algorithms, lookup tables for algorithm selection, detection rate, and false-positive rate have been established for different intensities of independent RFI sources and multiple nearby RFI sources in the above scenario. These lookup tables have universal guiding significance and provide reliability assurance in complex situations.

Radio frequency interference detection and mitigation in the DWD C-band weather radar network

Abstract. C-band weather radar data are commonly compromised by interference from external sources even though weather radars are the primary and therefore privileged user of this frequency band. This is also the case for the radar network of the German Meteorological Service (Deutscher Wetterdienst, DWD). Theoretically, dynamic frequency selection (DFS) by devices operating in the C band should prevent any disturbance of the primary user. In practice, this does not always work as intended under the current regulations. As it is not possible to separate a superimposed interference signal from measured weather radar data, the protection of the frequency band is of utmost importance. Currently, the only available option is to discard the compromised portions of the radar data. Therefore, the current best course of action is to shut down radio frequency interference (RFI) sources as quickly as possible. The automated RFI detection algorithm for the German C-band weather radar network, operational since July 2017, makes use of routinely measured radar moments. Built upon data gathered since 2017, an RFI classification with respect to the severity and duration of RFI sources was first implemented in 2019. An independent verification of the RFI detection algorithm was performed by using a commercially available Wi-Fi adapter, which is directly integrated into the radar receiver. Subsequently, a mitigation workflow was implemented to efficiently identify and shut down detected RFI sources by the German Federal Network Agency (Bundesnetzagentur, BNetzA). By following this workflow with great effort, the number of persistent RFI sources has been decreasing since October 2019, while a steady increase in short-lived RFI sources over the last 5 years also exists. In total, 11 889 RFI sources have been identified from July 2017 to May 2022. Most of these (94.8 %) are such short-lived sources that an unambiguous identification by the BNetzA is, in general, not feasible. However, as stated by the C-band regulations, any non-compliant transmitter compromising the operation of a weather radar has to be shut down. This is important, as even these short-lived RFI sources negatively affect meteorological product generation.

A Novel RFI Detection Method for Microwave Radiometers Using Multilag Correlators

Radio frequency interference (RFI) is an increasing problem in microwave radiometry, particularly for Earth observation, because the antennas are pointing toward the Earth. RFI has been observed at the L-band in European Space Agency’s (ESA’s) Soil Moisture and Ocean Salinity (SMOS) Earth Explorer mission, National Aeronautics and Space Administration’s (NASA’s) Soil Moisture Active and Passive (SMAP) and Aquarius missions, and Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) and WindSat missions at 10.7 and 18.7 GHz (Misra and de Matthaeis, 2014). Therefore, dedicated onboard systems are, nowadays, a must to detect and remove contaminated measurements, improving radiometric accuracy and increasing the spatial coverage. In this work, a novel detection technique, specially tailored for synthetic aperture interferometric radiometers (SAIRs), is proposed and its performance analyzed. It is based on the change of the shape of the cross correlation function at lags different from zero under the presence of RFI. The performance of the proposed technique is compared to other common RFI detection algorithms, and its limitations and advantages are discussed. Postcorrelation detection performance is found to outperform other commonly used algorithms, such as kurtosis. In addition, it presents some convenient properties for its practical application in correlation and real aperture radiometers.

Performance Evaluation of Radio Frequency Interference Measurements from Microwave Links in Dense Urban Cities

Radio frequency interference (RFI) constitutes a significant problem in achieving a good quality of service in radio links. Several techniques have been proposed to identify and mitigate RFI in wireless networks. However, most of these techniques are not generalized for all propagation environments due to varying geographical features from one environment to another. The need for extensive frequency scan measurements on the links to identify the available channels, evaluate the performances of the links, and detect RFI in the channels becomes imperative. This study presents a performance evaluation of frequency scan measurements from active microwave links comprising eighteen base stations. The measurements equipment included a spectrum analyzer and a 0.6 m antenna dish. The frequency scans were taken at 6 GHz, 7 GHz, and 8 GHz with full azimuth coverage of the horizontal and vertical polarization. Measured data were processed to determine the available frequencies and RFI in the channels. The histogram and probability density function of the frequency scans were computed. The cumulative distribution functions were determined, and the statistical error characteristics of the frequency scans for the estimated normal distribution and the estimated fitness curve were derived. The short-time Fourier transform of the noisy signal was obtained, and the signal without noise was recovered using the inverse short-time Fourier transform. Analysis of the scanned signals before and after the noise removal is demonstrated. The denoised signals compare favorably with related results in the preliminary literature. Overall, these frequency scans would be beneficial in evaluating RFI measurements and spectrum planning and hold great promise for designing robust RFI detection algorithms for future wireless systems.

