Data Transfer Research Articles

A coupled and parallel peridynamics–SPH modeling and simulation of buried explosion induced soil fragmentation and cratering

In this work, we develop a coupled peridynamics (PD) and smoothed particle hydrodynamics (SPH) model for simulating soil fragmentation, ejection, and cratering under buried explosions. Specifically, the non-ordinary state-based PD theory describes the dynamic response of soil, while the SPH method governs the motion of the explosive gas products. A stable and efficient coupling algorithm for data transfer between PD and SPH is adopted. To accurately capture soil behaviors under blast loading, we propose a modified Drucker–Prager plasticity model incorporating an additional high-pressure compaction equation of state (EoS). Additionally, a parallel code using the OpenMP algorithm is developed for the coupled PD–SPH model to handle large-scale, long-duration soil explosions. PD–SPH predictions are compared with two sets of well-designed centrifugal model tests. The gravitational effect is considered in the simulation to replicate the true physical process of buried explosions in soil. It is found that the coupled PD–SPH model successfully captures the entire physical process of soil fragmentation, ejection, and cratering induced by buried explosions. The predicted soil ejection and cratering morphologies, as well as quantitative results such as ejection speed, ejection height, and crater diameter, align well with the test results. Furthermore, the results indicate that the gravitational effect significantly influences the soil ejection and cratering processes. With increased gravitational acceleration, the soil ejection height and crater size decrease markedly.

Advancing privacy-aware machine learning on sensitive data via edge-based continual µ-training for personalized large models

Abstract This paper introduces an innovative method for fine-tuning a large multi-label model for abnormality detection, utilizing a smaller trainer and advanced knowledge distillation techniques. It studies the effects of fine-tuning on various abnormalities, noting different improvements based on the Original Model’s performance in specific tasks. The experimental setup, optimized for on-device inference and fine-tuning with limited computational resources, demonstrates moderate yet promising enhancements in model performance post-fine-tuning. Key insights from the study include the significance of aligning the μ-Trainer’s behavior with the Original Model and the influence of hyper-parameters like batch size on fine-tuning outcomes. The research acknowledges limitations such as the limited exploration of loss functions in multi-label models and constraints in architectural design, suggesting potential avenues for future investigation. While the proposed Naive Continual Fine-tuning Process is in its early stages, we highlight this paper’s potential model personalization on long-term data. Moreover, weight transfer in our system is exclusively for fine-tuning; hence, it improves user privacy protection by failing data reconstruction attempts from weights, like an issue with Federated Learning models. Our on-device fine-tuning prevents the transferring of data or gradients from the edge of the network to their server. Despite modest performance improvements after fine-tuning, these working layers represent a small fraction (0.7%) of the total weights in the Original Model and 1.6% in the μ-Trainer. This study establishes a foundational framework for advancing personalized model adaptation, on-device inference and fine-tuning while emphasizing the importance of safeguarding data privacy in model development.

Enhancing machine learning models: addressing challenges and future directions

Machine learning is considered as a core of modern artificial intelligence with progressive advancements throughout a spectrum including but not limited to healthcare and finance, natural language processing and self-driving cars. However, several problems remain to affect the efficiency, equal opportunities of users, and adaptability of ML models for an even faster-growing era. The limitations include shortage of high quality and access to training data, model complexity that can lead to overfitting, built in bias of the algorithm, interpretability and finally, the computational density needed for such big data models. These problems posed challenges to translate the knowledge derived from the ML systems into real-world use as well as hindering generalization of ML systems, particularly the medical and legal fields that have requirements of fairness and interpretability. These are the basic issues this journal addresses and provides possible ways of enhancing the performance of the ML models. To mitigate the problem of data deficiency, we present various techniques including data augmentation and transfer learning. To mitigate this issue, we present regularization strategies and methods of model validation. Several prevention methods are also mentioned including biasing of Algorithm and models using adversarial biasing, and Fairness-aware learning methods. Furthermore, we explore the increasing relevance of post-hoc model interpretability such as the SHAP and LIME methods which explain model’s outputs in a more detailed manner. The objective of the present Journal is to support further development of more stable, efficient and fair Machine Learning systems. In this paper, recent developments and long-term solutions are discussed to prepare the way for better and more responsible use of AI in the future.

