August 2024
Applications of BIG DATA in renewable energy systems based on cloud computing
2024, vol.16 , no.3, pp. 121-128
Article [2024-03-12]
This study examines the potential of microgrids (MG), which utilize renewable energy sources to provide sustainable power solutions. To conduct the analysis, we examined load and photovoltaic (PV) data, calculated minimum and maximum averages, and visualized the correlation using big data tools. We cleaned the data by removing unnecessary rows, merged the tables, converted them into CSV format, and uploaded them to the Databricks file distribution system (DBFS). Subsequently, we processed the data by creating a pipeline and using ETL (extract, transform, load) processes. We analyzed and visualized the data using tools such as Power BI and Tableau. The analysis identified the maximum and minimum PV production, assessed the impact of weather patterns on production, and measured the energy shortage between load demand and PV generation. Our research demonstrates the steps involved in handling and analyzing data, uploading it to the Hadoop ecosystem, transforming it into different file formats, connecting it to a relational database management system (RDBMS), and visualizing it using BI tools. In this study, we utilized cloud infrastructure to perform analytical tasks, including the use of business intelligence (BI) tools.
Tarun Shakthi Sreedhar, Saiful Islam, Meron Atomsa, Elaheh Yazdan Doust, Mohamed Suliman Elnaim, Shomesh Mishra, Venkata Naresh Vemparala, Rupali Bajpai. Applications of BIG DATA in renewable energy systems based on cloud computing. International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 121-128.
Preliminary organization of virtual network based on program monitoring
2024, vol.16 , no.3, pp. 111-120
Article [2024-03-11]
Virtual Local Area Network (VLAN) is a logical organization of communication resources and objects for maintaining efficient information processes based on network traffic. It is known that in the contemporary global network, traffic can be generated from various physical sources, which requires preliminary analysis and adequate management with the provision of the necessary information security. In this aspect, the proposed paper presents an opportunity to analyse supported network traffic in a virtual network environment by using program monitoring. In the computer field, various products are offered for measurement of parameters of information processes, including network communications. In this sense, a preliminary review of the main features of network traffic was conducted and possible solutions for the organization of adequate program measurement of parameters of information processes were discussed.
Radi Romansky . Preliminary organization of virtual network based on program monitoring. International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 111-120.
Analysing the electricity load and production by means of different machine learning methods: A case study of a MG system
2024, vol.16 , no.3, pp. 101-110
Article [2024-03-10]
Renewable energy is a promising solution to combat the scarcity of electricity, particularly in isolated and rural areas. Microgrids (MG) can be employed for installing systems with different energy sources, such as renewable energy components and conventional energy sources like utility grids or grid-connected inverter systems. The amount of energy produced by renewable sources depends on their location, which has implications for energy production. This research aims to explore MG and their challenges for efficient operation. The study discusses various AI models used by researchers to mitigate problems associated with MG planning. Additionally, the paper presents a case study based on the most beneficial ML tool like clustering to gain insights into an existing MG system. The paper also delves into the issues related to PV, a connected distributed energy resource (DER), such as forecasting, and predictive management to reduce maintenance costs, and how AI tools can address them. Furthermore, forecasting methods such as LSTM and GRU models are discussed because of the stochastic nature of PV production.
Saiful Islam , Amin Suaad , Michael Hartmann , Goran Rafajlovski . Analysing the electricity load and production by means of different machine learning methods: A case study of a MG system. International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 101-110.
A hybrid method for designing power electronic devices based on their life cycle
2024, vol.16 , no.3, pp. 91-100
Article [2024-03-09]
The work presents a hybrid approach to the design of power electronic devices, which combines various techniques with the aim of achieving optimal coverage of certain indicators under set restrictions and requirements. Thus, by jointly using state-of-the-art means of applied mathematics, computer modelling, artificial intelligence techniques and information and communication technologies, a new method for the optimal design of power electronic devices, oriented to their life cycle, considering the specifics is described for different applications.
Nikolay Hinov. A hybrid method for designing power electronic devices based on their life cycle . International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 91-100.
Modified genetic algorithm for unit commitment of grid connected microgrids under real time pricing conditions
2024, vol.16 , no.3, pp. 81-90
Article [2024-03-08]
This paper introduces a modification of the genetic algorithm aimed at enhancing the selection process for reproducing the next generation. This modification accelerates the optimization process and improves the outcome. The case study analyzes a grid-connected microgrid comprising renewable energy sources, a battery storage system, prosumers with installed photovoltaic generators, and consumers. The effectiveness of the proposed modification is validated through comparison with two selection algorithms commonly used in the standard genetic algorithms.
Natasha Dimishkovska Krsteski, Atanas Iliev. Modified genetic algorithm for unit commitment of grid connected microgrids under real time pricing conditions. International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 81-90.
