Hai Liang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Hai Liang has a total of 47 co-authors. Here are the top co-authors by number of shared publications:
Yong Ding
Guilin University of Electronic Technology
10 shared publicationsYujue Wang
Hangzhou Innovation Institute of Beihang University
8 shared publicationsChangsong Yang
Guilin University of Electronic Technology
7 shared publicationsHuiyong Wang
GuilinUniversity of Electronic Technology
6 shared publications- Yong Ding
Guilin University of Electronic Technology
5 shared publications - Changsong Yang
Guangxi Key Laboratory of Cryptography and Information Security, Guilin University of Electronic Technology, Guilin, China
3 shared publications
Latest Publications
- 10 August 2025
- Published by Springer Nature
- p. 161-180
FRBFT: A BFT Consensus Protocol Supporting Fault Removal for Industrial Internet of Things
- 10 August 2025
- Published by Springer Nature
- p. 161-180
FRBFT: A BFT Consensus Protocol Supporting Fault Removal for Industrial Internet of Things
- 1 May 2025
- Published by Springer Nature
- p. 19-34
A Novel Network Security Situation Prediction Scheme for IIoT by Using LGF-Net
- 1 May 2025
- Published by Springer Nature
- p. 19-34
A Novel Network Security Situation Prediction Scheme for IIoT by Using LGF-Net
- 28 January 2025
- Published by Springer Nature
- p. 111-124
A Cross-chain Model Based on Credit Hierarchical Notary Group for IoT Roaming Settlement
- 28 January 2025
- Published by Springer Nature
- p. 111-124
A Cross-chain Model Based on Credit Hierarchical Notary Group for IoT Roaming Settlement
- Open Access
- 13 August 2024
- Published by Springer Nature in Research Square
Deep Learning Model for Intrusion Detection Based on Weighted Cross Entropy Loss Function and OBiLSTM
Industrial Internet of Things has become an important tool to promote the transformation and upgrading of the manufacturing industry and enhance it's competitiveness. With the popularization of IIOT, the cyber security threats faced by industrial systems are also increasing, and intrusion detection techniques have emerged. However, due to the complex structure of IIOT and the variety of network traffic data generated, the Intrusion detection that has been studied accuracy is not high. To improve the accuracy of intrusion detection, we introduce an intrusion detection method for IIoT utilizing an Optimized Bidirectional Long Short-Term Memory network. We first enhance BiLSTM by substituting the inverse Long Short-term Memory network with a Gated Recurrent Unit, followed by designing a novel neural network, OBiLSTM. We also designed a weighting method for the Cross-entropy Loss function, which makes the detection accuracy further improved. Finally, We implemented the proposed intrusion detection prototype and tested it based on two datasets, UNSW-NB15 and NSL-KDD. The experimental outcomes indicate that our proposed prototype surpasses previous methods in terms of accuracy, recall, precision, and F1_score.
- 13 August 2024
- Published by Springer Nature in Research Square
Open AccessDeep Learning Model for Intrusion Detection Based on Weighted Cross Entropy Loss Function and OBiLSTM
Industrial Internet of Things has become an important tool to promote the transformation and upgrading of the manufacturing industry and enhance it's competitiveness. With the popularization of IIOT, the cyber security threats faced by industrial systems are also increasing, and intrusion detection techniques have emerged. However, due to the complex structure of IIOT and the variety of network traffic data generated, the Intrusion detection that has been studied accuracy is not high. To improve the accuracy of intrusion detection, we introduce an intrusion detection method for IIoT utilizing an Optimized Bidirectional Long Short-Term Memory network. We first enhance BiLSTM by substituting the inverse Long Short-term Memory network with a Gated Recurrent Unit, followed by designing a novel neural network, OBiLSTM. We also designed a weighting method for the Cross-entropy Loss function, which makes the detection accuracy further improved. Finally, We implemented the proposed intrusion detection prototype and tested it based on two datasets, UNSW-NB15 and NSL-KDD. The experimental outcomes indicate that our proposed prototype surpasses previous methods in terms of accuracy, recall, precision, and F1_score.
