Soft Bit Reference Level Calibration Using Mutual Information
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The quality of soft reliability information in non-volatile memory systems is compromised due to incorrect sensing of soft bit reference voltages, leading to suboptimal error correction capabilities, as the placement of soft bit thresholds is a trade-off between marking errors as low reliability and flagging correct bits as unreliable.
Innovation Solution
Calibrating the soft bit reference levels by determining the mutual information between original programmed data and data sensed at candidate soft bit reference levels, selecting the soft bit reference levels that maximize mutual information to improve the quality of reliability information for error correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If soft bit reference voltages are set close to hard bit reference voltage, then fewer correct bits are marked as unreliable, but many errors are not marked as low reliability
Solution Approach 1:
The patent performs preliminary sensing of memory cells at multiple candidate soft bit reference voltage levels before final decoding. By pre-sensing at various candidate voltages and evaluating the quality of soft reliability information at each level, the system determines the optimal soft bit reference voltage in advance, thereby resolving the trade-off between marking errors correctly and avoiding false unreliable markings.
Solution Approach 2:
The patent systematically varies the soft bit reference voltage parameter across multiple candidate levels to identify the optimal setting. By changing this voltage parameter and evaluating its effect on soft reliability information quality, the system finds the specific voltage level that maximizes error detection accuracy while minimizing false positives, thus resolving the contradiction between these two requirements.
2Reliability
If soft bit reference voltages are set far from hard bit reference voltage, then more errors are marked as low reliability, but too many correct bits are marked as unreliable
Solution Approach 1:
The patent performs preliminary sensing and evaluation at multiple candidate soft bit reference voltage levels before final decoding. By pre-assessing the quality of soft reliability information at each candidate voltage level, the system identifies the optimal voltage that maximizes error detection while preserving correct bit identification, thereby preventing loss of reliable information.
Solution Approach 2:
The patent uses feedback from preliminary sensing results to adjust and select the optimal soft bit reference voltage. By evaluating how well each candidate voltage level performs in terms of error detection and correct bit preservation, the system feeds this information back to determine the final optimal voltage setting, thus resolving the contradiction between error detection and information preservation.
3Reliability
If incorrect soft bit reference levels are used, then soft reliability information quality degrades, but error correction capability is reduced
Solution Approach 1:
The patent systematically tests multiple candidate soft bit reference voltage parameters to identify the optimal setting. By changing and comparing the performance at different voltage levels, the system determines which parameter value maximizes both sensing accuracy and error correction capability, thus resolving the contradiction between these two dependent qualities.
Solution Approach 2:
The patent performs preliminary sensing and quality evaluation at candidate voltage levels before final decoding operations. This preliminary assessment allows the system to identify and select the reference voltage level that ensures high-quality soft reliability information, thereby guaranteeing optimal error correction capability without sacrificing sensing accuracy.
Data Source
AI summary
Calibration of soft bit reference levels in a non-volatile memory system is disclosed. When the soft bit reference levels are to be calibrated, encoded data is read from a group of the memory cells. The encoded data is decoded and error corrected. Therefore, the original data that was programmed into the memory cells is recovered. The group of memory cells are sensed at candidate soft bit reference levels, and possibly other reference levels. For each candidate soft bit reference level, mutual information between the original programmed data and the data for that candidate soft bit reference level is determined. The mutual information serves as a good measure for how well the candidate soft bit reference level will aid in decoding the data. In an aspect, a soft bit reference level having the highest mutual information out of several candidates is selected as the calibrated soft bit reference level.