Memory system and control method

Abstract

According to one embodiment, a memory system includes a non-volatile memory, a memory interface that reads data recorded in the non-volatile memory as a received value, a converting unit that converts the received value read from the non-volatile memory to first likelihood information by using a first conversion table, a decoder that decodes the first likelihood information, a control unit that outputs an estimated value with respect to the received value, which is a decoding result obtained by the decoding, when decoding by the decoder has succeeded, and a generating unit that generates a second conversion table based on a decoding result obtained by the decoding, when decoding of the first likelihood information by the decoder has failed. When the generating unit generates the second conversion table, the converting unit converts the received value to the second likelihood information by using the second conversion table, and the decoder decodes the second likelihood information.

What is claimed is: 
     
         1 . A memory system comprising:
 a non-volatile memory;   a memory interface that reads data recorded in the non-volatile memory as a received value;   a converting unit that converts the received value read from the non-volatile memory to first likelihood information by using a first conversion table;   a decoder that decodes the first likelihood information;   a control unit that outputs an estimated value with respect to the received value, which is a decoding result obtained by the decoding, when decoding by the decoder has succeeded; and   a generating unit that generates a second conversion table based on a decoding result obtained by the decoding, when decoding of the first likelihood information by the decoder has failed, wherein   when the generating unit generates the second conversion table, the converting unit converts the received value to second likelihood information by using the second conversion table, and   the decoder decodes the second likelihood information.   
     
     
         2 . The memory system according to  claim 1 , wherein
 every time the decoder fails in decoding of the second likelihood information, the generating unit generates a new second conversion table based on a result of the failed decoding,   every time the generating unit generates the second conversion table, the converting unit converts the received value to the second likelihood information by using the second conversion table, and   every time the received value is converted to the second likelihood information, the decoder decodes the second likelihood information.   
     
     
         3 . The memory system according to  claim 1 , further comprising an aggregating unit that aggregates number of occurrences of a pair of the received value and the estimated value, when the decoder has failed in decoding of the first likelihood information, wherein
 the generating unit generates the second conversion table based on an aggregation result obtained by the aggregating unit.   
     
     
         4 . The memory system according to  claim 1 , further comprising an aggregating unit that aggregates number of occurrences of a pair of the received value and the estimated value, when the decoder has failed in decoding of the first likelihood information, wherein
 the decoder outputs a posteriori probability information as a result of the decoding,   the control unit generates the estimated value from a posteriori probability information indicating a value larger than a predetermined threshold, of the pieces of a posteriori probability information,   the aggregating unit aggregates number of occurrences of a pair of the received value and the estimated value with regard to the estimated value with the a posteriori probability information indicating a value larger than the predetermined threshold, and   the generating unit generates the second conversion table based on an aggregation result obtained by the aggregating unit.   
     
     
         5 . The memory system according to  claim 2 , further comprising an aggregating unit that aggregates number of occurrences of a pair of the received value and the estimated value, when the decoder has failed in decoding of the first likelihood information or the second likelihood information, wherein
 the generating unit generates the second conversion table based on an aggregation result obtained by the aggregating unit.   
     
     
         6 . The memory system accordingto  claim 2 , further comprising an aggregating unit that aggregates number of occurrences of a pair of the received value and the estimated value, when the decoder has failed in decodir, of the first likelihood information or the second likelihood information, wherein
 the decoder outputs a posteriori probability information as a result of the decoding,   the control unit generates the estimated value posteriori probability information indicating a value larger than a predetermined threshold, of the pieces f a posteriori probability information,   the aggregating unit aggregates number of occurrences of a pair of the received value and the estimated value with regard to the estimated value with the a posteriori probability information indicating a value larger than the predetermined threshold, and   the generating unit generates the second conversion table based on an aggregation result obtained by the aggregating unit.   
     
     
         7 . The memory system according to  claim 1 , further comprising a modification unit that modifies the second conversion table generated by the generating unit, wherein
 the converting unit converts the received value to the second likelihood information by using the second conversion table after modification by the modification unit.   
     
     
         8 . The memory system according to  claim 1 , wherein
 the converting unit converts a received value sequence configured by a predetermined number of the received values to a likelihood information sequence configured by the predetermined number of pieces of the first likelihood information or the second likelihood information, in unit of the received value sequence, and   the decoder performs the decoding in unit of the likelihood information sequence.   
     
