Guoqang Deng

Dr. Guoqang Deng

orcid

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 July 2022 in IEEE Transactions on Services Computing
Verified
Journal Article
Closed access
GDGuoqang Deng
Guoqang Deng
MTMin Tang
Min Tang
YXYuxing Xi
Yuxing Xi
MZMingwu Zhang
Mingwu Zhang
Online medical prediagnosis systems have already shown great achievement in providing the guidance of healthcare services with lower time and cost. Achieving a high-precision medical primary diagnosis system faces many severe challenges on the privacy of individual health information, the distributed storage of medical data and the diversity of the disease. In this paper, we propose an efficient and privacy-preserving framework for obtaining a pre-clinical guide model, which allows an authorized data analysis center to train a disease classifier using a combination of medical data gathered from different entities. Our proposed scheme is based on soft-margin support vector machine (SVM) which takes Taylor polynomial of exponential-loss as penalty. Our scheme achieves the following advantages: the trained model can tolerate some abnormal samples therefore has higher generalization ability, and the training process can constraint the inefficient operations in the encrypted domain thus leads to the availability of partial homomorphic encryption system. Lately, we prove that the proposed scheme achieves the goal of medical prediagnosis system construction and data without privacy leakage to data analysis center and model parameters without exposure to data providers, as well as demonstrating its utility and efficiency using real-world medical datasets.

5 April 2023 in Future Generation Computer Systems
Verified
Journal Article
Closed access
GDGuoqang Deng
Guoqang Deng
XDXuefeng Duan
Xuefeng Duan
MTMin Tang
Min Tang
YZYuhao Zhang
Yuhao Zhang
YHYing Huang
Ying Huang

23 September 2024 in IEEE Transactions on Intelligent Transportation Systems
Verified
Journal Article
Closed access
MTMin Tang
Min Tang
ZHZengyi Huang
Zengyi Huang
GDGuoqang Deng
Guoqang Deng
Deep learning is increasingly utilized in data-driven tasks in Vehicular Ad-hoc Networks (VANETs) such as traffic sign recognition or pedestrian detection, and is expected to fulfill autonomous driving. These applications involve large amounts of discrete private data. Federated learning can effectively address the issues of data islands and security. However, it is challenging to apply centralized framework to meet the rigorous requirements of intelligent services in VANETs, such as ultra-low latency, high security and reliability. In this work, we propose a privacy-preserving deep learning scheme named FEDL, which is built on a decentralized framework (abbreviated as EDVAN). This customized EDVAN structure decomposes computing tasks into each edge computing node equally. By combining the Voronoi diagram the network communication delay and overhead are alleviated tremendously. Furthermore, the FEDL scheme utilizes functional encryption to establish secure data communication and drop-allowance parameter aggregation module, achieving to resist inference attacks and eliminate the straggler effect. Finally, we apply the NS-3 simulator to simulate EDVAN’s communication environment. Three real traffic datasets are utilized to show the FEDL model is accurate lossless and efficient. Our code is available at https://github.com/HuangZengyi/FEDL.


4 December 2024 in IEEE Transactions on Consumer Electronics
Verified
Journal Article
Closed access
ZHZengyi Huang
Zengyi Huang
MTMin Tang
Min Tang
YWYuxing Wei
Yuxing Wei
GDGuoqang Deng
Guoqang Deng
As the Internet of Things progresses, the volume of information stored in electronic devices has surged dramatically. Employing Graph Convolutional Networks (GCNs) to model graph data derived from link relationships has led to widespread applications in consumer electronics, such as user preference analysis and personalized recommendations. However, with practical graph data scattered among various owners, traditional centralized GCN methods falter owing to privacy constraints. Thus, creating a distributed GCN framework that upholds model quality and user privacy remains a daunting task. In this paper, we pioneer the use of functional encryption in a federated learning GCN framework, termed FPGCN. This innovative technique achieves performance parity with centralized GCNs, capable of producing equivalent node representations without the need for direct access to raw data. By employing a hierarchical split learning framework coupled with privacy-preserving aggregation mechanisms, it enables secure and collaborative multi-party GCN training, ensuring confidentiality while maintaining model efficacy. Comprehensive experimental results in three real-world datasets demonstrate the capacity of FPGCN to match the predictive accuracy observed in centralized scenarios, while also showcasing its performance efficiency.

14 December 2022 in Applied Sciences
Verified
Journal Article
Open access
PDF file
GDGuoqang Deng
Guoqang Deng
MTMin Tang
Min Tang
YZYuhao Zhang
Yuhao Zhang
YHYing Huang
Ying Huang
XDXuefeng Duan
Xuefeng Duan
Artificial neural network (ANN) is powerful in the artificial intelligence field and has been successfully applied to interpret complex image data in the real world. Since the majority of images are commonly known as private with the information intended to be used by the owner, such as handwritten characters and face, the private constraints form a major obstacle in developing high-precision image classifiers which require access to a large amount of image data belonging to multiple users. State-of-the-art privacy-preserving ANN schemes often use full homomorphic encryption which result in a substantial overhead of computation and data traffic for the data owners, and are restricted to approximation models by low-degree polynomials which lead to a large accuracy loss of the trained model compared to the original ANN model in the plain domain. Consequently, it is still a huge challenge to train an ANN model in the encrypted-domain. To mitigate this problem, we propose a privacy-preserving ANN system for secure constructing image classifiers, named IPPNN, where the server is able to train an ANN-based classifier on the combined image data of all data owners without being able to observe any images using primitives, such as randomization and functional encryption. Our system achieves faster training time and supports lossless training. Moreover, IPPNN removes the need for multiple communications among data owners and servers. We analyze the security of the protocol and perform experiments on a large scale image recognition task. The results show that the IPPNN is feasible to use in practice while achieving high accuracy.