EAI Endorsed Transactions on Industrial Networks and Intelligent Systems

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

Thang Nguyen, Hoa Le, Hien Pham, Ngoc Dang — Wed, 23 Aug 2023 00:00:00 +0000

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.

Attention ConvMixer Model and Application for Fish Species Classification

Thanh Viet Le , Hoang-Minh-Quang Le , Van Yem Vu, Thi-Thao Tran, Van-Truong Pham — Wed, 06 Sep 2023 00:00:00 +0000

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.

SHELF: Combination of Shape Fitting and Heatmap Regression for Landmark Detection in Human Face

Ngo Thi Ngoc Quyen, Tran Duy Linh, Vu Hong Phuc, Nguyen Van Nam — Tue, 26 Sep 2023 00:00:00 +0000

Today, facial emotion recognition is widely adopted in many intelligent applications including the driver monitoring system, the smart customer care as well as the e-learning system. In fact, the human emotions can be well represented by facial landmarks which are hard to be detected from images, due to the high number of discrete landmarks, the variation of shapes and poses of the human face in real world. Over decades, many methods have been proposed for facial landmark detection including the shape fitting, the coordinate regression such as ASMNet and AnchorFace. However, their performance is still limited for real-time applications in terms of both accuracy and efficiency. In this paper, we propose a novel method called SHELF which is the first to combine the shape fitting and heatmap regression approaches for landmark detection in human face. The heatmap model aims to generate the landmarks that fit to the common shapes. The method has been evaluated on three datasets 300W-Challenging, WFLW, 300VW-E with 31557 images and achieved a normalized mean error (NME) of 6.67% , 7.34%, 12.55% correspondingly, which overcomes most existing methods. For the first two datasets, the method is also comparable to the state of the art AnchorFace with a NME of 6.19%, 4.62%, respectively.

A Multi-objective optimization model for sustainable production planning in textile MSMEs

Pablo Flores-Siguenza, Jose Antonio Marmolejo-Saucedo, Rodrigo Guamán — Wed, 27 Sep 2023 00:00:00 +0000

Textile MSMEs are characterized by their high influence on the economy of the countries, both for their contribution to the gross domestic product as well as for the generation of employment, in recent years the complexity of their operations, instability and lack of balance between economic, environmental and social factors, axes of sustainable development, stand out. It is necessary to implement approaches such as sustainable manufacturing and production planning, which seeks the creation of products with minimal environmental impact, safe for workers, and economically robust. In this context, this study aims to develop a multi-objective optimization model that enhances sustainable production planning in textile MSMEs. The methodology is based on two phases, the first one focused on the acquisition of information and the second one dedicated to the mathematical formulation of the model, where three objective functions focused on economic, environmental and social factors are proposed. The model is validated with real data from a textile MSME in Ecuador and different production alternatives are generated by proposing the implementation and use of photovoltaic energy as well as a greater use of personal protective equipment. One of the relevant conclusions of the study is the contribution to the textile industry with a sustainable decision support tool, where different scenarios for production planning and their respective economic, environmental and social consequences are shown.

Context-Aware Device Classification and Clustering for Smarter and Secure Connectivity in Internet of Things

Priyanka More, Sachin Sakhare — Mon, 02 Oct 2023 00:00:00 +0000

With the increasing prevalence of the Internet of Things (IoT), there is a growing need for effective access control methods to secure IoT systems and data. Traditional access control models often prove inadequate when dealing with the specific challenges presented by IoT, characterized by a variety of heterogeneous devices, ever-changing network structures, and diverse contextual elements. Managing IoT devices effectively is a complex task in maintaining network security.

This study introduces a context-driven approach for IoT Device Classification and Clustering, aiming to address the unique characteristics of IoT systems and the limitations of existing access control methods. The proposed context-based model utilizes contextual information such as device attributes, location, time, and communication patterns to dynamically establish clusters and cluster leaders. By incorporating contextual factors, the model provides a more accurate and adaptable clustering mechanism that aligns with the dynamic nature of IoT systems. Consequently, network administrators can configure dynamic access policies for these clusters.