ISSN: 2410-0218
Submit Article
Submission Instructions
Open Access Information
Ethics and Malpractice Statement
Most Recent Issue
2025
Issue 3Issue 2Issue 1
2024
Issue 4Issue 3Issue 2Issue 1
2023
Issue 4Issue 3Issue 2Issue 1
2022
Issue 4Issue 32Issue 31Issue 30
2021
Issue 29Issue 28Issue 27Issue 26
2020
Issue 25Issue 24Issue 23Issue 22
2019
Issue 21Issue 20Issue 19Issue 18
2018
Issue 17Issue 16Issue 15Issue 14
2017
Issue 13Issue 12Issue 11Issue 10
2016
Issue 7Issue 6
2015
Issue 5Issue 4Issue 3Issue 2
2014
Issue 1

Visit the new journal website to submit and consult our contents: https://publications.eai.eu/index.php/inis/index

Editor(s)-in-Chief: Trung Q. Duong, Le Nguyen Bao and Nguyen-Son Vo

EAI Endorsed Transactions on Industrial Networks and Intelligent Systems is open access, a peer-reviewed scholarly journal focused on ubiquitous computing, cloud computing, and cyber-physical system, all kinds of networks in large-scale factories, including a lot of traditional and new industries. The journal publishes research articles, review articles, commentaries, editorials, technical articles, and short communications with a quarterly frequency (four issues per year). This journal is  co-organized, and managed by Duy Tan University, Vietnam.

