Keywords
non-stationary time series
performance anomaly detection
Abstract
Large-scale cloud service platforms generate continuous performance time series from CPU utilization, memory usage, request latency, disk I/O, network throughput, queue length, and service error logs. These metrics are strongly coupled, but their distributions change frequently due to workload migration, elastic scaling, software release cycles, and diurnal user demand. This study proposes a cross-metric dependency fusion model for non-stationary cloud service performance anomaly detection. The method constructs a dynamic metric-dependency matrix through window-level mutual information and temporal attention. A multi-scale sequence encoder is used to capture short-term spikes, periodic workload changes, and long-range service degradation patterns. To reduce false alerts under normal workload shifts, a regime-conditioned residual scoring module adjusts anomaly thresholds according to recent workload states. Experiments are conducted on a cloud service monitoring dataset containing 2,460 microservice instances, 186 service groups, and 96 performance metrics collected at one-minute intervals over 120 days. The dataset contains 169 million timestamped metric records and 2,340 labeled incidents, including memory leakage, cascading latency, abnormal queue accumulation, failed deployment, and disk saturation. The proposed model reduces median incident detection delay from 14.2 minutes to 4.8 minutes compared with a static Transformer reconstruction model. The false alert volume decreases to 2.6 alerts per service group per week. The model reaches a Matthews correlation coefficient of 0.921 and a Cohen’s kappa of 0.903 across mixed workload regimes. Online inference processes 38,000 metric points per second, with a median scoring latency of 24 ms per service group. The results show that cross-metric dependency fusion can improve cloud performance anomaly detection under non-stationary service conditions.
References
1. Qi, C., & Qiao, X. (2026, May). Design Patterns for Multi-Agent Systems in Production. In 2026 International Conference on Intelligent Systems, Automation and Control (ISAC) (pp. 198-202). IEEE.
2. Shamsi, N., & Helmrich, A. (2025). Interdependency classification: a framework for infrastructure resilience. Environmental Research: Infrastructure and Sustainability, 5(1), 015009.
3. Su, D., & Shi, X. Optimizing Family–School Collaboration to Improve Educational and Social Outcomes for Children with Intellectual Disabilities in Inclusive Public Schools.
4. Yahya, M. A., Moya, A. R., & Ventura, S. (2025). Deep learning for multivariate time series anomaly detection: An evaluation of reconstruction-based methods. Artificial Intelligence Review, 58(12), 400.
5. Chen, X., Xiao, H., Zeng, Z., Zhang, S., & Xiao, R. (2025). Fine-Grained Multivariate Time Series Anomaly Detection via Causal Inference. Knowledge-Based Systems, 114765.
6. Ho, T. K. K., Karami, A., & Armanfard, N. (2025). Graph anomaly detection in time series: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence.
7. Shao, W. (2026, March). Interpretable Ensemble Learning for Network Traffic Anomaly Detection: A SHAP-based Explainable AI Framework for Embedded Systems Security. In 2026 14th International Symposium on Digital Forensics and Security (ISDFS) (pp. 1-4). IEEE.
8. Gao, G., Gao, R., Lu, C., Gao, R., & Kuang, Y. (2026, March). Security Governance Methods and Quantitative Evaluation for Enterprise SMS and Verification Code Systems. In 2026 International Conference on Generative Artificial Intelligence and Information Security (GAIIS) (pp. 455-458). IEEE.
9. Khanizadeh, F., Ettefaghian, A., Wilson, G., Shirazibeheshti, A., Radwan, T., & Luca, C. (2025). Smart data-driven medical decisions through collective and individual anomaly detection in healthcare time series. International Journal of Medical Informatics, 194, 105696.
10. Xu, T., Zhu, W., & Zhang, J. (2026). A Study on the Application of Alternative Data in Credit Assessment for the Unbanked Population.
11. Fu, Y., Gui, H., Li, W., & Wang, Z. (2020, August). Virtual Material Modeling and Vibration Reduction Design of Electron Beam Imaging System. In 2020 IEEE International Conference on Advances in Electrical Engineering and Computer Applications (AEECA) (pp. 1063-1070). IEEE.