High-Fidelity CFD Simulation of Mixed Convection and Forced Convection in a Pebble Bed Test Reactor Core

The Hermes low-power [35-MW(thermal)] reactor will be built and operated by Kairos Power LLC (KP) to demonstrate its fluoride salt–cooled high-temperature reactor (FHR) technology. In the KP FHR, the reactor core is composed of randomly packed pebbles with TRISO fuel particles inside with FLiBe flow upward through the core acting as a coolant. Previous numerical and experimental studies have been limited to either a small-size bed or to a lack of detailed measurements for heat transfer. To address the lack of high-fidelity heat transfer data in a real-size FHR core, in this study, we simulated a pebble bed core with 34 374 pebbles randomly packed, similar to the Hermes reactor’s size. The core radius was 14 times that of the pebble diameter, while the core height was 45 times. In this work, we were particularly interested in a mixed convection regime, where buoyancy is important. Therefore, we performed several large-eddy simulations at different Reynolds numbers (160 to 1000) with gravitational force included. The spectral element computational fluid dynamics code NekRS with graphics processing unit acceleration was used for this study. The low-Mach number approximation was applied to address property changes in the FLiBe and to account for buoyancy. A pure hexahedral mesh with 60 million elements was generated by the Voronoi cell method. At the polynomial order of 5, the total degrees of freedom was 7.5 billion. The developed case in this work is the first of its kind in terms of size and complexity. The local numerical data across the domain were obtained and compared with empirical correlations. After examining the data, we found the following conclusions. For pressure drop, the Reger correlation predicted less than a 5% error. On the other hand, for heat transfer, the Wakao correlation outperformed the others. Based on our findings, we recommend the use of the Wakao correlation for the Nusselt number calculation, and for pressure drop, the KTA (Kerntechnischer Ausschuss) correclation, among the available experimental correlations. The Reger direct numerical simulation–driven correlation for pressure drops should also be considered, given its best agreement with our calculations.

Cloth-net: improved hybrid adversarial network with dense vision transformer for 2D-3D image classification for accurate cloth recommendation engine

PurposeThe aim of this study is to enhance the performance of cloth recommendation systems by proposing a hybrid adversarial network called Cloth-Net, which integrates Dense Vision Transformers for effective 2D-3D image classification.Design/methodology/approachCloth-Net combines the strengths of adversarial networks with Dense Vision Transformers to process both 2D and 3D images for improved classification. The model was trained on a large-scale dataset of clothing images, using a hybrid adversarial approach that enhances both feature extraction and image classification accuracy. The methodology also includes data augmentation and transfer learning techniques to optimize the model’s generalization capability.FindingsExperimental results demonstrate that Cloth-Net significantly outperforms traditional convolutional neural network-based methods in terms of accuracy, precision, and recommendation quality. The hybrid adversarial framework, together with Dense Vision Transformers, enables the model to better understand complex clothing images, leading to more accurate and personalized recommendations.Originality/valueThis study introduces a novel hybrid adversarial model, Cloth-Net, that uniquely combines Dense Vision Transformers with traditional adversarial networks for the first time in the context of 2D-3D image classification. The findings present a substantial improvement in the performance of cloth recommendation engines, making the proposed model valuable for both academic research and practical applications in fashion technology.

Design of a Multi-Node Data Acquisition System for Logging-While-Drilling Acoustic Logging Instruments Based on FPGA

The logging-while-drilling (LWD) acoustic logging instrument is pivotal in unconventional oil and gas exploration, and in providing real-time assessments of subsurface formations. The acquisition system, a core component of the LWD acoustic logging suite, is tasked with capturing, transmitting, and processing acoustic signals from the formation, which directly affects the accuracy and timeliness of the logging data. Recognizing the constraints of current LWD acquisition systems, including limited data collection capabilities and inadequate precision, this study introduces an FPGA-based multi-node data acquisition system for LWD acoustic logging. This system increases sampling density and data accuracy, leading to a more comprehensive collection of formation information. The multi-node acquisition system is composed primarily of a main control circuit board and several acquisition circuit boards, all connected via an RS485 bus. The Field-Programmable Gate Array (FPGA) is utilized to develop the acquisition circuit board’s firmware, offering adjustable control over parameters, such as the AD7380’s operational mode, sampling rate, and depth, facilitating real-time and concurrent acquisition and storage of formation acoustic signals. The main control board communicates with the acquisition boards via the RS485 bus, issuing commands to enable autonomous data collection and transfer from each board, thus enhancing the system’s reliability and scalability. Experimental results confirm the system’s capacity to efficiently capture waveform signals and upload them in real-time, underscoring its dependability and timeliness. The findings suggest that the system is capable of high-speed, real-time acquisition and processing of acoustic signals, offering robust technical support for the continued application of LWD acoustic logging instruments.

