Journal of Military Science and Technology

A solution for analyzing and evaluating quantum communication networks through simulation

2025-10-30T07:58:27+00:00

This paper proposes a comprehensive solution for analyzing and evaluating the performance of quantum communication networks through the use of the open-source simulation software, QuNetSim. This method provides a flexible and cost-effective approach to studying network protocols and parameters prior to hardware implementation. The main content of the paper focuses on modeling realistic factors such as channel loss, noise, and eavesdropping behaviors. The research conducted detailed simulation scenarios with the BB84 and E91 protocols. In particular, a general network scenario, integrating quantum repeaters and teleportation, was analyzed through Fidelity. The results not only confirm the crucial role of repeaters in maintaining long-distance connectivity but also quantify the performance degradation caused by channel loss and eavesdropping. The simulation results also demonstrate that QuNetSim is a powerful tool, enabling researchers to easily build, test, and optimize complex quantum networks, thereby accelerating the transition from theory to practice.

A variational quantum eigensolver (VQE) algorithm for optimal target selection in two-dimensional space

2025-10-30T08:03:58+00:00

This paper applies the variational quantum eigensolver (VQE) algorithm to solve the problem of optimal target selection in two-dimensional space. Specifically, we consider a system comprising multiple moving targets, each traveling in a straight line at a different constant speed and heading towards a fixed object A. The objective of the problem is to find the target with the minimum time of approach to the fixed object A. Although this problem can be solved efficiently by classical methods, the quantum method is proposed here to demonstrate the potential of quantum computing in solving more complex versions of combinatorial optimization problems. We focus on an in-depth analysis of encoding the classical problem into a quantum problem, constructing the objective function H (Hamiltonian), and detailing the procedural steps of the VQE algorithm.

Research and evaluation of underwater acoustic signal classification capability using quantum K-mean algorithm on real dataset

2025-10-30T08:09:41+00:00

In the big data era, traditional machine learning algorithms like K-means face computational challenges in processing large, high-dimensional datasets, particularly for underwater acoustic signal classification. This study investigates the application of the quantum clustering algorithm Q-means, a quantum-enhanced variant of K-means, to real passive sonar datasets. The purpose is to evaluate Q-means' effectiveness in classifying propeller-driven ship signals under noisy conditions, leveraging quantum principles such as superposition and entanglement for exponential speedup. The research extends classical K-means to δ-k-means for robustness, implementing quantum subroutines including distance estimation, cluster assignment, and state tomography. Preprocessed features from power spectral density analysis of real datasets (from prior studies on varying ship speeds) are used as input to avoid issues with raw time-series data, such as high dimensionality and sensitivity to shifts. Simulations on 15,000 passive sonar samples demonstrate that Q-means achieves clustering quality comparable to K-means, with clear separation of three clusters and accurate centroids, while reducing complexity from O(n) to O(polylog(n)). This validates Q-means as a promising tool for large-scale, noisy acoustic data in national security applications, bridging quantum theory with practical machine learning.

Research on solutions to enhance trajectory quality in underwater autonomous navigation of unmanned underwater vehicles using IEKF SLAM algorithm for SONAR data

2025-10-30T08:13:35+00:00

The operation of remotely operated underwater vehicles in underwater environments always faces the challenge of lacking GPS signals, leading to the accumulation of positioning errors over time. This instability in motion significantly reduces the efficiency and safety of practical operations such as infrastructure inspection, seabed surveying, and search and rescue missions. This paper presents a Simultaneous Localization and Mapping (SLAM) method based on enhanced sonar data for the operational capability of underwater vehicles. The proposed algorithm fuses data from sonar with an inertial measurement unit (IMU) within an Iterated Extended Kalman Filter (IEKF) framework to optimize the vehicle's trajectory and correct for accumulated errors. By processing sonar data to extract features and then generating loop closure constraints to combine with motion estimates from odometry, the proposed model optimizes the entire trajectory of the vehicle and effectively corrects accumulated errors. The results obtained are highly accurate pose estimates and a consistent map of the operational environment throughout the voyage. The successful implementation of this algorithm demonstrates great potential in enhancing the autonomy, reliability, and operational efficiency of underwater vehicles in practical applications.