A machine learning and natural language processing-based smishing detection model for mobile money transactions
2024, vol.16 , no.3, pp. 69-80
Article [2024-03-07]
As mobile services proliferate to include financial transactions, the threat of phishing attacks targeting users has equally been escalating. Attackers have been using different kinds of phishing techniques, especially in third world where mobile services are prevalent. As such, this paper presents a Smishing (SMS phishing) Detection model leveraging Natural Language Processing (NLP) and Machine Learning (ML) techniques. It aims to detect smishing threats in real-time by the integrating NLP with ML The developed model harnesses NLP algorithms to analyse text- based messages, scrutinizing linguistic patterns and contextual clues indicative of smishing attempts. Through ML algorithms, the model learns to distinguish between legitimate (Non-Smishing) and fraudulent messages (Smishing), adapting dynamically to evolving smishing tactics. The model's efficacy is evaluated through comprehensive testing, demonstrating promising results of precision, recall, and accuracy with F-1 measure at 0.902 and AUC at 0.95. The Model stands as a proactive defence mechanism against smishing in mobile money environments, contributing to enhanced user security and trust in financial transactions.
Aaron Zimba, Katongo O. Phiri, Chimanga Kashale, Mwiza Norina Phiri. A machine learning and natural language processing-based smishing detection model for mobile money transactions . International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 69-80.
HEALTHSOLID 4.0: A novel solid-pod and blockchain-enabled framework for role-based access control and secure healthcare information exchange
2024, vol.16 , no.3, pp. 57-68
Article [2024-03-06]
Several blockchain-based models have surfaced in recent years to offer a safe method of storing and accessing delicate electronic medical records (EMRs) and role-based access control mechanisms across the healthcare industry. However, single points of failure, Security, privacy and unauthorized access of data are measure concerns for health records access management. Blockchain technology is a decentralized digital ledger that records transactions across multiple computers to ensure security, transparency, and immutability. Solid is a decentralized platform that provides patient centric access control of Electronic Medical Records (EMRs). In this study, a blockchain and solid-pod enabled secure framework has been proposed for EMR transactions and healthcare clouds. This has a significant impact on how quickly and easily emergency EMRs may be shared in a smart healthcare system. Solid and Blockchain prioritize privacy and security by enabling users to set immutable, transparent, and fine-grained access controls for their data using solidity-based smart contracts. For the performance evaluation of our proposed system, solid-pod computing and storage results are analysed.
Avani Dadhanial, Hiren Patel. HEALTHSOLID 4.0: A novel solid-pod and blockchain-enabled framework for role-based access control and secure healthcare information exchange. International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 57-68.
Modeling of neural network monitoring agent to predict traffic spikes and agent training
2024, vol.16 , no.3, pp. 49-56
Article [2024-03-05]
The article completes the study of solutions to the problem of creating distributed QoS monitoring control in IoT and IIoT telecommunications. A multi-agent architecture is used. For the proposed structure of the LSTM network, approaches to the process of its training (configuring the network configuration) are considered, based on the representation of this process in the form of a Markov decision-making process, which made it possible to solve the problem of determining the training times of the network within one epoch, as well as its retraining using a variant of the dynamic programming method – the iteration algorithm by values, which In the future, it allows you to proceed to the synthesis of an iterative algorithm for dynamically configuring the parameters of the developed LSTM network.
O. Ja. Kravets, I. A. Aksenov, Yu. V. Redkin, P. A. Rahman, M. V. Kochegarov, A. V. Gorshkov, S. A. Sorokin . Modeling of neural network monitoring agent to predict traffic spikes and agent training . International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 49-56.
Deep hybrid neural networks for prediction missing segments in sEMG time series data
2024, vol.16 , no.3, pp. 37-48
Article [2024-03-04]
Surface electromyography (sEMG) has illustrated noteworthy findings over different disciplines; however, it suffers from several issues like signal interference, noise, and interruptions. In this research, the Savitzky-Golay filter was first used to extract meaningful data while preserving the overall shape of the data, and then two hybrid neural network models based on Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) were performed to predict the missing sEMG data. Their performance was compared with independent LSTM and GRU models using the coefficient of determination (R-squared), Root Mean Square Error (RMSE), and correlation coefficient (ρ). All models were trained, validated, and tested on extended and limited datasets. In addition, the optimal number of hidden neurons was determined experimentally for each condition. The outcomes indicated that the deep learning architecture based on sequential GRU and LSTM models outperformed all competitors with a prediction accuracy of 99.25% and 98.91% for the long and short training datasets, respectively.
Jihane Ben Slimane. Deep hybrid neural networks for prediction missing segments in sEMG time series data. International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 37-48.
A neural network-based collaborative filtering model for social recommendation systems
2024, vol.16 , no.3, pp. 27-36
Article [2024-03-03]
The exponential rise in the popularity of social networks in recent years has led to a corresponding surge in online data, creating a vast pool of information. This growing dataset has captivated the interest of researchers, compelling them to devise recommendation algorithms designed to navigate through this immense sea of data to establish meaningful friendships. Despite the success of CF, there is a notable gap in the related research, as deep learning’s potential to generate friendship recommendations in social networks using CF has been neglected. This paper seeks to fill this gap by introducing a novel, collaborative friendship recommendation system. The proposed system leverages the power of Neural Collaborative Filtering (NCF) to offer users a platform for discovering new friends who are highly likely to engage interactively. To underscore the effectiveness of our model, we present experimental results using a real dataset, showcasing the system’s capability to deliver accurate and valuable friend recommendations
Aadil Alshammari, Rakan Alanazi. A neural network-based collaborative filtering model for social recommendation systems . International Journal on Information Technologies and Security, vol.16 , no.3, 2024, pp. 27-36.