|
Design of a High-Performance Instance Segmentation Model Using Uncertainty-Guided Non-Maximum Suppression and Normalization |
|
Seung Il Lee, Sangbeom Jeong(Seoul National University of Science and Technology, Korea), Hyun Kim(SeoulTech, Korea) |
| In this paper, we propose a robust instance segmentation model that efficiently tackles noise by leveraging normalized uncertainty in post-processing. Furthermore, we propose a novel technique called multi-uncertainty non-maximum suppression, which selects results with higher reliability compared to existing models. The experimental results show that the proposed method outperforms the existing baseline by achieving a performance improvement of 2.9%. |
|
|
Protection of Decentralized DC Microgrid System based on Detection of DC-bus Voltage Sensor Faults |
|
Dat Thanh Tran(SeoulTech, Korea), Seong Bae Jo, Kyeong-Hwa Kim(Seoul National University of Science and Technology, Korea) |
| Unique Contribution
In this study, authors presented a protection of a decentralized DC microgrid system based on the detection of DC-bus voltage sensor faults with multiple power sources, such as a utility grid (UG), a distributed generator (DG), and a photovoltaic (PV). In the proposed scheme, a voltage droop controller is first deployed to regulate the DCV to the nominal value and maintain the power balance in the decentralized DCMG under various uncertain conditions. To enhance the reliability of the decentralized DCMG under DCV sensor fault, a fault identification algorithm is presented for each power agent to detect the faulty DCV sensor. Then, a protection method for the decentralized |
|
|
Electrically Beam Steerable Phased Array Antenna Controlled by Varactors for Point-to-Point Wireless Communication |
|
Sebastian Verho, Van Thang Nguyen, Jae-Young Chung(SeoulTech, Korea) |
| This work proposes a different approach to design a phased array antenna by replacing a beamformer IC to more affordable varactors. They are controlled by altering their voltage with an open-loop control method. The challenge of this approach arises from the non-linearity of the varactor which requires a proper characterization to provide robust and predictable control. Therefore, we utilize an analytical method combined with measurement data to calculate the required control voltages for each varactor in the system. To demonstrate the feasibility of this approach, two different phased array designs along with measurement results are presented which operate in 3 GHz & 28 GHz, respectively. |
|
|
Performance Analysis and Deep-Learning-Based Real-Time Evaluation for Multihop MIMO Full-Duplex Relay Networks with Short-Packet URLLCs |
|
Ngo Hoang Tu(SeoulTech, Korea), Kyungchun Lee(Seoul National University of Science and Technology, Korea) |
| This work investigates multihop MIMO full-duplex relay (FDR) networks using short-packet communications in accordance with uRLLCs. The end-to-end block-error rate of FDR is analyzed and compared with that of half-duplex relay, from which the effective throughput, energy efficiency, reliability, and latency are also studied. However, the derived analytical expressions contain non-elementary functions, making them intricate for real-time configurations. Motivated by this, we introduce a deep multiple-output neural network with a short execution time, low computational complexity, and highly accurate estimation, which is demonstrated its efficiency by numerical results. |
|
|
Smart Energy Town: Constructing OPC UA-Based Network Architecture |
|
Changdae Lee, Young Il Lee(SeoulTech, Korea) |
| This paper explores how Open Platform Communication Unified Architecture (OPC UA) can be applied to smart energy towns for better management and integration of diverse, renewable energy sources. The study addresses challenges in system integration, network interoperability, and security by using OPC UA to coordinate various energy assets efficiently. The paper outlines the architecture of a smart energy town, details how to establish a network, and discusses the implementation of OPC UA-enabled devices. It also includes a case study on a SCADA system, contributing to the field by offering solutions for effectively managing distributed energy resources in smart towns. |
|
|
Performance Evaluation of 5G NR Vehicle-to-Vehicle Communication |
|
Derek Kwaku Pobi Asiedu(IMT-Atlantique, France), Kofi Amoako Ofori Amanfo, Maurice Nduwayezu(Seoul National University of Science and Technology, Korea), Ji-Hoon Yun(SeoulTech, Korea) |
| In this paper, we highlight the need for evaluation methodologies of fifth-generation (5G) new radio (NR) vehicle-to-vehicle (V2V) communication to better evaluate technologies for corresponding standards. To enable the evaluation of 5G NR V2V communication based on the established 3GPP evaluation methodologies, we develop a simulator named 5GModV2V. Leveraging the functionalities of the Matlab 5G communication toolbox, this simulator offers a platform for evaluating 5G NR V2V communication. We conduct performance assessments of 5G NR V2V communication in both highway and urban communication scenarios, utilizing the 5GModV2V simulator. |
|