Jae Hee Kim

This study proposes an antenna radome structure designed to support reliable long‐range wireless backhaul communication in extreme environments such as Antarctica (wind speeds of 30–40 m/s). Computational fluid dynamics (CFD) analysis based on three‐dimensional Reynolds‐averaged Navier–Stokes equations was conducted to compare the drag coefficients of four radome geometries: flat plate, dome, pyramid, and dimpled dome. The results show that the dimpled dome achieves the lowest drag coefficient of 0.189. However, the curved dome structure introduces phase distortion, leading to degradation of the antenna boresight gain. To address this issue, the dimpled dome was geometrically designed to exhibit a dielectric lens effect. Measurements at 5.8 GHz demonstrate that, when the proposed radome is applied, the antenna gain reaches up to 15.7 dBi, representing an improvement of more than 4 dB compared to the conventional structure. Furthermore, field measurements conducted in the Antarctic environment confirm that the proposed structure maintains strong wind resistance while delivering high gain performance, making it suitable for long‐distance communication links in extreme environments.

This paper proposes an artificial intelligence algorithm based on a mixture density network (MDN) to solve the inverse problem of estimating the excitation phases of antenna elements for array antenna radiation pattern synthesis. In array antenna radiation pattern synthesis, a nonuniqueness issue exists in which multiple phase solutions can produce an identical radiation pattern, which limits the effectiveness of conventional deep learning models in learning multiple solutions. The proposed MDN estimates the probability distribution of antenna excitation signals from a target radiation pattern, thereby effectively learning multiple feasible solutions. In addition, an iterative retraining framework based on the error between the target radiation pattern and the predicted radiation pattern is introduced to improve prediction accuracy. Simulation results for an eight‐element array antenna demonstrate that the proposed MDN achieves superior phase estimation performance for radiation patterns not included in the training dataset and arbitrarily generated radiation patterns.

Maglev trains require precise position data to control their propulsion systems. Specifically, sensing accurate phase based on position is necessary, for which cross‐inductive loop coils are used. To minimize phase errors, the envelope of the received voltage must be a sinusoidal wave depending on the position. This letter proposes the shapes of transmitting and receiving coils and presents, through simulation and experimentation, that the shapes of the transmitting and receiving coils and the induced voltage are in a convolution.

This paper presents a large-scale base station antenna assembly structure that is low-cost, reliable, and easy to manufacture. The antenna element is composed of a low-loss liquid crystal polymer based on a plastic molded module and a modified wideband stacked patch antenna. The base station antenna consists of a 4 × 8 antenna module, with each module comprising 3 × 1 subarrays along with dual-polarized antenna elements. The antenna element achieved an efficiency of 91% and an impedance bandwidth of 1.14 GHz within a height of 10 mm at 3.7 GHz. Furthermore, the fabricated array antenna structure was tested and verified to have an effective isotropic radiated power of 75.6 dBm at boresight and a steering range of less than ±60°. Therefore, the proposed structure meets the required antenna and beam-forming performance for commercial 5G active antenna systems.

In this study, we present a 1 MW-class wireless power transfer (WPT) system and test the operational data by applying it to an actual light-rail system for over 40,000 km. The data demonstrate that there are no reliability issues when applying the WPT system to light-rail systems. The power capacity of the WPT system was designed based on actual operating scenarios of light-rail systems. We propose a suitable transmitter and pick-up structure for real vehicle operation, and analyze the input/output voltages, current, power, and efficiency of the WPT system during operation. The applied WPT system achieved a maximum efficiency of 89%, which is considered high. The one-year operation data are expected to provide significant assistance in the future commercialization of wireless charging applications for railways.

This paper proposes a novel packaging and large-scale antenna-assembled structure for a printed circuit board (PCB) that reinforces productivity, facilitates cost reduction, and maintains reliability. This was achieved by splitting the antenna from the main board and packaging it into a radio-frequency integrated circuit. In addition, two innovative solutions—an externally attachable flexible PCB antenna and a PCB-embedded coaxial line—are introduced to overcome the degradation in antenna performance and vertical RF transition loss in the proposed low-cost hybrid PCB. First, the proposed externally attachable flexible PCB antenna and a parasitic aircoupled antenna, which were easily assembled on the PCB, achieved an antenna efficiency of 95% and an impedance bandwidth of 7 GHz. Second, the fabricated coaxial line exhibited enhanced impedance matching over a wide frequency range of 30–40 GHz and improved insertion loss of approximately 1.4 dB. Furthermore, the packaged antenna, composed of 256 dual-polarized antenna elements per stream, incorporated a 39 GHz CMOS-based 16-channel phased-array transceiver IC. The set-level beam-forming measurements were verified considering an effective isotropic radiated power of 55 dBm at boresight and a steering range >±60°. In addition to being suitable for mass production in terms of cost and reliability, the proposed structures and solutions met the required antenna and beam-forming performance for commercial 39 GHz base stations without sacrificing performance.

When a drone is used for inspection of facilities, there are often cases in which high-voltage power lines interfere, resulting in the drone being caught or falling. To prevent this type of incident, drones must be capable of detecting high-voltage power lines. Typically, a strong electric field is formed around the high-voltage lines. To detect the electric fields around high-voltage lines, this study proposes an electric field sensor that may be integrated within the body of a drone. In a laboratory environment, a voltage of 25 kV was applied to an overhead line, and the induced voltage in the proposed sensor was measured at various electric field intensities. Over an electric field range of 0.5 to 10.1 kV/m, a voltage of 0 to 0.77 V was measured with each proposed sensor. In addition, the electric field and the voltage induced in the sensor were measured in a real-world railway environment with overhead lines. Under these conditions, the proposed sensor has the compensated value of 4.5 when the measured electric field was 4.05 kV/m. Therefore, the proposed sensor may be applied in drones to measure large electric fields and to detect the presence of high-voltage lines in its vicinity.

We propose a deep neural network (DNN) to determine the matching circuit parameters for antenna impedance matching. The DNN determines the element values of the matching circuit without requiring a mathematical description of matching methods, and it approximates feasible solutions even for unimplementable inputs. For matching, the magnitude and phase of impedance should be known in general. In contrast, the element values of the matching circuit can be determined only using the impedance magnitude using the proposed DNN. A gamma-matching circuit consisting of a series capacitor and a parallel capacitor was applied to a conventional inverted-F antenna for impedance matching. For learning, the magnitude of input impedance S₁₁ of the antenna was extracted according to the element values of the matching circuit. A total of 377 training samples and 66 validation samples were obtained. The DNN was then constructed considering the magnitude of impedance S₁₁ as the input and the element values of the matching circuit as the output. During training, the loss converged as the number of epochs increased. In addition, the desired matching values for unlearned square and triangular waves were obtained during testing.