A5089450416- Privacy-Preserving Technologies in Data
- Sparse and Compressive Sensing Techniques
- Indoor and Outdoor Localization Technologies
- Advanced Optical Imaging Technologies
- IoT and Edge/Fog Computing
- Mobile Crowdsensing and Crowdsourcing
- Cryptography and Data Security
- Energy Efficient Wireless Sensor Networks
- Caching and Content Delivery
- Digital Holography and Microscopy
- Network Traffic and Congestion Control
- Peer-to-Peer Network Technologies
- Analog and Mixed-Signal Circuit Design
- Adversarial Robustness in Machine Learning
- Advanced X-ray Imaging Techniques
- Advanced Computational Techniques and Applications
- Optical measurement and interference techniques
- Image and Video Quality Assessment
- Internet Traffic Analysis and Secure E-voting
- Virtual Reality Applications and Impacts
- Software-Defined Networks and 5G
- Advanced Wireless Communication Technologies
- Advanced Data Storage Technologies
- Molecular Communication and Nanonetworks
- Underwater Vehicles and Communication Systems
Hunan University
2022-2025
Peng Cheng Laboratory
2024-2025
Beijing Institute of Technology
2024
Advanced Research Institute
2024
Capital Normal University
2024
Dali University
2024
Shanghai Jiao Tong University
2002-2024
State Forestry and Grassland Administration
2024
Henan Polytechnic University
2024
Zhengzhou University
2024
Federated learning (FL) is a privacy-preserving distributed machine framework, which involves training statistical models over number of mobile users (i.e., workers) while keeping data localized. However, recent works have demonstrated that workers engaged in FL are still susceptible to advanced inference attacks when sharing model updates or gradients, would discourage them from participating. Most the existing incentive mechanisms for mainly account workers' resource cost, cost incurred by...
Incentive mechanisms are essential for stimulating adequate worker participation to achieve good truth discovery performance in mobile crowdsensing (MCS) systems. However, most of existing incentive only consider compensating workers' sensing cost, while the cost incurred by potential privacy leakage has been largely neglected. Moreover, none privacy-preserving incorporated different preferences provide personalized payments them. In this paper, we propose a contract-based mechanism MCS...
Federated unlearning (FUL) is an emerging distributed machine learning paradigm which enables the removal or of specific training data effects from trained Learning (FL) models. While current studies mostly focus on client-side FUL to address "right be forgotten", and ignore server's right remove local models global model, particularly when clients are with low-quality data. In this paper, we introduce Server-Initiated Unlearning (SIFU) algorithm, devised eliminate model. SIFU consists two...
In this paper, a method is proposed to implement noises reduced three-dimensional (3D) holographic near-eye display by phase-only computer-generated hologram (CGH). The CGH calculated from double-convergence light Gerchberg-Saxton (GS) algorithm, in which the phases of two virtual convergence lights are introduced into GS algorithm simultaneously. first phase replacement random as iterative initial value and second will modulate distribution algorithm. Both simulations experiments carried...
With recent advances in communication technologies and Internet of Things (IoT) infrastructures, home automation (HA) systems have emerged as a new promising paradigm that provides convenient smart-home services to users. However, there exist various security risks during the deployment application HA systems, which pose severe threats On one hand, traditional IoT (e.g., device intrusion, protocol vulnerabilities, so on) are inherent systems. other core Trigger-Action Programming (TAP) model...
Transferability of adversarial examples is critical for black-box deep learning model attacks. While most existing studies focus on enhancing the transferability untargeted attacks, few them studied how to generate transferable targeted that can mislead models into predicting a specific class. Moreover, attacks usually fail sufficiently characterize target class distribution, thus suffering from limited transferability. In this paper, we propose Transferable Targeted Adversarial Attack...
As a privacy-preserving distributed learning paradigm, federated (FL) enables multiple client devices to train shared model without uploading their local data. To further enhance the privacy protection performance of FL, differential (DP) has been successfully incorporated into FL systems defend against attacks from adversaries. In with DP, how stimulate efficient collaboration is vital for server due nature DP and heterogeneity various costs (e.g., computation cost) participating clients....
Existing machine learning (ML) model marketplaces generally require data owners to share their raw data, leading serious privacy concerns. Federated (FL) can partially alleviate this issue by enabling training without exchange. However, are still susceptible leakage from gradient exposure in FL, which discourages participation. In work, we advocate a novel differentially private FL (DPFL)-based ML marketplace. We focus on the broker-centric design. Specifically, broker first incentivizes...
With the widespread application of 5G and Internet things (IoT), edge computing cloud have been collaboratively utilized for task offloading processing. However, though massive devices (e.g., smartphones) are organized into multi-cells, most existing works do not explore computation edge-cloud under inter-cell interference. Thus, decisions may be inappropriate as transmission rate is overestimated. To address this issue, we propose COMEC, a novel Computation Offloading scheme in Multi-cell...
Abstract Shanghai HIgh repetition rate XFEL aNd Extreme light facility (SHINE) is a free electron laser with high pulse of up to 1 MHz. STARLIGHT (SemiconducTor Array detectoR Large dynamIc ranGe and cHarge inTegrating readout) pixel array detector frame (≥ 10 kHz) large dynamic range (up 4 photons 12.4 keV), which being developed for SHINE. The readout ASIC utilized by in the front-end module HYLITE (High dYmamic Laser Imaging deTEctor). HYLITE200F 200 μm × size 64 pixels first full-scale...
A fundamental issue in multiaccess edge computing (MEC) is efficiently offloading multiple tasks to helper nodes (MTMH), i.e., MEC servers. However, most of the existing decentralized schemes do not consider interuser interference or merely adopt time division access (TDMA) as scheme for MTMH heterogeneous scenario, leading a large latency. To address these issues, we propose DOT, novel Decentralized Offloading Tasks orthogonal frequency (OFDMA)-based MEC, minimize sum cost terms energy...
Compressive data gathering (CDG) has been recognized as a promising technique to collect sensory in wireless sensor networks (WSNs) with reduced energy cost and better traffic load balancing. Besides, clustering is often integrated into CDG further facilitate the network performance. However, existing cluster-based methods generally require large number of nodes participate each compressive sensing (CS) measurement rarely consider possible node failures due power depletion or malicious...
The feeders in distribution networks can be generally classified into two types: homogenous and hybrid feeders. Hybrid consist of non-homogeneous line segments. In this paper, a fault location scheme is proposed for grids with both using the edge computing. standard transient amplitude ratio arrays are first constructed stored at nodes. Then, an accurate model trained by radial basis function neural network substation server. To achieve goal, transients sampled measurement nodes transmitted...
Fourier ptychographic microscopy (FPM) is a computational imaging technology used to achieve high-resolution with wide field-of-view. The existing methods of FPM suffer from the positional misalignment in system, by which quality recovered image determined. In this paper, forward neural network method correction (FNN-CP) proposed based on TensorFlow, consists two models. Both spectrum sample and four global position factors, are introduced describe positions LED elements, treated as...
Truth discovery is an effective tool to unearth truthful answers in crowdsourced question answering systems. Incentive mechanisms are necessary such systems stimulate worker participation. However, most of existing incentive only consider compensating workers' resource cost, while the cost incurred by potential privacy leakage has been rarely incorporated. More importantly, best our knowledge, how provide personalized payments for workers with different demands remains uninvestigated thus...