MPLS Stack-Bottom Header for Faster Ancillary Data Parsing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MPLS architectures face challenges in efficiently indicating and analyzing ancillary data carried at the stack bottom, leading to high complexity and low data processing efficiency, particularly due to the need for multiple indicator labels and frequent label stack regeneration.
Innovation Solution
Introduce an MPLS manage header located after the stack bottom label and prior to ancillary data, which includes data type and offset length information, allowing nodes to determine required processing without reading the ancillary data, thereby reducing indication complexity and improving analysis efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple indicator labels are used to indicate ancillary data at the stack bottom, then the indication capability is improved, but the label stack depth and processing complexity increase significantly
Solution Approach 1:
The patent merges multiple indicator labels into a single unified indicator label structure. Instead of using separate indicator labels for different ancillary data types, the invention creates one comprehensive indicator label that can represent multiple data types through bit fields, thereby reducing label stack depth while maintaining indication capability.
Solution Approach 2:
The indicator label is designed with multi-functionality to handle various ancillary data types (G-ACH, IOAM, SFC metadata, etc.) within a single label structure. The bit fields within the indicator label can be configured to represent different data types and their respective offsets, making the single label universal for multiple indication purposes.
2Loss of information
If offset information is stored in the label stack, then the ancillary data location can be indicated, but the solution becomes susceptible to label stack changes and loses expansibility
Solution Approach 1:
The indicator label is segmented into multiple bit fields, where each bit field represents a specific ancillary data type and its offset information. This segmentation allows the indicator label to independently manage information for different data types without being affected by changes in other parts, thereby improving robustness against label stack changes.
Solution Approach 2:
The indicator label acts as an intermediary between the label stack and ancillary data. It contains offset information that points to the actual ancillary data locations, decoupling the label stack structure from the ancillary data structure. This intermediary role allows the system to maintain indication capability even when label stacks change, as the indicator label adapts to reflect the new structure.
3Ease of operation
If chained analysis mode similar to IPv6 expansion header is used, then ancillary data can be processed, but message processing efficiency decreases due to inability to directly acquire interested data
Solution Approach 1:
The indicator label is placed at the beginning of the ancillary data section, preliminarily indicating the types and locations of all ancillary data before actual processing begins. This allows receiving nodes to quickly identify and directly access the specific ancillary data they are interested in without having to sequentially analyze all ancillary data in chained mode, thereby improving processing efficiency.
Solution Approach 2:
The patent replaces the mechanical sequential analysis approach (chained analysis mode) with a direct access approach based on offset information. Instead of mechanically processing each ancillary data item in sequence, nodes can use the offset fields in the indicator label to directly jump to and process only the relevant data, substituting the mechanical sequential system with a more efficient random access system.
Data Source
Figure 1~3
Figure 4~6
Figure 7~9
AI summary
Provided in the embodiments of the present disclosure is an MPLS message encapsulation method. The method comprises: setting an MPLS manage header in a stack bottom structure of an MPLS label stack, where the MPLS manage header is located after a stack bottom label and prior to ancillary data; the MPLS manage header includes: a data type and offset length information of the ancillary data; and the stack bottom structure includes: the MPLS manage header and the ancillary data. By means of the technical solution, the problems in the related art of high complexity, low efficiency, etc. of an indication mode of an MPLS label stack are solved.