Dehua Zhou
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Dehua Zhou has a total of 60 co-authors. Here are the top co-authors by number of shared publications:
- Beibei Song
College of Information Science and Technology/College of Cyber Security, Jinan University, Guangzhou 510632, China
2 shared publications - Chuansheng Wang
College of Information Science and Technology/College of Cyber Security, Jinan University, Guangzhou 510632, China
2 shared publications Haobin Ma
Guangdong Science Center, Guangzhou, China
2 shared publications- Ke Wang
Jinan University
2 shared publications - Shuai Pang
International School, Jinan University, Guangzhou, Guangdong, China
2 shared publications - Xiaoming Wang
Department of Computer Science, Jinan University, Guangzhou 510632, China
2 shared publications
Latest Publications
- 9 October 2026
- Published by Springer Nature in Automated Software Engineering
- Vol. 33 (4) , 146
OLASan: practical object-level sanitization through memory access aggregation
- 9 October 2026
- Published by Springer Nature in Automated Software Engineering
- Vol. 33 (4) , 146
OLASan: practical object-level sanitization through memory access aggregation
- Open Access
- 15 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 2266-2275
GAformer: Low-Light Image Enhancement Based on Gradient-Aware Kernel and Frequency-Modulated Transformer
- 15 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 2266-2275
Open AccessGAformer: Low-Light Image Enhancement Based on Gradient-Aware Kernel and Frequency-Modulated Transformer
- Open Access
- 6 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-7
AS-UAP:Attention-Shift Universal Adversarial Perturbation on Transformer-based Models
- 6 December 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessAS-UAP:Attention-Shift Universal Adversarial Perturbation on Transformer-based Models
- 1 May 2025
- Published by Elsevier in Journal of Systems Architecture
- Vol. 165
A traceable and revocable attribute-based encryption scheme with escrow-free in cloud storage
- 1 May 2025
- Published by Elsevier in Journal of Systems Architecture
- Vol. 165
A traceable and revocable attribute-based encryption scheme with escrow-free in cloud storage
- 30 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (2) , 3212-3224
BlockPPA: Blockchain-Assisted Privacy-Preserving Authentication for Cross-Domain Electric Vehicle Charging
Electric vehicles (EVs) bring convenience to public transport while having the potential to increase energy efficiency. With the rapid growth of the number of EVs, cross-domain charging will become a strong demand. Despite massive research on EV charging, existing schemes can hardly meet the security requirements of privacy protection and access control for crossdomain scenarios. Meanwhile, centralized methods bring the single point of failure and data modification. The problem of designing a secure cross-domain EV charging scheme to simultaneously ensure decentralization, privacy and access control has not been studied and remains open. Therefore, in this paper, we design BlockPPA, the first blockchain-assisted privacy-preserving authentication scheme for cross-domain EV charging in order to achieve this purpose. Our scheme makes each EV obtain an attribute-based secret key, which is used to authenticate itself and also anonymously illustrate that it really satisfies a designated access policy when cross-domain authentication happens. Only authorized EVs satisfying the policy do they have the ability to charge. We design three smart contracts and exploit Fabric blockchain to record the domain's detailed information and the authentication record of each cross-domain access from an EV, while achieving the decentralized verification of an authentication request. The results of security and performance analysis show the feasibility and effectiveness of the proposed cross-domain authentication mechanism
- 30 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 74 (2) , 3212-3224
BlockPPA: Blockchain-Assisted Privacy-Preserving Authentication for Cross-Domain Electric Vehicle Charging
Electric vehicles (EVs) bring convenience to public transport while having the potential to increase energy efficiency. With the rapid growth of the number of EVs, cross-domain charging will become a strong demand. Despite massive research on EV charging, existing schemes can hardly meet the security requirements of privacy protection and access control for crossdomain scenarios. Meanwhile, centralized methods bring the single point of failure and data modification. The problem of designing a secure cross-domain EV charging scheme to simultaneously ensure decentralization, privacy and access control has not been studied and remains open. Therefore, in this paper, we design BlockPPA, the first blockchain-assisted privacy-preserving authentication scheme for cross-domain EV charging in order to achieve this purpose. Our scheme makes each EV obtain an attribute-based secret key, which is used to authenticate itself and also anonymously illustrate that it really satisfies a designated access policy when cross-domain authentication happens. Only authorized EVs satisfying the policy do they have the ability to charge. We design three smart contracts and exploit Fabric blockchain to record the domain's detailed information and the authentication record of each cross-domain access from an EV, while achieving the decentralized verification of an authentication request. The results of security and performance analysis show the feasibility and effectiveness of the proposed cross-domain authentication mechanism
- Open Access
- 12 September 2024
- Published by MDPI AG in Cryptography
- Vol. 8 (3) , 42
A Multi-Candidate Self-Tallying Voting Scheme Based on Smart Contracts
In this paper, we propose a smart contract-based multi-candidate self-tallying voting scheme in order to guarantee the privacy of ballots in the case of multiple candidates. This scheme uses the ElGamal cryptosystem to ensure the security of the ballots, and combines it with a Distributed Encryption algorithm to make the voting scheme have self-tallying features, and guarantees the correctness of the intermediate data through zero-knowledge proofs. The experimental results show that the scheme improves the voting efficiency without compromising the security.
- 12 September 2024
- Published by MDPI AG in Cryptography
- Vol. 8 (3) , 42
Open AccessA Multi-Candidate Self-Tallying Voting Scheme Based on Smart Contracts
In this paper, we propose a smart contract-based multi-candidate self-tallying voting scheme in order to guarantee the privacy of ballots in the case of multiple candidates. This scheme uses the ElGamal cryptosystem to ensure the security of the ballots, and combines it with a Distributed Encryption algorithm to make the voting scheme have self-tallying features, and guarantees the correctness of the intermediate data through zero-knowledge proofs. The experimental results show that the scheme improves the voting efficiency without compromising the security.