INDEXING: Scopus (CiteScore: 3.1), Compendex, DOAJ, ProQuest, EBSCO, DBLP

  • Applications of wireless sensor networks, body area networks in large-scale industrial applications, such as fault theories of wireless networks, including routing, network control and management, reliable transmission and architectures, etc.
  • Applications of social networking, big data, ubiquitous computing, mobile computing, and cloud computing in various industries and services (e.g., intelligent systems enhanced by social networking, cloud-based industrial networks, cloud-assisted intelligent systems, etc.)
  • Analysis of industrial control and communication networks, including network lifetime, security, network scalability, reliability, stability, etc.
  • Design and choice of industrial, intelligent, application-specific network protocols and algorithms (e.g., EtherNet/IP, Ethernet Powerlink, EtherCAT, Modbus-TCP, Profinet, SERCOS III, etc.) at any communication layer
  • Opportunistic networks in the industry, such as underwater sensor networks in sewage treatment systems, including establishing a temporary data transmission structure using available devices (e.g., underwater robot, surface data station, surface sink and under water sink), optimizing horizontal multi-hop data links (e.g., 3D data transmission), etc.
  • Applications of intelligent systems in various industries, including collaborative systems, quality control, optimization, decision support, planning, high-level control concepts (e.g., multi-agent and holonic systems, service-oriented architectures), low-level control concepts (e.g., IEC 61131-3 and IEC 61499-based control), advanced system engineering concepts (e.g., model-driven development, component-based design), supply chains, value chains, virtual organizations, and virtual societies, emergency preparedness, crisis management, business channels, electronic marketplaces, enterprise resources planning, etc.
  • Design and analysis of real-time embedded industrial systems, including real-time computing, real-time operating systems, real-time communications, networked embedded systems technology, etc.
  • Novel control techniques, with respect to process control, equipment control, supervisory control, adaptive control, motion control, etc.
  • Automated manufacturing systems, regarding formal modeling and analysis of manufacturing systems, scheduling of manufacturing systems, queuing systems and petri nets in manufacturing systems, etc.
  • Computational intelligence in automation, including neural, fuzzy, evolutionary approaches in automation, ant colonies optimization and swarm intelligence in automation, machine learning, expert systems, etc.
  • Hardware and software design and development for intelligent systems, such as intelligent and humanized production monitoring and control, etc.
  • Big data analysis and processing in various industries and services, including constructing data analysis models, providing data analysis and processing tools and designing various optimization algorithms based on data analysis.
  • Crowd-sourced behavior analysis in various industry and services, such as measuring and calculating the diffusion direction and speed of gas in the petrochemical industry based on crowd-sourced data from a large number of and various types of sensors, as well as product and service evaluation.
  • Simulation and testbed of current industrial networks and intelligent systems, including network performance analysis, automated manufacturing, intelligent monitoring, disaster prevention, etc.
  • Vision of future smart factories, service, marketing, and their integration, incorporating current existing technologies.
  • Multimedia applications, content management, process management and knowledge management for various industries, services, and engineering education: including multimedia processing, multimedia retrieval, multimedia indexing, image sensing, image processing, image coding, image recognition, etc.
  • Pattern recognition methods for various industries and services: including statistical theory, clustering, similarity measures, unsupervised learning, supervised learning, etc.
  • Survey, review and essay of current industrial networks researches and intelligent systems development.
  • Ala Al-Fuqaha (Western Michigan University, USA)
  • Al-Sakib Khan Pathan (Southeast University, Bangladesh)
  • Ammar Rayes (Cisco Systems, USA)
  • Antonino Masaracchia (IIT-CNR, Italy)
  • Athanasios Maglaras (Dr, Prof . ofT.E.I. of Larissa)
  • Berk Canberk (Northeastern University, USA)
  • Ca V. Phan (Ho Chi Minh City University of Technology and Education, Vietnam)
  • Chau Yuen (Singapore University of Technology and Design, Singapore)
  • Chengfei Liu (Swinburne University of Technology, Australia)
  • Chinmoy Kundu (University of Texas at Dallas, USA)
  • Christer Carlsson (Åbo Akademi University, Finland)
  • Chunsheng Zhu (University of British Columbia)
  • Constandinos Mavromoustakis (University of Nicosia, Cyprus)
  • Der-Jiunn Deng (National Changhua University of Education, Taiwan)
  • Dickson Chiu (The University of Hong Kong)
  • Eleanna Kafeza (Athens University of Economics and Business, Greece)
  • Fu-ren Lin (National Tsing Hua University, Taiwan)
  • Gerhard Hancke (University of London, UK)
  • Guangjie Han (Hohai University, China)
  • Guojun Wang (Central South University, China)
  • Hacene Fouchal (University of Reims Champagne-Ardenne, France)
  • Haklae Kim (Chung-Ang University, South Korea)
  • Halil Yetgin (Bitlis Eren University, Turkey)
  • Hideyasu Sasaki (Ritsumeikan University, Kyoto, Japan)
  • Ho-fung Leung (Chinese University of Hong Kong, Hong Kong)
  • Honggang Wang (University of Massachusetts Dartmouth, USA)
  • Hua Hu (Hangzhou Dianzi University, China)
  • Ibrahim Kushchu (Mobile Government Consortium International, UK)
  • Irene Kafeza (Irene Law Office, Greece)
  • Isabelle Comyn-Wattiau (ESSEC Business School Paris, France)
  • Jaime Lloret- Mauri (Universitat Politècnica de València, Spain)
  • Javier M. Aguiar (Universidad de Valladolid, Valladolid, Spain)
  • Jesus Alonso-Zarate (Telecommunications Technology Center of Catalonia, Spain)
  • Jian Yang (Macquarie University, Australia)
  • Jiankun Hu (University of New South Wales, Australia)
  • Jianmin Jiang (Shenzhen University)
  • Jianwei Niu (Beihang University, China)
  • Jinlei Jiang (Tsinghua University, China)
  • Jinsong Wu (Bell Laboratory, China)
  • Joel Rodrigues (Inst. Telecomunicações, Univ. of Beira Interior, Portugal)
  • Juan Trujillo (University of Alicante, Spain)
  • Jucheng Yang (Tianjing University of Technology, China)
  • Junqing Zhang (Queen's University Belfast)
  • KUN WANG (Nanjing University of Posts and Telecommunications)
  • Kuo-Ming Chao (Leader – Distributed Systems and Modelling Research Group, UK)
  • Leandros A. Maglaras (De Montfort University, UK)
  • Lei Wang (Dalian University of Technology, China)
  • Liang Zhou (Nanjing University of Posts and Telecommunications, China)
  • Long D. Nguyen (Dong Nai University, Vietnam)
  • Maggie M. Wang (The University of Hong Kong, Hong Kong)
  • Nghia Duong-Trung (German Research Center for Artificial Intelligence, Germany)
  • Ngo Hoang Tu (Seoul National University of Science and Technology, South Korea)
  • Nguyen Van Nam (Viettel, Vietnam)
  • Nicholas C Romano (Oklahoma State University, USA)
  • Noel Crespi (Institut Mines-Telecom, Telecom SudParis, France)
  • Panlong Yang (PLA University of Science and Technology, China)
  • Pasi Tyrväinen (University of Jyväskylä, Finland)
  • Patrick C.K. Hung (University of Ontario Institute of Technology, Canada)
  • Periklis Chatzimisios (Alexander TEI of Thessaloniki, Greece)
  • Pierluigi Siano (Università degli Studi di Salerno, Italy)
  • Pirkko Walden (Abo Akademi University, Finland)
  • Phuong Bui (Duy Tan University, Vietnam)
  • Raymond Y.K Lau (City University of Hong Kong, Hong Kong)
  • Richard Yu (Carleton University, Canada)
  • Rong Yu (Guangdong University of Technology, China)
  • Rose Hu (Utah State University, USA)
  • Sammy Chan (City University of HongKong, HK)
  • Shing-Chi Cheung (Hong Kong University of Science and Technology, Hong Kong)
  • Stephen J. H. Yang (National Central University, Taiwan)
  • Syed Hassan Ahmed (University of Central Florida, USA)
  • Thanh-Phuong Nguyen (University of Toulon, France)
  • Tran Trung Duy (PTIT, VietNam)
  • Trang Hoang (Ho Chi Minh City University of Technology - Vietnam National University Ho Chi Minh City, Vietnam)
  • Tuan-Minh Pham (Phenikaa University, Vietnam)
  • Umar Zakir Abdul Hamid (Sensible 4 Oy, Helsinki)
  • Victor Leung (The University of British Columbia)
  • Vo Nguyen Son Dr. (Duy Tan University, Vietnam)
  • Wai-Wa Fung (Information Security and Forensics Society, Hong Kong)
  • Walid Gaaloul (Institut National des Télécommunications, France)
  • Weiwei Jiang, (Beijing University of Posts and Telecommunications (BUPT), China)
  • Wendy W. Y. Hui (University of Nottingham at Ningbo, China)
  • William Cheung (Hong Kong Baptist University, Hong Kong)
  • Xianfu Chen (VTT Technical Research Centre of Finland, Finland)
  • Xiang Gui (Massey University, New Zealand)
  • Xiaoling Wu (Chinese Academy of Sciences, China)
  • Xu Wang (Heriot Watt University, UK)
  • Yan Bai (University of Washington Tacoma, USA)
  • Yan Zhang (Simula Research Laboratory and University of Oslo, Norway)
  • Yi Zhuang (Zhejian Gongshang University, China)
  • Yong Li (Tsinghua University, China)
  • Yong Tang (South China Normal University, China)
  • Yuanfang Chen (Institute Mines-Telecom, University Pierre and Marie Curie )
  • Yuexing Peng (Beijing University of Posts and Telecommunications, China)
  • Yuqing Sun (Shangdong University, China)
  • Zakaria Maamar (Zayed University, UAE)
  • Zhangbing Zhou (China University of Geosciences, China)
  • Zhichao Sheng (Shanghai University, China)
  • ZhiMing Cai (Macau University of Science and Technology, Macau)
  • Mithun Mukherjee (Nanjing University of Information Science and Technology, China)
  •  

