Pub Date : 2026-02-16DOI: 10.1109/LCOMM.2026.3665355
Mengliang Li;Chenyu Wu;Shuai Han;Weixiao Meng
Terahertz (THz) communication, a promising 6G candidate boasting ultra-wide bandwidth, confronts critical challenges including severe path loss and the inherent beam splitting effect in wideband scenarios. Reconfigurable intelligent surfaces (RIS) offer an effective means to dynamically manipulate wireless channels for performance enhancement. This letter proposes a novel RIS-assisted terahertz physical layer security (PLS) precoding scheme aimed at 6G ultra-high-speed secure communication requirements, providing a core solution to the security transmission challenges in THz broadband scenarios. The scheme precisely resolves the double beam splitting bottleneck caused by the coupling between RIS and THz broadband through beam focusing and shaping technology, achieving directional concentration of signal energy and leakage suppression from a spatial perspective; further innovatively, the artificial noise (AN) is projected into the channel null space of legitimate users, leveraging the ultra-wide bandwidth advantage of THz to construct a full-band interference barrier. In the end, the simulation results demonstrate the effectiveness of the proposed method.
{"title":"A Security-Enhanced Precoding Scheme for RIS-Assisted THz Communication Systems","authors":"Mengliang Li;Chenyu Wu;Shuai Han;Weixiao Meng","doi":"10.1109/LCOMM.2026.3665355","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3665355","url":null,"abstract":"Terahertz (THz) communication, a promising 6G candidate boasting ultra-wide bandwidth, confronts critical challenges including severe path loss and the inherent beam splitting effect in wideband scenarios. Reconfigurable intelligent surfaces (RIS) offer an effective means to dynamically manipulate wireless channels for performance enhancement. This letter proposes a novel RIS-assisted terahertz physical layer security (PLS) precoding scheme aimed at 6G ultra-high-speed secure communication requirements, providing a core solution to the security transmission challenges in THz broadband scenarios. The scheme precisely resolves the double beam splitting bottleneck caused by the coupling between RIS and THz broadband through beam focusing and shaping technology, achieving directional concentration of signal energy and leakage suppression from a spatial perspective; further innovatively, the artificial noise (AN) is projected into the channel null space of legitimate users, leveraging the ultra-wide bandwidth advantage of THz to construct a full-band interference barrier. In the end, the simulation results demonstrate the effectiveness of the proposed method.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1166-1170"},"PeriodicalIF":4.4,"publicationDate":"2026-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299538","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-16DOI: 10.1109/LCOMM.2026.3665259
Mahmoud AlaaEldin;Amy S. Inwood;Peter J. Smith;Michail Matthaiou
This letter investigates a novel uplink (UL) system that integrates power-domain non-orthogonal multiple access (PD-NOMA) with a continuous reconfigurable intelligent surface (CRIS). We analyze the effective CRIS-assisted channels under spatially correlated fading to accurately approximate the characteristic function of the cascaded channel. This allows the derivation of an expression for the bit error rate (BER), a key performance metric for UL PD-NOMA. We further utilize the derived BER expressions to introduce a joint optimization framework that minimizes the average BER via UL power allocation and dynamic RIS partitioning among the users. The analytical results are validated by simulations, and show that the proposed optimization scheme eliminates the BER floors that are associated with UL NOMA. The results also confirm the superiority of the optimized CRIS-NOMA scheme over conventional orthogonal multiple access (OMA) and non-optimized UL NOMA schemes.
