ABSTRACT. This paper investigates the operating mechanism of a multiband antenna composed of a slotted metal case (SMC) and an internal folded dipole element (FDE) using Characteristic Mode Analysis (CMA). While the FDE exhibits multiband operation with limited bandwidth,integrating it into the SMC results in two wide operating bands. To clarify the bandwidth enhancement mechanism, CMA is applied to the SMC, the FDE, and the combined antenna structure. The modal characteristics are compared to identify the dominant modes and their interactions. The results provide physical insight into the wideband multiband behavior of the antenna.

ABSTRACT. In epidemic communication, information delivery is realized by direct wireless communication between mobile nodes and mobility. Information floating (IF) is a kind of epidemic communications and permits transmission from mobile nodes only in transmittable area (TA) to prevent meaningless spreading and transmissions. In this paper, we share answers to a question using IF (Q&A IF) and apply this Q&A IF to the collection of road damage information during disasters. We consider three methods based on Q&A IF and evaluate these methods by computer simulations. From the comparison of these methods, we show that encouraging behavior change of mobile nodes that find damaged roads and receive location of TA is effective on road damage information sharing.

ABSTRACT. Roads around the Sendai Castle Ruins in Aobayama, Sendai, experienced long-term closures following the Great East Japan Earthquake. Although these closures were lifted, multiple road sections are scheduled to be closed again by 2030 as part of the historic site redevelopment project. This study developed three road traffic network models representing the past (post-earthquake conditions), the present (current conditions), and the future (reflecting the planned redevelopment) using Discrete-Event Simulation (DES). Simulation experiments were conducted by gradually varying vehicle interarrival times, and average speed, traffic throughput, and traffic density were compared. The results indicate that traffic bottlenecks and network constraints shift in response to changes in road operations and traffic demand.

ABSTRACT. Coupled analysis requires CAD modeling, meshing, and CAE setup, but CAD model creation remains time-consuming and costly. This study proposes an AI-assisted CAD workflow for coupled analysis using large language models and open-source CAD/CAE tools. In the proposed workflow, AI generates 3D model data and Python code, which are reviewed in FreeCAD, converted into mesh data using Gmsh, and prepared for CAE. Preliminary results show that simple 3D models can be successfully generated and meshed, demonstrating the feasibility of the proposed approach.

ABSTRACT. We have previously investigated the influence of experimental conditions in sound lo-calization experiments on the accuracy of sound localization using simulation meth-ods[1,2]. The models we had been using up until then approximated the human head and the conductor with spheres. In this paper, we use data obtained by capturing the 3D shape of a commercial dummy head (Brüel & Kjær Type 4128C[3]) with an ul-tra-high-precision 3D scanner as a simulation model. The HRIR for a loudspeaker posi-tioned in front of the subject is calculated using non-stationary analysis with ADVEN-TURE_Sound[4]. The accuracy of the obtained hrir calculation will be verified.

ABSTRACT. In 3-D animation movie production, 3-D modeling and ren-dering of urban landscape background scene often require a large amount of manual work, resulting in significant produc-tion costs and time consumption. To address this issue, this research employs Python scripting on a 3-D CG software, Blender to generate buildings and urban scenes procedurally.

ABSTRACT. 3-D large-scale electro-magnetic analysis with more realistic complicated geometry and more detailed material modeling may be handled. However, its model preparation phase is now a clear bottleneck. A robust and fast mesh generation technique is desired. To overcome the situation, a simple tetrahedral mesh generator is developed. It is based on a voxel-based, with boundary fitting techniques. To connect voxel-patterned structural grid parts and free-form unstructured grid parts, advancing front method is introduced.

ABSTRACT. This paper proposes a position-dependent spatial audio presentation method designed to enhance the sense of immersion within virtual reality spaces. Specifically, physics-based acoustic simulations were utilized to replicate reflected sounds tailored to the architectural geometry. Furthermore, by employing the Adhan as the sound source, we recreated its char-acteristic reverberation both inside and outside the Kobe Muslim Mosque. Consequently, this approach provides a more realistic VR experience by integrating auditory information with visual data.

ABSTRACT. This paper focuses on a VR digital archive system for viewing the Yamadate process of Gion Festival Yamahoko floats as intangible cultural heritage. The proposed method extracts differences from three-dimensional measured point cloud data acquired at multiple time points and uses the extracted data to visualize temporal changes. This enables users to understand temporal changes in the Yamadate process within an immersive VR environment.