VoI-based Situation-Aware Routing Protocol for Non-linear Underwater Communication Networks

One of the main challenges for underwater applications, such as environmental monitoring and disaster management, is achieving efficient data transmission in environments where conditions change rapidly, and resources need for data transport are scarce. The capability of evaluating the Value of information (VoI) enables us to assess these problems by proposing a Value of Information-based Situation-Aware Non-Linear Routing (VoI SANLR/VoI SANL) method. It aims to deal with critical event scenarios using BDI (Belief-Desire-Intention) logic criteria and prioritizing the timely uploading of data-driven information towards the destination. SANLR of VoI is developed to reduce energy consumption, end-to-end latency, jitter, and improve Packet Delivery Ratio (PDR) in underwater communication networks. VoI SANLR introduces principles of priority-based methods and intends to address challenges in terms of underwater environment such as varying channel conditions, lack energy resources, and real-time decision requirements by using SANLR. Energy optimization analysis reveals consistent outperformance, achieving a remarkable 95% reduction in energy consumption compared to other techniques. Low latency is maintained, ranging from 2.5 to 0.5 seconds, showcasing enhanced efficiency and scalability. VoI SANLR demonstrates exceptional performance in both throughput and jitter. It achieves the highest data transfer rates, ranging from 100 kbps to 110 kbps, indicating outstanding efficiency. Additionally, the jitter remains consistently low, between 1.8 ms and 2 ms, ensuring minimal delay variability and improved communication stability. PDR consistently surpasses other techniques, reaching a maximum of 99%. Additionally, network lifetime analysis demonstrates VoI SANLR's superiority, exhibiting the highest network lifetime at each node and a significant 31.25% improvement at Node 100 compared to other methods.

Dose Accuracy and Reliability of a Connected Insulin Pen System.

For people with diabetes on insulin injection therapy, insulin pen dose accuracy and reliability are key features. Dose accuracy of the Tempo PenTM with attached Tempo Smart ButtonTM [Module] (the system) was tested for three doses under standard atmospheric conditions. Reliability and the ability of the Module to detect, store, and transmit dose-related data were also tested. The system met the International Organization for Standardization 11608-1:2014 requirements for dose accuracy at all doses. Mean (standard deviation) doses were 0.010 mL (0.002), 0.299 mL (0.002), and 0.597 mL (0.004) for the 0.010-mL, 0.300-mL, and 0.600-mL doses, respectively. The Module also met requirements for data transfer after every injection. The system accurately delivered doses and reliably captured and stored insulin dosing information.

Organizing, Transmitting, and Accessing Vector Data in DGGS Based on Progressive Tile Structure

ABSTRACTIn order to solve the efficiency problem of massive multiresolution vector data organization in DGGS field, we propose a vector data organization method based on progressive tile structure in DGGS. The method is a combination of DGGS with tile structure, which are traditionally used in the field of GIS. That structure can support tile scheduling and express vector objects asymptotically in the inner pairs of tiles, which reduces the amount of data transfer and improves the access and construction efficiency of vector objects while guaranteeing the efficiency of data scheduling. To validate the above design, we first resampled the global 1:100,000 vector data containing multiple topological types to multiresolution vector data with a data resolution of 1:559,000–1:35,000, the data volume is 22,312 MB after converting to tile structure. Then we carried out three experiments: data volume, data construction, and data access. The results of the experiments show that compared with the organization method based on the traditional tile‐pyramid structure, the method is able to achieve higher construction efficiency and access efficiency, as well as the smaller data storage volume. Therefore, we believe that this method provides an efficient method for the organization of massive multiresolution vector data based on DGGS and also provides an idea for the organization of massive multiresolution data in the field of GIS.

What Next for the Science of Patient Preference? Interoperability, Standardization, and Transferability.

