Pulei Xiong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Pulei Xiong has a total of 44 co-authors. Here are the top co-authors by number of shared publications:
- Arash Habibi Lashkari
Behaviour-Centric Cybersecurity Center (BCCC), School of Information Technology, York University, Toronto, Canada
16 shared publications - Dilli Prasad Sharma
Department of Electrical and Computer Engineering, University of Toronto, Toronto, ON, Canada
15 shared publications - Mahdi Daghmehchi Firoozjaei
MacEwan University
14 shared publications - Samaneh Mahdavifar
McGill University
14 shared publications - Shahrear Iqbal
National Research Council Canada, Fredericton, Canada
5 shared publications - Xiaodong Lin
University of Guelph
5 shared publications
Latest Publications
- 18 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cloud Computing
- Vol. PP (3) , 1-12
PLOSQ3D: Privacy-Preserving Location-Based Skyline Query over Encrypted 3D Spatial Data
Privacy-preserving location-based skyline queries (LBSQ) enable location-based services (LBS) over outsourced spatial data while preserving data privacy. Prior studies primar ily consider Euclidean distance as the spatial-attribute metric. However, this does not capture reachability in three-dimensional (3D) obstacle environments, where obstacles invalidate direct Euclidean distances and necessitate obstacle-aware path computations. In this work, we present PLOSQ3D, a privacy-preserving location-based skyline query framework in 3D obstacle space. We first formalize the problem, specify its privacy requirements, and identify the challenges introduced by obstacle-aware distances. Subsequently, we introduce a novel index, Secure Augmented R-Tree (SAR-Tree) that jointly encodes spatial and non-spatial at tributes, enabling effective pruning under 3D spatial constraints. To enforce privacy, we leverage Function Secret Sharing (FSS) to design three protocols: SLDC for secure dominance computation, SSPC for secure shortest-path computation, and OLDU for oblivious skyline-set update. Building on them, we carefully design the PLOSQ3D scheme. Our security analysis demonstrates strong privacy guarantees against honest-but-curious adversaries. We further conduct extensive experiments on both synthetic and real 3D datasets. Experimental results show that our framework achieves high efficiency in both computation and communication overhead, confirming its practicality for 3D spatial applications that demand spatial-awareness and privacy protection.
- 18 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cloud Computing
- Vol. PP (3) , 1-12
PLOSQ3D: Privacy-Preserving Location-Based Skyline Query over Encrypted 3D Spatial Data
Privacy-preserving location-based skyline queries (LBSQ) enable location-based services (LBS) over outsourced spatial data while preserving data privacy. Prior studies primar ily consider Euclidean distance as the spatial-attribute metric. However, this does not capture reachability in three-dimensional (3D) obstacle environments, where obstacles invalidate direct Euclidean distances and necessitate obstacle-aware path computations. In this work, we present PLOSQ3D, a privacy-preserving location-based skyline query framework in 3D obstacle space. We first formalize the problem, specify its privacy requirements, and identify the challenges introduced by obstacle-aware distances. Subsequently, we introduce a novel index, Secure Augmented R-Tree (SAR-Tree) that jointly encodes spatial and non-spatial at tributes, enabling effective pruning under 3D spatial constraints. To enforce privacy, we leverage Function Secret Sharing (FSS) to design three protocols: SLDC for secure dominance computation, SSPC for secure shortest-path computation, and OLDU for oblivious skyline-set update. Building on them, we carefully design the PLOSQ3D scheme. Our security analysis demonstrates strong privacy guarantees against honest-but-curious adversaries. We further conduct extensive experiments on both synthetic and real 3D datasets. Experimental results show that our framework achieves high efficiency in both computation and communication overhead, confirming its practicality for 3D spatial applications that demand spatial-awareness and privacy protection.
