Over-the-air computation (AirComp) enables spectral-efficient global model aggregation for federated learning (FL) by supporting concurrent transmission and harnessing co-channel interference. However, unfavorable channel conditions and inaccurate channel estimation are two performance-limiting factors of AirComp-assisted FL. In this paper, we leverage reconfigurable intelligent surface (RIS) to assist AirComp for gradient aggregation with imperfect cascaded channel state information (CSI), taking into account both the expectation-based and worst-case error models (i.e., stochastic and deterministic models). Guided by the convergence analysis, we minimize the gradient aggregation distortion measured by the mean-squared-error (MSE), taking into account the unit modulus constraints of RIS phase-shifts. To alleviate the performance degradation due to imperfect channel estimation, we propose two robust algorithms to optimize the receive beamforming at the edge server, RIS phase-shifts, and transmit power at the edge devices. Both algorithms are designed under an alternating optimization framework, where Schur’s complement and penalty convex–concave procedure are adopted for the worst-case error model, and Lagrange duality and difference-of-convex programming are used for the expectation-based error model. Simulations are conducted to validate the learning performance superiority of the proposed algorithms over baseline schemes, inducing the robustness against CSI errors.
{"title":"Robust Design for RIS-Assisted Over-the-Air Federated Learning With Imperfect Cascaded CSI","authors":"Qiaochu An;Hongbin Zhu;Qiang Ye;Ning Zhang;Yuanming Shi;Yong Zhou","doi":"10.1109/TCCN.2026.3661518","DOIUrl":"10.1109/TCCN.2026.3661518","url":null,"abstract":"Over-the-air computation (AirComp) enables spectral-efficient global model aggregation for federated learning (FL) by supporting concurrent transmission and harnessing co-channel interference. However, unfavorable channel conditions and inaccurate channel estimation are two performance-limiting factors of AirComp-assisted FL. In this paper, we leverage reconfigurable intelligent surface (RIS) to assist AirComp for gradient aggregation with imperfect cascaded channel state information (CSI), taking into account both the expectation-based and worst-case error models (i.e., stochastic and deterministic models). Guided by the convergence analysis, we minimize the gradient aggregation distortion measured by the mean-squared-error (MSE), taking into account the unit modulus constraints of RIS phase-shifts. To alleviate the performance degradation due to imperfect channel estimation, we propose two robust algorithms to optimize the receive beamforming at the edge server, RIS phase-shifts, and transmit power at the edge devices. Both algorithms are designed under an alternating optimization framework, where Schur’s complement and penalty convex–concave procedure are adopted for the worst-case error model, and Lagrange duality and difference-of-convex programming are used for the expectation-based error model. Simulations are conducted to validate the learning performance superiority of the proposed algorithms over baseline schemes, inducing the robustness against CSI errors.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6045-6060"},"PeriodicalIF":7.0,"publicationDate":"2026-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146134231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-02-03DOI: 10.1109/TCCN.2026.3660763
Shikhar Verma;Sayantan Bose;Mostafa M. Fouda;Zubair Md Fadlullah;Diptendu Sinha Roy
The rise of unmanned aerial vehicles (UAVs) and electric vertical take-off and landing (eVTOL) aircraft is accelerating the growth of the low-altitude economy (LAE), enabling mobility beyond conventional ground-based transport. However, aerial vehicles or flying vehicles (FVs) in LAE environments face significant communication challenges due to high mobility, frequent handovers between base stations (BSs), and the susceptibility of mmWave bands to blockage and path loss. Reactive handover mechanisms—triggered only after link degradation—often lead to disconnections and degraded service quality, particularly in dense urban areas. Moreover, uneven FV distribution can cause BS load imbalances, further compromising quality of service (QoS). To address these challenges, we propose a proactive BS association framework for intelligent handover management using artificial general intelligence (AGI). Our approach leverages deep learning to jointly predict future received signal strength indicator (RSSI) and BS load, enabling an autonomous decision algorithm to select optimal BSs for stable, high-throughput connectivity while minimizing unnecessary handovers. Simulation results demonstrate that the proposed joint prediction-based strategy significantly reduces handover frequency and improves average throughput compared to reactive, single-metric baselines and two additional benchmark predictors introduced for extended evaluation. These findings underscore the potential of predictive, AGI-driven mobility management to enhance the stability and performance of communication networks in the emerging LAE ecosystem.
