LDPC Decoder Dynamic Scaling for Faster Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data transfer systems face limitations in data decoding due to saturation issues and the inability to converge on a corrected data stream, particularly in systems using fixed or static scalars in encoding/decoding processes.
Innovation Solution
The implementation of data processing circuits with a decoder circuit and a scalar circuit that dynamically scales decoder messages using variable scalar values, allowing for rapid convergence in initial iterations and enhanced information delivery in subsequent iterations, particularly utilizing Low Density Parity Check (LDPC) decoders with variable nodes and check nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed or static scalar values are used in the decoding process, then the system design is simplified, but the decoding convergence speed decreases and saturation issues occur
Solution Approach 1:
The patent implements dynamic scalar values that change during the decoding process. The scalar circuit modifies scalar values based on iteration number and decoding state, allowing the system to adapt to changing conditions rather than using fixed values throughout the entire decoding process.
Solution Approach 2:
The patent changes the parameter of scalar values from fixed to variable. The scalar values are dynamically adjusted based on the decoding iteration and state, enabling the system to optimize convergence speed at different stages of the decoding process while avoiding saturation issues.
2Reliability
If multiple detection and decode iterations are performed, then the possibility of convergence is heightened, but the processing time increases
Solution Approach 1:
The patent uses dynamic scalar values that adapt during iterations. By adjusting scalars based on the current decoding state and iteration number, the system accelerates convergence in early iterations and maintains effectiveness in later iterations, reducing the total time required while maintaining high convergence reliability.
Solution Approach 2:
The patent applies different scalar values at different iteration stages. The scalar circuit periodically updates scalar values based on iteration count, creating a structured approach where early iterations use scalars optimized for rapid convergence and later iterations use scalars optimized for maintaining convergence progress.
3Ease of manufacture
If static scalar values are used, then the system is easier to implement, but data saturation occurs limiting system capability
Solution Approach 1:
The patent implements a scalar circuit that dynamically adjusts scalar values during decoding. This dynamic adjustment allows the system to adapt to different data conditions and prevent saturation, while the overall system structure remains relatively simple and buildable with standard circuit components.
Solution Approach 2:
The scalar circuit automatically adjusts scalar values based on the decoding state and iteration number without external intervention. The system self-regulates the scalar values to optimize performance for different data conditions, providing adaptability while maintaining ease of implementation through automated control.
Data Source
AI summary
Various embodiments of the present invention provide systems and methods for data processing using variable scaling.