Physical Communication最新文献

FedDAK: Distribution-aware personalized federated learning with dynamic knowledge distillation FedDAK：基于动态知识蒸馏的分布感知个性化联邦学习

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-24 DOI: 10.1016/j.phycom.2026.103015

Wenlong Lu , Ping Zhang , An Bao

With the rapid advancement of edge intelligence, conventional federated learning (FL) frameworks still struggle to achieve competitive performance under highly heterogeneous and class-imbalanced data distributions. To address these limitations, this paper presents FedDAK, a distribution-aware and adaptive knowledge distillation framework for personalized federated learning. FedDAK enhances both stability and personalization capability through three key designs: dynamic distillation weighting, adaptive rare-class enhancement, and distribution-aware global aggregation. Unlike existing distillation-based FL systems that rely on static or heuristic weighting, FedDAK introduces a KL-divergence–guided dynamic distillation coefficient, enabling each client to automatically regulate the strength of global knowledge constraints according to its divergence from the global data distribution. Furthermore, FedDAK integrates class-level scarcity modeling, assigning increased importance to underrepresented categories to alleviate bias under severe class imbalance. At the global level, FedDAK employs distribution-aware aggregation, reducing the negative influence of highly divergent clients and improving global stability and generalization. Extensive experiments on benchmark datasets demonstrate that FedDAK achieves significantly better personalized performance and global convergence than existing FL baselines under the standard federated learning setting, without requiring the sharing of raw data. The code is available at https://github.com/youmurong50-cmd/fedDAK.

随着边缘智能的快速发展，传统的联邦学习（FL）框架在高度异构和类不平衡的数据分布下仍然难以达到有竞争力的性能。为了解决这些限制，本文提出了FedDAK，一个用于个性化联邦学习的分布感知和自适应知识蒸馏框架。FedDAK通过三个关键设计增强了稳定性和个性化能力：动态蒸馏加权、自适应稀有类增强和分布感知全局聚合。与现有的依赖静态或启发式加权的基于蒸馏的FL系统不同，FedDAK引入了kl -发散引导的动态蒸馏系数，使每个客户端能够根据其与全局数据分布的偏离程度自动调节全局知识约束的强度。此外，FedDAK集成了类级稀缺性模型，增加了代表性不足的类别的重要性，以减轻严重的类不平衡下的偏见。在全球层面，FedDAK采用分布感知聚合，减少了高度分散客户的负面影响，提高了全球稳定性和泛化。在基准数据集上的大量实验表明，在标准联邦学习设置下，FedDAK在不需要共享原始数据的情况下，实现了比现有FL基线更好的个性化性能和全局收敛性。代码可在https://github.com/youmurong50-cmd/fedDAK上获得。

{"title":"FedDAK: Distribution-aware personalized federated learning with dynamic knowledge distillation","authors":"Wenlong Lu , Ping Zhang , An Bao","doi":"10.1016/j.phycom.2026.103015","DOIUrl":"10.1016/j.phycom.2026.103015","url":null,"abstract":"<div><div>With the rapid advancement of edge intelligence, conventional federated learning (FL) frameworks still struggle to achieve competitive performance under highly heterogeneous and class-imbalanced data distributions. To address these limitations, this paper presents FedDAK, a distribution-aware and adaptive knowledge distillation framework for personalized federated learning. FedDAK enhances both stability and personalization capability through three key designs: dynamic distillation weighting, adaptive rare-class enhancement, and distribution-aware global aggregation. Unlike existing distillation-based FL systems that rely on static or heuristic weighting, FedDAK introduces a KL-divergence–guided dynamic distillation coefficient, enabling each client to automatically regulate the strength of global knowledge constraints according to its divergence from the global data distribution. Furthermore, FedDAK integrates class-level scarcity modeling, assigning increased importance to underrepresented categories to alleviate bias under severe class imbalance. At the global level, FedDAK employs distribution-aware aggregation, reducing the negative influence of highly divergent clients and improving global stability and generalization. Extensive experiments on benchmark datasets demonstrate that FedDAK achieves significantly better personalized performance and global convergence than existing FL baselines under the standard federated learning setting, without requiring the sharing of raw data. The code is available at <span><span>https://github.com/youmurong50-cmd/fedDAK</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103015"},"PeriodicalIF":2.2,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146079110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A hybrid GAN and attention-based sparse autoencoder framework for robust end-to-end wireless communication 一种用于端到端无线通信的混合GAN和基于注意力的稀疏自编码器框架

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-22 DOI: 10.1016/j.phycom.2026.103008

Safalata S. Sindal, Y.N. Trivedi

This paper presents a hybrid deep learning framework that integrates a Generative Adversarial Network (GAN) with an Attention-based Sparse Autoencoder (GAN-AAE) for end-to-end wireless communication over Rayleigh fading channels with imperfect channel state information at the receiver (CSIR). Traditional autoencoder models lack the ability to learn underlying signal distributions or correct distortions caused by fading and noise. The proposed GAN-AAE addresses these limitations by using a generator as a learnable channel surrogate to refine encoded signals and an attention mechanism to dynamically prioritize relevant features for improved decoding. The imperfection in the CSI is quantified by a correlation coefficient ρ, where 0 ≤ ρ ≤ 1. Perfect channel knowledge is denoted by ρ = 1, and decreasing values of ρ correspond to increasingly inaccurate CSIR. The model is jointly trained using adversarial and reconstruction losses to enhance its adaptability. Simulation results show that the GAN-AAE framework significantly outperforms conventional Maximum Likelihood Detection and baseline deep and convolutional neural network-based models in terms of bit error rate (BER). The model is evaluated over M-ary phase shift keying (M-PSK) and M-ary Quadrature Amplitude Modulation (M-QAM) with Rayleigh fading channel. At ρ = 0.9 and a signal-to-noise ratio (SNR) of 10 dB, the conventional baseline model achieves a BER of 0.072, whereas the proposed GAN-AAE achieves a lower BER of 0.02404 for Binary phase shift keying (BPSK). A complexity analysis indicates that although the GAN-AAE model introduces some additional computational overhead, the performance gains in reconstruction justify the trade-off. Overall, the GAN-AAE offers a resilient and adaptive solution for end-to-end communication under realistic wireless impairments.

