		FeedBack \| Contact Us \| Links \| Site Map

Welcome to Journal of Theoretical and Applied Information Technology

Submit Paper / Call for Papers

Journal receives papers in continuous flow and we will consider articles from a wide range of Information Technology disciplines encompassing the most basic research to the most innovative technologies. Please submit your papers electronically to our submission system at http://jatit.org/submit_paper.php in an MSWord, Pdf or compatible format so that they may be evaluated for publication in the upcoming issue. This journal uses a blinded review process; please remember to include all your personal identifiable information in the manuscript before submitting it for review, we will edit the necessary information at our side. Submissions to JATIT should be full research / review papers (properly indicated below main title).

Journal of Theoretical and Applied Information Technology
September 2021 | Vol. 99 No.17

Title:	PERFORMANCE MEASUREMENT OF VOLLEYBALL PLAYER IN ATTACKER POSITION THROUGH OBJECT DETECTION AND TRACKING TECHNIQUES
Author:	BALAJI.S.R, S. KARTHIKEYAN, R. MANIKANDAN
Abstract:	Individual Performance Measurement of volleyball player is very much essential for the trainers to select the players for the tournaments. Usually, this kind of measurement is carried out by capturing the videos during the test session. But it involves plenty of challenges like occlusion, fast moving videos, complex body actions of players etc. In this paper, we proposed a method to measure the performance of the player when they are in test session. Videos are captured while the players are in test session. The players are detected and tracked from the videos for performance measurement. The work proposed here uses Metaheuristic algorithm for object detection and tracking. Later, we have taken two parameters such as height of jump measurement from ground level when the player leaps to hit the ball and then the arm speed of the player when the player tries to hit the ball. Finally, the average among these parameters for the five attempts is calculated for each player to measure the best performer easily. Based on these two parameters the trainer selects the player for attacker position in the tournament. From the result it was observed that the trainer can select Player 3 as the best player since this player has good average height of jump around 56.1cm and average arm speed of 17.9m/s which is better than the other players.
Keywords:	Metaheuristic algorithm, Height of jump, Arm speed, Attacker position, Player Detection, Player Tracking
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	ONTOLOGY FOR CORPUS OF OCCUPATION AND COMPETENCY STANDARDS
Author:	SYAMSI RUHAMA, DETTY PURNAMASARI, I MADE WIRYANA, ADANG SUHENDRA
Abstract:	The ability of personnel in the field of information and communication technology varies greatly, in line with the emergence of various formal and informal educational institutions in this field. Increasing the capabilities and competencies of Human Resources in accordance with the demands of the global labor market needs a reciprocal relationship between the providers of Human Resources and the industrial world in need. There are many occupations in the Information and Communication Technology (ICT) field, each occupation has several competencies. Each occupation has more than one competency. The many occupations and competencies, in Indonesia, are grouped under the name Occupation Map and Competency Standards (CS). The first thing to do is review the Occupational Map document and Competency Standards (CS) document. The results of the analysis of the two documents produce elements that can be used to build ontologies. So that based on this, in this study a method was developed to produce ontology from Occupational Map and Competency Standards (CS). The results obtained from this study are the Occupational Map ontology and Competency Standards (CS) ontology. The ontology concept of the Occupational Map and Competency Standards (CS) can be used as the basis for making the Occupational Map corpus and the Competency Standards (CS) corpus. The Occupational Corpus and Competency Standards (CS) Corpus can be used by governments, private companies or industry when making job vacancies information or looking for prospective employees and that can be used to see the suitability between occupation and competence.