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21, 7920-7935
Privacy-Preserving Reverse Spatial Skyline Query Scheme for Obstacle-Aware 3-D Space
- 1 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 21, 7920-7935
Privacy-Preserving Reverse Spatial Skyline Query Scheme for Obstacle-Aware 3-D Space
- Open Access
- 25 December 2025
- Published by Elsevier in Knowledge-Based Systems
- Vol. 335
Towards integration of privacy enhancing technologies in explainable artificial intelligence
- 25 December 2025
- Published by Elsevier in Knowledge-Based Systems
- Vol. 335
Open AccessTowards integration of privacy enhancing technologies in explainable artificial intelligence
- Open Access
Enhancing Adversarial Robustness of IoT Intrusion Detection via SHAP-Based Attribution Fingerprinting
The rapid proliferation of Internet of Things (IoT) devices has transformed numerous industries by enabling seamless connectivity and data-driven automation. However, this expansion has also exposed IoT networks to increasingly sophisticated security threats, including adversarial attacks targeting artificial intelligence (AI) and machine learning (ML)-based intrusion detection systems (IDS) to deliberately evade detection, induce misclassification, and systematically undermine the reliability and integrity of security defenses. To address these challenges, we propose a novel adversarial detection model that enhances the robustness of IoT IDS against adversarial attacks through SHapley Additive exPlanations (SHAP)-based fingerprinting. Using SHAP's DeepExplainer, we extract attribution fingerprints from network traffic features, enabling the IDS to reliably distinguish between clean and adversarially perturbed inputs. By capturing subtle attribution patterns, the model becomes more resilient to evasion attempts and adversarial manipulations. We evaluated the model on a standard IoT benchmark dataset, where it significantly outperformed a state-of-the-art method in detecting adversarial attacks. In addition to enhanced robustness, this approach improves model transparency and interpretability, thereby increasing trust in the IDS through explainable AI.
Open AccessEnhancing Adversarial Robustness of IoT Intrusion Detection via SHAP-Based Attribution Fingerprinting
The rapid proliferation of Internet of Things (IoT) devices has transformed numerous industries by enabling seamless connectivity and data-driven automation. However, this expansion has also exposed IoT networks to increasingly sophisticated security threats, including adversarial attacks targeting artificial intelligence (AI) and machine learning (ML)-based intrusion detection systems (IDS) to deliberately evade detection, induce misclassification, and systematically undermine the reliability and integrity of security defenses. To address these challenges, we propose a novel adversarial detection model that enhances the robustness of IoT IDS against adversarial attacks through SHapley Additive exPlanations (SHAP)-based fingerprinting. Using SHAP's DeepExplainer, we extract attribution fingerprints from network traffic features, enabling the IDS to reliably distinguish between clean and adversarially perturbed inputs. By capturing subtle attribution patterns, the model becomes more resilient to evasion attempts and adversarial manipulations. We evaluated the model on a standard IoT benchmark dataset, where it significantly outperformed a state-of-the-art method in detecting adversarial attacks. In addition to enhanced robustness, this approach improves model transparency and interpretability, thereby increasing trust in the IDS through explainable AI.
- Open Access
- 1 September 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 8, 6958-6966
Accelerating Adversarial Training on Under-Utilized GPU
Deep neural networks are vulnerable to adversarial attacks and adversarial training has been proposed to defend against such attacks by adaptively generating attacks, i.e., adversarial examples, during training. However, adversarial training is significantly slower than traditional training due to the search for worst attacks for each minibatch. To speed up adversarial training, existing work has considered a subset of a minibatch for generating attacks and reduced the steps in the search for attacks. We propose a novel adversarial training acceleration method, called AttackRider, by exploring under-utilized GPU hardware to reduce the number of calls to attack generation without increasing the time of each call. We characterize the extent of under-utilization of GPU for given GPU and model size, hence the potential for speedup, and present the application scenarios where this opportunity exists. The results on various machine learning tasks and datasets show that AttackRider can speed up state-of-the-art adversarial training algorithms with comparable robust accuracy. The source code of AttackRider is available at https://github.com/zxzhan/AttackRider.
