Method and system of dynamic data storage for error correction in a memory device
View Patent ↗A method of dynamic data storage for error correction in a memory device is disclosed. Data for storage is received, and the received data is then encoded and associated error correction code (ECC) is generated. The encoded data is stored in a portion of a data partition of the memory device, wherein percentage of the stored data in the data partition is determined according to an amount of corrected errors associated with the data partition or is predetermined.
1. A method of dynamic data storage for error correction in a memory device, comprising:
receiving data for storage;
determining whether a destination page for storing the received data has been marked previously, wherein the marked page is generated by:
(a) reading and decoding a page of data in the memory device that includes a plurality of pages each having a plurality of data partitions in a data storage area;
(b) comparing a number of corrected errors with a predetermined threshold; and
(c) marking the page when the number of corrected errors exceeds the predetermined threshold;
encoding the received data and generating error correction code (ECC); and
storing the encoded data and the ECC in the data partitions of the marked page, wherein more corrected errors a page has, a smaller portion of each of the corresponding data partitions associated with said page is allocated to store the encoded data, while a size of the ECC is fixed, thereby increasing capability of correcting errors in said page.
2. The method of claim 1 , further comprising a step of replicating and storing the marked page in a reserved storage area.
3. The method of claim 2 , further comprising a step of recording a location of the replicated page in the reserved storage area.
4. The method of claim 1 , wherein the encoded data are stored to fill up the data partitions of an un-marked page.
5. The method of claim 1 , wherein the data partitions are placed in the page in succession, followed by a data header, and subsequently the associated ECCs and an ECC header are placed in succession.
6. The method of claim 1 , wherein the data partitions and an associated header and the ECC are stored in succession in the page.