Radio Site Condition Prediction Using Selective ML Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting operational conditions of radio network nodes in cellular networks are inadequate, failing to accurately forecast issues like power outages and sleeping cells, which can lead to downtime and service disruptions.
Innovation Solution
A method utilizing machine learning models, activated in an inference engine, to predict future operational conditions based on input properties such as RATs, power sources, and geographical data, allowing for the collective application of models to enhance prediction accuracy, with the option to filter and weight models and receive feedback for improvement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional monitoring methods are used in NOC, then system complexity is low, but prediction accuracy of operational conditions is insufficient
Solution Approach 1:
The system segments the prediction task by selecting and activating multiple specialized machine learning models based on input properties. Each model is trained for specific operational conditions (e.g., power outage prediction, sleeping cell detection), allowing the system to achieve high prediction accuracy for different failure modes without requiring a single overly complex model
Solution Approach 2:
The inference engine provides universal functionality by dynamically selecting and coordinating multiple machine learning models based on input properties. This multi-functional approach enables the system to handle various prediction tasks (power outages, sleeping cells, latency degradation) through a single unified platform that adapts to different operational scenarios
2Measurement precision
If multiple machine learning models are collectively applied, then prediction accuracy improves, but computational resources and processing time increase
Solution Approach 1:
The system performs preliminary action by pre-selecting relevant machine learning models based on input properties before actual prediction is needed. The inference engine evaluates input properties and activates only the necessary models, avoiding the computational overhead of running all possible models and reducing processing time while maintaining high accuracy
Solution Approach 2:
The system applies partial action by activating only the subset of machine learning models that are relevant to the specific input properties and prediction task at hand. Rather than running all available models, the inference engine selectively engages only those needed, reducing computational resource consumption and processing time while still achieving accurate predictions
Data Source
AI summary
It is provided a method for enabling prediction of a future operational condition for at least one site, each site comprising at least one radio network node of a radio access technology, RAT, of a cellular network. The method comprises the steps of: obtaining input properties of the at least one site; selecting a plurality of machine learning models based on the input properties; and activating the selected plurality of machine learning models in an inference engine, such that all of the selected plurality of machine learning models are collectively applicable to enable prediction of a future operational condition of the at least one site.