Soft decoding parameter determination method, electronic equipment and storage medium

Abstract

The invention discloses a soft decoding parameter determination method, electronic equipment and a storage medium. The soft decoding parameter determination method comprises the following steps: acquiring a plurality of test deviation values of a memory and a plurality of interval ranges corresponding to the test deviation values; performing soft reading on the memory for multiple times by utilizing each test deviation value, so as to determine mutual information between a unit of the memory and a plurality of interval ranges corresponding to each test deviation value; determining the test deviation value corresponding to the maximum mutual information as a target deviation value of soft decoding; and performing soft decoding on the memory by using the target deviation value to obtain a target log-likelihood ratio of soft decoding. According to the scheme, the parameter configuration of soft decoding can be optimized, the soft decoding capability is improved, the soft decoding success rate and efficiency are improved, and the reliability of data is ensured.

Description

Technical Field

[0001] The present application relates to the field of storage technology, and in particular to a method for determining soft decoding parameters, an electronic device, and a storage medium. Background Art

[0002] With the development of technology, decoders are more and more widely used. For example, low-density parity check code (LDPC) decoders are divided into hard decoding and soft decoding during decoding.

[0003] In a flash memory system, the voltage offset value and log-likelihood ratio of the divided intervals need to be configured during the soft decoding process of the LDPC code.

[0004] Currently, different main control manufacturers have different configuration methods. Summary of the invention

[0005] The main technical problem solved by the present application is to provide a method for determining soft decoding parameters, an electronic device and a storage medium to optimize the configuration of the soft decoding parameters.

[0006] In order to solve the above problems, the present application provides a method for determining soft decoding parameters, the soft decoding parameters of the memory include a target offset value and a target log-likelihood ratio, and the method for determining the soft decoding parameters includes: obtaining multiple test offset values ​​of the memory and multiple interval ranges corresponding to each test offset value; using each test offset value to soft-read the memory multiple times respectively to determine the mutual information between the unit of the memory and the multiple interval ranges corresponding to each test offset value; determining the test offset value corresponding to the maximum mutual information as the target offset value of soft decoding; using the target offset value to soft-decode the memory to obtain the target log-likelihood ratio of soft decoding.

[0007] Among them, the steps of performing soft readings on the memory multiple times using each test offset value to determine the mutual information between the unit of the memory and the multiple interval ranges corresponding to each test offset value include: performing soft readings on the memory multiple times using each test offset value to obtain the probability that the unit of the memory falls into the multiple interval ranges corresponding to each test offset value, and obtaining a transition probability matrix; based on the transition probability matrix, respectively calculating the mutual information between the unit of the memory and the multiple interval ranges corresponding to each test offset value.

[0008] The memory is soft-read multiple times using each test offset value to obtain the probability that the memory unit falls into the multiple intervals corresponding to each test offset value, and the step of obtaining the transfer probability matrix includes: determining multiple read voltage values ​​corresponding to the test offset value based on the test offset value and the reference voltage of the memory; performing soft reading on the memory based on the multiple read voltage values ​​to obtain soft reading results; determining the probability that each unit of the memory falls into the multiple intervals corresponding to each test offset value based on the soft reading results; and obtaining the transfer probability matrix by synthesizing the probability that each unit falls into the multiple intervals corresponding to each test offset value.

[0009] The step of determining the probability that each unit of the memory falls into a plurality of intervals corresponding to each test offset value based on the soft read result includes: determining the probability of the corresponding interval based on the number of memory cells in each interval in the soft read result; wherein the probability is proportional to the number of corresponding memory cells.

[0010] The steps of respectively calculating the mutual information between the memory unit and the multiple intervals corresponding to each test offset value based on the transition probability matrix include: using a preset formula based on the probability in the transition probability matrix, the number of columns of the probability in the transition probability matrix, the number of rows of the probability in the transition probability matrix, the order of the memory and the information entropy function, to calculate the mutual information corresponding to the probability; the preset formula includes Among them, I is the mutual information, H is the information entropy function, X is the unit, Y is the interval range, P nm is the probability, n is P nm The number of rows in the transition probability matrix, m is P nm In the number of columns of the transfer probability matrix, i is the total number of rows of the transfer probability matrix, and j is the total number of columns of the transfer probability matrix.