Keywords:	Ontology, Occupation, Competency Occupation Map, Competency Standards, Corpus
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	ENHANCING DISTRIBUTED AGILE TEAMS’ ADOPTION DURING THE COVID-19 PANDEMIC USING TAGICK ACTIVITIES
Author:	MONA RASMY, OSAMA ABU-ELNASR, SAMIR ELMOUGY
Abstract:	Due to the viral outbreak of the coronavirus (COVID-19) pandemic, organizations have adhered to social distancing and/or lockdown measurements. Project teams have shifted from direct communication in the workplace to remote working. Obviously, such sudden changes have led the organizations to face many challenges. A number of these challenges are related to the communication and cooperation among the team members and also to fulfilling the code documentation process that helps the team later in the maintenance phase. In this paper, the researchers propose a modification to the Scrum methodology called the Distributed Scrum (Di-Scrum). It is designed based on suggesting activities called the TAGICK activities (Timekeeping; Aggregation; Groupthink; Interconnectedness; Continuous documentation; and Knowledge transfer). Moreover, each activity supports the agile principles and rules. Additionally, these activities help the team members improve their performance. They also enable them to overcome all the obstacles mentioned above which they face while working remotely. We have used a questionnaire to evaluate the suggested activities. The questionnaire is filled in by 40 different employees in four software development companies applying the Di-Scrum model. The results of the evaluation indicate the effectiveness of the TAGICK activities with remote teams. They have led to enhancement of group communication, cooperation and ability of continuous documentation.
Keywords:	Distributed agile, Distributed Teams, Work Remotely, COVID-19, Software Development, Communications Theory, Scrum Methodology
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	A THEORETICAL DISCUSSION OF FACTORS AFFECTING THE ACCEPTANCE OF M-LEARNING BY INTEGRATING IS SUCCESS MODEL WITH TAM MODEL
Author:	MOHAMMAD ELAYAN AYED ALSHARAFAT
Abstract:	The There have been tremendous developments in different industry fields, including the higher education sector. In fact, such sector has been considered as one of the sectors that attracted the special focus of the government. There are several benefits of M-learning that are attractive to both educator and learner circles, with the top advantage being its mobility. In other words, M-learning allows users to exchange information at any time and at any location, removing the issue of locality and traveling to access learning. However, M-learning in the institutions of higher learning is still in its infancy and tangible M-learning initiatives have yet to be set up in the sector. Added to this, the factors impacting the acceptance of such technology are still ambiguous. This study provides advanced knowledge by explaining the importance, adaption strategies and determinants of mobile Learning in the higher education. This in turn would play an important role in increasing the acceptance level of M-learning among universities.
Keywords:	Higher Education, Information and Communication Technology, Electronic Learning, Mobile Learning, Technology Acceptance Model, Information Systems Success Model
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	AN OPTIMUM LOCATION OF DISTRIBUTED GENERATION FOR SOCIAL SURPLUS MAXIMIZATION IN DEREGULATED ELECTRICITY MARKET WITH INELASTIC LOADS
Author:	RACHAPPA CHIMIRELA, GOPALA RAO JAMMI, VIJAYA KUMAR M
Abstract:	This paper presents the impact of Distribution Generation (DG) on congestion, loss, Locational Marginal Pricing (LMP), and Social Surplus in the Optimum Power Flow (OPF) based restructured electricity market. The issue of perfect placement of DG to reduce congestion and also lower LMPs is formulated with the objective of social surplus maximization. In this work, the seed genetic algorithm method by using DC Optimal Power Flow (DCOPF) is proposed to calculate LMPs at all buses while maximizing social surplus or minimizing fuel cost. Different scenarios for LMP determination i.e. not considering losses, losses are considered but concentrated at reference bus, and losses are distributed at all buses have been examined. Linear bids are assumed for generators. Here, the load is considered as fixed i.e. inelastic. The impact of DG on loss, congestion, LMP, and social surplus has been presented in IEEE 14- Bus system.
