Abstract:
Disclosed are a method for estimating a deterioration state of a memory device, which estimates threshold voltage distributions using data read from the memory device and determines a deterioration state of the memory device, and a wear leveling method of the memory system using the same. The method for estimating a deterioration state of a memory device comprises the steps of: reading data from selected word lines of a memory cell array by applying, to the selected word line, a plurality of read voltages having values corresponding to at least one valley of threshold voltage distributions of the selected memory cells; generating quality estimation information indicating states of the threshold voltage distributions using the data read from the selected memory cells; and determining a deterioration state of a storage area including the selected word lines based on the generated quality estimation information.
Abstract:
PURPOSE: A decoding method and a memory system device for using the method are provided to improve decoding throughput by increasing the successful rate of a decoding operation while reducing an error-floor region. CONSTITUTION: Plural parameter nodes and plural check nodes are updated by using a provability value of the received data(S120). If the updating is not successful, at least one variable node is selected from the variable nodes. The provability value of the data received to the selected variable node is corrected. The variable nodes and the check nodes are updated by using probability values.
Abstract:
메모리 시스템에서의 LDPC 디코딩 방법 및 이를 이용한 LDPC 디코더에 관하여 개시한다. 메모리 시스템에서의 LDPC 디코딩 방법은 저밀도 패리티 체크(LDPC) 디코딩을 위한 체크 노드들과 변수 노드들 간의 메시지 교환 순서를 나타내는 스케줄링 정보에 기초하여 상기 체크 노드들과 상기 변수 노드들 간의 메시지들을 교환하는 단계 및 상기 교환된 메시지들에 기초하여 LDPC 디코딩 처리를 수행하는 단계를 포함하고, 상기 스케줄링 정보는 LDPC 이분 그래프로부터 체크 노드들의 순서 조작 또는 변수 노드들의 순서 조작을 통하여 결정되는 것을 특징으로 한다.
Abstract:
PURPOSE: An iterative decoding method and an iterative decoding apparatus are provided to improve a bit error rate by outputting estimated values of all corresponding variable nodes as a final decoding result when a minimum number of check nodes dose not satisfy a parity check formula. CONSTITUTION: An iterative decoding method and an iterative decoding apparatus include a data buffer and a decoding block. The first estimated values corresponding to all variable nodes are stored into a buffer when an error bit number is minimum by using an iterative decoding algorithm while increasing the number of iterative decoding from one to a maximum iteration number gradually(S10-S30). The first estimated values stored in the buffer are outputted as a final decoding result when the iteration number is reached to the maximum iteration number(S40-S50).