[0011] Among them, the step of soft-decoding the memory using the target offset value to obtain a target log-likelihood ratio of the soft decoding includes: determining multiple target read voltage values ​​based on the target offset value, and dividing the multiple target voltage values ​​into corresponding multiple target intervals; soft-decoding the memory based on the multiple target voltage values ​​to obtain a log-likelihood ratio of each target interval; calculating the bit error rate of the log-likelihood ratio of each target interval, and determining the log-likelihood ratio with the smallest bit error rate as the target log-likelihood ratio of the soft decoding.

[0012] Among them, the steps of calculating the bit error rate of the log-likelihood ratio of each target interval and determining the log-likelihood ratio with the smallest bit error rate as the target log-likelihood ratio of soft decoding include: scaling and quantizing the log-likelihood ratio of each target interval using a scaling factor; calculating the original bit error rate of the log-likelihood ratio of each target interval after scaling and quantization, and determining the scaled and quantized log-likelihood ratio with the smallest original bit error rate as the target log-likelihood ratio of soft decoding.

[0013] Among them, the steps of calculating the original bit error rate of the log-likelihood ratio of each target interval after scaling and quantization, and determining the scaled and quantized log-likelihood ratio with the smallest original bit error rate as the target log-likelihood ratio for soft decoding also include: in response to the existence of a log-likelihood ratio after scaling and quantization greater than a preset threshold, determining the preset threshold as the target log-likelihood ratio for soft decoding.

[0014] To solve the above technical problems, the present invention further provides an electronic device, comprising a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory to implement any of the above soft decoding parameter determination methods.

[0015] In order to solve the above technical problems, the present invention further provides a computer-readable storage medium having program instructions stored thereon, and when the program instructions are executed by a processor, the method for determining soft decoding parameters as described in any one of the above items is implemented.

[0016] The beneficial effects of the present invention are as follows: Different from the prior art, the present application first obtains multiple test offset values ​​of the memory and multiple interval ranges corresponding to each test offset value; uses each test offset value to perform soft reading on the memory multiple times to determine the mutual information between the unit of the memory and the multiple interval ranges corresponding to each test offset value; determines the test offset value corresponding to the maximum mutual information as the target offset value of soft decoding, and finally uses the target offset value to perform soft decoding on the memory to obtain the target log-likelihood ratio of soft decoding, thereby using the mutual information to determine the test offset value with the highest soft decoding success rate among the multiple test offset values ​​as the target offset value, and then obtains the target log-likelihood ratio of soft decoding based on the target offset value, thereby optimizing the parameter configuration of LDPC soft decoding, improving the LDPC soft decoding capability, improving the soft decoding success rate and efficiency, and ensuring the reliability of data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of an embodiment of a method for determining soft decoding parameters of the present application;

[0018] Figure 2 It is a flowchart of another embodiment of the method for determining soft decoding parameters of the present application;

[0019] Figure 3 It is a schematic diagram of an implementation method of dividing the SLC Nand Flash threshold voltage interval range;

[0020] Figure 4 yes Figure 3 A schematic diagram of a channel model for an implementation of unit probability in an embodiment;

[0021] Figure 5 is a schematic diagram of an implementation method of a soft read result;

[0022] Figure 6 It is a schematic diagram of the framework of an embodiment of the electronic device of the present application;

[0023] Figure 7 A schematic diagram of a framework of an embodiment of a computer-readable storage medium of the present application. DETAILED DESCRIPTION

[0024] The scheme of the embodiment of the present application is described in detail below in conjunction with the drawings of the specification.

[0025] In the following description, for the purpose of explanation rather than limitation, specific details such as specific system structures, interfaces, and technologies are provided to facilitate a thorough understanding of the present application.

[0026] The terms "system" and "network" are often used interchangeably in this article. The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship. In addition, "many" in this article means two or more than two.

[0027] See also Figure 1 , Figure 1 1 is a flow chart of an embodiment of a method for determining soft decoding parameters of the present application. The method for determining soft decoding parameters in this embodiment specifically includes the following steps:

[0028] Step S11: obtaining a plurality of test offset values ​​from the memory and a plurality of interval ranges corresponding to each test offset value.

[0029] The soft decoding parameters of the memory of this embodiment include a target offset value and a target log-likelihood ratio. Before soft decoding, it is necessary to determine the target offset value and the target log-likelihood ratio of the memory, that is, this embodiment specifically determines the target offset value and the target log-likelihood ratio of the memory. Among them, the memory may include various types of Nand Flash non-volatile memories. For example, SLC (Single Level Cell), MLC (Multi-Level Cell), TLC (Triple-Level Cell), QLC (Quad-Level Cell), and Nand Flash with larger single cell storage capacity in the future, etc., are not limited here.