{"title":"Proactive Handover Management in 6G Low-Altitude Economy Networks for Aerial Vehicles Using Artificial General Intelligence","authors":"Shikhar Verma;Sayantan Bose;Mostafa M. Fouda;Zubair Md Fadlullah;Diptendu Sinha Roy","doi":"10.1109/TCCN.2026.3660763","DOIUrl":"10.1109/TCCN.2026.3660763","url":null,"abstract":"The rise of unmanned aerial vehicles (UAVs) and electric vertical take-off and landing (eVTOL) aircraft is accelerating the growth of the low-altitude economy (LAE), enabling mobility beyond conventional ground-based transport. However, aerial vehicles or flying vehicles (FVs) in LAE environments face significant communication challenges due to high mobility, frequent handovers between base stations (BSs), and the susceptibility of mmWave bands to blockage and path loss. Reactive handover mechanisms—triggered only after link degradation—often lead to disconnections and degraded service quality, particularly in dense urban areas. Moreover, uneven FV distribution can cause BS load imbalances, further compromising quality of service (QoS). To address these challenges, we propose a proactive BS association framework for intelligent handover management using artificial general intelligence (AGI). Our approach leverages deep learning to jointly predict future received signal strength indicator (RSSI) and BS load, enabling an autonomous decision algorithm to select optimal BSs for stable, high-throughput connectivity while minimizing unnecessary handovers. Simulation results demonstrate that the proposed joint prediction-based strategy significantly reduces handover frequency and improves average throughput compared to reactive, single-metric baselines and two additional benchmark predictors introduced for extended evaluation. These findings underscore the potential of predictive, AGI-driven mobility management to enhance the stability and performance of communication networks in the emerging LAE ecosystem.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"5953-5965"},"PeriodicalIF":7.0,"publicationDate":"2026-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110330","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-02-03DOI: 10.1109/TCCN.2026.3660777
Xinyi Wei;Ruoguang Li;Yingyang Chen;Lianming Xu;Li Wang;Zhu Han
In disaster scenarios, infrastructure damage and wireless resource scarcity pose significant challenges for providing prompt and reliable communication and sensing (C&S) services. Recently, unmanned aerial vehicle (UAV) enabled integrated sensing and communication (ISAC) has emerged as a promising technique to tackle the above issues by leveraging flexibility and mobility of multiple UAVs to offer high-quality and cost-efficient C&S services. In parallel, rate-splitting multiple access (RSMA) facilitates customized transmission by partitioning messages into private and common parts with adjustable rates, thereby making it well-suited for on-demand data transmission in disaster scenarios. In this paper, we propose a framework that utilizes coordinated RSMA for ISAC (Coordinated RSMA-ISAC) in an emergency UAV system. This framework enables multiple transmit UAVs to simultaneously communicate with several communication survivors (CSs) and detect a potentially trapped survivor (TS) in a coordinated manner with imperfect channel state information (CSI). In addition, an optimization problem is formulated to jointly optimize the UAV-CS association, UAV deployment, and transmit beamforming to maximize the weighted sum rate (WSR) of the system, subject to the sensing signal-to-noise ratio (SNR) requirement. To efficiently solve such a mixed-integer non-convex programming (MINCP) problem, an iterative algorithm is proposed by applying the generalized Benders decomposition (GBD) technique. Specifically, the original problem is decoupled into a master problem for pure integer programming and a primal problem for non-convex programming. Then, we further use successive convex approximation (SCA), semi-definite relaxation (SDR), and cutting-plane techniques to solve the decoupled problems. Simulation results verify the effectiveness of the proposed algorithm, and demonstrate that the coordinated RSMA-ISAC framework outperforms conventional space division multiple access (SDMA), non-orthogonal multiple access (NOMA), and orthogonal multiple access (OMA) in terms of both C&S performance.