Using patient preference information (PPI) to incorporate patient voices into the drug development lifecycle can help align therapies with the needs and values of patients. However, several barriers have limited the use of PPI, including a lack of clarity on its use by decision-makers, a need for greater decision-maker trust in PPI, and a lack of time, budgets, and access to specialist expertise. The value proposition for PPI could be enhanced by making it FAIR: Findable, Accessible, Interoperable, and Reusable. To support the development of a research agenda to deliver FAIR PPI, we reviewed related endeavors in the development of repositories of existing studies, disease models, benefit transfer, and common data standards. We concluded that developing FAIR PPI would require advances in the science of PPI, including the establishment of a consortium, mirroring the Clinical Data Interchange Standards Consortium (CDISC) or Observational Medical Outcomes Partnership (OPOM), to develop PPI data standards, and research into the sources of variation in patient preferences. This will require the science of PPI to graduate from being a body of empirical observations to developing theories that explain variations in patient preferences, simultaneously driving both efficiency in the generation of PPI and trust in PPI.

Intrusion detection in metaverse environment internet of things systems by metaheuristics tuned two level framework

Internet of Things (IoT) is one of the most important emerging technologies that supports Metaverse integrating process, by enabling smooth data transfer among physical and virtual domains. Integrating sensor devices, wearables, and smart gadgets into Metaverse environment enables IoT to deepen interactions and enhance immersion, both crucial for a completely integrated, data-driven Metaverse. Nevertheless, because IoT devices are often built with minimal hardware and are connected to the Internet, they are highly susceptible to different types of cyberattacks, presenting a significant security problem for maintaining a secure infrastructure. Conventional security techniques have difficulty countering these evolving threats, highlighting the need for adaptive solutions powered by artificial intelligence (AI). This work seeks to improve trust and security in IoT edge devices integrated in to the Metaverse. This study revolves around hybrid framework that combines convolutional neural networks (CNN) and machine learning (ML) classifying models, like categorical boosting (CatBoost) and light gradient-boosting machine (LightGBM), further optimized through metaheuristics optimizers for leveraged performance. A two-leveled architecture was designed to manage intricate data, enabling the detection and classification of attacks within IoT networks. A thorough analysis utilizing a real-world IoT network attacks dataset validates the proposed architecture’s efficacy in identification of the specific variants of malevolent assaults, that is a classic multi-class classification challenge. Three experiments were executed utilizing data open to public, where the top models attained a supreme accuracy of 99.83% for multi-class classification. Additionally, explainable AI methods offered valuable supplementary insights into the model’s decision-making process, supporting future data collection efforts and enhancing security of these systems.

Group Transfer Learning for High‐Dimensional Linear Regression

ABSTRACTTransfer learning is a powerful technique for enhancing parameter estimation in target datasets by incorporating information from related source datasets. However, traditional transfer learning methods often assess the overall transferability of source data without fully considering partially transferable components. To address this, we propose a novel group transfer learning (GTL) algorithm that efficiently integrates both fully and partially transferable data by grouping parameters based on their transferability. Our approach includes an adaptive mechanism to identify transferable parameter groups when their transferability is unknown. We provide theoretical guarantees for the effectiveness of the GTL algorithm in parameter estimation and its ability to accurately detect transferable segments. Extensive simulations and real‐world case studies demonstrate that our method surpasses existing approaches in estimation accuracy and robustness.

Thermometer Ions, Internal Energies, and In-Source Fragmentation in Ambient Ionization.

Ionization and fragmentation are at the core of mass spectrometry. But they are not necessarily separated in space, as in-source fragmentation can also occur. Here, we survey the literature published since our 2005 review on the internal energy and fragmentation in electrospray ionization sources. We present new thermometer molecules to diagnose and quantify source heating, provide tables of recommended threshold (E0) and appearance energies (Eapp) for the survival yield method, and attempt to compare the softness of a variety of ambient pressure ionization sources. The droplet size distribution and desolvation dynamics play a major role: lower average internal energies are obtained when the ions remain protected by a solvation shell and spend less time nakedly exposed to activating conditions in the transfer interface. Methods based on small droplet formation without charging can thus be softer than electrospray. New dielectric barrier discharge sources can gas-phase ionize small molecules while conferring barely more internal energy than electrospray ionization. However, the tuning of the entire source interface often has an even greater influence on ion internal energies and fragmentation than on the ionization process itself. We hope that this review will facilitate further research to control and standardize in-source ion activation conditions, and to ensure the transferability of data and research results in mass spectrometry.

