IDS-INT:Intrusion detection system using transformer-based transfer learning for imbalanced network traffic

A network intrusion detection system is critical for cyber security against llegitimate attacks.In terms of feature perspectives,network traffic may include a variety of elements such as attack reference,attack type,a subcategory of attack,host information,malicious scripts,etc.In terms of network perspectives,network traffic may contain an imbalanced number of harmful attacks when compared to normal traffic.It is challenging to identify a specific attack due to complex features and data imbalance issues.To address these issues,this paper proposes an Intrusion Detection System using transformer-based transfer learning for Imbalanced Network Traffic(IDS-INT).IDS-INT uses transformer-based transfer learning to learn feature interactions in both network feature representation and imbalanced data.First,detailed information about each type of attack is gathered from network interaction descriptions,which include network nodes,attack type,reference,host information,etc.Second,the transformer-based transfer learning approach is developed to learn detailed feature representation using their semantic anchors.Third,the Synthetic Minority Oversampling Technique(SMOTE)is implemented to balance abnormal traffic and detect minority attacks.Fourth,the Convolution Neural Network(CNN)model is designed to extract deep features from the balanced network traffic.Finally,the hybrid approach of the CNN-Long Short-Term Memory(CNN-LSTM)model is developed to detect different types of attacks from the deep features.Detailed experiments are conducted to test the proposed approach using three standard datasets,i.e.,UNsWNB15,CIC-IDS2017,and NSL-KDD.An explainable AI approach is implemented to interpret the proposed method and develop a trustable model.

A single micro-LED manipulation system based on micro-gripper

Micro-LEDs(μLEDs)have advantages in terms of brightness,power consumption,and response speed.In addition,they can also be used as micro-sensors implanted in the body via flexible electronic skin.One of the key techniques involved in the fabrication ofμLED-based devices is transfer printing.Although numerous methods have been proposed for transfer printing,improving the yield ofμLED arrays is still a formidable task.In this paper,we propose a novel method for improving the yield ofμLED arrays transferred by the stamping method,using an innovative design of piezoelectrically driven asymmetric micro-gripper.Traditional grippers are too large to manipulateμLEDs,and therefore two micro-sized cantilevers are added at the gripper tips.AμLED manipulation system is constructed based on the micro-gripper together with a three-dimensional positioning system.Experimental results using this system show that it can be used successfully to manipulateμLED arrays.

Effect of surfactant frequently used in soil flushing on oxygen mass transfer in micro-nano-bubble aeration system

In-site soil flushing and aeration are the typical synergetic remediation technology for contaminated sites.The surfactant present in flushing solutions is bound to affect the aeration efficiency.The purpose of this study is to evaluate the effect of surfactant frequently used in soil flushing on the oxygen mass transfer in micro-nano-bubble(MNB)aeration system.Firstly,bio-surfactants and chemical surfactants were used to investigate their effects on Sauter mean diameter of bubble(dBS),gas holdup(ε),volumetric mass-transfer coefficient(kLa)and liquid-side mass-transfer coefficient(kL)in the MNB aeration system.Then,based upon the experimental results,the Sardeing's and Frossling's models were modified to describe the effect of surfactant on kL in the MNB aeration.The results showed that,for the twenty aqueous surfactant solutions,with the increase in surfactant concentration,the value of dBS,kLa and kL decreased,while the value ofεand gas-liquid interfacial area(a)increased.These phenomena were mainly attributed to the synergistic effects of immobile bubble surface and the suppression of coalescence in the surfactant solutions.In addition,with the presence of electric charge,MNBs in anionic surfactant solutions were smaller and higher in number than in non-ionic surfactant solutions.Furthermore,the accumulation of surfactant on the gas-liquid interface was more conspicuous for small MNB,so the reduction of kL in anionic surfactant solutions was larger than that in non-ionic surfactant solutions.Besides,the modified Frossling's model predicted the effect of surfactant on kL in MNB aeration system with reasonable accuracy.