本文提出了一种混合深度学习框架，该框架将生成对抗网络（GAN）与基于注意力的稀疏自编码器（GAN- aae）集成在瑞利衰落信道上，用于端到端无线通信，接收器（CSIR）的信道状态信息不完善。传统的自编码器模型缺乏学习潜在信号分布或纠正由衰落和噪声引起的失真的能力。提出的GAN-AAE通过使用生成器作为可学习的通道代理来改进编码信号，并使用注意机制来动态优先考虑相关特征以改进解码，从而解决了这些限制。CSI的不完全性通过相关系数ρ来量化，其中0 ≤ ρ ≤ 1。完美通道知识用ρ = 1表示，ρ值越小，CSIR越不准确。利用对抗损失和重建损失对模型进行联合训练，增强模型的适应性。仿真结果表明，GAN-AAE框架在误码率（BER）方面明显优于传统的最大似然检测和基于基线深度和卷积神经网络的模型。对该模型进行了基于瑞利衰落信道的M-ary相移键控（M-PSK）和M-ary正交调幅（M-QAM）的评估。在ρ = 0.9，信噪比（SNR）为10 dB时，传统基线模型的误码率为0.072，而GAN-AAE的二进制相移键控（BPSK）误码率较低，为0.02404。复杂性分析表明，尽管GAN-AAE模型引入了一些额外的计算开销，但重构中的性能增益证明了这种权衡是合理的。总体而言，GAN-AAE为实际无线损伤下的端到端通信提供了弹性和自适应解决方案。

{"title":"A hybrid GAN and attention-based sparse autoencoder framework for robust end-to-end wireless communication","authors":"Safalata S. Sindal, Y.N. Trivedi","doi":"10.1016/j.phycom.2026.103008","DOIUrl":"10.1016/j.phycom.2026.103008","url":null,"abstract":"<div><div>This paper presents a hybrid deep learning framework that integrates a Generative Adversarial Network (GAN) with an Attention-based Sparse Autoencoder (GAN-AAE) for end-to-end wireless communication over Rayleigh fading channels with imperfect channel state information at the receiver (CSIR). Traditional autoencoder models lack the ability to learn underlying signal distributions or correct distortions caused by fading and noise. The proposed GAN-AAE addresses these limitations by using a generator as a learnable channel surrogate to refine encoded signals and an attention mechanism to dynamically prioritize relevant features for improved decoding. The imperfection in the CSI is quantified by a correlation coefficient <em>ρ</em>, where 0 ≤ <em>ρ</em> ≤ 1. Perfect channel knowledge is denoted by <em>ρ</em> = 1, and decreasing values of <em>ρ</em> correspond to increasingly inaccurate CSIR. The model is jointly trained using adversarial and reconstruction losses to enhance its adaptability. Simulation results show that the GAN-AAE framework significantly outperforms conventional Maximum Likelihood Detection and baseline deep and convolutional neural network-based models in terms of bit error rate (BER). The model is evaluated over <em>M</em>-ary phase shift keying (<em>M</em>-PSK) and <em>M</em>-ary Quadrature Amplitude Modulation (<em>M</em>-QAM) with Rayleigh fading channel. At <em>ρ</em> = 0.9 and a signal-to-noise ratio (SNR) of 10 dB, the conventional baseline model achieves a BER of 0.072, whereas the proposed GAN-AAE achieves a lower BER of 0.02404 for Binary phase shift keying (BPSK). A complexity analysis indicates that although the GAN-AAE model introduces some additional computational overhead, the performance gains in reconstruction justify the trade-off. Overall, the GAN-AAE offers a resilient and adaptive solution for end-to-end communication under realistic wireless impairments.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103008"},"PeriodicalIF":2.2,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024339","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

MA-PPO driven autonomous decision system for UAV swarms: Integrating semantic parsing and anti-jamming RL control 基于MA-PPO驱动的无人机群自主决策系统：集成语义解析和抗干扰RL控制

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-21 DOI: 10.1016/j.phycom.2026.103016

Yiming Xiang , Han Yang

The operational effectiveness of drone swarms in complex electromagnetic environments is fundamentally limited by autonomous decision-making capabilities, particularly under dynamic interference and stringent real-time constraints. This study develops an autonomous decision-making system centered on an enhanced Multi-Agent Proximal Policy Optimization (MA-PPO) algorithm, with a core focus on joint communication-control optimization. A hierarchical policy network architecture is designed to tightly couple global task planning with local anti-interference control. Crucially, a dynamic interference model is integrated to co-optimize communication power allocation and flight trajectory planning in real-time, thereby enhancing robustness against channel uncertainty and adversarial jamming. Experimental results under -80 dBm interference intensity demonstrate a 9.4% improvement in task completion rate over MADDPG, a communication interruption rate reduced to 7.1% (19.5% of traditional PID methods), and a 107% enhancement in energy efficiency (8.9 tasks/kWh). The primary contributions are threefold: (1) a hierarchical decision architecture that enables deep coupling between planning and interference-aware control; (2) a joint optimization framework that dynamically balances communication quality with motion constraints; (3) quantitative validation in a realistic electromagnetic environment, confirming the engineering feasibility of the proposed approach for reliable swarm operations. This work provides a scalable and robust solution for autonomous drone swarms, advancing the state-of-the-art in physical-layer aware cooperative control.

