Connection Lookup Index and Way Storage for Fast Peer Matching

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In connection-oriented communication, the existing methods require running a hash function and comparing connection information multiple times, which is time-consuming, especially when a server connects with a large number of clients, leading to inefficiencies in building and maintaining connections.

Innovation Solution

Generating and storing connection-associated information, including an index and a way, within the packet headers to facilitate fast look-up by using fewer bits to represent these elements, thereby reducing the time needed to search and compare connection information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If hash function and connection information comparison are performed multiple times to establish connections, then connection reliability is ensured, but time consumption increases significantly

Engineering Contradiction:

Improveconnection reliabilityVSAvoidtime consumption

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent pre-calculates and stores hash values of connection information (source IP, destination IP, source port, destination port) in a connection table before actual connection establishment. When a connection request arrives, the system performs a direct hash value comparison instead of running the hash function multiple times, significantly reducing time consumption while maintaining connection reliability through pre-prepared connection data.

Inventive Principle:

Principle #10Preliminary action

2Productivity

If connection table size is increased to handle more clients (64k+), then system capacity is improved, but lookup complexity and time increase

Engineering Contradiction:

Improvesystem capacityVSAvoidlookup complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the connection lookup process into two independent parts: hash value comparison (using pre-calculated hashes) and connection information verification. This segmentation allows the system to handle large numbers of clients (64k+) by breaking down the complex lookup operation into simpler, faster sub-operations, reducing overall lookup complexity while maintaining high system capacity.

Inventive Principle:

Principle #1Segmentation

3Measurement precision

If full connection information comparison is performed for each connection lookup, then accuracy is maintained, but processing speed decreases

Engineering Contradiction:

Improvecomparison accuracyVSAvoidprocessing speed

Core Design Contradiction:

Measurement precisionVSSpeed

Solution Approach 1:

The patent extracts the hash value from the connection information (source IP, destination IP, source port, destination port) and performs comparison using only this extracted hash value rather than comparing all connection information fields. This extraction approach maintains comparison accuracy by using the unique hash identifier while dramatically increasing processing speed by avoiding redundant comparisons of full connection data.

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS7684402B2Method and network device for fast look-up in a connection-oriented communication

Publication Date: 2010.03.23 FARADAY TECH CORP

AI summary

A method for fast look-up in a connection-oriented communication includes generating a connection associated information according to peer information of a first peer and a second peer, storing the connection associated information into the first peer and the second peer, and building a connection between the first peer and the second peer according to the connection associated information. Generating the connection associated information includes generating an index and a way.