Interpolated Bumpless Transfer Control for Asynchronously Switched Linear Systems

This paper revisits the problem of bumpless transfer control(BTC) for discrete-time nondeterministic switched linear systems. The general case of asynchronous switching is considered for the first time in the field of BTC for switched systems. A new approach called interpolated bumpless transfer control(IBTC) is proposed, where the bumpless transfer controllers are formulated with the combination of the two adjacent modedependent controller gains, and are interpolated for finite steps once the switching is detected. In contrast with the existing approaches, IBTC does not necessarily run through the full interval of subsystems, as well as possesses the time-varying controller gains(with more flexibility and less conservatism) achieved from a control synthesis allowing for the stability and other performance of the whole switched system. Sufficient conditions ensuring stability and H_(∞) performance of the underlying system by IBTC are developed, and numerical examples verify the theoretical findings.

Deep Learning Based Efficient Crowd Counting System

Estimation of crowd count is becoming crucial nowadays,as it can help in security surveillance,crowd monitoring,and management for different events.It is challenging to determine the approximate crowd size from an image of the crowd’s density.Therefore in this research study,we proposed a multi-headed convolutional neural network architecture-based model for crowd counting,where we divided our proposed model into two main components:(i)the convolutional neural network,which extracts the feature across the whole image that is given to it as an input,and(ii)the multi-headed layers,which make it easier to evaluate density maps to estimate the number of people in the input image and determine their number in the crowd.We employed the available public benchmark crowd-counting datasets UCF CC 50 and ShanghaiTech parts A and B for model training and testing to validate the model’s performance.To analyze the results,we used two metrics Mean Absolute Error(MAE)and Mean Square Error(MSE),and compared the results of the proposed systems with the state-of-art models of crowd counting.The results show the superiority of the proposed system.

Response analysis of NMRG system considering Rb-Xe coupling

The dynamic range of the nuclear magnetic resonance gyroscope can be effectively improved through the closedloop control scheme,which is crucial to its application in inertial measurement.This paper presents the analytical transfer function of Xe closed-loop system in the nuclear magnetic resonance gyroscope considering Rb–Xe coupling effect.It not only considers the dynamic characteristics of the system more comprehensively,but also adds the influence of the practical filters in the gyro signal processing system,which can obtain the accurate response characteristics of signal frequency and amplitude at the same time.The numerical results are compared with an experimentally verified simulation program,which indicate great agreement.The research results of this paper are of great significance to the practical application and development of the nuclear magnetic resonance gyroscope.

Mass Transfer-Promoted Fe^(2+)/Fe^(3+)Circulation Steered by 3D Flow-Through Co-Catalyst System Toward Sustainable Advanced Oxidation Processes

Realizing fast and continuous generation of reactive oxygen species(ROSs)via iron-based advanced oxidation processes(AOPs)is significant in the environmental and biological fields.However,current AOPs assisted by co-catalysts still suffer from the poor mass/electron transfer and non-durable promotion effect,giving rise to the sluggish Fe^(2+)/Fe^(3+)cycle and low dynamic concentration of Fe^(2+)for ROS production.Herein,we present a three-dimensional(3D)macroscale co-catalyst functionalized with molybdenum disulfide(MoS_(2))to achieve ultra-efficient Fe^(2+)regeneration(equilibrium Fe^(2+)ratio of 82.4%)and remarkable stability(more than 20 cycles)via a circulating flow-through process.Unlike the conventional batch-type reactor,experiments and computational fluid dynamics simulations demonstrate that the optimal utilization of the 3D active area under the flow-through mode,initiated by the convectionenhanced mass/charge transfer for Fe^(2+)reduction and then strengthened by MoS_(2)-induced flow rotation for sufficient reactant mixing,is crucial for oxidant activation and subsequent ROS generation.Strikingly,the flow-through co-catalytic system with superwetting capabilities can even tackle the intricate oily wastewater stabilized by different surfactants without the loss of pollutant degradation efficiency.Our findings highlight an innovative co-catalyst system design to expand the applicability of AOPs based technology,especially in large-scale complex wastewater treatment.