Physical unclonable in-memory computing for simultaneous protecting private data and deep learning models

Compute-in-memory based on resistive random-access memory has emerged as a promising technology for accelerating neural networks on edge devices. It can reduce frequent data transfers and improve energy efficiency. However, the nonvolatile nature of resistive memory raises concerns that stored weights can be easily extracted during computation. To address this challenge, we propose RePACK, a threefold data protection scheme that safeguards neural network input, weight, and structural information. It utilizes a bipartite-sort coding scheme to store data with a fully on-chip physical unclonable function. Experimental results demonstrate the effectiveness of increasing enumeration complexity to 5.77 × 1075 for a 128-column compute-in-memory core. We further implement and evaluate a RePACK computing system on a 40 nm resistive memory compute-in-memory chip. This work represents a step towards developing safe, robust, and efficient edge neural network accelerators. It potentially serves as the hardware infrastructure for edge devices in federated learning or other systems.

Texture Size Reduction Through Symmetric Overlap and Texture Carving

Maintaining memory efficient 3D assets is critical for game development due to size constraints for applications, and runtime costs such as GPU data transfers. While most prior work on 3D modeling focuses on reducing triangle count, few works focus on reducing texture sizes. We propose an automatic approach to reduce the texture size for 3D models while maintaining the rendered appearance of the original input. The two core components of our approach are: (1) Overlapping identical UV charts and folding mirrored regions within charts through an optimal transport optimization, and (2) Carving redundant and void texels in a UV-aware and texture-aware way without inverting the UV mesh. The first component creates additional void space, while the second removes void space, and their combination can greatly increase texels utilized by the UV mesh at lower texture resolutions. Our method is robust and general, and can process a 3D model with arbitrary UV layout and multiple textures without modifying the 3D mesh. We evaluate our approach on 110 models from the Google Scanned Object dataset and 64 models from Sketchfab. Compared to other approaches, ours has on average 1-3 dB PSNR higher rendering similarity and reduces pixelation in visual comparisons.

A Descriptive Study on Interconnection Networks for Parallel Computing and Algorithm Models in Parallel Computing

In parallel computing, Interconnection networks are very crucial for efficient communication among all processors within a similar system. Parallel computing has become a crucial topic in the concern of computer science and also it is revealed to be critical when researching in high performance. The evolution of computer architectures towards an improved number of nodes, where parallelism could be the approach to option for speeding up an algorithm within the last few decades. Efficient data transfer between processors is an essential component in any large scale parallel computation. Motivated by the growing interest in parallel computers, a significant amount of theoretical research has been devoted to the area of interconnection networks for parallel computers, most of it to the packet routing (or store-and-forward) model of communication. We survey some of the major developments in this field, and discuss several new alternative models of communication, In many large scale applications, communication time dominates the execution time of the whole parallel computation. Thus, the performance of a large scale parallel computer is highly correlated with the efficiency of its network and communication algorithm. The combination of processing units build a model of computation (circuits) has gained an essential place in the area of high performance computing (HPC) due to its configuration and considerable processing supremacy that is parallel, series, etc. The aim of the Presenting this paper is study on the idea of parallel computing and its programming models and also explore some theoretical and technical concepts which can be often needed to understand the Interconnection network. In particular, we show how this technology is new in assisting the field of computational physics, especially when the issue is data parallel. In the real-life example of parallel computing, there are two queues to get a ticket of anything; if two cashiers are giving tickets to 2 persons simultaneously, it helps to save time as well as reduce complexity

Analysis of "Dose Accuracy and Reliability of a Connected Insulin Pen System".

In an article in the Journal of Diabetes Science and Technology, Backfish and coauthors examined the dose accuracy and reliability of the Tempo Pen and Tempo Smart Button connected insulin pen system. This study sponsored by Eli Lilly and Company found that this system met the International Organization for Standardization 11608-1:2014 requirements for dose accuracy at a range of doses, as well as the data transfer requirements after all injections. While these results are very encouraging, they were based on simulated human factors data while data from a human factors validation study where individuals successfully dialed and administered correct doses was not reported. There is a need for further studies in a variety of populations that would greatly benefit from this enhanced insulin pen technology to facilitate optimization of dose adjustments based on accurate and reliable insulin dosing data.

