Root-Cause Diagnosis for Communication Network Events
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current monitoring systems for communication networks, particularly mobile telephony networks, face challenges in efficiently processing large amounts of real-time data to detect and diagnose root causes of anomalies, often requiring skilled operators and lacking effective automated tools for simplifying investigations and prioritizing events.
Innovation Solution
A method and system that implement a two-phase processing approach for root-cause diagnosis, using machine learning clustering and business logic rules to analyze network data, identifying single event and group root-causes by determining distributions on network dimensions and applying adaptive anomaly threshold values to minimize errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If monitoring systems gather real-time data from multiple sources, then the ability to detect network anomalies is improved, but the complexity of processing and comprehending the data increases
Solution Approach 1:
The patent segments the data processing into two distinct phases: first determining single event diagnoses for individual events, then determining group root-cause diagnoses for correlated events. This segmentation allows the system to handle complex multi-source data by processing it in manageable steps, reducing overall processing complexity while maintaining comprehensive anomaly detection capability.
Solution Approach 2:
The patent introduces an intermediary processing layer that uses machine learning algorithms and business logic rules to transform raw real-time data into structured diagnoses. This intermediary layer acts as a mediator between the complex data collection phase and the final root-cause analysis, simplifying the processing complexity while preserving detection reliability.
2Productivity
If automated tools are used to analyze network events, then the speed of root-cause identification is improved, but the accuracy and reliability of diagnosis may deteriorate
Solution Approach 1:
The patent employs machine learning algorithms that dynamically adjust their analysis parameters based on the characteristics of detected events. By changing analysis parameters adaptively, the system maintains high diagnosis accuracy while operating at automated speed, resolving the contradiction between productivity and reliability.
Solution Approach 2:
The system incorporates feedback mechanisms where the results of single event diagnoses are used to inform group root-cause analysis. This feedback loop allows the automated tool to refine its diagnoses continuously, maintaining accuracy while achieving rapid root-cause identification through automation.
3Reliability
If the system analyzes groups of correlated events, then the completeness of root-cause diagnosis is improved, but the processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary action by first determining single event diagnoses for all detected events before proceeding to group root-cause analysis. This preliminary processing organizes the data structure and identifies potential correlations in advance, reducing the computational time required for the subsequent group analysis phase while maintaining complete diagnosis.
Solution Approach 2:
By segmenting the analysis into sequential phases (single event diagnosis first, then group root-cause diagnosis), the system processes correlated events in manageable batches rather than analyzing all events simultaneously. This segmentation reduces overall processing time while preserving the completeness of root-cause diagnosis for event groups.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention concerns a method and a system for determining root-cause diagnosis of events occurring during the operation of a communication network, implemented by a processor of a programmable device, comprising monitoring (52) time signals representative of the operation of the network to detect the occurrence of an event relative to the network traffic. The method further comprises a first processing phase, wherein for each detected event, during the duration of said event: - obtaining (58) distributions of data on several dimensions of the network linked to said event, - automatically determining (60, 62) an event root-cause diagnosis of the detected event, called single event diagnosis, comprising at least one element of said distributions, an element being a value taken by a network dimension having a contribution in said distributions of data, the single event diagnosis determination using a first set of rules of business logic configuration organized hierarchically, which are applied according to said hierarchy to select at least one element of said distributions, the selection of more than one element comprising machine learning clustering.