LDPC Packet Rearrangement for Reliable Erasure Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing erasure correction methods, such as Reed-Solomon codes, are inadequate for handling a large number of packet erasures in applications like moving image streaming, especially when erasures occur due to fading in radio communication paths, and require increased block lengths that lead to higher calculation costs and circuit complexity.
Innovation Solution
The implementation of a Low-Density Parity-Check Convolutional Code (LDPC-CC) system that arranges information data according to a specific constraint length and coding rate, generating parity packets through erasure correction coding to improve decoding capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the block length of Reed-Solomon code is increased to improve correction performance, then the erasure correction capability is improved, but the calculation amount and circuit scale increase
Solution Approach 1:
The patent changes the fundamental parameter of the error correction code from Reed-Solomon to LDPC code, which has different structural characteristics. LDPC codes use a sparse parity-check matrix with low density, enabling efficient encoding and decoding algorithms that achieve high correction performance without requiring large block lengths, thus reducing circuit complexity while maintaining reliability
Solution Approach 2:
The patent segments the parity-check matrix of the LDPC code into a structured form that can be efficiently implemented in hardware. By organizing the matrix with specific patterns and properties, the decoding process can be divided into manageable stages, reducing the overall circuit scale while maintaining the ability to correct a large number of erasures
2Reliability
If the block length of Reed-Solomon code is increased to improve correction performance, then the erasure correction capability is improved, but the calculation amount increases
Solution Approach 1:
The patent changes the code type from Reed-Solomon to LDPC, which fundamentally alters the computational complexity characteristics. LDPC codes support iterative decoding algorithms whose calculation amount grows linearly with block length rather than quadratically, enabling high correction performance with reduced computational burden and power consumption
Solution Approach 2:
The patent employs iterative decoding for LDPC codes where the decoding process dynamically adjusts based on feedback from previous iterations. This dynamic approach allows the system to achieve high correction performance with adaptive calculation effort, avoiding the fixed high computational cost required by increased block length in Reed-Solomon codes
Data Source
AI summary
A loss correction encoding device having an improved capability of loss correction using LDPC-CC includes a rearranging unit that rearranges information data contained in n information packets according to the constraint length Kmax and the encoding rate (q−1)/q of a check polynomial of the loss correction code used in a loss correction encoding unit. Specifically, the rearranging unit rearranges the information data in such a way that continuous Kmax×(q−1) pieces of information data after rearrangement are contained in different information packets. The rearranging unit distributes the information data to information blocks from n information packets, where n satisfies the formula Kmax×(q−1)≤n.