5G Slice Usage Prediction with Clustered ML Resource Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 5G networking lacks effective solutions for predicting network usage and traffic patterns across multiple slices or cells, and there is a lack of mechanisms to handle multiple key performance indicators through a single or cluster of models, leading to inefficient resource management.
Innovation Solution
A machine learning system is trained using a clustering algorithm to standardize and predict future usage of 5G network slices or cells, enabling dynamic resource allocation and efficient management by creating a model-efficient framework that handles multiple correlated performance metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional resource management methods are used in 5G networks, then network operations can be maintained, but resource allocation efficiency deteriorates leading to overutilization and underutilization
Solution Approach 1:
The system implements feedback mechanisms by continuously monitoring network usage patterns and performance metrics, then using this information to dynamically adjust resource allocation decisions. The machine learning models process historical and real-time data to provide feedback-driven optimization of network slice resource distribution, preventing both overutilization and underutilization of network resources.
Solution Approach 2:
The machine learning system enables self-service resource management by autonomously predicting network slice usage patterns and automatically adjusting resource allocation without manual intervention. The system serves itself by using its own predictions to trigger resource allocation changes, creating a self-optimizing loop that continuously improves resource distribution efficiency.
2Measurement precision
If multiple separate models are used to predict different performance metrics, then each metric can be predicted, but system complexity increases making management difficult
Solution Approach 1:
The patent merges multiple prediction models into a unified machine learning system that simultaneously predicts multiple network performance metrics. Instead of maintaining separate models for different metrics, the system integrates them into a single framework that processes data once and generates multiple predictions, reducing system complexity while maintaining prediction accuracy across all metrics.
Solution Approach 2:
The machine learning system is designed with multi-functionality to handle various network slice types, performance metrics, and prediction scenarios through a single universal platform. The system can adapt to different network conditions and slice requirements without requiring separate specialized models, thereby reducing complexity while preserving measurement precision across diverse use cases.
3Adaptability or versatility
If resources are statically allocated to network slices, then resource management is simple, but network adaptability deteriorates preventing efficient handling of dynamic traffic patterns
Solution Approach 1:
The system implements dynamic resource allocation by continuously adjusting network slice resources based on real-time usage patterns and predictions. Instead of static allocation, the system dynamically modifies resource distribution to match actual network demands, enabling high adaptability to changing traffic patterns while using automated machine learning processes to manage the complexity of dynamic adjustments.
Data Source
AI summary
A machine learning system is trained to predict resource usage by cells or network slices of a mobile network. For example, a computing system obtains respective datasets for the cells or network slices. Each dataset comprises time steps and respective values for a performance metric of the corresponding one of the cells or network slices. The computing system groups, based on a clustering algorithm applied to (1) the datasets, or (2) the cells or network slices, the datasets into clusters of datasets. The computing system applies, to a subset of most-recent time steps and corresponding values of each dataset of a first cluster of the clusters, a transformation to obtain a set of time steps and corresponding standardized values with which a machine learning system is trained to generate predicted values at future time steps of the datasets of the first cluster.