A5052951255- Privacy-Preserving Technologies in Data
- Cryptography and Data Security
- Cloud Data Security Solutions
- Internet Traffic Analysis and Secure E-voting
- Blockchain Technology Applications and Security
- Chaos-based Image/Signal Encryption
- Adversarial Robustness in Machine Learning
- Stochastic Gradient Optimization Techniques
- Cryptographic Implementations and Security
- Advanced Malware Detection Techniques
- Advanced Steganography and Watermarking Techniques
- Intelligence, Security, War Strategy
- Physical Unclonable Functions (PUFs) and Hardware Security
- Age of Information Optimization
- Neural Networks and Applications
- Advanced Memory and Neural Computing
- UAV Applications and Optimization
- Face and Expression Recognition
- Network Security and Intrusion Detection
- Cognitive Computing and Networks
- Text and Document Classification Technologies
- Blood donation and transfusion practices
- Advanced Graph Neural Networks
- Satellite Communication Systems
- Digital Media Forensic Detection
Shandong University
2023-2025
Peng Cheng Laboratory
2025
University of Wollongong
2023
Cloud computing, which provides adequate storage and computation capability, has been a prevalent information infrastructure. Secure data sharing is basic demand when was outsourced to cloud server. Attribute-based proxy re-encryption promising approach that allows secure encrypted on clouds. With attribute-based re-encryption, delegator can designate set of shared users through issuing key will be used by the server transform delegator's users'. However, existing schemes lack mechanism...
As a privacy-preserving distributed learning paradigm, federated (FL) has been proven to be vulnerable various attacks, among which backdoor attack is one of the toughest. In this attack, malicious users attempt embed triggers into local models, resulting in crafted inputs being misclassified as targeted labels. To address such several defense mechanisms are proposed, but may lose effectiveness due following drawbacks. First, current methods heavily rely on massive labeled clean data, an...
Contrastive learning has recently emerged as a powerful technique for graph self-supervised pretraining (GSP). By maximizing the mutual information (MI) between positive sample pair, network is forced to extract discriminative from graphs generate high-quality representations. However, we observe that, in process of MI maximization (Infomax), existing contrastive GSP algorithms suffer at least one following problems: 1) treat all samples equally during optimization and 2) fall into single...
6G-based wireless communication system is poised to redefine the next-generation network landscape by enabling novel services and applications, such as intelligent link establishment, power control, data collection, transmission, distribution. However, security issues, particularly recently revealed channel access attack (CAA), present significant challenges performance optimization tasks in heterogeneous networks of 6G, namely, Age Information (AoI) oriented Network (AoN), Throughput (ToN),...
Recent studies have demonstrated that backdoor attacks can cause a significant security threat to federated learning. Existing defense methods mainly focus on detecting or eliminating the patterns after model is backdoored. However, these either performance degradation heavily rely impractical assumptions, such as labeled clean data, which exhibit limited effectiveness in To this end, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Website fingerprinting attack is an extensively studied technique used in a web browser to analyze traffic patterns and thus infer confidential information about users. Several website attacks based on machine learning deep tend use the most typical features achieve satisfactory performance of attacking rate. However, these suffer from several practical implementation factors, such as skillfully pre-processing step or clean dataset. To defend against attacks, random packet defense (RPD) with...
The replication-Based Outsourced Computation (RBOC) mechanism allows a client to outsource the same computing job multiple contractors and honest will get paid in incentivized system based on fact that majority of honestly perform computation. As self-executing contracts, smart contracts are utilized decentralized blockchain networks execute coded programs automatically transparently, publicly. It is natural apply RBOC improve performance by setting as converter between reduce load client....
The decentralized Federated Learning (FL) paradigm built upon blockchain architectures leverages distributed node clusters to replace the single server for executing FL model aggregation. This tackles vulnerability of centralized malicious in vanilla and inherits trustfulness robustness offered by blockchain. However, existing blockchain-enabled schemes face challenges related inadequate confidentiality on models limited computational resources blockchains perform large-scale computations....
Attribute-based encryption that supports computation outsourcing has been a promising approach to implement fine-grained access control in the Internet of Things (IoT). However, existing schemes only focus on minimizing lightweight user terminals while overlooking significant computational burden faced by devices. In IoT scenarios with large volume concurrent requests, such can overwhelm devices, leading inefficient decryption execution. This paper, for first time, proposes ORR-CP-ABE, an...
With the advancement of Internet Things (IoT), numerous machine learning applications on IoT are encountering performance bottlenecks. Graph embedding is an emerging type that has achieved commendable results in areas such as network anomaly detection, malware device management, and service recommendation within Things. However, for some resource-constrained devices, computing graph algorithms highly complex time-consuming. In this paper, we introduce efficient secure distributed outsourcing...