Pansharpening, a pivotal task in remote sensing, involves integrating low-resolution multispectral images with high-resolution panchromatic images to synthesize an image that is both high-resolution and retains multispectral information. These pansharpened images enhance precision in land cover classification, change detection, and environmental monitoring within remote sensing data analysis. While deep learning techniques have shown significant success in pansharpening, existing methods often face limitations in their evaluation, focusing on restricted satellite data sources, single scene types, and low-resolution images. This paper addresses this gap by introducing PanBench, a high-resolution multi-scene dataset containing all mainstream satellites and comprising 5898 pairs of samples. Each pair includes a four-channel (RGB + near-infrared) multispectral image of 256 × 256 pixels and a mono-channel panchromatic image of 1024 × 1024 pixels. To avoid irreversible loss of spectral information and achieve a high-fidelity synthesis, we propose a Cascaded Multiscale Fusion Network (CMFNet) for pansharpening. Multispectral images are progressively upsampled while panchromatic images are downsampled. Corresponding multispectral features and panchromatic features at the same scale are then fused in a cascaded manner to obtain more robust features. Extensive experiments validate the effectiveness of CMFNet.
3 August 2022 in IEEE Transactions on Network Science and Engineering
Mobile edge caching has great advantages in alleviating video traffic pressure and reducing transmission delay, which is considered as a hopeful solution to improve video resource utilization and achieve faster service response. However, due to insufficient and non-real time acquisition of the status between edge servers and mobile users, it is still difficult for existing solutions to improve the video quality of experience (VQoE) for mobile users. Besides, due to the mobility of users and malicious third-party attacks, users may suffer the risk of personal privacy (e.g. location, trajectory, preference, etc.) leakage when requesting video services that are scattered in different areas. In order to efficiently deal with these problems, this paper proposes a privacy-preserving Q-learning based video caching optimization framework in mobile edge networks (VC-PPQ), including a privacy scheme for user data and a dynamic video caching algorithm. Specifically, in the first task, the local differential privacy model is leveraged to protect the locations and preferences of users. Then, a data aggregation model is provided to make trade-off between aggregation accuracy and privacy protection. In the second task, we further obtain the video popularity in user's area by a computation service model, then formulate a caching objective function combined with transcoding technologies. Afterwards, Q-learning is adopted to obtain the goal of caching optimization, which can achieve low delay and high hit ratio. Finally, extensive simulations prove that our VC-PPQ framework delivers prominent performance advantages in terms of user privacy and caching services, which greatly improves the VQoE of users.
5 January 2024 in IEEE Transactions on Computational Social Systems
Due to the rapid expansion of the Internet of Vehicles (IoVs), service providers deploy roadside units (RSUs), and base stations (BSs) close to vehicles. They can provide vehicles with computational offloading services quickly. In the context of vehicle social networks, where vehicles can communicate and share data with each other, the security and efficiency of data distribution are crucial. Unfortunately, the open nature of RSU BSs makes them vulnerable to malicious attackers, hence affecting the quality of the user experience. This article proposes a security trust degree incentive-based evaluation mechanism that calculates the security trust degree of vehicle users to RSU BSs through the continuous interaction between them in order to effectively address the aforementioned issues. Additionally, taking into account the competitive nature of task computation offloading between vehicle users and BSs, a stable matching algorithm is used to match each vehicle user with the most appropriate BS so that they can work together to prevent competition in task offloading and improve task offloading efficiency. Due to the limited number of BS matches and the dynamic position changes of vehicle users, we further increase the data distribution efficiency by calculating the vehicle user degree of relationship and connection probability to match vehicle users with similar preferences. Finally, our proposed scheme is validated via numerous simulations with enhanced security service performance in terms of vehicle task offloading, while data distribution efficiency are effectively improved.
5 January 2024 in IEEE Transactions on Sustainable Computing
5G vehicular networks provide abundant multimedia services among mobile vehicles. However, due to the mobility of vehicles, large-scale mobile traffic poses a challenge to the core network load and transmission latency. It is difficult for existing solutions to guarantee the quality of service (QoS) of vehicular networks. Besides, the sensitivity of vehicle trajectories also brings privacy concerns in vehicular networks. To address these problems, we propose a privacy-preserving and trajectory clustering-based framework for cooperative caching optimization (PTCC) in vehicular networks, which includes two tasks. Specifically, in the first task, we first apply differential privacy technologies to add noise to vehicle trajectories. In addition, a data aggregation model is provided to make the trade-off between aggregation accuracy and privacy protection. In order to analyze similar behavioral vehicles, trajectory clustering is then achieved by utilizing machine learning algorithms. In the second task, we construct a cooperative caching objective function with the transmission latency. Afterwards, the multi-agent deep Q network (MADQN) is leveraged to obtain the goal of caching optimization, which can achieve low delay. Finally, extensive simulation results verify that our framework respectively improves the QoS up to $9.8\%$ and $12.8\%$ with different file numbers and caching capacities, compared with other state-of-the-art solutions.
20 May 2025 in IEEE Transactions on Network Science and Engineering
With the rapid advancement of intelligent transportation systems (ITS), vehicle edge caching (VEC) technology shows great potential in processing large-scale vehicle data and providing instant services. The high mobility of vehicles restricts the duration of their connection to a vehicle roadside unit (RSU) to a short interval within the unit's service region. It becomes a challenge to accurately predict and update the popular contents during this period and decide the caching location of the popular contents. The transfer of substantial data volumes concurrently presents a significant danger of privacy breaches. To address these difficulties effectively, we proposed a collaborative caching scheme for vehicular edges that considers mobility of vehicles, employing federated distillation (FD) and deep reinforcement learning (DRL). First, we establish a vehicle mobility model using the real-world vehicle movement dataset T-Drive. After that, we propose a framework that integrates recommender systems with federated distillation instead of the traditional autoencoder (AE)-based federated learning scheme to obtain more accurate global models to predict content popularity. Due to the limited storage capacity of a single RSU, in order to maximise the use of edge cache resources, we propose a cooperative caching algorithm based on deep reinforcement learning, which caches these popular contents in multiple RSUs close to the vehicle. Finally, we evaluate the performance of our proposed algorithm on the vehicle mobility model, and the experimental results show that our proposed scheme improves the caching efficiency by 6.7% and reduces the content delivery latency by 11.3% relative to the five baseline schemes on the MovieLens-1M dataset.