Large Noise Research Articles

Estimation of AMSU-A and MHS Antenna Emission from MetOp-A End-of-Life Deep Space View Test

A unique End-of-Life (EOL) Deep Space View Test (DSVT) was performed on 27 November 2021 for the Advanced Microwave Sounding Unit-A (AMSU-A) and the Microwave Humidity Sounder (MHS) onboard the first EUMETSAT MetOp-A satellite in the deorbiting process. The purpose of this test is to recalibrate the antenna sidelobe, to derive antenna emission, and to quantify the in-orbit asymmetric scan biases of AMSU-A and MHS to, ultimately, improve Near Real-Time (NRT) products for MetOp-B and -C and the entire Fundamental Climate Data Records (FCDR). In this study, MetOp-A AMSU-A and MHS EOL DSVT data on 27 November 2021 have been analyzed. The deep space scene antenna temperatures were first applied for the antenna pattern correction; then, the antenna reflector channel emissivity values were derived from the corrected temperatures. For the MHS, the observed scan-angle-dependent brightness temperatures (BTs) for all channels were well behaved after the antenna pattern correction, except for channel 1. The derived antenna reflector emissivity values from this test are 0.0016, 0.0036, 0.0036, and 0.0019 for channels 1, 3, 4, and 5, respectively. For AMSU-A, the deep space view counts were not homogeneous during the test period, exhibiting large variations in the along-track and cross-track directions, mainly due to the instrument temperature’s rapid change during the test period. The large relative noise in the deep space view observations negatively impacted the data quality and limits the value of this test. The large relative noise may contribute to the different emissivity values derived from the same frequency for channels 9 to 14. We also found unexpected scan-angle-dependent BT after antenna pattern correction for quasi-vertical (QV) channels 1 and 2 when compared to the emission model. Further investigation using a simulation confirmed that channels 1 and 2 are QV channels, as designed.

Evaluation of measurement errors in the Patient-Oriented Eczema Measure (POEM) outcome.

The Patient-Oriented Eczema Measure (POEM) is the recommended core outcome instrument for atopic dermatitis (AD) symptoms. POEM is reported by recalling the presence/absence of seven symptoms in the last 7 days. To evaluate measurement errors in POEM recordings due to imperfect recall. Using data from a clinical trial of 247 AD patients aged 12-65 years, we analysed the reported POEM score (r-POEM) and the POEM derived from the corresponding daily scores for the same seven symptoms without weekly recall (d-POEM). We quantified recall error by comparing the r-POEM and d-POEM for 777 patient-weeks collected from 207 patients, and estimated two components of recall error: (1) recall bias due to systematic errors in measurements and (2) recall noise due to random errors in measurements, using a bespoke statistical model. POEM scores have a relatively low recall bias, but a high recall noise. Recall bias was estimated at 1.2 points lower for the r-POEM on average than the d-POEM, with a recall noise of 5.7 points. For example, a patient with a recall-free POEM of 11 (moderate) could report their POEM score anywhere from 5 to 14 (with 95% probability) because of recall error. Model estimates suggested that patients tend to recall itch and dryness more often than experienced (positive bias of less than 1 day), but less often for the other symptoms (bleeding, cracking, flaking, oozing/weeping and sleep disturbance; negative bias ranging 1-4 days). In this clinical trial data set, we found that patients tended to slightly underestimate their symptoms when reporting POEM, with significant variation in how well they were able to recall the frequency of their symptoms every time they reported POEM. A large recall noise should be taken into consideration when interpreting POEM scores.

Behavior of a chain of coupled self-propelled particles with two-dimensional potential and noise

The behavior of an elastic chain of coupled self-propelled particles in a two-dimensional potential with Gaussian colored noise is investigated. The channel and colored noise regulates the directed transport of the system. The chain of particles changes the moving direction with increasing modulation constant of potential. The average velocity of the chain shows periodic fluctuation with increasing phase shift of the potential. Strong coupling between particles inhibits the directed transport speed of the system. Large translation noise intensity is good for the directed transport in the - or , but large self-correlation time inhibits this phenomenon. The effect of the self-propulsion speed is influenced by the modulation parameter of the potential. The elastic chain moves in the when the angle noise intensity is zero. Noise intensity and self-correlation time of the angle noise have a complicated effect on the system.