  • FedNDA: Enhancing Federated Learning with Noisy Client Detection and Robust Aggregation

    Appears in:
    inis 25(3):

    Authors:
    Trung-Kien Tran, Tuan Dung Kieu, Hai Vu, Thi-Lan Le, Huu-Thanh Nguyen, Charles Fonbonne, Thanh-Hai Tran

    Published:
    3rd Jul 2025

    Abstract:
    Federated Learning is a novel decentralized methodology that enables multiple clients to collaboratively train a global model while preserving the privacy of their local data. Although federated learning enhances data privacy, it faces challenges related to data quality and client behavior. A fundamental issue is the presence of noisy labels in certain clients, which damages the global model's performance. To address this problem, this paper introduces a Federated learning framework with Noisy client Detection and robust Aggregation, FedNDA. In the first stage, FedNDA detects noisy clients by analyzing the distribution of their local losses. A noisy client exhibits a loss distribution distinct from that of clean clients. To handle class imbalance issue in local data, we utilize per-class losses instead of the total loss. We then assign each client a noisiness score, calculated as the Earth Mover’s Distance between the per-class loss distribution of the client and the average distribution of all clean clients. This noisiness metric is more sensitive for detecting noisy clients compared to conventional metrics such as Euclidean distance or L1 norm. The noisiness score is subsequently transfered to and used in the server-side aggregation function to prioritize clean clients while reducing the influence of noisy clients. Experimental results demonstrate that FedNDA outperforms FedAvg and FedNoRo by 4.68% and 3.6% on the CIFAR-10 dataset, and by 10.65% and 0.48% on the ICH dataset, respectively, in a high noisy setting.