ABSTRACT. Three-dimensional Gaussian Splatting (3DGS) enables high-quality novel-view rendering, but conventional pipelines rely on multi-view images and camera poses typically estimated through Structure-from-Motion (SfM). This requirement makes it difficult to construct 3DGS representations directly from existing point-cloud data, particularly when a transparent visualization style is desired. We present an SfM-free pipeline for generating transparent 3DGS representations from point clouds. Virtual cameras are automatically distributed around the target object using spherical sampling, and Stochastic Point-Based Rendering (SPBR) is used to synthesize transparent multi-view images from the input point cloud. For each viewpoint, the pipeline also generates a binary foreground mask and records the corresponding camera parameters, which are converted into a format compatible with 3DGS training. In parallel, the input point cloud is downsampled and used to initialize the Gaussian primitives. A preliminary experiment on a cultural-heritage point-cloud dataset demonstrates that the proposed pipeline can reconstruct a transparent 3DGS representation without manual image acquisition or SfM-based pose estimation. The results indicate the feasibility of converting point-cloud data into continuously viewable transparent 3DGS representations.

ABSTRACT. hree-dimensional Gaussian Splatting (3DGS) has recently attracted considerable attention for real-time photorealistic rendering. However, conventional 3DGS methods require original multi-view photographs, making them difficult to apply to legacy datasets con-sisting only of colored point clouds. This paper proposes a novel pipeline that generates Gaussian Splatting models from colored point clouds using Stochastic Point-Based Ren-dering (SPBR). High-quality virtual images are rendered from multiple viewpoints and subsequently reconstructed into Gaussian Splatting models using RealityScan. Experi-mental results demonstrate that SPBR-generated images enable successful Gaussian Splat-ting reconstruction, indicating the feasibility of applying 3DGS to existing colored point cloud archives.

ABSTRACT. Although “Digital Twin” has gained widespread attention in the AI era, its foundational concept is already deeply embedded in many scientific fields. Tokamak physics is particularly well suited to this paradigm, as accurate fusion plasma modeling demands high-fidelity representation down to complex engineering details. In this talk, we introduce the Formosa Integrated Research Spherical Tokamak (FIRST)—Taiwan’s pioneering tokamak project—and outline our workflow for building its digital twin using NVIDIA Omniverse. A modern digital twin requires high-quality 3D rendering alongside real-time user interaction, both of which Omniverse uniquely provides. Furthermore, because full-physics simulations are computationally intensive, integrating AI surrogates is essential. We will highlight our current progress and future directions in developing a complete digital twin for FIRST.

ABSTRACT. Against the backdrop of global environmental changes, unprecedented extreme phenomena have been occurring one after another in various regions around the world. To obtain further deep understandings of the future of the Earth environment and decide which actions we should take, there are still various challenges that must be addressed. Simulations for the atmosphere, oceans, and the global environment have contributed to the accumulation of a wealth of knowledge on our future. However, considering the future development of AI, we are faced with the practical problem of a severe shortage of diverse types of data, including observational data. In this lecture, I will highlight examples of extreme phenomena, outline status of simulations related to the atmosphere, oceans, and the global environment, and explore future directions.

ABSTRACT. A Wireless Sensor Network WSN is widely used in various applications. There in classic WSN models, each Sensor Node SN is forced to operate under severe resource limita-tions. Many SNs are randomly deployed in a field. There the base station BS to collect all sensed data and to connect to the outside user is located far away from such field. A layer 3 routing in a WSN is typically classified into either centralized or distributed by where data transmission routes are constructed. In the former the BS constructs such routes, while in the latter each SN itself does in a self-organized way. In this paper, we propose a novel layer 3 routing scheme, termed a switch routing, from a distributed way to centralized aiming at longer WSN operations. There at first PEGASIS (Power Efficient in GAthering Sensor Information Systems) operates halfway to collect sensed data as well as in-trees representing data transmission. Next, using those in-trees, the BS solves an integer programming problem to maximize the rest WSN operations. Simulation ex-periments show our proposal valid providing a 1.8 to 60.8%, 15.4% on average, longer first dead node time than PEGASIS alone.

ABSTRACT. The relation of channel assignment problems in wireless communications and coloring problems of graph theory is well-known. A channel in the wireless communication is corresponding to a color assigned to a vertex (an edge) in a graph. It is popular to minimize the number of colors assigned to vertices of the graph. However, the maximum number of colors called Grundy Number is also important. In greedy colorings of a graph, it is well-known that the minimum number of colors is the chromatic number and the maximum number of colors is Grundy number. In the previous study, for any value k between the chromatic number and Grundy number of graph G, there is a greedy coloring of G with k colors. In this paper, we consider this property holds or not in a model close to real wireless communication.

