Network Node Prediction Model for RAN Performance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mobile Broadband operators face challenges in accurately measuring user experience metrics like Page Load Time (PLT) due to reliance on costly client-based measurements and the complexity of deep dive data, which provides limited visibility into real user experience, especially for encrypted services.
Innovation Solution
A method using a network node to predict performance indicators by obtaining measurement data from a radio access network node and an end node, training a prediction model to forecast PLT without client-based measurements, and selecting predictors that are independent of the service, allowing for attribution of changes to RAN node performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If client-based measurements are used to measure performance indicators like Page Load Time, then measurement accuracy is improved, but cost and complexity increase significantly
Solution Approach 1:
The patent introduces a prediction model as an intermediary between RAN node measurements and performance indicators. The model translates simple RAN measurements (bandwidth, latency, packet loss) into accurate performance predictions without requiring complex client-based measurements. This mediator approach resolves the contradiction by providing accurate performance indication through a simpler intermediate measurement system.
Solution Approach 2:
The patent creates a virtual copy of performance indicator measurements through prediction models. Instead of directly measuring performance indicators like Page Load Time which requires complex client-based systems, the invention creates accurate predictions (copies) of these indicators using simpler RAN node measurements fed into trained prediction models, thereby reducing complexity while maintaining measurement accuracy.
2Loss of information
If deep dive data is collected to analyze user experience, then visibility into user experience is improved, but data volume and processing complexity become unmanageable
Solution Approach 1:
The patent extracts only the essential features needed for performance prediction from the complex deep dive data. Instead of collecting and processing all available deep dive data, the invention extracts key RAN node measurements (bandwidth, latency, packet loss) that are sufficient for accurate performance indication, thereby reducing data volume while maintaining user experience visibility.
Solution Approach 2:
The patent segments the measurement process into two parts: simple RAN node measurements that are easily obtainable, and complex performance indicator calculations that are performed separately through prediction models. This segmentation allows the system to maintain comprehensive user experience visibility without needing to collect and process all the raw deep dive data simultaneously.
3Measurement precision
If service-specific measurements are used to predict performance, then prediction accuracy for that service is improved, but the solution becomes less versatile across different services
Solution Approach 1:
The patent creates a universal prediction framework where the same RAN node measurements (bandwidth, latency, packet loss) can be used to predict performance across multiple different services. The prediction models are trained on service-specific data but can be applied universally to various services including web browsing, video streaming, and voice calls, resolving the contradiction between service-specific accuracy and cross-service versatility.
Data Source
AI summary
The present invention relates to a method for predicting a performance indicator for a service in a network. The method is performed by a network node of the network, and the method comprises obtaining measurement data of a metric affecting a service communicating via a radio access network, RAN, node, wherein the metric is independent of the service communicating via the RAN node, inputting the obtained measurement data into a prediction model for performance of the service communicating via the RAN node, wherein the prediction model has been trained with measurement data from the RAN node and measurement data from an end node, and predicting the performance indicator for performance of the service in the network. A network node, a computer program and a computer program product are also presented.