一种基于时间序列分解网络流量预测系统
By combining multi-scale feature extraction, time-series data augmentation, and neural network denoising algorithms with the MCLSTM model, the problem of excessive noise and abrupt changes in network traffic prediction is solved, achieving higher prediction accuracy and lower error, and improving the ability to detect complex network traffic.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIHANG UNIV
- Filing Date
- 2022-04-18
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies are affected by noise and a large number of abrupt changes when processing network traffic prediction, resulting in poor prediction performance, especially for the detection of unknown threats. Furthermore, traditional methods suffer from imbalanced dataset distribution and high false alarm rates.
A network traffic prediction system based on time series decomposition is adopted. Through multi-scale feature extraction, time series data augmentation and neural network denoising algorithm, combined with MCLSTM model, trend and residual subsequence are jointly modeled, and the denoising threshold is dynamically adjusted to improve prediction accuracy.
It improves the prediction accuracy of network traffic feature sequences, reduces prediction errors, enhances the model's generalization ability, reduces the risk of overfitting, and significantly improves the prediction performance for complex network traffic.
Smart Images
Figure QLYQS_1 
Figure QLYQS_7 
Figure QLYQS_8