Bit Indexed Explicit Replication for Network Debugging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In deterministic networking, existing technologies face challenges in identifying which node generates a data packet copy and tracking the replication progress, especially in cases where replication occurs in the middle of a network path, making it difficult to determine the direction of transmission for debugging and fault identification.
Innovation Solution
The implementation of a bit index within a packet header that controls replication operations and identifies replication progress through a Bit Indexed Explicit Replication (BIER) structure, where each bit corresponds to a deterministic segment, allowing network devices to selectively replicate or eliminate packets based on set bits, and a Path Computation Element (PCE) generates and updates the bit index to manage replication points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If replication operations are performed in the middle of a network path, then packet delivery reliability is improved, but the ability to identify which node generates a data packet copy and track replication progress deteriorates
Solution Approach 1:
The patent uses bit index values (0 and 1) as visual markers analogous to color changes to indicate replication status. Each bit position corresponds to a specific deterministic segment, and the bit value indicates whether replication occurred at that segment's endpoint. This allows network devices to visually distinguish between different packet copies and their replication origins, solving the tracking difficulty while maintaining reliability.
Solution Approach 2:
The bit index acts as an intermediary data structure between the replication operation and the packet header. It mediates the information flow by explicitly recording which segments require replication and which nodes should forward packets, enabling reliable intermediate replication while providing clear identification of replication points for debugging purposes.
2Measurement precision
If a bit index is added to the packet header to control replication, then replication control precision is improved, but packet header complexity increases
Solution Approach 1:
The bit index is segmented into individual bits, where each bit corresponds to a specific deterministic segment. This segmentation allows the system to control replication at granular levels without requiring a completely new header structure. The bit index can be implemented as a compact field within the existing packet header, adding precision without excessive complexity.
Solution Approach 2:
The bit index serves multiple functions: it controls replication operations, identifies replication points, tracks packet transmission directions, and provides debugging information. By making the bit index multi-functional, the patent avoids adding separate control mechanisms for each function, thereby limiting the increase in header complexity while achieving high precision replication control.
Data Source
Figure 1A
Figure 1B
Figure 2
AI summary
In one embodiment, a method comprises receiving, by a network device in a deterministic data network comprising deterministic segments, a data packet comprising a packet header having a bit index, each bit in the bit index associated with a corresponding one of the deterministic segments; the network device identifying, from the bit index, a first bit for a corresponding first deterministic segment and a second bit for a corresponding second deterministic segment, wherein the network device is an endpoint for transmitting deterministic traffic into each of the first and second deterministic segments; the network device selectively executing a replication operation of the data packet for transmission of a replicated data packet into each of the first and second deterministic segments, based on the network device detecting the corresponding first and second bits set for replication.