Existing model poisoning attacks on federated learning (FL) assume that an adversary has access to the full data distribution. In reality, an adversary usually has limited prior knowledge about clients’ data distributions. In such a case, a poorly chosen target class renders an attack less effective. In particular, we considered a semi-targeted situation where the source class is predetermined but the target class is not. The goal is to cause the global classifier to misclassify data of the source class. Approaches such as label flipping have been adopted to inject poisoned parameters into FL. Nevertheless, it has shown that their performances are usually class-sensitive, varying with different target classes applied. Typically, an attack can become less effective when shifting to a different target class. To overcome this challenge, we propose the Attacking Distance-aware Attack (ADA) that enhances a poisoning attack by finding the optimized target class in the feature space. ADA deduces pair-wise distances between classes in the latent feature space using the Fast LAyer gradient MEthod (FLAME). We performed extensive evaluations, varying the attacking frequency in five benchmark image classification tasks with three model architectures. Furthermore, ADA’s efficacy was studied under different defense strategies in FL. ADA succeeded in increasing attack performance to 2.8 times in the most challenging case with an attacking frequency of 0.01 and bypassed existing defenses, where differential privacy that was the most effective defense still could not reduce the attack performance to below 50%.

Camera-based remote photoplethysmography (rPPG) technology has shown a promising future in contact-free cardiac and other smart health applications. The rPPG technology typically requires facial videos as a source input, which may lead to identity-privacy concerns. Facial videos are sensitive and contain subjects’ identifiable appearance features. Coupled with the health information potentially revealed by rPPG techniques, the compounding sensitivity has been a major obstacle to encouraging the sharing of facial rPPG video datasets in the research community to foster the advancement of the field. This paper investigates a suite of anonymization transforms that remove the identifiable appearance features in facial videos and retain the physiological signals for rPPG analysis. After the transformation, the facial videos are de-identified and may be shared in public with little risk of identity-privacy leakage. The proposed algorithm offers tunable options to balance the physiological fidelity and the identity-protecting strength to meet different levels of privacy requirements. A human subject study has been carried out to understand – both qualitatively and quantitatively – the perceived strength and efficacy of privacy protection by these anonymization techniques in de-identifying the facial videos and maintaining the physiological signals.

While user’s perception and performance are predominantly examined independently in virtual reality, the Action-Specific Perception (ASP) theory postulates that the performance of an individual on a task modulates this individual’s spatial and time perception pertinent to the task’s components and procedures. This paper examines the association between performance and perception and the potential effects that tactile feedback modalities could generate. This paper reports a user study (N=24), in which participants performed a standardized Fitts’s law target acquisition task by using three feedback modalities: visual, visuo-electrotactile, and visuo-vibrotactile. The users completed 3 Target Sizes X 2 Distances X 3 feedback modalities = 18 trials. The size perception, distance perception, and (movement) time perception were assessed at the end of each trial. Performance-wise, the results showed that electrotactile feedback facilitates a significantly better accuracy compared to vibrotactile and visual feedback, while vibrotactile provided the worst accuracy. Electrotactile and visual feedback enabled a comparable reaction time, while the vibrotactile offered a substantially slower reaction time than visual feedback. Although amongst feedback types the pattern of differences in perceptual aspects were comparable to performance differences, none of them was statistically significant. However, performance indeed modulated perception. Significant action-specific effects on spatial and time perception were detected. Changes in accuracy modulate both size perception and time perception, while changes in movement speed modulate distance perception. Also, the index of difficulty was found to modulate all three perceptual aspects. However, individual differences appear to affect the magnitude of action-specific effects. These outcomes highlighted the importance of haptic feedback on performance, and importantly the significance of action-specific effects on spatial and time perception in VR, which should be considered in future VR studies.

In this paper, we propose a novel three-dimensional (3D) near-field beamforming (BF) design for Large Intelligent Surface (LIS). We firstly investigate the definitions of near-field and far-field of LIS, and derive the Fresnel near-field region where amplitudes variations are negligible and only phase variations worsen the harvested array-gains. We show that the Fresnel region which covers the majority part of near-field, can be enlarged by a factor of four when considering possible imperfectness from a conventional two-dimensional (2D) far-field BF. Therefore, it is of interest to design analog 3D-BF than can recover array-gain losses in this region. Secondly, with the proved decomposition theorem we show that the optimal 3D-BF can be decomposed into a 2D far-field BF and a one-dimensional (1D) near-field BF. The 2D far-field BF compensates phase variations from mismatches in the azimuth and elevation angles, while the 1D near-field BF compensates remaining phases variations caused by distance differences from the UE to different antenna-elements on LIS. Such a proposed “2D+1D” BF design reduces codebook size significantly and is fully compatible with the existing far-field BF in the fifth-generation new-radio (5G-NR) system. Thirdly, we analyze an optimal codebook design for the 1D near-field BF, and show that with a small codebook it can perform close to optimal. Numerical results further verify that our proposal is effective and robust to recover array-gains in the near-field of LIS.

