This letter presents a reconfigurable circularly polarized (CP) microstrip antenna that achieves polarization switching using only positive voltages. The antenna maintains a single-layer structure integrating a ring slot and a central via hole, enabling intrinsic DC-RF isolation within the antenna structure, eliminating the need for on-board choke components. A varactor diode and a chip capacitor are employed to control the ±90° phase difference between two orthogonal modes, allowing left-hand circular polarization (LHCP) and right-hand circular polarization (RHCP) switching without voltage polarity inversion. While maintaining the same circuit configuration as the previous bipolar-voltage design, the proposed antenna enables unipolar operation, providing easier integration with low-voltage control systems. A 5.8-GHz prototype was fabricated and measured, confirming CP reconfigurability with a low axial ratio (AR) and good impedance matching. The proposed element exhibits high implementation compatibility for high-density array and low-voltage-driven applications.

In mission-critical networked systems, such as motion control and automotive systems, high-capacity and high-reliability optical networks are required. This study proposes a method to improve the reliability and efficiency of motion control systems over optical ring networks based on Si-photonics-based in-vehicle optical network (SiPhON) architecture. The proposed method includes network redundancy using network coding and communication traffic reduction using event-triggered technology based on quality of performance, i.e., tracking error. Simulations confirm that the proposed method reduces communication traffic while maintaining control performance in SiPhON-based motion control systems.

This study investigates the feasibility of increasing transmission capacity by reducing adjacent channel spacing in wavelength division multiplexing (WDM) systems using hollow core fiber (HCF), which exhibits extremely low optical nonlinearity. Numerical simulations were conducted to evaluate dense coherent WDM transmission over HCF and compare its performance with conventional silica-core single-mode fiber (SMF). The results show that HCF enables stable high-density WDM transmission while maintaining narrow channel spacing. Furthermore, due to the high power tolerance of HCF, multi-span transmission with repeater spacing exceeding 160km is achievable while maintaining stable signal quality.

The beamspace (BS) method is an effective preprocessing technique for improving the efficiency of direction-of-arrival (DOA) estimation using array antennas. However, the conventional BS method sometimes results in an excess or shortage of beams depending on the observed data. This paper proposes an Adaptive BS (ABS) method that adaptively determines the number of beams in the BS method according to the observed single snapshot signal. We show the result of the initial evaluation of the ABS method, which indicates that the ABS method can appropriately reduce the number of input beams for the subsequent DOA estimation.

Automated ICT system design struggles with slow convergence because of large state spaces and sparse rewards. We propose a new framework that combines Double DQN with Truncated Noisy Networks to enhance learning stability. Double DQN helps prevent overestimating values, while truncated noise encourages more reliable exploration without destabilizing the GNN-based agent. Experimental results show this approach reduces the number of episodes needed for convergence by 22% compared to traditional methods. This indicates the approach’s strong potential to significantly accelerate complex network configuration tasks.

Ranging techniques using orthogonal frequency-division multiplexing (OFDM) signals play an important role in realizing integrated sensing and communication (ISAC) systems. However, conventional methods face challenges in separating closely spaced targets due to bandwidth limitations. To address this issue, we propose a novel ranging method that applies phase-compensated circle fitting to the trajectory of the channel frequency response on the complex plane. The proposed method is validated through experimental measurements involving two closely spaced targets. The results demonstrate that the proposed method achieves significantly improved performance, particularly when a power difference exists between the paths.

This letter proposes a graph-based method for estimating radio-wave propagation paths. Building-layout images are converted into graph structures, and a graph attention network (GAT) is trained to generate path images that represent reflection and diffraction routes. The estimated path images are then used as features for a neural-network-based path-loss estimator. Simulation results utilizing ray-tracing data show that the proposed method improves path-loss prediction accuracy compared to conventional convolutional neural network-based(CNN-based) approaches, thereby demonstrating the effectiveness of incorporating path information into propagation-loss estimation.

This paper investigates a low-latency decoding approach for convolutional polar codes using a neural network decoder (NND). Although NNDs enable one-shot decoding and significantly reduce latency, their error-rate performance can be further improved. To address this issue, we introduce a training strategy combining importance sampling, a ResNet-based architecture, and ensemble learning. Simulation results obtained using a convolutional polar code and an NND over an AWGN channel are presented. The proposed NND achieves stepwise improvements in BER compared with a conventional NND baseline, while preserving low decoding latency.

