Flash Memory Controller Gaussian Threshold Voltage Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage devices face challenges in accurately reading data due to variations in threshold voltage distributions of memory cells, leading to errors that exceed the error correction capability, especially when read voltages are not positioned between the distributions.
Innovation Solution
A data storage device with a controller that performs Gaussian modeling to estimate average threshold voltages and optimal read voltages by calculating section cell numbers, area ratios, and adjusting reference voltages to position them within reliable sections, minimizing error bits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If read voltage is applied to memory cells for data reading, then data can be retrieved from storage, but threshold voltage variations cause error bits that exceed correction capability
Solution Approach 1:
The controller performs preliminary Gaussian modeling to estimate threshold voltage distributions before actual data reading. By calculating section cell numbers and determining average threshold voltages in advance, the system prepares optimal read voltages that account for threshold variations, thereby reducing error bits before they occur during data retrieval operations
Solution Approach 2:
The system dynamically adjusts read voltage parameters based on estimated threshold voltage distributions. By changing the read voltage to match the optimal voltage derived from Gaussian modeling, the controller compensates for threshold voltage shifts and maintains accurate data reading despite variations in memory cell characteristics
2Measurement precision
If Gaussian modeling is performed to estimate average threshold voltage, then optimal read voltage can be determined, but additional processing time and complexity are required
Solution Approach 1:
The controller divides the memory cell population into multiple sections based on threshold voltage ranges and calculates section cell numbers for each segment. This segmentation approach simplifies the Gaussian modeling process by handling smaller subsets of data separately, making the overall estimation more manageable and less computationally intensive while maintaining accuracy
Solution Approach 2:
The patent introduces section cell numbers as an intermediary parameter that bridges raw read data and threshold voltage estimation. By first counting cells in different voltage sections and then using these counts to determine average threshold voltages, the system creates a simplified intermediate representation that reduces the complexity of direct Gaussian fitting on raw data
3Ease of operation
If reference voltage is used for reading data, then simple voltage application is possible, but reference voltage may not be positioned within the threshold voltage distribution leading to reading errors
Solution Approach 1:
The system transitions from using a fixed reference voltage to a dynamic read voltage that adapts to threshold voltage distribution changes. By continuously estimating the threshold voltage distribution through Gaussian modeling and adjusting the read voltage accordingly, the system maintains optimal reading conditions even as memory cell characteristics drift over time or due to wear
Data Source
AI summary
A data storage device includes a memory device including a plurality of memory cells; and a controller suitable for determining, based on data read from the plurality of memory cells, section cell numbers corresponding to threshold voltage sections, and for determining an average threshold voltage of a threshold voltage distribution selected among a plurality of threshold voltage distributions of the memory cells which are estimated based on the section cell numbers, based on a Gaussian distribution function.