LDPC-CC Data Rearrangement for Sequential Packet Erasures
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 packet erasures exceed the correction capability or occur sequentially 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) 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 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 the mechanical/computational approach of Reed-Solomon decoding with a probabilistic belief propagation mechanism. Instead of performing complex algebraic operations over finite fields, the system uses iterative message passing between check nodes and variable nodes, substituting deterministic calculation with probabilistic inference that achieves comparable correction performance with reduced computational complexity.
Solution Approach 2:
The patent transforms the error correction problem from a block code framework to a convolutional code framework by changing the temporal parameter structure. By introducing a time dimension and using convolutional encoding with memory, the system achieves better correction performance for burst errors without requiring large block lengths, thus avoiding the complexity increase associated with long block Reed-Solomon codes.
2Reliability
If Reed-Solomon code block length is increased to improve correction performance, then erasure correction capability improves, but calculation amount increases
Solution Approach 1:
The patent replaces the algebraic decoding mechanism of Reed-Solomon codes with a probabilistic belief propagation system. Instead of performing Gaussian elimination or other complex algebraic operations that scale poorly with block length, the system uses iterative message passing with simple sum-product operations at each node, reducing the computational burden while maintaining correction capability.
Solution Approach 2:
The patent introduces dynamic iterative processing where the decoding process adapts its behavior based on the received signal quality and error patterns. The belief propagation algorithm dynamically adjusts message passing between nodes based on current estimates, allowing the system to achieve good correction performance with fewer operations compared to static algebraic decoding methods.
3Adaptability or versatility
If LDPC-CC is used for encoding and decoding of information sequences of arbitrary length, then flexibility in code configuration is achieved, but multiple parity check matrices are needed to support different coding lengths and rates
Solution Approach 1:
The patent creates a universal LDPC-CC encoding framework that can handle arbitrary information sequence lengths and coding rates using a single parity check polynomial structure. By using convolutional encoding with a fixed parity check polynomial, the system achieves adaptability to different code lengths and rates without requiring separate parity check matrices for each configuration, unlike LDPC block codes.
Solution Approach 2:
The patent employs dynamic code construction where the effective code length and rate are adjusted by controlling the encoding duration and information bit input rate, rather than by changing the underlying parity check structure. This allows a single static parity check polynomial to generate dynamic code configurations adapted to different transmission requirements.
Data Source
AI summary
A loss correction encoding device having an improved capability of loss correction using LDPC-CC is disclosed. In the loss correction encoding device (120), a rearranging unit (122) rearranges information data contained in n information packets according to the constraint length Kmax and the encoding rate (q−1)/q of a cheek polynomial of the loss correction code used in a loss correction encoding unit (123). Specifically, the rearranging unit (122) 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 (122) distributes the information data to information blocks from n information packets (n satisfies formula (1)).Kmax×(q−1)≦n (1).