Properties of the RFI Environment at 1400–1427 MHz as Observed by the Soil Moisture Active/Passive Mission Microwave Radiometer

The soil moisture active/passive (SMAP) satellite microwave radiometer has been providing global measurements of L-band thermal emission from Earth since April 2015. Although the radiometer operates in the protected 1400-1427 MHz portion of the radio spectrum, its measurements are still corrupted by either radio frequency interference (RFI) from out-of-band emissions via legal sources or by sources operating in-band illegally. The SMAP radiometer includes a digital backend that enables implementation of multiple ground-based RFI detection and filtering algorithms. This data is used to collect statistics and trends of Earth's RFI environment. This article examines properties of the global RFI environment as observed by SMAP, including information on RFI source properties (obtained from analysis of SMAP multiple detector outputs) and the evolution of the RFI environment in time. Residual RFI contributions after the application of SMAP RFI processing are also examined as preliminary information for the development of future methods to address their effect.

Radio frequency interference mitigation based on the asymmetrically reweighted penalized least squares and SumThreshold method

ABSTRACT As radio telescopes become more sensitive, radio frequency interference (RFI) is becoming more important for interesting signals of radio astronomy. There is a demand for developing an automatic, accurate and efficient RFI mitigation method. Therefore, we have investigated an RFI detection algorithm. First, we introduce an asymmetrically reweighted penalized least squares (ArPLS) method to estimate the baseline more accurately. After removing the estimated baseline, several novel strategies were proposed based on the SumThreshold algorithm for detecting different types of RFI. The threshold parameter in SumThreshold can be determined automatically and adaptively. The adaptiveness is essential for reducing human intervention and for the online RFI processing pipeline. Applications to data from the Five-hundred-meter Aperture Spherical Telescope (FAST) show that the proposed scheme based on ArPLS and SumThreshold is superior to some typically available methods for RFI detection with respect to efficiency and performance.

An RFI Detection and Mitigation Algorithm for an L-Band Phased Array Radiometer

To monitor sea surface salinity (SSS) in coastal zones with high temporal and spatial resolution, an L-band phased array radiometer is developed. However, it is well known that the L-band radiometers suffer from serious radio frequency interference (RFI) effects. To mitigate the RFI effects, an RFI detection and mitigation algorithm is proposed for the L-band phased array radiometer. The experiments in coastal zones are performed to assess the performance of the RFI algorithm. The results indicate that the detection and mitigation algorithm exhibits good performance and can effectively mitigate the RFI effects.

Comparison of Polarimetric Algorithm With Other Algorithms in Detecting Radio Frequency Interference

Anthropogenic radio frequency interference (RFI) is an increasing problem in microwave remote sensing radiometry. Therefore, there is growing interest in RFI counteraction methods, including detection and blanking. In this study, the performance of polarimetric algorithm has been studied and compared with several other detection algorithms with respect to their capability to detect different potential types of RFI, such as continuous wave (CW), pulse modulated, and quadrature phase shift keying (QPSK) modulated. The mission scenario and parameters applied correspond to a typical space-borne, polar-orbiting, conically scanning microwave radiometer. Therefore, the results obtained help to estimate the achievable RFI detection limits (in terms of the brightness temperature) for similar future satellite instruments if polarimetry or other algorithms analyzed here were applied. The results obtained suggest that the RFI detection limit in the 1-K range can be achieved in the scenario under study. Incorporating multiple and complementary RFI detection algorithms in the overall system, e.g., in a digital RFI processor, enables a wide range of different RFI types to be detected. The results suggest that polarimetry and spectral density estimation are sensitive in detecting constant envelope signals (such as CW or QPSK), while kurtosis and spectral kurtosis are effective in detecting pulsed RFI. A detection limit as low as 0.1 K can be achieved for pulsed RFI with low duty cycles. These findings confirm and extend earlier results. To the best knowledge of the authors, this is the first quantitative study on the performance of the RFI detection algorithms to include polarimetry.

Detection and mitigation of radiometers radio-frequency interference by using the local outlier factor

ABSTRACTLarge amounts of radio-frequency interference (RFI) are present in Earth observations at the L-band frequencies of European Space Agency’s Soil Moisture and Ocean Salinity, the National Aeronautics and Space Administration’s Aquarius and Soil Moisture Active and Passive missions. Multiple approaches have been proposed to detect and eliminate the RFI signals in the past few decades, including time, statistical, polarimetric and frequency domain methods. This letter focuses on a new potential RFI detection and mitigation algorithm that is based on the local outlier factor (LOF). Experimental results show that a satisfactory performance can be obtained by an LOF algorithm even in detecting moderate RFI.

