Packet Distribution Core Selection via Decryption Hashing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

As packet-based telecommunications networks experience increasing volumes, bottlenecks arise in components that analyze packets for routing, leading to significant processing efforts and delays due to repeated access and use of packet headers by multiple devices along the network path.

Innovation Solution

Generating a header hash from identifiers in the packet header during decryption and using it to distribute packets among multiple processing cores for further processing and routing, reducing the need for direct access to the larger identifiers and minimizing processing resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple devices along the network path each perform access and use of the packet header, then routing decisions can be made at each device, but processing effort and delays become significant as packet volume scales

Engineering Contradiction:
Improverouting decision capabilityVSAvoidprocessing throughput
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent extracts only the essential identifier information from the packet header and creates a simplified representation (hash value) that can be processed efficiently. Instead of each device accessing and analyzing the complete header, the system extracts minimal necessary information for routing decisions, reducing processing overhead while maintaining routing capability.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms the packet header information through a hash function, changing the parameter representation from the original header format to a compact hash value. This parameter transformation enables faster comparison and matching operations at each network device, improving processing throughput while preserving the ability to make routing decisions based on destination identification.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If complete packet headers are accessed and processed by each device, then accurate routing information is obtained, but processing resources and time are significantly consumed

Engineering Contradiction:
Improverouting information accuracyVSAvoidprocessing delay
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent creates a copy of the essential routing information in the form of a hash value that can be quickly generated and compared. Instead of processing the complete original header at each device, the system uses this copied representation to make routing decisions, significantly reducing processing time while maintaining routing accuracy through proper hash table matching.

Inventive Principle:
Principle #26Copying

3Quantity of substance

If packet volume through the network continues to scale up, then network capacity increases, but processing effort and delays from header analysis become significant bottlenecks

Engineering Contradiction:
Improvepacket volumeVSAvoidprocessing complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent segments the packet processing task into two distinct parts: (1) generation of a compact hash value from the packet header at the source, and (2) efficient matching of this hash value against hash tables at intermediate devices. This segmentation allows each device to handle only the simplified hash matching operation, reducing processing complexity and enabling the network to scale to higher packet volumes without proportionally increasing device complexity.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS10686763B2Techniques for load balancing in a packet distribution system
Publication Date: 2020.06.16 INTEL CORP
  • US10686763B2 patent drawing
  • US10686763B2 patent drawing
  • US10686763B2 patent drawing

AI summary

Various embodiments are generally directed to techniques to distribute encrypted packets among multiple cores in a load-balanced manner for further processing. An apparatus may include a processor component; a decryption component to decrypt an encrypted packet to generate a decrypted packet from the encrypted packet, the encrypted packet comprising a header that comprises at least one field of information; a hash component to generate a header hash from the at least one field of information during decryption of at least a portion of the encrypted packet by the decryption component, the header hash comprising a smaller quantity of bits than the at least one field of information; and a distribution component to select a first core of multiple cores coupled to the processor component based on the header hash and to transmit the decrypted packet to the first core from the processor component. Other embodiments are described and claimed.