  • Return Loss Optimization in Rectangular Microstrip Patch Antennas Using Response Surface Methodology (RSM) for 5G Applications

    Appears in:
    inis 25(3):

    Authors:
    Thi Bich Ngoc Tran, Van Su Dang

    Published:
    12nd Jun 2025

    Abstract:
    In recent decades, wireless communication has advanced significantly. People increasingly rely on the Internet of Things, cloud computing, and big data analytics. These services require higher data rates, faster transmission and reception times, greater coverage, and increased throughput. 5G technology supports all of these features. Antennas, essential components of modern wireless devices, must be designed to meet the growing demand for fast and intelligent products. This study aims to optimize the dimensions and characteristics of a rectangular patch antenna. To examine the impact of independent variables (such as patch length, patch width, inset slot length, and inset slot width) on the response variables (return loss and resonant frequency), Response Surface Methodology (RSM) combined with Central Composite Design (CCD) was applied. The findings of the RSM analysis indicated that the experimental data were best represented by a quadratic polynomial model, with regression coefficients exceeding 0.970 for all responses. The optimized parameters identified are as follows: a patch length of 4.7 mm, a patch width of 4.7 mm, an inset slot length of 0.8 mm, and an inset slot width of 1.0 mm. The antenna designed using these optimized parameters achieved a target return loss of -45.865 dB at a frequency of 28.122 GHz. Finally, the results were validated using CST Studio Suite, which demonstrated good agreement with the experimental data.

  • QoE-Energy Consumption Optimization for End-User Devices in Adaptive Bitrate Video Streaming Using the Lagrange Multiplier Method

    Appears in:
    inis 25(3):

    Authors:
    Thao Nguyen Thi Huong, Tien Vu Huu

    Published:
    14th Apr 2025

    Abstract:
    The reduction of greenhouse gas emissions in the Internet and ICT sectors has become a critical challenge. According to recent research, the key contributors to greenhouse gas emissions in Internet include high energy consumption factors such as data centers, transmission network devices, and end-user devices. Among Internet services, video streaming is one of the services having the highest traffic volume and number of users. Consequently, developing energy-efficient solutions for video streaming networks, particularly for end-user devices, is an urgent research priority. Reducing energy consumption in end-user devices in a video streaming system often requires compromises in parameters that impact the quality of user experience (QoE). Therefore, achieving an optimal trade-off between minimizing energy consumption and maintaining an acceptable QoE is a key objective. In this study, a cost function that integrates QoE and energy consumption is developed using the Lagrange multiplier method. Based on this function, an adaptive bitrate algorithm is proposed to select optimal video segments for video players, ensuring maximum QoE while minimizing energy consumption. The performance of the proposed method is evaluated using various types of video samples under varying network bandwidth conditions. Experimental results show that the proposed method reduces energy consumption of end-user devices by up to 6.7% and enhances QoE by 20% compared to previous methods.

  • Enhancing AI-Inspired Analog Circuit Design: Optimizing Component Sizes with the Firefly Algorithm and Binary Firefly Algorithm

    Appears in:
    inis 25(2):

    Authors:
    Trang Hoang

    Published:
    11th Apr 2025

    Abstract:
    This paper explores the use of the Firefly Algorithm (FA) and its binary variant (BFA) in optimizing analog circuit component sizing, specifically as a case study for a two-stage operational amplifier (op-amp) designed with a 65nm CMOS process. Recognizing the limitations of traditional optimization approaches in handling complex analog design requirements, this study implements both FA and BFA to enhance convergence speed and accuracy within multi-dimensional search spaces. The Python-Spectre framework in this paper facilitates automatic, iterative simulation and data collection, driving the optimization process. Through extensive benchmarking, the BFA outperformed traditional FA, balancing exploration and exploitation while achieving superior design outcomes across key parameters such as voltage gain, phase margin, and unity-gain bandwidth. Comparative analysis with existing optimization methods, including Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), underscores the efficiency and accuracy of BFA in optimizing circuit metrics, particularly in power-constrained environments. This study demonstrates the potential of swarm intelligence in advancing automatic analog design and establishes a foundation for future enhancements in analog circuit automation.