Intelligent Power Grid Load Transferring Based on Safe Action-Correction Reinforcement Learning

When a line failure occurs in a power grid, a load transfer is implemented to reconfigure the network by changingthe states of tie-switches and load demands. Computation speed is one of the major performance indicators inpower grid load transfer, as a fast load transfer model can greatly reduce the economic loss of post-fault powergrids. In this study, a reinforcement learning method is developed based on a deep deterministic policy gradient.The tedious training process of the reinforcement learning model can be conducted offline, so the model showssatisfactory performance in real-time operation, indicating that it is suitable for fast load transfer. Consideringthat the reinforcement learning model performs poorly in satisfying safety constraints, a safe action-correctionframework is proposed to modify the learning model. In the framework, the action of load shedding is correctedaccording to sensitivity analysis results under a small discrete increment so as to match the constraints of line flowlimits. The results of case studies indicate that the proposed method is practical for fast and safe power grid loadtransfer.

Deep Learning-Based Mask Identification System Using ResNet Transfer Learning Architecture

Recently,the coronavirus disease 2019 has shown excellent attention in the global community regarding health and the economy.World Health Organization(WHO)and many others advised controlling Corona Virus Disease in 2019.The limited treatment resources,medical resources,and unawareness of immunity is an essential horizon to unfold.Among all resources,wearing a mask is the primary non-pharmaceutical intervention to stop the spreading of the virus caused by Severe Acute Respiratory Syndrome Coronavirus 2(SARS-CoV-2)droplets.All countries made masks mandatory to prevent infection.For such enforcement,automatic and effective face detection systems are crucial.This study presents a face mask identification approach for static photos and real-time movies that distinguishes between images with and without masks.To contribute to society,we worked on mask detection of an individual to adhere to the rule and provide awareness to the public or organization.The paper aims to get detection accuracy using transfer learning from Residual Neural Network 50(ResNet-50)architecture and works on detection localization.The experiment is tested with other popular pre-trained models such as Deep Convolutional Neural Networks(AlexNet),Residual Neural Networks(ResNet),and Visual Geometry Group Networks(VGG-Net)advanced architecture.The proposed system generates an accuracy of 98.4%when modeled using Residual Neural Network 50(ResNet-50).Also,the precision and recall values are proved as better when compared to the existing models.This outstanding work also can be used in video surveillance applications.

Blockchain-Based System for Secure and Efficient Cross-Border Remittances: A Potential Alternative to SWIFT

This paper proposes a blockchain-based system as a secure, efficient, and cost-effective alternative to SWIFT for cross-border remittances. The current SWIFT system faces challenges, including slow settlement times, high transaction costs, and vulnerability to fraud. Leveraging blockchain technology’s decentralized, transparent, and immutable nature, the proposed system aims to address these limitations. Key features include modular architecture, implementation of microservices, and advanced cryptographic protocols. The system incorporates Proof of Stake consensus with BLS signatures, smart contract execution with dynamic pricing, and a decentralized oracle network for currency conversion. A sophisticated risk-based authentication system utilizes Bayesian networks and machine learning for enhanced security. Mathematical models are presented for critical components, including transaction validation, currency conversion, and regulatory compliance. Simulations demonstrate potential improvements in transaction speed and costs. However, challenges such as regulatory hurdles, user adoption, scalability, and integration with legacy systems must be addressed. The paper provides a comparative analysis between the proposed blockchain system and SWIFT, highlighting advantages in transaction speed, costs, and security. Mitigation strategies are proposed for key challenges. Recommendations are made for further research into scaling solutions, regulatory frameworks, and user-centric designs. The adoption of blockchain-based remittances could significantly impact the financial sector, potentially disrupting traditional models and promoting financial inclusion in underserved markets. However, successful implementation will require collaboration between blockchain innovators, financial institutions, and regulators to create an enabling environment for this transformative system.