Keywords:	DC Optimal Power Flow, Distributed Generation, Electricity Market, Locational Marginal Pricing, and Social Welfare.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	PROPOSED EFFECTIVE LEARNING DESIGN MODEL ON LANGUAGE SKILLS USING PARTIAL LEAST SQUARES STRUCTURAL EQUATION MODELING (PLS-SEM)
Author:	HASMAINI HASHIM, SAZILAH SALAM, SITI NURUL MAHFUZAH MOHAMAD, MOHD MAWARDY ABDULLAH, JACK FEBRIAN RUSDI
Abstract:	The success in learning is the most significant issue in the realm of language learning such as lack of studies focusing on the efficacy of fully online language learning. Mostly existing researches were conducted for face-to-face and blended learning modes. Activities are an important part in language learning which involve skills for reading, writing, listening, and speaking. Existing studies also mostly focus only on partial skill activities. Therefore, this study proposes an effective learning design model based on the relationship between learner’s characteristics and their self-efficacy in language skills. A survey design was developed to collect data and test the proposed model. A Partial Least Squares Structural Equation Modeling (PLS-SEM) was used in order to answer research questions. The survey was conducted on-line involving 130 respondents after completing Mandarin MOOC course. The results of the model shows two constructs from learner characteristics have positive relationships and significant to the effective learning design model, that one dimension for learning styles (visual, p value = 0.000) and one dimension for cognitive styles, (thinking, p value = 0.000). The findings show that visual and thinking dimensions contribute to the effective learning design model that shows improvement in student performance for all language skills: listening, speaking, reading and writing.
Keywords:	MOOCs, learner characteristics, learning design, student performance, self-efficacy
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	BIOINSPIRED RELIABLE SUPPORT VECTOR MACHINE FOR ANALYZING SENTIMENTS IN BIG DATA
Author:	J. RATHIKA, M. SORANAMAGESWARI
Abstract:	Real-time shopping and E-commerce benefit greatly from big data analytics and it is employed to increase sales of items and improve customer-retailer interaction. Shops increasingly employ internet marketing to identify top places to acquire quality items. The shopper's buying experience and thoughts about the retailer might be detected by observing social media activity on both sides. Sentimental Analysis is a great tool for identifying the emotional or emotional impact of the contents. SA investigates people's feelings and thoughts on various things. Weak sentimental analysis worsens the significance of gaining insight into customers' feelings when purchasing products and also it diminishes the impact of knowing the customers' perception about a product. This paper proposes a Bioinspired Reliable Support Vector Machine (BRSVM) for performing the sentiment analysis in big data. BRSVM is inspired from foraging behavior of ants and it is used to identify the sentiments in big review dataset. Lagrangian strategy utilized in BRSVM assist in achieving better optimization. Proposed classifier has been evaluated with fourproduct review big dataset using accuracy and f-measure performance metrics. Results make an indication that the proposed classifier attains better classification accuracy than existing classifiers.
Keywords:	Ant Colony, Classification, Sentiment Analysis, SVM, Amazon, Optimization
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	A MODEL OF VIDEO WATCHING CONCENTRATION LEVEL MEASUREMENT AMONG STUDENTS USING HEAD POSE AND EYE TRACKING DETECTION
Author:	AYAT ALRAWAHNEH, DR. SUHAILAN BIN SAFEI
Abstract:	E-learning has become popular across countries in recent years because of its flexibility, availability, and accessibility. During the COVID-19 pandemic, most schools and universities have switched to embrace E-learning for teaching purposes. One of the big challenges in online teaching is to monitor student’s learning engagement especially in watching video learning materials. The concentration level of students in watching the video is among the important factors for having effective learning. While this aspect is essential, there has also been limited work to develop it. Although there have been a variety of studies and implementations for student concentration monitoring and evaluation in pervasive learning, the majority of the current works are either based on questionnaires, self-reports, instructor introspective assessments, and assignments or are assisted by commercial eye tracking devices/software. In this paper, we propose a model for measure a student's concentration level in an e-learning environment, using hybrid methods combining head pose and eye tracking detection. Specifically, obtain the information through the webcam of the student's head pose rotation and eye-tracking technique to discriminate the behavior of the students during the E-learning classes. For the performance evaluation experiment, we used a 5-minute E-learning video watched by 10 participants. We observe the scenario from an expert perspective and evaluate the students' high and low concentration levels. The video is then benchmarked with our model to see the model accuracy with respect to the expert decision. Based on this experiment, 80% of the result were similarly matched with the expert's opinion. Our extensive evaluations, which included different scenarios, and an expert perspective, demonstrate that the proposed method is effective and usable. This technique will enhance the student’s concentration and make the session more interesting and updated. This approach could readily be expanded to generate real-time alerts for remote live monitoring in E-learning classes.
