Fake Signals Research Articles

A Deep Learning Approach based on Ensemble Classification Pipeline and Interpretable Logical Rules for Bilingual Fake Speech Recognition

The essential steps of our study are to quantify and classify the differences between real and fake speech signals. In this scope, the main aim is to use the salient feature learning ability of deep learning in our study. With the use of ensemble classification pipeline, the interpretable logical rules were used for generalized reasoning with the class activation maps to discriminate the different speech classes as correctly. Fake audio samples were generated by using Deep Convolutional Generative Adversarial Neural Network. Our experiments were conducted on three different language dataset such as Turkish, English languages and Bilingual. As a result of higher classification and recognition accuracy with the use of classification pipeline as compiled into a majority voting-based ensemble classifier, the experimental results were obtained for each individual language performance approximately as 90% for training and as 80.33% for testing stages for pipeline, and it reached as 73% for majority voting results considered together with the appropriate test cases as well. To extract semantically rich rules, an interpretable logical rules infrastructure was used to infer the correct fake speech from class activations of deep learning’s generative model. Discussion and conclusion based on scientific findings are included in our study.

True image construction in quantum-secured single-pixel imaging under spoofing attack

In this paper, we introduce a quantum-secured single-pixel imaging technique designed to withstand spoofing attacks, wherein adversaries attempt to deceive imaging systems with fake signals. Unlike previous quantum-secured protocols that impose a threshold error rate limiting their operation, even with the existence of true signals, our approach not only identifies spoofing attacks but also facilitates the reconstruction of a true image. Our method involves the analysis of a specific mode correlation of a photon-pair, which is independent of the mode used for image construction, to check security. Through this analysis, we can identify both the targeted image region of the attack and the type of spoofing attack, enabling reconstruction of the true image. A proof-of-principle demonstration employing the polarization-correlation of a photon-pair is provided, showcasing successful image reconstruction even under the condition of spoofing signals that are 2000 times stronger than true signals. We expect our approach to be applied to quantum-secured signal processing, such as quantum target detection or ranging.

Cue-driven microbial cooperation and communication: evolving quorum sensing with honest signaling

BackgroundQuorum sensing (QS) is the ability of microorganisms to assess local clonal density by measuring the extracellular concentration of signal molecules that they produce and excrete. QS is also the only known way of bacterial communication that supports the coordination of within-clone cooperative actions requiring a certain threshold density of cooperating cells. Cooperation aided by QS communication is sensitive to cheating in two different ways: laggards may benefit from not investing in cooperation but enjoying the benefit provided by their cooperating neighbors, whereas Liars explicitly promise cooperation but fail to do so, thereby convincing potential cooperating neighbors to help them, for almost free. Given this double vulnerability to cheats, it is not trivial why QS-supported cooperation is so widespread among prokaryotes.ResultsWe investigated the evolutionary dynamics of QS in populations of cooperators for whom the QS signal is an inevitable side effect of producing the public good itself (cue-based QS). Using spatially explicit agent-based lattice simulations of QS-aided threshold cooperation (whereby cooperation is effective only above a critical cumulative level of contributions) and three different (analytical and numerical) approximations of the lattice model, we explored the dynamics of QS-aided threshold cooperation under a feasible range of parameter values. We demonstrate three major advantages of cue-driven cooperation. First, laggards cannot wipe out cooperation under a wide range of reasonable environmental conditions, in spite of an unconstrained possibility to mutate to cheating; in fact, cooperators may even exclude laggards at high cooperation thresholds. Second, lying almost never pays off, if the signal is an inevitable byproduct (i.e., the cue) of cooperation; even very cheap fake signals are selected against. And thirdly, QS is most useful if local cooperator densities are the least predictable, i.e., if their lattice-wise mean is close to the cooperation threshold with a substantial variance.ConclusionsComparing the results of the four different modeling approaches indicates that cue-driven threshold cooperation may be a viable evolutionary strategy for microbes that cannot keep track of past behavior of their potential cooperating partners, in spatially viscous and in well-mixed environments alike. Our model can be seen as a version of the famous greenbeard effect, where greenbeards coexist with defectors in a evolutionarily stable polymorphism. Such polymorphism is maintained by the condition-dependent trade-offs of signal production which are characteristic of cue-based QS.

