Memory Block Page Grouping for Word Line Failure Recovery

The patent segments data pages into distinct groups where each group is independently protected by its own redundancy mechanism. This segmentation ensures that a word line failure affecting multiple data pages does not propagate errors across all groups, as each group's redundancy can independently recover its affected pages. The segmentation principle directly resolves the contradiction by limiting the scope of error propagation while maintaining comprehensive data protection.

The patent applies preliminary action by pre-calculating and storing redundancy information (parity data or encoded redundant pages) for each data page group before actual data storage or retrieval operations. This preliminary redundancy preparation ensures that when a word line failure occurs, the recovery process can immediately utilize the pre-computed redundancy without requiring complex real-time calculations, thereby achieving complete data recovery while maintaining system reliability.

By segmenting data into groups and applying redundancy at the group level rather than across the entire storage space, the patent reduces the total redundancy overhead. Each group requires redundancy only for its own pages, not for all pages in the memory block. This segmented approach provides effective protection against word line failures while minimizing the quantity of redundant storage required, thus resolving the contradiction between reliability and storage capacity.

The patent applies local quality by tailoring the redundancy scheme to the specific characteristics of each data page group and its associated word lines. Instead of using a uniform redundancy approach across all data, the system optimizes redundancy placement and calculation for each local group, ensuring that redundancy is applied only where needed to protect against word line failures. This localized approach improves protection efficiency while reducing overall redundancy requirements.

The patent manages system complexity through segmentation by dividing the redundancy management task into independent, manageable groups. Each group can be processed, protected, and recovered independently without affecting other groups. This modular segmentation simplifies the overall system architecture by breaking down the complex task of protecting against word line failures into smaller, more manageable units, thereby achieving reliable data recovery while controlling system complexity.

The patent reduces operational complexity through preliminary action by pre-establishing the grouping structure and pre-calculating redundancy information during data initialization or idle periods. This preliminary setup eliminates the need for complex real-time calculations during data recovery operations, as the redundancy relationships are already established and stored. The system simply needs to retrieve and apply the pre-computed redundancy, significantly simplifying the recovery process while maintaining high reliability.