- 1 September 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 8, 6958-6966
Open AccessAccelerating Adversarial Training on Under-Utilized GPU
Deep neural networks are vulnerable to adversarial attacks and adversarial training has been proposed to defend against such attacks by adaptively generating attacks, i.e., adversarial examples, during training. However, adversarial training is significantly slower than traditional training due to the search for worst attacks for each minibatch. To speed up adversarial training, existing work has considered a subset of a minibatch for generating attacks and reduced the steps in the search for attacks. We propose a novel adversarial training acceleration method, called AttackRider, by exploring under-utilized GPU hardware to reduce the number of calls to attack generation without increasing the time of each call. We characterize the extent of under-utilization of GPU for given GPU and model size, hence the potential for speedup, and present the application scenarios where this opportunity exists. The results on various machine learning tasks and datasets show that AttackRider can speed up state-of-the-art adversarial training algorithms with comparable robust accuracy. The source code of AttackRider is available at https://github.com/zxzhan/AttackRider.
- Top Cited PapersOpen Access
- 1 September 2025
- Published by KEAI Publishing in Journal of Information and Intelligence
- Vol. 3 (5) , 375-400
An efficient self attention-based 1D-CNN-LSTM network for IoT attack detection and identification using network traffic
- 1 September 2025
- Published by KEAI Publishing in Journal of Information and Intelligence
- Vol. 3 (5) , 375-400
Top Cited PapersOpen AccessAn efficient self attention-based 1D-CNN-LSTM network for IoT attack detection and identification using network traffic
- Open Access
Towards integration of Privacy Enhancing Technologies in Explainable Artificial Intelligence
Explainable Artificial Intelligence (XAI) is a crucial pathway in mitigating the risk of non-transparency in the decision-making process of black-box Artificial Intelligence (AI) systems. However, despite the benefits, XAI methods are found to leak the privacy of individuals whose data is used in training or querying the models. Researchers have demonstrated privacy attacks that exploit explanations to infer sensitive personal information of individuals. Currently there is a lack of defenses against known privacy attacks targeting explanations when vulnerable XAI are used in production and machine learning as a service system. To address this gap, in this article, we explore Privacy Enhancing Technologies (PETs) as a defense mechanism against attribute inference on explanations provided by feature-based XAI methods. We empirically evaluate 3 types of PETs, namely synthetic training data, differentially private training and noise addition, on two categories of feature-based XAI. Our evaluation determines different responses from the mitigation methods and side-effects of PETs on other system properties such as utility and performance. In the best case, PETs integration in explanations reduced the risk of the attack by 49.47%, while maintaining model utility and explanation quality. Through our evaluation, we identify strategies for using PETs in XAI for maximizing benefits and minimizing the success of this privacy attack on sensitive personal information.
Open AccessTowards integration of Privacy Enhancing Technologies in Explainable Artificial Intelligence
Explainable Artificial Intelligence (XAI) is a crucial pathway in mitigating the risk of non-transparency in the decision-making process of black-box Artificial Intelligence (AI) systems. However, despite the benefits, XAI methods are found to leak the privacy of individuals whose data is used in training or querying the models. Researchers have demonstrated privacy attacks that exploit explanations to infer sensitive personal information of individuals. Currently there is a lack of defenses against known privacy attacks targeting explanations when vulnerable XAI are used in production and machine learning as a service system. To address this gap, in this article, we explore Privacy Enhancing Technologies (PETs) as a defense mechanism against attribute inference on explanations provided by feature-based XAI methods. We empirically evaluate 3 types of PETs, namely synthetic training data, differentially private training and noise addition, on two categories of feature-based XAI. Our evaluation determines different responses from the mitigation methods and side-effects of PETs on other system properties such as utility and performance. In the best case, PETs integration in explanations reduced the risk of the attack by 49.47%, while maintaining model utility and explanation quality. Through our evaluation, we identify strategies for using PETs in XAI for maximizing benefits and minimizing the success of this privacy attack on sensitive personal information.
- 27 May 2025
- Published by Springer Nature
- p. 3-13
Why AI and Security?