{"title":"BER Analysis and Optimization for Continuous RIS-Enabled NOMA","authors":"Mahmoud AlaaEldin;Amy S. Inwood;Peter J. Smith;Michail Matthaiou","doi":"10.1109/LCOMM.2026.3665259","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3665259","url":null,"abstract":"This letter investigates a novel uplink (UL) system that integrates power-domain non-orthogonal multiple access (PD-NOMA) with a continuous reconfigurable intelligent surface (CRIS). We analyze the effective CRIS-assisted channels under spatially correlated fading to accurately approximate the characteristic function of the cascaded channel. This allows the derivation of an expression for the bit error rate (BER), a key performance metric for UL PD-NOMA. We further utilize the derived BER expressions to introduce a joint optimization framework that minimizes the average BER via UL power allocation and dynamic RIS partitioning among the users. The analytical results are validated by simulations, and show that the proposed optimization scheme eliminates the BER floors that are associated with UL NOMA. The results also confirm the superiority of the optimized CRIS-NOMA scheme over conventional orthogonal multiple access (OMA) and non-optimized UL NOMA schemes.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1161-1165"},"PeriodicalIF":4.4,"publicationDate":"2026-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-16DOI: 10.1109/LCOMM.2026.3664877
Yunlong Li;Jun Wu;Jiabao Yu;Zhiguang Yang;Mingkun Su;Yanrong Zhai;Xu Bai;Xiaorong Xu;Jianrong Bao
Cognitive radio (CR) technology aims to address spectrum scarcity by allowing sensor nodes (SNs) to detect and opportunistically access channels that are not utilized by primary users (PUs), thereby improving spectrum utilization. By applying cooperative spectrum sensing (CSS), CR improves spectrum detection accuracy, but also introduces security risks such as spectrum sensing data falsification (SSDF) attack. This letter proposes a defense strategy based on the improved isolation forest (IIF) for detecting malicious sensor nodes (MSNs) in cognitive wireless sensor networks (CWSNs). We analyze the distinct behavior patterns of the two types of SNs, and detect MSNs by isolating data points. Furthermore, we present an enhanced version of the traditional isolation forest to address its limitations. Finally, comparative experiments confirm the superiority of the proposed approach over six other anomaly detection algorithms, demonstrating the effectiveness of the improvements.
{"title":"Improved Isolation Forest-Based Anomaly Detection for Cooperative Spectrum Sensing Against SSDF Attack in CWSNs","authors":"Yunlong Li;Jun Wu;Jiabao Yu;Zhiguang Yang;Mingkun Su;Yanrong Zhai;Xu Bai;Xiaorong Xu;Jianrong Bao","doi":"10.1109/LCOMM.2026.3664877","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3664877","url":null,"abstract":"Cognitive radio (CR) technology aims to address spectrum scarcity by allowing sensor nodes (SNs) to detect and opportunistically access channels that are not utilized by primary users (PUs), thereby improving spectrum utilization. By applying cooperative spectrum sensing (CSS), CR improves spectrum detection accuracy, but also introduces security risks such as spectrum sensing data falsification (SSDF) attack. This letter proposes a defense strategy based on the improved isolation forest (IIF) for detecting malicious sensor nodes (MSNs) in cognitive wireless sensor networks (CWSNs). We analyze the distinct behavior patterns of the two types of SNs, and detect MSNs by isolating data points. Furthermore, we present an enhanced version of the traditional isolation forest to address its limitations. Finally, comparative experiments confirm the superiority of the proposed approach over six other anomaly detection algorithms, demonstrating the effectiveness of the improvements.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1151-1155"},"PeriodicalIF":4.4,"publicationDate":"2026-02-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299560","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-13DOI: 10.1109/LCOMM.2026.3664435
Bingqing Wang;Xiaoxia Xu;Man Ouyang;Jiang Liu;Yuanwei Liu
Mobile edge generation (MEG) enables low-latency AI-generated content (AIGC) by decentralizing large generative AI (GAI) models across the edge server (ES) and user equipment (UE). However, the distributed collaborative generation raises significant privacy concerns under potential communication eavesdropping. To address this issue, a novel secure MEG framework is proposed for image generation, which integrates a customized privacy-aware encoder-decoder module into the decentralized GAI architecture. Then, latent diffusion features can be securely transmitted from the ES to the UE via joint source-channel coding (JSCC), which supports cost-efficient and privacy-preserving ES-UE collaboration. To assess privacy leakage, a new metric termed privacy-preserving structural similarity (PP-SSIM) is introduced. The proposed modules are jointly trained in an end-to-end manner to optimize both reconstruction fidelity and privacy preservation. Experimental results show that the proposed method achieves low-latency and secure image generation for legitimate users, while effectively mitigating information leakage to eavesdroppers. Despite a slight degradation in visual quality, the perceptual fidelity is well preserved.
