.png){kind=icon}
.png){kind=icon}
Return
Deep Q-learning with feature extraction and prioritized experience replay for edge node overload in edge computing
.png){kind=icon}
PRE
AI
DOI：10.1016/j.engappai.2025.112124.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Novel DQN algorithm to learn edge resource patterns and to predict the overload. • A system model to converted high-dimensional features into lower-dimensional features. • DQN-PER+FE, a DRL based edge node resource overload estimation scheme for edge nodes. • The combination of feature extraction-based LDA and DQN feature in estimating Q-value. • The DQN with prioritized experience replay for classifying resources overloading.
Journal
IF:
8
Papers:
5.3K
Citations:
3.5W