A method to build feature descriptor for GNSS spoofing detection by carrier phase double difference measurement

GNSS spoofing is a type of attack that aims to deceive GNSS receivers by transmitting fake signals that imitate the authentic ones. To detect such attacks, a possible solution is to calculate the double difference (DD) of carrier phase measurements between two antennas of two separated receivers. This DD measurement represents the angle of arrival (AoA) information of the signal. In the case of authentic signals, each signal comes from a different satellite, so the AoA values are different for each satellite. However, in the case of a spoofing attack, all the counterfeit signals from the same broadcaster come from the same direction, and thus the AoA values are the same for all DD measurements. Sum of Squares (SoS) detector and Dispersion of Double Difference (D3) detector are known as simple but effective methods for spoofing detection. However, these detectors are not robust in some special regions where cycle slips are large and vary rapidly. This paper proposes a new method for detecting GNSS spoofing by using carrier phase double difference measurements to generate a feature vector as a descriptor for each satellite in a set of observed satellites. The feature vectors are formed by statistical parameters of a series of consecutive double difference values computed for each reference satellite. Our proposed feature descriptor has been evaluated on a dataset containing both real and fake signals using an SVM classifier. The study shows that the proposed feature descriptor is highly effective in detecting spoofing satellite signals.

GMM-based GNSS spoofing detector using double differential phase measurement

GNSS spoofing is an intentional interference wherein a GNSS receiver tracks counterfeit signals, resulting in incorrect outcome positions. This is the most dangerous type of intervention in the GNSS technology. In this study, we introduce a novel method for detecting spoofing signals using a Gaussian mixture model (GMM) on the double-carrier phase difference (DD) created by two independent receivers. The DD can completely eliminate errors caused by the satellite clock, receiver clock, and atmospheric layers; hence, the signal angle of arrival (AoA) is clearly expressed in the DD measurement. We utilized the GMM to model the probability density function of the DD measurement computed from the phase measurement of the receivers. Theoretically, AoA values of an authentic signal change over time owing to the nature of the signal broadcast from an orbiting satellite. However, fake signals are often transmitted from the generator, resulting in a central distribution of the corresponding AoA values. Existing studies deal with the spoofing detection problem using the above theoretical assumptions. However, it is practical to broadcast spoofing signals from several sources, and random noise can be mixed into the generated phase to render the DD measurement noisy. In such complicated scenarios, existing approaches are not sufficiently robust to detect non-authentic signals. Another observation is that because real satellites are moving in fixed orbits, there should be a correlation among the AoA values of the signals coming from these satellites. In contrast, counterfeit signals (with the main purpose of causing the wrong position or noisy phase) do not follow the pattern of the real signals. Therefore, instead of using the above hard assumption about the signal AoA, we propose a statistical model GMM to learn the hidden relationship of the AoA values among real signals from different real satellites, which are then used to detect spoofing signals.

STUDYING THE RESISTANCE OF BIOMETRIC AUTHENTICATION SYSTEMS TO ATTACKS USING VOICE CLONING TECHNOLOGY BASED ON DEEP NEURAL NETWORKS

With the development of voice synthesis technologies based on deep neural networks, the security threats to biometric authentication systems that use voice recognition have increased. These systems, which were considered reliable, can be easily compromised by fake voices created using advanced models such as WaveNet, Tacotron 2, and other modern algorithms. In today's cybersecurity environment, such attacks jeopardize the confidentiality of personal data, which necessitates the improvement of protection methods. The purpose of this article is to study the resilience of biometric authentication systems to attacks using voice cloning technology, to analyze the effectiveness of modern synthesis methods for circumventing such systems, and to provide a comparative overview of various approaches to protect voice biometric data. The article discusses technologies that allow for the creation of accurate and realistic synthetic voices, as well as methods for detecting and protecting against fake signals. The article also analyzes the current vulnerabilities of voice systems and suggests strategies to increase resistance to such attacks, providing users with greater security and privacy.