{"title":"Secure Mobile Edge Generation (MEG): Privacy Preservation in the Presence of Eavesdroppers","authors":"Bingqing Wang;Xiaoxia Xu;Man Ouyang;Jiang Liu;Yuanwei Liu","doi":"10.1109/LCOMM.2026.3664435","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3664435","url":null,"abstract":"Mobile edge generation (MEG) enables low-latency AI-generated content (AIGC) by decentralizing large generative AI (GAI) models across the edge server (ES) and user equipment (UE). However, the distributed collaborative generation raises significant privacy concerns under potential communication eavesdropping. To address this issue, a novel secure MEG framework is proposed for image generation, which integrates a customized privacy-aware encoder-decoder module into the decentralized GAI architecture. Then, latent diffusion features can be securely transmitted from the ES to the UE via joint source-channel coding (JSCC), which supports cost-efficient and privacy-preserving ES-UE collaboration. To assess privacy leakage, a new metric termed privacy-preserving structural similarity (PP-SSIM) is introduced. The proposed modules are jointly trained in an end-to-end manner to optimize both reconstruction fidelity and privacy preservation. Experimental results show that the proposed method achieves low-latency and secure image generation for legitimate users, while effectively mitigating information leakage to eavesdroppers. Despite a slight degradation in visual quality, the perceptual fidelity is well preserved.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1136-1140"},"PeriodicalIF":4.4,"publicationDate":"2026-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-12DOI: 10.1109/LCOMM.2026.3664146
Rui Guo;Yulin Wang;Meilin He
Polarization-Adjusted Convolutional (PAC) codes combine the polarization effect of polar codes with the memory structure of convolutional precoding, offering strong error-correcting performance. However, the performance of PAC codes heavily depends on the selection of information set. Existing rate profiling methods such as Reed Muller-Polar (RM-Polar) may inadvertently assign an excessively large information set to low-capacity bit-channels. This tends to result in degraded decoding performance especially when using decoders with small to moderate list sizes. This letter proposes a rate profiling refinement strategy for PAC codes that reassigns certain information and frozen bit positions to better match the underlying channel capacity. Simulation results indicate that the proposed strategy significantly improves the decoding performance of PAC codes over various code rates.
{"title":"Capacity-Constrained Rate Profiling Refinement for PAC Codes via Bit Position Reassignment","authors":"Rui Guo;Yulin Wang;Meilin He","doi":"10.1109/LCOMM.2026.3664146","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3664146","url":null,"abstract":"Polarization-Adjusted Convolutional (PAC) codes combine the polarization effect of polar codes with the memory structure of convolutional precoding, offering strong error-correcting performance. However, the performance of PAC codes heavily depends on the selection of information set. Existing rate profiling methods such as Reed Muller-Polar (RM-Polar) may inadvertently assign an excessively large information set to low-capacity bit-channels. This tends to result in degraded decoding performance especially when using decoders with small to moderate list sizes. This letter proposes a rate profiling refinement strategy for PAC codes that reassigns certain information and frozen bit positions to better match the underlying channel capacity. Simulation results indicate that the proposed strategy significantly improves the decoding performance of PAC codes over various code rates.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1146-1150"},"PeriodicalIF":4.4,"publicationDate":"2026-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-11DOI: 10.1109/LCOMM.2026.3663709
Xiaoyu Yao;Quan Gao;Zhanwen Wang;Li Wei;Lan Lan;Wei E. I. Sha
This work investigates a deterministic transmit beamforming algorithm for sidelobe suppression under a predefined mainlobe gain loss. By enforcing the unit-norm constraint via the Lagrange multiplier method, the problem is formulated as a quadratic eigenvalue problem that yields a numerical solution without iterative optimization. The algorithm also supports asymmetric sidelobe control and null steering, offering flexible beampattern shaping. Simulation results under various gain loss constraints demonstrate that the proposed method achieves up to 28.70% relative reduction in the sidelobe power compared with classical tapering window functions and 16.13% compared with gradient-based algorithms, while achieving comparable peak sidelobe levels.
