Pub Date : 2025-11-22DOI: 10.1016/j.phycom.2025.102929
Maryam Najimi
The cognitive radio network (CRN) consists of multiple secondary users (SUs) and primary users (PUs). In fact, the spectrum of each PU is shared by the several SUs without or with determined interference with PUs (i.e., licensed users), which enhances spectral efficiency. Spectrum sensing is one of the solutions to improve the secondary network (SN) performance and primary network (PN) in the opportunistic spectrum access of the CRN. In this case, by assistance of intelligent reflecting surface (IRS), the received signal strength and therefore, the communication of SNs and PNs are enhanced according to the spectrum sensing results. On the other words, IRS utilization leads to the coverage, data rates and power efficiency improvement in cognitive radio networks. In this work, we consider multi IRS-assisted CRN with multiple SNs, PNs and IRSs. The main goal of this study is maximizing the SNs’ sum rates by obtaining the user-IRS association and optimizing secondary transmitters (STs) beamforming, sensing duration of each PU and three stages IRS phase shifts matrices for each SN and each IRS, respectively while average achievable rate requirement of PUs, the STs’ maximum transmit power and the detection performance constraints are maintained. For solving the non-convex optimization problem, an iterative algorithm is utilized based on the Gradient Descent framework. Simulation results verify that the user-IRS assignment leads to more spectral efficiencies of both primary and secondary transmissions in the proposed IRS-assisted CRN compared to the benchmark algorithms.
认知无线网络（CRN）由多个secondary user （su）和primary user （pu）组成。实际上，每个PU的频谱是由几个su共享的，没有或确定与PU（即授权用户）的干扰，从而提高了频谱效率。在CRN的机会频谱接入中，频谱感知是提高从网（SN）和主网（PN）性能的解决方案之一。在这种情况下，通过智能反射面（IRS）的辅助，根据频谱感知结果增强接收信号强度，从而增强SNs和pn的通信。换句话说，IRS的使用导致了认知无线网络的覆盖、数据速率和功率效率的提高。在这项工作中，我们考虑了多个irs辅助CRN与多个SNs， pn和irs。本研究的主要目标是通过获取用户-IRS关联和优化二次发射机（STs）波束形成、每个PU的感知持续时间和每个SN和每个IRS的三级IRS相移矩阵来最大化SNs的和速率，同时保持PU的平均可达速率要求、STs的最大发射功率和检测性能约束。对于非凸优化问题，采用了基于梯度下降框架的迭代算法。仿真结果表明，与基准算法相比，用户- irs分配导致所提出的irs辅助CRN中主传输和二次传输的频谱效率更高。
{"title":"IRS-user association, spectrum sensing and primary-secondary transmission in multi-IRS assisted multi- band cognitive radio networks","authors":"Maryam Najimi","doi":"10.1016/j.phycom.2025.102929","DOIUrl":"10.1016/j.phycom.2025.102929","url":null,"abstract":"<div><div>The cognitive radio network (CRN) consists of multiple secondary users (SUs) and primary users (PUs). In fact, the spectrum of each PU is shared by the several SUs without or with determined interference with PUs (i.e., licensed users), which enhances spectral efficiency. Spectrum sensing is one of the solutions to improve the secondary network (SN) performance and primary network (PN) in the opportunistic spectrum access of the CRN. In this case, by assistance of intelligent reflecting surface (IRS), the received signal strength and therefore, the communication of SNs and PNs are enhanced according to the spectrum sensing results. On the other words, IRS utilization leads to the coverage, data rates and power efficiency improvement in cognitive radio networks. In this work, we consider multi IRS-assisted CRN with multiple SNs, PNs and IRSs. The main goal of this study is maximizing the SNs’ sum rates by obtaining the user-IRS association and optimizing secondary transmitters (STs) beamforming, sensing duration of each PU and three stages IRS phase shifts matrices for each SN and each IRS, respectively while average achievable rate requirement of PUs, the STs’ maximum transmit power and the detection performance constraints are maintained. For solving the non-convex optimization problem, an iterative algorithm is utilized based on the Gradient Descent framework. Simulation results verify that the user-IRS assignment leads to more spectral efficiencies of both primary and secondary transmissions in the proposed IRS-assisted CRN compared to the benchmark algorithms.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"74 ","pages":"Article 102929"},"PeriodicalIF":2.2,"publicationDate":"2025-11-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145694730","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}
Pub Date : 2025-11-21DOI: 10.1016/j.phycom.2025.102925
Xinru Song , Yi Wu , Jun Zhang , Qi Zhang
This paper investigates the unmanned aerial vehicle (UAV)-borne simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS)-assisted non-orthogonal multiple access system under the influence of UAV jitter. Since jitter leads to beam misalignment and hence significant user ergodic rate degradation, this paper formulates the weighted ergodic sum-rate maximization optimization problem. To cope with the challenges posed by coupled variables, non-convex constraints and jitter characteristics, this paper proposes a scheme that combines jitter-free optimization with jitter adaptive optimization. The first stage under jitter-free assumptions optimizes UAV hovering positions, base station beamforming, and power allocation strategies via linear search and closed-form solutions, reducing computational complexity while ensuring baseline performance. The second stage adaptively adjusts STAR-RIS phases based on real-time jitter states, employing wide beams to compensate for beam misalignment. Simulation results demonstrate that the proposed scheme dynamically adjusts beamwidth based on jitter intensity, effectively mitigating rate degradation from jitter and enhancing system robustness.