Numerical optimisation of a classical stochastic system for targeted energy transfer

The paper studies stochastic dynamics of a two-degree-of-freedom system,where a primary linear system is connected to a nonlinear energy sink with cubic stiffness nonlinearity and viscous damping.While the primary mass is subjected to a zero-mean Gaussian white noise excitation,the main objective of this study is to maximise the efficiency of the targeted energy transfer in the system.A surrogate optimisation algorithm is proposed for this purpose and adopted for the stochastic framework.The optimisations are conducted separately for the nonlinear stiffness coefficient alone as well as for both the nonlinear stiffness and damping coefficients together.Three different optimisation cost functions,based on either energy of the system’s components or the dissipated energy,are considered.The results demonstrate some clear trends in values of the nonlinear energy sink coefficients and show the effect of different cost functions on the optimal values of the nonlinear system’s coefficients.

Nonreciprocity of energy transfer in a nonlinear asymmetric oscillator system with various vibration states

The nonreciprocity of energy transfer is constructed in a nonlinear asymmetric oscillator system that comprises two nonlinear oscillators with different parameters placed between two identical linear oscillators.The slow-flow equation of the system is derived by the complexification-averaging method.The semi-analytical solutions to this equation are obtained by the least squares method,which are compared with the numerical solutions obtained by the Runge-Kutta method.The distribution of the average energy in the system is studied under periodic and chaotic vibration states,and the energy transfer along two opposite directions is compared.The effect of the excitation amplitude on the nonreciprocity of the system producing the periodic responses is analyzed,where a three-stage energy transfer phenomenon is observed.In the first stage,the energy transfer along the two opposite directions is approximately equal,whereas in the second stage,the asymmetric energy transfer is observed.The energy transfer is also asymmetric in the third stage,but the direction is reversed compared with the second stage.Moreover,the excitation amplitude for exciting the bifurcation also shows an asymmetric characteristic.Chaotic vibrations are generated around the resonant frequency,irrespective of which linear oscillator is excited.The excitation threshold of these chaotic vibrations is dependent on the linear oscillator that is being excited.In addition,the difference between the energy transfer in the two opposite directions is used to further analyze the nonreciprocity in the system.The results show that the nonreciprocity significantly depends on the excitation frequency and the excitation amplitude.

Electro-enzyme coupling systems for selective reduction of CO_(2)

To address the energy crisis and alleviate the rising level of CO_(2)in the atmosphere,various CO_(2)capture and utilization(CCU)technologies have been developed.The use of electro-enzyme coupling systems is a promising strategy for the sustainable production of fuels,chemicals and materials using CO_(2)as the feedstock.In this review,the recent progresses in the development of electro-enzyme coupling systems for the selective reduction of CO_(2)are systematically summarized.We first provide a brief background about the significance and challenges in the direct conversion of CO_(2)into value-added chemicals.Next,we describe the materials and strategies in the design of electrodes,as well as the common enzymes used in the electro-enzyme coupling systems.Then,we focus on the state-of-the-art routes for the electro-enzyme coupling conversion of CO_(2)into a variety of compounds(formate,CO,methanol,C≥2chemicals)by a single enzyme or multienzyme systems.The emerging approaches and materials used for the construction of electro-enzyme coupling systems to enhance the electron transfer efficiency and the catalytic activity/stability are highlighted.The main challenges and perspectives in the integration of enzymatic and electrochemical strategies are also discussed.

UAV-Aided Data and Energy Integrated Network: System Design and Prototype Development

Terminal devices deployed in outdoor environments are facing a thorny problem of power supply.Data and energy integrated network(DEIN)is a promising technology to solve the problem,which simultaneously transfers data and energy through radio frequency signals.State-of-the-art researches mostly focus on theoretical aspects.By contrast,we provide a complete design and implementation of a fully functioning DEIN system with the support of an unmanned aerial vehicle(UAV).The UAV can be dispatched to areas of interest to remotely recharge batteryless terminals,while collecting essential information from them.Then,the UAV uploads the information to remote base stations.Our system verifies the feasibility of the DEIN in practical applications.