[0030] When performing soft decoding, the base voltage (base read) is used as the center, and multiple read voltage values ​​are determined before and after it, so as to perform multiple soft reads on the memory based on the multiple read voltage values. The base voltage is generally provided by the Flash manufacturer or obtained based on the test experience of the R&D personnel, and is not limited here.

[0031] The target offset value refers to the offset value between the reference voltage and the voltage values ​​before and after it. In a specific application scenario, when the memory is soft-decoded N4, its read voltage values ​​are base (reference voltage), base-delta, and base+delta, respectively, and delta is the target offset value to be determined in this embodiment.

[0032] The target log-likelihood ratio (Log Likelihood Ratios) is an indicator that reflects authenticity and is a composite indicator that reflects both sensitivity and specificity. To use LDPC codes more efficiently, LDPC decoders use "soft" data, such as the analog probability that each bit of the data being decoded has a given value, which is the log-likelihood ratio (LLR).

[0033] In this step, a plurality of test offset values ​​of the memory and a plurality of interval ranges corresponding to each test offset value are first obtained.

[0034] Among them, multiple test offset values ​​can be determined based on the type of memory and manual experience. In a specific application scenario, a range of test offset values ​​can be determined based on the type of memory and manual experience, and multiple test offset values ​​can be selected within the range of the test offset value. The selection method can be random selection, equidistant selection based on interval values, or special value selection, etc., which is not limited here. The interval value can be determined based on the range size and manual experience.

[0035] After determining multiple test offset values, multiple interval ranges corresponding to each test offset value are determined. Specifically, a test offset value and a reference voltage are used to determine multiple corresponding read voltage values, and then multiple corresponding interval ranges are obtained based on the multiple read voltage values. In a specific application scenario, when the multiple read voltage values ​​of the soft decoding are base, base-delta, base+delta, base-2*delta, and base+2*delta, respectively. Then, six regions can be obtained, namely, region 0, region 1, region 2, region 3, region 4, and region 5. Among them, the voltage range of region 0 is less than base-2*delta; the voltage range of region 1 is between base-2*delta and base-delta; the voltage range of region 2 is between base-delta and base; the voltage range of region 3 is between base and base+delta; the voltage range of region 4 is between base+delta and base+2*delta; and the voltage range of region 5 is greater than base+2*delta. In other application scenarios, the regional division method of multiple read voltage values ​​is similar to this and will not be repeated.

[0036] Step S12: performing multiple soft reads on the memory using each test offset value to determine mutual information between the memory unit and multiple intervals corresponding to each test offset value.

[0037] Mutual Information is a useful information measure in information theory. It can be viewed as the amount of information contained in a random variable about another random variable, or the uncertainty of a random variable reduced by knowing another random variable.

[0038] In a specific application scenario, the memory can be soft-read based on multiple read voltage values ​​corresponding to each test offset value, and the mutual information between each interval range corresponding to the test offset value can be determined based on the soft read result fed back by each interval range in the soft read result.

[0039] In a specific application scenario, the probability that the memory unit (state) falls into the multiple intervals corresponding to each test offset value can also be determined to obtain a transition probability matrix; based on the transition probability matrix, the mutual information between the memory unit and the multiple intervals corresponding to each test offset value is calculated. Specifically, the method for determining the mutual information is not limited here.

[0040] In a specific application scenario, assuming that there are 10 test offset values, each of which can determine 4 intervals, then in this step, the mutual information between the memory unit and each interval corresponding to each test offset value is determined to be 10*4=40 in total.

[0041] Step S13: Determine the test offset value corresponding to the maximum mutual information as the target offset value for soft decoding.

[0042] After obtaining the mutual information between the memory unit and the multiple intervals corresponding to each test offset value, the maximum mutual information is selected, and the test offset value corresponding to the maximum mutual information is determined as the target offset value of soft decoding. That is, the test offset value corresponding to the multiple read voltage values ​​corresponding to the maximum mutual information is determined as the target offset value of soft decoding.

[0043] In a specific application scenario, assume that there are 10 test offset values, namely, C1, C2, C3, ..., C10, and the mutual information between the 10 test offset values ​​and the corresponding interval ranges is 40. Among them, the test offset value corresponding to the largest mutual information is C2, and C2 is used as the target offset value for soft decoding.