We propose a linear state estimation (SE) model with complex coefficients and variables suitable for processing large-scale data in electric power systems observable by phasor measurement units. The presented model is based on factor graphs and solved using the belief propagation (BP) algorithm. The proposed algorithm is placed in the non-overlapping multi-area SE scenario without a central coordinator. The communication between areas is asynchronous, where neighboring areas exchange only “beliefs” about specific state variables. Presented architecture directly exploits system sparsity, can be flexibly paralellized and results in substantially lower computational complexity compared to traditional SE solutions. Finally, we discuss performances of the BP-based SE algorithm using power systems with 118, 1354 and 9241 buses.

In this paper, we consider the problem of removing clouds and recovering ground cover information from remote sensing images by proposing novel framework based on a deep unfolded and prior-aided robust principal component analysis (DUPA-RPCA) network. Clouds, together with their shadows, usually occlude ground-cover features in optical remote sensing images. This hinders the utilization of these images for a range of applications such as earth observation, land-cover classification and urban planning. We model these cloud-contaminated images as a sum of low rank and sparse elements and then unfold an iterative RPCA algorithm that has been designed for reweighted l1-minimization. As a result, the activation function in DUPA-RPCA adapts for every input at each layer of the network. Our experimental results on both Landsat and Sentinel images indicate that our method gives better accuracy and efficiency when compared with existing state of the art methods.

The conventional virtual synchronous generator (VSG) is normally designed to meet certain operational and control requirements in the islanded mode. However, once the VSG is switched to grid-connected mode (GCM), the robust operation cannot be guaranteed under different grid conditions. It can lead to, especially in strong grids, poor dynamic performance such as significant oscillation, long settling time, and high overshoot. In order to improve the VSG performance in the GCM, this article first analyzes in depth the inherent coupling between the active and reactive power and its dependence on the grid conditions, such as the short circuit ratio, grid impedance ratio Xg=Rg, and the wide grid impedance variations. Afterwards, an online grid impedance estimation-based adaptive VSG (AVSG) control strategy is proposed to ensure the robust operation of the VSG considering both strong and weak grid conditions. This technique allows the operator to specify the desired settling time of output power and damping ratio.To estimate the grid impedance in real time without extra hardware and reduce the associated impacts on power quality, an online event-based grid impedance estimation algorithm is embedded into the control loop of the AVSG. Both the simulation and experimental results indicate that, compared to the conventional fixed-parameters-based controller design method, the AVSG enables the desired performance such as no oscillation, specified time duration for the settling time, and minimal overshoot regardless of the grid conditions.

Automated individual tree crown (ITC) delineation plays an important role in forest remote sensing. Accurate ITC delineation benefits biomass estimation, allometry estimation, and species classification among other forest related tasks, all of which are used to monitor forest health and make important decisions in forest management.  In this paper, we introduce Neuro-Symbolic DeepForest,  a convolutional neural network (CNN) based ITC delineation algorithm that uses a neuro-symbolic framework to inject domain knowledge (represented as rules written in probabilistic soft logic) into a CNN.  We create rules that encode concepts for competition, allometry, constrained growth, mean ITC area, and crown color.  Our results show that the delineation model learns from the annotated training data as well as the rules and that under some conditions, the injection of rules improves model performance and affects model bias.  We then analyze the effects of each rule on its related aspects of model performance.

Performance Monitor Counter(PMC)-based and circuit signal-based approaches are the two primary methodologies for online power estimation of processors. PMC-based methods will not bring hardware overhead, but the observation granularity is too large to be applied to fine-grained power management strategies. Meanwhile, signal-based methods bring high hardware overhead, and current power management strategies cannot effectively utilize clock-level precision. In this paper, we propose Power Abstract Counter(PAC). This counter records the change times of circuit signal closely related to the processor’s power consumption. We also developed an online power estimation method based on PAC. This method combines the advantages of PMC-based and signal-based methods. It can estimate the processor’s power at a smaller observation granularity and has a low hardware overhead. We validate the approach on the Loongson GS364 processor. The results show that using 16 PACs can achieve an estimation error of less than 1% within a granularity of 512 cycles.