Potential Impacts of Exogenous Nitrogen Enrichment on Distribution and Transfer of Nitrogen in Plant-Soil System of Suaeda salsa Marsh in the Yellow River Estuary, China

To determine the potential impacts of exogenous nitrogen(N)enrichment on distribution and transfer of N in Suaeda salsa marsh in the Yellow River Estuary,the variations of N in plant-soil system during the growing season were investigated by field N addition experiment.The experiment included four treatments:NN(no N input treatment,0gNm^(−2) yr^(−1)),LN(low N input treatment,3.0 gNm^(−2) yr^(−1)),MN(medium N input treatment,6 gNm^(−2) yr^(−1))and HN(high N input treatment,12 gNm^(−2) yr^(−1)).Results showed that N additions generally increased the contents of total nitrogen(TN),ammonia nitrogen(NH_(4)^(+)-N)and nitrate nitrogen(NO_(3)^(−)-N)in different soil layers and the increasing trend was particularly evident in topsoil.Compared with the NN treatment,the average contents of TN in topsoil in the LN,MN and HN treatments during the growing season increased by 10.85%,30.14%and 43.98%,the mean contents of NH_(4)^(+)-N increased by 8.56%,6.96%and 14.34%,and the average contents of NO_(3)^(−)-N increased by 35.73%,45.99%and 46.66%,respectively.Although exogenous N import did not alter the temporal variation patterns of TN contents in organs,the N transfer and accumulation differed among tissues in different treatments.With increasing N import,both the N stocks in soil and plant showed increasing trend and the values in N addition treatments increased by 9.43%–38.22%and 13.40%–62.20%,respectively.It was worth noting that,compared with other treatments,the S.salsa in the MN treatments was very likely to have special response to N enrichment since not only the period of peak growth was prolonged by about 20 days but also the maximum of TN content in leaves was advanced by approximately one month.This paper found that,as N loading reached MN level in future,the growth rhythm of S.salsa and the accumulation and transference of N in its tissues would be altered significantly,which might generate great impact on the stability and health of S.salsa marsh ecosystem.

High-fidelity topological quantum state transfers in a cavity-magnon system

We propose a scheme for realizing high-fidelity topological state transfer via the topological edge states in a onedimensional cavity-magnon system.It is found that the cavity-magnon system can be mapped analytically into the generalized Su-Schrieffer-Heeger model with tunable cavity-magnon coupling.It is shown that the edge state can be served as a quantum channel to realize the photonic and magnonic state transfers by adjusting the coupling strength between adjacent cavity modes.Further,our scheme can realize the quantum state transfer between photonic state and magnonic state by changing the cavity-magnon coupling strength.With the numerical simulation,we quantitatively show that the photonic,magnonic and magnon-to-photon state transfers can be achieved with high fidelity in the cavity-magnon system.Spectacularly,three different types of quantum state transfer schemes can be even transformed into each other in a controllable fashion.The Su-Schrieffer-Heeger model based on the cavity-magnon system provides us a tunable platform to engineer the transport of photon and magnon,which may have potential applications in topological quantum processing.

Developing a Breast Cancer Resistance Protein Substrate Prediction System Using Deep Features and LDA

Breast cancer resistance protein(BCRP)is an important resistance protein that significantly impacts anticancer drug discovery,treatment,and rehabilitation.Early identification of BCRP substrates is quite a challenging task.This study aims to predict early substrate structure,which can help to optimize anticancer drug development and clinical diagnosis.For this study,a novel intelligent approach-based methodology is developed by modifying the ResNet101 model using transfer learning(TL)for automatic deep feature(DF)extraction followed by classification with linear discriminant analysis algorithm(TLRNDF-LDA).This study utilized structural fingerprints,which are exploited by DF contrary to conventional molecular descriptors.The proposed in silico model achieved an outstanding accuracy performance of 98.56%on test data compared to other state-of-the-art approaches using standard quality measures.Furthermore,the model’s efficacy is validated via a statistical analysisANOVAtest.It is demonstrated that the developedmodel can be used effectively for early prediction of the substrate structure.The pipeline of this study is flexible and can be extended for in vitro assessment efficacy of anticancer drug response,identification of BCRP functions in transport experiments,and prediction of prostate or lung cancer cell lines.