SMAP L-Band Microwave Radiometer: RFI Mitigation Prelaunch Analysis and First Year On-Orbit Observations

The National Aeronautics and Space Administration's (NASA) Soil Moisture Active and Passive (SMAP) mission, which was launched on January 31, 2015, is providing global measurements of soil moisture and freeze/thaw state. The SMAP radiometer operates within the protected Earth Exploration Satellite Service passive frequency allocation of 1400-1427 MHz. However, unauthorized in-band transmitters and out-of-band emissions from transmitters operating at frequencies adjacent to this allocated spectrum are known to cause interference to microwave radiometry in this band. Because measurement corruption by these terrestrial transmissions, which is referred to as radio-frequency interference (RFI), threatens mission success, the SMAP radiometer includes special flight hardware to enable the detection and filtering of RFI. Results from the first year of SMAP data show the presence of RFI with frequent occurrence over Asia and Europe. During the calibration/validation stage of the mission, the RFI detection and mitigation algorithms were modified to provide enhanced performance. Analysis of the L1B_TB products indicates good algorithmic performance with respect to RFI detection and removal. However, some regions of the globe (e.g., Japan) continue to experience complete data loss. This paper summarizes updates to the SMAP RFI processing algorithms based on prelaunch tests and on-orbit measurements, as well as RFI information obtained in SMAP's first year on orbit.

Low-level radio-frequency interference detection algorithm based on European Centre for Medium-Range Weather Forecasting for Soil Moisture and Ocean Salinity

At present, the Soil Moisture and Ocean Salinity (SMOS) mission is severely affected by radio frequency interferences (RFIs), and the detection of low-level RFI-contamination brightness temperatures (BTs) is still a challenge in SMOS. A low-level RFI detection algorithm is proposed, which is based on the soil surface temperature products provided by the European Centre for Medium-Range Weather Forecasting. The algorithm is analyzed in terms of RFI-flagged snapshot, RFI-flagged probability, and localization accuracy. The performance of the algorithm is demonstrated by SMOS data. The results show this algorithm can detect and flag more low-level RFI-contamination BTs and show a better performance.

Adaptive Control of Undetected Radio Frequency Interference With a Spaceborne Microwave Radiometer

In microwave radiometric remote sensing, undetected radio frequency interference (RFI) can adversely affect the accuracy of the science products. A method is presented to adaptively tune the parameters of an RFI detection algorithm which controls the equivalent brightness temperature of undetected RFI. The method is adaptive in the sense that it adjusts to variations in the RFI environment, e.g., from high RFI conditions near some population centers to low RFI conditions in the tropical Pacific Ocean. The RFI environment is characterized by inferring the distribution of low-level undetected RFI from that of high-level detected RFI using appropriate scaling arguments. The resulting tuned algorithm adjusts its detection threshold to equalize the brightness temperature calibration bias due to RFI at the expense of the now variable measurement precision (noise equivalent delta temperature). This tradeoff between calibration bias and measurement precision can be represented as a modified version of the classic receiver-operating-characteristic curve. The radiometer on the Aquarius/SAC-D mission is used as an example.

Aquarius RFI Detection and Mitigation Algorithm: Assessment and Examples

Aquarius is an L-band radiometer system designed to map sea surface salinity from space. This is a sensitive measurement, and protection from radio frequency interference (RFI) is important for success. An initial look at the performance of the Aquarius RFI detection and mitigation algorithm is reported together with examples of the global distribution of RFI at the L-band. To protect against RFI, Aquarius employs rapid sampling (10 ms) and a “glitch” detection algorithm that looks for outliers among the samples. Samples identified as RFI are removed, and the remainder is averaged to produce an RFI-free signal for the salinity retrieval algorithm. The RFI detection algorithm appears to work well over the ocean with modest rates for false alarms (5%) and missed detection. The global distribution of RFI coincides well with population centers and is consistent with observations reported by the Soil Moisture and Ocean Salinity mission.