This paper presents a novel proximal planning and control (PPC) formulation for an unmanned ground vehicle (UGV) affected by skidding and slip disturbances. The control approach also considers the presence of moving and static obstacles in the context of operation. The PPC technique is divided into three parts;  first, a nonlinear model predictive control (NMPC) based path-planner is designed to periodically generate an updated feasible trajectory to prevent the vehicle from colliding with other objects from start to the goal pose. In particular, a proximal averaged Newton-type method for optimal control (PANOC) is used to design NMPC. Second, evolutionary programming (EP) based kinematic control (KC) is designed to generate the velocity commands. Third, a dynamic control law is proposed with an extended state observer (ESO) to estimate the effects of bounded but unknown perturbations. Finally,  simulations are performed in the presence of linear and nonlinear trajectories of moving obstacles (MO) and static obstacles (SO) to verify the performance of the proposed scheme. Additionally, we have investigated and confirmed that the proposed PPC could run in real-time on a CPU with limited resources.

This paper studies a high step-up DC-DC converter, derived from the DC-DC voltage-fed push-pull converter, which main features are magnetic and electric symmetry of the internal transformer, transfer to the load of the energy trapped in the leakage inductances, natural clamping of the voltages across the power switches and independent of the duty cycle, voltage across the active switches five times lower than the load voltage, high efficiency, and high power density. Operation stages, main waveforms, and design-oriented analysis and equations are included. A 1-kW, 50-V DC input, 500-V DC output prototype, operating at 40 kHz, was designed, fabricated, and tested in the laboratory, with a measured efficiency of 95% at rated power. The studied converter is suitable for renewable and vehicle electric power conversion systems.

Seafloor massive sulfide deposits have attracted attention as a mineral resource, as they contain a wide variety of base, precious, and other valuable critical metals. Previous studies have shown that signatures of hydrothermal activity can be detected by a multi-beam echo sounder (MBES), which would be beneficial for exploring sulfide deposits. Although detecting such signatures from acoustic images is currently performed by skilled humans, automating this process could lead to improved efficiency and cost effectiveness of exploration for the seafloor deposits. Herein, we attempted to establish a method for automated detection of MBES water column anomalies using deep learning models. First, we compared the “Mask R-CNN” and “YOLO-v5” detection model architectures, wherein YOLO-v5 yielded higher F1 scores. We then compared the number of training classes and found that models trained with two classes (signal and noise) exhibited superior performance compared with models trained with only one class (signal). Finally, we examined the number of trainable parameters and obtained the best model performance when the YOLO-v5l model with a large trainable parameters was used in the two-class training process. The best model had a precision of 0.928, a recall of 0.881, and an F1 score of 0.904. Moreover, this model achieved a low false alarm rate (less than 0.7%) and had a high detection speed (20−25 ms per frame), indicating that it can be applied in the field for automatic and real-time exploration of seafloor hydrothermal deposits. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible. 2023.2.24: This work was published by IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. https://ieeexplore.ieee.org/document/10052641

As a result of the worldwide energy transition, reactive power generation has started to become a more scarce resource in the power grid. Until recently, reactive power has been an auxiliary grid service that classical power generation facilities have provided without necessarily allocating any cost for this valuable service. In this paper, a new approach for predicting the additional costs of reactive power services delivered by large hydrogenerators is proposed. We derive the optimal reactive power with minimal losses as a function of the active power level within the generator’s capability diagram. This pathway can then be used to calculate additional losses from operational regimes deviating from the optimal reactive power for each active power level. To back up the analysis, a dedicated population study was handpicked consisting of four real-world generators scaled in terms of power rating, i.e., 15 MVA, 47 MVA, 103 MVA, and 160 MVA. The objective was to identify how the optimal reactive power scale from smaller to larger MVA-sized generators. Moreover, a sensitivity analysis explores the link between the standard parameters, the stator losses, the rotor losses, the optimal reactive power, and the optimal efficiency. We find the ratio between the rotor and stator losses as the determining factor. Finally, the operational pathway introduces a new way to allocate the power producer’s cost associated with their reactive power services and can be used to justify potential profit for this service, especially considering that the intermittent reactive power needs are projected to increase in the future.