- 23 February 2024
- Published by Springer Nature
- p. 167-186
Efficient and Revocable Anonymous Account Guarantee System Based on Blockchain
- 23 February 2024
- Published by Springer Nature
- p. 167-186
Efficient and Revocable Anonymous Account Guarantee System Based on Blockchain
- 23 February 2024
- Published by Springer Nature
- p. 321-340
A Lightweight PUF-Based Group Authentication Scheme for Privacy-Preserving Metering Data Collection in Smart Grid
- 23 February 2024
- Published by Springer Nature
- p. 321-340
A Lightweight PUF-Based Group Authentication Scheme for Privacy-Preserving Metering Data Collection in Smart Grid
- 25 November 2023
- Published by Springer Nature
- p. 242-255
Sharding Technologies in Blockchain: Basics, State of the Art, and Challenges
- 25 November 2023
- Published by Springer Nature
- p. 242-255
Sharding Technologies in Blockchain: Basics, State of the Art, and Challenges
- Open Access
- 10 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 29-35
Blockchain-based Secure Key Distribution for WSNs in Soil Monitoring
With the rapid development of society and economy, soil issues have been emphasized. As the front-end of data collection, sensor networks are used in various soil monitoring applications. However, current wireless sensor networks face the difficulty of sensor authentication when forming clusters, and lack effective authentication methods and efficient and secure key distribution mechanisms. To address these issues, this paper proposes a blockchain-based polynomial key distribution scheme (PCD). During the initialization process, the identification information of the sensor is uploaded into the blockchain to ensure the consistency of the authentication information and provide security for the authentication process. The polynomial-based distribution mechanism further improves the efficiency of cluster key distribution. Experimental analysis and comparison prove that our construction is efficient and practical.
- 10 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 29-35
Open AccessBlockchain-based Secure Key Distribution for WSNs in Soil Monitoring
With the rapid development of society and economy, soil issues have been emphasized. As the front-end of data collection, sensor networks are used in various soil monitoring applications. However, current wireless sensor networks face the difficulty of sensor authentication when forming clusters, and lack effective authentication methods and efficient and secure key distribution mechanisms. To address these issues, this paper proposes a blockchain-based polynomial key distribution scheme (PCD). During the initialization process, the identification information of the sensor is uploaded into the blockchain to ensure the consistency of the authentication information and provide security for the authentication process. The polynomial-based distribution mechanism further improves the efficiency of cluster key distribution. Experimental analysis and comparison prove that our construction is efficient and practical.
- Open Access
- 10 July 2023
- Published by Emerald Publishing in International Journal of Web Information Systems
- Vol. 19 (2) , 61-79
Output regeneration defense against membership inference attacks for protecting data privacy
Purpose: Recently, deep learning (DL) has been widely applied in various aspects of human endeavors. However, studies have shown that DL models may also be a primary cause of data leakage, which raises new data privacy concerns. Membership inference attacks (MIAs) are prominent threats to user privacy from DL model training data, as attackers investigate whether specific data samples exist in the training data of a target model. Therefore, the aim of this study is to develop a method for defending against MIAs and protecting data privacy. Design/methodology/approach: One possible solution is to propose an MIA defense method that involves adjusting the model’s output by mapping the output to a distribution with equal probability density. This approach effectively preserves the accuracy of classification predictions while simultaneously preventing attackers from identifying the training data. Findings: Experiments demonstrate that the proposed defense method is effective in reducing the classification accuracy of MIAs to below 50%. Because MIAs are viewed as a binary classification model, the proposed method effectively prevents privacy leakage and improves data privacy protection. Research limitations/implications: The method is only designed to defend against MIA in black-box classification models. Originality/value: The proposed MIA defense method is effective and has a low cost. Therefore, the method enables us to protect data privacy without incurring significant additional expenses.
