Graph-Based Line Code Conversion for Error Correction and DC Balance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing line codes, such as 4b/10b, suffer from low coding rate, high overhead, and limited ability to implement additional functions due to constraint conditions, making error correction and processing performance suboptimal.
Innovation Solution
A conversion device and method that utilizes a graph-based approach to create a line code with error correction capabilities by converting bit strings through a clique search, ensuring predetermined constraint conditions are met, including DC-balancing, clock recovery, and error detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If 4b/10b line code is used to achieve error correction function, then error correction capability is improved, but coding rate deteriorates (only 60%)
Solution Approach 1:
The patent changes the fundamental parameters of the line code system by using 5b6b encoding instead of 4b10b, achieving a 50% coding rate that provides better error correction capability while maintaining DC-balancing and other constraint conditions. This parameter change resolves the contradiction by finding an optimal middle ground between error correction and coding efficiency.
2Reliability
If 4b/10b line code is used to satisfy constraint conditions, then DC-balancing and error detection are improved, but overhead increases significantly
Solution Approach 1:
The patent changes the encoding parameters from 4b10b to 5b6b, reducing the overhead portion while maintaining DC-balancing capability. The 5b6b code structure provides sufficient control characters for DC-balancing with less redundant information, thereby reducing overhead while preserving the essential constraint satisfaction.
3Reliability
If 4b/10b line code is used to meet constraint conditions, then error detection is improved, but processing performance deteriorates
Solution Approach 1:
The patent changes the code structure from 4b10b to 5b6b, creating a more efficient encoding scheme that maintains error detection capability through appropriate constraint conditions while reducing processing complexity. The shorter code length and optimized structure enable faster encoding and decoding operations, improving processing performance without sacrificing error detection reliability.
4Reliability
If only 20 codes are used in 4b/10b line code for control information, then constraint condition satisfaction is improved, but adaptability deteriorates (difficult to implement additional functions)
Solution Approach 1:
The patent changes the code allocation parameters by using 5b6b encoding with a different distribution of control characters and data characters. This provides a more flexible framework for implementing additional functions while maintaining constraint condition satisfaction, as the encoding structure allows for more versatile control information representation.
Data Source
AI summary
One aspect of the present disclosure relates to a conversion device including an acquisition unit configured to acquire a first bit string having a first bit length L1; a conversion unit configured to convert, in accordance with conversion information that associates respective bit strings each having the first bit length L1 with bit strings each having a second bit length L2 uniquely assigned to the respective bit strings, the first bit string into a second bit string having the second bit length L2. The conversion information is created by searching for a clique that includes 2L1 or more nodes, from a graph including nodes and an edge representing the bit strings each having the second bit length L2 that satisfy a predetermined constraint condition.