General three-dimensional thermal illusion metamaterials

Abstract Thermal illusion aims to create fake thermal signals or hide the thermal target from the background thermal field to mislead infrared observers, and illusion thermotics was proposed to regulate heat flux with artificially structured metamaterials for thermal illusion. Most theoretical and experimental works on illusion thermotics focus on two-dimensional materials, while heat transfer in real three-dimensional (3D) objects remains elusive, so the general 3D illusion thermotics is urgently demanded. In this study, we propose a general method to design 3D thermal illusion metamaterials with varying illusions at different sizes and positions. To validate the generality of the 3D method for thermal illusion metamaterials, we realize thermal functionalities of thermal shifting, splitting, trapping, amplifying and compressing. In addition, we propose a special way to simplify the design method under the condition that the size of illusion target is equal to the size of original heat source. The 3D thermal illusion metamaterial paves a general way for illusion thermotics and triggers the exploring of illusion metamaterials for more functionalities and applications.

Unveiling Chiral Perovskite CD Signal Scaling: Discerning Authentic and Counterfeit Signals through Sample-State Analysis.

Circular dichroism (CD) spectroscopy is a well-known and powerful technique widely used for distinguishing chiral enantiomers based on their differential absorbance of the right and left circularly polarized light. With the increasing demand for solid-state chiral optics, CD spectroscopy has been extended to elucidate the chirality of solid-state samples beyond the traditional solution state. However, due to the sample preparation differential, the CD spectra of the same compound measured by different researchers may not be mutually consistent. In this study, we employ solution, powder, thin-film, and single-crystal samples to explore the challenges associated with CD measurements and distinguish between genuine and fake signals. Rational fabrication of the solid-state samples can effectively minimize the macroscopic anisotropic nature of the samples and thereby mitigate the influence of linear dichroism (LD) and linear birefringence (LB) effects, which arise from anisotropy-induced differences in the absorbances and refractive indices. The local anisotropic and overall isotropic features of the high-quality thin-film sample achieve an optically isotropic state, which exhibits superior CD signal repeatability at the front and back sides at different angles by rotating the sample along the light path. In addition, sample thickness-induced CD signal overload and absorption saturation pose more severe challenges than the LBLD-induced amplified CD signal but are rarely focused on. The CD signal overload in the deep UV region leads to the presence of fake signals, while absorption saturation results in a complete loss of the CD signal. These findings help obtain accurate CD signals by a well-fabricated optically isotropic sample to avoid LDLB and optimize the sample thickness to avoid fake signals and no signals.

The use of Wasserstein Generative Adversarial Networks in searches for new resonances at the LHC.

In the search for physics beyond the standard model, machine learning classifiers provide methods for extracting signals from background processes in data produced at the LHC. Semi-supervised machine learning models are trained on a labeled background and unlabelled signal. When using semi-supervised techniques in the training of machine learning models, over-training can lead to background events incorrectly being labeled as signal events. The extent of false signals generated must therefore be quantified before semi-supervised techniques can be used in resonance searches. In this study, a frequentest methodology is presented to quantify the extent of fake signals generated in the training of semi supervised DNN classifiers when confronting side-bands and the signal regions. The use of a WGAN is explored as a machine learning based data generator.

