KPI Trend Deviation Normalization for Mobile Network Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network performance prediction in mobile networks is challenging due to exponential traffic growth, dynamic changes, long capacity addition cycles, and the need for accurate future network performance estimation and what-if evaluations, compounded by the difficulty in processing large volumes of Call Detail Record (CDR) data to derive actionable intelligence.
Innovation Solution
A method involving a Smart Service Analyzer that uses Machine Learning to estimate performance indices like CEI and SQI, with a Trend Deviation-Based Quantitative KPI Normalizer to normalize User Level KPIs based on deviations from Network Level benchmarks, enabling automated, end-to-end analysis and notification of significant changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If User Level KPIs are aggregated to generate Network Level KPIs, then network performance estimation accuracy is improved, but data processing complexity increases
Solution Approach 1:
The patent segments the network performance analysis into User Level KPIs and Network Level KPIs, processing them at different aggregation levels. This allows accurate performance estimation through hierarchical aggregation while managing complexity by handling data in structured segments rather than as a monolithic processing task.
Solution Approach 2:
The patent introduces an intermediary processing layer that aggregates User Level KPIs to generate Network Level KPIs. This intermediary layer simplifies the overall processing complexity by creating a structured intermediate representation that bridges raw user-level data and high-level network performance metrics.
2Loss of information
If large volumes of CDR data are processed to derive actionable intelligence, then performance measurement capability is improved, but processing time and resources increase
Solution Approach 1:
The patent extracts only the essential and relevant features from large volumes of CDR data to create KPIs. By taking out only the critical information needed for performance measurement rather than processing all raw data, the system derives actionable intelligence while significantly reducing processing time and computational resources.
Solution Approach 2:
The patent applies partial action by processing a representative subset of CDR data through aggregation and normalization to generate Network Level KPIs. This partial processing approach provides sufficient actionable intelligence for capacity planning without the excessive time and resources required for complete data processing.
3Productivity
If capacity is added to mobile networks to meet growing traffic demand, then network performance is improved, but capital expenditure increases
Solution Approach 1:
The patent performs preliminary action by using Network Level KPIs and trend analysis to predict future network performance and identify capacity requirements in advance. This allows operators to plan capacity additions proactively based on data-driven insights, optimizing capital expenditure by adding capacity only when and where it is truly needed rather than reacting to performance degradation.
Solution Approach 2:
The patent implements feedback mechanisms where normalized Network Level KPIs continuously monitor network performance and feed insights back to capacity planning decisions. This feedback loop enables operators to adjust capital expenditure based on actual performance trends and predictive analytics, ensuring capacity additions are justified by measured needs rather than arbitrary timelines.
Data Source
AI summary
A Trend Deviation-Based Quantitative Key Performance Indictors (KPI) Normalizer and method for a Smart Service Analyzer. Trend Deviation-Based Quantitative Key Performance Indictors (KPI) Normalizer receives User Level Quantitative Key Performance Indictors (KPIs) at a Cluster Based Aggregator. Cluster-Based Aggregation of the User Level Quantitative KPIs is performed to generate Network Level Quantitative KPIs. A forecasted trend is determined based on the Network Level Quantitative KPIs. The User Level Quantitative KPIs are processed by comparing against a forecasted Network Level Quantitative KPI to generate Normalized Quantitative KPIs at the User Level based on the forecasted trend, wherein the Normalized Quantitative KPIs at the User Level include a deviation from the forecasted trend.