Communication Traffic Analysis Device Using Stream Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques face challenges in accurately analyzing traffic in carrier networks due to difficulties in identifying changes in feature amounts and performing high-accuracy classification, especially when large-capacity traffic is composed of multiple flows and classification is based on tendencies rather than definite patterns.
Innovation Solution
An analysis device that divides communication traffic into streams and calculates an abnormality degree through autoregressive analysis and comparison between similar streams, using a division unit to split traffic into flows and a calculation unit to determine abnormality scores based on normalized vectors and threshold comparisons.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stream mining technique is used to analyze communication traffic, then traffic analysis can be performed, but it is difficult to accurately ascertain changes in feature amounts because large-capacity traffic is obtained by superimposing multiple flows
Solution Approach 1:
The patent divides the communication traffic into multiple streams based on flow characteristics. By segmenting the superimposed traffic into distinct streams, the system can accurately detect changes in feature amounts for each stream individually, overcoming the difficulty of analyzing mixed traffic composition.
2Measurement precision
If classification method based on tendency is used, then traffic classification can be performed, but classification accuracy decreases because it relies on tendency rather than definite classification
Solution Approach 1:
The patent segments traffic into distinct streams with clear identification criteria, enabling definite classification rather than tendency-based classification. Each stream can be reliably identified and classified based on its specific characteristics, improving both accuracy and reliability.
Solution Approach 2:
The patent incorporates abnormality detection feedback into the classification process. By continuously monitoring stream characteristics and comparing them against established patterns, the system can confirm or adjust classifications, ensuring high accuracy and reliability of traffic classification results.
3Reliability
If conventional stream mining is used, then traffic analysis can be performed, but it cannot accurately detect abnormalities in complex network traffic
Solution Approach 1:
The patent divides complex network traffic into manageable streams, making abnormality detection more reliable. By analyzing each stream separately rather than treating traffic as a monolithic complex mass, the system can accurately identify abnormalities even in complex network environments.
Solution Approach 2:
The patent introduces an intermediary analysis layer that processes and simplifies complex traffic before abnormality detection. This intermediary processing step breaks down complex traffic patterns into analyzable stream components, enabling accurate abnormality detection without being overwhelmed by traffic complexity.
Data Source
AI summary
A division unit divides communication traffic in a communication network into a plurality of streams. A calculation unit calculates a first abnormality degree based on a result of autoregressive analysis of a first stream among the plurality of streams, and a second abnormality degree based on a comparison result between the first stream and a second stream similar to the first stream. A determination unit determines whether the cause of the abnormality is the first stream, a stream different from the first stream, or a communication network based on the first abnormality degree and the second abnormality degree.