Improved fast neutron detection using CNN-based pulse shape discrimination

The importance of fast neutron detection for nuclear safeguards purposes has increased due to its potential advantages such as reasonable cost and higher precision for larger sample masses of nuclear materials. Pulse-shape discrimination (PSD) is inevitably used to discriminate neutron- and gamma-ray- induced signals from organic scintillators of very high gamma sensitivity. The light output (LO) threshold corresponding to several MeV of recoiled proton energy could be necessary to achieve fine PSD performance. However, this leads to neutron count losses and possible distortion of results obtained by neutron multiplicity counting (NMC)-based nuclear material accountancy (NMA). Moreover, conventional PSD techniques are not effective for counting of neutrons in a high-gamma-ray environment, even under a sufficiently high LO threshold. In the present work, PSD performance (figure-of-merit, FOM) according to LO bands was confirmed using a conventional charge comparison method (CCM) and compared with results obtained by convolution neural network (CNN)-based PSD algorithms. Also, it was attempted, for the first time ever, to reject fake neutron signals from distorted PSD regions where neutron-induced signals are normally detected. The overall results indicated that higher neutron detection efficiency with better accuracy could be achieved via CNN-based PSD algorithms.

Deep Einstein@Home All-sky Search for Continuous Gravitational Waves in LIGO O3 Public Data

We present the results of an all-sky search for continuous gravitational waves in the public LIGO O3 data. The search covers signal frequencies 20.0 Hz ≤ f ≤ 800.0 Hz and a spin-down range down to −2.6 × 10−9 Hz s−1, motivated by detectability studies on synthetic populations of Galactic neutron stars. This search is the most sensitive all-sky search to date in this frequency/spin-down region. The initial search was performed using the first half of the public LIGO O3 data (O3a), utilizing graphical processing units provided in equal parts by the volunteers of the Einstein@Home computing project and by the ATLAS cluster. After a hierarchical follow-up in seven stages, 12 candidates remain. Six are discarded at the eighth stage, by using the remaining O3 LIGO data (O3b). The surviving six can be ascribed to continuous-wave fake signals present in the LIGO data for validation purposes. We recover these fake signals with very high accuracy with our last stage search, which coherently combines all O3 data. Based on our results, we set upper limits on the gravitational-wave amplitude h 0 and translate these into upper limits on the neutron star ellipticity and on the r-mode amplitude. The most stringent upper limits are at 203 Hz, with h 0 = 8.1 × 10−26 at the 90% confidence level. Our results exclude isolated neutron stars rotating faster than 5 ms with ellipticities greater than within a distance d from Earth and r-mode amplitudes for neutron stars spinning faster than 150 Hz.

Evolutionary generative adversarial network based end-to-end learning for MIMO molecular communication with drift system

Molecular communication (MC) is a novel paradigm for nano-communication networks. Compared with diffusion-based single-input single-out (SISO) systems, multiple-input multiple-output (MIMO) MC with drift systems can effectively mitigate the negative effects of inter symbol interference (ISI), inter link interference (ILI) and noise, further improving transmission efficiency. The modeling complexity of MIMO MC systems inspires the application of deep learning (DL) techniques to establish end-to-end architectures for signal recovery. However, training of the entire end-to-end system is limited by the unknown channel and small training sample size. In this paper, aiming at signal recovery of the newly developed mathematical MIMO MC with drift system model, a Kullback–Leibler divergence (KLD) evolutionary generative adversarial network (EGAN)-based end-to-end learning method is proposed. The end-to-end architecture can be trained offline with both the sampled and fake signals generated by KLD EGAN, even with a small training sample size, and then used to recover online transmitted signals directly. Besides, two traditional detection algorithms denoted as the maximum a posterior (MAP) detector and fixed threshold (FT) detector, are proposed as well for theoretical comparison. Experiments of the effect of different model parameters on the system performance have been carried out. Results validate the effectiveness and robustness of our proposed method compared to other DL-based methods, including the deep neural networks (DNN)-based, the original GAN-based, and the original EGAN-based, in terms of transmission accuracy.

