Network Analysis Apparatus Bursty Traffic Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network monitoring systems face challenges in efficiently detecting bursty traffic and maintaining network quality due to high processing loads caused by the need to store and analyze statistical information at both infinitesimal and relatively long time intervals, leading to increased computational costs and memory requirements.
Innovation Solution
A method that updates and processes statistical information in continuous storage areas, allowing for efficient analysis at multiple cycles, including a first cycle for storing communication state and a second cycle for partial information, reducing memory usage and processing time by switching between update and read candidates, and ensuring data consistency through exclusive control of storage areas.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If statistical information is stored at an infinitesimal time interval to detect bursty traffic, then detection precision is improved, but processing load and memory requirements increase excessively
Solution Approach 1:
The patent segments statistical information storage into two distinct time intervals: infinitesimal time interval storage for bursty traffic detection, and relatively long time interval storage for overall network quality monitoring. This segmentation allows the system to maintain high detection precision for bursty traffic while avoiding the excessive processing load that would result from storing all statistical information at the infinitesimal interval, as each segment serves its specific purpose with appropriate storage frequency.
2Reliability
If statistical information is stored at an infinitesimal time interval, then bursty traffic detection capability is improved, but memory usage becomes unrealistic
Solution Approach 1:
The patent extracts only the necessary statistical information for bursty traffic detection at the infinitesimal time interval, rather than storing all possible statistical data. By taking out only the critical metrics needed for burst detection, the system maintains reliable bursty traffic detection capability while keeping memory usage at realistic levels, avoiding the storage of redundant or less critical information.
3Reliability
If both short-term and long-term statistical information storage functions are provided, then network quality monitoring is improved, but processing performance degrades due to multiple memory accesses
Solution Approach 1:
The patent implements periodic action by storing statistical information at two different periodic intervals: infinitesimal intervals for bursty traffic detection and relatively long intervals for overall network quality monitoring. This periodic approach allows the system to maintain comprehensive network quality monitoring capabilities while optimizing processing performance by avoiding continuous simultaneous updates to multiple storage structures, instead using periodic consolidation at the longer interval.
Data Source
AI summary
In an analysis apparatus, when detecting each packet communicated via a network, an update unit updates statistical information that is stored in a continuous storage area of a storage unit and that indicates the communication state of the network for a first cycle. Then, a first processing unit reads the statistical information from the storage unit at every first cycle, and processes the statistical information and initializes the statistical information in the storage unit. A second processing unit reads partial statistical information that is part of the statistical information from the storage unit at every second cycle shorter than the first cycle, and processes the partial statistical information and initializes the partial statistical information in the storage unit.