{"title":"An Efficient Fast Beamforming Algorithm for Sidelobe Suppression and Gain Loss Limitation","authors":"Xiaoyu Yao;Quan Gao;Zhanwen Wang;Li Wei;Lan Lan;Wei E. I. Sha","doi":"10.1109/LCOMM.2026.3663709","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3663709","url":null,"abstract":"This work investigates a deterministic transmit beamforming algorithm for sidelobe suppression under a predefined mainlobe gain loss. By enforcing the unit-norm constraint via the Lagrange multiplier method, the problem is formulated as a quadratic eigenvalue problem that yields a numerical solution without iterative optimization. The algorithm also supports asymmetric sidelobe control and null steering, offering flexible beampattern shaping. Simulation results under various gain loss constraints demonstrate that the proposed method achieves up to 28.70% relative reduction in the sidelobe power compared with classical tapering window functions and 16.13% compared with gradient-based algorithms, while achieving comparable peak sidelobe levels.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1131-1135"},"PeriodicalIF":4.4,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-10DOI: 10.1109/LCOMM.2026.3663440
Hua Zhang;Zijian Cao;Le Liang;Hao Ye;Shi Jin;Geoffrey Ye Li
To enable multi-task semantic communication (SC) for image transmission, this letter proposes a prompt-guided multi-task SC system, termed PGMT-SC. PGMT-SC introduces a task-oriented hypernetwork that transforms natural-language description into prompt generator parameters, enabling flexible task adaptation without retraining the backbone. To enhance zero-shot generalization, we design a mixture-of-prompts (MOP) mechanism with Mixup training, which synthesizes prompts for unseen tasks by interpolating trained task embeddings. Simulation results under Rayleigh fading channels demonstrate that PGMT-SC outperforms existing schemes in multi-task performance and exhibits superior adaptability to unseen tasks.
{"title":"Prompt-Guided Multi-Task Semantic Communication for Image Transmission","authors":"Hua Zhang;Zijian Cao;Le Liang;Hao Ye;Shi Jin;Geoffrey Ye Li","doi":"10.1109/LCOMM.2026.3663440","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3663440","url":null,"abstract":"To enable multi-task semantic communication (SC) for image transmission, this letter proposes a prompt-guided multi-task SC system, termed PGMT-SC. PGMT-SC introduces a task-oriented hypernetwork that transforms natural-language description into prompt generator parameters, enabling flexible task adaptation without retraining the backbone. To enhance zero-shot generalization, we design a mixture-of-prompts (MOP) mechanism with Mixup training, which synthesizes prompts for unseen tasks by interpolating trained task embeddings. Simulation results under Rayleigh fading channels demonstrate that PGMT-SC outperforms existing schemes in multi-task performance and exhibits superior adaptability to unseen tasks.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1111-1115"},"PeriodicalIF":4.4,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-10DOI: 10.1109/LCOMM.2026.3663346
Ehsan Nassaji;Dmitri Truhachev
A fixed-to-fixed (F2F) length distribution matching (DM) based on Huffman coding and a random transformation of the input sequence is presented. The proposed algorithm necessitates small amounts of memory, is low-complexity, and delivers nearly the same performance as the enumerative sphere shaping (ESS). Furthermore, the proposed algorithm increases the speeds of matching and dematching and decreases the latency significantly, especially at the receiver.
