LDPC-CC Data Rearrangement for Sequential Packet Erasures
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 erasures occur due to fading in radio communication paths, leading to ineffective error correction and increased computational burdens.
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 changes the fundamental parameters 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 reduce calculation complexity while maintaining or improving correction performance for packet erasure channels
Solution Approach 2:
The patent replaces the traditional Reed-Solomon code mechanism with an LDPC code mechanism that uses iterative belief propagation decoding. This substitution transforms the error correction approach from algebraic decoding to probabilistic iterative decoding, reducing the computational burden and circuit complexity while handling burst erasures more effectively
2Reliability
If Reed-Solomon code is used for erasure correction, then correction capability can be enhanced, but calculation amount increases
Solution Approach 1:
The patent changes the code structure parameters from Reed-Solomon's dense generator matrix to LDPC's sparse parity-check matrix. This parameter change enables the use of iterative decoding algorithms that converge quickly with fewer calculations, reducing energy consumption while maintaining correction capability
Solution Approach 2:
The patent applies iterative decoding with a predetermined number of iterations, performing only the necessary calculations to achieve convergence or reach the iteration limit. This partial action approach avoids exhaustive computation while achieving sufficient correction performance for the given channel conditions
3Productivity
If LDPC code is used for packet erasure correction, then encoding and decoding can be performed with feasible time and calculation cost, but correction performance may be insufficient for large number of sequential erasures
Solution Approach 1:
The patent performs preliminary arrangement of information packets before LDPC encoding, organizing them in a specific sequence that optimizes the distribution of erasures across code blocks. This preliminary action ensures that even when burst erasures occur, they are distributed in a manner that the iterative decoder can effectively handle
Solution Approach 2:
The patent employs dynamic iterative decoding with adaptive termination, where the decoding process continues until convergence or a predetermined iteration limit is reached. This dynamic approach allows the system to adapt to different erasure patterns and channel conditions, maintaining good correction performance for burst erasures while preserving fast decoding speed
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.