Packet Rearrangement for LDPC Erasure Correction in Communication Links
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 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 packet erasures exceed the correction capability or occur over a long period due to fading in radio communication paths, leading to ineffective error correction.
Innovation Solution
The implementation of a low-density parity-check convolutional code (LDPC-CC) erasure correction coding apparatus and method, which arranges information data according to a constraint length and coding rate to generate parity packets, improving erasure correction capability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Reed-Solomon code is used for erasure correction, then correction performance can be improved by increasing block length, but calculation amount and circuit scale increase
Solution Approach 1:
The patent replaces traditional Reed-Solomon code mechanisms with LDPC-CC (Low-Density Parity-Check Convolutional Code) mechanisms. LDPC-CC uses a different mathematical approach based on sparse parity-check matrices and belief propagation algorithms, which achieves comparable or superior erasure correction performance with reduced computational complexity and smaller circuit implementation requirements.
2Reliability
If Reed-Solomon code is used for erasure correction, then correction capability can be improved by increasing block length, but calculation amount increases
Solution Approach 1:
The patent substitutes Reed-Solomon decoding calculations with LDPC-CC belief propagation algorithms. The LDPC-CC approach uses iterative message passing on a factor graph derived from the parity-check matrix, which converges faster and requires fewer operations to achieve the same correction capability, thereby reducing the calculation amount and power consumption.
Solution Approach 2:
The patent changes the fundamental parameters of the error correction system by adopting LDPC-CC with specific parity-check matrix structures and coding rates. This parameter change enables achieving high erasure correction capability with more efficient encoding and decoding operations compared to traditional Reed-Solomon codes.
3Productivity
If LDPC-CC is used for erasure correction, then feasible time and calculation cost are achieved, but handling extensive packet erasures requires improved arrangement methods
Solution Approach 1:
The patent segments information packets into multiple code blocks for LDPC-CC encoding. Each code block is processed independently with its own parity packets, allowing the system to handle extensive erasures by ensuring that erasures in one block do not propagate to other blocks. This segmentation maintains high correction effectiveness while preserving encoding/decoding speed.
Solution Approach 2:
The patent performs preliminary arrangement of information packets before LDPC-CC encoding, organizing them in a specific sequence that optimizes the distribution of information across code blocks. This preliminary action ensures that even when extensive erasures occur, the remaining packets are optimally positioned for successful decoding, thereby improving reliability without sacrificing productivity.
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.