无人机群在复杂电磁环境中的作战效能从根本上受到自主决策能力的限制，特别是在动态干扰和严格的实时性约束下。本研究开发了一个以增强型多智能体近端策略优化（MA-PPO）算法为中心的自主决策系统，其核心是联合通信控制优化。设计了一种分层策略网络结构，将全局任务规划与局部抗干扰控制紧密耦合。关键是，集成了动态干扰模型，实时优化通信功率分配和飞行轨迹规划，从而增强了对信道不确定性和对抗性干扰的鲁棒性。实验结果表明，在-80 dBm干扰强度下，与madpg相比，任务完成率提高了9.4%，通信中断率降低到7.1%（传统PID方法的19.5%），能源效率提高了107%（8.9个任务/千瓦时）。主要贡献有三个方面：(1)分层决策架构，实现了计划和干扰感知控制之间的深度耦合；(2)动态平衡通信质量和运动约束的联合优化框架；(3)在现实电磁环境中进行定量验证，验证了所提出的方法在可靠的群体作战中的工程可行性。这项工作为自主无人机群提供了一个可扩展和强大的解决方案，推进了物理层感知协同控制的最新技术。

{"title":"MA-PPO driven autonomous decision system for UAV swarms: Integrating semantic parsing and anti-jamming RL control","authors":"Yiming Xiang , Han Yang","doi":"10.1016/j.phycom.2026.103016","DOIUrl":"10.1016/j.phycom.2026.103016","url":null,"abstract":"<div><div>The operational effectiveness of drone swarms in complex electromagnetic environments is fundamentally limited by autonomous decision-making capabilities, particularly under dynamic interference and stringent real-time constraints. This study develops an autonomous decision-making system centered on an enhanced Multi-Agent Proximal Policy Optimization (MA-PPO) algorithm, with a core focus on joint communication-control optimization. A hierarchical policy network architecture is designed to tightly couple global task planning with local anti-interference control. Crucially, a dynamic interference model is integrated to co-optimize communication power allocation and flight trajectory planning in real-time, thereby enhancing robustness against channel uncertainty and adversarial jamming. Experimental results under -80 dBm interference intensity demonstrate a 9.4% improvement in task completion rate over MADDPG, a communication interruption rate reduced to 7.1% (19.5% of traditional PID methods), and a 107% enhancement in energy efficiency (8.9 tasks/kWh). The primary contributions are threefold: (1) a hierarchical decision architecture that enables deep coupling between planning and interference-aware control; (2) a joint optimization framework that dynamically balances communication quality with motion constraints; (3) quantitative validation in a realistic electromagnetic environment, confirming the engineering feasibility of the proposed approach for reliable swarm operations. This work provides a scalable and robust solution for autonomous drone swarms, advancing the state-of-the-art in physical-layer aware cooperative control.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103016"},"PeriodicalIF":2.2,"publicationDate":"2026-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146079109","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Effective degrees of freedom maximization for XL-RIS-assisted near-field communication via hybrid learning-driven optimization 通过混合学习驱动优化实现xml - ris辅助近场通信的有效自由度最大化

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-20 DOI: 10.1016/j.phycom.2026.103012

Yaru Li , Wengen Gao , Yunfei Li , Wenlong Jiang

Extremely Large Reconfigurable Intelligent Surfaces (XL-RIS) have emerged as a transformative technology for controlling electromagnetic propagation in near-field wireless communication. However, optimizing their performance is challenging due to the complex spatial coupling and polarization effects in this regime-physical phenomena that are not fully captured by conventional models and result in intractable high-dimensional optimization problems. This paper proposes a hybrid learning-driven framework for maximizing the Effective Degrees of Freedom (EDoF) of XL-RIS-assisted systems. The proposed framework is grounded in an electromagnetically rigorous dyadic Green’s function-based channel model that accurately captures these critical near-field environment. To tackle the high-dimensional optimization problem efficiently, we introduce a novel method that combines a Multi-Layer Perceptron (MLP) as a fast performance surrogate with a Genetic Algorithm (GA) for global search. Comprehensive simulations demonstrate that the proposed framework achieves superior performance in achievable EDoF and channel capacity compared to existing benchmarks, effectively reveals the saturation behavior of spatial degrees of freedom and highlights the substantial gains enabled by polarization diversity. The results indicate that the integration of precise physical modeling with learning-based optimization offers an efficient and scalable approach for enhancing the performance of near-field XL-RIS.