Keywords:	E-learning, Concentration Level, Facial landmark, Head Pose estimation, Eye Aspect Ratio.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	SENTIMENT ANALYSIS OF INDONESIAN E-COMMERCE PRODUCT REVIEWS USING SUPPORT VECTOR MACHINE BASED TERM FREQUENCY INVERSE DOCUMENT FREQUENCY
Author:	SITI FIDYANTI NURFADILA HADJU, RIYANTO JAYADI
Abstract:	E-commerce transaction has grown eight times every year in Indonesia. Meanwhile, the number of Indonesian e-commerce customers grows twice every year. The Indonesian e-commerce market has attracted prospects both in terms of management strategies and customer opinions. Therefore, this study aims to provide another perspective to e-commerce management that product reviews reviewed by users can be parameters in determining management strategy. Existing reviews can be collected, processed, and predicted into product segmentation data using the CRISP-DM method. To achieve this goal, the authors compare several machine learning algorithms to determine the best way for sentiment analysis on product reviews contained in five E-Commerce. in Indonesia. Data were collected from product reviews in 3 broad categories: telephone, groceries, and fashion. Then this data is placed by comparing four algorithms, i.e., Decisions Tree, Random Forest, Gradient Boost, and Super Vector Machine). It can be denied that the SVM method is the best method with an accuracy rate of 95.875%. After the prediction model is modelled, the deployment involves connecting the prediction to the dashboard software to display it to E-commerce management and related sellers.
Keywords:	Sentiment Analysis, Product Review, E-commerce.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	DVFS ENABLED GREEDY MECHANISM FOR ENERGY AWARE TASK SCHEDULING IN HIGH PERFORMANCE COMPUTING SYSTEMS
Author:	SIDDESHA K , JAYARAMAIAH G V
Abstract:	Nowadays, the high performance computing (HPC) systems are widely adopted in data centers to process and accomplish the tasks which consume more resources. These tasks increase load on the data centers however, HPC systems utilize multicore processors which facilitate faster computing. However, inappropriate use of these multicore systems leads to excessive energy consumption and wastage of computing resource. Dynamic voltage and frequency scaling is one of the promising solution to overcome energy consumption related issue in most of the processor based systems. In this paper, we present a combined energy aware scheduling model to schedule multiple types of heterogeneous tasks along with DVFS for high performance computing platform. We introduce a schedule architecture model which acts as a central scheduler along with DVFS. First of all, the incoming tasks are analyzed based on their deadline constraints and resource requirement. DVFS scheme is applied to adjust the voltage and frequency based on the core’s configuration. Later, the proposed greedy energy aware scheduling approach assigns the tasks to the specific processor, based on the response time and probability value of the task. This problem is formulated as an optimization process. Moreover, the proposed technique minimizes the queuing delay, which helps to minimize the delay in scheduling. The comparative analysis shows that, the average waiting time and average energy saving ratio of the proposed approach is improved.
Keywords:	DVFS, Task scheduling, Multicore processors, Energy aware, Greedy mechanism.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	DEVELOPING A BILINGUAL MODEL OF WORD EMBEDDING FOR DETECTING INDONESIAN ENGLISH PLAGIARISM
Author:	YULYANI ARIFIN, SANI M. ISA, LILI AYU WULANDHARI, EDI ABDURACHMAN
Abstract:	The Internet users that increasing can make it easier to access information even in different languages. Also, the translation application can help users to translate some idea or document without proper citation or acknowledge their idea So, plagiarism is increasing not only in the academic field but also in the industry. A lot of researchers already propose some method to detect plagiarism, but mostly in the European language. Previous research in Indonesian-English plagiarism has already proposed some methods but it is still dependent on machine translation. So, from this research, we purpose a model that can be used to detect cross-language plagiarism without depending on machine translation. The model's purpose is to use combination canonical correlation analysis with the paragraph to vector. Evaluation will be done with the monolingual task and cross-language detection plagiarism. The model evaluation has a good result in monolingual word similarity also when detecting cross-language plagiarism without depending on machine translation. After comparing with the benchmark that using Fingerprint Method with machine translation, the proposed method can detect plagiarism type with paraphrasing more accurately than the benchmark. Even the improvement compared with the benchmark not so significantly but through this proposed method can detect cross-language plagiarism in Indonesian-English language without depending on machine translation. For future work, it needs to enlarge the parallel corpus for Indonesian-English to improve the accuracy of the proposed method.