A Simple Method to Mitigate Large Background Noises by Common Mode EMI in a Modern Power Converter Network

A Simple Method to Mitigate Large Background Noises by Common Mode EMI in a Modern Power Converter Network

Pixel-Perfect Structure-From-Motion With Featuremetric Refinement.

Finding local features that are repeatable across multiple views is a cornerstone of sparse 3D reconstruction. The classical image matching paradigm detects keypoints per-image once and for all, which can yield poorly-localized features and propagate large errors to the final geometry. In this paper, we refine two key steps of structure-from-motion by a direct alignment of low-level image information from multiple views: we first adjust the initial keypoint locations prior to any geometric estimation, and subsequently refine points and camera poses as a post-processing. This refinement is robust to large detection noise and appearance changes, as it optimizes a featuremetric error based on dense features predicted by a neural network. This significantly improves the accuracy of camera poses and scene geometry for a wide range of keypoint detectors, challenging viewing conditions, and off-the-shelf deep features. Our system easily scales to large image collections, enabling pixel-perfect crowd-sourced localization at scale. Our code is publicly available athttps://github.com/cvg/pixel-perfect-sfm as an add-on to the popular Structure-from-Motion software COLMAP.

Ultra-reliable intelligent link scheduling based on DRL for manned/unmanned aerial vehicle cooperative scenarios

When manned/unmanned aerial vehicle (MAV–UAV) perform cooperative missions in complex and variable situations, the communication link between manned aerial vehicle (MAV) and unmanned aerial vehicle (UAV) must be extremely dependable to guarantee the reliable transfer of command and control information. To address the above issues, this study provides an intelligent link scheduling algorithm based on MAVlink protocol that leverages deep reinforcement learning (DRL) for highly reliable communication in MAV–UAV multiple input multiple output-orthogonal frequency division multiplexing (MIMO-OFDM) communication systems. This algorithm combines a deep neural network (DNN) with a Q-learning (QL) algorithm to get a deep Q-network (DQN). The downlink communication link between MAV and UAV uses it to intelligently schedule the modulation and coding scheme (MCS) and the number of spatial multiplexing layers. Results from system simulations indicate that the proposed algorithm outperforms other representative scheduling algorithms in the communication between MAV and UAV in complex scenarios with high-speed movement and large noise interference, improving the reliability of communication in MAV–UAV cooperative missions.

How to Launch a Powerful Side-Channel Collision Attack?

A cryptographic implementation produces very similar power leakages when fed with the same input. Side-channel collision attacks exploit these similarities to establish the relationship between sub-keys and improve the efficiency of key recovery. Benefiting from independence of leakage model, they play an important role in non-profiled setting. However, performance of existing approaches against single collision value is still sub-optimal and optimization is promising. Motivated by this, we first theoretically analyze the mathematical dependency between the number of collisions and the number of encryptions, and propose an efficient side-channel attack named Collision-Paired Correlation Attack (CPCA) to guarantee that the side with fewer samples in a collision is completely paired in low noise scenario. This allows overcoming the inefficient utilization of information in existing works. Moreover, to further employ underlying informativeness, we maximize collision pairs as many as possible. This optimization significantly improves performance of CPCA and thereby extends it to large noise scenarios. Finally, to achieve moderate computational complexity, two equivalent variants of CPCA are investigated to address the potential problem of limited computing resources. Our further theoretical study illustrates that CPCA provides the upper security bound of Correlation-Enhanced Collision Attack (CECA), and experimental results fully verify its superiority.