- 10 July 2023
- Published by Emerald Publishing in International Journal of Web Information Systems
- Vol. 19 (2) , 61-79
Open AccessOutput regeneration defense against membership inference attacks for protecting data privacy
Purpose: Recently, deep learning (DL) has been widely applied in various aspects of human endeavors. However, studies have shown that DL models may also be a primary cause of data leakage, which raises new data privacy concerns. Membership inference attacks (MIAs) are prominent threats to user privacy from DL model training data, as attackers investigate whether specific data samples exist in the training data of a target model. Therefore, the aim of this study is to develop a method for defending against MIAs and protecting data privacy. Design/methodology/approach: One possible solution is to propose an MIA defense method that involves adjusting the model’s output by mapping the output to a distribution with equal probability density. This approach effectively preserves the accuracy of classification predictions while simultaneously preventing attackers from identifying the training data. Findings: Experiments demonstrate that the proposed defense method is effective in reducing the classification accuracy of MIAs to below 50%. Because MIAs are viewed as a binary classification model, the proposed method effectively prevents privacy leakage and improves data privacy protection. Research limitations/implications: The method is only designed to defend against MIA in black-box classification models. Originality/value: The proposed MIA defense method is effective and has a low cost. Therefore, the method enables us to protect data privacy without incurring significant additional expenses.
- Open Access
- 30 May 2023
- Published by Wiley in Security and Communication Networks
- Vol. 2023, 1-10
An Anomaly Detection Approach Based on Integrated LSTM for IoT Big Data
Due to the expanding scope of Industry 4.0, the Internet of Things has become an important element of the information age. Cyber security relies heavily on intrusion detection systems for Internet of Things (IoT) devices. In the face of complex network data and diverse intrusion methods, today’s network security environment requires more suitable machine learning methods to meet its security needs, and the current machine learning methods are hardly competent. In part because of network attacks by intruders using cutting-edge techniques and the constrained environment of IoT devices themselves, the most widely used algorithms in recent years include CNN and LSTM, with the former being particularly good at extracting features from the original data space and the latter concentrating more on temporal features of the data. We aim to address the issue of merging spatial and temporal variables in intrusion detection models by introducing a fusion model CNN and C-LSTM in this paper. Fusion features enhanced parallelism in the training process and better results without a very deep network, giving the model a shorter training time, fast convergence, and computational speed for emerging resource-limited network entities. This model is more suitable for anomaly detection tasks in the resource-constrained and time-sensitive big data environment of the Internet of Things. KDDCup-99, a publicly available IBD dataset, was applied in our experiments to demonstrate the model’s validity. In comparison to existing deep learning implementations, our proposed multiclass classification model delivers higher accuracy, precision, and recall.
- 30 May 2023
- Published by Wiley in Security and Communication Networks
- Vol. 2023, 1-10
Open AccessAn Anomaly Detection Approach Based on Integrated LSTM for IoT Big Data
Due to the expanding scope of Industry 4.0, the Internet of Things has become an important element of the information age. Cyber security relies heavily on intrusion detection systems for Internet of Things (IoT) devices. In the face of complex network data and diverse intrusion methods, today’s network security environment requires more suitable machine learning methods to meet its security needs, and the current machine learning methods are hardly competent. In part because of network attacks by intruders using cutting-edge techniques and the constrained environment of IoT devices themselves, the most widely used algorithms in recent years include CNN and LSTM, with the former being particularly good at extracting features from the original data space and the latter concentrating more on temporal features of the data. We aim to address the issue of merging spatial and temporal variables in intrusion detection models by introducing a fusion model CNN and C-LSTM in this paper. Fusion features enhanced parallelism in the training process and better results without a very deep network, giving the model a shorter training time, fast convergence, and computational speed for emerging resource-limited network entities. This model is more suitable for anomaly detection tasks in the resource-constrained and time-sensitive big data environment of the Internet of Things. KDDCup-99, a publicly available IBD dataset, was applied in our experiments to demonstrate the model’s validity. In comparison to existing deep learning implementations, our proposed multiclass classification model delivers higher accuracy, precision, and recall.