Implementation Of Network Research Articles

Implementation of near-infrared spectroscopy and convolutional neural networks for predicting particle size distribution in fluidized bed granulation

Monitoring the particle size distribution (PSD) is crucial for controlling product quality during fluidized bed granulation. This paper proposed a rapid analytical method that quantifies the D10, D50, and D90 values using a Convolutional Block Attention Module-Convolutional Neural Network (CBAM-CNN) framework tailored for deep learning with near-infrared (NIR) spectroscopy. This innovative framework, which fuses CBAM with CNN, excels at extracting intricate features while prioritizing crucial ones, thereby facilitating the creation of a robust multi-output regression model. To expand the training dataset, we incorporated the C-Mixup algorithm, ensuring that the deep learning model was trained comprehensively. Additionally, the Bayesian optimization algorithm was introduced to optimize the hyperparameters, improving the prediction performance of the deep learning model. Compared with the commonly used Partial Least Squares (PLS), Support Vector Machine (SVM), and Artificial Neural Network (ANN) models, the CBAM-CNN model yielded higher prediction accuracy. Furthermore, the CBAM-CNN model avoided spectral preprocessing, preserved the spectral information to the maximum extent, and returned multiple predicted values at one time without degrading the prediction accuracy. Therefore, the CBAM-CNN model showed better prediction performance and modeling convenience for analyzing PSD values in fluidized bed granulation.

Enhancing 802.1X authentication with identity providers using EAP-OAUTH and OAuth 2.0

EAP-OAUTH is a novel Extensible Authentication Protocol (EAP) method that integrates the OAuth 2.0 framework to provide a secure and flexible authentication mechanism for LANs and WLANs that implement the IEEE 802.1X framework. EAP-OAUTH leverages existing, OAuth 2.0-enabled Identity Providers (IdPs) and their single sign-on (SSO) capabilities, thus offering a streamlined authentication experience for both users and organizations. The advantages of EAP-OAUTH for users include an SSO experience and enhanced privacy, while organizations benefit from simplified identity management, reduced operational costs, consistent security policies, and easier compliance. Furthermore, EAP-OAUTH represents a promising solution for addressing the challenges of authentication in modern wireless networks, such as the deployment of various multi-factor or risk-based, adaptive authentication strategies. This article presents an in-depth analysis of the EAP-OAUTH method, its design, implementation, and use cases in enterprise networks and public hotspots. It explores the OAuth 2.0 Device Authorization Grant flow and allows network clients to perform fast re-authentications without resorting to sessions on IdPs or even their SSO features. The implementation of EAP-OAUTH is demonstrated in real-world scenarios, using two IdPs (Google and Auth0), confirming its effectiveness, suitable performance and compatibility with various components of typical Wi-Fi infrastructures.

Efficient Neural Networks on the Edge with FPGAs by Optimizing an Adaptive Activation Function.

The implementation of neural networks (NNs) on edge devices enables local processing of wireless data, but faces challenges such as high computational complexity and memory requirements when deep neural networks (DNNs) are used. Shallow neural networks customized for specific problems are more efficient, requiring fewer resources and resulting in a lower latency solution. An additional benefit of the smaller network size is that it is suitable for real-time processing on edge devices. The main concern with shallow neural networks is their accuracy performance compared to DNNs. In this paper, we demonstrate that a customized adaptive activation function (AAF) can meet the accuracy of a DNN. We designed an efficient FPGA implementation for a customized segmented spline curve neural network (SSCNN) structure to replace the traditional fixed activation function with an AAF. We compared our SSCNN with different neural network structures such as a real-valued time-delay neural network (RVTDNN), an augmented real-valued time-delay neural network (ARVTDNN), and deep neural networks with different parameters. Our proposed SSCNN implementation uses 40% fewer hardware resources and no block RAMs compared to the DNN with similar accuracy. We experimentally validated this computationally efficient and memory-saving FPGA implementation of the SSCNN for digital predistortion of radio-frequency (RF) power amplifiers using the AMD/Xilinx RFSoC ZCU111. The implemented solution uses less than 3% of the available resources. The solution also enables an increase of the clock frequency to 221.12 MHz, allowing the transmission of wide bandwidth signals.