Deep-blur: Blind identification and deblurring with convolutional neural networks

Abstract We propose a neural network architecture and a training procedure to estimate blurring operators and deblur images from a single degraded image. Our key assumption is that the forward operators can be parameterized by a low-dimensional vector. The models we consider include a description of the point spread function with Zernike polynomials in the pupil plane or product-convolution expansions, which incorporate space-varying operators. Numerical experiments show that the proposed method can accurately and robustly recover the blur parameters even for large noise levels. For a convolution model, the average signal-to-noise ratio of the recovered point spread function ranges from 13 dB in the noiseless regime to 8 dB in the high-noise regime. In comparison, the tested alternatives yield negative values. This operator estimate can then be used as an input for an unrolled neural network to deblur the image. Quantitative experiments on synthetic data demonstrate that this method outperforms other commonly used methods both perceptually and in terms of SSIM. The algorithm can process a 512 $ \times $ 512 image under a second on a consumer graphics card and does not require any human interaction once the operator parameterization has been set up.1

A general martingale approach to large noise homogenization

A general martingale approach to large noise homogenization

A Method for Clustering and Analyzing Vessel Sailing Routes Efficiently from AIS Data Using Traffic Density Images

A vessel automatic identification system (AIS) provides a large amount of dynamic vessel information over a large coverage area and data volume. The AIS data are a typical type of big geo-data with high dimensionality, large noise, heterogeneous densities, and complex distributions. This poses a challenge for the clustering and analysis of vessel sailing routes. This study proposes an efficient vessel sailing route clustering and analysis method based on AIS data that uses traffic density images to transform the clustering problem of complex AIS trajectories into an image processing problem. First, a traffic density image is constructed based on the statistics of the preprocessed AIS data. Next, the main sea route regions of traffic density images are extracted based on local image features, geometric structures, and spatial features. Finally, the sailing trajectories are clustered using the extracted sailing patterns. Based on actual vessel AIS data, multimethod comparisons and performance analysis experiments are conducted to verify the feasibility and effectiveness of the proposed method. These experimental results reveal that the proposed method displays potential for the clustering task of challenging vessel sailing routes.

Speckle suppression in holographic phase fringe patterns with different level noises based on FFDNet.

In this paper, an ANLVENet speckle suppression method in holographic phase fringe patterns with different level noises is proposed based on FFDNet, combined with asymmetric pyramid non-local block with a verge extraction module. The experimental results are compared to three network models and several representative algorithms. It is shown that the ANLVENet method not only has better superiority in the speckle suppression with different noise levels, but also preserves more details of the image edge. In addition, another speckle noise model is applied in the phase fringe patterns to prove the stronger generalization of the ANLVENet algorithm. The proposed method is suitable for suppressing the speckle with different levels in a large noise range under complex environmental conditions.

Safeguarding Multi-Fidelity Bayesian Optimization Against Large Model Form Errors and Heterogeneous Noise

Abstract Bayesian optimization (BO) is a sequential optimization strategy that is increasingly employed in a wide range of areas such as materials design. In real-world applications, acquiring high-fidelity (HF) data through physical experiments or HF simulations is the major cost component of BO. To alleviate this bottleneck, multi-fidelity (MF) methods are used to forgo the sole reliance on the expensive HF data and reduce the sampling costs by querying inexpensive low-fidelity (LF) sources whose data are correlated with HF samples. However, existing multi-fidelity BO (MFBO) methods operate under the following two assumptions that rarely hold in practical applications: (1) LF sources provide data that are well correlated with the HF data on a global scale, and (2) a single random process can model the noise in the MF data. These assumptions dramatically reduce the performance of MFBO when LF sources are only locally correlated with the HF source or when the noise variance varies across the data sources. In this paper, we view these two limitations and uncertainty sources and address them by building an emulator that more accurately quantifies uncertainties. Specifically, our emulator (1) learns a separate noise model for each data source, and (2) leverages strictly proper scoring rules in regularizing itself. We illustrate the performance of our method through analytical examples and engineering problems in materials design. The comparative studies indicate that our MFBO method outperforms existing technologies, provides interpretable results, and can leverage LF sources which are only locally correlated with the HF source.

Robust Point Cloud Registration Network for Complex Conditions.