12. Pelosi, D., Cacciagrano, D., & Piangerelli, M. (2025). Explainability and interpretability in concept and data drift: a systematic literature review. Algorithms, 18(7), 443.
13. Chen, F., Yue, L., Xu, P., Liang, H., & Li, S. (2025). Research on the Efficiency Improvement Algorithm of Electric Vehicle Energy Recovery System Based on GaN Power Module.
14. Liang, R., Feifan, F. N. U., Liang, Y., & Ye, Z. (2025). Emotion-Aware Interface Adaptation in Mobile Applications Based on Color Psychology and Multimodal User State Recognition. Frontiers in Artificial Intelligence Research, 2(1), 51-57.
15. Gajanin, R., Marcelino, C., & Nastic, S. (2025). Performance Isolation for Serverless Functions. IEEE Transactions on Services Computing.
16. Wang, Y., Chen, J., Arias, R., Wang, Y., & Yin, X. (2026). Development and Validation of a Patient-Friendly Digital Assessment Platform for Precision Screening of Oral Anti-Obesity Medications (AOMs).
17. Jiao, Y., Zhao, B., Wang, A., & Shi, T. (2026). Construction and Empirical Study of a Modularized Teaching System for Art Courses Based on a Unified Training Pathway.
18. Ram, A., Chakraborty, S. K., Banerjee, A., & Mahato, G. K. (2025). Formulating and analysis of traffic flow to secure software-defined network (SDN) using recursive network (RN) learning method: A. Ram et al. The Journal of Supercomputing, 81(5), 654.
19. Yin, J., Rao, H., & Huang, Y. (2026, March). Dynamic Modeling and Heterogeneity Analysis of Platform User Behavior Time Series. In 2026 International Conference on AI in Education Technology and Applications (AIETA) (pp. 30-33). IEEE.
20. Shakil, S. M., Hossain, A., & Rahman, M. M. (2025). An empirical evaluation of anomaly detection techniques in smart grid systems using real-time operational data. American Journal of Advanced Technology and Engineering Solutions, 1(02), 95-134.
21. Yin, J., Huang, Y., & Rao, H. (2026). A Study on the Dynamic Evolution of Learning Behavior and Outcome Prediction in Digital Educational Environments. Available at SSRN 6607139.
22. Yang, J. (2026). Computational Analysis of How Digital Non-Clinical Communication Tools Influence Social Participation Among Older Adults in Community-Based Elderly Care. Available at SSRN 6682838.
23. Tariq, S., Baruwal Chhetri, M., Nepal, S., & Paris, C. (2025). Alert fatigue in security operations centres: Research challenges and opportunities. ACM Computing Surveys, 57(9), 1-38.
24. Zhang, Z. (2026). A Study on Multimodal Product Understanding and Assembly Guidance Optimization in e-commerce for Complex Consumer Goods. Available at SSRN 6734559.
25. Bittla, S. R. (2025). Predicting Failures with AI/ML Analytics. In AI-Driven Software Testing: Transforming Software Testing with Artificial Intelligence and Machine Learning (pp. 435-454). Berkeley, CA: Apress.
26. Johnphill, O., Sadiq, A. S., Kaiwartya, O., & Taheir, M. A. (2026). Cross: a cloud-native approach to automated remediation and self-healing in cyber-physical systems. Cybersecurity, 9(1), 124.
27. Zhang, Z., Gao, Y., & Tong, Y. (2026). CausalML4CX: A Causal Inference and Machine Learning Framework for Customer Experience Optimization with Application in Banking.
28. Qi, C., & Qiao, X. (2026). Building and Operating a Large Scale Multi-Agent System: A Case Study from Industry. Available at SSRN 6795198.
29. Korostenskyi, R., & Olenych, I. (2025, October). EDGE Approach to Early Detection of Anomalies in Electricity Consumption. In 2025 IEEE 6th KhPI Week on Advanced Technology (KhPIWeek) (pp. 1-5). IEEE.
30. Xu, T., Zhang, J., & Zhu, W. (2026). Fairness-Accuracy Trade-offs and Calibration Mechanisms in Machine Learning-Based Subprime Credit Scoring. Available at SSRN 6893759.