{"title":"Coordinated Rate-Splitting Multiple Access for Emergency UAV-Enabled Integrated Sensing and Communication","authors":"Xinyi Wei;Ruoguang Li;Yingyang Chen;Lianming Xu;Li Wang;Zhu Han","doi":"10.1109/TCCN.2026.3660777","DOIUrl":"10.1109/TCCN.2026.3660777","url":null,"abstract":"In disaster scenarios, infrastructure damage and wireless resource scarcity pose significant challenges for providing prompt and reliable communication and sensing (C&S) services. Recently, unmanned aerial vehicle (UAV) enabled integrated sensing and communication (ISAC) has emerged as a promising technique to tackle the above issues by leveraging flexibility and mobility of multiple UAVs to offer high-quality and cost-efficient C&S services. In parallel, rate-splitting multiple access (RSMA) facilitates customized transmission by partitioning messages into private and common parts with adjustable rates, thereby making it well-suited for on-demand data transmission in disaster scenarios. In this paper, we propose a framework that utilizes coordinated RSMA for ISAC (Coordinated RSMA-ISAC) in an emergency UAV system. This framework enables multiple transmit UAVs to simultaneously communicate with several communication survivors (CSs) and detect a potentially trapped survivor (TS) in a coordinated manner with imperfect channel state information (CSI). In addition, an optimization problem is formulated to jointly optimize the UAV-CS association, UAV deployment, and transmit beamforming to maximize the weighted sum rate (WSR) of the system, subject to the sensing signal-to-noise ratio (SNR) requirement. To efficiently solve such a mixed-integer non-convex programming (MINCP) problem, an iterative algorithm is proposed by applying the generalized Benders decomposition (GBD) technique. Specifically, the original problem is decoupled into a master problem for pure integer programming and a primal problem for non-convex programming. Then, we further use successive convex approximation (SCA), semi-definite relaxation (SDR), and cutting-plane techniques to solve the decoupled problems. Simulation results verify the effectiveness of the proposed algorithm, and demonstrate that the coordinated RSMA-ISAC framework outperforms conventional space division multiple access (SDMA), non-orthogonal multiple access (NOMA), and orthogonal multiple access (OMA) in terms of both C&S performance.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"5999-6015"},"PeriodicalIF":7.0,"publicationDate":"2026-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-02-03DOI: 10.1109/TCCN.2026.3660764
Xingcan Chen;Chenglin Li;Wei Meng;Wendong Xiao
WiFi channel state information (CSI)-based human activity recognition (HAR) approaches face some fundamental limitations, such as high computational cost of the model due to irrelevant signal components obscure human-related features, position-sensitive representations cause activity misclassification under spatial variance, and environmental heterogeneity induces domain shifts that degrade generalization. To overcome these challenges, we propose a lightweight HAR approach based on WiFi CSI component decomposition and triple feature fusion (WiDeFus). Specifically, WiDeFus first isolates human-related components via quantum-inspired sparse decomposition, and leverage Hermite-Gaussian bases with group-element sparsity constraints to eliminate dynamic interference and noises. WiDeFus then introduces a triple-feature adaptive fusion network to achieve dynamic frequency-domain selection, automatically extract temporal features, and perform environment-robust calibration. These purified features are processed by a dendrite net (DD) that replaces nonlinear activations with multiplicative interactions for efficient inference. Experimental results show that WiDeFus is a lightweight HAR approach with high recognition accuracy and satisfactory cross-domain generalization performance.
{"title":"WiDeFus: A Wi-Fi-Based Lightweight Human Activity Recognition via CSI Component Decomposition and Adaptive Feature Fusion","authors":"Xingcan Chen;Chenglin Li;Wei Meng;Wendong Xiao","doi":"10.1109/TCCN.2026.3660764","DOIUrl":"10.1109/TCCN.2026.3660764","url":null,"abstract":"WiFi channel state information (CSI)-based human activity recognition (HAR) approaches face some fundamental limitations, such as high computational cost of the model due to irrelevant signal components obscure human-related features, position-sensitive representations cause activity misclassification under spatial variance, and environmental heterogeneity induces domain shifts that degrade generalization. To overcome these challenges, we propose a lightweight HAR approach based on WiFi CSI component decomposition and triple feature fusion (WiDeFus). Specifically, WiDeFus first isolates human-related components via quantum-inspired sparse decomposition, and leverage Hermite-Gaussian bases with group-element sparsity constraints to eliminate dynamic interference and noises. WiDeFus then introduces a triple-feature adaptive fusion network to achieve dynamic frequency-domain selection, automatically extract temporal features, and perform environment-robust calibration. These purified features are processed by a dendrite net (DD) that replaces nonlinear activations with multiplicative interactions for efficient inference. Experimental results show that WiDeFus is a lightweight HAR approach with high recognition accuracy and satisfactory cross-domain generalization performance.