{"title":"Adaptive robust optimization for UAV-borne STAR-RIS-assisted NOMA system: Mitigating jitter effects on ergodic rate","authors":"Xinru Song , Yi Wu , Jun Zhang , Qi Zhang","doi":"10.1016/j.phycom.2025.102925","DOIUrl":"10.1016/j.phycom.2025.102925","url":null,"abstract":"<div><div>This paper investigates the unmanned aerial vehicle (UAV)-borne simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS)-assisted non-orthogonal multiple access system under the influence of UAV jitter. Since jitter leads to beam misalignment and hence significant user ergodic rate degradation, this paper formulates the weighted ergodic sum-rate maximization optimization problem. To cope with the challenges posed by coupled variables, non-convex constraints and jitter characteristics, this paper proposes a scheme that combines jitter-free optimization with jitter adaptive optimization. The first stage under jitter-free assumptions optimizes UAV hovering positions, base station beamforming, and power allocation strategies via linear search and closed-form solutions, reducing computational complexity while ensuring baseline performance. The second stage adaptively adjusts STAR-RIS phases based on real-time jitter states, employing wide beams to compensate for beam misalignment. Simulation results demonstrate that the proposed scheme dynamically adjusts beamwidth based on jitter intensity, effectively mitigating rate degradation from jitter and enhancing system robustness.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"74 ","pages":"Article 102925"},"PeriodicalIF":2.2,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145625381","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}
Pub Date : 2025-11-20DOI: 10.1016/j.phycom.2025.102920
Guoxin Zhou , Ting Liu , Zonghui Yang , Kai Zhang , Yang Xu
This paper presents a novel cooperative search framework for multiple Autonomous Underwater Vehicles (AUVs) operating in uncertain marine environments. Unlike most existing works that focus on static or aerial scenarios, this study targets the dynamic and probabilistic nature of underwater target search tasks, where environmental uncertainty, communication constraints, and ocean currents significantly affect performance. A Cooperative Pigeon-Inspired Optimization (CPIO) algorithm is proposed to improve the global search capabilities and convergence stability of traditional PIO. The CPIO integrates chaotic initialization, bounded velocity correction, and elite retention mechanisms. In addition, an environmental modeling framework is designed based on a Gaussian target probability map, a certainty-aware information graph, and a digital pheromone mechanism, enabling collaborative, efficient, and non-redundant exploration. Extensive simulations under realistic marine constraints demonstrate that the proposed method outperforms several advanced bio-inspired algorithms, including IWOA, SHHO and DWOLF, in terms of search accuracy, coverage rate, and robustness.
