Soft-Information Decoding for Uncorrectable Flash Memory Errors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory devices experience data loss due to various error-inducing factors such as pseudo-random electrical fluctuations, defects, and degradation, leading to uncorrectable errors that existing technologies struggle to recover.
Innovation Solution
A method is introduced to update soft information values for error correction codes in non-volatile memory devices, allowing for additional decoding attempts by adjusting soft information values based on parity and magnitude from other units of memory, enabling recovery of data that would otherwise be lost.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional decoding operations are used on memory units, then decoding speed is maintained, but data recovery capability for uncorrectable errors is insufficient
Solution Approach 1:
The patent applies preliminary action by performing syndrome decoding and identifying uncorrectable errors before attempting traditional decoding. This preliminary error identification allows the system to prepare updated soft information values in advance, improving the chances of successful recovery in subsequent decoding attempts without compromising overall decoding speed.
Solution Approach 2:
The patent introduces an intermediary mechanism by using updated soft information values derived from syndrome decoding results. These updated soft information values act as a mediator between the initial decoding failure and the recovery process, providing enhanced information that bridges the gap and enables successful data recovery from uncorrectable errors.
2Reliability
If soft information values are updated based on syndrome decoding, then data recovery from uncorrectable errors improves, but computational overhead increases
Solution Approach 1:
The patent applies partial action by selectively updating soft information values only for bits identified as uncorrectable through syndrome decoding. Rather than processing all data bits uniformly, the system focuses computational resources only on the specific bits that require recovery, reducing overall computational overhead while maintaining data integrity for affected portions.
3Reliability
If multiple decoding operations are performed on the same memory unit, then data recovery success rate increases, but processing time increases
Solution Approach 1:
The patent applies periodic action by implementing a structured multi-stage decoding process. The system performs syndrome decoding periodically to identify uncorrectable errors, then conducts targeted updates of soft information values, followed by recovery decoding attempts. This periodic, rhythmic approach to decoding operations optimizes the balance between recovery success rate and processing time by avoiding continuous, redundant decoding attempts.
Data Source
AI summary
A method is provided that includes performing first decoding operations on data obtained from a plurality of units of memory using soft information values for the plurality of units of memory, where the plurality of units of memory includes an error correction stripe. The method further includes determining that two or more units of memory have uncorrectable errors. The method further includes updating a soft information value for a first unit of memory in accordance with a magnitude of a soft information value for a second unit and a direction based on parity of the error correction stripe excluding the first unit, where the first unit of memory and the second unit of memory are included in the two or more units of memory that have uncorrectable errors. The method further includes performing a second decoding operation on data obtained from the first unit using the updated soft information value.