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6237-6246"},"PeriodicalIF":7.0,"publicationDate":"2026-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110912","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we consider an uncrewed aerial vehicle-assisted integrated sensing and communication system, consisting of an aerial base station (ABS) and a mobile radar (MR) equipped with a radar receiver. The ABS hovers at a specific position to provide communication services to user clusters, while the MR flies along a predefined trajectory to receive communication reflection signals for target sensing. By taking into account user communication and target sensing performance, as well as flight energy consumption of the MR, the system utility function is modeled as a weighted sum of the minimum average rate of user clusters, sensing channel gain, and flight energy consumption of the MR. The joint communication precoding design, ABS deployment, MR trajectory planning, and communication sensing scheduling problem is modeled as a constrained system utility function maximization problem. Given that the modeled optimization problem is a highly coupled and non-convex mixed-integer optimization problem, it is challenging to solve directly. To tackle this problem, we decompose the original problem into four subproblems and sequentially solve each subproblem using an alternating iteration algorithm. Specifically, for the communication precoding design subproblem, the zero-forcing algorithm is used to eliminate the interference among users and the original problem is transformed into a semi-definite programming problem. For the ABS deployment subproblem and the MR trajectory planning subproblem, the Taylor expansion and the successive convex approximation are employed and slack variables are introduced to convert the original problems into convex optimization problems. For the communication sensing scheduling subproblem, the variable relaxation method is adopted to relax the binary variables into continuous variables, and the optimization tool is used to obtain the solution. Then, two heuristic algorithms are proposed to restore the communication and sensing scheduling variables. Finally, the effectiveness of the proposed algorithms is verified through simulations.
{"title":"System Utility Function Optimization-Based Flight Trajectory and Resource Allocation for UAV-Assisted Integrated Sensing and Communication Systems","authors":"Zida Guo;Rong Chai;Ruijin Sun;Chengchao Liang;Qianbin Chen","doi":"10.1109/TCCN.2026.3660231","DOIUrl":"10.1109/TCCN.2026.3660231","url":null,"abstract":"In this paper, we consider an uncrewed aerial vehicle-assisted integrated sensing and communication system, consisting of an aerial base station (ABS) and a mobile radar (MR) equipped with a radar receiver. The ABS hovers at a specific position to provide communication services to user clusters, while the MR flies along a predefined trajectory to receive communication reflection signals for target sensing. By taking into account user communication and target sensing performance, as well as flight energy consumption of the MR, the system utility function is modeled as a weighted sum of the minimum average rate of user clusters, sensing channel gain, and flight energy consumption of the MR. The joint communication precoding design, ABS deployment, MR trajectory planning, and communication sensing scheduling problem is modeled as a constrained system utility function maximization problem. Given that the modeled optimization problem is a highly coupled and non-convex mixed-integer optimization problem, it is challenging to solve directly. To tackle this problem, we decompose the original problem into four subproblems and sequentially solve each subproblem using an alternating iteration algorithm. Specifically, for the communication precoding design subproblem, the zero-forcing algorithm is used to eliminate the interference among users and the original problem is transformed into a semi-definite programming problem. For the ABS deployment subproblem and the MR trajectory planning subproblem, the Taylor expansion and the successive convex approximation are employed and slack variables are introduced to convert the original problems into convex optimization problems. For the communication sensing scheduling subproblem, the variable relaxation method is adopted to relax the binary variables into continuous variables, and the optimization tool is used to obtain the solution. Then, two heuristic algorithms are proposed to restore the communication and sensing scheduling variables. Finally, the effectiveness of the proposed algorithms is verified through simulations.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6331-6343"},"PeriodicalIF":7.0,"publicationDate":"2026-02-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146101487","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ray-tracing (RT) channel simulation has been widely used for simulating and analyzing propagation of electromagnetic waves in complex environments. Accuracy of RT simulations depends on environment construction, including both scene structure and material classification. However, existing RT studies rely on manual material segmentation and presuppose idealized material parameters, which highly overlooks critical challenges of material recognition from real-world data. This prevents RT from being applied in complex scenarios and leads to inaccurate simulation results. To address this issue, we use point cloud measurements to capture real-world environment information and propose a dual-branch network based on PointNet model to automatically classify environmental materials by integrating point cloud data and LiDAR-derived feature parameters. The proposed network significantly enhances material classification accuracy within complex scenes, thereby delivering more precise and computationally efficient input data for RT simulations. Furthermore, we analyze influence of material recognition accuracy on simulation parameters, such as path loss and delay spread. The results demonstrate that the proposed network achieves high classification performance and meets accuracy requirements of RT, thereby contributing to more realistic and reliable predictions for wireless systems. This approach lays a crucial foundation for development of environment-aware models for 6G networks, enabling more effective simulation of outdoor communication environments.