{"title":"Uncertainty-aware cooperative target search in marine environments: An enhanced Pigeon-Inspired Optimization for multi-AUV systems","authors":"Guoxin Zhou , Ting Liu , Zonghui Yang , Kai Zhang , Yang Xu","doi":"10.1016/j.phycom.2025.102920","DOIUrl":"10.1016/j.phycom.2025.102920","url":null,"abstract":"<div><div>This paper presents a novel cooperative search framework for multiple Autonomous Underwater Vehicles (AUVs) operating in uncertain marine environments. Unlike most existing works that focus on static or aerial scenarios, this study targets the dynamic and probabilistic nature of underwater target search tasks, where environmental uncertainty, communication constraints, and ocean currents significantly affect performance. A Cooperative Pigeon-Inspired Optimization (CPIO) algorithm is proposed to improve the global search capabilities and convergence stability of traditional PIO. The CPIO integrates chaotic initialization, bounded velocity correction, and elite retention mechanisms. In addition, an environmental modeling framework is designed based on a Gaussian target probability map, a certainty-aware information graph, and a digital pheromone mechanism, enabling collaborative, efficient, and non-redundant exploration. Extensive simulations under realistic marine constraints demonstrate that the proposed method outperforms several advanced bio-inspired algorithms, including IWOA, SHHO and DWOLF, in terms of search accuracy, coverage rate, and robustness.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102920"},"PeriodicalIF":2.2,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145579002","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}
Pub Date : 2025-11-20DOI: 10.1016/j.phycom.2025.102924
Zihao Pan, Daoxing Guo, Bangning Zhang, Pan Zhen, Ning Wang, Heng Wang
Robust Adaptive Beamforming (RAB) based on interference plus noise covariance matrix (INCM) reconstruction suffers significant performance degradation in uncalibrated arrays, where each sensor has unknown gain-phase errors. In this paper, we propose an INCM reconstruction-based RAB framework, where the reconstructed INCM is corrected by the sparsity-based array calibration. First, we present the sparse representation of uncalibrated array output and integrate the gain-phase errors into the model. Then, we perform spatial sparse reconstruction using sparse Bayesian learning (SBL), and the gain-phase errors and interference DOA can be obtained. The accuracy of estimated results can be guaranteed because of the implicit high power of interference, and the estimated results are directly used to reconstruct INCM. Finally, the INCM reconstruction-based RAB is proposed. Note that the potential high power of interference is applicable in practice. By using the proposal on simulation experiments, we observed significant improvements over the compared methods in uncalibrated arrays.
{"title":"Robust adaptive beamforming in uncalibrated arrays: Sparsity-based array calibration for covariance matrix reconstruction","authors":"Zihao Pan, Daoxing Guo, Bangning Zhang, Pan Zhen, Ning Wang, Heng Wang","doi":"10.1016/j.phycom.2025.102924","DOIUrl":"10.1016/j.phycom.2025.102924","url":null,"abstract":"<div><div>Robust Adaptive Beamforming (RAB) based on interference plus noise covariance matrix (INCM) reconstruction suffers significant performance degradation in uncalibrated arrays, where each sensor has unknown gain-phase errors. In this paper, we propose an INCM reconstruction-based RAB framework, where the reconstructed INCM is corrected by the sparsity-based array calibration. First, we present the sparse representation of uncalibrated array output and integrate the gain-phase errors into the model. Then, we perform spatial sparse reconstruction using sparse Bayesian learning (SBL), and the gain-phase errors and interference DOA can be obtained. The accuracy of estimated results can be guaranteed because of the implicit high power of interference, and the estimated results are directly used to reconstruct INCM. Finally, the INCM reconstruction-based RAB is proposed. Note that the potential high power of interference is applicable in practice. By using the proposal on simulation experiments, we observed significant improvements over the compared methods in uncalibrated arrays.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"74 ","pages":"Article 102924"},"PeriodicalIF":2.2,"publicationDate":"2025-11-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145584505","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}
Pub Date : 2025-11-17DOI: 10.1016/j.phycom.2025.102897
Lan Xu , Zhongqiang Luo , Chengyang Kang
Conventional visual or radar-based drone detection systems often fail under adverse conditions such as low light, adverse weather, or dense urban environments, and are further limited by high deployment costs and privacy concerns. To address these challenges, we explore the underutilized potential of temporal audio signals for robust, low-cost, and privacy-preserving UAV identification. We propose Light Temporal-Frequency Mamba (LTFM), a novel lightweight architecture that, for the first time, enables simultaneous drone classification and 3D trajectory prediction using only acoustic input. Our key innovation lies in a multi-scale time-frequency fusion mechanism integrated with a hybrid convolutional-recurrent structure, which captures complex acoustic dynamics with minimal computational overhead. A knowledge distillation framework further enhances the compact model’s discriminative power without sacrificing efficiency. Evaluated on a large-scale multimodal dataset, LTFM achieves over 95% classification accuracy while reducing model size and computational cost by more than 60%, and accelerating inference by 1.5 compared to the state-of-the-art TFMamba. It also delivers consistent and precise 3D trajectory estimation, particularly in the X–Y plane. These advancements make LTFM a highly efficient and scalable solution for real-time, edge-based drone monitoring.
