An archetype signal dependent noise (SDN) model is a component used in modeling image acquisition systems. This component may share properties with stationary normal white noise (WN). The characterization of WN images was used as a reference for making comparisons with SDN in both the image and Fourier domains (FDs). The image domain (ID) wavelet expansion was applied to WN images (n = 1000). Expansion-image orthogonality conditions were used to model and characterize the variance decomposition in both domains. Aside from the wavelet analysis, FD components of WN were characterized with summarized distributions and probability density function modeling. SDN was constructed by either multiplying simulated mammograms (i.e., stochastic images with 1/f2α power spectra) or clinical mammograms (both with n = 1000) with WN. SDN was analyzed with the same techniques applied to WN. The variance decomposition for WN decreases exponentially as a function of the ID wavelet expansion level as did SDN; expansion images for both types of noise were distributed similarly in the Fourier plane. Theoretically, the Fourier transform of normal WN produces a complex normally distributed image in the FD with a uniform power spectrum distributed exponentially; SDN has similar attributes. These findings are counterintuitive as SDN can be nonstationary in the ID but have stationary attributes in the FD.

This paper examines the steady-state impacts of cyberattacks, communication packet losses, and communication breakdowns on the consensus secondary control of islanded Microgrids (MGs). Four (4) scenarios are analyzed, encompassing various communication dysfunctions, such as Denial-of-Service (DoS) and False Data Injection (FDI) attacks, communication packet loss, and complete breakdowns of communication links. Our goal is to offer a clear and concise interpretation of how these communication disruptions, if undetected, can influence the steady-state performance of the consensus-based distributed secondary control of islanded MGs. Based on the simulation results, communication dysfunctions can deteriorate the secondary control of islanded MGs, causing considerable degradation of the uniform power sharing of DGs and compromising the restoration of frequency.

With the growing interest that is being shown in marine resources, the concept of the internet of things (IoT) has been extended to underwater scenarios, which has given rise to the internet of underwater things (IoUT). The IoUT encompasses a network of interconnected intelligent underwater devices that can be used to monitor underwater environments and support various applications, such as underwater exploration, disaster prevention, and environmental monitoring. Advances in underwater wireless communication and sensor technologies have propelled the IoUT concept forward. However, the IoUT faces significant challenges. The harsh and vast underwater environment make information sensing particularly difficult and lead to insufficient or inaccurate data being collected. Additionally, underwater conditions like pressure variation, hydrological characteristics, temperature changes, water currents, and topography hinder conventional communication models and make data transmission difficult and inefficient. Security in IoUT networks is a critical concern due to hardware limitations and seawater channel imperfections. Constrained sensor nodes and spatial-temporal uncertainty introduced by node mobility further complicate security provisioning. This survey paper addresses these challenges by offering a comprehensive overview of IoUT security. The investigation thoroughly examines both traditional and classic machine learning techniques, and focuses on deploying advanced technologies such as federated learning and digital twin. The study effectively addresses integration challenges and open issues, and provides a roadmap for future directions to play a pivotal role in formulating robust security mechanisms for IoUT networks.

Reliability assessment is mandatory to guarantee the correct behavior of Deep Neural Network (DNN) hardware accelerators in safety-critical applications. While fault injection stands out as a well-established, practical and robust method for reliability assessment, it is still a very time-consuming process. This paper contributes with three recipes for optimizing the efficiency of the reliability assessment: a) hybrid analytical and hierarchical FI-based reliability assessment for systolic-array-based DNN accelerators; b) mixing techniques for the reliability assessment of in-chip AI accelerators in GPUs; c) reliability assessment of DNN hardware accelerators through physical fault injection. The experimental results demonstrate the efficiency of the proposed methods applied to their target DNN HW accelerator platforms.  

Deep Neural Network (DNN) hardware accelerators are essential in a spectrum of safety-critical edge-AI applications with stringent reliability, energy efficiency, and latency requirements. Multiplication is the most resource-hungry operation in the neural network's processing elements. This paper proposes a scalable adaptive fault-tolerant approximate multiplier (AdAM) tailored for ASIC-based DNN accelerators at the algorithm and circuit levels. AdAM employs an adaptive adder that relies on an unconventional use of input Leading One Detector (LOD) values for fault detection by optimizing unutilized adder resources. A gate-level optimized LOD design and a hybrid adder design are also proposed as a part of the adaptive multiplier to improve the hardware performance. The proposed architecture uses a lightweight fault mitigation technique that sets the detected faulty bits to zero. The hardware resource utilization and the DNN accelerator's reliability metrics are used to compare the proposed solution against the Triple Modular Redundancy (TMR) in multiplication, unprotected exact multiplication, and unprotected approximate multiplication. It is demonstrated that the proposed architecture enables a multiplication with a reliability level close to the multipliers protected by TMR while at the same time utilizing 2.74 × less area and with 39.06% less power-delay product compared to the exact multiplier. Moreover, it has similar area, delay, and power consumption parameters compared to the state-of-the-art approximate multipliers with similar accuracy while providing fault detection and mitigation capability.