Hybrid neural network-based metaheuristics for prediction of financial markets: a case study on global gold market

Abstract Technical analysis indicators are popular tools in financial markets. These tools help investors to identify buy and sell signals with relatively large errors. The main goal of this study is to develop new practical methods to identify fake signals obtained from technical analysis indicators in the precious metals market. In this paper, we analyze these indicators in different ways based on the recorded signals for 10 months. The main novelty of this research is to propose hybrid neural network-based metaheuristic algorithms for analyzing them accurately while increasing the performance of the signals obtained from technical analysis indicators. We combine a convolutional neural network and a bidirectional gated recurrent unit whose hyperparameters are optimized using the firefly metaheuristic algorithm. To determine and select the most influential variables on the target variable, we use another successful recently developed metaheuristic, namely, the moth-flame optimization algorithm. Finally, we compare the performance of the proposed models with other state-of-the-art single and hybrid deep learning and machine learning methods from the literature. Finally, the main finding is that the proposed neural network-based metaheuristics can be useful as a decision support tool for investors to address and control the enormous uncertainties in the financial and precious metals markets.

GEM foil characterization in cost-effective, and efficient way and attempt made to use it as an imaging detector

Increasing demand for security scanners and medical imaging techniques has risen with the advancement of technology based on silicon sensors. However, these technologies are much expensive and require critical handling. The Gas Electron Multiplier (GEM) foil, based detector has been attempted to use as an imaging detector. GEM foil is generally constructed using 50 μm highly insulating film coated with 5 μm copper on both sides and a network of highly dense double conical holes of size 50-70 μm in it. Due to the microscopic structure of holes and the dependence of the electric field inside them, it becomes essential to study the defects and uniformity of holes along with the electrical property of foils in various conditions. In our work, we have tested GEM foils both optically and electrically and saw the response of the triple-layered GEM detector towards fake signal and gain.

Detection of Fake Replay Attack Signals on Remote Keyless Controlled Vehicles Using Pre-Trained Deep Neural Network

Keyless systems have replaced the old-fashioned methods of inserting physical keys into keyholes to unlock the door, which are inconvenient and easily exploited by threat actors. Keyless systems use the technology of radio frequency (RF) as an interface to transmit signals from the key fob to the vehicle. However, keyless systems are also susceptible to being compromised by a threat actor who intercepts the transmitted signal and performs a replay attack. In this paper, we propose a transfer learning-based model to identify the replay attacks launched against remote keyless controlled vehicles. Specifically, the system makes use of a pre-trained ResNet50 deep neural network to predict the wireless remote signals used to lock or unlock doors of a remote-controlled vehicle system. The signals are finally classified into three classes: real signal, fake signal high gain, and fake signal low gain. We have trained our model with 100 epochs (3800 iterations) on a KeFRA 2022 dataset, a modern dataset. The model has recorded a final validation accuracy of 99.71% and a final validation loss of 0.29% at a low inferencing time of 50 ms for the model-based SGD solver. The experimental evaluation revealed the supremacy of the proposed model.

The impact of detection rate changes and correlations on random-coincidence background measurements

Coincidence detection of multiple particles emitted during an experiment can yield a new depth of understanding of the underlying process under study. However, the probability of detecting particles that are generated from the same physical event within a given coincidence time window is generally much lower than that of detecting particles that appear in the same coincidence time window, but were not created from the same physical event, and are therefore detected randomly in coincidence with each other. Thus, accurate and precise methods of measuring this random-coincidence background are essential for a wide variety of fields of science. A method to determine this background directly using the data themselves without any additional experimental run time or fake signals introduced in the data was recently established (O’Donnell, 2016). This method yields a statistical uncertainty on the random-coincidence background that is orders of magnitude smaller than that of the true coincidence data, though the potential for systematic errors of backgrounds from this method was never explored. In this work, we discuss common varieties of correlated and uncorrelated changes in the detection rates of each particle detected in an experiment. We demonstrate here that a correlation between particle detection rates from, for example, an incident particle beam that initiates a physical process of interest, creates systematic errors in the random-coincidence background measurement. We also discuss the impact of a variety of other realistic scenarios for rate changes in experiments. Lastly, a method is introduced to correct for errors in the random-coincidence background from any source, yielding an optimization between statistical precision and eliminating potential lingering systematic errors.

