Time-Varying QC-LDPC Convolutional Coding With Periodic Check Matrices

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing FEC technologies, such as LDPC block codes and time-invariant LDPC convolutional codes, face challenges in achieving high-gain performance and high throughput, especially in high-speed optical transmission systems, due to high complexity, error floor issues, and low parallelism, making them unsuitable for optical transmission.

Innovation Solution

A time-varying periodic LDPC convolutional code with a quasi-cyclic LDPC structure is developed, where check matrix parameters are determined based on system performance and complexity, and a QC-LDPC block code is constructed to form a check matrix that splits into sub-matrices, reducing implementation complexity and enhancing decoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If LDPC block codes are used to achieve satisfactory error correction performance, then a very great code length is required, but this leads to very high implementation complexity

Engineering Contradiction:
Improveerror correction performanceVSAvoidimplementation complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent segments the LDPC block code into multiple smaller sub-blocks that are processed as separate LDPC convolutional code units. Each sub-block is encoded independently with its own check matrix, allowing parallel processing and reducing the complexity of any single encoding operation while maintaining overall error correction performance through the convolutional structure.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from static LDPC block codes to dynamic LDPC convolutional codes where the check matrix changes over time according to a time-varying periodic pattern. This dynamic structure allows the system to achieve good error correction performance with shorter effective code lengths at any given time step, reducing implementation complexity while maintaining reliability.

Inventive Principle:
Principle #15Dynamics

2Device complexity

If time-invariant LDPC convolutional code is used, then the structure is simple, but the performance is weak and error floor risk is high

Engineering Contradiction:
Improvecheck matrix structure simplicityVSAvoiderror correction performance
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent introduces time-varying periodicity into the LDPC convolutional code by periodically changing the check matrix according to a predetermined pattern. This dynamic variation prevents error floors by ensuring that the same error patterns do not repeatedly occur, while the periodic nature maintains reasonable structural simplicity and implementation feasibility.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the parameters of the check matrix over time, specifically the connection patterns between variable nodes and check nodes in the Tanner graph representation. By varying these parameters periodically, the system achieves better error correction performance and avoids error floors while maintaining a systematic approach to code design.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If time-varying periodic LDPC convolutional code is used, then error correction performance improves, but decoding complexity increases

Engineering Contradiction:
Improveerror correction performanceVSAvoiddecoding complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent employs periodic action in the check matrix variation, where the time-varying pattern repeats after a fixed period. This periodicity allows the decoder to reuse computation patterns and storage structures across different time periods, reducing the effective decoding complexity compared to fully time-varying codes. The receiver can implement efficient buffering and processing by exploiting this periodic structure.

Inventive Principle:
Principle #19Periodic action

4Reliability

If LDPC block code is used, then error correction capability is strong, but transmission efficiency is reduced due to great code length requirements

Engineering Contradiction:
Improveerror correction capabilityVSAvoidtransmission efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent segments the long LDPC block code into multiple shorter LDPC convolutional code units that are transmitted sequentially. This segmentation allows for more efficient transmission by reducing the latency associated with encoding and decoding very long codes, while the convolutional structure enables continuous encoding and decoding operations that improve transmission efficiency.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary organization of the code structure by pre-defining the time-varying periodic pattern and sub-block divisions. This preliminary structuring allows the transmitter and receiver to efficiently manage the encoding and decoding processes without requiring buffering of extremely long code sequences, thereby improving transmission efficiency while maintaining strong error correction capability.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2790327B1Adaptive coding using time-varying periodic LDPC convolutional codes based on quasi-cyclic LDPC block codes
Publication Date: 2016.08.24 HUAWEI TECH CO LTD
  • EP2790327B1 patent drawingFigure 1
  • EP2790327B1 patent drawingFigure 2
  • EP2790327B1 patent drawingFigure 3~5

AI summary

The present invention discloses a coding and decoding method, apparatus, and system for forward error correction, and pertains to the field of communications. The method includes: determining check matrix parameters of time-varying periodic LDPC convolutional code according to performance a transmission system, complexity of the transmission system, and a synchronization manner for code word alignment, constructing a QC-LDPC check matrix according to the determined check matrix parameters, and obtaining a check matrix Hc of the time-varying periodic LDPC convolutional code according to the QC-LDPC check matrix; de-blocking, according to requirements of the Hc, data to be coded, and coding data of each sub-block according to the Hc, so as to obtain multiple code words of the LDPC convolutional code; and adding the multiple code words of the LDPC convolutional code in a data frame and sending the data frame. In the present invention, forward error correction is performed by using time-varying periodic LDPC convolutional code with a QC-LDPC structure. The check characteristics of the QC-LDPC may decrease the complexity of the check; furthermore, the performance of LDPC convolutional code is better than that of LDPC block codes, and the performance of time-varying LDPC convolutional code is better than that of time-invariant LDPC convolutional code. Therefore, the present invention may be applicable to the high-speed optical transmission system and meet the requirements of high-gain performance and high throughput.