This study investigated the use of local 5G technology to transmit platform camera footage to train drivers during driver-only operations. A test train was equipped with a local 5G terminal and monitored to evaluate the session establishment time, monitor processing time, and video latency. The results showed that the establishment and processing times were sufficiently short for practical use, and that real-time video transmission is achievable with the appropriate selection of the encoder and decoder.

We propose a model that combines a neural network composed of LSTM units with polynomials corresponding to each average input power level. This combination enables high-precision modeling of the rectifier’s output signals in response to modulated signal inputs at multiple average power levels. The accuracy of the proposed model was evaluated using the Normalized Mean Square Error (NMSE). The evaluation results confirmed that, for all average input power levels, the proposed model reproduces the actual rectifier’s output signals more accurately than a model using that uses only the a neural network.

Packet classification is used to determine the behavior of packets incoming to network devices based on a predefined policy. In linear search, the increasing number of comparisons between packets and rules can cause communication latency. The problem of optimizing the rule order to minimize the number of comparisons has been actively studied. Since this problem is known to be NP-hard, various heuristic approaches have been proposed. In this paper, we propose a new approach to optimizing communication latency by focusing on rule order that reduces latency involved in parts of the rule list. We conduct computational experiments to demonstrate that the proposed algorithm is faster and more efficient than existing algorithms.

BBR is a new congestion-based congestion control algorithm proposed by Google to address the bufferbloat problem. It is designed to achieve the maximum throughput with the minimum RTT by continuously estimating the available bandwidth. Our preliminary evaluation results in this paper show that delay jitters in ACK packets cause overestimation of available bandwidth and accordingly increase of RTT. In this paper, we propose a new BBR rate calculation algorithm, Low-Pass-Filter-based BBR (LPF-BBR), which mitigating fluctuations of calculated delivery rate. Our evaluation results by computer simulation show that our proposing LPF-BBR can estimate bottleneck bandwidth accurately and prevent increase of RTT caused by overestimation, even in the case that ACK packets suffer delay jitter. As far as we know, this is the first work focusing on influence of backward path congestion in BBR.

A propeller LED transmitter (P-Tx) further boosts image sensor communication (ISC) throughput by mounting LEDs on a rotating arm and encoding data via their On/Off state. Reducing the angular interval per bit increases data rate but also raises the risk of overlapping “light trails” and ensuing interference. To date, no analytical model has linked the control angle to inter-trail interference in P-Tx-based ISC. This paper proposes a fundamental model of light trail generation and blinking control conditions to prevent interference between adjacent light trails. In the proposed model, the luminous energy distribution on the image plane is calculated based on the LED rotation trajectory. We also provide preliminary experimental results to validate the proposed model.

In this study, we propose a text-based CAPTCHA that utilizes human visual completion to prevent automated programs from getting through, without affecting human cognition. This CAPTCHA requires users to enter the reading of the kanji strings, whose centers are hidden behind a black bar. This method makes it challenging for machines to recognize the characters, while still allowing humans to identify the original text through their visual completion ability. We evaluate both the usability and security of the proposed CAPTCHA. To assess usability, we observed how humans respond to the CAPTCHA challenge. For the security evaluation, we tested the machine’s ability to read the kanji using a Convolutional Neural Network (CNN). The results showed that the proposed CAPTCHA achieves a success rate comparable to traditional CAPTCHAs, with a slightly shorter response time than a similar type of CAPTCHA. Furthermore, security experiments have shown that it exhibits a certain degree of resistance against simple convolutional neural networks.

This paper proposes a novel method for controlling the directional pattern of array antennas using a quantum computer. Phase shifters are attached to each element of the array antenna, and the problem of controlling the directional pattern is formulated as a QUBO (Quadratic Unconstrained Binary Optimization) format. The optimal weight coefficients of the phase shifters are determined using quantum annealing. It is shown that a high-gain beam is formed in the desired direction by the proposed method. Also, a formula for forming multiple beams has been proposed. Furthermore, it is confirmed that the executing time remains stable even as the number of antenna elements is increased. Effectiveness of the proposed method for high-speed control of large-scale array antennas is demonstrated.

Elastic Optical Networks (EON) can support high data bit rates efficiently and network virtualization technology manages network resources flexibly by running multiple virtual networks (VN) on a substrate network. However, it is difficult to provide an acceptable Quality of Service (QoS) during network congestion due to bandwidth unavailability. Bandwidth degradation, which corresponds to allocating a smaller amount of bandwidth than requested, is one solution to this problem. In this paper, we propose an adaptive bandwidth degradation algorithm for VN construction in EONs that is QoS assured. The performance of the proposed method has been evaluated by simulation, and numerical results show that the proposed method has significantly reduced the blocking probability.