A smart roadside unit with integrated vision and 24 GHz IOT radar for real-time behavior prediction and V2X communication

2025-10-30T08:18:29+00:00

This paper proposes a Smart Roadside Unit (RSU) architecture designed for intelligent transportation systems in high-density urban environments with heterogeneous traffic. The system integrates V2X communication, 24 GHz radar, and AI-based vision to enable autonomous, context-aware decision-making for traffic regulation and incident response. Leveraging a machine learning model trained on localized traffic behavior, the RSU not only detects violations but also predicts potentially hazardous actions specific to each area. Real-time sensor fusion between radar and camera enhances the accuracy of object detection, tracking, and behavioral analysis under various environmental conditions. The proposed system improves traffic flow management, congestion mitigation, and proactive control capabilities. The solution is well-suited for deployment in smart cities, particularly in the context of increasing urbanization and the growing demand for data-driven mobility infrastructure.

A 32×32 16-bit integer matrix multiplication hardware acceleration design and evaluation on 45nm PDK

2025-10-30T08:25:45+00:00

This paper presents the design and implementation of a 32×32 matrix multiplier using 16-bit integer data, targeted for hardware acceleration applications. The design is described in VHDL and synthesized using Cadence EDA toolchain with a FreePDK45nm CMOS process for ASIC implementation. The proposed architecture employs pipelining and parallelism techniques to optimize speed and power consumption. Post-layout Place and Rout results demonstrate that the design achieves a maximum operating frequency of 200 MHz, occupies an area of 107,240 μm², and consumes 350.24 mW of power under typical conditions. The experimental results validate the feasibility of the design for high-performance embedded systems, edge devices and digital signal processing applications.

A double split-ring resonator based tunable filter for frequency detection and monitoring system

2025-10-30T08:31:22+00:00

This paper presents a compact, continuously tunable bandpass filter based on a double split-ring resonator (DSRR) loaded with two varactor diodes for modern wireless communication systems. The proposed design achieves a wide tuning range of 8 - 12 GHz (40% fractional bandwidth) through precise capacitance adjustment, addressing critical limitations of conventional tunable filters in bandwidth, reliability, and control complexity. By leveraging the DSRR's coupled resonance characteristics and independent varactor tuning, the filter maintains stable performance with return loss exceeding 10 dB across a 250-MHz passband. The architecture eliminates mechanical switches, offering improved reliability compared to MEMS/p-i-n diode alternatives while greatly reducing component count compared to traditional multi-bank filters. This work provides a foundation for developing reconfigurable RF front-ends that balance wide tunability, miniaturization, and power efficiency for 5G/6G and cognitive radio applications.

Enhancing performance of GaN MMIC power amplifiers through transistor sizing and bias voltage optimization

2025-10-30T08:35:58+00:00

The paper presents a method for optimizing bias voltage and transistor size to improve power conversion efficiency (PAE) and output power in high-power amplifiers using GaN MMIC technology. Using simulation with the MP2500S transistor model in the WIN NP25 technology library on ADS software, the paper analyzed the effects of different V_GS, V_DS voltage levels and gate sizes on PAE, output power and Gain. The simulation and analysis results are applied to satellite communication uplink amplifiers (14 GHz ÷ 14.5 GHz) with gate width configurations from 2×75 µm to 8×125 µm. The results show that, with appropriate gate width, the selection of bias point in AB mode can significantly improve PAE while still ensuring the required P_out. However, the study also shows that excessive gate width expansion can reduce performance due to the trapping effect, parasitic leakage current and self-heating effect. The paper proposes a method to investigate and select the optimal transistor size and bias voltage in GaN PA design, particularly for Ku-band applications (12÷18 GHz) requiring high power and efficiency. This study contributes to improving the performance of GaN HEMT in power electronics applications.

Outage probability analysis of RIS-assisted RSMA systems over Nakagami-m fading channels

2025-10-30T08:40:14+00:00

In this study, we propose and analyze an RIS-assisted rate-splitting multiple access (RSMA) system. The framework enhances transmission efficiency by simultaneously exploiting both the direct and RIS-reflected links to serve two users. For practical considerations, imperfect successive interference cancellation (SIC) is taken into account. Under the Nakagami-m fading channel model, closed-form expressions for the outage probability (OP) and energy efficiency (EE) are derived and verified through Monte Carlo simulations with a large number of iterations. Furthermore, a comprehensive comparative analysis with a benchmark RIS-assisted non-orthogonal multiple access (NOMA) scheme is conducted. Simulation results show that the proposed system significantly outperforms its NOMA counterpart in terms of OP. The impacts of key system parameters, including the number of RIS reflecting elements, transmit power, carrier frequency, fading severity, and SIC imperfection level, are thoroughly investigated. These findings demonstrate the strong potential of RIS-assisted RSMA as a promising solution to enhance the reliability of future sixth-generation (6G) wireless networks.

