Filtering Noisy Application Signatures for Accurate Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large and inefficient lookup tables in application-based routing systems are populated with rarely-used or peer-to-peer connection details, leading to increased memory requirements and slowed queries, as well as incorrect routing of network traffic due to misidentification of peer IP addresses.
Innovation Solution
A method and system that filter out noisy application signatures by determining their validity based on frequency across distinct network environments, using a two-stage filtering process to increment or decrement signature scores, and promote only valid signatures to lookup tables, thereby reducing memory and processing burdens.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the lookup table is populated with all application signatures from peer-to-peer connections, then the table becomes comprehensive for routing purposes, but the table size increases and query performance deteriorates
Solution Approach 1:
The patent extracts and removes noisy peer-to-peer connection signatures from the lookup table while retaining valid application signatures. This is achieved through a filtering mechanism that identifies and eliminates signatures associated with peer-to-peer connections, thereby maintaining routing accuracy for legitimate applications while reducing table size and improving query performance.
2Adaptability or versatility
If the lookup table includes all connection details, then routing coverage is complete, but memory requirements increase
Solution Approach 1:
The patent removes unnecessary peer-to-peer connection details from the lookup table while preserving essential application routing information. This extraction process reduces the quantity of data stored in memory while maintaining the table's ability to route legitimate application traffic effectively.
Solution Approach 2:
The patent applies different quality standards to different types of signatures in the lookup table. Valid application signatures are retained with full detail, while peer-to-peer signatures are filtered out or marked differently. This local differentiation optimizes memory usage by storing only the necessary information for each signature type.
3Adaptability or versatility
If peer IP addresses are included in the lookup table, then peer-to-peer routing is enabled, but misidentification of peer IP addresses occurs
Solution Approach 1:
The patent extracts and separates peer IP addresses from the main application signature database. By removing peer-to-peer connection signatures from the lookup table, the system prevents misidentification of peer IP addresses while maintaining the ability to route legitimate application traffic through accurate signature matching.
Data Source
AI summary
The system and methods discussed herein provide for filtering out noisy application signatures to improve the precision of first packet application classification. In some implementations, the system receive application signatures from devices along with their network identifiers. Based upon the frequency at which identical application signatures appear as originating from distinct network environments, the system determines the validity of application signatures and avoids storing irrelevant information for routing network traffic.