Pengfei Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Pengfei Liu has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Hong Song
School of Computer Science and engineering, Central South University, Changsha, China
2 shared publications - Shigeng Zhang
Central South University
2 shared publications - Weiping Wang
School of Computer Science and engineering, Central South University, Changsha, China
2 shared publications - Congmin Ren
School of Computer Science and Engineering
1 shared publication - Guangquan Xu
Tianjin University
1 shared publication Jian Sun
1 shared publication
Latest Publications
- Open Access
- 1 January 2025
- Published by Institution of Engineering and Technology (IET) in IET Information Security
- Vol. 2025 (1)
AEDroid: Adaptive Enhanced Android Malware Detection‐Based on Interpretability of Deep Learning
As the most widely used operating system in the world, Android has naturally become the main target of malicious hackers. The current research on Android malware detection relies on manually defined sensitive API feature sets. With the continuous innovation and change of malicious behavior, new threats and attack methods have emerged. If we still rely on the original sensitive API set, malicious applications will not be discovered. To address this issue, we do not use the existing sensitive API feature set but instead design a key activation mechanism (KAM) based on convolutional neural networks (CNNs) to obtain sensitive API. We use this mechanism to automatically mine API features that play an important role in determining maliciousness from application datasets. And we use the API group (ApiG) obtained through this mechanism for template generalization, and obtain a method called AEDroid that can delay model aging. By analyzing these API features, it was found that they not only cover the existing sensitive API feature types but also include sensitive APIs for seven new types of malicious behavior. The experimental results show that with the addition of the newly discovered sensitive API, the Android malware detection rate has increased by more than 5%, especially on newly emerged malicious datasets, where the effect is more pronounced.
- 1 January 2025
- Published by Institution of Engineering and Technology (IET) in IET Information Security
- Vol. 2025 (1)
Open AccessAEDroid: Adaptive Enhanced Android Malware Detection‐Based on Interpretability of Deep Learning
As the most widely used operating system in the world, Android has naturally become the main target of malicious hackers. The current research on Android malware detection relies on manually defined sensitive API feature sets. With the continuous innovation and change of malicious behavior, new threats and attack methods have emerged. If we still rely on the original sensitive API set, malicious applications will not be discovered. To address this issue, we do not use the existing sensitive API feature set but instead design a key activation mechanism (KAM) based on convolutional neural networks (CNNs) to obtain sensitive API. We use this mechanism to automatically mine API features that play an important role in determining maliciousness from application datasets. And we use the API group (ApiG) obtained through this mechanism for template generalization, and obtain a method called AEDroid that can delay model aging. By analyzing these API features, it was found that they not only cover the existing sensitive API feature types but also include sensitive APIs for seven new types of malicious behavior. The experimental results show that with the addition of the newly discovered sensitive API, the Android malware detection rate has increased by more than 5%, especially on newly emerged malicious datasets, where the effect is more pronounced.
- Open Access
- 7 April 2023
- Published by Wiley in Security and Communication Networks
- Vol. 2023, 1-11
ImageDroid: Using Deep Learning to Efficiently Detect Android Malware and Automatically Mark Malicious Features
The popularity of the Android platform has led to an explosion in malware. The current research on Android malware mainly focuses on malware detection or malware family classification. These studies need to extract a large number of features, which consumes a lot of manpower and material resources. Moreover, some malware use obfuscation to evade decompiler tools extracting features. To address these problems, we propose ImageDroid, a method based on the image format of Android applications that can not only detect and classify malware without prior knowledge but also detect the obfuscated malware. Furthermore, we utilize the Grad-CAM interpretable mechanism of the deep learning model to automatically label the image that play a key role in determining maliciousness in a visual way. We evaluate ImageDroid over 10,000 Android applications. Experimental results show that the accuracy of malicious detection and multifamily classification achieve 97.2% and 95.1%, respectively, and the detection accuracy of obfuscated malware achieves 94.6%.
- 7 April 2023
- Published by Wiley in Security and Communication Networks
- Vol. 2023, 1-11
Open AccessImageDroid: Using Deep Learning to Efficiently Detect Android Malware and Automatically Mark Malicious Features
The popularity of the Android platform has led to an explosion in malware. The current research on Android malware mainly focuses on malware detection or malware family classification. These studies need to extract a large number of features, which consumes a lot of manpower and material resources. Moreover, some malware use obfuscation to evade decompiler tools extracting features. To address these problems, we propose ImageDroid, a method based on the image format of Android applications that can not only detect and classify malware without prior knowledge but also detect the obfuscated malware. Furthermore, we utilize the Grad-CAM interpretable mechanism of the deep learning model to automatically label the image that play a key role in determining maliciousness in a visual way. We evaluate ImageDroid over 10,000 Android applications. Experimental results show that the accuracy of malicious detection and multifamily classification achieve 97.2% and 95.1%, respectively, and the detection accuracy of obfuscated malware achieves 94.6%.
- Open Access
- 28 April 2022
- Published by Wiley in Security and Communication Networks
- Vol. 2022, 1-11
FGL_Droid: An Efficient Android Malware Detection Method Based on Hybrid Analysis
With the popularity of Android intelligent terminals, malicious applications targeting Android platform are growing rapidly. Therefore, efficient and accurate detection of Android malicious software becomes particularly important. Dynamic API call sequences are widely used in Android malware detection because they can reflect the behaviours of applications accurately. However, the raw dynamic API call sequences are very usually too long to be directly used, and most existing works just use a truncated segment of the sequence or statistical features of the sequence to perform malware detection, which loses the execution order information of applications and consequently results in high false alarm rate. In this work, we propose a method that transforms the dynamic API call sequence into a function call graph, which retains most of the application execution order information with significantly reduced sequence size. To compensate for the missed behaviour information during the transformation, the advanced features of permission requests extracted from the application are utilized. We then propose FGL_Droid, which fusions the transformed function call graph feature and the extracted permission request feature to perform accurate malware detection. Experiments on benchmark dataset show that FGL_Droid achieves a high detection accuracy of 0.975 and a high F-score of 0.978, which are better than the existing methods.
- 28 April 2022
- Published by Wiley in Security and Communication Networks
- Vol. 2022, 1-11
Open AccessFGL_Droid: An Efficient Android Malware Detection Method Based on Hybrid Analysis
With the popularity of Android intelligent terminals, malicious applications targeting Android platform are growing rapidly. Therefore, efficient and accurate detection of Android malicious software becomes particularly important. Dynamic API call sequences are widely used in Android malware detection because they can reflect the behaviours of applications accurately. However, the raw dynamic API call sequences are very usually too long to be directly used, and most existing works just use a truncated segment of the sequence or statistical features of the sequence to perform malware detection, which loses the execution order information of applications and consequently results in high false alarm rate. In this work, we propose a method that transforms the dynamic API call sequence into a function call graph, which retains most of the application execution order information with significantly reduced sequence size. To compensate for the missed behaviour information during the transformation, the advanced features of permission requests extracted from the application are utilized. We then propose FGL_Droid, which fusions the transformed function call graph feature and the extracted permission request feature to perform accurate malware detection. Experiments on benchmark dataset show that FGL_Droid achieves a high detection accuracy of 0.975 and a high F-score of 0.978, which are better than the existing methods.