IEEE Transactions on Machine Learning in Communications and Networking最新文献

{"title":"ARCS: Adversarial RF Map Categorization and Synthesis Through Unsupervised Feature Extraction","authors":"Sopan Sarkar;Marwan Krunz;David Manzi;Rajesh Kulkarni","doi":"10.1109/TMLCN.2026.3676377","DOIUrl":"https://doi.org/10.1109/TMLCN.2026.3676377","url":null,"abstract":"In wireless networks, radio-frequency coverage maps (RF maps) are critical for tasks such as capacity planning, coverage estimation, and localization. Traditional methods for obtaining these maps, including site surveys and ray-tracing simulations, are either labor-intensive or computationally expensive, particularly at high frequencies. Generative AI offers a promising alternative for RF map synthesis and data augmentation. However, supervised generative approaches are often infeasible due to the lack of labeled training data, while unsupervised methods typically lack control over the generation process, limiting their practical utility. To overcome these challenges, we propose ARCS (Adversarial RF Map Categorization and Synthesis), a novel generative adversarial network (GAN)-based framework for unsupervised RF map categorization and synthesis. ARCS leverages the principles of information maximizing GAN (InfoGAN) to learn the latent structure of RF maps in an unsupervised manner during training and, through manipulation of the learned and interpretable latent codes, enables controlled generation of RF maps across floor plan regions and transmitter (Tx) locations during inference. To improve training stability and synthesis quality, we integrate a gradient penalty-based Wasserstein GAN objective function along with a customized gradient-based loss function. Extensive experiments on both experimental and simulated datasets show that ARCS generates high-quality RF maps, associates them with discrete regions of the floor plan, and provides fine-grained control of Tx location within each region. Compared with a UNet-based conditional GAN and a conditional diffusion model, ARCS attains the best scores across structural similarity index metric (SSIM), PSNR, MAPE, RMSE, cosine similarity (CS), and Jensen–Shannon divergence (JSD). Moreover, ARCS is extremely fast, requiring a few milliseconds per synthetic map, compared to over 14 seconds with ray-tracing.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"612-628"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11450453","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147606246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Adversarially Optimized Multi-Space Prototypical Network for Intrusion Detection in 5G-Enabled IoT Systems","authors":"Shahid Latif;Jawad Ahmad;Wadii Boulila;Djamel Djenouri","doi":"10.1109/TMLCN.2026.3673694","DOIUrl":"https://doi.org/10.1109/TMLCN.2026.3673694","url":null,"abstract":"Existing state-of-the-art IDS solutions in 5G-enabled IoT, including deep convolutional architectures, are constrained by data scarcity, limited edge resources, and poor robustness in few-shot learning (FSL) scenarios with scarce attack samples. To address these limitations, we propose MSPL-IDS, a robust FSL-based intrusion detection framework tailored for 5G-enabled IoT environments. Unlike existing methods that rely on a single embedding space and are highly vulnerable to adversarial perturbations, MSPL-IDS employs multi-space prototypical learning to project traffic features into multiple complementary embedding spaces, generating diverse class prototypes that strengthen decision boundaries against evasion attacks. A dual-loss adversarial training strategy jointly improves classification accuracy and robustness, while an attention mechanism adaptively selects the most reliable embedding spaces during inference, enabling low-latency deployment at the network edge. Comprehensive experiments on the 5G-NIDD benchmark demonstrate that MSPL-IDS significantly outperforms state-of-the-art deep learning and FSL-based IDSs under adversarial conditions. While representative models such as ResNet and Inception experience catastrophic accuracy drops to as low as 21.2% under evasion attacks, MSPL-IDS consistently maintains detection accuracy above 87% in identical few-shot settings. On average, the proposed framework achieves accuracy gains of 31.8% under FGSM and 26.4% under PGD attacks, with improvements exceeding 54% at high perturbation levels (<inline-formula> <tex-math>$epsilon ge 0.15$ </tex-math></inline-formula>), while preserving near-optimal clean-data accuracy (88.2–94.9%).","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"575-590"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11433812","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147558030","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Equivariant Multi-Agent Reinforcement Learning for Multimodal Vehicle-to-Infrastructure Systems","authors":"Charbel Bou Chaaya;Mehdi Bennis","doi":"10.1109/TMLCN.2026.3683038","DOIUrl":"https://doi.org/10.1109/TMLCN.2026.3683038","url":null,"abstract":"In this article, we study a vehicle-to-infrastructure (V2I) system where distributed base stations (BSs) acting as road-side units (RSUs) collect multimodal (wireless and visual) data from moving vehicles. We consider a decentralized rate maximization problem, where each RSU relies on its local observations to optimize its resources, while all RSUs must collaborate to guarantee favorable network performance. We recast this problem as a distributed multi-agent reinforcement learning (MARL) problem, by incorporating rotation symmetries in terms of vehicles’ locations. To exploit these symmetries, we propose a novel self-supervised learning framework