  • Security-Reliability Analysis of NOMA-Assisted Hybrid Satellite-Terrestrial Relay Multi-Cast Transmission Networks Using Fountain Codes and Partial Relay Selection with Presence of Multiple Eavesdroppers

    Appears in:
    inis 25(3):

    Authors:
    Ngoc Son Pham, Trung Duy Tran, Lam Thanh Tu, Viet Tuan Pham, Toan Van Nguyen

    Published:
    11th Apr 2025

    Abstract:
    This article proposes a hybrid satellite-terrestrial relaying network (HSTRN) that integrates physical-layer security (PLS), Fountain codes (FCs), non-orthogonal multiple access (NOMA), and partial relay selection (PRS) to enhance system performance in terms of reliability, data rate, and security. In the proposed system, a satellite uses NOMA to simultaneously transmit Fountain packets to two clusters of terrestrial users. Data transmission is assisted by one of the terrestrial relay stations, selected by the PRS algorithm. We derive exact expressions for outage probability (OP) and system outage probability (SOP) at the legitimate users, as well as intercept probability (IP) and system intercept probability (SIP) at eavesdroppers. Monte Carlo simulations are realized to validate the accuracy of the analytical results, illustrate performance trends, and analyze the impact of key parameters on the considered performance.

  • A novel approach for graph-based real-time anomaly detection from dynamic network data listened by Wireshark

    Appears in:
    inis 25(2):

    Authors:
    Resul Das, Mehmet Ozdem, Muhammed Onur Kaya

    Published:
    11th Apr 2025

    Abstract:
    This paper presents a novel approach for real-time anomaly detection and visualization of dynamic network data using Wireshark, globally's most widely utilized network analysis tool. As the complexity and volume of network data continue to grow, effective anomaly detection has become essential for maintaining network performance and enhancing security. Our method leverages Wireshark’s robust data collection and analysis capabilities to identify anomalies swiftly and accurately. In addition to detection, we introduce innovative visualization techniques that facilitate the intuitive representation of detected anomalies, allowing network administrators to comprehend network conditions and make informed decisions quickly. The results of our study demonstrate significant improvements in both the efficacy of anomaly detection and the practical applicability of visualization tools in real-time scenarios. This research contributes valuable insights into network security and management, highlighting the importance of integrating advanced analytical methods with effective visualization strategies to enhance the overall management of dynamic networks.

  • Integrated Cloud-Twin Synchronization for Supply Chain 5.0

    Appears in:
    inis 25(2):

    Authors:
    Divya Sasi Latha, Tartat Mokkhamakkul

    Published:
    12nd Mar 2025

    Abstract:
    The digital twin is thus emerging means of improving real-world performance from virtual spaces, especially relatedto Supply Chain 5.0 in Industry 5.0. This framework employs the integration of cloud computing and digital twin technologies to secure data storage, trusted tracking, and high reliability, is architectural for the integration of supply-chain sustainable enterprises. In this work, we introduce a high level architecture of cloud-based digital twin model for supply chain 5.0 , which was created to align the system of supply chain through real-time observation as well as real-timesupply chain 5.0 decision-making and control. This study introduces a cloud-based twin optimization model for Supply Chain 5.0, validated through genetic algorithm (GA) simulations. The model determines optimal weights to balance objectives, achieving an optimal objective function value that reflects trade-offs among operational efficiency, cost, and sustainability. A convergence plot illustrates the model’s iterative solution improvements, demonstrating its dynamic adaptability. Lastly, the proposed model defines and test a supply chain performance analysis through dynamic simulations.

  • Single-level Discrete Two Dimensional Wavelet Transform Based Multiscale Deep Learning Framework for Two-Wheeler Helmet Detection

    Appears in:
    inis 25(2):

    Authors:
    Subhra Rani Patra, Amrutha Annadurai, Trilok Nath Pandey, Manas Ranjan Prusty

    Published:
    11th Mar 2025

    Abstract:
    INTRODUCTION: A robust method is proposed in this paper to detect helmet usage in two-wheeler riders to enhance road safety. OBJECTIVES: This involves a custom made dataset that contains 1000 images captured under diverse real-world scenarios, including variations in helmet size, colour, and lighting conditions. This dataset has two classes namely with helmet and without helmet. METHODS: The proposed helmet classification approach utilizes the Multi-Scale Deep Convolutional Neural Network (CNN) framework cascaded with Long Short-Term Memory (LSTM) network. Initially the Multi-Scale Deep CNN extracts modes by applying Single-level Discrete 2D Wavelet Transform (dwt2) to decompose the original images. In particular, four different modes are used for segmenting a single image namely approximation, horizontal detail, vertical detail and diagonal detail. After feeding the segmented images into a Multi-Scale Deep CNN model, it is cascaded with an LSTM network. RESULTS: The proposed model achieved accuracies of 99.20% and 95.99% using both 5-Fold Cross-Validation (CV) and Hold-out CV methods, respectively. CONCLUSION: This result was better than the CNN-LSTM, dwt2-LSTM and a tailor made CNN model.