{"title":"LTFM: A lightweight model for unmanned aerial vehicle classification and trajectory prediction based on temporal audio","authors":"Lan Xu , Zhongqiang Luo , Chengyang Kang","doi":"10.1016/j.phycom.2025.102897","DOIUrl":"10.1016/j.phycom.2025.102897","url":null,"abstract":"<div><div>Conventional visual or radar-based drone detection systems often fail under adverse conditions such as low light, adverse weather, or dense urban environments, and are further limited by high deployment costs and privacy concerns. To address these challenges, we explore the underutilized potential of temporal audio signals for robust, low-cost, and privacy-preserving UAV identification. We propose Light Temporal-Frequency Mamba (LTFM), a novel lightweight architecture that, for the first time, enables simultaneous drone classification and 3D trajectory prediction using only acoustic input. Our key innovation lies in a multi-scale time-frequency fusion mechanism integrated with a hybrid convolutional-recurrent structure, which captures complex acoustic dynamics with minimal computational overhead. A knowledge distillation framework further enhances the compact model’s discriminative power without sacrificing efficiency. Evaluated on a large-scale multimodal dataset, LTFM achieves over 95% classification accuracy while reducing model size and computational cost by more than 60%, and accelerating inference by 1.5<span><math><mo>×</mo></math></span> compared to the state-of-the-art TFMamba. It also delivers consistent and precise 3D trajectory estimation, particularly in the X–Y plane. These advancements make LTFM a highly efficient and scalable solution for real-time, edge-based drone monitoring.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102897"},"PeriodicalIF":2.2,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145579005","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}
Pub Date : 2025-11-17DOI: 10.1016/j.phycom.2025.102914
Ruizheng Chen, Yan Guo, Jianyu Wei, Junhui Huang, Jiawei Yi
As unmanned aerial vehicles (UAVs) are widely applied in open environments, multiple security risks are exposed during their information transmission process. To address this issue, this paper proposes a blockchain-assisted lightweight secure authentication and key agreement protocol for UAVs. The protocol implements access control through blockchain smart contracts, allowing only authorized entities to access secret information stored on the chain. It adopts hash functions to construct a dynamic pseudonym mechanism, enabling dynamic hiding and updating of UAV identities. It also relies on hash operations to ensure the integrity of authentication messages. Moreover, the protocol combines lightweight symmetric encryption and decryption algorithms with XOR functions to guarantee data confidentiality during authentication. It integrates Physical Unclonable Function (PUF) with fuzzy extractor, effectively resolving the inherent noise sensitivity of PUF. Verified by Mao-Boyd logic, the AVISPA tool, and informal security analysis, the protocol can effectively resist various attacks. Performance evaluations demonstrate that it achieves the best comprehensive performance in computational efficiency, communication overhead, and other aspects, providing secure and reliable communication support for resource-constrained UAV networks.
{"title":"Blockchain-assisted distributed lightweight anonymous two-way authentication protocol for UAV-UAV communication","authors":"Ruizheng Chen, Yan Guo, Jianyu Wei, Junhui Huang, Jiawei Yi","doi":"10.1016/j.phycom.2025.102914","DOIUrl":"10.1016/j.phycom.2025.102914","url":null,"abstract":"<div><div>As unmanned aerial vehicles (UAVs) are widely applied in open environments, multiple security risks are exposed during their information transmission process. To address this issue, this paper proposes a blockchain-assisted lightweight secure authentication and key agreement protocol for UAVs. The protocol implements access control through blockchain smart contracts, allowing only authorized entities to access secret information stored on the chain. It adopts hash functions to construct a dynamic pseudonym mechanism, enabling dynamic hiding and updating of UAV identities. It also relies on hash operations to ensure the integrity of authentication messages. Moreover, the protocol combines lightweight symmetric encryption and decryption algorithms with XOR functions to guarantee data confidentiality during authentication. It integrates Physical Unclonable Function (PUF) with fuzzy extractor, effectively resolving the inherent noise sensitivity of PUF. Verified by Mao-Boyd logic, the AVISPA tool, and informal security analysis, the protocol can effectively resist various attacks. Performance evaluations demonstrate that it achieves the best comprehensive performance in computational efficiency, communication overhead, and other aspects, providing secure and reliable communication support for resource-constrained UAV networks.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102914"},"PeriodicalIF":2.2,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145579003","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}
Orthogonal time sequency multiplexing (OTSM), a recently proposed novel modulation technique, outperforms orthogonal frequency division multiplexing (OFDM) in high mobility and doubly-spread channel conditions. Intelligent reflecting surface (IRS) gains attention as a potential technology in wireless communication research. By adaptively adjusting the phases of the reflecting elements, the IRS can significantly modify the channel conditions. In this paper, we integrate OTSM with IRS to further improve the performance of the OTSM system. We derive the input–output relationship in both the time domain and the delay sequency domain. Then, we present a matched filter Gauss–Seidel method by leveraging zero-padding to reduce time domain interference. Additionally, using a zero-padding pilot approach, we estimate channel parameters and compute their mean square error. Findings show improved BER performance with more reflecting elements, fading parameters, and lower speed, but degradation with higher order modulation. These contributions would enhance IRS-assisted OTSM communication systems, particularly in high-mobility scenarios, benefiting the advancement of wireless technologies, 5G, and beyond communication systems.