where each BS agent aligns the latent features of its multimodal observation to extract the positions of the vehicles in its local region. Equipped with this sensing data at each RSU, we train an equivariant policy network using a graph neural network (GNN) with message passing layers, such that each agent computes its policy locally, while all agents coordinate their policies via a signaling scheme that overcomes partial observability and guarantees the equivariance of the global policy. We present numerical results carried out in a simulation environment, where ray-tracing and computer graphics are used to collect wireless and visual data. Results show the generalizability of our self-supervised and multimodal sensing approach, achieving more than two-fold accuracy gains over baselines, and the efficiency of our equivariant MARL training, attaining more than 50% performance gains over standard approaches.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"706-732"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11480198","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147737180","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"TDoA-Based Self-Supervised Channel Charting With NLoS Mitigation","authors":"Mohsen Ahadi;Omid Esrafilian;Florian Kaltenberger;Adeel Malik","doi":"10.1109/TMLCN.2026.3688619","DOIUrl":"https://doi.org/10.1109/TMLCN.2026.3688619","url":null,"abstract":"Channel Charting (CC) has emerged as a promising framework for data-driven radio localization, yet existing approaches often struggle to scale globally and to handle the distortions introduced by non-line-of-sight (NLoS) conditions. In this work, we propose a novel CC method that leverages Channel Impulse Response (CIR) data enriched with practical features such as Time Difference of Arrival (TDoA) and Transmission Reception Point (TRP) locations, enabling a TDoA-based self-supervised localization function on a global scale. The proposed framework is further enhanced with short-interval User Equipment (UE) displacement measurements, which improve the continuity and robustness of the learned positioning function. Our algorithm incorporates a mechanism to identify and mask NLoS-induced noisy measurements, leading to significant performance gains. We present the evaluations of our proposed models in a real 5G testbed and benchmarked against centimeter-accurate Real-Time Kinematic (RTK) positioning, in an O-RAN–based 5G network by OpenAirInterface (OAI) software at EURECOM. It demonstrates results that outperform the state-of-the-art semi-supervised and self-supervised CC approaches in a real-world scenario. The results show localization accuracies of 2–4 meters in 90% of cases, across varying NLoS ratios. Furthermore, we provide public datasets of CIR recordings, along with the true position labels used in this paper’s evaluation.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"780-793"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11500565","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147828987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reliable and Energy-Efficient MAC Protocols in Industrial IoT Networks via Multi-Agent Reinforcement Learning","authors":"Luciano Miuccio;Daniela Panno;Salvatore Riolo;Anita Schilirò","doi":"10.1109/TMLCN.2026.3683651","DOIUrl":"https://doi.org/10.1109/TMLCN.2026.3683651","url":null,"abstract":"The transition from wired to wireless communications in industrial Internet of Things (IIoT) networks introduces stringent challenges in terms of reliability and energy efficiency, aggravated by harsh propagation conditions and contention for a shared radio medium. These constraints require advanced medium access control (MAC) protocols capable of jointly managing channel access, packet retransmissions, and buffer operations while accounting for the battery limitations of IoT devices (IoTDs). This paper proposes a multi-agent reinforcement learning (MARL) framework for the autonomous design of energy-efficient and reliable MAC protocols in uplink wireless IIoT networks supporting time–frequency multiplexing. Moving away from conventional decentralized partially observable Markov decision process (Dec-POMDP)-based MARL designs, the framework adopts a partially observable Markov game (POMG), thereby enabling per-device policy learning. A novel reward mechanism is introduced, in which the base station broadcasts a resource-level feedback, and each device constructs a local reward based solely on its own observations and past actions, ensuring feasibility in real deployments. Simulation results show that the proposed framework achieves maximum reliability, whereas state-of-the-art MARL benchmarks based on global rewards fail to meet the required target, highlighting the importance of POMG modeling and local reward structures for reliable wireless IIoT networks. Furthermore, a comparison with conventional grant-based protocols, which inherently achieve maximum reliability, demonstrates that the proposed solution significantly reduces the active-mode duration, thereby improving overall energy efficiency.