Non-Volatile Memory Decoding Using Multi-Read Soft Reliability Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Non-volatile storage devices face challenges in reliably decoding data due to noise, particularly 1/f noise from electron trapping and de-trapping in trap sites, which affects the accuracy of read operations.
Innovation Solution
A method involving multiple read operations and iterative probabilistic decoding using reliability metrics, such as logarithmic likelihood ratios, to improve decoding accuracy by adjusting and refining probability metrics based on subsequent read operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple read operations are performed to improve decoding reliability, then data retrieval accuracy is improved, but read operation time and system complexity increase
Solution Approach 1:
The patent applies preliminary action by performing multiple read operations before the final decoding process. Specifically, it conducts an initial read to obtain first data, performs probabilistic decoding to generate first probability metrics, then performs a second read to obtain second data, and finally uses both datasets to generate improved probability metrics for conclusive decoding. This preliminary gathering of multiple readings before final decision-making improves reliability without requiring all reads to complete before decoding begins.
Solution Approach 2:
The patent implements dynamics by making the decoding process adaptive and iterative rather than static. The system dynamically adjusts the decoding strategy based on intermediate results: it performs probabilistic decoding after the first read, evaluates the results, then performs a second read, and finally combines both readings with updated probability metrics. This dynamic, multi-stage approach allows the system to adapt its reading and decoding behavior based on evolving information, improving reliability while managing time efficiently.
2Object-affected harmful factors
If multiple read operations with iterative probabilistic decoding are used to improve decoding accuracy, then noise resistance is improved, but computational complexity increases
Solution Approach 1:
The patent applies feedback by using the results of intermediate decoding operations to guide subsequent read and decode operations. After the first read and initial probabilistic decoding, the system generates first probability metrics that serve as feedback. This feedback informs the second read operation and the final decoding process, where both readings and their respective probability metrics are combined. The feedback loop allows the system to iteratively improve decoding accuracy by learning from intermediate results, effectively resisting noise through adaptive refinement.
Solution Approach 2:
The patent introduces another dimension by transitioning from single-read hard decisions to multi-read soft decision decoding with probability metrics. Instead of making binary decisions based on single readings, the system operates in a probabilistic domain, maintaining and combining probability metrics from multiple reads. This dimensional shift from discrete to continuous probabilistic space allows for more nuanced noise resistance, where uncertainty is explicitly modeled and combined, improving robustness against noise while managing complexity through mathematical frameworks.
3Measurement precision
If probability metrics are adjusted based on subsequent read operations to improve decoding precision, then measurement precision is improved, but processing time increases
Solution Approach 1:
The patent applies preliminary action by performing the initial read and generating first probability metrics before the final decoding operation. This preliminary processing prepares the data in advance, organizing it into a probabilistic framework that facilitates the final high-precision decoding. By doing this preparation work upfront, the system enables faster final processing while maintaining high precision, as the foundational probabilistic analysis is already complete when the second read and final combination occur.
Data Source
AI summary
Data stored in non-volatile storage is decoded using iterative probabilistic decoding and multiple read operations to achieve greater reliability. An error correcting code such as a low density parity check code may be used. In one approach, initial reliability metrics, such as logarithmic likelihood ratios, are used in decoding read data of a set of non-volatile storage element. The decoding attempts to converge by adjusting the reliability metrics for bits in code words which represent the sensed state. If convergence does not occur, e.g., within a set time period, the state of the non-volatile storage element is sensed again, current values of the reliability metrics in the decoder are adjusted, and the decoding again attempts to converge. In another approach, the initial reliability metrics are based on multiple reads. Tables which store the reliability metrics and adjustments based on the sensed states can be prepared before decoding occurs.