Flash Memory Decoder Error Floor Reduction via Adaptive Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Flash memory devices, especially multi-level cell (MLC) and triple-level cell (TLC) devices, face increased error rates due to noise sensitivity and physical degradation over program-erase cycles, leading to decoder error floors and reduced reliability.
Innovation Solution
A method involving the measurement and adaptive application of log-likelihood ratio scaling factors based on program-erase cycle counts, where initial scaling factors are used until degradation is detected, then re-measured and updated at predetermined intervals to maintain accurate decoding performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a fixed scaling factor is used for log-likelihood ratio quantization, then the decoding process is simple and fast, but decoder error floor increases due to physical degradation over program-erase cycles
Solution Approach 1:
The patent applies dynamics by transitioning from a fixed scaling factor to a dynamic scaling factor that changes based on the program-erase cycle count. The system monitors the number of program-erase cycles and adjusts the scaling factor accordingly, allowing the decoding process to adapt to physical degradation over time, thereby reducing decoder error floor while managing complexity through predefined thresholds and intervals.
Solution Approach 2:
The patent implements parameter changes by modifying the scaling factor parameter based on the program-erase cycle count. Different scaling factors are applied at different stages of the memory device's lifecycle, with re-measurement occurring at predetermined intervals. This parameter adjustment optimizes the balance between decoding accuracy and complexity management throughout the device's operational life.
2Reliability
If log-likelihood ratio is re-measured frequently to track degradation, then decoding accuracy is maintained, but throughput is reduced due to additional measurement operations
Solution Approach 1:
The patent applies periodic action by re-measuring the log-likelihood ratio at predetermined intervals based on program-erase cycle count rather than continuously or frequently. This periodic re-measurement strategy maintains decoding accuracy by updating scaling factors at appropriate intervals while minimizing the overhead of frequent measurements, thereby preserving data throughput efficiency.
Solution Approach 2:
The patent implements preliminary action by pre-determining the program-erase cycle threshold and re-measurement intervals before actual decoding operations. This allows the system to prepare scaling factor adjustment schedules in advance, avoiding the need for frequent real-time measurements and maintaining high throughput while ensuring decoding accuracy through planned, periodic updates.
3Reliability
If multiple scaling factors are applied at different endurance points, then decoder performance is optimized over time, but device complexity increases due to tracking and switching mechanisms
Solution Approach 1:
The patent applies local quality by optimizing decoding performance for specific local conditions (different program-erase cycle ranges) using appropriate scaling factors for each stage. Instead of a single universal scaling factor, the system uses locally optimized factors tailored to the current endurance state, improving overall decoder performance while managing complexity through stage-specific parameters.
Solution Approach 2:
The patent implements feedback by monitoring the program-erase cycle count and using this information to determine when to switch between different scaling factors. This feedback mechanism allows the system to adapt to physical degradation and maintain optimal decoding performance throughout the device's lifecycle, balancing improved reliability with acceptable complexity through cycle-count-based decision making.
Data Source
AI summary
An electronic non-volatile computer storage apparatus and methods for reducing decoder error floor for such a storage apparatus are disclosed. An analysis process it utilized to study one or more performance metrics of a decoder of the storage apparatus in order to determine various endurance points throughout the lifetime of that particular type of storage apparatus. Theses endurance points indicate when different scaling factors should be applied and/or when log-likelihood ratio should be re-measured to accommodate physical degradations over time.