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"677-705"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11481086","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147737179","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Deep Reinforcement Learning-Based Scheduling for Wi-Fi Multi-Access Point Coordination","authors":"David Nunez;Francesc Wilhelmi;Maksymilian Wojnar;Katarzyna Kosek-Szott;Szymon Szott;Boris Bellalta","doi":"10.1109/TMLCN.2026.3682239","DOIUrl":"https://doi.org/10.1109/TMLCN.2026.3682239","url":null,"abstract":"Multi-access point coordination (MAPC) is a key feature of IEEE 802.11bn, with a potential impact on future Wi-Fi networks. MAPC enables joint scheduling decisions across multiple access points (APs) to improve throughput, latency, and reliability in dense Wi-Fi deployments. However, implementing efficient scheduling policies under diverse traffic and interference conditions in overlapping basic service sets (OBSSs) remains a complex task. This paper presents a method to minimize the network-wide worst-case latency by formulating MAPC scheduling as a sequential decision-making problem and proposing a deep reinforcement learning (DRL) mechanism to minimize worst-case delays in OBSS deployments. Specifically, we train a DRL agent using proximal policy optimization (PPO) within an 802.11bn-compatible Gymnasium environment. This environment provides observations of queue states, delay metrics, and channel conditions, enabling the agent to schedule multiple AP-station pairs to transmit simultaneously by leveraging spatial reuse (SR) groups. Simulations demonstrate that our proposed solution outperforms state-of-the-art heuristic strategies across a wide range of network loads and traffic patterns. The trained machine learning (ML) models consistently achieve lower 99th-percentile delays, showing up to a 30% improvement over the best baseline.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"744-757"},"PeriodicalIF":0.0,"publicationDate":"2026-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11478468","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147796246","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Clustering-Assisted Deep Reinforcement Learning for Joint Trajectory Design and Resource Allocation in Two-Tier-Cooperated UAVs Communications","authors":"Shujun Zhao;Simeng Feng;Chao Dong;Xiaojun Zhu;Qihui Wu","doi":"10.1109/TMLCN.2025.3647806","DOIUrl":"https://doi.org/10.1109/TMLCN.2025.3647806","url":null,"abstract":"Considering their high mobility and relatively low cost, uncrewed aerial vehicles (UAVs) equipped with mobile base stations are regarded as a potential technological approach. However, the dual pressures of limited onboard resources of UAVs and the demand for high-quality services in dynamic low-altitude applications jointly form a bottleneck for system performance. Although multi-UAVs communication networks can provide higher system performance through coordinated deployment, the challenges of cooperation and competition among UAVs, as well as more complex optimization problems, significantly increase costs and pose formidable challenges. To overcome the challenges of low coordination efficiency and intense resource competition among multiple UAVs, and to ensure the timely and efficient satisfaction of ground users (GUs) communication service demands, this paper conceives a centralized-controlled two-tier-cooperated UAVs communication network. The network comprises a central UAV (C-UAV) tier as control center and a marginal UAV (M-UAV) tier to serve GUs. In response to the increasingly dynamic and complex scenarios, along with the challenge of insufficient generalization ability in Deep Reinforcement Learning (DRL) algorithms, we propose a clustering-assisted dual-agent soft actor critic (CDA-SAC) algorithm for trajectory design and resource allocation, aiming to maximize the fair energy efficiency of the system. Specifically, by integrating a clustering-matching method with a dual-agent strategy, the proposed CDA-SAC algorithm achieves significant improvements in generalization ability and exploration capability. Simulation results demonstrate that the proposed CDA-SAC algorithm can be deployed without retraining in scenarios with different numbers of GUs. Furthermore, the CDA-SAC algorithm outperforms both the multi-UAV scenarios based on the MADDPG algorithm and the FDMA scheme in terms of fairness and total energy efficiency.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"178-197"},"PeriodicalIF":0.0,"publicationDate":"2025-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11313631","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145886662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Data-Driven Cellular Mobility Management Via Bayesian Optimization and Reinforcement Learning","authors":"Mohamed Benzaghta;Sahar Ammar;David López-Pére;Basem Shihada;Giovanni Geraci","doi":"10.1109/TMLCN.2025.3647807","DOIUrl":"https://doi.org/10.1109/TMLCN.2025.3647807","url":null,"abstract":"Mobility management in cellular networks faces increasing complexity due to network densification and heterogeneous user mobility characteristics. Traditional handover (HO) mechanisms, which rely on predefined parameters such as A3-offset and time-to-trigger (TTT), often fail to optimize mobility performance across varying speeds and deployment conditions. Fixed A3-offset and TTT configurations either delay HOs, increasing radio link failures (RLFs), or accelerate them, leading to excessive ping-pong effects. To address these challenges, we propose two distinct data-driven mobility management approaches leveraging high-dimensional Bayesian optimization (HD-BO) and deep reinforcement learning (DRL). While HD-BO optimizes predefined HO parameters such as A3-offset and TTT, DRL provides a parameter-free alternative by allowing an agent to select serving cells based on real-time network conditions. We systematically compare these two approaches in real-world site-specific deployment scenarios (employing Sionna ray tracing for site-specific channel propagation modeling), highlighting their complementary strengths. Results show that both HD-BO and DRL outperform 3GPP set-1 (TTT of 480 ms and A3-offset of 3 dB) and set-5 (TTT of 40 ms and A3-offset of −1 dB) benchmarks. We augment HD-BO with transfer learning so it can generalize across a range of user speeds. Applying the same transfer-learning strategy to the DRL method reduces its training time by a factor of 2.5 while preserving optimal HO performance, showing that it adapts efficiently to the mobility of aerial users such as UAVs. Simulations further reveal that HD-BO remains more sample-efficient than DRL, making it more suitable for scenarios with limited training data.