Point cloud registration is widely used in autonomous driving, SLAM, and 3D reconstruction, and it aims to align point clouds from different viewpoints or poses under the same coordinate system. However, point cloud registration is challenging in complex situations, such as a large initial pose difference, high noise, or incomplete overlap, which will cause point cloud registration failure or mismatching. To address the shortcomings of the existing registration algorithms, this paper designed a new coarse-to-fine registration two-stage point cloud registration network, CCRNet, which utilizes an end-to-end form to perform the registration task for point clouds. The multi-scale feature extraction module, coarse registration prediction module, and fine registration prediction module designed in this paper can robustly and accurately register two point clouds without iterations. CCRNet can link the feature information between two point clouds and solve the problems of high noise and incomplete overlap by using a soft correspondence matrix. In the standard dataset ModelNet40, in cases of large initial pose difference, high noise, and incomplete overlap, the accuracy of our method, compared with the second-best popular registration algorithm, was improved by 7.0%, 7.8%, and 22.7% on the MAE, respectively. Experiments showed that our CCRNet method has advantages in registration results in a variety of complex conditions.

Convergence Analysis of Accelerated Stochastic Gradient Descent Under the Growth Condition

We study the convergence of accelerated stochastic gradient descent (SGD) for strongly convex objectives under the growth condition, which states that the variance of stochastic gradient is bounded by a multiplicative part that grows with the full gradient and a constant additive part. Through the lens of the growth condition, we investigate four widely used accelerated methods: Nesterov’s accelerated method (NAM), robust momentum method (RMM), accelerated dual averaging method (DAM+), and implicit DAM+ (iDAM+). Although these methods are known to improve the convergence rate of SGD under the condition that the stochastic gradient has bounded variance, it is not well understood how their convergence rates are affected by the multiplicative noise. In this paper, we show that these methods all converge to a neighborhood of the optimum with accelerated convergence rates (compared with SGD), even under the growth condition. In particular, NAM, RMM, and iDAM+ enjoy acceleration only with a mild multiplicative noise, whereas DAM+ enjoys acceleration, even with a large multiplicative noise. Furthermore, we propose a generic tail-averaged scheme that allows the accelerated rates of DAM+ and iDAM+ to nearly attain the theoretical lower bound (up to a logarithmic factor in the variance term). We conduct numerical experiments to support our theoretical conclusions.

Environmental Knowledge-Driven Over-the-Horizon Propagation Loss Prediction Based on Short- and Long-Term Parallel Double-Flow TrellisNets

Accurate perceptual knowledge of atmospheric characteristics in the propagation path is of great significance for the design of communication systems. However, the atmosphere above the ocean is inhomogeneous, which brings challenges to accurate prediction of propagation loss. Moreover, the atmospheric refractive index distribution model calculated from atmospheric data requires at least two stations with near-sea meteorological data. In real maritime over-the-horizon communication or detection, only one station transmits and receives meteorological data, and the received meteorological data and propagation loss have large temporal noise. To address these issues, first, a denoising model based on the one-dimensional convolution autoencoder (1DCAE) is constructed to filter out the temporal noise of input environmental meteorological data and propagation loss. Second, to accurately predict the influence of environmental factors on the prediction of propagation loss, a deep-learning framework called short- and long- term parallel double-flow TrellisNets (SL-TrellisNets) is proposed to predict the loss. Finally, the extensive experiments demonstrated that the root mean square and mean absolute errors of the proposed 1DCAE are dramatically reduced. Our proposed SL-TrellisNets outperforms the other state-of-the-art techniques in terms of propagation loss prediction. In addition, we analyzed the impact of the environmental factors on the accuracy of over-the-horizon propagation loss prediction.

A Mechanism and Data Hybrid-Driven Method for Main Circuit Ground Fault Diagnosis in Electrical Traction System

Ground faults can be caused by abnormal grounding of several different modules in the electrical traction drive system, leading to vulnerable train operation. Therefore, rapid and accurate diagnosis of ground faults is of great significance to improve safety and reliability of electric locomotives and electric multiple units (EMU). This paper proposes a mechanism and data hybrid driven method to diagnose the ground faults. Firstly, according to ground fault mechanism analysis, characteristic variables are constructed and several fault features are extracted by using signals selected from existing control system. Secondly, according to historical data, the probability membership function associated with fault features and typical ground fault types is established offline by fuzzy logic theory. Then, the probability information of fault features is fused using Dempster-Shafer (D-S) evidence theory and a decision-making method based on the basic probability number is applied to classify faults. Finally, field experiments are performed and the results show that compared with canonical correlation analysis (CCA) based method and on-board fault diagnosis method, the proposed method has better performance in the presence of large measurement noise and transient condition changes, and it is also promising for practical applications.