Decarbonization is important for reducing environmental impact and improving sustainability across all commercial enterprizes including hard to decarbonize sectors such as marine. A hybrid seagoing marine vehicle model was created and used to investigate the fuel savings potential of hybrid electric powertrains for commercial fishing boats in Alaska and the sensitivity of those potential fuel savings to component sizing and specific fishery participation. Data collected from the subject conventional diesel-powered boat were used to create and validate a flexible physics-based energy model of the conventional powertrain, which served as the foundation of an adaptable hybrid electric energy model. A linear programming formulation was developed for energy management in the hybrid model and used to optimize fuel savings from the hybrid system.  To generalize the results to a broader population of commercial fishing boats, A Markov random walk method was used to generate duty cycles of entire fishing seasons for boats working in various fisheries over the course of a year. A simple payback analysis considers which powertrain options are financially viable for each year-long duty cycle. Simulation results show that an aggressive hybrid powertrain could result in up to 3 days of full electric operation while a mild hybrid would result in a 50% fuel savings on the same 3-day mission. Missions out to 12 days still resulted in 12-19% savings for those same hybrid powertrains,  but the at-sea recharging and longer transits inherent in longer voyages reduced the maximum savings compared to shorter missions that include recharging batteries at dock.

The full realization of smart city technology is dependent on the secure and honest collaboration between IoT applications and edge-computing. In particular, resource constrained IoT devices may rely on fog-computing to alleviate the computing load of IoT tasks. Mutual authentication is needed between IoT and fog to preserve IoT data security, and monetization of fog services is needed to promote the fog service ecosystem. However, there is no guarantee that fog nodes will always respond to IoT requests correctly, either intentionally or accidentally. In the public decentralized IoT-fog environment, it is crucial to enforce integrity among fog nodes. In this paper, we propose a blockchain-based system that 1) streamlines the mutual authentication service monetization between IoT and fog, 2) verifies the integrity of fog nodes via service audits, and 3) discourages malicious activity and promotes honesty among fog nodes through incentives and penalties.

This paper presents a residential, round-trip, mixed ownership, electric car-sharing scheme with an optimized charging schedule, expressed as a mixed integer linear programming model. Representing the given travel demand as a graph, a mixed (shared and owned cars) ownership fleet configuration, with different trip fulfillment possibilities, is determined using maximal cliques. Next, a car-to-trip allocation model for feasible trip plans, with considerations for equal car utilization and charging price, is used to generate charging profiles for the fleet. The model and the outputs can be variously integrated into existing energy system modeling frameworks. A case study, using data from a test building with a residential car-sharing fleet in Munich, demonstrates the potential of the proposed models to derive meaningful operational insights for fleet and energy system planners. 92% of the recorded trips at the test building could have been feasibly accomplished with only 60% of the available cars.

Predicting user-item interactions in e-commerce platforms plays an important role in improving recommendation systems and optimizing user experience. This paper presents an in-depth exploration of using temporal graph networks (TGNs) for user-item interaction prediction. Using TGN on datasets, Amazon, LastFM, and RetailMarket, we experiment node classification and edge prediction tasks. Our approach achieves an accuracy of 80% and an AUC score of 0.93. Our findings prove the effectiveness of TGNs in identifying patterns and predicting interactions, thus advocating for their integration into e-commerce platforms to improve recommendation systems and user satisfaction.

A document by Kangjie Cao. Click on the document to view its contents.

A simple numerical solution for general relativity in the Newtonian weak field regime is provided -- no tensor calculus is required. The topics under consideration are the deflection of photons and neutrinos by the Sun, the relativistic rotation of Mercury's orbit plane, the Shapiro time delay, and the gravitational redshift -- namely, the four Solar System tests of general relativity. To simplify the calculations, Euler integration is used. Hopefully, this numerical solution can be found to be an important tool, primarily due to this simplicity.