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"228-244"},"PeriodicalIF":0.0,"publicationDate":"2025-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11313634","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145929689","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transforming Indoor Localization: Advanced Transformer Architecture for NLOS Dominated Wireless Environments With Distributed Sensors","authors":"Saad Masrur;Jung-Fu Cheng;Atieh R. Khamesi;İsmail Güvenç","doi":"10.1109/TMLCN.2025.3647376","DOIUrl":"https://doi.org/10.1109/TMLCN.2025.3647376","url":null,"abstract":"Indoor localization in challenging non-line-of-sight (NLOS) environments often leads to poor accuracy with traditional approaches. Deep learning (DL) has been applied to tackle these challenges; however, many DL approaches overlook computational complexity, especially for floating-point operations (FLOPs), making them unsuitable for resource-limited devices. Transformer-based models have achieved remarkable success in natural language processing (NLP) and computer vision (CV) tasks, motivating their use in wireless applications. However, their use in indoor localization remains nascent, and directly applying Transformers for indoor localization can be both computationally intensive and exhibit limitations in accuracy. To address these challenges, in this work, we introduce a novel tokenization approach, referred to as Sensor Snapshot Tokenization (SST), which preserves variable-specific representations of power delay profile (PDP) and enhances attention mechanisms by effectively capturing multi-variate correlation. Complementing this, we propose a lightweight Swish-Gated Linear Unit-based Transformer (L-SwiGLU-T) model, designed to reduce computational complexity without compromising localization accuracy. Together, these contributions mitigate the computational burden and dependency on large datasets, making Transformer models more efficient and suitable for resource-constrained scenarios. Experimental results on simulated and real-world datasets demonstrate that SST and L-SwiGLU-T achieve substantial accuracy and efficiency gains, outperforming larger Transformer and CNN baselines by over 40% while using significantly fewer FLOPs and training samples.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"161-177"},"PeriodicalIF":0.0,"publicationDate":"2025-12-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11313538","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145886663","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Explainable Multi-Agent Reinforcement Learning for Extended Reality Codec Adaptation","authors":"Pedro Enrique Iturria-Rivera;Raimundas Gaigalas;Medhat Elsayed;Majid Bavand;Yigit Ozcan;Melike Erol-Kantarci","doi":"10.1109/TMLCN.2025.3646125","DOIUrl":"https://doi.org/10.1109/TMLCN.2025.3646125","url":null,"abstract":"Extended Reality (XR) services are set to transform applications over <inline-formula> <tex-math>${mathbf {5}}^{th}$ </tex-math></inline-formula> and <inline-formula> <tex-math>${mathbf {6}}^{th}$ </tex-math></inline-formula> generation wireless networks, delivering immersive experiences. Concurrently, Artificial Intelligence (AI) advancements have expanded their role in wireless networks, however, trust and transparency in AI remain to be strengthened. Thus, providing explanations for AI-enabled systems can enhance trust. We introduce Value Function Factorization (VFF)-based Explainable (X) Multi-Agent Reinforcement Learning (MARL) algorithms, explaining reward design in XR codec adaptation through reward decomposition. We contribute four enhancements to XMARL algorithms. Firstly, we detail architectural modifications to enable reward decomposition in VFF-based MARL algorithms: Value Decomposition Networks (VDN), Mixture of Q-Values (QMIX), and Q-Transformation (Q-TRAN). Secondly, inspired by multi-task learning, we reduce the overhead of vanilla XMARL algorithms. Thirdly, we propose a new explainability metric, Reward Difference Fluctuation Explanation (RDFX), suitable for problems with adjustable parameters. Lastly, we propose adaptive XMARL, leveraging network gradients and reward decomposition for improved action selection. Simulation results indicate that, in XR codec adaptation, the Packet Delivery Ratio reward is the primary contributor to optimal performance compared to the initial composite reward, which included delay and Data Rate Ratio components. Modifications to VFF-based XMARL algorithms, incorporating multi-headed structures and adaptive loss functions, enable the best-performing algorithm, Multi-Headed Adaptive (MHA)-QMIX, to achieve significant average gains over the Adjust Packet Size baseline up to 10.7%, 41.4%, 33.3%, and 67.9% in XR index, jitter, delay, and Packet Loss Ratio (PLR), respectively.","PeriodicalId":100641,"journal":{"name":"IEEE Transactions on Machine Learning in Communications and Networking","volume":"4 ","pages":"245-264"},"PeriodicalIF":0.0,"publicationDate":"2025-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11303975","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145929688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}