Towards accurate anomaly detection for cloud system via graph-enhanced contrastive learning
Abstract As a critical technology, anomaly detection ensures the smooth operation of cloud systems while maintaining the market competitiveness of cloud service providers. However, the resource data in real-world cloud systems is predominantly unannotated, leading to insufficient supervised signals...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2024-11-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-024-01659-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571168856473600 |
|---|---|
| author | Zhen Zhang Zhe Zhu Chen Xu Jinyu Zhang Shaohua Xu |
| author_facet | Zhen Zhang Zhe Zhu Chen Xu Jinyu Zhang Shaohua Xu |
| author_sort | Zhen Zhang |
| collection | DOAJ |
| description | Abstract As a critical technology, anomaly detection ensures the smooth operation of cloud systems while maintaining the market competitiveness of cloud service providers. However, the resource data in real-world cloud systems is predominantly unannotated, leading to insufficient supervised signals for anomaly detection. Moreover, complicated topological associations existed between cloud servers (e.g., computation, storage, and communication). While acquiring resource information, correlating the system topology is challenging. To this end, we propose the GCAD for cloud system anomaly detection, which integrates data augmentation, GraphGRU, contrastive learning, and reconstruction. First, GCAD constructs positive and negative sample pairs through the masking and Gaussian noise data augmentation. Then, the GraphGRU processes extended temporal graph data, extracting and fusing spatiotemporal features from resource status and system topology. In addition, GCAD introduces linear attention for encoding spatiotemporal representations to capture their global correlation information. The weight parameters of the encoder are optimized using a contrastive learning mechanism. Finally, GCAD utilizes a reconstruction technique to calculate anomaly scores, facilitating the evaluation of the state of the cloud system at each time point. Experimental results indicate that GCAD outperforms state-of-the-art compared methods on two real-world datasets that contain topology information. |
| format | Article |
| id | doaj-art-6053aa3a9bb84cc283ac99a2664e6c2a |
| institution | Kabale University |
| issn | 2199-4536 2198-6053 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | Springer |
| record_format | Article |
| series | Complex & Intelligent Systems |
| spelling | doaj-art-6053aa3a9bb84cc283ac99a2664e6c2a2025-02-02T12:49:10ZengSpringerComplex & Intelligent Systems2199-45362198-60532024-11-0111111910.1007/s40747-024-01659-xTowards accurate anomaly detection for cloud system via graph-enhanced contrastive learningZhen Zhang0Zhe Zhu1Chen Xu2Jinyu Zhang3Shaohua Xu4School of Computer Science and Engineering, Shandong University of Science and TechnologyShandong Artificial Intelligence Institute, Qilu University of Technology (Shandong Academy of Sciences)School of Computer Science, Fudan UniversitySchool of Computer Science and Engineering, Shandong University of Science and TechnologySchool of Computer Science and Engineering, Shandong University of Science and TechnologyAbstract As a critical technology, anomaly detection ensures the smooth operation of cloud systems while maintaining the market competitiveness of cloud service providers. However, the resource data in real-world cloud systems is predominantly unannotated, leading to insufficient supervised signals for anomaly detection. Moreover, complicated topological associations existed between cloud servers (e.g., computation, storage, and communication). While acquiring resource information, correlating the system topology is challenging. To this end, we propose the GCAD for cloud system anomaly detection, which integrates data augmentation, GraphGRU, contrastive learning, and reconstruction. First, GCAD constructs positive and negative sample pairs through the masking and Gaussian noise data augmentation. Then, the GraphGRU processes extended temporal graph data, extracting and fusing spatiotemporal features from resource status and system topology. In addition, GCAD introduces linear attention for encoding spatiotemporal representations to capture their global correlation information. The weight parameters of the encoder are optimized using a contrastive learning mechanism. Finally, GCAD utilizes a reconstruction technique to calculate anomaly scores, facilitating the evaluation of the state of the cloud system at each time point. Experimental results indicate that GCAD outperforms state-of-the-art compared methods on two real-world datasets that contain topology information.https://doi.org/10.1007/s40747-024-01659-xCloud computingAnomaly detectionGraph neural networkContrastive learningLinear attention mechanism |
| spellingShingle | Zhen Zhang Zhe Zhu Chen Xu Jinyu Zhang Shaohua Xu Towards accurate anomaly detection for cloud system via graph-enhanced contrastive learning Complex & Intelligent Systems Cloud computing Anomaly detection Graph neural network Contrastive learning Linear attention mechanism |
| title | Towards accurate anomaly detection for cloud system via graph-enhanced contrastive learning |
| title_full | Towards accurate anomaly detection for cloud system via graph-enhanced contrastive learning |
| title_fullStr | Towards accurate anomaly detection for cloud system via graph-enhanced contrastive learning |
| title_full_unstemmed | Towards accurate anomaly detection for cloud system via graph-enhanced contrastive learning |
| title_short | Towards accurate anomaly detection for cloud system via graph-enhanced contrastive learning |
| title_sort | towards accurate anomaly detection for cloud system via graph enhanced contrastive learning |
| topic | Cloud computing Anomaly detection Graph neural network Contrastive learning Linear attention mechanism |
| url | https://doi.org/10.1007/s40747-024-01659-x |
| work_keys_str_mv | AT zhenzhang towardsaccurateanomalydetectionforcloudsystemviagraphenhancedcontrastivelearning AT zhezhu towardsaccurateanomalydetectionforcloudsystemviagraphenhancedcontrastivelearning AT chenxu towardsaccurateanomalydetectionforcloudsystemviagraphenhancedcontrastivelearning AT jinyuzhang towardsaccurateanomalydetectionforcloudsystemviagraphenhancedcontrastivelearning AT shaohuaxu towardsaccurateanomalydetectionforcloudsystemviagraphenhancedcontrastivelearning |