极大可重构智能表面（XL-RIS）作为一种变革性的技术出现在近场无线通信中，用于控制电磁传播。然而，优化它们的性能是具有挑战性的，因为在这种情况下，复杂的空间耦合和极化效应——传统模型不能完全捕捉到的物理现象，并导致难以处理的高维优化问题。本文提出了一个混合学习驱动框架，用于最大化xml - ris辅助系统的有效自由度（EDoF）。所提出的框架是基于电磁严格的并矢格林函数的信道模型，该模型准确地捕获了这些关键的近场环境。为了有效地解决高维优化问题，我们引入了一种新的方法，将多层感知器（MLP）作为快速性能代理与遗传算法（GA）相结合进行全局搜索。综合仿真表明，与现有基准相比，该框架在可实现的EDoF和信道容量方面具有优越的性能，有效地揭示了空间自由度的饱和行为，并突出了极化分集所带来的实质性收益。结果表明，将精确物理建模与基于学习的优化相结合，为提高近场xml - ris的性能提供了一种高效、可扩展的方法。

{"title":"Effective degrees of freedom maximization for XL-RIS-assisted near-field communication via hybrid learning-driven optimization","authors":"Yaru Li , Wengen Gao , Yunfei Li , Wenlong Jiang","doi":"10.1016/j.phycom.2026.103012","DOIUrl":"10.1016/j.phycom.2026.103012","url":null,"abstract":"<div><div>Extremely Large Reconfigurable Intelligent Surfaces (XL-RIS) have emerged as a transformative technology for controlling electromagnetic propagation in near-field wireless communication. However, optimizing their performance is challenging due to the complex spatial coupling and polarization effects in this regime-physical phenomena that are not fully captured by conventional models and result in intractable high-dimensional optimization problems. This paper proposes a hybrid learning-driven framework for maximizing the Effective Degrees of Freedom (EDoF) of XL-RIS-assisted systems. The proposed framework is grounded in an electromagnetically rigorous dyadic Green’s function-based channel model that accurately captures these critical near-field environment. To tackle the high-dimensional optimization problem efficiently, we introduce a novel method that combines a Multi-Layer Perceptron (MLP) as a fast performance surrogate with a Genetic Algorithm (GA) for global search. Comprehensive simulations demonstrate that the proposed framework achieves superior performance in achievable EDoF and channel capacity compared to existing benchmarks, effectively reveals the saturation behavior of spatial degrees of freedom and highlights the substantial gains enabled by polarization diversity. The results indicate that the integration of precise physical modeling with learning-based optimization offers an efficient and scalable approach for enhancing the performance of near-field XL-RIS.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103012"},"PeriodicalIF":2.2,"publicationDate":"2026-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

FMCW radar implementation on RF sampling transceiver with signal processing techniques for enhanced range accuracy FMCW雷达在射频采样收发机上的实现，采用信号处理技术提高距离精度

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-20 DOI: 10.1016/j.phycom.2026.103013

S. Reshma , Indu Gopan , M.J. Lal , S. Sreelal , M. Vani Devi

In this work, we present the implementation of an improved signal processing algorithm on an FMCW radar system realized using a direct RF sampling transceiver with greater flexibility compared to traditional radars built with several independent analog and digital components. In addition to the sweep bandwidth, the range resolution of an FMCW radar relies on the beat frequency estimation technique used in the signal processing stage. For improved range accuracy, we propose a signal processing algorithm based on the Chirp Z transform (CZT) combined with an inter-bin interpolation technique that outperforms the conventional FFT in beat frequency estimation. To further enhance the accuracy of radar range, the wavelet denoising technique is applied to the beat signal prior to beat frequency estimation. The effectiveness of this algorithm is validated through hardware experiment in radiated mode using the FMCW signal impaired with simulated noisy scenarios, especially phase noise. The range estimation accuracy of the radar system was evaluated based on Root Mean Square Error (RMSE) as the performance metric. From the test results, it was found that the CZT technique with the Jacobsen estimator and wavelet denoising resulted in the least RMSE value and provided the best accuracy in range measurements, even under noisy conditions.

在这项工作中，我们提出了一种改进的信号处理算法在FMCW雷达系统上的实现，该系统使用直接射频采样收发器实现，与使用多个独立模拟和数字组件构建的传统雷达相比，该系统具有更大的灵活性。除了扫描带宽外，FMCW雷达的距离分辨率还依赖于信号处理阶段使用的拍频估计技术。为了提高距离精度，我们提出了一种基于Chirp Z变换（CZT）和帧间插值技术的信号处理算法，该算法在拍频估计方面优于传统的FFT。为了进一步提高雷达距离的精度，在拍频估计之前对拍频信号进行小波去噪处理。通过硬件实验验证了该算法在辐射模式下的有效性，实验中，FMCW信号受到模拟噪声特别是相位噪声的干扰。以均方根误差（RMSE）为性能指标，对雷达系统的距离估计精度进行了评价。从测试结果中可以发现，使用Jacobsen估计器和小波去噪的CZT技术即使在噪声条件下，也能产生最小的RMSE值，并提供最佳的距离测量精度。