  • Coverage Probability of EH-enabled LoRa networks - A Deep Learning Approach

    Appears in:
    inis 24(2):

    Authors:
    Tran Cong-Hung, Thi-Tuyet-Hai Nguyen, Tran Trung Duy, Nguyen Hong-Son, Lam-Thanh Tu, Tan Hanh

    Published:
    5th Dec 2024

    Abstract:
    The performance of energy harvesting (EH)-enabled long-range (LoRa) networks is analyzed in this work. Specifically, we employ deep learning (DL) to estimate the coverage probability (Pcov) of the considered networks. Our study incorporates a general fading distribution, specifically the Nakagami-m distribution, and utilizes tools from stochastic geometry (SG) to model the spatial distributions of all nodes and end-devices (EDs) with EH capability. The DL approach is employed to overcome the limitations of model-based methods that can only evaluate the Pcov under simplified network conditions. Therefore, we propose a deep neural network (DNN) that estimates the Pcov with high accuracy compared to the ground truth values. Additionally, we demonstrate that DL significantly outperforms the Monte Carlo simulation approach in terms of resource consumption, including time and memory.

  • Drug classification system based on drug composition and usage instructions

    Appears in:
    inis 24(1):

    Authors:
    Quang-Dung Le, Hoang-Dieu Vu, Vu-Hien Pham

    Published:
    4th Dec 2024

    Abstract:
    This study presents a natural language processing (NLP) approach to classify drugs based on compositional and usage descriptions. NLP techniques including text preprocessing, word embedding, and deep learning models were applied to a Vietnamese drug dataset. Traditional machine learning models like Support Vector Machines (SVM) and deep models including Bidirectional Long Short-Term Memory (BiLSTM) and PhoBERT were evaluated. Besides, since there is a limitation in the information of our own collected data, some data augmentation techniques were applied to increase the variation of the dataset. Results show PhoBERT achieving 95% accuracy, highlighting the benefits of transferring knowledge from large language models. Errors primarily occurred between similar drug categories, suggesting taxonomy refinement could improve performance. In summary, an automated drug classification framework was developed leveraging state-of- the-art NLP, validating the feasibility of analyzing drug data at scale and aiding therapeutic understanding. This supports NLP’s potential in pharmacovigilance applications.

  • Internet Traffic Prediction Using Recurrent Neural Networks

    Appears in:
    inis 22(4): e1

    Authors:
    Mircea Eugen Dodan, Quoc-Tuan Vien, Tuan Thanh Nguyen

    Downloads:
    1144

    Abstract:
    Network traffic prediction (NTP) represents an essential component in planning large-scale networks which are in general unpredictable and must adapt to unforeseen circumstances. In small to medium-size networks, the administrator can anticipate the fluctuations in traffic without the need of using forecasting tools, but in the scenario of large-scale networks where hundreds of new users can be added in a matter of weeks, more efficient forecasting tools are required to avoid congestion and over provisioning. Network and hardware resources are however limited; and hence resource allocation is critical for the NTP with scalable solutions. To this end, in this paper, we propose an efficient NTP by optimizing recurrent neural networks (RNNs) to analyse the traffic patterns that occur inside flow time series, and predict future samples based on the history of the traffic that was used for training. The predicted traffic with the proposed RNNs is compared with the real values that are stored in the database in terms of mean squared error, mean absolute error and categorical cross entropy. Furthermore, the real traffic samples for NTP training are compared with those from other techniques such as auto-regressive moving average (ARIMA) and AdaBoost regressor to validate the effectiveness of the proposed method. It is shown that the proposed RNN achieves a better performance than both the ARIMA and AdaBoost regressor when more samples are employed.

