RAN Network Element Alarms for Low-Latency SON Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cellular network performance optimization systems face latency issues due to repeated processing of performance metrics at each hop enroute the Self-Organizing Network (SON) server, and transient issues may not be addressed in time by current methods.
Innovation Solution
Implementing a system where network elements transmit performance alarms directly to a SON server, reducing latency by bypassing intermediate processing, and utilizing AI/ML algorithms to identify and initiate mitigation actions based on these alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If performance metrics are processed at each hop enroute the SON server, then comprehensive performance analysis is achieved, but latency increases
Solution Approach 1:
The patent extracts the performance alarm generation function from the intermediate network elements and consolidates it at the SON server. Network elements generate performance alarms based on monitored metrics and transmit them directly to the SON server, eliminating the need for repeated processing at each hop. This extraction reduces latency while maintaining analysis accuracy through direct alarm transmission and AI/ML-based processing at the server.
Solution Approach 2:
The patent introduces performance alarms as an intermediary mechanism between network elements and the SON server. Instead of transmitting and reprocessing raw performance metrics at each hop, network elements convert metrics into condensed performance alarms that carry essential information. These alarms serve as mediators that reduce data volume and processing requirements while preserving critical performance insights for timely analysis.
2Reliability
If current performance monitoring methods are used, then existing issues are detected, but transient issues are not addressed in time
Solution Approach 1:
The patent implements preliminary action by having network elements continuously monitor performance metrics and pre-generate performance alarms when predefined thresholds are breached. This allows the system to detect transient issues at their inception and transmit alarms immediately to the SON server, enabling proactive response before issues escalate. The AI/ML algorithms further enhance this by analyzing alarm patterns to predict potential problems before they manifest.
Solution Approach 2:
The patent establishes a feedback loop where the SON server receives performance alarms, analyzes them using AI/ML algorithms, and generates mitigation actions that are transmitted back to network elements. This closed-loop feedback system ensures rapid response to transient issues by continuously monitoring performance, detecting anomalies through alarms, and automatically implementing corrective actions, thereby improving both detection reliability and response time.
Data Source
AI summary
Techniques for optimizing cellular network performance of network elements of a Radio Access Network (RAN) are disclosed. In operation, Performance Metrics (PM) corresponding to a network element of the RAN are monitored, where the PM includes a plurality of network performance indicators indicative of cellular network performance of the network element. It is then determined that a first network performance indicator from amongst the plurality of network performance indicators is meeting a predetermined condition. A performance alarm is then transmitted to a Self-Organizing Network (SON) server, where the performance alarm is indicative of at least one issue corresponding to the cellular network performance of the network element.