{"title":"Point Cloud-Based Environmental Material Classification for Wireless Channel Ray-Tracing Simulations","authors":"Zhuoyin Li;Ruisi He;Mi Yang;Ziyi Qi;Zhong Zhang;Haoxiang Zhang;Jiahui Han;Bo Ai;Zhangdui Zhong","doi":"10.1109/TCCN.2026.3659825","DOIUrl":"10.1109/TCCN.2026.3659825","url":null,"abstract":"Ray-tracing (RT) channel simulation has been widely used for simulating and analyzing propagation of electromagnetic waves in complex environments. Accuracy of RT simulations depends on environment construction, including both scene structure and material classification. However, existing RT studies rely on manual material segmentation and presuppose idealized material parameters, which highly overlooks critical challenges of material recognition from real-world data. This prevents RT from being applied in complex scenarios and leads to inaccurate simulation results. To address this issue, we use point cloud measurements to capture real-world environment information and propose a dual-branch network based on PointNet model to automatically classify environmental materials by integrating point cloud data and LiDAR-derived feature parameters. The proposed network significantly enhances material classification accuracy within complex scenes, thereby delivering more precise and computationally efficient input data for RT simulations. Furthermore, we analyze influence of material recognition accuracy on simulation parameters, such as path loss and delay spread. The results demonstrate that the proposed network achieves high classification performance and meets accuracy requirements of RT, thereby contributing to more realistic and reliable predictions for wireless systems. This approach lays a crucial foundation for development of environment-aware models for 6G networks, enabling more effective simulation of outdoor communication environments.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"5878-5890"},"PeriodicalIF":7.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089962","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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 : 2026-01-30DOI: 10.1109/TCCN.2026.3659849
Minhyeok Jang;Jalel Ben-Othman;Hyunbum Kim
Object detection is a fundamental task in computer vision with broad applications in autonomous transportation driving, smart surveillance, and traffic monitoring. In the context of Computing Power Networks (CPNs), which interconnect cloud, edge, and terminal nodes to support distributed AI services, deploying efficient object detection models under constrained computational resources is a critical challenge, particularly at the edge and terminal layers. This study investigates backbone optimization for the YOLOv11M object detection framework to improve computational efficiency while maintaining detection performance. We propose two lightweight variants, YOLOv11M-MN and YOLOv11M-Shuffle, by replacing the original backbone with MobileNetV3-Small and ShuffleNetV2, respectively. All Edge AI-enabled models share an identical detection head and training pipeline to ensure fair and controlled comparisons. To reflect resource-limited CPN environments, all experiments are conducted under CPU-only settings with staged training budgets. Performance is evaluated using the COCO128 dataset in terms of FLOPs, parameter count, inference latency, and detection accuracy. Experimental results demonstrate that the proposed lightweight backbones significantly reduce computational overhead and inference time, while exhibiting different accuracy–efficiency trade-offs, highlighting their suitability for selective deployment across heterogeneous CPN layers.
{"title":"Edge AI-Enabled Backbone Optimization for Real-Time Object Detection in Computing Power Networks","authors":"Minhyeok Jang;Jalel Ben-Othman;Hyunbum Kim","doi":"10.1109/TCCN.2026.3659849","DOIUrl":"10.1109/TCCN.2026.3659849","url":null,"abstract":"Object detection is a fundamental task in computer vision with broad applications in autonomous transportation driving, smart surveillance, and traffic monitoring. In the context of Computing Power Networks (CPNs), which interconnect cloud, edge, and terminal nodes to support distributed AI services, deploying efficient object detection models under constrained computational resources is a critical challenge, particularly at the edge and terminal layers. This study investigates backbone optimization for the YOLOv11M object detection framework to improve computational efficiency while maintaining detection performance. We propose two lightweight variants, YOLOv11M-MN and YOLOv11M-Shuffle, by replacing the original backbone with MobileNetV3-Small and ShuffleNetV2, respectively. All Edge AI-enabled models share an identical detection head and training pipeline to ensure fair and controlled comparisons. To reflect resource-limited CPN environments, all experiments are conducted under CPU-only settings with staged training budgets. Performance is evaluated using the COCO128 dataset in terms of FLOPs, parameter count, inference latency, and detection accuracy. Experimental results demonstrate that the proposed lightweight backbones significantly reduce computational overhead and inference time, while exhibiting different accuracy–efficiency trade-offs, highlighting their suitability for selective deployment across heterogeneous CPN layers.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"5891-5902"},"PeriodicalIF":7.