Experimental demonstration of sub-100 picometer level signal extraction with time-delay interferometry technique

In this work, we have built an experimental setup to simulate the clock noise transmission between two spacecrafts, and further demonstrated the extraction of picometer level signal drowned by the large laser frequency noise and clock noise with the data post-processing method. Laser frequency noise is almost eliminated by using the idea of time-delay interferometry (TDI) to construct an equal arm interferometer. Clock asynchronism and clock jitter noise are significantly suppressed by laser sideband transmitting the clock noise with an electro-optic modulator (EOM). Experimental results show a reduction in laser frequency noise by approximately 105 and clock noise by 102, recovering a weak displacement signal with an average amplitude about 60 picometer and period 1 s. This work has achieved the principle verification of the noise reduction function of TDI technique to some extent, serving the data processing research of space-borne gravitational wave detection.

Noise Reduction Method for the Vibration Signal of Reactor CRDM Based on CEEMDAACN-SK

The reactor control rod drive mechanism (CRDM) controls the startup, shutdown and power of the reactor; it is one of the key pieces of equipment to ensure the normal operation of the reactor. CRDM is complex, mainly composed of stator, rotor, bearing, roller, etc. The characteristic analysis of the vibration monitoring signal is one of the important methods of the CRDM state evaluation. In view of the characteristics of large noise interference and the difficulty in analyzing the vibration monitoring signal of CRDM, this paper proposes a noise reduction method for the vibration signal of CRDM based on complete ensemble empirical mode decomposition with adaptive amplitude correction noise and spectral kurtosis (CEEMDAACN-SK), which can deeply reduce the vibration signal of CRDM. Firstly, the proposed CEEMDAACN algorithm is used to decompose the vibration signal of CRDM to obtain multiple intrinsic mode functions (IMF). Then, the spectral kurtosis of each IMF component is analyzed to obtain the spectral kurtosis map of each IMF component, which is compared with the spectral kurtosis map of the original signal. Finally, the denoising reconstruction of the signal is carried out to obtain the final denoising signal. Through experimental analysis, the performance of the proposed CEEMDAACN-SK denoising algorithm is better than the complete ensemble empirical mode decomposition with adaptive noise and spectral kurtosis (CEEMDAN-SK) algorithm in terms of results. The method proposed in this paper can be applied not only to the vibration signal noise reduction of CRDM, but also to other equipment and fields, such as nuclear power main circulation pump and the chemical industry.

Collective behavior of self-propelled particles with heading estimation via focal observation

Animals, including humans, often achieve velocity alignment by observing their neighbors from their own perspective rather than form a bird’s-eye view. In this study, we explore a particular implementation of this idea, where particles estimate the heading of their neighbors based on their own focal view observations, thereby achieving velocity alignment. Two types of observations are used to estimate the heading of neighbors in this study: the relative position of neighbors at two adjacent moments, and the movement of the focal particle itself. Simulation results show that noise amplitude η arising during observations and the weight parameter s denoting the ratio of noise contained in the two types of observations can induce continuous order–disorder phase transition. The probability density distribution of the noise contained in different models is statistically analyzed, and results show that the type of phase transition mainly depends on the variance of the noise. Furthermore, an optimal value of s is found in our model that makes the model tolerate large noise amplitude. Our findings suggest that the focal view observation plays an important role in collective motion.

Electromagnetic Compatibility Testing Method and Application of Generators Based on Flexible Loading System

As an important part of the extended-range new-energy vehicle, the electromagnetic compatibility of the generator system will affect the overall performance of the vehicle. Therefore, it is necessary to conduct electromagnetic compatibility testing on the generator system, which includes two parts: electromagnetic emission and electromagnetic immunity. By implementing EMC testing of generators based on flexible loading equipment, we simulate the working state of the generator system during the real vehicle operation in the 3m anechoic chamber and evaluate the electromagnetic compatibility performance of the generator system. Flexible loading equipment can solve the adverse effects of large engine system volume, high noise, and exhaust emissions on the anechoic chamber environment in extended-range new-energy vehicles, and obtain real-time testing information of the generator system, achieving the standard requirements of quantifiable and evaluable test results.