{"title":"Fast Fixed-to-Fixed Distribution Matching via Huffman Coding","authors":"Ehsan Nassaji;Dmitri Truhachev","doi":"10.1109/LCOMM.2026.3663346","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3663346","url":null,"abstract":"A fixed-to-fixed (F2F) length distribution matching (DM) based on Huffman coding and a random transformation of the input sequence is presented. The proposed algorithm necessitates small amounts of memory, is low-complexity, and delivers nearly the same performance as the enumerative sphere shaping (ESS). Furthermore, the proposed algorithm increases the speeds of matching and dematching and decreases the latency significantly, especially at the receiver.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1141-1145"},"PeriodicalIF":4.4,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147299589","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Truncating the channel impulse response (CIR) is a critical step for non-line-of-sight (NLOS) identification within ultra-wideband (UWB) systems to manage data complexity. This letter formulates the process as a constrained and quantifiable optimization problem, thereby achieving an optimal balance between computational performance and classification accuracy. To implement this methodology, we first designed a lightweight attention-assisted network to serve as an effective evaluation tool. Leveraging this network, we then conducted a systematic search that identified an optimal input configuration: a CIR subset containing only 63 sample points. This co-optimized input model pair achieves a remarkable NLOS recall of 93.68%. More importantly, this result provides a critical a priori determination for deploying high precision localization algorithms on resource constrained MCUs. This co-optimized system has been successfully quantized and deployed on an STM32 MCU, verifying its high efficiency and feasibility for industrial grade applications.
{"title":"Optimizing UWB CIR Truncation for NLOS Identification on MCU Deployment","authors":"Yuqing Jiang;Hanbo Liu;Zhangyi Yang;Jiaxin Wu;Wei Dong","doi":"10.1109/LCOMM.2026.3662481","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3662481","url":null,"abstract":"Truncating the channel impulse response (CIR) is a critical step for non-line-of-sight (NLOS) identification within ultra-wideband (UWB) systems to manage data complexity. This letter formulates the process as a constrained and quantifiable optimization problem, thereby achieving an optimal balance between computational performance and classification accuracy. To implement this methodology, we first designed a lightweight attention-assisted network to serve as an effective evaluation tool. Leveraging this network, we then conducted a systematic search that identified an optimal input configuration: a CIR subset containing only 63 sample points. This co-optimized input model pair achieves a remarkable NLOS recall of 93.68%. More importantly, this result provides a critical a priori determination for deploying high precision localization algorithms on resource constrained MCUs. This co-optimized system has been successfully quantized and deployed on an STM32 MCU, verifying its high efficiency and feasibility for industrial grade applications.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1091-1095"},"PeriodicalIF":4.4,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146223572","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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 : 2026-02-09DOI: 10.1109/LCOMM.2026.3658941
Ende Wang;Taiyang Ma;Jiageng Liu;Yifan Ji;Suyao Sun
This letter presents a modulation recognition model to address LPI waveform recognition (LWR) through time-frequency images (TFIs). The model introduces a Dilated-MBConv module to analyze the structural disparities between TFIs and natural images. By introducing dilated convolutions with enlarged receptive fields, the module enables the network to extract both fine-grained and spatially contextualized features for reliable discrimination. To further enhance representational capacity, a Convolutional Block Attention Module (CBAM) is incorporated to selectively emphasize informative regions while suppressing redundant activations. Experimental results demonstrate that model outperforms state-of-the-art methods in LPI waveform recognition and validate the transferability of the Dilated-MBConv module.
{"title":"DCENet: A Dilated Convolution and Attention-Based Network for LPI Radar Signal Modulation Recognition","authors":"Ende Wang;Taiyang Ma;Jiageng Liu;Yifan Ji;Suyao Sun","doi":"10.1109/LCOMM.2026.3658941","DOIUrl":"https://doi.org/10.1109/LCOMM.2026.3658941","url":null,"abstract":"This letter presents a modulation recognition model to address LPI waveform recognition (LWR) through time-frequency images (TFIs). The model introduces a Dilated-MBConv module to analyze the structural disparities between TFIs and natural images. By introducing dilated convolutions with enlarged receptive fields, the module enables the network to extract both fine-grained and spatially contextualized features for reliable discrimination. To further enhance representational capacity, a Convolutional Block Attention Module (CBAM) is incorporated to selectively emphasize informative regions while suppressing redundant activations. Experimental results demonstrate that model outperforms state-of-the-art methods in LPI waveform recognition and validate the transferability of the Dilated-MBConv module.","PeriodicalId":13197,"journal":{"name":"IEEE Communications Letters","volume":"30 ","pages":"1046-1050"},"PeriodicalIF":4.4,"publicationDate":"2026-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146223731","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：