{"title":"FMCW radar implementation on RF sampling transceiver with signal processing techniques for enhanced range accuracy","authors":"S. Reshma , Indu Gopan , M.J. Lal , S. Sreelal , M. Vani Devi","doi":"10.1016/j.phycom.2026.103013","DOIUrl":"10.1016/j.phycom.2026.103013","url":null,"abstract":"<div><div>In this work, we present the implementation of an improved signal processing algorithm on an FMCW radar system realized using a direct RF sampling transceiver with greater flexibility compared to traditional radars built with several independent analog and digital components. In addition to the sweep bandwidth, the range resolution of an FMCW radar relies on the beat frequency estimation technique used in the signal processing stage. For improved range accuracy, we propose a signal processing algorithm based on the Chirp Z transform (CZT) combined with an inter-bin interpolation technique that outperforms the conventional FFT in beat frequency estimation. To further enhance the accuracy of radar range, the wavelet denoising technique is applied to the beat signal prior to beat frequency estimation. The effectiveness of this algorithm is validated through hardware experiment in radiated mode using the FMCW signal impaired with simulated noisy scenarios, especially phase noise. The range estimation accuracy of the radar system was evaluated based on Root Mean Square Error (RMSE) as the performance metric. From the test results, it was found that the CZT technique with the Jacobsen estimator and wavelet denoising resulted in the least RMSE value and provided the best accuracy in range measurements, even under noisy conditions.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103013"},"PeriodicalIF":2.2,"publicationDate":"2026-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024416","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Covert underwater acoustic communication: Joint spectral mimicry and soft-Limiting peak suppression 隐蔽水声通信：联合频谱模拟和软限制峰值抑制

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-17 DOI: 10.1016/j.phycom.2026.103009

Ebrahim Raeisian Dashtaki , Ehsan Moradi , Mohammadreza Jalili

The broadcast nature of underwater acoustic channels necessitates low probability of detection communication to secure critical sensor networks. This paper proposes a hardware-aware steganographic transceiver that reconciles high reliability with strict covertness constraints under practical hardware impairments. To ensure reliable transmission amidst severe multipath fading, we adopt an orthogonal frequency division multiplexing framework. However, practical deployment faces challenges from the high peak-to-average power ratio and hardware non-idealities, including in-phase/quadrature imbalance and residual hardware impairments. Our architecture addresses these by disguising the signal via Wenz-based spectral amplitude shaping and random phase scrambling to mimic ambient ocean noise. Furthermore, we introduce a symmetric frequency diversity scheme that transforms in-phase/quadrature imbalance-induced interference into constructive diversity gain and employ a soft-limiting suppression mechanism based on a hyperbolic-tangent profile. This technique smoothly compresses signal peaks to a target of 7 dB, mitigating non-linear distortion while preserving the Gaussian statistical integrity of the cover signal. Simulation results demonstrate that the proposed system can achieve a bit error rate

< 10^{- 2}

across transmission ranges from 100 m to 7 km. The effective data rate is scalable between 3.31 kbps and 52 bps, while maintaining a negligible Kullback-Leibler divergence ( ≈ 0.03) relative to the Gaussian background noise, validating its feasibility for hardware-constrained underwater covert operations.

水声信道的广播性质要求低探测通信概率以保证关键传感器网络的安全。本文提出了一种硬件感知的隐写收发器，该收发器在实际硬件缺陷的情况下能够兼顾高可靠性和严格的隐蔽性约束。为了保证在严重多径衰落情况下的可靠传输，我们采用了正交频分复用框架。然而，实际部署面临着来自高峰值平均功率比和硬件非理想性的挑战，包括同相/正交不平衡和剩余的硬件损伤。我们的架构通过基于wenz的频谱幅度整形和随机相位置乱来模拟环境海洋噪声来掩盖信号，从而解决了这些问题。此外，我们还引入了一种对称频率分集方案，该方案将同相/正交不平衡引起的干扰转换为建设性分集增益，并采用了基于双曲-切线轮廓的软限制抑制机制。该技术平滑地将信号峰值压缩到7 dB的目标，减轻了非线性失真，同时保持了覆盖信号的高斯统计完整性。仿真结果表明，该系统在100 ~ 7 km的传输范围内可以实现10−2的误码率。有效数据速率可在3.31 kbps和52 bps之间扩展，同时相对于高斯背景噪声保持可忽略的kullbackleibler散度（ ≈ 0.03），验证了其在硬件受限的水下隐蔽操作中的可行性。

{"title":"Covert underwater acoustic communication: Joint spectral mimicry and soft-Limiting peak suppression","authors":"Ebrahim Raeisian Dashtaki , Ehsan Moradi , Mohammadreza Jalili","doi":"10.1016/j.phycom.2026.103009","DOIUrl":"10.1016/j.phycom.2026.103009","url":null,"abstract":"<div><div>The broadcast nature of underwater acoustic channels necessitates low probability of detection communication to secure critical sensor networks. This paper proposes a hardware-aware steganographic transceiver that reconciles high reliability with strict covertness constraints under practical hardware impairments. To ensure reliable transmission amidst severe multipath fading, we adopt an orthogonal frequency division multiplexing framework. However, practical deployment faces challenges from the high peak-to-average power ratio and hardware non-idealities, including in-phase/quadrature imbalance and residual hardware impairments. Our architecture addresses these by disguising the signal via Wenz-based spectral amplitude shaping and random phase scrambling to mimic ambient ocean noise. Furthermore, we introduce a symmetric frequency diversity scheme that transforms in-phase/quadrature imbalance-induced interference into constructive diversity gain and employ a soft-limiting suppression mechanism based on a hyperbolic-tangent profile. This technique smoothly compresses signal peaks to a target of 7 dB, mitigating non-linear distortion while preserving the Gaussian statistical integrity of the cover signal. Simulation results demonstrate that the proposed system can achieve a bit error rate <span><math><mrow><mo><</mo><msup><mn>10</mn><mrow><mo>−</mo><mn>2</mn></mrow></msup></mrow></math></span> across transmission ranges from 100 m to 7 km. The effective data rate is scalable between 3.31 kbps and 52 bps, while maintaining a negligible Kullback-Leibler divergence ( ≈ 0.03) relative to the Gaussian background noise, validating its feasibility for hardware-constrained underwater covert operations.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103009"},"PeriodicalIF":2.2,"publicationDate":"2026-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

