IEEE Transactions on Cognitive Communications and Networking最新文献

{"title":"Intelligent Resource Allocation for Satellite-RIS-Assisted URLLC-ISAC Systems Using Deep Learning Approach","authors":"Debbarni Sarkar;Keshav Singh;Satyendra Singh Yadav;Chih-Peng Li","doi":"10.1109/TCCN.2026.3667139","DOIUrl":"10.1109/TCCN.2026.3667139","url":null,"abstract":"In sixth-generation (6G) wireless communication, cognitive satellite-terrestrial networks play a crucial role in ensuring seamless and reliable connectivity for satellite users (SUs) and concurrently enable efficient spectrum sharing among cellular users (CUs). However, the coexistence of SUs and CUs introduces challenges such as interference management, latency constraints, privacy threats from potential eavesdroppers (Eves), and the need for accurate sensing to support dynamic spectrum access. To address these challenges, the reconfigurable intelligent surface (RIS) emerges as a promising and cost-efficient technology, and its performance can be further enhanced through integration with ultra-reliable low-latency communication (URLLC). URLLC ensures fast, reliable, and secure communication for SUs and CUs and supports both real-time and mission-critical applications. Along with these, integrated sensing and communication (ISAC) is a pivotal technology that enhances the system performance through environmental monitoring, detecting Eve’s activity, and strengthening both satellite and terrestrial links. Recently, deep learning (DL) techniques have gained significant attention as powerful tools in optimizing network performance, improving signal detection, enabling smarter resource allocation, and adaptive beamforming. Therefore, this paper introduces an intelligent resource allocation approach in a satellite-RIS-assisted URLLC-ISAC system. A DL model has been proposed to optimize the sensing beam, maximizing the minimum beam gain across multiple targets to enhance resource allocation. The proposed DL model jointly optimizes transmit beamforming, the reflecting units, and the sensing matrix under practical constraints. Simulation results demonstrate the model’s effectiveness in handling the complex optimization in the system under consideration. Simulation results show that the convergence and feasibility of the proposed model, and the proposed DL model outperform in optimizing the beam gain compared to the benchmark models. The performance has also been analyzed for both active and passive RIS configurations. Furthermore, to test the model’s efficacy in changing channel conditions, the performance of the proposed model has been validated across different channels. The time complexity analysis of the proposed model has also been evaluated.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6492-6508"},"PeriodicalIF":7.0,"publicationDate":"2026-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147279096","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":"Optimizing System Performance for Autonomous Recovery in AI-Native Networks Using Large Language Models","authors":"Liming Huang;Yulei Wu;Dimitra Simeonidou","doi":"10.1109/TCCN.2026.3667165","DOIUrl":"10.1109/TCCN.2026.3667165","url":null,"abstract":"Future networks are expected to be AI-native, where AI is no longer an external tool but an integral element of the network infrastructure, with a complete AI lifecycle management. This lifecycle spans data collection, model training, inference, and evaluation, and must be coordinated alongside network infrastructure and application performance to maintain reliable end-to-end system behavior. However, such deep integration introduces new challenges, particularly in managing performance degradation caused by complex dependencies across lifecycle components in the system level. Traditional optimization methods, which isolate AI components as well as their interactions with networks and applications, fail to provide system-level assurance in AI-native environments. To address this, we propose a system-level performance optimization framework based on Large Language Models (LLMs) as autonomous reasoning agents within an Agentic AI paradigm. Our approach introduces a two-phase learning strategy: 1) fine-tuning the LLM with knowledge from AI solutions, networks, and applications to enable autonomous identification of performance-affecting factors, especially performance bottlenecks; and 2) real-world refinement using interaction-based learning and a reward-driven alignment mechanism that ties LLM decisions to application performance improvements. By embedding embodied intelligence into the network operation, the framework enables adaptive and lifecycle-aware performance recovery. We validate our method on a real-time 3D reconstruction task within an AI-native network, achieving a 94.5% performance recovery rate with minimal correction latency. The results demonstrate the effectiveness of LLM-powered agentic reasoning and adaptation in maintaining system performance and mitigating performance degradation in complex AI-native networking scenarios.