Millimeter-wave (mmWave) communication systems utilize phased-array antennas to generate highly directional beams, effectively reducing the signal footprint. Nonetheless, eavesdropping, particularly within the main-lobe, remains a significant concern. This paper introduces BeamSec, a novel beam hopping approach to maximize absolute secrecy rates with no information about the Channel State Information (CSI) or location of the eavesdroppers. Methodologically, BeamSec identifies diverse beam-pairs between transceivers by analyzing signal characteristics, such as Angle of Departure (AoD) and Angle of Arrival (AoA). To prevent the complete message from being eavesdropped, BeamSec splits and encodes data between selected beams. Depending on the availability of full/partial or transmitter (TX)/receiver (RX)-only channel knowledge, Beam-Sec optimizes time allocation using analytical models. Our experimental validation using an 802.11ad-compatible 60 GHz phased-array testbed demonstrates a significant increase in absolute secrecy rate showcasing resilience against at least four colluding eavesdroppers. Specifically, BeamSec demonstrates a non-zero absolute secrecy rate even for the simplistic uniform time allocation approach. Radio map (partial channel knowledge) and known room geometry (instantaneous TX/RX) based schemes provide further improvement of 124.8% and 58.13%, respectively, as compared to uniform time allocation.

Contribution: This study uses a qualitative research method to analyze interviews where participants simplified an electric circuit while explaining their thought process. Background: Rearranging circuit diagrams is a fundamental skill in electrical and electronic engineering, yet students can struggle with unfamiliar configurations. Current research in the discipline is often quantitative, centered on conceptual understanding. By using a qualitative method, the process of 'how' students interact with circuit diagrams is investigated. Research Question: How do students approach circuit diagram simplifications? Methodology: 15-minute individual discussions with 10 participants (undergraduate Years 1-4) simplifying an unconventionally presented circuit diagram were recorded. Reflexive thematic analysis was used to identify common themes. Findings: 1. Participants initially rely upon pattern recognition to solve circuit problems before applying other analysis techniques. 2. Two rearrangement methods were identified: 'component focused', where combinations of components are grouped and then connected together, and 'ground focused' where components in the circuit are related to ground and then connected together. 3. Students using a ground focused strategy were less hesitant in their circuit rearrangement process. 4. Students broadly used mechanicalistic methods of error checking, selecting software tools rather than applying conceptual understanding.

Heart failure is a leading cause of death among Diabetic and Obese patients globally, contributing to 8.5% of all heart disease deaths and potentially 36% of cardiovascular disease deaths. Early detection is crucial for timely intervention, reducing symptoms, lowering hospitalizations, and improving patient outcomes through personalized management. This paper presents the use of Semi-Parametric and Accelerated Failure Time survival models (AFT) on Heart failure prediction. Cox Proportional Hazard model from the family of Semi-Parametric survival models and Weibull's Accelerated Failure Time models from the family of Accelerated Failure Time models have been compared and contrasted. The Cox PH model excels in its ability to adapt to various survival time distributions avoiding any assumptions on survival time distribution. The Cox proportional hazard allows for the examination of covariate effects on the hazard function, making it a widely used survival model. Weibull AFT model follows a parametric approach, directly estimating the distribution of the survival time. Medical records data of close to 299 patients, who had heart failure, collected during their clinical followup period is used for building and evaluating the model. The final evaluation of model performance was conducted, focusing on their capacity to predict the probability of patient survival beyond 250 days from the clinical visit. Among the Cox proportional hazard model and Weibull's AFT model, the Weibull's AFT model demonstrated superior performance compared to the Cox model. Remarkably, Weibull's model exhibited consistently exceptional performance in both train and test validations.

The 1D inversion method is widely used in airborne transient electromagnetic method (ATEM) but often encounter inaccurate imaging and overlook important geological information. The three-dimensional inversion is relatively robust and provide more detailed information on subsurface resistivity distributions but limited by significant computational burdens. In this study, we develop a rapid 3D imaging scheme based on deep learning for subsurface resistivity structures with lateral heterogeneity, which performs better than traditional 1D inversion but slightly inferior to 3D inversion, but its efficiency makes it more practical than traditional 3D inversion. To create a reasonably large training set of different resistivity models that generalizes well. Here we propose three strategies that focus on training datasets, to improve the performance of DL models, including divide-and-conquer strategy, random models generating and depth constraints. Through a large of reasonable structural models, appropriate network setups, a generalized result can be obtained through our proposed UNet framework, which has been demonstrated to be effective on both synthetic and field data. The results of gradient-based 3D and 1D inversion method are compared and analyzed, respectively, which demonstrates the reliability of the proposed method. To flexibly apply the deep learning model to field data, we propose a model scaling scheme to make pre-trained UNet model compatible with different inversion specifications and survey configurations, without retraining a new DL model. The inverted structures from field data clearly delineate the geometries of the lakes, faults and surrounding mountains. The inversion operator can support instantaneous 3D subsurface resistivity imaging for ATEM observations.