In display-camera communication (DCC), asynchronism and sampling rate mismatch between a transmitter and a receiver cause temporal inter-symbol interference and necessitate synchronization headers (preambles), which severely degrade effective throughput. This paper proposes a preamble-free spatio-temporal self-synchronization and maximal-ratio combining (MRC) demodulation scheme. By reverse-exploiting the spatial expansion of the temporal phase caused by the rolling shutter phenomenon as a spatio-temporal fingerprint, along with zero-sum orthogonal codes superimposed on the transmission data, the proposed scheme simultaneously estimates the temporal parameters (e.g., initial phase) from merely a single captured frame. Furthermore, MRC based on the estimated time parameters optimally recovers the dispersed signal energy to maximize the output SNR. Theoretical analysis clarifies the mathematical uniqueness and convergence of the proposed model, and quantitatively demonstrates that it achieves Mbps-class transmission, overwhelmingly surpassing conventional schemes in asynchronous oversampling environments.

Device-Free Localization (DFL) is a promising approach for smart environments and disaster response because it enables non-intrusive tracking without requiring users to carry radio devices. Radio Tomographic Imaging (RTI) enables accurate DFL by exploiting shadowing effects induced by targets on wireless links. This method models link paths between anchor nodes using a voxel-based map and estimates target locations by superimposing shadowed paths without prior environmental measurements or fingerprinting. However, robust performance depends strongly on the selection of model parameters. Because determining these parameters requires extensive calibration measurement, efficient procedures are essential for practical use. To address this issue, this letter proposes a measurement-based training strategy for Bayesian-optimized multipath-RTI and investigates how different training-point configurations affect localization performance. Experimental evaluations at 25 indoor positions demonstrate that the proposed strategy achieves stable and generalizable localization performance across different spatial conditions with reduced calibration effort.

In this paper, we propose a received signal strength indicator (RSSI) estimation model using sparse measurements as input. In smart factories, there are numerous wireless communication devices, so optimization of access point placement is required. Previously, an RSSI estimation model using a 3D U-Net with a convolutional neural network (CNN) was proposed. The proposed model aims to improve estimation accuracy by providing sparse measurements to the Non-Line-of-Sight (NLoS) area. As a result, estimation accuracy improved by an average of 1.24dB compared to conventional methods, and estimation time was reduced to 1/82 of that required for simulation.

This letter implements an ultra-low-power IoT wireless sensor driven by energy harvesting (EH) from 2.45 GHz Wi-Fi and verifies its performance. First, a planar Yagi-Uda antenna was used to extend transmission distance, and its maximum gain was 12.2 dBi. Next, a high-conversion-efficiency rectifier was implemented using surface-mount components to achieve ultra-low input power operation and achieved 15.7% conversion efficiency at -20 dBm input power. Finally, connecting these components to the sensor module enabled sensor operation at a distance of 3.0 m from the Wi-Fi router. This shows that the proposed IoT wireless sensor offers high flexibility and enables semi-permanent operation.

Out-of-distribution (OOD) detection is critical for the reliable deployment of radio-frequency (RF) signal recognition systems, as novel signals may severely degrade performance. We propose a multi-expert framework that integrates time-, frequency-, and time-frequency-domain experts to construct a richer feature space. Within this space, a normalized Mahalanobis distance is employed to jointly assess feature similarity and cross-view consistency. This design enables more robust separation between known and unknown signals compared with single-domain approaches. Experiments on both a synthetic Automatic Modulation Recognition (AMR) dataset and a real-world Specific Emitter Identification (SEI) dataset show that our method significantly outperforms comparison approaches in OOD detection while maintaining high closed-set classification accuracy.

This letter analyzes the test reproducibility of the BCI method, which is one of the immunity tests for automotive components. The analysis does not take a probability-based approach, but is carried out by using an experimental jig that ensures reproducibility for the wire placement and layout. The results of evaluation and electromagnetic field simulation show that sufficient reproducibility is guaranteed under conditions where the position and layout are guaranteed. Based on this data, the appropriate amplitude of the injected current when applying the BCI method to a single wire is also discussed.

This paper presents the uplink coverage expansion effects achieved by employing relay stations equipped with true-time-delay (TTD) arrays in the millimeter wave band. While millimeter waves provide the advantage of wide bandwidth availability, they also suffer from significant propagation losses, making beamforming techniques using array antennas essential. However, for users located at geographically distant positions, it is necessary to switch beams temporally, which restricts throughput per user terminal and results in shortened uplink coverage. Recently, the focus has shifted to joint phase-time array (JPTA), which allows for the formation of frequency-dependent analog beams by integrating TTD circuits into the antenna elements of conventional phased array antennas, thus enabling the expansion of uplink coverage. Nevertheless, even with JPTA implemented at the base station, the extension of uplink coverage remains constrained in the millimeter wave band due to the impact of obstructions. Therefore, this paper provides a simplified evaluation of the uplink coverage expansion effects when applying JPTA in relay stations, comparing these effects with the coverage provided by traditional relay stations equipped with phased arrays.