GMSANet: A hybrid CNN-transformer network for CSI feedback GMSANet：一种用于CSI反馈的混合cnn -变压器网络

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-17 DOI: 10.1016/j.phycom.2026.103010

Wenhong Pan, Junjie Wu, Jiangnan Yuan

In massive multiple-input multiple-output (MIMO) systems operating under the frequency division duplexing (FDD) mode, user equipment (UE) is required to feed back downlink channel state information (CSI) to the base station (BS). As the number of antennas increases, CSI feedback suffers from excessive bandwidth overhead. To reduce the feedback cost and facilitate practical deployment, both convolutional neural network (CNN)-based and Transformer-based methods have achieved remarkable success in CSI feedback. However, CNN-based approaches are inherently limited by convolution operations and struggle to effectively capture global contextual information. Meanwhile, Transformer-based approaches often face insufficient local feature modeling and high computational complexity caused by the self-attention mechanism. To address these limitations, this paper proposes a novel CNN-Transformer hybrid architecture Gated Multi-Scale Additive Attention Network, termed GMSANet. The proposed framework is built upon two key ideas. First, we introduce a new Multi-Scale Additive Attention (MSAA) mechanism that can effectively extract multi-scale features from the CSI matrix while modeling long-range correlations with reduced computational complexity. Second, by employing gated linear units (GLU) as channel mixers, the model captures frequency-domain correlations among adjacent subcarriers, thereby enhancing local modeling capability and improving robustness. Simulation results demonstrate that the proposed network achieves superior CSI reconstruction performance while reducing computational complexity.

在工作在频分双工（FDD）模式下的大规模多输入多输出（MIMO）系统中，用户设备（UE）需要向基站（BS）反馈下行信道状态信息（CSI）。随着天线数量的增加，CSI反馈受到带宽开销过大的困扰。为了降低反馈成本和便于实际部署，基于卷积神经网络（CNN）和基于transformer的方法在CSI反馈中都取得了显著的成功。然而，基于cnn的方法固有地受到卷积操作的限制，并且难以有效地捕获全局上下文信息。同时，基于transformer的方法往往面临局部特征建模不足和自关注机制导致的计算复杂度高的问题。为了解决这些限制，本文提出了一种新颖的CNN-Transformer混合架构门控多尺度加性注意力网络，称为GMSANet。拟议的框架建立在两个关键思想之上。首先，我们引入了一种新的多尺度加性注意（MSAA）机制，该机制可以有效地从CSI矩阵中提取多尺度特征，同时降低了计算复杂度。其次，通过采用门控线性单元（GLU）作为信道混频器，该模型捕获了相邻子载波之间的频域相关性，从而增强了局部建模能力并提高了鲁棒性。仿真结果表明，该网络在降低计算复杂度的同时取得了较好的CSI重构性能。

{"title":"GMSANet: A hybrid CNN-transformer network for CSI feedback","authors":"Wenhong Pan, Junjie Wu, Jiangnan Yuan","doi":"10.1016/j.phycom.2026.103010","DOIUrl":"10.1016/j.phycom.2026.103010","url":null,"abstract":"<div><div>In massive multiple-input multiple-output (MIMO) systems operating under the frequency division duplexing (FDD) mode, user equipment (UE) is required to feed back downlink channel state information (CSI) to the base station (BS). As the number of antennas increases, CSI feedback suffers from excessive bandwidth overhead. To reduce the feedback cost and facilitate practical deployment, both convolutional neural network (CNN)-based and Transformer-based methods have achieved remarkable success in CSI feedback. However, CNN-based approaches are inherently limited by convolution operations and struggle to effectively capture global contextual information. Meanwhile, Transformer-based approaches often face insufficient local feature modeling and high computational complexity caused by the self-attention mechanism. To address these limitations, this paper proposes a novel CNN-Transformer hybrid architecture Gated Multi-Scale Additive Attention Network, termed GMSANet. The proposed framework is built upon two key ideas. First, we introduce a new Multi-Scale Additive Attention (MSAA) mechanism that can effectively extract multi-scale features from the CSI matrix while modeling long-range correlations with reduced computational complexity. Second, by employing gated linear units (GLU) as channel mixers, the model captures frequency-domain correlations among adjacent subcarriers, thereby enhancing local modeling capability and improving robustness. Simulation results demonstrate that the proposed network achieves superior CSI reconstruction performance while reducing computational complexity.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103010"},"PeriodicalIF":2.2,"publicationDate":"2026-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024418","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Investigation of MIMO channel model for ultra supersonic unmanned aerial vehicles 超声速无人机MIMO信道模型研究

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-16 DOI: 10.1016/j.phycom.2026.103005

Lei Xu , Jie Zhou , Rui Liu , Manyan Zhao , Gen-Fu Shao

In this paper, a three-dimensional multi-input multi-output (MIMO) random channel model based on unmanned aerial vehicle (UAV) is proposed to address the communication challenges between supersonic UAVs and ground stations. During the flight of supersonic UAVs, a unique plasma sheath phenomenon is observed. The plasma sheath, located beneath the UAV, initially affects the signal emitted by the UAVs transmitter before it reaches the scatterers or ground receiver. Consequently, we develop a model that incorporates the signal passing through the plasma sheath and couples it with the air-to-ground channel model. The envelope and phase correlation coefficients of the channel, as well as the spatial-temporal frequency correlation function, are derived and described in detail. The statistical characteristics of the channel are examined by varying the incidence angles of the plasma layer and the UAV flight altitude. The results demonstrate that the plasma sheath significantly reduces the channel correlation coefficient within a short time frame, while a higher channel correlation coefficient is observed when the signal vertically impacts the plasma layer.