In recent years, the Internet of Things (IoT) has been increasingly deployed in various domains. Message Queuing Telemetry Transport (MQTT), which features low overhead, has been adopted as a communication protocol for IoT to replace conventional HTTP. MQTT is a representative protocol of the publish/subscribe (Pub/Sub) communication model and enables high-speed, low-latency data transmission. However, in wireless environments, transmission errors and handovers may cause packet loss, thereby degrading the communication quality. This paper investigates the impact of such wireless transmission impairments on IoT communications in mobile environments using the Gilbert-Elliott model, a well-known model for characterizing bursty transmission errors in wireless channels.

This study obtained the far-field radiation pattern of a reflector antenna from the near-field amplitude distribution using the Phase Retrieval (PR) method in the plane perpendicular to the axis tilted from the optical axis. The PR method usually uses the near-field amplitude distribution in the plane perpendicular to the optical axis of the antenna. However, due to the design of the measurement system and antenna, the measurements had to be made in a plane perpendicular to an axis tilted from the optical axis. Since the literature has not reported such cases, this study investigated the determination of phase retrieval from the near-field amplitude distribution in a tilted plane by simulation.

In this paper, we derive analytical expressions for the cumulative distribution function (CDF) of channel capacity achievable by suboptimal joint transmit-receive diversity (JTRD) for perfectly correlated and uncorrelated fading cases in a strongly asymmetric MIMO channel. The derived CDF of channel capacity is validated by Monte-Carlo numerical computation. It is confirmed that in a strongly asymmetric MIMO channel, the fading correlation has a favorable impact on suboptimal JTRD based on maximal-ratio diversity (MRD), while it only slightly degrades suboptimal JTRD based on selection diversity (SD). The fading correlation has also a favorable impact on optimal JTRD, which is confirmed by investigating the behavior of eigenmode transmission by using Monte-Carlo numerical computation.

This paper proposes a dual-frequency polarization-multiplexed metasurface for Orbital angular momentum (OAM) divergence control using phased-array principles. Compared to FSS-loaded designs, the unit achieves ultrathin profile, full-phase coverage, and angle insensitivity. By implementing phased-array-based divergence tuning, we demonstrate four-channel high-purity vortex beams with equal divergence angles across modes. Fabricated prototypes validate the OAM control capability, showing good agreement between simulated and measured results. The design shows strong potential for vortex-beam communications and radar detection.

To reduce the risks of natural disasters, proactive use of environmental data is important. However, securing electricity in rural areas is often challenging. Thus, IoT gateways powered by power-over-fiber (PoF) are required. In this work, we develop an intermittent IoT gateway by implementing a task-aware module operation algorithm suitable for intermittent IoT sensing with power-limited PoF. The algorithm selectively powers the necessary modules and minimizes their operating time during the active state. As a result, power consumption in the active state is reduced by 56% compared with previous work, in which almost all modules were activated. Furthermore, the algorithm suppresses the temperature increase of the transmitter, one of the most power-hungry modules, by ∼11K.

This study investigates the effect of equalization-enhanced phase noise (EEPN) on fiber-optic transmission systems that employ a phase-sensitive amplifier (PSA) based on double-sideband (DSB) modulation. The DSB-PSA scheme enables digital chromatic dispersion (CD) pre-compensation, facilitating the universal deployment of phase-sensitive pre-amplification; however, this process introduces EEPN. Through spectral analysis and visualization, we show that EEPN can be mitigated by phase-sensitive amplification and manifests as frequency-dependent coherent gain fluctuations. Numerical simulations with a 96-Gbaud 16QAM signal confirmed that the PSA significantly improves signal quality, even under severe EEPN conditions with large laser linewidths and/or high CD values.

This study applies Bayesian learning techniques, specifically Variational Inference (VI) and Monte Carlo Dropout (MC Dropout) to Automatic Modulation Classification (AMC). Both methods are built upon a Convolutional Neural Network (CNN) framework and are capable of detecting certain out-of-domain (OOD) modulations through threshold-based decision making. The results indicate that the confidence scores assigned to OOD modulations are influenced by their proximity to the trained in-domain classes. Higher confidence scores are typically observed for modulation types that closely resemble known modulation classes. This approach enables effective classification by training only on essential modulation types, thereby reducing the need for exhaustive class coverage and supporting the constraints of edge computing environments.

Image overlay uplink screen camera communication consists of a small screen with LED backlight modulation and a rolling shutter telephoto camera. In a previous study on symbol synchronization, object detection deep neural network (DNN) detected the screen, and autocorrelation detected the preamble separately. However, both object detection DNN and autocorrelation can detect the preamble. First, in this study, object detection DNN detects the screen and preamble. Next, autocorrelation precisely detects the preamble assisted with object detection DNN, improving the efficiency and accuracy of preamble detection. Also, preamble pattern suitable for object detection DNN is revealed.