  • An Intelligent Fashion Object Classification Using CNN

    Appears in:
    inis 23(4): e2

    Authors:
    Jay Sanghvi, Debabrata Swain, Yugandhar Manchala, Kaxit Pandya

    Downloads:
    892

    Abstract:
    Every year the count of visually impaired people is increasing drastically around the world. At present time, approximately 2.2 billion people are suffering from visual impairment. One of the major areas where our model will affect public life is the area of house assistance for specially-abled persons. Because of visual improvement, these people face lots of issues. Hence for this group of people, there is a high need for an assistance system in terms of object recognition. For specially-abled people sometimes it becomes really difficult to identify clothing-related items from one another because of high similarity. For better object classification we use a model which includes computer vision and CNN. Computer vision is the area of AI that helps to identify visual objects. Here a CNN-based model is used for better classification of clothing and fashion items. Another model known as Lenet is used which has a stronger architectural structure. Lenet is a multi-layer convolution neural network that is mainly used for image classification tasks. For model building and validation MNIST fashion dataset is used.

  • Human Activity Recognition System For Moderate Performance Microcontroller Using Accelerometer Data And Random Forest Algorithm

    Appears in:
    inis 22(4): e4

    Authors:
    To-Hieu Dao, Duc-Nghia Tran, Hai-Yen Hoang, Van-Nhat Hoang, Duc-Tan Tran

    Downloads:
    824

    Abstract:
    There has been increasing interest in the application of artificial intelligence technologies to improve the quality of support services in healthcare. Some constraints, such as space, infrastructure, and environmental conditions, present challenges with assistive devices for humans. This paper proposed a wearable-based real-time human activity recognition system to monitor daily activities. The classification was done directly on the device, and the results could be checked over the internet. The accelerometer data collection application was developed on the device with a sampling frequency of 20Hz, and the random forest algorithm was embedded in the hardware. To improve the accuracy of the recognition system, a feature vector of 31 dimensions was calculated and used as an input per time window. Besides, the dynamic window method applied by the proposed model allowed us to change the data sampling time (1-3 seconds) and increase the performance of activity classification. The experiment results showed that the proposed system could classify 13 activities with a high accuracy of 99.4%. The rate of correctly classified activities was 96.1%. This work is promising for healthcare because of the convenience and simplicity of wearables.

  • Deep Reinforcement Learning for Intelligent Reflecting Surface-assisted D2D Communications

    Appears in:
    inis 23(1): e1

    Authors:
    Antonino Masaracchia, Khoi Khac Nguyen, Cheng Yin

    Downloads:
    790

    Abstract:
    In this paper, we propose a deep reinforcement learning (DRL) approach for solving the optimisation problem of the network’s sum-rate in device-to-device (D2D) communications supported by an intelligent reflecting surface (IRS). The IRS is deployed to mitigate the interference and enhance the signal between the D2D transmitter and the associated D2D receiver. Our objective is to jointly optimise the transmit power at the D2D transmitter and the phase shift matrix at the IRS to maximise the network sum-rate. We formulate a Markov decision process and then propose the proximal policy optimisation for solving the maximisation game. Simulation results show impressive performance in terms of the achievable rate and processing time.

  • An Accurate Viewport Estimation Method for 360 Video Streaming using Deep Learning

    Appears in:
    inis 22(4): e2

    Authors:
    Thu Huong Truong, Tran Long Dang, Ngoc Son Pham, Thu Ngan Dao, Trung Dung Nguyen, Hung Nguyen

    Downloads:
    754

    Abstract:
    Nowadays, Virtual Reality is becoming more and more popular, and 360 video is a very important part of the system. 360 video transmission over the Internet faces many difficulties due to its large size. Therefore, to reduce the network bandwidth requirement of 360-degree video, Viewport Adaptive Streaming (VAS) was proposed. An important issue in VAS is how to estimate future user viewing direction. In this paper, we propose an algorithm called GLVP (GRU-LSTM-based-Viewport-Prediction) to estimate the typical view for the VAS system. The results show that our method can improve viewport estimation from 9.5% to near 20%compared with other methods.

  • Functional Security and Trust in Ultra-Connected 6G Ecosystem

    Appears in:
    inis 22(4): e5

    Authors:
    Vishal Sharma

    Downloads:
    726

    Abstract:
    Security and trust are the entangled role players in the future generation of wireless networks. Security in 5G networks is currently supported using several functions. Given the advantages of such a system, this article explores the functional security and trust for the 6G ecosystem with ultra-connectivity. Several associated challenges, application-specific domains, and consumer issues related to 6G security are discussed. The article highlights the network security-by-design and trust-by-design principles and performance expectations from the security protocols in supporting handover in an ultra-connected scenario. Finally, potential research directions are presented for a road towards the 6G ecosystem.