{"title":"Channel estimation and data detection for IRS-Aided SISO OTSM communication system under high mobility scenarios","authors":"Sanjeet Kumar Bhagat , Sapta Girish Neelam , P.R. Sahu","doi":"10.1016/j.phycom.2025.102922","DOIUrl":"10.1016/j.phycom.2025.102922","url":null,"abstract":"<div><div>Orthogonal time sequency multiplexing (OTSM), a recently proposed novel modulation technique, outperforms orthogonal frequency division multiplexing (OFDM) in high mobility and doubly-spread channel conditions. Intelligent reflecting surface (IRS) gains attention as a potential technology in wireless communication research. By adaptively adjusting the phases of the reflecting elements, the IRS can significantly modify the channel conditions. In this paper, we integrate OTSM with IRS to further improve the performance of the OTSM system. We derive the input–output relationship in both the time domain and the delay sequency domain. Then, we present a matched filter Gauss–Seidel method by leveraging zero-padding to reduce time domain interference. Additionally, using a zero-padding pilot approach, we estimate channel parameters and compute their mean square error. Findings show improved BER performance with more reflecting elements, fading parameters, and lower speed, but degradation with higher order modulation. These contributions would enhance IRS-assisted OTSM communication systems, particularly in high-mobility scenarios, benefiting the advancement of wireless technologies, 5G, and beyond communication systems.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102922"},"PeriodicalIF":2.2,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145579004","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}
Pub Date : 2025-11-12DOI: 10.1016/j.phycom.2025.102921
Ken Long, Qianwen Bai, Guoxu Xia
This paper proposes an autoencoder (AE)-based end-to-end (E2E) learning framework for multiple-input multiple-output (MIMO) systems that integrates adaptive modulation and dynamic power control under a unified multi-objective design. Unlike existing AE-based schemes that optimize constellation shaping or detection under fixed modulation and power settings, the proposed framework jointly learns the modulation order, transmit constellation, and power allocation strategy under the assumption of known channel state information (CSI). A differentiable Lagrangian-based power control mechanism is embedded during training to ensure compliance with power constraints, while an adaptive modulation selection strategy during inference maximizes throughput subject to a target bit error rate (BER). While each channel is quasi-static per transmission, SNR and fading variations enable generalization across channel conditions. Simulation results demonstrate that the proposed method significantly outperforms fixed-modulation baselines in terms of both spectral efficiency and BER across varying antenna configurations.
