Gain control for read operations in flash memory
A technique for performing read operations with reduced errors in a memory device such as flash memory. An automatic gain control approach is used in which cells which have experienced data retention loss are read by a fine M-level quantizer which uses M−1 read threshold voltage levels. In one approach, M-quantized threshold voltage values are multiplied by a gain to obtain gain-adjusted threshold voltage values, which are quantized by an L-level quantizer, where L<M, by comparing the gain-adjusted threshold voltage values to read threshold voltage levels of a fresh memory device. In another approach, the read threshold voltage levels of the fresh device are gain adjusted for reading non-gain-adjusted threshold voltage values from the cells which have experienced data retention loss.
1. A method of reading data from a unit of a flash memory device, the method comprising:
reading data from cells of the flash using a first quantizer with M−1 read threshold voltage levels, thereby assigning to each read cell a representation value from a predefined set of M values;
modifying the representation value of each cell in the read unit, thereby generating a modified representation value for each cell; and
assigning a read threshold voltage level for each cell, each said read threshold voltage level computed as a function of said modified representation value associated with the cell and a set of predefined L−1 read threshold voltage levels, where L is less than M.
2. The method of claim 1 , wherein said modifying step comprises modifying said representation value according to a function that depends on read threshold values from a set of cells, thereby generating a modified representation number for each cell.
3. The method of claim 2 wherein said set of cells is a set of cells containing data stored in the flash.
4. The method of claim 2 wherein said set of cells is separate from the cells containing data of the flash.
5. The method of claim 2 , wherein said modifying step comprises dividing said representation value by a function of a magnitude measure of said set of cells, thereby generating said modified representation value for each cell.
6. The method of claim 5 further comprising multiplying said modified representation value by a value R.
7. The method of claim 6 , wherein R is a function of a magnitude measure of a set of reference values associated with said set of cells.
8. The method of claim 5 , wherein said magnitude measure is a Euclidian norm
9. The method of claim 5 , wherein said magnitude measure is an L 1 norm.
10. The method of claim 5 , wherein said magnitude measure is an L ∞ norm.
11. The method of claim 1 wherein L is a power of 2.
12. The method of claim 11 wherein each said cell read from the flash device is programmed to one of L states, representing log 2 (L) bits.
13. A method of reading data from a unit of a flash memory device, the method comprising:
reading data from cells of the flash using a first quantizer with M−1 read threshold voltage levels, thereby assigning to each set of N read cells a representation vector from a predefined set of M N vectors;
modifying the representation vector of each set of N cells in the read unit, thereby generating a modified representation vector for each set of N cells;
assigning a vector of read values for each set of N cells, each said vector of read values is computed as a function of said modified representation vector associated with the set of N cells and a set of predefined L read threshold vectors, where L is less than M N .
14. The method of claim 13 wherein said L is a power of 2.
15. The method of claim 14 wherein each said set of N cells read from the flash device is programmed to one of L states, representing log 2 (L) bits.