  • A Fully Convolutional Network with Waterfall Atrous Spatial Pooling and Localized Active Contour Loss for Fish Segmentation

    Appears in:
    inis 23(1): e4

    Authors:
    Van Truong Pham, Van Yem Vu, Thanh Viet Le, Thi-Thao Tran

    Downloads:
    671

    Abstract:
    Accurate measurements and statistics of fish data are important for sustainable development of aqua-enviroment and marine fisheries. For data measurements and statistics, automatic segmentation of fish is one of key tasks. The fish segmentation however is a challenging task due to arterfacts in underwater images. In this study, we introduce a deep-learning approach, namely FCN-WRN-WASP for automatic fish segmentation from the underwater images. In particular, we introduce a computational-efficient variation called Waterfall Atrous Spatial Pooling (WASP) module into a Fully convolutional network with Wide ResNet baseline. We also proposed a loss function inspired from active contour approach that can exploit the local intensity information from the input image. The approach has been validated on the DeepFish data and the SIUM data set. The results are promissing for fish segmentation, with higher Intersection over Union (IoU) scores compared to state of the arts. The evaluation results showed that the incorporation of the image based active contour loss helps increase the segmentation performance. In addition, the use of the WASP in the architecture is effective especially for forground fish segmentation.

  • Availability of Free-Space Laser Communication Link with the Presence of Clouds in Tropical Regions

    Appears in:
    inis 23(3): e1

    Authors:
    Thang Nguyen, Ngoc Dang, Hien Pham, Hoa Le

    Downloads:
    654

    Abstract:
    Free-space laser communication (lasercom), a great application of using free-space optics (FSO) for satellite communication, has been gaining significant attraction. However, despite of great potential of lasercom, its performance is limited by the adverse effects of atmospheric turbulence and cloud attenuation, which directly affect the quality and availability of lasercom links. The paper, therefore, concentrates on evaluating the cloud attenuation in the FSO downlinks between satellite and ground stations in tropical regions. The meteorological ERA-Interim database provided by the European Center for Medium-Range Weather Forecast (ECMWF) from 2015 to 2020 is used to get the cloud database in several areas in tropical regions. This study proposed a novel probability density function of cloud attenuation, which is validated by using a well-known curve-fitting method. Moreover, we derive a closed-form of satellite-based FSO link availability by applying the site diversity technique to improve the system performance. Numerical results, which demonstrate the urgency of the paper, reveal that the impact of clouds on tropical regions is more severe than in temperate regions.

  • Performance Analysis of NOMA Over n-m Fading Channels with imperfect SIC

    Appears in:
    inis 23(1): e5

    Authors:
    Trần Quý Hữu, Mau Dung Ong

    Downloads:
    650

    Abstract:
    In this paper, a downlink non-orthogonal multiple access (NOMA) network with two users is considered. In particular, the performance of NOMA is evaluated by assuming perfect and imperfect channel state information (CSI). We derive the closed-form expressions for the outage probability over η − µ fading channels in the special case of two users. Moreover, the proposed system model-based NOMA always achieves better performance than that with perfect CSI in the medium SNR region. Monte Carlo simulations are then performed to confirm a good match with the analytical results.

  • Attention ConvMixer Model and Application for Fish Species Classification

    Appears in:
    inis 23(3): e2

    Authors:
    Van-Truong Pham, Van Yem Vu, Hoang-Minh-Quang Le, Thanh Viet Le, Thi-Thao Tran

    Downloads:
    635

    Abstract:
    Exploring the ocean has always been one of the foremost challenges for humankind, and fish classification is one of the crucial tasks in this endeavor. Manual fish classification methods, although accurate, consume significant time, money, and effort, while computer-based methods such as image processing and traditional machine learning often fall short of achieving high accuracy. Recently, deep convolutional neural networks have demonstrated their capability to ensure both time efficiency and accuracy in this task. However, deep convolutional networks typically have a large number of parameters, requiring substantial training time, and the convolutional operations lack attentional mechanisms. Therefore, in this paper, we propose the AttentionConvMixer neural network with Priority Channel Attention (PCA) and Priority Spatial Attention (PSA). The proposed approach exhibits good performance across all three fish classification datasets without introducing any additional parameters, thus demonstrating the effectiveness of our proposed method.
Publisher
EAI
ISSN
2410-0218
Number of Volumes
12
Last Published
2025-04-11
EAI Logo

About EAI

Community

Publish with EAI