Investigation of hybrid detector structures for surveillance radar with correlated target fluctuations

2025-10-30T08:45:25+00:00

This paper investigates the effect of temporally correlated target fluctuations on pulse-train integration efficiency in surveillance radar. A proper complex Gaussian signal model in white noise is assumed. Two hybrid detector structures are synthesized using principal component analysis, where coherent integration is limited to selected subspaces and the remaining components are processed noncoherently. Hybrid-1 restricts coherent integration to the dominant eigen-direction, whereas Hybrid-90 extends it to the subspace capturing 90% of the signal energy. Their performance, measured in terms of detection probability and computational complexity, is evaluated against two widely used reference classes in radar detection: noncoherent energy detection and matched-filter–based coherent integration schemes. The maximum-likelihood detector serves as the optimal benchmark. Monte Carlo simulations across different target fluctuation characteristics and signal-to-noise ratios show that the proposed hybrids achieve a more stable detection probability than conventional schemes while approaching the optimal bound. Hybrid-1, in particular, offers a favorable trade-off between detection reliability and algorithmic complexity. The results provide practical guidelines for surveillance radar design and suggest directions for further research.

Deep residual regression network for underwater acoustic source number estimation

2025-10-30T08:51:44+00:00

Source Number Estimation (SNE) is a crucial task in underwater acoustic array signal processing, as it significantly affects the performance of subsequent algorithms. Traditional methods, such as the Akaike Information Criterion (AIC) and Minimum Description Length (MDL), often perform poorly in challenging underwater environments, especially under low Signal-to-Noise Ratio (SNR) conditions, with a limited number of snapshots and complex noise structures. To tackle these issues, this paper presents an Eigenvalue-based Residual Regression Network (EResNet) designed for robust source number estimation. Comprehensive simulations conducted in various complex noise scenarios have shown that EResNet is notably effective. The results reveal that the proposed model achieves higher accuracy and demonstrates greater robustness compared to the AIC and MDL methods and other baseline neural network architectures.

Impact of Doppler effect on pilot augmentation techniques for channel estimation in UWA-OFDM systems

2025-10-30T08:56:37+00:00

This paper investigates the impact of the Doppler effect on pilot insertion techniques for channel estimation in Underwater Acoustic Orthogonal Frequency Division Multiplexing (UWA-OFDM) systems. Accurate channel estimation is critical for reliable data transmission in underwater acoustic communications, where the propagation environment is highly dynamic and susceptible to Doppler-induced distortions. To address this challenge, several pilot augmentation strategies are evaluated with the objective of enhancing the robustness and accuracy of channel estimation. The study provides a comparative analysis of these techniques under varying Doppler conditions, highlighting their effectiveness in compensating for frequency and time shifts caused by relative motion in the underwater environment. The results demonstrate the selection of appropriate pilot structures to mitigate Doppler effects and improve overall system performance.

Underwater communication system for shallow water environments using FSK modulation

2025-10-30T08:59:09+00:00

Underwater acoustic communication presents significant challenges due to complex environmental factors such as noise, multipath propagation, and time-varying channel conditions. This paper presents an underwater communication system employing Frequency Shift Keying (FSK) modulation with matched filter-based demodulation to mitigate the impact of noise in shallow water environments, suited to Vietnam’s marine area. By leveraging the optimal signal detection capability of the matched filter in noisy and time-varying channels, the system significantly improves communication reliability under harsh underwater conditions. Simulation results demonstrate that the proposed FSK-based system achieves a bit error rate (BER) below 10⁻³ at a signal-to-noise ratio (SNR) exceeding 20 dB. These findings suggest that FSK modulation, combined with matched filter demodulation, is a suitable solution for underwater communication in shallow water scenarios.

Broadband signal source localization using convolutional neural networks combined with coherent signal subspace methods

2025-10-30T09:03:00+00:00

In array signal processing, the Direction Of Arrival (DOA) of the wideband radio frequency signal source plays a key role in fields such as sonar, radar, communications, and electronic warfare. Root Multiple Signal Classification (Root-MUSIC) is one of the good algorithms in the subspace method. This algorithm is based on some assumptions, including the initial angular, but their performance is strongly degraded when these assumptions do not hold. The article presents a novel approach that uses Deep Neural Network (DNN) in combination with Root-MUSIC algorithm for improvement in DOA estimation. Observation samples are divided into distinguishable subspaces. The input autocorrelation matrix is trained by DNN to analyze real values. An estimator based on Super-resolution subspace improving the accuracy of direction-of-arrival estimation is presented and analyzed in the paper. Our simulation results show that wideband signal sources can be calculated in low SNR and limited snapshot conditions.