ABSTRACT. The chemical evolution of organic molecules in interstellar and prebiotic environments is essential for understanding the origin of life and the nature of cosmic dust. In this study, we utilize Reactive Force Field Molecular Dynamics (ReaxFF-MD) to evaluate two fundamental evolutionary processes: (1) the polymerization of alanine under thermal cycles mimicking vacuum ultraviolet (VUV) irradiation, and (2) the fragmentation of coronene (C$_{24}$H$_{12}$) induced by energetic nitrogen impacts to clarify the synthesis of Quenched Nitrogen-included Carbonaceous Composite (QNCC). Our simulations demonstrate that high molecular density and repeated thermal pulses are critical for generating high-mass products (up to 315 g/mol), with stable peptide bonds identified via DFTB$+$ analysis. For the coronene system, we first identify favorable adsorption sites via systematic potential energy mapping, classifying the coronene surface into three distinct regions (R$_1$, R$_2$, R$_3$) with different adsorption characteristics. Single-collision simulations confirm that adsorption preferentially occurs in the low-energy outer region (R$_3$), consistent with the potential energy landscape. Detailed energy-dependent fragmentation behavior and multi-collision dynamics are reported in a companion study which will be presented by Takeuchi et al. These results provide a unified atomistic understanding of how cosmic organic matter complexifies or fragments under non-equilibrium energy inputs.

ABSTRACT. Catechins, the primary polyphenols in green tea, exhibit potent antioxidant activity, with the scavenging of highly reactive hydroxyl (OH) radicals being of paramount importance. However, the transient nature of OH radicals complicates experimental observation of their scavenging mechanisms. In this study, we employed molecular dynamics (MD) simulations to elucidate the early-stage scavenging mechanisms of four major catechins (EC, EGC, ECg, and EGCg), focusing on radical approach and the formation of pre-reactive complexes. We quantitatively evaluated the access probability of OH radicals to specific oxygen sites and the formation of hydrogen bonds between catechins and OH radicals. Our findings demonstrate that both access probability and hydrogen bonding frequency follow the order EGCg > ECg > EGC > EC, which correlates strongly with established in vitro antioxidant activity data. Structural analysis reveals that pyrogallol moieties, characterized by three adjacent hydroxyl groups, significantly enhance radical recognition and complex stability.

ABSTRACT. Multiple ionization plays an important role in high-linear energy transfer (LET) particle irra-diation and may affect water radiolysis and subsequent DNA damage. In this study, Geant4-DNA was modified to incorporate multiple ionization processes, and water radiolysis simulations were performed for proton, alpha-particle, and carbon-ion irradiation. The LET dependence of the G values of hydroxyl radicals (•OH) and hydrogen peroxide (H₂O₂) was evaluated and compared with experimental data. The simulated G values showed good agree-ment with reported measurements, demonstrating the validity of the developed simulation framework. The influence of multiple ionization on radiation-induced DNA damage is currently under quantitative investigation.

ABSTRACT. Application of optical vortex to large capacity optical fiber communication has been discussed mainly based on the Laguerre-Gaussian mode. Then the Laguerre-Gaussian is free space solution to paraxial approximation of wave equation and stable propagation is not easy in core of optical fiber with finite radius. In this work, it is considered to ap-ply a hybrid-mode optical vortex to the optical fiber communication. The propagation of the hybrid-mode optical beam is numerically simulated by using FDTD method and sta-bility of the hybrid-mode optical vortex propagation is evaluated.

ABSTRACT. To aim to effectively use optical vortex for various applications, we discuss creation of focused optical vortex beam using Schwarzschild-type lenses in this work. Then, it is nec essary to carefully design the mirror surface shape of the Schwarzschild-type lens to maintain proper phase structure (spiral wavefront) of the optical vortex at the focus point. This paper presents full-wave electromagnetic field analysis for focusing phenomena of the optical vortex in the Schwarzschild-type lens, by using a 2.5-dimensional method of moments (MoM).

ABSTRACT. Finite-difference time-domain (FDTD) simulations of optical vortices generated by helical filters are affected by non-physical reflections caused by staircase approximation. Although bidimensional empirical mode decomposition (BEMD) can reduce these artifacts, it does not explicitly provide temporal or physical consistency. This paper proposes a conditioned Maxwell physics-informed neural network (PINN) that combines BEMD-based supervision with Maxwell residuals, orbital angular momentum (OAM) constraints, and consecutive-frame consistency. Ours present results show smoother electric-field distributions than BEMD and phase patterns consistent with the topological charges.