Low-complexity two-dimensional data encoding for memory inter-cell interference reduction

Inter-cell coupling interference (ICI) affects capacity, performance, reliability and endurance of semiconductor memories, particularly NAND Flash. Constrained coding of a row of cells has been proposed for ICI mitigation by forbidding data patterns that induce interference above a chosen threshold level. We expand the interference model to two dimensions, and derive the severity function for row-by-row data programming. We propose a two-dimensional interference-constrained coding algorithm, based on adaptive one-dimensional encoding. Adaptation occurs at each encoding step, based on the data in the current row and in the previous one. Similarly, decoding is done in reverse. We explore the capacity bounds with the proposed constraints.