This research paper explores the transformative potential of Metaverse technologies in enhancing operational efficiency and safety within the oil and gas industry. It investigates how virtual reality (VR), Augmented Reality (AR), and comprehensive digital integration can revolutionize traditional practices in this sector. The study methodically analyzes the adoption, impact, and future trajectory of Metaverse applications, focusing on their capacity to improve simulations, training, and real-time asset monitoring through digital twins. Critical insights are offered on the integration challenges, including technological, security, and workforce adaptation barriers that must be navigated to effectively leverage these advanced digital tools. The paper advocates for ongoing research and dialogue to drive innovation, improve safety protocols, and ensure environmental sustainability in oil and gas operations, emphasizing the Metaverse's role in pushing the boundaries of digital transformation in industrial settings. The Metaverse, a sophisticated amalgamation of the real world, virtual reality (VR), augmented reality (AR), and the internet, stands as a modern marvel that has captured the imagination of innovators and industry leaders alike[1]. More than a mere collection of virtual spaces, the Meta- verse constitutes a series of interlinked realms, offering rich, immersive experiences that transcend traditional digital interfaces. Its significance in the oil and gas sector—a field historically beset with complex CSR evaluations due to vast operational scales, hazardous environments, and intricate global supply chains—is increasingly acknowledged [2]. Metaverse technology is poised to revolutionize these industries, promising to enhance operational simulations and safety training, foster collaborative virtual teams, and enable real-time monitoring of assets through digital twins[3]. This study examines the adoption and impact of Metaverse technologies in the oil and gas sectors, focusing on enhancing operational efficiency, safety, and sustainable production through innovative digital approaches. Our objective is to methodically assess the integration of these technologies, identify drivers of innovation, and evaluate their current applications and outcomes. This includes a critical analysis of the frameworks for technological integration and highlighting the gaps in existing research [1]. The incorporation of Metaverse technology is pivotal for several reasons. It provides stakeholders with a clear understanding of the technological and organizational needs required to leverage these advanced tools effectively. This review also explores the potential of Metaverse technologies to tackle industry-specific challenges such as safety, efficiency, and environmental sustainability. By identifying existing shortcomings and proposing directions for future research, this study encourages an ongoing dialogue on digital transformation within the industry, which is crucial for informed decision-making and strategic planning in the digital era [1].

.

Carbon dioxide (CO2) fluxes in regulated Alpine rivers are driven by multiple biogeochemical and anthropogenic processes, acting on different spatiotemporal scales. We quantified the relative importance of these drivers and their effects on the dynamics of CO2 concentration and atmospheric exchange fluxes in a representative Alpine river segment regulated by a cascading hydropower system with diversion, which includes two residual flow reaches and a reach subject to hydropeaking. We combined instantaneous and time-resolved water chemistry and hydraulic measurements at different times of the year, and quantified the main CO2 fluxes by calibrating a one-dimensional transport-reaction model with measured data. As a novelty compared to previous inverse modelling applications, the model also included carbonate buffering, which contributed significantly to the CO2 budget of the case study. The spatiotemporal distribution and drivers of CO2 fluxes depended on hydropower operations. Along the residual flow reaches, CO2 fluxes were directly affected by the upstream dams only in the first ~ 2.5 km, where the supply of supersaturated water from the reservoirs was predominant. Downstream of the hydropower diversion outlets, the CO2 fluxes were dominated by systematic sub-daily fluctuations in CO2 transport and evasion fluxes (`carbopeaking') driven by hydropeaking. Hydropower operational patterns and regulation approaches in Alpine rivers affect CO2 fluxes and their response to biogeochemical drivers significantly across different temporal scales. Our findings highlight the importance of considering all scales of CO2 variations for accurate quantification and understanding of these impacts, to clarify the role of natural and anthropogenic drivers in global carbon cycling.