Ionic Diffusive Nanomemristors with Dendritic Competition and Cooperation Functions for Ultralow Voltage Neuromorphic Computing.

Realization of dendric signal processing in the human brain is of great significance for spatiotemporal neuromorphic engineering. Here, we proposed an ionic dendrite device with multichannel communication, which could realize synaptic behaviors even under an ultralow action potential of 80 mV. The device not only could simulate one-to-one information transfer of axons but also achieve a many-to-one modulation mode of dendrites. By the adjustment of two presynapses, Pavlov's dog conditioning experiment was learned successfully. Furthermore, the device also could emulate the biological synaptic competition and synaptic cooperation phenomenon through the comodulation of three presynapses, which are crucial for artificial neural network (ANN) implementation. Finally, an ANN was further constructed to realize highly efficient and anti-interference recognition of fashion patterns. By introducing the cooperative device, synaptic weight updates could be improved for higher linearity and larger dynamic regulation range in neuromorphic computing, resulting in higher recognition accuracy and efficiency. Such an artificial dendric device has great application prospects in the processing of more complex information and the construction of an ANN system with more functions.

Artificial neural networks optimize the establishment of a Brazilian germplasm core collection of winter squash (Cucurbita moschata D.)

With widespread cultivation, Cucurbita moschata stands out for the carotenoid content of its fruits such as β and α-carotene, components with pronounced provitamin A function and antioxidant activity. C. moschata seed oil has a high monounsaturated fatty acid content and vitamin E, constituting a lipid source of high chemical–nutritional quality. The present study evaluates the agronomic and chemical–nutritional aspects of 91 accessions of C. moschata kept at the BGH-UFV and propose the establishment of a core collection based on multivariate approaches and on the implementation of Artificial Neural Networks (ANNs). ANNs was more efficient in identifying similarity patterns and in organizing the distance between the genotypes in the groups. The averages and variances of traits in the CC formed using a 15% sampling of accessions, were closer to those of the complete collection, particularly for accumulated degree days for flowering, the mass of seeds per fruit, and seed and oil productivity. Establishing the 15% CC, based on the broad characterization of this germplasm, will be crucial to optimize the evaluation and use of promising accessions from this collection in C. moschata breeding programs, especially for traits of high chemical–nutritional importance such as the carotenoid content and the fatty acid profile.

An Energy-Efficient Cluster Head Selection and Secure Data Transmission in WSN using Spider Monkey Optimized Algorithm and Hybrid Cryptographic with Security

To conserve energy in wireless sensor networks, clustering is the well-known strategies. However, choosing a cluster head that is energy efficient is crucial for the best clustering. Because data packets must be transmitted between cluster members and the sink node, improper cluster head selection (CHs) uses more energy than other sensor nodes. As a result, it lowers the network's performance and lifespan. Due to the requirement that this network implement appropriate security measures to guarantee secure communication. This paper provides a novel cluster head selection technique that addresses issues of networks’ lives and energy usages using Spider Monkey Optimised Fuzzy C-Means Algorithm (SMOFCM). The CH is chosen using the Spider Monkey Optimisation method in the proposed SMOFCM approach, which builds on the Fuzzy C-means clustering framework. The hybrid cryptographic technique is appropriate for WSN for safe data transmission because it can address sensor challenges such processing power, storage capability, and energy. The Rivest-Shamir-Adleman (RSA), advanced encryption standards (AES), and the suggested algorithm are all used at various stages. Because asymmetric key cryptography makes key management simpler but symmetric key cryptography offers a high level of security. The AES algorithm has been created for phase 1. Phase 2 employed RSA, and all phases were carried out concurrently. According to the simulation results, it reduces energy use, lengthens the network's lifespan, and offers faster encryption, decryption, and execution times for secure data transmission

An efficient hyperpath-based algorithm for the capacitated transit equilibrium assignment problem

