Energy saving through flexible kubernetes pod capacity selection during horizontal pod autoscaling (HPA)

A device, comprising:

• at least one processor configured to implement:
• a KPI predictor configured to receive performance metrics regarding resource allocation and capacity of pods, analyze a traffic pattern, measure a current system capacity, determine a traffic curve representing the traffic pattern associated with a current traffic demand on the current system capacity, select an Artificial Intelligence/Machine Learning (AI/ML) model based on the traffic curve, and predict a future traffic demand using the selected AI/ML model;
• a scaling decision configured to select a pod capacity scaling in terms of a number of the pods and a pod capacity version based on the predicted future traffic demand verses the current system capacity to provide optimal performance for the future traffic demand according to pod capacity categories;
• a container manager configured to scale the pods to fulfill the future traffic demand according to the pod capacity categories;
• and
• a Non-Real-Time Radio Access Network Intelligent Controller (Non-RT RIC), wherein the Non-RT RIC uses rApps to apply scaling commands received from the scaling decision to the pods deployed in an Open-Cloud (0-Cloud) System;
• wherein the Non-RT RIC is configured to collect Open-Cloud Fault, Configuration, Accounting, Performance, Security (FCAPS) data over an 02 interface, and to collect E2 node data over an 01 interface, wherein the performance metrics includes the performance metrics for the Non-RT RIC based on collection of the FCAPS data over the 02 interface, and collection of E2 node data over the 01 interface, and wherein the Non-RT RIC trains and deploys one or more AI/ML models to generate scaling guidance for the O-Cloud System or E2 Nodes based on priority, load, energy consumption, and service quality specifications.

A non-transitory computer-readable media having computer-readable instructions stored thereon for performing operations comprising:

• implementing an Artificial Intelligent/Machine Learning (AI/ML)-based horizontal pod autoscaler (HPA);
• receiving, at the AI/ML-based HPA, performance metrics regarding resource allocation and capacity of pods;
• analyzing a traffic pattern and measuring a current system capacity;
• determining a traffic curve representing the traffic pattern associated with a current traffic demand on the current system capacity;
• selecting an Artificial Intelligence/Machine Learning (AI/ML) model based on the traffic curve;
• predicting, by the AI/ML-based HPA, a future traffic demand using the selected AI/ML model;
• selecting pod capacity scaling in terms of a number of the pods and a pod capacity version based on the predicted future traffic demand verses the current system capacity to provide optimal performance for the future traffic demand according to pod capacity categories;
• and
• scaling the pods to fulfill the future traffic demand according to the pod capacity categories, wherein the implementing the AI/ML-based HPA includes implementing a Non-Real-Time Radio Access Network Intelligent Controller (Non-RT RIC), wherein the Non-RT RIC uses rApps to apply scaling commands to the pods deployed in an Open-Cloud (O-Cloud) System;
• wherein the receiving the performance metrics includes obtaining the performance metrics by the Non-RT RIC by collecting O-Cloud Fault, Configuration, Accounting, Performance, Security (FCAPS) data over an O2 interface, collecting E2 node data over an O1 interface, and wherein the Non-RT RIC trains and deploys AI/ML models to generate scaling guidance for the O-Cloud System or E2 Nodes based on priority, load, energy consumption, and service quality specifications;
• and
• wherein the training and deploying the AI/ML models by the Non-RT RIC includes training and deploying at least one of a Linear Regression model, a Feedforward Neural Networks (FNN), a Convolutional Neural Networks (CNN) model, or a Long Short Term Memory model.