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6451-6465"},"PeriodicalIF":7.0,"publicationDate":"2026-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147279098","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":"A Non-Stationary Channel Model for RIS-Mounted UAV-Assisted ISAC Systems","authors":"Abhinav Singh Parihar;Keshav Singh;Vimal Bhatia;Chih-Peng Li;Derrick Wing Kwan Ng","doi":"10.1109/TCCN.2026.3667173","DOIUrl":"10.1109/TCCN.2026.3667173","url":null,"abstract":"This paper introduces a novel unmanned aerial vehicle (UAV)-mounted reconfigurable intelligent surface (RIS)-assisted integrated sensing and communication (ISAC) framework that synergistically amalgamates data transmission with environmental awareness to enhance both spectral efficiency and system performance. However, the practical deployment of this system encounters challenges such as severe path loss and frequent line-of-sight (LoS) obstructions, especially in dynamic environments. To address these issues, we employ an RIS-equipped UAV that dynamically adapts its reflection coefficients, optimizing both communication and sensing in a multiple-input multiple-output (MIMO)-based wireless network, where a base station (BS) serves a user and simultaneously detects a target. This framework is further extended to a multi-user downlink scenario, where the BS serves multiple users via the RIS-mounted UAV while jointly performing target sensing. By leveraging the UAV’s mobility and considering multiple scatterers, the system optimally reconfigures its RIS phase shifts to improve sensing accuracy and ensure communication reliability. We derive analytical expressions for both communication and sensing rates to evaluate system performance and examine the impact of RIS reflection phase configurations on the statistical properties of the channel, providing valuable design insights for effective deployments. To accommodate the non-stationary effects in UAV-assisted RIS channels, our model incorporates real-time UAV velocity variations and their influence on beamforming. Furthermore, we characterize key propagation metrics, including spatial cross-correlation, temporal autocorrelation, and frequency correlation functions, along with channel capacity analysis. Theoretical and simulation results validate the proposed channel model, demonstrating improvements in system performance.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6358-6375"},"PeriodicalIF":7.0,"publicationDate":"2026-02-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147279097","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":"Joint Subarray Selection and Precoding Design for mmWave MIMO Systems: Deep Unfolding With GCN Approach","authors":"Lan Wu;Wei Huang;Shiwen He;Li Ping Qian;Caihong Kai","doi":"10.1109/TCCN.2026.3666539","DOIUrl":"10.1109/TCCN.2026.3666539","url":null,"abstract":"Millimeter-wave (mmWave) communication has been identified as a key enabler for future wireless systems due to its potential to offer high spectral efficiency with large antenna arrays. However, the high hardware cost and power consumption associated with the full digital architecture pose challenges in practical deployment. Hybrid precoding with antenna or subarray selection (AS/SS) offers a cost-effective solution but introduces challenging non-convex optimization problems. Traditional iterative algorithms are often computationally intensive and difficult to scale, while data-driven deep learning (DL) methods suffer from poor interpretability and limited generalization. To address these limitations, we propose a model-driven deep unfolding framework for joint precoding and AS/SS in mmWave systems. By using the equivalence between sum-rate maximization and mean square error (MSE) minimization, we unfold the weighted minimum MSE (WMMSE) algorithm into a trainable network with embedded sparsity modules. For AS, we incorporate a sparse mapping function with gradient-based matrix inversion to reduce complexity. For SS, a graph convolutional network (GCN) is integrated to exploit the spatial correlation among subarrays. Simulation results show that the proposed methods achieve near-optimal sum-rate with significantly lower complexity and good generalization, outperforming conventional optimization and learning approaches.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6391-6404"},"PeriodicalIF":7.0,"publicationDate":"2026-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146230624","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":"Joint Semantic Compression and Massive MIMO Transmission for Edge Video Analytics","authors":"Peng Cui;Shujun Han;Ze Liu;Yunfei Luo;Xiaodong Xu;Xiaoyu Chi;Bingxuan Xu;Ping Zhang","doi":"10.1109/TCCN.2026.3666538","DOIUrl":"10.1109/TCCN.2026.3666538","url":null,"abstract":"The advancement of artificial intelligence (AI) technologies has catalyzed widespread deployment of emerging video analytics applications, particularly in edge computing environments requiring real-time processing capabilities. However, achieving optimal trade-off between analytical accuracy and real-time performance for edge video analytics in bandwidth-constrained conditions poses a critical challenge. To address this problem, this paper proposes a joint semantic compression and massive multiple-input and multiple-output (massive-MIMO) transmission scheme for edge video analytics. Specifically, we design a task-oriented performance metric named task reliability, jointly evaluating the stochastic delay bound of intelligent task transmission and the accuracy of intelligent task execution. By optimizing pilot sequence length, pilot power, data power, and semantic compression ratio jointly, we maximize the weighted sum task reliability. Furthermore, we propose a power and semantic compression ratio allocation (PSCRA) algorithm, which utilizes maximum entropy multi-agent reinforcement learning to balance exploration and exploitation. Numerical results demonstrate that our resource allocation scheme achieves superior system utility compared to baseline methods, while the PSCRA algorithm outperforms conventional multi-agent reinforcement learning algorithms.