{"title":"Learning-based MIMO adaptive modulation: Dynamic power control and non-regular constellation optimization","authors":"Ken Long, Qianwen Bai, Guoxu Xia","doi":"10.1016/j.phycom.2025.102921","DOIUrl":"10.1016/j.phycom.2025.102921","url":null,"abstract":"<div><div>This paper proposes an autoencoder (AE)-based end-to-end (E2E) learning framework for multiple-input multiple-output (MIMO) systems that integrates adaptive modulation and dynamic power control under a unified multi-objective design. Unlike existing AE-based schemes that optimize constellation shaping or detection under fixed modulation and power settings, the proposed framework jointly learns the modulation order, transmit constellation, and power allocation strategy under the assumption of known channel state information (CSI). A differentiable Lagrangian-based power control mechanism is embedded during training to ensure compliance with power constraints, while an adaptive modulation selection strategy during inference maximizes throughput subject to a target bit error rate (BER). While each channel is quasi-static per transmission, SNR and fading variations enable generalization across channel conditions. Simulation results demonstrate that the proposed method significantly outperforms fixed-modulation baselines in terms of both spectral efficiency and BER across varying antenna configurations.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102921"},"PeriodicalIF":2.2,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145528361","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}
Pub Date : 2025-11-12DOI: 10.1016/j.phycom.2025.102919
Girma G. Amlaku , Fikreselam G. Mengistu , Pushparaghavan Annamalai
The persistent need for higher data rates and greater capacity has made it necessary to use millimeter wave (mmWave) signals, where larger, uncongested bandwidths are available. However, mmWave signals experience severe atmospheric absorption and higher attenuation from obstacles compared to lower-frequency signals. This necessitates differentiation of indoor and outdoor scenarios resulting in higher deployment and maintenance costs, reduced user quality of experience and diminished operational flexibility. These challenges could be reduced by realizing a unified, seamless wireless network architecture. The foundation of unified and seamless wireless communication systems lies in low-loss outdoor-to-indoor (O2I) signal propagation. To address this challenge, we propose an aerial reconfigurable intelligent surface (ARIS) framework, which synergistically combines unmanned aerial vehicles (UAVs) and reflective RIS technology to enhance mmWave O2I communications. Furthermore, we introduce two distinct algorithms to optimize ARIS performance: first, a search-based method for exhaustive solution space exploration, and second, a deep reinforcement learning (DRL)-based approach that adaptively learns optimal configurations while offering scalability and robustness to dynamic environmental conditions. Together, these contributions enable significant gains in spectral efficiency and link reliability without requiring structural modifications to existing infrastructure. To show the potential of the proposed approach and algorithms, we simulated a wireless communication system consisting of an outdoor mmWave base station operating at 28 GHz and an indoor receiver which resides in a studio apartment. With the search-based algorithm, performance improvement in terms of spectral efficiency is simulated for four different indoor receiver locations, which exhibit maximum improvements from 0.2 bps/Hz to 14.4 bps/Hz, 0.5 bps/Hz to 13.8 bps/Hz, 10.7 bps/Hz to 14.2 bps/Hz and 0.2 bps/Hz to 8.1 bps/Hz, comparing scenarios without an ARIS to those with an optimally deployed ARIS. The proposed DRL-based approach achieves performance comparable to the search-based method, without requiring an exhaustive search of the solution space. Furthermore, it offers enhanced adaptability to dynamic environmental changes, underscoring its practical utility in real-world deployments.
{"title":"Aerial reconfigurable intelligent surface-assisted outdoor-to-indoor mmWave communications","authors":"Girma G. Amlaku , Fikreselam G. Mengistu , Pushparaghavan Annamalai","doi":"10.1016/j.phycom.2025.102919","DOIUrl":"10.1016/j.phycom.2025.102919","url":null,"abstract":"<div><div>The persistent need for higher data rates and greater capacity has made it necessary to use millimeter wave (mmWave) signals, where larger, uncongested bandwidths are available. However, mmWave signals experience severe atmospheric absorption and higher attenuation from obstacles compared to lower-frequency signals. This necessitates differentiation of indoor and outdoor scenarios resulting in higher deployment and maintenance costs, reduced user quality of experience and diminished operational flexibility. These challenges could be reduced by realizing a unified, seamless wireless network architecture. The foundation of unified and seamless wireless communication systems lies in low-loss outdoor-to-indoor (O2I) signal propagation. To address this challenge, we propose an aerial reconfigurable intelligent surface (ARIS) framework, which synergistically combines unmanned aerial vehicles (UAVs) and reflective RIS technology to enhance mmWave O2I communications. Furthermore, we introduce two distinct algorithms to optimize ARIS performance: first, a search-based method for exhaustive solution space exploration, and second, a deep reinforcement learning (DRL)-based approach that adaptively learns