SISR is an important research topic in computer vision. Its goal is to reconstruct HR images with rich details from LR inputs. Early methods based on CNNs made some progress, but their performance reached a limit due to limited model capacity and expressiveness. Recently, methods based on Transformers have shown significant improvements in this field. Their ability to capture long-range dependencies makes them well-suited for image reconstruction. However, these models often require high computational resources, which limits their practical use. This paper presents a detailed analysis of SISR and provides one key insight: the reconstruction performance depends on both low-level and high-level features. Then, we propose a novel SR model called HGFormer. This model uses a shallow architecture and introduces a Dynamic Spatial Information Compression (DSIC) module, which reduces computational complexity by converting the spatial information of mid-level features into the channel dimension. This improves both the efficiency and effectiveness of the model. HGFormer is the first method to expand the self-attention window to \(\varvec{32 \times 32}\) while keeping low computational costs. It achieves a performance gain of 0.69 dB on the Urban100 dataset. Extensive experiments on public datasets show that HGFormer outperforms existing methods in both objective metrics and visual quality.
11 February 2026 in IEEE Communications Standards Magazine
Existing Internet of Medical Things (IoMT) systems often struggle with the high latency and energy constraints required for real-time physio-logical monitoring. To address these challenges, this paper proposes a “Terminal-Edge-Cloud” collaborative architecture specifically designed for 6G-enabled medical applications. Leveraging 6G’s Ultra-Reliable Low-Latency Communication (URLLC) and Terahertz bandwidth, we devise a hierarchical task offloading strategy: terminal nodes perform lightweight signal preprocessing to conserve energy; edge nodes execute time-critical tasks such as fatigue detection using pre-trained models; and the cloud handles computation-intensive deep learning training and multi-sensor fusion. Theoretical analysis and architectural comparison demonstrate that this dynamic collaboration significantly reduces system latency compared to cloud-only models and improves resource utilization over static edge paradigms. The proposed framework offers a scalable solution for next-generation intelligent healthcare, ensuring both the timeliness of emergency alerts and the accuracy of long-term health analysis.
27 January 2026 in IEEE Communications Standards Magazine
Digital Twin (DT) is gradually becoming the core of future communication infrastructure and has been widely applied in key areas such as intelligent transportation, manufacturing, and healthcare. However, with the deep integration of 5G/6G and edge computing, real-time communication in DT systems faces significant cybersecurity threats and data privacy challenges. To address this issue, this paper proposes a robust and privacy-aware network watermarking framework (RPAW) for edge environments, designed for tracing communication paths and enabling real-time detection of data leakage. The framework innovatively combines time-domain and sequence-domain features to form a hybrid-domain embedding strategy, ensuring low latency and high concealment. To tackle the synchronization loss problem caused by time disturbances, we develop an adaptive synchronization mechanism based on the sequence domain, enabling stable watermark extraction in any network flow segment. Additionally, we design an intelligent watermark extraction model based on TabNet, significantly enhancing robustness under packet loss and delay jitter conditions. Extensive experiments demonstrate that the RPAW excels in terms of concealment and extraction robustness, and effectively supports secure and traceable data exchange in edge DT communication scenarios, offering a new approach to building a trustworthy DT communication ecosystem.
5 January 2026 in IEEE Transactions on Consumer Electronics
In latency-sensitive IoT Consumer Electronics (CE) scenarios, effective task offloading is pivotal for ensuring interaction quality. However, existing approaches often overlook fine-grained task urgency, while standard Deep Reinforcement Learning (DRL) suffers from slow convergence and high inference latency. To address these challenges, this paper proposes a semantic-aware optimization framework grounded in the Holographic Counterpart (HC) model. First, the Historical-Experience-Driven Dynamic Priority Task Allocation and Offloading (HE-DPTAO) method is developed, which leverages semantic priorities and edge caching to minimize queuing delays and task drop rates. Subsequently, the Offloading Decision Method based on MDPRL (ODM-MDPRL) is introduced, integrating self-attention mechanisms and asynchronous prioritized experience replay to accelerate training efficiency and decision precision. Simulation results demonstrate that this coordinated framework significantly outperforms representative baselines, achieving a superior balance between timeliness, stability, and resource utilization.
Existing autonomous aerial vehicle (AAV) clusters lack a formalized model, fail to consider external interference factors, and overlook the need for dynamic cluster maintenance to ensure stability and coordination in open scenarios. To address these issues, we propose an AAV collaborative cluster formation method suitable for open scenarios, which can form an energy-efficient, reliable, and stable AAV cluster even in external interferences. First, we present a AAV node promotion method based on mobility similarity and connectivity. Then, we formalize a collaborative AAV cluster model based on the energy efficiency, reliability, and stability among AAV nodes. Next, we propose a formation method for AAV clusters based on Pareto optimality and provide a maintenance method for clusters. Extensive simulation results demonstrate that the proposed method significantly outperforms state-of-the-art (SOTA) approaches. Specifically, in open scenarios, our method improves average cluster efficiency (ACE) by up to 6.9%, enhances average cluster reliability (ACR) by 12.4%, enhances average cluster stability (ACS) by 14.2%, and extends the average cluster survival time (ANST) and the average node survival time (ANST) by 18.1% and 17.2% compared to the best-performing baseline, verifying the superior effectiveness and robustness of the proposed framework.