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6509-6523"},"PeriodicalIF":7.0,"publicationDate":"2026-02-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146230625","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":"Joint Beamwidth and Power Density Allocation for ISAC-Enabled AAVs in Low-Altitude Economy","authors":"Fengyi Wang;Shilian Wang;Zhu Xiao;Tong Li;Zhu Han","doi":"10.1109/TCCN.2026.3665914","DOIUrl":"10.1109/TCCN.2026.3665914","url":null,"abstract":"This paper investigates integrated sensing and communication (ISAC) in low-altitude economy (LAE) networks, which is crucial for the flight safety of low-altitude autonomous aerial vehicles (AAVs). We derive the expected achievable rate (EAR) as a novel performance metric to evaluate the ISAC system performance under uncertain predicted direction. Both accurate beam direction prediction and beam resource allocation are essential for achieving fair EAR performance for multiple AAVs. However, the trajectories of the AAVs exhibit non-Gaussian deviations from the predefined trajectories. Moreover, the beamwidth and power density are interdependent for EAR performance. Motivated by this, we propose a transformer-based multitype weighting network (TMW-Net) to obtain accurate AAV state prediction under non-Gaussian scenarios. Building upon the state prediction results, we develop a joint beamwidth and power density allocation (JBPDA) method to maximize the minimum EAR for multiple AAVs. The proposed method searches for an auxiliary variable by establishing a power consumption-based feasibility evaluation problem, which is transformed into a simpler form through the cases of symmetric and asymmetric beam allocation. Finally, the numerical results show that the proposed TMW-Net has higher beam direction prediction accuracy. Furthermore, the designed JBPDA method can achieve a trade-off between communication and sensing in the ISAC system.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6421-6435"},"PeriodicalIF":7.0,"publicationDate":"2026-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146230631","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":"MT-KD: Multi-Teacher Knowledge Distillation for Privacy-Aware Unlearning in IIoT Intrusion Detection Systems","authors":"Shigen Shen;Jiashuo Niu;Yizhou Shen;Jingnan Dong;Wenlong Ke;Tian Wang;Ruidong Li","doi":"10.1109/TCCN.2026.3665917","DOIUrl":"10.1109/TCCN.2026.3665917","url":null,"abstract":"The growth of Industrial Internet of Things (IIoT) systems creates a direct conflict for intrusion detection systems: while essential for security, their inability to selectively forget specific attack patterns hinders adaptability and impedes compliance with privacy regulations like General Data Protection Regulation. To resolve this conflict, we introduce the Multi-Teacher Knowledge Distillation (MT-KD) unlearning framework, designed to erase knowledge of specified attack classes without the prohibitive cost of full model retraining. Our approach employs deep convolutional neural networks, multi-teacher guidance, and gradient-based noise injection to train a student model that forgets targeted attacks while retaining high accuracy on remaining classes. We validate this by converting network traffic from the CIC-IDS2017 and USTC-TFC2016 benchmark datasets into grayscale images for extensive experiments. The results demonstrate that our MT-KD framework achieves complete forgetting of targeted classes, verified by Membership Inference Attack resistance, while maintaining over 95% accuracy on retained classes. By outperforming existing unlearning baselines in both effectiveness and efficiency, our work delivers a practical and scalable solution for privacy-preserving intrusion detection in real-world IIoT scenarios. The source code is publicly available at <uri>https://github.com/NJS-HZ/MT-KD</uri>","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6344-6357"},"PeriodicalIF":7.0,"publicationDate":"2026-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146231071","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":"Multi-Stage Survivable Network Slicing With Load Balancing for LEO Mega-Constellation","authors":"Ning Chen;Ailing Xiao;Sheng Wu;Zhe Ji;Haoge Jia;Linling Kuang","doi":"10.1109/TCCN.2026.3665891","DOIUrl":"10.1109/TCCN.2026.3665891","url":null,"abstract":"In the low earth orbit (LEO) mega-constellation, dynamically reconstructing resources to enhance the on-orbit efficiency and flexibility of payloads remains a critical challenge. Network slicing (NS), a core concept in 5G communications, offers a promising solution by deploying end-to-end customized and logically isolated virtual