Radio Access Network Load Change Detection for Accurate Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication networks face challenges in accurately forecasting key performance indicators (KPIs) due to unidentified load changes, which can lead to inaccurate rescaling of historical data, resulting in unsatisfactory forecasts when seasonal and network changes are not differentiated.
Innovation Solution
A system and method for analyzing loading data from cell sites to identify changepoints, distinguishing between seasonal and network-related changes, and rescaling historical data only for network-related changes to improve forecasting accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If historical data is rescaled without differentiating between seasonal and network changes, then the rescaling process is simple, but forecasting accuracy deteriorates
Solution Approach 1:
The patent segments load changes into two distinct categories: seasonal changes and network-related changes. By detecting changepoints and analyzing their characteristics (such as time of occurrence, magnitude, and pattern), the system separates these different types of changes and applies rescaling only to network-related changes, thereby improving forecasting accuracy without unnecessarily complicating the overall process
Solution Approach 2:
The patent performs preliminary analysis of load changes before rescaling historical data. By detecting changepoints and determining their causes in advance, the system prepares the data by identifying which portions require rescaling, ensuring that the rescaling process is applied correctly and improves forecasting accuracy
2Measurement precision
If all load changes are treated as network-related for rescaling, then rescaling is applied consistently, but seasonal patterns cause forecasting errors
Solution Approach 1:
The patent segments load changes into seasonal and non-seasonal categories by analyzing changepoint characteristics. Seasonal changes (such as those occurring at regular intervals like daily or weekly patterns) are identified and excluded from rescaling, while only non-seasonal network-related changes are rescaled, thereby maintaining forecast reliability
Solution Approach 2:
Instead of applying rescaling to all load changes as conventionally done, the patent inverts the approach by applying rescaling only to the subset of changes that are not seasonal. This inverted logic prevents seasonal patterns from causing forecasting errors while still capturing network-related changes
Data Source
AI summary
One or more methods and/or systems for detecting loading changes are provided. First data is gathered from a wireless cell site. The first data may be indicative of loading of the wireless cell site. Changepoints in the first data may be detected. The first data may be analyzed to obtain indications of causes of the changepoints. The causes may be seasonal events and/or non-seasonal events. A determination of the causes may be made using the indications.