Keywords:	Word Embeddings, Plagiarism, Bilingual Model, Cross-Lingual, Canonical Correlation Analysis
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	SARDINE FEAST METAHEURISTIC OPTIMIZATION: AN ALGORITHM BASED ON SARDINE FEEDING FRENZY
Author:	MOHAMMAD FAIDZUL NASRUDIN, FITRANTO KUSUMO, DWI YANUAR PANJI TRESNA, MOHD SAIFUL SYAHMI SAIFUDDIN, LIZAWATI MI YUSUF
Abstract:	Many metaheuristics mimic biological interaction metaphors, such as ant colony, particle swarm, bee foraging, eagle predator behavior, and cuckoo brood parasitism, to solve complex optimization problems. Another type of biological interaction is commensalism, where one species obtains food from the other without harming or benefiting the latter. One of the great objective-driven commensalism phenomena that amazes scientists and has not yet been modeled is the sardine feast. In this study, we create an optimization algorithm, the sardine feast metaheuristic algorithm (SFMO), based on the ecological relationship between all predators involved in the feast. In this initial work, the algorithm is based on the behavior of dolphins and two types of sea birds, blue-footed boobies and brown pelicans, which prey on a school of sardines. We demonstrate the usefulness of the algorithm for solving several standard benchmark functions and compare the results with those obtained by using another metaheuristic algorithm, namely the Genetic Algorithm (GA), Bat-inspired Algorithm (BA) and Cuckoo Search (CS). The results of the tests show that the SFMO is better in terms of number of evaluations compared with the other algorithms. Further refinement of the model is needed to fully develop the algorithm.
Keywords:	Sardine Feast, Metaheuristics, Nature-inspired, Optimization
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	MUSIC RECOMMENDATION SYSTEM USED EMOTIONS TO TRACK AND CHANGE NEGATIVE USERS MOOD
Author:	MARWA HUSSIEN MOHAMED, MOHAMED HELMY KHAFAGY, MOHAMED HASAN IBRAHIM, KHALED ELMENSHAWY, HAITHAM RIZK FADLALLAH
Abstract:	Recently, the Recommender system is the most important research area with the advent of e-commerce and e-business on the web. Emotion-based music recovery will have extraordinary potential in catering nowadays, digital music archives quickly extending in the developing smartphones and ubiquitous environments. Many types of research are conducted to improve the music recommendation to users based on their emotions. Human emotions have much difficulty due to the subjective perception of emotions and accuracy challenges. In this paper, we need to solve the problem of recommending songs to the user based on his selection if it was bad, sad, or angry mood by using our system we will recommend to the user songs from pleasant mood to try changing him to the good mood and track if user listen to this song or scaped it. Our new algorithm, "Hybrid emotion-based music recommendation system," will recommend music to the next level, generating playlist which suits and matches your mood of listening to music. The user can try three choices to get the emotion by using face recognition, choosing three colors, and using the arousal map to select the emotion will appear to users then recommended songs according to his status we merge the output of the system to detect the right mood. Our new system has good novelty and diversity of songs recommended to users and changes the user's mood to the pleasure. At our experimental results We are using precision, recall and f-measure accuracy equations to calculate the effective of our system. To gain high results we apply different experiments detect users’ emotions like using face only, colors, arousal map then let users select to types of emotion like face and colors or colors and arousal and finally apply hybrid emotions system. Every time we measure the accuracy of the results. Based on the experiments results using our new hybrid emotions model is best accuracy in surprised, anger, natural and relaxed. While user’s emotion sadness using face. arousal map has high accuracy with happy emotions.
