RNA Optimization Through AI/ML Anomaly Cell Exclusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 5G RAN-based Notification Areas (RNA) may include anomaly cells with poor performance, leading to communication issues and UE unreachability due to RRC connection failures and paging problems when UEs transition to RRC_INACTIVE state.
Innovation Solution
Utilizing a radio intelligent controller (RIC) with AI/ML techniques to detect and analyze anomaly cells based on KPIs, PMs, CMs, FM data, and trace data, recommending actions like excluding or disabling RRC connections to improve RNA optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If anomaly cells are included in RNA configuration, then RNA coverage is improved, but UE reachability and communication reliability deteriorate due to paging issues and RRC connection failures
Solution Approach 1:
The RIC extracts and identifies anomaly cells from the RNA configuration using AI/ML analysis of KPIs, PMs, CMs, FM data, and trace data. Once identified, these anomaly cells are excluded from the RNA configuration to prevent UE reachability issues while maintaining coverage from healthy cells.
Solution Approach 2:
The system performs preliminary identification and exclusion of anomaly cells before UEs are assigned to RNA. The RIC proactively analyzes cell performance data and prevents UE assignment to anomaly cells, avoiding communication failures rather than reacting to them afterward.
2Reliability
If AI/ML-based anomaly detection is implemented, then UE communication reliability is improved, but system complexity increases due to RIC and data analysis requirements
Solution Approach 1:
The RIC acts as an intermediary component that centralizes the complex AI/ML-based anomaly detection functionality. Instead of distributing complexity across multiple network elements, the RIC consolidates data collection, analysis, and decision-making in a single dedicated entity, simplifying the overall system architecture.
Solution Approach 2:
The RIC autonomously performs anomaly cell identification and recommends exclusions without requiring manual intervention. The system self-services by automatically collecting data from multiple sources, analyzing it using AI/ML algorithms, and generating configuration recommendations, reducing operational complexity.
3Measurement precision
If real-time anomaly detection is performed using multiple data sources, then cell performance accuracy is improved, but data processing time and computational resources increase
Solution Approach 1:
The RIC is designed as a multi-functional platform that simultaneously collects and analyzes multiple data types (KPIs, PMs, CMs, FM data, trace data) from various network sources. This universal approach consolidates multiple data processing functions into a single system, improving efficiency despite the diverse data requirements.
Solution Approach 2:
The system prioritizes analysis of the most critical data sources and KPIs that provide the highest value for anomaly detection. Rather than processing all available data with equal depth, the RIC focuses on key indicators that most strongly correlate with cell performance anomalies, reducing processing time while maintaining detection accuracy.
Data Source
AI summary
A method of optimizing Radio Access Network-based Notification Area (RNA) in the case a user equipment (UE) transitions to radio resource control inactive (RRC_INACTIVE) state includes: detecting, by a radio intelligent controller (RIC) using artificial intelligence (AI) and/or machine learning (ML) technique (e.g., State Vector Machine or Isolation Forest), an anomaly cell based on at least one of the following factors: key performance indicators (KPIs), performance measurements (PMs), configuration parameters (CMs), fault management (FM) data, and trace data; determining, by the RIC, based on the at least one of the factors, a cause for the detected anomaly cell; and recommending, by the RIC to a gNodeB associated with the RNA having the anomaly cell, at least one of the following actions: a) exclude the anomaly cell from the RNA, and b) disable new radio resource control (RRC) connections from the UE.