Hardware-efficient matrix multiplication core optimization for edge AI on FPGA

2025-10-30T09:08:34+00:00

This paper presents an optimization approach for matrix multiplication IP cores on FPGA by transforming convolution operations into matrix multiplications. The proposed method leverages parallel computation combined with simultaneous data loading within the same processing cycle, thereby reducing memory requirements and computational latency. Furthermore, casting the output data from 64-bit to 32-bit effectively shrinks the output buffer, resulting in significant hardware resource savings. Simulation results on ModelSim and Vivado–Vitis demonstrate that the design achieves higher computational efficiency and resource utilization compared to traditional implementations, while maintaining stable processing time. This work contributes to the design of CNN inference accelerators on FPGA for edge AI applications, where resource constraints and power consumption are critical factors..

A deep learning solution for compressed semantic segmentation of LiDAR point cloud maps

2025-10-30T16:53:35+00:00

Navigating Unmanned Aerial Vehicles (UAVs) in Global Navigation Satellite System (GNSS)-denied environments often relies on pre-built Light Detection and Ranging (LiDAR) maps. However, the large memory footprint and high computational cost of these point cloud maps pose significant challenges for resource-constrained UAVs. This paper proposes a deep learning solution using a lightweight, modified RandLA-Net architecture to efficiently compress and semantically segment these maps. Our results demonstrate a significant reduction in model size and memory usage while maintaining competitive segmentation accuracy, presenting a viable solution for real-time, on-board processing on embedded systems.

A solution for preprocessing to select discrete spectral features of propeller-equipped marine targets to enhance passive acoustic direction-finding accuracy

2025-10-30T17:04:42+00:00

The low-frequency discrete spectral components generated by the propeller system of marine targets play a crucial role in signal processing, particularly in the problem of underwater acoustic direction finding. The preprocessing step to select these spectral components helps enhance the signal-to-noise ratio (SNR) of the signal input to direction of arrival (DOA) algorithms. This paper proposes a novel solution to improve the selectivity and accumulation of low-frequency line spectral components, which are key features in the detection and parameter estimation of underwater targets. The proposed method combines several classical signal processing techniques, including low-frequency analysis and recording (LOFAR), detection of envelope modulation on noise (DEMON), and adaptive recursive comb filtering (AR-COMB). This combination enhances the input signal's SNR, providing improved quality for subsequent processing stages, especially DOA estimation. Simulation results demonstrate a significant improvement in the accuracy of the DOA algorithm, suggesting a new approach for detecting and characterizing this type of underwater target.

EB-UNet++: An enhanced crack segmentation network combining EfficientNet-B2 and UNet++ with boundary extraction module

2025-10-30T17:08:51+00:00

Pavement crack detection is a crucial task in intelligent transportation systems and infrastructure maintenance. However, accurate segmentation of cracks remains challenging due to their irregular shapes, low contrast against the background, and varying lighting or surface conditions. In this study, we propose EB-UNet++, a novel deep learning architecture designed to enhance crack segmentation performance. EB-UNet++ integrates the powerful feature encoding capabilities of EfficientNet-B2 into the UNet++ encoder structure, enabling more efficient and robust multi-scale feature extraction. To further refine the crack boundaries and suppress false detections, we incorporate a Boundary Extraction Module into the network. Experimental results on benchmark pavement crack datasets demonstrate that EB-UNet++ outperforms several state-of-the-art models in both segmentation accuracy and boundary delineation, achieving higher IoU and F1-scores. The proposed architecture shows strong potential for practical deployment and scalability in automated road inspection and infrastructure monitoring systems.

Data augmentation for UAV-captured vessel images in maritime surveillance using multimodal language and diffusion models

2025-10-30T17:14:29+00:00

In maritime surveillance, UAV-based vessel detection is essential for ensuring security and safety at sea. However, limited and non-diverse annotated data often restrict model performance in complex maritime environments. This study introduces a novel data augmentation pipeline using multimodal generative models to enhance training datasets with realistic synthetic images. Scene descriptions are automatically generated from UAV imagery using Gemma, a lightweight multimodal language model, and then used to guide FLUX, a text-to-image diffusion model, in creating diverse vessel-centric scenes under varying environmental conditions. A hybrid annotation strategy combines YOLO-World for initial object proposals with manual refinement to ensure label accuracy. The augmented dataset is integrated with the original data to train a vessel detection model. Experiments on the VESSELImg benchmark demonstrate that the proposed approach improves the YOLOv11 detector’s mean average precision (mAP) from 0.775 to 0.805 at IoU thresholds of 0.50:0.95. These results validate the effectiveness of combining multimodal diffusion and language models for domain-specific data synthesis, offering improved generalization and robustness in UAV-based maritime vessel detection.