This paper studied the capacitated transit equilibrium assignment problem (CTEAP), which particularly accounts for the capacity effect that impacts passengers' route choices in a transit network. Specifically, we impose strict capacity constraints on line segments to ensure that the passenger flow does not exceed the line capacity. By introducing associated Lagrange multipliers, we derived the generalised hyperpath cost function and established an equivalent variational inequality formulation. A solution framework is developed to solve CTEAP based on the Method of Multipliers. To handle destroyed Cartesian product structures, we transformed the master problem into a sequence of uncapacitated subproblems, which can be tackled by two modified Newton-type hyperpath-based algorithms. Numerical analyses were performed for a small Gentile network and a large Shenzhen transit network to evaluate the impacts of the capacity constraint on passenger flows and computation costs. Our results demonstrate the superiority of the proposed CTEAP model to prevent over-loaded flows, and show the promise of applying the hyperpath-based algorithm in the implementation of real-world networks.

WiMAX Network Optimization Using Frame Period with Channel Allocation Techniques

Internet connectivity in WiMAX networks, along with various applications, is increasing rapidly, so the connectivity of internet and data transfer speed are always challenge for effective data transmission in wireless networks. Several factors affect the performance of networks. One important factor is to choose a suitable frame period for effective data transmissions. The performance of different frame period with round-robin and strict priority are evaluated in this work. A frame period in round robin performs better than a strict priority in terms of throughput, but a strict priority performs better in terms of drop rates is observed. This paper also demonstrates that an effective frame period, when combined with a proper bandwidth allocation algorithm, yields better results. This work gives the analysis that round robin performs 83.8847% better while strict priority performs 86.0020% better than the earliest deadline first algorithms for 10 subscriber stations in terms of throughput. This work is helpful to researchers and industrialists for actual implementations in WiMAX networks.

Challenges in Implementing Online Examination amidst COVID-19 Pandemic

The COVID-19 outbreak led to a remarkable shift in teaching methods, forcing educational institutions to swiftly move from physical classrooms to digital platforms. In this regard, E-examinations became a crucial mode of evaluation during pandemic. This research aims to explore challenges faced by educators in higher education institutions, and examine potential gender-based differences in perceived difficulties. From 51 Higher Education Institutions (HEI) in West Bengal, 116 teachers (49 female and 67 male) participated in the study. Using Google Forms, data were collected and analyzed with six distinct codes to categorize the challenges expressed by the teachers. During online examinations, six major challenges encountered were: managing answer scripts and administering exams, invigilation and evaluation, network connectivity, overall planning and implementation, and technological equity and access. Predominant tools used for online examinations: laptops, mobile devices, and computers or tablets. Primary challenges highlighted by teachers are mainly: administration of examinations and management of response scripts and the ethical implications of online examinations. This research offers insights highlighting the need for further in-depth research, possibly integrating qualitative or mixed-methods approaches. The implications of these findings may inspire practitioners, and policymakers to enhance online examination procedures with cutting-edge tools and techniques. Moreover, the study strongly advocates for crucial measures to address identified challenges, fostering a seamless and equitable online examination experience for educators and students.

Enhancing Open-Loop Wavefront Prediction in Adaptive Optics through 2D-LSTM Neural Network Implementation

Enhancing Open-Loop Wavefront Prediction in Adaptive Optics through 2D-LSTM Neural Network Implementation

Buckling variability analysis in damaged composite laminates subjected to thermally varying environment

The present study investigates the effect of internal defects on the composite laminated (CL) plates comprised of two distinct graphite-epoxy materials (AS4/3501-6 and T300/5208). The centrally damaged laminates stacked in two distinct lamination sequences are exposed in thermally varying environments. The CL plates are considered to experience damage in two different combinations of damage parameters. The improved first-order shear deformation theory (IFSDT) based finite element (FE) model is employed to estimate the thermal buckling of CL plates comprised of multiple fiber orientation angles and constrained to C-S-C-S and S-C-S-C boundary restrictions. The implementation of a Radial basis function network (RBFN) surrogate model has demonstrated its potential as a feasible substitute for the computationally demanding Monte-Carlo simulations (MCS) in the context of stochastic exploration. The present investigation scrutinizes the probabilistic properties of diverse materials and their associated damage characteristics. Subsequently, the RBFN-based surrogate model is employed to evaluate the likelihood of failure.

