Ethernet Control Tag Classification for Parsing Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Interpreting control tags in Ethernet packets is challenging due to their varying number and size, leading to difficulties in determining proper parsing operations in network nodes.
Innovation Solution
A method and device that classify control tags into categories such as unambiguous interposable (UI) and ambiguous interposable (AI) based on EtherType fields, using content addressable memories (CAMs) and logic components to determine parsing operations and identify control packets, enabling flexible processing and offloading of control packets to the control plane.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If control tags are classified into multiple categories (UI, AI, UN) based on EtherType fields, then parsing accuracy and network node efficiency are improved, but device complexity increases due to the need for classification logic and CAMs
Solution Approach 1:
The patent segments control tags into distinct categories (UI, AI, UN) based on their EtherType fields and parsing characteristics. This segmentation allows each tag type to be processed with specialized logic, improving parsing accuracy while managing complexity through structured organization of the classification system.
Solution Approach 2:
The patent introduces content addressable memories (CAMs) as intermediary components to facilitate efficient classification of control tags. The CAMs act as mediators between the incoming packet data and the classification logic, enabling fast lookup and determination of tag categories without complex real-time analysis.
2Productivity
If control packets are identified and offloaded to the control plane, then network node efficiency is improved, but loss of time increases due to additional processing steps
Solution Approach 1:
The patent performs preliminary classification of control tags and identification of control packets within the data plane processing pipeline. By pre-categorizing tags into UI, AI, or UN types and identifying control packets early in the process, the system prepares packets for efficient offloading to the control plane, reducing overall processing time despite the additional offloading step.
3Reliability
If ambiguous interposable tags are processed by analyzing tag sequences, then parsing reliability is improved, but device complexity increases due to sequence analysis requirements
Solution Approach 1:
The patent segments the parsing process into distinct phases: classification of individual control tags into categories (UI, AI, UN), determination of tag sequences, and parsing based on the determined sequence. This segmentation allows reliable parsing of ambiguous tags through systematic analysis while managing complexity by breaking down the overall parsing function into manageable components.
Data Source
AI summary
A system may include receiving a packet, of a packet stream, including control tags in a header portion of the packet and classifying each of the control tags into a category selected from a set of possible categories. The set of possible categories may include an unambiguous interposable (UI) category that is assigned to a control tag that corresponds to an unambiguous parsing interpretation and that is interposable within a sequence of the control tags, and an ambiguous interposable (AI) category that is assigned to a control tag in which the control tag has an ambiguous parsing interpretation and in which the control tag is interposable within the sequence of the control tags. The method may further include determining parsing operations to perform for the packet based on the classified categories of the control tags and based on the packet stream of the packet.