Benchmarking Information Scrambling.

Information scrambling refers to the rapid spreading of initially localized information over an entire system, via the generation of global entanglement. This effect is usually detected by measuring a temporal decay of the out-of-time order correlators. However, in experiments, decays of these correlators suffer from fake positive signals from various sources, e.g., decoherence due to inevitable couplings to the environment, or errors that cause mismatches between the purported forward and backward evolutions. In this Letter, we provide a simple and robust approach to single out the effect of genuine scrambling. This allows us to benchmark the scrambling process by quantifying the degree of the scrambling from the noisy backgrounds. We also demonstrate our protocol with simulations on IBM cloud-based quantum computers.

スプーフィング検知に向けたMUSIC法による信号到来方向推定

GNSS signals are weak and can easily be affected by radio interference. Also, since the signal format is open to the public, the signal can be created easily. Thus, a fake signal (spoofing signal) imitating a satellite signal may be generated and transmitted. As the signal has the same structure as the satellite signal, it is difficult for general receivers to distinguish them. However, the direction in which a spoofing signal arrives is usually different from that in which the satellite signal does. The purpose of this research is to calculate the direction of arrival (DOA) by using an array antenna and to determine whether the received signals are spoofing signals or not. The MUSIC algorithm can estimate the DOA with high accuracy, but this algorithm can be applied only when the number of arriving signals is less than that of antenna elements. Consequently, it is usually difficult to apply it to the satellite signals. Therefore, in this research, we proposed the MUSIC algorithm based on “despreading” of the satellite signals and verified the effectiveness of this method. As a results of experiment, the spoofing signal was detected by estimating the DOA of the received signals.

Identification of PPG Measurement Sites Toward Countermeasures Against Biometric Presentation Attacks

In this study, we propose a countermeasure against a presentation attack on a photoplethysmogram (PPG)-based biometric authentication system. The countermeasure identifies PPG measurement sites on the body based on the difference between PPG waveforms recorded at genuine measurement and non-genuine sites. In an experiment, we computed the correlation coefficients as the similarity indices between PPG waveforms using two datasets, i.e., PPG signals recorded at multiple measurement sites on participants and mapped signals to generate fake signals for authentication. We then evaluated the proposed countermeasure using the feature values extracted from the PPG signals to identify the measurement sites on the body. The experimental results indicated that the identification of measurement sites as a countermeasure operated successfully for both PPG datasets, regardless of the presence or absence of waveform mapping, and exhibited an identification accuracy of more than 90 % regardless of the elapsed time.

A Cybersecure P300-Based Brain-to-Computer Interface against Noise-Based and Fake P300 Cyberattacks

In a progressively interconnected world where the Internet of Things (IoT), ubiquitous computing, and artificial intelligence are leading to groundbreaking technology, cybersecurity remains an underdeveloped aspect. This is particularly alarming for brain-to-computer interfaces (BCIs), where hackers can threaten the user’s physical and psychological safety. In fact, standard algorithms currently employed in BCI systems are inadequate to deal with cyberattacks. In this paper, we propose a solution to improve the cybersecurity of BCI systems. As a case study, we focus on P300-based BCI systems using support vector machine (SVM) algorithms and EEG data. First, we verified that SVM algorithms are incapable of identifying hacking by simulating a set of cyberattacks using fake P300 signals and noise-based attacks. This was achieved by comparing the performance of several models when validated using real and hacked P300 datasets. Then, we implemented our solution to improve the cybersecurity of the system. The proposed solution is based on an EEG channel mixing approach to identify anomalies in the transmission channel due to hacking. Our study demonstrates that the proposed architecture can successfully identify 99.996% of simulated cyberattacks, implementing a dedicated counteraction that preserves most of BCI functions.