     
         9 . The memory system according to  claim 8 , further comprising an aggregating unit that aggregates number of occurrences of a pair of the received value and the estimated value, when the decoder has failed in decoding of the likelihood information sequence, wherein
 the aggregating unit aggregates number of occurrences of a pair of the received value and the estimated value obtained by decoding performed by the decoder with respect to the plurality of likelihood information sequences, and   the generating unit generates the second conversion table based on an aggregation result obtained by the aggregating unit.   
     
     
         10 . The memory system according to  claim 1 , wherein after the generating unit has performed generation of the second conversion table for a predetermined number of times, when the decoder has failed in decoding of the second likelihood information, the control unit outputs a decoding failure. 
     
     
         11 . The memory system according to  claim 1 , wherein the first likelihood information and the second likelihood information are a log-likelihood ratio. 
     
     
         12 . The memory system according to  claim 1 , wherein the data recorded in the non-volatile memory is data encoded by an error correction code that can be used in soft-decision decoding. 
     
     
         13 . A control method of a non-volatile memory, the control method comprising:
 reading data recorded in the non-volatile memory as received value;   converting the received value read from the non-volatile memory to first likelihood information by using first conversion table;   decoding the first likelihood information;   when decoding of the first likelihood information has succeeded, outputting an estimated value with respect to the received value, which is a decoding result obtained by the decoding;   when decoding of the first likelihood information has failed, generating a second conversion table based on a decoding result obtained by the decoding;   converting the received value to second likelihood information by using the generated second conversion table; and   decoding the second likelihood information.   
     
     
         14 . The control method according to  claim 13 , wherein
 every time the decoding of the second likelihood information is failed, a new second conversion table is generated based on a result of the failed decoding,   every time the second conversion table is generated, the received value is converted to the second likelihood information by using the second conversion table, and   every time the received value is converted to the second likelihood information, the second likelihood information is decoded.   
     
     
         15 . The control method according to  claim 13 , further comprising aggregating number of occurrences of a pair of the received value and the estimated value, when the decoding of the first likelihood information has been failed, wherein
 the second conversion table is generated based on an aggregation result obtained by the aggregating.   
     
     
         16 . The control method according to  claim 13 , further comprising aggregating number of occurrences of a pair of the received value and the estimated value, when the decoding of the first likelihood information has been failed, wherein
 a posteriori probability information is outputted as a result of the decoding,   the estimated value is generated from a posteriori probability information indicating a value larger than a predetermined threshold, of the pieces of a posteriori probability information,   the number of occurrences of a pair of the received value and the estimated value is aggregated with regard to the estimated value with the a posteriori probability information indicating a value larger than the predetermined threshold, and   the second conversion table is generated based on an aggregation result obtained by the aggregating unit.   
     
     
         17 . The control method according to  claim 14 , further comprising aggregating number of occurrences of a pair of the received value and the estimated value, when the decoding of the first likelihood information or the second likelihood information has been falied, wherein
 the second conversion table is generated based on an aggregation result obtained by the aggregating unit.   
     
     
         18 . The control method according to  claim 14 , further comprising aggregating number of occurrences of a pair of the received value and the estimated value, when the decoding of the first likelihood information or the second likelihood information has been failed, wherein
 a posteriori probability information is outputted as a result of the decoding,   the estimated value is generated from a posteriori probability information indicating a value larger than a predetermined threshold, of the pieces of a posteriori probability information,   the number of occurrences of a pair of the received value and the estimated value is aggregated with regard to the estimated value with the a posteriori probability information indicating a value larger than the predetermined threshold, and   the second conversion table is generated based on an aggregation result obtained by the aggregating unit.   
     
     
         19 . The control method according to  claim 13 , further comprising modifying the second conversion table generated by the generating, wherein
 the received value is converted to the second likelihood information by using the second conversion table after modification by the modification unit.   
     
     
         20 . The control method according to  claim 13 , wherein
 a received value sequence configured by a predetermined number of the received values is converted to a likelihood information sequence configured by the predetermined number of pieces of the first likelihood information or the second likelihood information, in unit of the received value sequence, and   the decoding is performed in unit of the likelihood information sequence.