Adaptive Memory Subsystem LDPC Decoding Across Varying HRER Ranges
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory sub-systems face challenges in effectively handling a wide range of high-reliability error rates, particularly due to varying high readability error rates (HRER) that impact the decoding performance and reliability of LDPC codes, leading to increased codeword error rates (CWER).
Innovation Solution
A memory sub-system controller iteratively performs a decoding algorithm with multiple LLR sets and transformation sets optimized for different HRER ranges, using a combination of likelihood ratios (LLR) and transformation parameters to enhance error correction capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single decoding algorithm with fixed parameters is used, then the device complexity is low, but the reliability deteriorates when facing varying HRER conditions
Solution Approach 1:
The patent implements dynamic parameter adjustment by maintaining multiple LLR sets and transformation sets that can be adaptively selected based on detected error conditions. The decoding algorithm transitions from static to dynamic operation, where parameters such as LLR thresholds and transformation matrices are adjusted in real-time according to the observed error patterns and HRER levels, thereby improving reliability without requiring a completely separate decoder for each condition
Solution Approach 2:
The patent systematically varies key decoding parameters including LLR (log-likelihood ratio) thresholds, transformation set selections, and iteration limits to optimize performance across different HRER ranges. By pre-computing and storing multiple parameter sets corresponding to different error rate scenarios, the system can switch between parameter configurations to match current operating conditions, resolving the contradiction between maintaining low complexity and achieving high reliability
2Reliability
If multiple LLR sets and transformation sets are used to handle varying HRER, then the reliability improves, but the device complexity increases
Solution Approach 1:
The patent divides the decoding parameter space into discrete segments or sets, where each LLR set and transformation set corresponds to a specific HRER range or error pattern type. This segmentation allows the system to manage complexity by organizing parameters into manageable groups that can be independently stored and selected, rather than maintaining a continuous parameter space. The segmented approach enables efficient memory utilization and faster parameter selection during decoding operations
Solution Approach 2:
The patent performs preliminary computation and optimization of multiple LLR sets and transformation sets during system initialization or manufacturing, storing the pre-computed parameter sets in memory. This preliminary action eliminates the need for real-time computation of complex transformations during actual decoding operations, significantly reducing processing complexity while maintaining the ability to handle varying HRER conditions. The pre-computed sets are readily available for immediate selection based on detected error conditions
Data Source
AI summary
A first decoding parameter is obtained responsive to the determining that the codeword contains errors. The first decoding parameter includes a first likelihood set and a first transformation set. A decoding operation with the first decoding parameter is performed. A second decoding parameter is obtained responsive to determining that the decoding operation did not correct the errors of the codeword. The second decoding parameter includes a second likelihood set and a second transformation set. The decoding operation with the second decoding parameter is performed.