Flash Memory Soft Data Generation via Statistical Hard Read Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current flash memory devices face limitations in decoding performance due to the lack of soft data, which can improve error rate performance, and existing methods for mitigating intercell interference and distortions are not fully effective.
Innovation Solution
The development of methods and apparatus for generating soft data in flash memory devices by obtaining hard read values and generating associated soft data values based on statistics, including bit-based, cell-based, and pattern-dependent statistics, to enhance decoding performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If hard data only is provided to the flash control system, then the device complexity is reduced, but the decoding performance and error rate performance deteriorate
Solution Approach 1:
A soft data generation module is introduced as an intermediary component between the flash memory array and the flash control system. This module receives hard data values from the memory array, applies statistical models and algorithms to generate corresponding soft data values, and provides both hard and soft data to the controller. This intermediary structure enables soft data generation without requiring fundamental changes to the existing flash memory architecture, thus improving decoding performance while maintaining reasonable device complexity
Solution Approach 2:
The patent generates soft data as a copy or derivative of the hard data by applying statistical transformations. Instead of requiring completely new sensing mechanisms, the system creates soft data representations (such as probability distributions or reliability metrics) based on the existing hard data readings. This copying approach allows the system to benefit from soft data for error correction while avoiding the need for entirely new hardware structures
2Reliability
If soft data generation based on statistics is implemented, then the decoding performance improves, but the device complexity increases
Solution Approach 1:
The patent pre-calculates and stores statistical parameters (such as mean values, standard deviations, and probability distribution characteristics) that are needed for soft data generation. These statistical models are prepared in advance based on memory array characteristics and stored for later use during read operations. This preliminary action reduces the computational burden during actual soft data generation, as the system only needs to apply pre-computed statistical transformations rather than performing complex real-time analysis
Solution Approach 2:
The patent transforms hard data values into soft data by changing the parameter representation from discrete binary values to continuous probability distributions or reliability metrics. This parameter transformation allows the system to encode additional information about data quality and reliability without requiring fundamentally new processing architectures. The statistical models enable this parameter change to be performed using relatively simple mathematical operations
Data Source
AI summary
Methods and apparatus are provided for soft data generation for memory devices. At least one soft data value is generated for a memory device, by obtaining at least one hard read value; and generating the soft data value associated with the at least one hard read value based on statistics for reading the hard read value. The hard read value may be one or more of data bits, voltage levels, current levels and resistance levels. The generated soft data value may be one or more of (i) a soft read value that is used to generate one or more log likelihood ratios, and (ii) one or more log likelihood ratios. The statistics comprise one or more of bit-based statistics and cell-based statistics. The statistics may also optionally comprise pattern-dependent disturbance of at least one aggressor cell on the target cell, as well as location-specific statistics. At least one soft data value can be generated for a memory device, by obtaining a soft read value; and generating the soft data value associated with the soft read value based on statistics for reading the soft read value, wherein the statistics comprise one or more of location-specific statistics and pattern-dependent statistics.