Re-configurable parallel Feed-Forward Neural Network implementation using FPGA

This paper proposes a novel hardware architecture for a Feed-Forward Neural Network (FFNN) with the objective of minimizing the number of execution clock cycles needed for the network’s computation. The proposed architecture depends mainly on using two physical layers that are multiplexed and reused during the computation of the FFNN to achieve an efficient parallel design. Two physical layers are designed to handle the computation of different sizes of Neural Networks (NN). The proposed FFNN architecture hardware resources are independent of the NN’s number of layers, instead, they depend only on the number of neurons in the largest layer. This versatile architecture serves as an accelerator in Deep Neural Network (DNN) computations as it exploits parallelism by making the two physical layers work in parallel through the computations. The proposed implementation was implemented with 18-bit fixed point representation reaching 200 MHz clock speed on a Spartan7 FPGA. Furthermore, the proposed architecture achieves a lower neuron computation factor compared to previous works in the literature.

Retraction: “Implementation of deep neural networks for classifying electroencephalogram signal using fractional S‐transform for epileptic seizure detection”

Retraction: Ashokkumar SR, Anupallavi S, Premkumar M, Jeevanantham V. Implementation of deep neural networks for classifying electroencephalogram signal using fractional S‐transform for epileptic seizure detection. Int J Imaging Syst Technol. 2021;31:895–908. https://doi.org/10.1002/ima.22565The above article, published online on 02 March 2021 in Wiley Online Library (wileyonlinelibrary.com), has been retracted by agreement between the journal Editor‐in‐Chief Dr. Mohamed Seghier, and Wiley Periodicals, LLC. The retraction has been agreed due to major overlap between this article and the following article: Liu G, Zhou W, Geng M. Automatic Seizure Detection Based on S‐Transform and Deep Convolutional Neural Network. Int. J. Neural Syst. 2020;30:1950024. https://doi.org/10.1142/S0129065719500242. Following a complaint from a third party, an investigation by the editorial office and Wiley concluded that Figures 5 and 10 were duplicated from the previously‐published article and that the degree of overlap between the articles' topics, preprocessing methods, and proposed AI tools constitutes plagiarism. The authors responded to an inquiry by the publisher, but did not provide original data. The authors were notified about the retraction but did not respond.

FPGA Implementation of Classical Dynamic Neural Networks for Smooth and Nonsmooth Optimization Problems

FPGA Implementation of Classical Dynamic Neural Networks for Smooth and Nonsmooth Optimization Problems

Multi-bit MRAM based high performance neuromorphic accelerator for image classification

Binary neural networks (BNNs) are the most efficient solution to bridge the design gap of the hardware implementation of neural networks in a resource-constrained environment. Spintronics is a prominent technology among emerging fields for next-generation on-chip non-volatile memory. Spin transfer torque (STT) and spin-orbit torque (SOT) based magnetic random-access memory (MRAM) offer non-volatility and negligible static power. Over the last few years, STT and SOT-based multilevel spintronic memories have emerged as a promising solution to attain high storage density. This paper presents the operation principle and performance evaluation of spintronics-based single-bit STT and SOT MRAM, dual-level cells, three-level cells (TLCs), and four-level cells. Further, multi-layer perceptron architectures have been utilized to perform MNIST image classification with these multilevel devices. The performance of the complete system level consisting of crossbar arrays with various MRAM bit cells in terms of area, energy, and latency is evaluated. The throughput efficiency of the BNN accelerator using TLCs is 26.6X, and 3.61X higher than conventional single-bit STT-MRAM, and SOT-MRAM respectively.

