Taesoo Jun’s research while affiliated with Kumoh National Institute of Technology and other places

... The interconnected networks enable communication and data exchange between nodes, such as shipboard systems, aerial platforms, satellites, underwater sensors, submarines, naval combat systems, autonomous underwater vehicles, and other components. The marine industry is experiencing a growing trend towards decentralization, connectivity, and automation, increasing its vulnerability to cyber threats [2]. The World Economic Forum's Global Risks Report 2020 identifies cyber-attacks on maritime infrastructure as the fifth most significant threat [3]. ...

Reference:

Enhanced Secure Communication in Maritime Tactical Networks Using Quantum Cryptography

... Developing intelligent manufacturing has become a consensus of the international community because of the widespread popularity of fifth-generation mobile communication, artificial intelligence (AI) and big data [7,8]. Intelligent optimisation and decision-making are the main focuses of intelligent manufacturing [9,10]. ...

Reference:

Integrated Optimisation of Shop Scheduling and Machine Layout for Discrete Manufacturing Considering Uncertain Events Based on an Improved Immune Genetic Algorithm

... The rapid progress in machine learning [1] and deep learning has made significant contributions to addressing various challenges faced by humanity. Brain tumors, classified as either malignant or benign growths found in the brain, pose significant challenges in terms of detection and treatment. ...

Reference:

A Blockchain-Based Framework for Distributing and Managing CNN-Derived Brain Tumor Detection Models

... Internet of Things (IoT) sinkhole and selective forwarding attack detection has received little attention in the literature. To protect networks from selective forwarding attacks and identify topological disruptions, Wallgren et al. [10] suggested the Heartbeat self-healing protocol. One intrusion detection system (IDS) that can handle various threats, such as sinkhole and selective forwarding, is SVELTE, which is introduced in [7]. ...

Reference:

An efficient internet of things based intrusion detection and optimization algorithm for smart networks

... Texture information is insufficient to fully depict the pattern of anomalies, particularly logical abnormalities. In [4], authors discuss structural anomaly detection, while [5] explains both structural and logical anomalies using a balanced dataset. To overcome this challenge, this work focuses on enhancing the MES operations through efficient and cost-sensitive prediction of faults using the You Only Look Once version 8 (YOLOv8) paradigm to extract complex texture features of MES objects. ...

Reference:

Deep Learning for Multivariate Anomaly Detection in Advanced Manufacturing Execution Systems

... Based on experimental findings using the real Doppler radar with digital array receiver (RAD-DAR) database, the proposed method can identify UAVs with up to 99.5% accuracy. Additionally, in [189], the authors proposed employing a CNN to detect UAVs using data from radar images. The microwave and radar group developed the real Doppler RAD-DAR radar technology, a range-Doppler system. ...

Reference:

A Comprehensive Survey of Unmanned Aerial Vehicles Detection and Classification Using Machine Learning Approach: Challenges, Solutions, and Future Directions

... The model reveals that pollution sources in western Chaohu Lake significantly impact overall water quality, which helps trace pollution sources. In terms of edge device deployment, the model can be integrated into internet of things (IoT) sensors to reduce data transmission delays and improve response speed (Ajakwe et al., 2023). ...

Reference:

An improved graph neural network integrating indicator attention and spatio-temporal correlation for dissolved oxygen prediction

... The error between the two signals can be quantified using measures such as root mean square error (RMSE), 25 mean absolute error (MAE), 26 or mean absolute percentage error (MAPE). 27 However, as cross correlation relies on the time delay between the reconstructed and original signals to compare their degree of correlation, it is highly sensitive to changes in the flow pattern of the fluid, making it less suitable for the measurement of multiphase fluid in the petroleum and natural gas industry. The second method involves comparing the performance of the reconstructed signal and the original signal in terms of flow velocity measurement or other parameters. ...

Reference:

A model for multiphase flow velocity calculation in pipelines based on ultrasonic sensors

... For each of these activities, the underlying AI models carry out specific operations whose result is a determinant for triggering other activities. Various atmospheric and weather conditions, such as rain, snow, fog, dust, changing sunlight intensity, etc. significantly impact UAV smart mobility and, by extension, the functional capacity of CUAS for monitoring UAV safety [43,44]. ...

Reference:

Facets of security and safety problems and paradigms for smart aerial mobility and intelligent logistics

... We use TensorFlow for the purposes of machine learning and deep learning [1], but it necessitates a substantial amount of computational power and resources. TensorFlow Lite is particularly well-suited for deployment on edge devices. ...

Reference:

Real-time Object Detection on a Custom Built Edge Device Using TensorFlow Lite