International Journal of Electrical and Computer Engineering Systems

A New Proposed Triple Active Bridge Converter for Fuel Cell Applications: Study, Control and Energy Management

2024-07-19T21:35:20+02:00

This paper deals with a new proposed three port converter structure dedicated for two-input source hybrid systems especially for fuel cell applications. This converter is made up of three-phase triple active bridges which are galvanically isolated by means of three single phase high frequency transformers. The present converter integrates a fuel cell as the primary power source with a battery that stores energy, harnessing the unique benefits of both sources to deliver reliable power to a DC load through a single power conversion stage. In order to control the power flow between the ports, a phase shift control technique has been carried out to generate the control signals of the load and battery side bridges in reference with those of the fuel cell bridge. A detailed analysis of the proposed converter has been presented in this paper. A novel proposed energy management algorithm has been developed. This algorithm provides a robust solution for managing and distributing power flow between the converter's ports, ensuring an optimal balance of power delivery. The algorithm has been rigorously validated through simulations and experimental test, using Dspace 1104 board.

Implementation of Cyber Network’s Attacks Detection System with Deep Learning Designing Algorithms

2024-07-17T10:44:04+02:00

The internet has become indispensable for modern communication, playing a vital role in the development of smart cities and communities. However, its effectiveness is contingent upon its security and resilience against interruptions. Intrusions, defined as unauthorized activities that compromise system integrity, pose a significant threat. These intrusions can be broadly categorized into host intrusions, which involve unauthorized access and manipulation of data within a system, and network intrusions, which target vulnerabilities within the network infrastructure. To mitigate these threats, system administrators rely on Network Intrusion Detection Systems (NIDS) to identify and respond to security breaches. However, designing an effective and adaptable NIDS capable of handling novel and evolving attack strategies presents a significant challenge. This paper proposes a deep learning-based approach for NIDS development, leveraging Self-Taught Learning (STL) and the NSL-KDD benchmark dataset for network intrusion detection. The proposed approach is evaluated using established metrics, including accuracy, F-measure, recall, and precision. Experimental results demonstrate the effectiveness of STL in the 5-class categorization, achieving an accuracy of 79.10% and an F-measure of 75.76%. This performance surpasses that of Softmax Regression (SMR), which attained 75.23% accuracy and a 72.14% F-measure. The paper concludes by comparing the proposed approach's performance with existing state-of-the-art methods.

Mask FORD-NET: Efficient Detection of Digital Image Forgery using Hybrid REG-NET based Mask-RCNN

2024-07-12T17:11:05+02:00

Digital image is a binary representation of visual data which provides a rapid method for analyzing large quantities of data. Furthermore, digital images are more vulnerable to fraud when distributed over an open channel via information and communication technology. However, the image data can be modified fraudulently by intruders using vulnerabilities in telecommunications infrastructure. To overcome these issues, this paper proposes a novel Mask-RCNN based Image FORgery Detection (Mask FORD-NET) which is developed for digital image forgery detection. Initially, the input image is passed beyond the recompression module to reduce the insignificance and complexity of the image to preserve or transfer the data efficiently. After image recompression, the recompressed image is transferred to the feature extraction phase which is done by using REG-NET. The extracted features are received to the noise cancellation and ELA converter module to analyze and reduce the ambient noise. After noise cancellation, the data are passed to the MASK-RCNN module, to detect and classify the forged images and finally provide the segmented output. The Mask FORD-NET framework is simulated by using MATLAB. The efficiency of the proposed Mask FORD-NET framework is assessed by using accuracy, precision, recall, and F1-measure. The experimental results show that the accuracy of the Mask FORD-NET framework has increased to up to 98.72% for digital image forgery detection. The accuracy of the proposed Mask FORD-NET framework is 80.72%, 86.32%, and 95.00% better than existing ASCA, VixNet, and MiniNet techniques respectively.

Deep Reinforcement Learning for Dynamic Task Scheduling in Edge-Cloud Environments

2024-06-03T17:43:40+02:00

With The advent of the Internet of Things (IoT) and its use cases there is a necessity for improved latency which has led to edgecomputing technologies. IoT applications need a cloud environment and appropriate scheduling based on the underlying requirements of a given workload. Due to the mobility nature of IoT devices and resource constraints and resource heterogeneity, IoT application tasks need more efficient scheduling which is a challenging problem. The existing conventional and deep learning scheduling techniques have limitations such as lack of adaptability, issues with synchronous nature and inability to deal with temporal patterns in the workloads. To address these issues, we proposed a learning-based framework known as the Deep Reinforcement Learning Framework (DRLF). This is designed in such a way that it exploits Deep Reinforcement Learning (DRL) with underlying mechanisms and enhanced deep network architecture based on Recurrent Neural Network (RNN). We also proposed an algorithm named Reinforcement Learning Dynamic Scheduling (RLbDS) which exploits different hyperparameters and DRL-based decision-making for efficient scheduling. Real-time traces of edge-cloud infrastructure are used for empirical study. We implemented our framework by defining new classes for CloudSim and iFogSim simulation frameworks. Our empirical study has revealed that RLbDS out performs many existing scheduling methods.