Keywords:	Recommender System; Emotions; Face Recognition; Content-Based Filtering; Collaborative Filtering.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	ANALYSIS OF FACTORS AFFECTING THE INTENTION TO USE AUGMENTED REALITY TECHNOLOGY IN INDONESIA’S ONLINE RETAIL CUSTOMER
Author:	AGNY MONICA CANDRAPUTRI, VIANY UTAMI TJHIN
Abstract:	The goals of the thesis were to determine the main factors that affect the intention to use Augmented Reality technology in Indonesia’s Online Retail Customers and to find out which factors have the most influence on the intention to use this technology. Data were compiled by distributing questionnaires to 400 respondents who had used Augmented Reality technology at Online Retail Indonesia. Analysis data using SmartPLS 3.0 Software to test the validity and reliability of the System Quality, Information Quality, Service Quality, Usefulness, and Satisfaction factors. Results of Analysis were System Quality, Information Quality, Service Quality, Satisfaction and Usefulness have a positive effect on Intention to Use Augmented Reality technology. It can be concluded that all the factors tested in this study were significant or had an effect on Intention to Use Augmented Reality with the most influencing factor was Information Quality and the most influencing for mediation factor was Usefulness.
Keywords:	Information System, Augmented Reality, Human Computer Interaction, Intention To Use, Usefulness
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	EEG SIGNAL DE-NOISING BASED ON THE FEJER- KOROVKIN WAVELET FILTER
Author:	B.V.V.S.R.K.K. Pavan, Dr.P. Esther Rani
Abstract:	The procedure used to measure electrical activity in the brain is an electroencephalogram (EEG). Via electrical impulses, brain cells interact with each other. The EEG can be used to help identify possible issues associated with this operation. Processing of the EEG signals in a noisy environment is a major problem in biomedical signal processing. Especially, Electroencephalography (EEG) acquisition and processing is crucial and difficult concept to step forward. Many neurogenic sounds and non-neurogenic interferences follow the actual EEG signal. The calculated brain response is an event-related potential (ERP) that is the direct product of a sensory, cognitive, or engine event which is low. The challenging task is that rebuilding of ERP signal from contaminated EEG signal. In this article, we propose an efficient approach that combines the decomposition of the empirical mode of the ensemble (EEMD) and fejer- korovkin filtering. The proposed algorithm began with the decomposition of noisy signals Using EEMD Using the comparison between the decomposed IMFs and the original signal several intrinsic mode function (IMF) components are obtained. Then by using fejer-korovkin filtering IMF components are de- noised. From these de-noised IMFs required ERP signal is reassembled. The suggested algorithm is tested using real EEG signals and compared with existing methods of Hybrid Mother Wavelet Using DWT and EMD-IIT with the Performance parameters of Standard deviation, Peak signal to noise ratio (PSNR), Pearson correlation coefficient (PCC) and Root mean square error (RMSE).
Keywords:	Electroencephalography, Event-Related Potential, Ensemble Empirical Mode Decomposition, Intrinsic Mode Functions, Fejer-Korovkin.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	A NOVEL DETECTION OF ISLANDING IN DISTRIBUTION NETWORKS WITH MULTIPLE DISTRIBUTED GENERATIONS
Author:	B. BALA SAI BABU, M. HEMANTH SAI, B. BADDU NAIK
Abstract:	As the demand for electrical energy increases, traditional power generation will no longer be a sustainable source. Distributed generation is emerging technology in which the generation sources are placed near to load centers. Performance and security of the new distributed generation systems needs special attention therefore, a specialized analysis is required to detect different conditions. For safe operation of distributed generation, one of the main requirements is the detection of islands. In this paper for detection of islands a novel method is proposed for highly penetrated wind power distributed power system. Wavelet transform is selected as an appropriate tool to discriminate islanding condition from other disturbances. Energy and standard deviation has been calculated for islanding and other disturbance conditions like normal operation, other DG’s tripping and instant load change etc. The proposed approach is capable of differentiate islanding condition from non-islanding conditions and is very effective and vigorous for different operating conditions.