Applications of an AMSR-E RFI detection and correction algorithm in 1-DVAR over land

Land retrievals using passive microwave radiometers are sensitive to small fluctuations in land brightness temperatures. As such, the radio-frequency interference (RFI) signals emanating from man-made microwave radiation transmitters can result in large errors in land retrievals. RFI in C- and X-band channels can contaminate remotely sensed measurements, as experienced with the Advanced Microwave Scanning Radiometer (AMSR-E) and the WindSat sensor. In this work, applications of an RFI detection and correction algorithm in retrieving a comprehensive suite of geophysical parameters from AMSR-E measurements using the one-dimensional variational retrieval (1-DVAR) method are described. The results indicate that the values of retrieved parameters, such as land skin temperature (LST), over these areas contaminated by RFI are much higher than those from the global data assimilation system (GDAS) products. The results also indicate that the differences between new retrievals and GDAS products are decreased evidently through taking into account the RFI correction algorithm. In addition, the convergence metric (χ2) of 1-DVAR is found to be a new method for identifying regions where land retrievals are affected by RFI. For example, in those regions with much stronger RFI, such as Europe and Japan, χ 2 of 1-DVAR is so large that convergence cannot be reached and retrieval results may not be reliable or cannot be obtained. Furthermore, χ 2 also decreases with the RFI-corrected algorithm for those regions with moderate or weak RFI. The results of RFI detected by χ 2 are almost consistent with those identified by the spectral difference method.

RFI Detection Algorithm: Accurate Geolocation of the Interfering Sources in SMOS Images

The Soil Moisture and Ocean Salinity (SMOS) satellite is strongly affected by radio-frequency interference (RFI). A detection algorithm has been developed to accurately obtain the coordinates of the interfering source emitters from the SMOS images. The results obtained from this detection algorithm are regularly used to locate the on-ground sources of interference. This has allowed the identification and termination of over 200 RFI sources observed by SMOS. In the majority of cases, the accuracy of the coordinates provided was better than 4 km, which is a very important achievement considering that SMOS spatial resolution is larger than 35 km and that the contamination of a single RFI can extend to several thousands of kilometers in some cases.

Experimental Study on the Performance of RFI Detection Algorithms in Microwave Radiometry: Toward an Optimum Combined Test

Radio-frequency interference (RFI) is probably today's most serious limitation to the accurate retrieval of geophysical parameters from microwave radiometric measurements. Strong RFI inducing a change in the detected power larger than the natural variability is simple to detect. Moderate or weak RFI can be masked by the natural variability of the measurements, passing undetected and corrupting them. A number of techniques have been devised in the past years to detect and, eventually, mitigate RFI present in microwave radiometry measurements: 1) time domain; 2) frequency domain; 3) spectrogram techniques looking for anomalously high power peaks; 4) statistical techniques testing the hypothesis of Gaussianity of the received signal; 5) polarimetric techniques looking for anomalous signatures in the third and fourth Stokes parameters; or 6) wavelet techniques to estimate the RFI signal and cancel it (if any). In this paper, the first four techniques are evaluated with real data gathered with a multifrequency microwave radiometer. It will be shown how spectrogram techniques can detect RFI signals concentrated in narrow frequency bands and/or time intervals that may pass undetected with time-domain and/or frequency-domain techniques alone or with statistical methods. A combined approach is proposed to take advantage of the best performance of each technique. On one side, for strong localized RFI, the approach is spectrogram blanking or, if it is too demanding in terms of computational resources, simple time- and frequency-domain blanking. On the other side, for weak RFI, the approach is the Kurtosis statistical test, which exhibits the best performance among the ten normality tests evaluated, in conjunction with the Anderson–Darling test to detect potential RFI in the blind spots of the Kurtosis test.

Detection of Radio-Frequency Interference Signal Over Land From FY-3B Microwave Radiation Imager (MWRI)

The MicroWave Radiation Imager (MWRI) onboard the FengYun (FY)-3B satellite has five frequencies at 10.65, 18.7, 23.8, 36.5, and 89.0 GHz, each having dual channels at vertical and horizontal polarization states, respectively. It is found that radio-frequency interference (RFI) is present in MWRI data over land. The RFI signals are, in general, detectable from a spectral difference method and a principal component analysis (PCA) method. In particular, the PCA method is applied to derive RFI signals from natural radiations by using the characteristics of natural radiation measurements having all-channel correlations. In the area where data have a higher projection onto the first principle component (PC) mode, RFI is, in general, present. However, both the spectral and PCA methods cannot detect RFI reliably over frozen grounds and scattering surfaces, where the brightness temperature difference between 10.65 and 18.7 GHz is large. Thus, detection is improved through the use of normalized PCA. The new RFI detection algorithm is now working reliably for MWRI applications. It is found that RFI at 10.65 GHz distributes widely over Europe and Japan, and is less popular over the United States and China.