EMF Exposure Reduction Using Weighted Angle Model for Multi-Technology Sectorized BS

2024-08-21T10:19:40+02:00

Mobile networks are now growing quickly due to major advancements in wireless technology especially with the introduction of the Fifth Generation New Radio (5G NR). A greater risk of exposure to electromagnetic field radiation (EMF) is being raised by the widespread deployment of base stations (BSs). Standard guidelines are set to control the amount of EMF radiation. This paper proposed a design model to de-concentrate and reduce the total exposure of the multi-technology BS with no drawbacks on network coverage level and key performance indicators (KPIs). The proposed solution applies the concept of weighted antenna’s azimuth to spread the total exposure by separating the antennas in the same sector. A set of simulations is carried out to calculate the reduction in total exposure ratio (TER) and the Compliance Distance (CD). Also, A field measurement test was done in a life network to validate and evaluate the proposed model under real conditions. Furthermore, the network operation support system (OSS) records were analyzed to evaluate the impact on the network coverage and capacity behavior. The pre- and post-results demonstrate that using the proposed model enhanced the CD and TER., the results show using two azimuths reduces the CD by 23% and by 43.4% when using six antennas. Also, the field test result demonstrated a 19.23% reduction in the Total Exposure Ratio. Overall, the system records show no significant impacts were registered on network coverage level and capacity performance for the sites involved in the test.

Comparison Between Different Source Localization and Connectivity Metrics of Spiky and Oscillatory MEG Activities

2024-08-23T15:46:54+02:00

Epilepsy is considered the second neurological disease in a coma after stroke. Famous markers of epilepsy are repetitive seizures, their origin is stroma and cortical deformation. A neurologist would be assisted by identifying Epileptogenic Zones EZ when diagnosing epilepsy.. Source localization is utilized to identify regions known as EZ, which are of excessive discharges. It consists of both forward and inverse problems. The forward problem models the head through analytical and numerical methods. The inverse problem can be resolved using several techniques to locate the cerebral abnormal sources, via the electrophysiological recording biomarkers. In our study, we will investigate four distributed inverse problem methods: minimum norm estimation MNE, standardized low-resolution brain electromagnetic tomography sLORETA, maximum entropy on the mean MEM, Dynamic statistical parametric maps dSPM, to define epileptic networks connectivity of spiky and oscillatory events. We will examine the epileptic network connectivity using Phase Locking Value (PLV), Phase Transfert Entropy (PTE) for oscillatory events, cross-correlation (CC), and Granger Causality (GC) for spiky events applied on 5 pharmaco resistant subjects. We suggest rating the effectiveness of these networks in locating EZ through a phase of confrontation within iEEG transitory and oscillatory networks connectivity by exploring concordant nodes, their distance, propagation delays connection strength, and their cooperation in recognition of seizure onset zone. All studied techniques of the inverse problem, connection metrics, for both biomarkers of the 5 patients succeed in detecting at least one part of SOZ. sLORETA provides the highest concordant nodes and the closed one for spiky events using CC and GC. sLORETA also depicts the lowest propagation delay for oscillatory events using PTE. Through the 5 patients, MEM, dSPM, and MNE using CC, CG for spiky events, and PTE, PLV for oscillatory activities provide about 72 % of concordant nodes between MEG and iEEG.

An Enhancement of Grid Integration in Renewable Energy Systems Using Multi- Objective Multi-Verse Optimization

2024-07-02T13:49:32+02:00

Going by the recent trends, the application of Renewable Energy Sources (RESs) has grown significantly across the world. Still, the integration of grid with photovoltaic (PV), wind and battery, remains a critical challenge as it results into power quality issues. To address the increasing need for electricity caused by industrialization and growing population, hybrid PV, wind, and battery combinations are used in this study. To accomplish an optimal energy management, this research proposes the multi- objective multi-verse optimization (MOMVO) approach, along with the modified perturb and observe (MP&O) technique. The proposed MOMVO-MP&O controller operates between the wind turbine and the battery storage system, for providing optimal power distribution and stability. The suggested model is evaluated alongside three other popular controller combinations that are, multi-verse optimization-perturb and observe (MVO-P&O), MVO-MP&O, and MOMVO-P&O. A comparative analysis is conducted, with existing methods namely, Modified-Fuzzy Direct Power Control (MF-DPC) and Adaptive Neuro-Fuzzy Inference System (ANFIS) also. From the analyzed results of this comparison, the proposed MOMVO-MP&O achieves lesser Total Harmonic Distortion (THD) of 1.86%, which demonstrates its efficiency in addressing power quality issues using hybrid RES systems.

Scaling and Dynamic Resource Reallocation in NFV: Challenges and Research Perspectives

2024-08-09T13:42:50+02:00

Network Function Virtualization (NFV) has brought incredible experiences for Internet users and network operators. NFV enables the implementation of Virtualized Network Functions (VNFs) as software running in High Volume Servers (HVSs) to execute a Service Function Chain (SFC) to satisfy service demands of Internet users. During the execution of SFCs, VNFs and Virtual Links (VLs) tend to change their resource requirements due to the dynamic nature of the end user's demands. In this paper, we focus on dynamic resource allocation to the elements of SFC throughout the SFC process to adapt to the elasticity in demand from users by providing an overall picture of NFV and the scaling problem of SFC. We then review and analyze related studies on dynamic resource allocation of NFV systems during SFC operation and analyze the results of these projects. The most recent works are also classified based on several criteria to highlight their approaches, achievements, and also shortcomings. Finally, we introduce some research directions to deal with the scaling problem during SFC operation that needs more attention from researchers to inspire future work in the elastic operation of NFV-enabled systems.