Flash Memory Soft Demapping for Intercell Interference Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory devices face significant challenges in mitigating intercell interference (ICI), noise, and back pattern dependency, which distort the threshold voltage distribution and limit the number of reliable voltage levels that can be stored, especially as transistor sizes decrease below 65 nm.
Innovation Solution
The implementation of soft demapping techniques using probability density functions that account for pattern-dependent disturbances from aggressor cells, allowing for the computation of log likelihood ratios to improve data detection and mitigation of ICI, noise, and other distortions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If transistor sizes are reduced below 65 nm to increase storage capacity, then storage density is improved, but intercell interference and noise increase causing data reliability deterioration
Solution Approach 1:
The patent introduces soft demapping as an intermediary processing step between reading raw voltage values and making hard data decisions. This soft demapping process computes reliability metrics that account for ICI and noise, effectively mediating the conflict between high-density storage and data reliability by providing nuanced information about the quality of each stored value.
Solution Approach 2:
The patent changes the parameter representation from simple hard decisions (0 or 1) to soft decisions with associated reliability metrics (log-likelihood ratios). This parameter transformation allows the system to capture the degree of confidence in each stored value, enabling better error correction and reliability assessment even when physical transistor dimensions are reduced.
2Quantity of substance
If the number of voltage levels per cell is increased to store more bits per cell, then storage capacity is improved, but the impact of ICI and noise on each level increases causing measurement precision deterioration
Solution Approach 1:
The patent implements feedback through iterative soft demapping and decoding processes. The system repeatedly refines its estimates of stored values by using decoded information to inform subsequent demapping iterations, progressively improving measurement precision even in the presence of ICI and noise that worsen with increased voltage levels per cell.
Solution Approach 2:
The patent replaces direct mechanical/voltage-based threshold detection with a computational soft demapping approach. Instead of relying solely on physical voltage threshold comparisons that become increasingly difficult as more levels are packed together, the system uses probabilistic computations and log-likelihood ratio calculations to infer stored values, substituting physical measurement precision with computational analysis.
3Measurement precision
If pattern-dependent disturbances from aggressor cells are considered in demapping, then data retrieval accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the complex demapping problem into manageable components by processing cells in groups and considering only relevant aggressor cells (those with non-zero coupling coefficients). This segmentation approach allows the system to account for pattern-dependent disturbances without requiring computation across the entire memory array, balancing accuracy with computational feasibility.
Data Source
AI summary
Methods and apparatus are provided for soft demapping and intercell interference mitigation in flash memories. In one variation, a target cell in a flash memory device capable of storing at least two data levels, s, per cell is read by obtaining a measured read value, r, for at least one target cell in the flash memory; obtaining a value, h, representing data stored for at least one aggressor cell in the flash memory; selecting one or more probability density functions based on a pattern of values stored in at least a portion of the flash memory, wherein the probability density functions comprises pattern-dependent disturbance of one or more aggressor cells on the at least one target cell in the flash memory; evaluating at least one selected probability density function based on the measured read value, r; and computing one or more log likelihood ratios based on a result of the evaluating step.