针对超声速无人机与地面站之间的通信问题，提出了一种基于无人机的三维多输入多输出（MIMO）随机信道模型。在超声速无人机飞行过程中，观察到一种独特的等离子鞘层现象。等离子护套，位于无人机下方，在到达散射器或地面接收器之前，最初影响由无人机发射器发射的信号。因此，我们开发了一个模型，该模型包含了通过等离子体护套的信号，并将其与空对地通道模型耦合。推导并详细描述了信道的包络和相位相关系数以及时空频率相关函数。通过改变等离子体层入射角和无人机飞行高度，考察了通道的统计特性。结果表明，等离子体鞘层在短时间内显著降低了通道相关系数，而当信号垂直撞击等离子体层时，通道相关系数更高。

{"title":"Investigation of MIMO channel model for ultra supersonic unmanned aerial vehicles","authors":"Lei Xu , Jie Zhou , Rui Liu , Manyan Zhao , Gen-Fu Shao","doi":"10.1016/j.phycom.2026.103005","DOIUrl":"10.1016/j.phycom.2026.103005","url":null,"abstract":"<div><div>In this paper, a three-dimensional multi-input multi-output (MIMO) random channel model based on unmanned aerial vehicle (UAV) is proposed to address the communication challenges between supersonic UAVs and ground stations. During the flight of supersonic UAVs, a unique plasma sheath phenomenon is observed. The plasma sheath, located beneath the UAV, initially affects the signal emitted by the UAVs transmitter before it reaches the scatterers or ground receiver. Consequently, we develop a model that incorporates the signal passing through the plasma sheath and couples it with the air-to-ground channel model. The envelope and phase correlation coefficients of the channel, as well as the spatial-temporal frequency correlation function, are derived and described in detail. The statistical characteristics of the channel are examined by varying the incidence angles of the plasma layer and the UAV flight altitude. The results demonstrate that the plasma sheath significantly reduces the channel correlation coefficient within a short time frame, while a higher channel correlation coefficient is observed when the signal vertically impacts the plasma layer.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103005"},"PeriodicalIF":2.2,"publicationDate":"2026-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Measurement-based analysis of RIS-assisted MIMO channel characteristics and capacity at 6 GHz in an outdoor scenario 基于测量的室外场景下6 GHz下ris辅助MIMO信道特性和容量分析

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-15 DOI: 10.1016/j.phycom.2026.103007

Shiyu Zhang , Yuxiang Zhang , Jianhua Zhang , Huiwen Gong , Yutong Sun , Jialin Wang , Tao Jiang , Zheng Hu

Reconfigurable intelligent surfaces (RIS) are a promising technology for future wireless networks. However, the channel characteristics of RIS-assisted multiple-input multiple-output (MIMO) systems remain insufficiently studied, especially in the 6 GHz band. To address this, an outdoor street-corner measurement campaign was conducted using a 32  ×  56 MIMO array and an RIS operating at 6 GHz with 200 MHz bandwidth. The measured power-angle-delay profiles (PADPs) show that RIS reinforcement strengthens reflected paths and increases received power, while reduced root mean square (RMS) delay and angular spreads indicate a mitigation of temporal and angular dispersion. Channel capacity and eigenvalue analyses further reveal that although RIS improves capacity, it slightly limits spatial multiplexing. These findings suggest that the primary advantage of RIS-assisted MIMO lies in enhancing power and link robustness under non-line-of-sight (NLoS) conditions, rather than universally improving multiplexing capability. This emphasizes the need for environment-aware RIS deployment strategies that balance power gain and spatial richness in future wireless systems.

可重构智能表面（RIS）是未来无线网络的一项很有前途的技术。然而，ris辅助多输入多输出（MIMO）系统的信道特性研究仍然不够充分，特别是在6 GHz频段。为了解决这个问题，使用32 × 56 MIMO阵列和6 GHz、200 MHz带宽的RIS进行了室外街角测量活动。测量的功率-角-延迟曲线（PADPs）表明，RIS增强增强了反射路径并增加了接收功率，而减少的均方根（RMS）延迟和角扩散表明时间和角色散得到缓解。信道容量和特征值分析进一步表明，虽然RIS提高了容量，但它略微限制了空间复用。这些发现表明，ris辅助MIMO的主要优势在于增强非视距（NLoS）条件下的功率和链路鲁棒性，而不是普遍提高多路复用能力。这强调了在未来的无线系统中，需要有环境意识的RIS部署策略来平衡功率增益和空间丰富度。

