Network Quality Analysis With Adaptive Packet Loss Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network analysis devices face challenges in distinguishing between different abnormal states in communication networks, leading to incorrect alarm notifications and adjustments in abnormality determination thresholds, which can result in missed or unnecessary alarms.
Innovation Solution
The implementation of a network analysis method using just-in-time (JIT) modeling to dynamically adjust abnormality determination thresholds based on historical network quality trends and variations, allowing for more accurate abnormality detection even with limited packet data, by calculating an estimated packet loss rate and adjusting the normal range based on operator feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the abnormality determination threshold is set based on average packet loss rates from aggregated network analysis data across multiple connections, then the device can perform collective analysis to reduce monitoring costs, but the device generates exaggerated alarms when the number of packets decreases during out-of-service periods
Solution Approach 1:
The patent applies dynamics by making the abnormality determination threshold adaptive rather than static. The threshold dynamically adjusts based on the current number of packets and historical data, allowing the system to accommodate varying network traffic conditions. When packet numbers decrease during out-of-service periods, the threshold automatically adapts to prevent exaggerated alarms while maintaining sensitivity to actual abnormalities.
Solution Approach 2:
The patent changes the parameter of the abnormality determination threshold based on the number of packets and historical network analysis data. Instead of using a fixed threshold, the system modifies the threshold parameter dynamically according to current traffic conditions, thereby resolving the contradiction between collective analysis efficiency and accurate abnormality detection.
2Reliability
If the abnormality determination threshold is increased to prevent exaggerated alarms during low-traffic periods, then false alarms are reduced, but the device cannot generate necessary alarms when actual network abnormalities occur
Solution Approach 1:
The system dynamically adjusts the abnormality determination threshold based on the current number of packets and historical data patterns. During low-traffic periods, the threshold adapts to prevent false alarms, while during normal traffic conditions, it maintains high sensitivity to detect actual abnormalities. This dynamic adjustment resolves the contradiction between alarm accuracy and detection sensitivity.
Solution Approach 2:
The patent implements feedback mechanisms using historical network analysis data to continuously refine the abnormality determination threshold. The system learns from past patterns and adjusts the threshold accordingly, ensuring that alarm accuracy is maintained during low-traffic periods while preserving the ability to detect genuine abnormalities when they occur.
3Reliability
If JIT modeling is used to estimate current network quality based on previous network quality trends, then appropriate abnormality determination can be performed even with small packet numbers, but the system complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-processing and storing historical network analysis data in advance. This historical data is prepared and organized beforehand, enabling the JIT modeling to efficiently estimate current network quality without requiring complex real-time computations. The pre-prepared historical context allows accurate abnormality determination even with limited current packet data.
Solution Approach 2:
The system uses copying by creating a simplified statistical model (local model) that replicates the essential characteristics of network quality behavior based on historical data. This local model copies the key patterns from extensive historical data, allowing the system to perform accurate anomaly detection with reduced computational complexity during real-time operation.
Data Source
Figure 1A
Figure 2~3
Figure 4
AI summary
A computer readable network analysis program of performing local modeling analysis of determining an estimated value of a current network quality corresponding to explanatory variable vector in current aggregated data based on a local model including local training data; determining an abnormality in the network based on whether or not a measured value of the current network quality is lower than a threshold; determining whether or not a distribution of the connections having the measured value of the network quality exceeding the threshold is present in a large size; extracting an individual-analysis-target connection group including more than predetermined proportions of connections in the distribution of the connections having the large size; and performing the local modeling analysis to the individual-analysis-target connection group and the remaining connection groups to determine the abnormality in the network.