PEaF-Production Environment Analyzer Framework: Assisting Continuous Deployment of 5G Workloads Using AI/ML

Abstract

A Production Environment Analyzer Framework (PEaF) is proposed to address the limitations of the Continuous Deployment (CD) process for 5G workflow lifecycle management. By integrating an AI/ML-based PEaF into the CD pipeline, we aim to ensure reliable deployments. PEaF uses AI/ML techniques to analyze the production environment and predict the health status of the hardware components. It collects raw data, applies K-Means clustering to group similar data points, and assigns scores to each cluster. These scores serve as features for training Support Vector Machine (SVM) and Random Forest (RF) classifiers to classify hardware health status. Experimental results show that PEaF achieves high classification accuracies of 97.26% and 96.44% for SVM and RF, respectively, with clustering. By analyzing the production environment and excluding deteriorating hardware from the CD, service failures are reduced by at least 27.04%. Moreover, PEaF decreases the polling frequency of hardware status by 48.7%, enhancing operational efficiency. Overall, PEaF contributes to advancing Continuous Integration/Continuous Deployment (CI/CD) practices in the 5G ecosystem, ensuring the reliability and stability of the production environment before deploying/upgrading services.

Keywords

• 5G
• AI/ML
• CI/CD
• classification
• feature selection
• K-means clustering