networks over shared infrastructures. However, existing methods overlook the survivability of satellite nodes and the impact of load balancing of satellites and inter-satellite links (ISLs) on subsequent slicing processes. Therefore, this paper proposes a multi-stage survivable NS (MS-SNS) method based on load balancing for LEO mega-constellations. The method adopts a hybrid heuristic-reinforcement learning (RL) framework, decomposing the problem into multiple stages. First, a multi-constraint “load coefficient” indicator is proposed for constellation target area planning and selection using a greedy algorithm. Second, a soft actor-critic (SAC)-based algorithm is proposed for slice mapping within the selected target region. Finally, a survivable NS remapping algorithm is proposed, where a multi-constraint “remapping cost” indicator is proposed to guide the selection of remappable LEO satellite nodes, ensuring service continuity during node failures. Extensive simulations demonstrate the effectiveness of MS-SNS.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6405-6420"},"PeriodicalIF":7.0,"publicationDate":"2026-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146231072","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":"A YOLO-Driven Adaptive Focusing Network for Multi-Source Overlapping Modulation Perception","authors":"Xuan Zhu;Jiangjun Ruan;Hao Wu;Xun Ni;Shiao Zhu;Jin Meng","doi":"10.1109/TCCN.2026.3663595","DOIUrl":"10.1109/TCCN.2026.3663595","url":null,"abstract":"Towards the challenges of high dynamic signal-to-noise ratio (SNR) variation and multi-source channel overlapping, a YOLO-driven adaptive focusing network (YOLO-AFNet) is proposed for overlapping modulation signal perception. Firstly, a time-frequency spectrum dataset of single and overlapping modulation signals is constructed. Moreover, based on adaptive multi-scale feature extraction, Focal Modulation module and region-focused feature fusion network embedded with area-attention mechanism, a lightweight YOLO-AFNet recognition model is designed. Furthermore, a Kullback-Leibler divergence-optimized adaptive random magnitude channel pruning (KL-ARMCP) algorithm is proposed for model compression, assisted by masked generative distillation to restore accuracy, enabling exceptional computational efficiency for edge deployment with limited computing resources. Additionally, the deep semantic features extracted by the pruned YOLO-AFNet are combined with random forest regression for the end-to-end SNR estimation. Experimental results demonstrate that our method achieves a high recognition accuracy of 99.28% comparable to the state-of-the-art models, with a drastic reduction of 74.39% in FLOPs and 82.39% in Params, and achieves an SNR estimation error of 0.95%. Semi-physical experimental results further validate its practicality, confirming this efficient and lightweight end-to-end visual perception paradigm as a practical solution for intelligent signal processing on resource-limited edge platforms.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6376-6390"},"PeriodicalIF":7.0,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146161360","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":"SCAW: Exploring Semantic Structures of Radio Signals via Soft Top-k Contrastive and Adaptive Consistency With Wavelet Refinement","authors":"Yong Zu;Shuyuan Yang;Zhixi Feng;Xinlei Zhang;Yanbin Lu;Shuai Chen;Jun Liu;Qiukai Pan","doi":"10.1109/TCCN.2026.3663600","DOIUrl":"10.1109/TCCN.2026.3663600","url":null,"abstract":"Self-supervised learning (SSL) methods have recently been applied to cognitive radio technology (CRT), yielding promising results by utilizing unlabeled data. Nevertheless, most existing SSL approaches primarily focus on a single aspect, either adopting contrastive paradigms that emphasize discriminative feature learning or generative paradigms that excel at extracting the semantic structure of fine-grained information. In this paper, we introduce a novel SSL method, Soft Top-k Contrastive and Adaptive Consistency with Wavelet-Based Refinement (SCAW), which effectively integrates the advantages of both discriminative feature learning and semantic, fine-grained structural information extraction, thereby enhancing robustness and generalization in CRT scenarios. First, a wavelet-based feature refinement module is incorporated into the encoder to enable full-frequency and fine-grained semantic feature extraction. Second, a soft top-k contrastive loss is introduced to exploit inherent inter-sample relationships, thereby preserving discriminative capability and enhancing the semantic structural integrity of the learned representations. Moreover, an adaptive consistency loss is formulated to mitigate noise-induced perturbations while maintaining the underlying structure of the representations. The effectiveness of SCAW is evaluated on four public datasets across multiple downstream tasks. Experimental results show that SCAW consistently outperforms existing state-of-the-art methods.","PeriodicalId":13069,"journal":{"name":"IEEE Transactions on Cognitive Communications and Networking","volume":"12 ","pages":"6436-6450"},"PeriodicalIF":7.0,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146161359","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}