Cross-Domain TSK Fuzzy System Based on Semi-Supervised Learning for Epilepsy Classification

Through semi-supervised learning and knowledge inheritance,a novel Takagi-Sugeno-Kang(TSK)fuzzy system framework is proposed for epilepsy data classification in this study.The new method is based on the maximum mean discrepancy(MMD)method and TSK fuzzy system,as a basic model for the classification of epilepsy data.First,formedical data,the interpretability of TSK fuzzy systems can ensure that the prediction results are traceable and safe.Second,in view of the deviation in the data distribution between the real source domain and the target domain,MMD is used to measure the distance between different data distributions.The objective function is constructed according to the MMD distance,and the distribution distance of different datasets is minimized to find the similar characteristics of different datasets.We introduce semi-supervised learning to further explore the relationship between data.Based on the MMD method,a semi-supervised learning(SSL)-MMD method is constructed by using pseudo-tags to realize the data distribution alignment of the same category.In addition,the idea of knowledge dissemination is used to learn pseudo-tags as additional data features.Finally,for epilepsy classification,the cross-domain TSK fuzzy system uses the cross-entropy function as the objective function and adopts the back-propagation strategy to optimize the parameters.The experimental results show that the new method can process complex epilepsy data and identify whether patients have epilepsy.

A Deep Learning-Based Crowd Counting Method and System Implementation on Neural Processing Unit Platform

In this paper, a deep learning-based method is proposed for crowdcountingproblems. Specifically, by utilizing the convolution kernel densitymap, the ground truth is generated dynamically to enhance the featureextractingability of the generator model. Meanwhile, the “cross stage partial”module is integrated into congested scene recognition network (CSRNet) toobtain a lightweight network model. In addition, to compensate for the accuracydrop owing to the lightweight model, we take advantage of “structuredknowledge transfer” to train the model in an end-to-end manner. It aimsto accelerate the fitting speed and enhance the learning ability of the studentmodel. The crowd-counting system solution for edge computing is alsoproposed and implemented on an embedded device equipped with a neuralprocessing unit. Simulations demonstrate the performance improvement ofthe proposed solution in terms of model size, processing speed and accuracy.The performance on the Venice dataset shows that the mean absolute error(MAE) and the root mean squared error (RMSE) of our model drop by32.63% and 39.18% compared with CSRNet. Meanwhile, the performance onthe ShanghaiTech PartB dataset reveals that the MAE and the RMSE of ourmodel are close to those of CSRNet. Therefore, we provide a novel embeddedplatform system scheme for public safety pre-warning applications.

A Novel Efficient Patient Monitoring FER System Using Optimal DL-Features

Automated Facial Expression Recognition(FER)serves as the backbone of patient monitoring systems,security,and surveillance systems.Real-time FER is a challenging task,due to the uncontrolled nature of the environment and poor quality of input frames.In this paper,a novel FER framework has been proposed for patient monitoring.Preprocessing is performed using contrast-limited adaptive enhancement and the dataset is balanced using augmentation.Two lightweight efficient Convolution Neural Network(CNN)models MobileNetV2 and Neural search Architecture Network Mobile(NasNetMobile)are trained,and feature vectors are extracted.The Whale Optimization Algorithm(WOA)is utilized to remove irrelevant features from these vectors.Finally,the optimized features are serially fused to pass them to the classifier.A comprehensive set of experiments were carried out for the evaluation of real-time image datasets FER-2013,MMA,and CK+to report performance based on various metrics.Accuracy results show that the proposed model has achieved 82.5%accuracy and performed better in comparison to the state-of-the-art classification techniques in terms of accuracy.We would like to highlight that the proposed technique has achieved better accuracy by using 2.8 times lesser number of features.