{"title":"Measurement-based analysis of RIS-assisted MIMO channel characteristics and capacity at 6 GHz in an outdoor scenario","authors":"Shiyu Zhang , Yuxiang Zhang , Jianhua Zhang , Huiwen Gong , Yutong Sun , Jialin Wang , Tao Jiang , Zheng Hu","doi":"10.1016/j.phycom.2026.103007","DOIUrl":"10.1016/j.phycom.2026.103007","url":null,"abstract":"<div><div>Reconfigurable intelligent surfaces (RIS) are a promising technology for future wireless networks. However, the channel characteristics of RIS-assisted multiple-input multiple-output (MIMO) systems remain insufficiently studied, especially in the 6 GHz band. To address this, an outdoor street-corner measurement campaign was conducted using a 32  ×  56 MIMO array and an RIS operating at 6 GHz with 200 MHz bandwidth. The measured power-angle-delay profiles (PADPs) show that RIS reinforcement strengthens reflected paths and increases received power, while reduced root mean square (RMS) delay and angular spreads indicate a mitigation of temporal and angular dispersion. Channel capacity and eigenvalue analyses further reveal that although RIS improves capacity, it slightly limits spatial multiplexing. These findings suggest that the primary advantage of RIS-assisted MIMO lies in enhancing power and link robustness under non-line-of-sight (NLoS) conditions, rather than universally improving multiplexing capability. This emphasizes the need for environment-aware RIS deployment strategies that balance power gain and spatial richness in future wireless systems.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 103007"},"PeriodicalIF":2.2,"publicationDate":"2026-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024414","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhanced AI–simulink hybrid framework for low-latency interference mitigation and performance optimization in 5 G RF receivers 增强的AI-simulink混合框架，用于5g射频接收器的低延迟干扰缓解和性能优化

IF 2.2 4区计算机科学 Q3 ENGINEERING, ELECTRICAL & ELECTRONIC

Physical Communication

Pub Date : 2026-01-13 DOI: 10.1016/j.phycom.2026.102995

Shatha Kareem Abbas, Osman Nuri Uçan, AlSeddiq Oday

'The rapid expansion of 5 G wireless communication systems has accelerated the need for intelligent receiver architectures capable of adaptively mitigating interference while preserving signal integrity. This study presents an enhanced hybrid AI–Simulink framework that integrates deep learning-based estimation with a parametric Simulink RF signal chain to detect, predict, and suppress jamming and noise-induced distortion in 5 G RF receivers. A processed dataset of approximately 250,000 labeled signal samples was generated from Simulink simulations and used to train and validate the proposed model. The framework demonstrates substantial improvements in signal-to-noise ratio (SNR) and root mean square error (RMSE) when compared with traditional filtering and baseline machine learning approaches. Experimental results show an SNR enhancement of over +14 dB and consistently low error metrics across multiple interference power levels and frequency configurations. The proposed architecture maintains a compact computational footprint (≈8.4 MB) and supports low-latency inference suitable for integration into hardware-accelerated or embedded execution environments. These outcomes confirm the potential of the proposed hybrid approach as a precise and efficient solution for AI-assisted interference mitigation in 5 G receivers, while also outlining future directions toward over-the-air validation and FPGA-based deployment.

“5g无线通信系统的快速扩展加速了对智能接收器架构的需求，该架构能够自适应地减轻干扰，同时保持信号完整性。本研究提出了一个增强型混合AI-Simulink框架，该框架将基于深度学习的估计与参数化Simulink射频信号链集成在一起，以检测、预测和抑制5g射频接收器中的干扰和噪声引起的失真。从Simulink模拟中生成了大约25万个标记信号样本的处理数据集，并用于训练和验证所提出的模型。与传统的滤波和基线机器学习方法相比，该框架在信噪比（SNR）和均方根误差（RMSE）方面有了实质性的改进。实验结果表明，在多种干扰功率水平和频率配置下，信噪比增强超过+14 dB，误差指标始终较低。所提出的体系结构保持紧凑的计算占用（≈8.4 MB），并支持适合集成到硬件加速或嵌入式执行环境中的低延迟推理。这些结果证实了所提出的混合方法作为5g接收器中人工智能辅助干扰缓解的精确有效解决方案的潜力，同时也概述了空中验证和基于fpga的部署的未来方向。

{"title":"Enhanced AI–simulink hybrid framework for low-latency interference mitigation and performance optimization in 5 G RF receivers","authors":"Shatha Kareem Abbas, Osman Nuri Uçan, AlSeddiq Oday","doi":"10.1016/j.phycom.2026.102995","DOIUrl":"10.1016/j.phycom.2026.102995","url":null,"abstract":"<div><div>'The rapid expansion of 5 G wireless communication systems has accelerated the need for intelligent receiver architectures capable of adaptively mitigating interference while preserving signal integrity. This study presents an enhanced hybrid AI–Simulink framework that integrates deep learning-based estimation with a parametric Simulink RF signal chain to detect, predict, and suppress jamming and noise-induced distortion in 5 G RF receivers. A processed dataset of approximately 250,000 labeled signal samples was generated from Simulink simulations and used to train and validate the proposed model. The framework demonstrates substantial improvements in signal-to-noise ratio (SNR) and root mean square error (RMSE) when compared with traditional filtering and baseline machine learning approaches. Experimental results show an SNR enhancement of over +14 dB and consistently low error metrics across multiple interference power levels and frequency configurations. The proposed architecture maintains a compact computational footprint (≈8.4 MB) and supports low-latency inference suitable for integration into hardware-accelerated or embedded execution environments. These outcomes confirm the potential of the proposed hybrid approach as a precise and efficient solution for AI-assisted interference mitigation in 5 G receivers, while also outlining future directions toward over-the-air validation and FPGA-based deployment.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"75 ","pages":"Article 102995"},"PeriodicalIF":2.2,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146024415","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0