NAND Flash Read Correction Using RBER-Based Block Characterization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The proliferation of intelligent devices has led to denser NAND flash memory, but this has resulted in reduced readback reliability, necessitating a mechanism to enhance data reliability while minimizing repetitive reads.
Innovation Solution
A storage system comprising a control processor unit, read/write channel, and non-volatile storage array that calculates a raw bit error rate (RBER) by correcting errors and applying a correction model characterization to raw data pages, thereby improving data reliability and reducing the need for repetitive reads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If NAND flash memory is made denser by scaling cell size and storing more bits per cell, then cost per gigabyte is reduced, but readback reliability deteriorates
Solution Approach 1:
The patent performs preliminary error correction by calculating RBER and generating correction model characterization data during initial reads. This preparatory action creates correction profiles that are stored and applied to subsequent reads of the same physical block, preventing error correction failures before they occur during actual data retrieval operations.
Solution Approach 2:
The system implements feedback by using the calculated RBER and correction model characterization from initial reads to improve subsequent read operations. The correction model is continuously refined based on feedback from error patterns observed in raw data, creating a closed-loop system that adapts to and compensates for the deteriorating reliability of scaled flash cells.
2Reliability
If error correction is performed on every read operation, then data reliability is improved, but the number of repetitive reads increases
Solution Approach 1:
The patent performs error correction analysis and RBER calculation in advance during initial read operations, storing the correction model characterization data for later use. This preliminary action eliminates the need to repeat full error correction processing on every subsequent read of the same physical block, significantly improving read operation efficiency while maintaining data reliability.
Solution Approach 2:
The system creates a copy of the correction model characterization data from initial reads and stores it separately in the storage medium. This copied correction information is then applied to subsequent reads without requiring repeated full error correction processing, reducing the number of repetitive read operations needed while maintaining correction effectiveness.
3Reliability
If correction model characterization is applied to all raw data pages in a physical block, then data recovery is improved, but processing complexity increases
Solution Approach 1:
The patent applies correction model characterization locally to each physical block based on its specific error patterns. Instead of using a universal correction approach, the system calculates RBER and generates correction models specific to each physical block's characteristics, improving data recovery effectiveness while managing processing complexity through localized rather than global processing.
Solution Approach 2:
The system dynamically adjusts correction parameters based on the calculated RBER and observed error patterns in each physical block. By changing correction model parameters according to the specific characteristics of each block rather than using fixed parameters, the system improves data recovery while the automated parameter adjustment manages processing complexity.
Data Source
AI summary
A storage system includes: a control processor unit, configured to: initiate a read of a raw data page, having correctable errors, calculate a raw bit error rate (RBER) (EQ1) by correcting the correctable errors to become corrected data and comparing raw data with the corrected data, and calculate a correction model characterization based on the RBER (EQ1); and a non-volatile storage array, coupled to the control processor unit, configured to store a processed data page in a physical block with the raw data page; and wherein the control processor unit is further configured to apply the correction model characterization to the raw data page in the physical block.