Keywords:	Distributed generation, multiple DGs, wavelet transform, negative sequence components, islanding
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	AN EFFICIENT APPROACH FOR DETECTING TINY OBJECTS IN MASSIVE BACKGROUND BASED ON SPLIT-ATTENTION NETWORK
Author:	HOANH NGUYEN
Abstract:	With the fast development of deep learning, object detection based on vision has achieved great progress in recent years. Though considerable progress has been made, there still exist challenges for objects with tiny size. One of the main reason is that feature representatives of tiny objects become sparse and weak due to their tiny size in an enormous background. This makes tiny objects difficult to be detected with state-of-the-art object detectors. This paper proposes an efficient method for detecting tiny objects in massive background. First, ResNest architecture is adopted as the backbone to extract features from input images. ResNest captures cross-channel feature correlations, while preserving independent representation in the meta structure. As a result, ResNest architecture achieves better speed-accuracy trade-offs than state-of-the-art deep CNN-based models without incurring excessive computational costs. Next, feature maps generated by the backbone are used to build feature pyramid following FPN network. Finally, this paper proposes an attention network in the detection part to solve problems of occlusion, noise, and blurring and effectively enhance the representations of tiny objects in complex backgrounds based on multi-dimensional attention network and inception module. Experiment results on the AI-TOD dataset show that the proposed method is very efficient in terms of the detection ability of very tiny and tiny objects.
Keywords:	Tiny Objects Detection, Convolutional Neural Network, Deep Learning, Object Detection, Split-Attention Network
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	DIVERSITY COMPARISON OF VARIOUS COMBINING SCHEMES IN IDMA FOR RAYLEIGH AND RICIAN FADING CHANNEL
Author:	DOLLY SHARMA, Dr. RABINDRA SINGH, Dr. HIMANSHU KATIYAR
Abstract:	Recent advancements and demand of wireless communication technique have gained huge attraction in cellular communication standards. Several standards have been developed such as 1G, 2G, 3G, 4G and 5G communication. However, these techniques suffer from Multiple Access Interference (MAI) and Inter Symbol Interference (ISI). Several multiple access schemes have been introduced such as FDMA, TDMA and CDMA but achieving higher efficiency still remains a challenging task. Currently, Interleave-Division Multiple Access (IDMA) is considered as a promising technique to combat with these issues. However, fading is a challenging scenario in IDMA systems. To overcome this issue, several diversity schemes have been developed. These diversities require a combiner module to mitigate the fading effect. In this work, we focus on combining schemes such as selective combining, equal gain combining and maximal ratio combining for IDMA systems. With the help of these combining schemes we measured the performance of IDMA for Rayleigh and Rician fading. The experimental study shows that maximal ratio combining scheme achieves better performance in both fading schemes.
Keywords:	IDMA, Combining Schemes, Diversity Comparison, Rayleigh Fading, Rician Fading
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text

Title:	A MODIFIED CASCADED CELLS MULTILEVEL INVERTER WITH BATTERIES STATE OF CHARGE ADJUSTMENT
Author:	RAFIK Mohamed, KHALILI Tajeddine, BOUAMRANE Omar, RAIHANI Abdelhadi
Abstract:	Power quality enhancement is a critical factor during the design of new multilevel inverters. These topologies are generally modified in order to improve the operating of the converter. In this paper a detailed analysis of a new modified cascaded cells multilevel inverter architecture with power bank adjustment is presented. This new topology provides an enhanced power output through the combination of state of charge balancing in the power bank and the adaptation of firing angles. The present work uses an adapted genetic algorithm for determining the optimal firing angles. The equilibration method assesses the state of charge of the batteries in the power bank in order to generate a balanced discharge. This approach showed promising results related to the voltage output quality, functioning reliability and batteries life span. Each battery string in the power bank is connected to a level generation switch. The monitoring and control algorithm performs a variation of combinations of the power bank’s DC storage units in order to control their discharge, while the control algorithm operates independently of the adjustment algorithm. The proposed multilevel inverter topology is first presented, then the discharge method of the power bank is analyzed under different circumstances. All tests on the batteries state of charge and voltage output generation are achieved in a simulation environment.
Keywords:	Multilevel Inverter (MLI), Cascaded Cells, Power Bank, Power Management, New MLI Architecture, SoC Balancing, Discharge Control.
Source:	Journal of Theoretical and Applied Information Technology 15^th September 2021 -- Vol. 99. No. 17 -- 2021
Full Text