Implementation of deep neural networks learning on unmanned aerial vehicle based remote-sensing

Due to efficient and adaptable data collecting, unmanned aerial vehicle (UAV) has been a popular topic in computer vision (CV) and remote sensing (RS) in recent years. Inspiring by the recent success of deep learning (DL), several enhanced object identification and tracking methods have been broadly applied to a variety of UAV-related applications, including environmental monitoring, precision agriculture, and traffic management. In this research, we present efficient neural network (ENet), a unique deep neural network architecture designed exclusively for jobs demanding low latency operation. ENet is up to quicker, takes fewer floating-point operations per second (FLOPs), has fewer parameters, and offers accuracy comparable to or superior to that of previous models. We have tested it on the street and cityscapes reports on comparisons with current state-of-the-art approaches and the tradeoffs between a network's processing speed and accuracy. We give measurements of the proposed architecture's performance on embedded devices and offer software enhancements that might make ENet even quicker.

Implementation of a Convolutional Neural Network Into an Embedded Device for Polyps Detection

The increasing rates of colorectal cancer and associated mortality have attracted interest in the use of computer-aided diagnosis tools based on artificial intelligence (AI) for the detection of polyps at an early stage. Most AI models are implemented on software platforms; however, due to the demands of embedded devices, hardware implementations have to fulfil the demands of real-time applications with better accuracy and low-power consumption. In this letter, we propose an optimized four-layer network that can be implanted into an embedded device and determine the feasibility of implanting our convolutional neural network (CNN) into a microprocessor. The essential functions of the CNN (i.e., padding, convolution, ReLU, max-pooling, fully-connected, and softmax layers) are implemented in the microprocessor. The proposed method achieves efficient classification with high performance and takes only 2.5488mW at a working frequency of 8MHz. We conclude this letter with a discussion of the results and future direction of research.

Design and implementation of neural network based conditions for the CMS Level-1 Global Trigger upgrade for the HL-LHC

Abstract The CMS detector will be upgraded to maintain, or even improve, the physics acceptance under the harsh data taking conditions foreseen during the High-Luminosity LHC operations. In particular, the trigger system (Level-1 and High Level Triggers) will be completely redesigned to utilize detailed information from sub-detectors at the bunch crossing rate: the upgraded Global Trigger will use high-precision trigger objects to provide the Level-1 decision. Besides cut-based algorithms, novel machine-learning-based algorithms will also be included in the Global Trigger to achieve a higher selection efficiency and detect unexpected signals. Implementation of these novel algorithms is presented, focusing on how the neural network models can be optimized to ensure a feasible hardware implementation. The performance and resource usage of the optimized neural network models are discussed in detail.

Wireless Sensor Network Implementation to Prevent Railway Route Accident

Wireless Sensor Network Implementation to Prevent Railway Route Accident

ДОСЛІДЖЕННЯ ЕФЕКТИВНОСТІ ВИКОРИСТАННЯ PNN ДЛЯ РОЗВ’ЯЗАННЯ ЗАДАЧ МЕДИЧНОЇ ДІАГНОСТИКИ В УМОВАХ АНАЛІЗУ МАЛИХ ДАНИХ ВИСОКОЇ РОЗМІРНОСТІ

Parkinson's disease is one of the illnesses that cause certain difficulties at the stage of diagnosis. Recently, there has been a tendency to increase the popularity of the use of artificial intelligence methods as an auxiliary diagnostic tool. A probabilistic neural network (PNN) has become widely used for solving problems in the field of medicine. Despite the rather high efficiency of its use for certain tasks, some aspects of its functioning remain insufficiently researched in practice. Existing scientific works do not pay due attention to the issue of using the optimal distance as a measure of similarity between objects. Calculating the distance between the current data vector and each reference sample vector is the first step in implementing a PNN. The classification accuracy of a neural network of this type depends on its efficiency. In this work, the effectiveness of using different distances in the algorithm of different implementations of a probabilistic neural network for detecting Parkinson's disease based on biomedical voice indicators in the case of small high-dimensional data was investigated. Experimental modeling of three different variants of PNN implementation was carried out using the following distances: Chebyshov, Manhattan, Minkowski, cosine, and Canberra. The results of the study showed different values of the F1-measure when applying different distances. It was found that the use of the Euclidean metric in the structure of a probabilistic neural network is not always the best option. In particular, the application of non-Euclidean metrics provided a significant increase in accuracy for the analyzed dataset. This indicates the need for correctly selecting this parameter of the probabilistic neural network to obtain the highest accuracy when solving medical diagnosis problems.