HYBRID DEEP LEARNING ALGORITHM FOR FRACTAL HUMAN ACTIVITY RECOGNITION USING SMART IOT-EDGE-CLOUD CONTINUUM

Human activity recognition (HAR) employs a broad range of sensors that generate massive volumes of data. Traditional server-based and cloud computing methods require all sensor data to be sent to servers or clouds for processing, which leads to high latency and bandwidth costs. The long-term data transfer between servers and sensors maximizes the cost of latency and bandwidth. Real-time processing is, nevertheless, highly required for human action identification. By bringing processing and quick data storage to the sensors instead of depending on a central database, edge computing is rapidly emerging as a solution to this issue. Artificial intelligence is responsible for most HAR, which demands a lot of processing power and calculation. Artificial intelligence (AI) needs more computation which is not allowed by edge computing. So Edge intelligence, which allows AI to operate at the network edge for actual-time applications, has been made possible by the advent of binarized neural networks. To provide less latency and less memory for human activity identification at the edge network, we construct a hybrid deep learning-based binarized neural network (HDL-Binary Dilated DenseNet) in this research. Fractal HAR optimization algorithms could be applied to these algorithms. For example, fractal-HAR optimization techniques might be used to provide less latency and less memory human activity identification at the edge network. Using three sensors-based human activity detection datasets such as Radar HAR dataset, UCI HAR dataset and UniMib-SHAR dataset, we implemented the Hybrid Binary Dilated Dense Net. It is then assessed using four criteria. Comparatively, the Hybrid Binary Dilated DenseNet performs better with 99.6% radar HAR dataset which is highest than other models like CNN-BiLSTM and GoogLeNet.

Performance comparison in workflow efficiency between a remotely installed 3D workstation and an on-premises image processing workstation for dental cone-beam CT image reconstruction.

This study aims to compare the image processing times of dental cone beam CT (CBCT) images using a remote medical image processing workstation (RW) versus on-premises image processing (OP) and assess its impact on workflow efficiency. Data from 100 CBCT cases were randomly selected and processed using the OP3D VISION 17-19DX (EH Japan Co., Ltd.). In the OP environment, OnDemand 3D Dental (Cybermed Inc.) was used on a local terminal, while the RW setup involved a remote workstation-ZIO STATION (Ziosoft Inc.) connected via a 2 Gbps network. Seven experienced dentists processed the same data in both environments, and various processing times, including data transfer, re-slicing, 3D reconstruction, and PACS transfer, were compared. The RW environment showed significantly shorter data transfer and re-slicing times than the OP environment. However, 3D image reconstruction times were similar between the two setups. Overall, processing time was significantly reduced in the RW environment. Variability in processing times among operators was observed, with most achieving reductions in the RW environment. Remote processing of dental CBCT images using a dedicated image processing device offers equivalent or improved performance compared to on-premises processing. This approach can enhance workflow efficiency by reducing processing times and freeing up local resources, although further research is needed to optimize remote display protocols and multi-client access.

Deep Learning for Risky Cardiovascular and Cerebrovascular Event Prediction in Hypertensive Patients

In this comprehensive study, we employed a versatile approach to tackle the prediction challenges associated with atrial fibrillation (AF) and cardiovascular events (CE). Exploiting the Gaussian copula synthesizer technique for data generation, we created high-quality synthetic data to overcome the limitations posed by scarce patient records. Heart rate variability (HRV), known to be an efficient indicator of cardiac health often used with artificial intelligence (AI), was used to train and optimize custom-built deep learning (DL) models. Additionally, we explored transfer learning (TL) to enhance the model capabilities by adapting our AF classification model to address CE classification challenges, effectively transferring learned features and patterns, without extensive retraining. As a result, our models achieved accuracy rates of 77% for AF and 82% for CEs, with high sensitivity, highlighting the efficacy of synthetic data generation and transfer learning in improving classification performance across diverse medical datasets. These findings hold significant promise for enhancing diagnostic and predictive capabilities in clinical settings, ultimately contributing to improved patient care and outcomes.