Twitter sentiment analysis has come a long way in recent years, and many experiments have been conducted in this domain. Many machine learning techniques have been used for sentiment analysis in the past and will be utilized in the future to achieve more improved results. People nowadays express their sentimental emotions more openly since they have access to   forums where they can freely express themselves, hence this gives us an opportunity where sentimental analysis could be performed and its results can be used for various different objectives as it could give information about an individual’s emotional state. This work contributes to the sentiment analysis for customer review classification, which is useful for analyzing information in the form of a large number of tweets with highly unstructured thoughts that are either good or bad. Sentiment analysis is usually done in three ways: machine learning-based, sentiment lexicon-based, or a hybrid approach. The proposed approach for sentiment analysis consists of first pre-processing tweets and using a feature extraction method. Further, Logistic regression, SVM, Decision Trees (DTs), Random Forest (RF), and XBG Classifier are used for sentiment analysis classification in the proposed framework and then all the results are compared to find out the best techniques. Out of all the techniques used, SVM is working best with an accuracy of 95.21%.

In this paper, we develop a data-driven performance-based evaluation framework for a novel electric vertical takeoff and landing aircraft (eVTOL) in the context of Urban Air Mobility (UAM) applications. First, a two-stage comprehensive simulation framework is developed to generate a benchmark database for the performance evaluation of both UAS Traffic Management (UTM) algorithms and high-fidelity eVTOL dynamical models. In the developed simulation framework, we implement UTM algorithms and incorporate real-world constraints, e.g., vertiport infrastructures and different wind conditions. From the developed simulation framework, we generate 1,213,010 flight profiles. These flight profiles are used in a model based eVTOL performance evaluation tool as inputs to compute the physical performance of 3 types of eVTOLs. Due to the high computational cost of model-based eVTOL performance evaluation approaches, a clustering-based sampling procedure is employed to reduce the redundancy in the generated flight profiles and utilize the re-sampled flight profiles to form an eVTOL performance analysis dataset. We then train and compare several machine learning models on the eVTOL performance analysis dataset to predict: performance variables-flight conditions, aerodynamic coefficients, aircraft electronics, electric motor and propeller efficiencies. Finally, we deploy the proposed data-driven models in the framework and reduce the eVTOL performance inference time to real-time. The implementation of the proposed framework can be found on GitHub: https://github.com/mrinmoysarkar/eVTOL_performance_evaluation.git

This paper investigates the performance of a non-orthogonal multiple access (NOMA) enabled full-duplex (FD) device-to-device (D2D) communication network that is underlaid in the cellular uplink. The uplink cellular user is licensed to serve the BS. For improving spectrum utilization efficiency, it also concurrently relays D2D information in FD mode. The transmit powers of D2D nodes are chosen such that the cumulative interference at the BS is below the Interference Temperature Limit (ITL). Using the statistical channel knowledge and successive interference cancellation at the D2D nodes, an expression for the D2D throughput is derived in closed-form. Thereafter, using the approximate throughput expressions, a closed-form expression of ITL apportioning parameter is derived in closed form that maximizes the D2D throughput while ensuring desired quality of service at the BS. Computer simulations demonstrate accuracy of the derived expressions.

This paper investigates the performance of high altitude platforms (HAPS) based free-space optical (FSO) communication links including HAPS-to-ground station (downlink), ground-to-HAPS (uplink) and HAPS-to-HAPS (horizontal link) communications. The effects of  attenuation loss, atmospheric turbulence, pointing error and angle-of-arrival (AOA) are taken into account. Also, the application of adaptive optics correction, one of the most effective turbulence mitigation techniques, is analyzed using the Zernike polynomials representation. Closed-form expressions are obtained for probability density function (PDF), cumulative distribution function (CDF), Rytov variance, adaptive optics filter function and outage probability mainly in terms of Meijer’s G function when both no adaptive optics correction is used and adaptive optics correction is applied. Some selected results are presented depending on the various  parameters such as the HAPS altitude, the ratio of vertical and horizontal deviations, beam waist, Zenith angle, height of ground station, receiver aperture diameter, channel state threshold and wind speed. The performance improvement with adaptive optics correction is investigated by removing different Zernike modes. We show that the downlik performance outperforms the uplink, the performance of the horizontal link sharply increases above certain altitude and communication links benefit from the adaptive optics correction up to a certain level in terms of performance improvement.

We present a soliton order preservation mechanism leading to broad tunability with high energy conversion efficiency (ECE) from a fiber-based soliton self-frequency shift (SSFS) system. Here, a pulse compressing fiber, placed before the Raman shifting nonlinear fiber (NLF), acts as a soliton order preserver. Thanks to this passive mechanism, the soliton order at the input of the NLF is preserved within 2.080% is maintained over a tunable range as large as 180 nm in a passive fiber-based SSFS system. Solitons are also tunable up to 320 nm in the wavelength range of1.96-2.28 µm with ECE>50%. Such a soliton order preservation mechanism that enables a high ECE while maintaining a broad soliton tunability is of utmost practical interest for wavelength conversion systems.