optimal configurations while offering scalability and robustness to dynamic environmental conditions. Together, these contributions enable significant gains in spectral efficiency and link reliability without requiring structural modifications to existing infrastructure. To show the potential of the proposed approach and algorithms, we simulated a wireless communication system consisting of an outdoor mmWave base station operating at 28 GHz and an indoor receiver which resides in a studio apartment. With the search-based algorithm, performance improvement in terms of spectral efficiency is simulated for four different indoor receiver locations, which exhibit maximum improvements from 0.2 bps/Hz to 14.4 bps/Hz, 0.5 bps/Hz to 13.8 bps/Hz, 10.7 bps/Hz to 14.2 bps/Hz and 0.2 bps/Hz to 8.1 bps/Hz, comparing scenarios without an ARIS to those with an optimally deployed ARIS. The proposed DRL-based approach achieves performance comparable to the search-based method, without requiring an exhaustive search of the solution space. Furthermore, it offers enhanced adaptability to dynamic environmental changes, underscoring its practical utility in real-world deployments.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102919"},"PeriodicalIF":2.2,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145579007","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}
The burgeoning Internet of Things (IoT) landscape demands innovative solutions for massive connectivity and energy efficiency, where backscatter communication excels in enabling low-power, cost-effective passive device interactions, and non-orthogonal multiple access (NOMA) boosts spectral efficiency via power-domain multiplexing. The synergy of these technologies in backscatter NOMA systems addresses key IoT challenges but introduces complexities such as inter-user interference, eavesdropping vulnerabilities, and residual SIC artifacts. This paper examines a backscatter NOMA system featuring two NOMA user groups, i.e., near-end and far-end, an embedded backscatter device (BD), a dual-antenna relay, and an eavesdropper, operating under a time-sharing mechanism with Rayleigh fading and AWGN. We derive closed-form expressions for outage probability (OP), asymptotic OP and intercept probability (IP) at each node. Meanwhile, the secrecy throughput and energy efficiency (EE) of the system in the delay-limited mode under the conditions of imperfect/perfect successive interference cancellation (ipSIC/pSIC) are evaluated. Validated by Monte Carlo simulations, our analysis reveals: Thresholds and interference levels critically modulate performance; System configurations can be optimized to balance the reliability-security tradeoff, where reduced reliability may enhance security; Transmit power escalation enhances throughput but excessive interference hampers gains; An “EE-optimal power range” exists, balancing conservation and performance. These insights advance secure, efficient backscatter NOMA for IoT, paving the way for relay-assisted optimizations.
{"title":"Secure backscatter NOMA systems with dual-antenna relay: Reliability, security, and efficiency analysis","authors":"Liyao Ma , Yuhui Zhou , Gaojian Huang , Mengyan Huang","doi":"10.1016/j.phycom.2025.102917","DOIUrl":"10.1016/j.phycom.2025.102917","url":null,"abstract":"<div><div>The burgeoning Internet of Things (IoT) landscape demands innovative solutions for massive connectivity and energy efficiency, where backscatter communication excels in enabling low-power, cost-effective passive device interactions, and non-orthogonal multiple access (NOMA) boosts spectral efficiency via power-domain multiplexing. The synergy of these technologies in backscatter NOMA systems addresses key IoT challenges but introduces complexities such as inter-user interference, eavesdropping vulnerabilities, and residual SIC artifacts. This paper examines a backscatter NOMA system featuring two NOMA user groups, i.e., near-end and far-end, an embedded backscatter device (BD), a dual-antenna relay, and an eavesdropper, operating under a time-sharing mechanism with Rayleigh fading and AWGN. We derive closed-form expressions for outage probability (OP), asymptotic OP and intercept probability (IP) at each node. Meanwhile, the secrecy throughput and energy efficiency (EE) of the system in the delay-limited mode under the conditions of imperfect/perfect successive interference cancellation (ipSIC/pSIC) are evaluated. Validated by Monte Carlo simulations, our analysis reveals: <span><math><mrow><mo>(</mo><mn>1</mn><mo>)</mo></mrow></math></span> Thresholds and interference levels critically modulate performance; <span><math><mrow><mo>(</mo><mn>2</mn><mo>)</mo></mrow></math></span> System configurations can be optimized to balance the reliability-security tradeoff, where reduced reliability may enhance security; <span><math><mrow><mo>(</mo><mn>3</mn><mo>)</mo></mrow></math></span> Transmit power escalation enhances throughput but excessive interference hampers gains; <span><math><mrow><mo>(</mo><mn>4</mn><mo>)</mo></mrow></math></span> An “EE-optimal power range” exists, balancing conservation and performance. These insights advance secure, efficient backscatter NOMA for IoT, paving the way for relay-assisted optimizations.</div></div>","PeriodicalId":48707,"journal":{"name":"Physical Communication","volume":"73 ","pages":"Article 102917"},"PeriodicalIF":2.2,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145528242","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}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号