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089963","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Waveform recognition is used to identify classes of different electrical signal waveforms. It is widely used in medical diagnosis, wireless communication, signal video processing, audio processing, and other fields. However, conventional approaches to waveform recognition perform poorly under low signal-to-noise ratio conditions and with easily confusable signals of different types. To address this challenge, we introduce automatic machine learning (AutoML) into the waveform recognition task to automatically obtain a high-performance recognition network. Among AutoML, Differentiable Neural Architecture Search (DARTS) has been widely used due to its short training time, low hardware requirements, and efficient neural networks generation ability. However, DARTS cannot guide the sparsity of the architecture parameters, leading to performance loss in discretization. To this end, we propose an end-to-end trainable evolutionary method called Directional Differentiable Architecture Search (DDAS) in this paper. First, during the architecture search, the training data is augmented with more easily confusable signals samples to enhance the model’s ability to distinguish ambiguous patterns. Second, the soft-max function delays the non-zero-is-one operation selection to a weighted sum of the different operations. This makes the operation architectural parameters differentiable and greatly reduces the training cost. Third, a directional pruning-based optimization method is used to bring the highest weight of operations closer to 1 to reduce discretization loss, where the operation with the highest weight is selected as the final operation in the generated network. Experiments on two benchmark waveform recognition datasets show that the resulting network outperforms both the traditional manually designed network and the network obtained by directly applying existing architectural search methods, achieving higher accuracy. The obtained network also has a higher ability to recognize easily confusable signals. Notably, the generated network performance performs equally well under all signal-to-noise (SNR) ratios, offering new insights for waveform recognition. Code and datasets are available on https://github.com/tju-xm/DDAS.
{"title":"Directional Differentiable Architecture Search for Waveform Recognition","authors":"Xuemin Sun;Qing Wang;Xiaofeng Liu;Zhiming Zhan;Haozhi Wang;Qi Chen;Yifang Zhang","doi":"10.1109/TCCN.2026.3658773","DOIUrl":"10.1109/TCCN.2026.3658773","url":null,"abstract":"Waveform recognition is used to identify classes of different electrical signal waveforms. It is widely used in medical diagnosis, wireless communication, signal video processing, audio processing, and other fields. However, conventional approaches to waveform recognition perform poorly under low signal-to-noise ratio conditions and with easily confusable signals of different types. To address this challenge, we introduce automatic machine learning (AutoML) into the waveform recognition task to automatically obtain a high-performance recognition network. Among AutoML, Differentiable Neural Architecture Search (DARTS) has been widely used due to its short training time, low hardware requirements, and efficient neural networks generation ability. However, DARTS cannot guide the sparsity of the architecture parameters, leading to performance loss in discretization. To this end, we propose an end-to-end trainable evolutionary method called Directional Differentiable Architecture Search (DDAS) in this paper. First, during the architecture search, the training data is augmented with more easily confusable signals samples to enhance the model’s ability to distinguish ambiguous patterns. Second, the soft-max function delays the non-zero-is-one operation selection to a weighted sum of the different operations. This makes the operation architectural parameters differentiable and greatly reduces the training cost. Third, a directional pruning-based optimization method is used to bring the highest weight of operations closer to 1 to reduce discretization loss, where the operation with the highest weight is selected as the final operation in the generated network. Experiments on two benchmark waveform recognition datasets show that the resulting network outperforms both the traditional manually designed network and the network obtained by directly applying existing architectural search methods, achieving higher accuracy. The obtained network also has a higher ability to recognize easily confusable signals. Notably, the generated network performance performs equally well under all signal-to-noise (SNR) ratios, offering new insights for waveform recognition. Code and datasets are available on <uri>https://github.com/tju-xm/DDAS</uri>.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6309-6319"},"PeriodicalIF":7.0,"publicationDate":"2026-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146089964","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"AGI-Inspired Digital Twin Framework for UAV-BS Deployment in Disaster Communication Recovery","authors":"Luyu Qi, Yulei Wu, Shuping Dang, Zhuhui Li, Dimitra Simeonidou","doi":"10.1109/tccn.2026.3658781","DOIUrl":"https://doi.org/10.1109/tccn.2026.3658781","url":null,"abstract":"","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"217 1","pages":""},"PeriodicalIF":8.6,"publicationDate":"2026-01-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146070220","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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：