- 26 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (23) , 38552-38568
VSAFL: Verifiable and Secure Aggregation With (Poly) Logarithmic Overhead in Federated Learning
Federated learning is a distributed machine learning framework that enables multiple participants to train models without directly sharing local data. However, sensitive information about participants may still be leaked through their gradients. Furthermore, centralized servers used for aggregating these gradients can be vulnerable to compromise, leading to privacy violations or other malicious attacks. Therefore, it is essential to verify the integrity of the aggregation. In this work, we focus on designing communication-efficient and fast verifiable aggregations for federated learning. We propose VSAFL, a verifiable secure aggregation protocol specifically designed for cross-device federated learning. VSAFL achieves computation and communication cost of O(log2n+log n + 1) and O(logn+l), respectively, for secure aggregation and verification for each user in each epoch, where n represents the number of clients and l denotes the dimension of the gradient vector. By employing a lightweight cryptographic primitive pseudorandom generator, VSAFL enables central servers and clients to prove and verify the correctness of model aggregations, significantly reducing verification costs. Our polynomial logarithmic overhead is particularly advantageous for clients with limited resources and high-dimensional gradients. Additionally, the proposed protocol is to be fully robust to clients dropping at any point. Through experimental evaluation, we demonstrate that VSAFL outperforms prior work in terms of verification speed by orders of magnitude.
- 26 August 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Internet of Things Journal
- Vol. 11 (23) , 38552-38568
VSAFL: Verifiable and Secure Aggregation With (Poly) Logarithmic Overhead in Federated Learning
Federated learning is a distributed machine learning framework that enables multiple participants to train models without directly sharing local data. However, sensitive information about participants may still be leaked through their gradients. Furthermore, centralized servers used for aggregating these gradients can be vulnerable to compromise, leading to privacy violations or other malicious attacks. Therefore, it is essential to verify the integrity of the aggregation. In this work, we focus on designing communication-efficient and fast verifiable aggregations for federated learning. We propose VSAFL, a verifiable secure aggregation protocol specifically designed for cross-device federated learning. VSAFL achieves computation and communication cost of O(log2n+log n + 1) and O(logn+l), respectively, for secure aggregation and verification for each user in each epoch, where n represents the number of clients and l denotes the dimension of the gradient vector. By employing a lightweight cryptographic primitive pseudorandom generator, VSAFL enables central servers and clients to prove and verify the correctness of model aggregations, significantly reducing verification costs. Our polynomial logarithmic overhead is particularly advantageous for clients with limited resources and high-dimensional gradients. Additionally, the proposed protocol is to be fully robust to clients dropping at any point. Through experimental evaluation, we demonstrate that VSAFL outperforms prior work in terms of verification speed by orders of magnitude.
- 25 November 2023
- Published by Elsevier in Journal of Systems Architecture
- Vol. 146
Flexible and secure access control for EHR sharing based on blockchain
- 25 November 2023
- Published by Elsevier in Journal of Systems Architecture
- Vol. 146
Flexible and secure access control for EHR sharing based on blockchain
- Open Access
- 29 May 2023
- Published by MDPI AG in Future Internet
- Vol. 15 (6) , 197
How Can We Achieve Query Keyword Frequency Analysis in Privacy-Preserving Situations?
Recently, significant progress has been made in the field of public key encryption with keyword search (PEKS), with a focus on optimizing search methods and improving the security and efficiency of schemes. Keyword frequency analysis is a powerful tool for enhancing retrieval services in explicit databases. However, designing a PEKS scheme that integrates keyword frequency analysis while preserving privacy and security has remained challenging, as it may conflict with some of the security principles of PEKS. In this paper, we propose an innovative scheme that introduces a security deadline to query trapdoors through the use of timestamps. This means that the keywords in the query trapdoor can only be recovered after the security deadline has passed. This approach allows for keyword frequency analysis of query keywords without compromising data privacy and user privacy, while also providing protection against keyword-guessing attacks through the dual-server architecture of our scheme. Moreover, our scheme supports multi-keyword queries in multi-user scenarios and is highly scalable. Finally, we evaluate the computational and communication efficiency of our scheme, demonstrating its feasibility in practical applications.
- 29 May 2023
- Published by MDPI AG in Future Internet
- Vol. 15 (6) , 197
Open AccessHow Can We Achieve Query Keyword Frequency Analysis in Privacy-Preserving Situations?
Recently, significant progress has been made in the field of public key encryption with keyword search (PEKS), with a focus on optimizing search methods and improving the security and efficiency of schemes. Keyword frequency analysis is a powerful tool for enhancing retrieval services in explicit databases. However, designing a PEKS scheme that integrates keyword frequency analysis while preserving privacy and security has remained challenging, as it may conflict with some of the security principles of PEKS. In this paper, we propose an innovative scheme that introduces a security deadline to query trapdoors through the use of timestamps. This means that the keywords in the query trapdoor can only be recovered after the security deadline has passed. This approach allows for keyword frequency analysis of query keywords without compromising data privacy and user privacy, while also providing protection against keyword-guessing attacks through the dual-server architecture of our scheme. Moreover, our scheme supports multi-keyword queries in multi-user scenarios and is highly scalable. Finally, we evaluate the computational and communication efficiency of our scheme, demonstrating its feasibility in practical applications.