[0044] When the mutual information is the largest, it means that at this time, the read voltage values ​​divided into multiple intervals are optimal and the success rate of soft decoding is the highest. The test offset value corresponding to the maximum mutual information is determined as the target offset value of soft decoding, which can optimize the parameter configuration of LDPC soft decoding and improve the LDPC soft decoding capability.

[0045] Step S14: soft-decoding the memory using the target offset value to obtain a soft-decoded target log-likelihood ratio.

[0046] After obtaining the target offset value of the memory, the target offset value is used to determine the corresponding multiple read voltage values, and then the memory is soft-decoded based on the multiple read voltage values, and the log-likelihood ratio of each interval range is calculated, and the optimal value is selected and determined as the target log-likelihood ratio.

[0047] In a specific application scenario, the corresponding LLR value can be calculated for each divided interval range, and then the value of the LLR sequence can be multiplied by a certain coefficient until the LDPC soft decoding capability is no longer improved. At this time, the LLR sequence is used as the target log-likelihood ratio.

[0048] In a specific application scenario, the memory can be soft-decoded to obtain the log-likelihood ratio of each target interval; the bit error rate of the log-likelihood ratio of each target interval is calculated, and the log-likelihood ratio with the smallest bit error rate is determined as the target log-likelihood ratio of soft decoding. The specific method for determining the optimal target log-likelihood ratio is not limited here.

[0049] The soft decoding parameters of the memory are obtained based on the target offset value and the target log-likelihood ratio, thereby completing the determination of the soft decoding parameters, and then the soft decoding parameters of the memory can be configured based on the soft decoding parameters of the memory.

[0050] Through the above steps, the method for determining the soft decoding parameters of the present embodiment first obtains multiple test offset values ​​of the memory and multiple interval ranges corresponding to each test offset value; uses each test offset value to perform soft reading on the memory multiple times to determine the mutual information between the unit of the memory and the multiple interval ranges corresponding to each test offset value; determines the test offset value corresponding to the maximum mutual information as the target offset value of soft decoding, and finally uses the target offset value to perform soft decoding on the memory to obtain the target log-likelihood ratio of soft decoding, so as to first use the mutual information to determine the test offset value with the highest soft decoding success rate among the multiple test offset values ​​as the target offset value, and then obtain the target log-likelihood ratio of soft decoding based on the target offset value, thereby optimizing the parameter configuration of LDPC soft decoding, improving the LDPC soft decoding capability, improving the soft decoding success rate and efficiency, reducing the occurrence of soft decoding failures, and ensuring data reliability.

[0051] See also Figure 2 , Figure 2 FIG. 1 is a flow chart of another embodiment of the method for determining soft decoding parameters of the present application. The method for determining soft decoding parameters in this embodiment specifically includes the following steps:

[0052] Step S21: obtaining a plurality of test offset values ​​from the memory and a plurality of interval ranges corresponding to each test offset value.

[0053] The multiple test offset values ​​can be determined based on the type of memory and manual experience. After the multiple test offset values ​​are determined, the multiple interval ranges corresponding to the multiple test offset values ​​are determined based on the test offset values. Specifically, a test offset value and a reference voltage are used to determine the corresponding multiple read voltage values, and then the corresponding multiple interval ranges are obtained based on the multiple read voltage values.

[0054] This step is the same as step S11 in the aforementioned embodiment, please refer to the above text.

[0055] See also Figure 3 , Figure 3 It is a schematic diagram of an implementation method of dividing the SLC Nand Flash threshold voltage interval range.

[0056] This diagram shows the division of the N4 interval range of the SLC Nand Flash threshold voltage distribution. The threshold voltage distribution is divided into four intervals by three read voltage values. The three read voltage values ​​of the divided intervals and the LLR values ​​corresponding to each interval are the configuration parameters for LDPC to perform N4 soft decoding. For SLC type Nand Flash, there are two units (states), which can be considered as two units A and B. After the three read voltage values ​​are divided, they are divided into region1, region2, region3, and region4. Among them, the voltage range of region1 is less than base-delta, the voltage range of region2 is between base-delta and base; the voltage range of region3 is between base and base+delta; the voltage range of region4 is greater than base+delta.

[0057] In other implementations, when the memory is of other types or the intervals of N6 and N8 are divided, the division is similar to the above implementation and will not be described in detail.

[0058] Step S22: Soft-read the memory multiple times using each test offset value to obtain the probability that the cells of the memory fall into multiple intervals corresponding to each test offset value, and obtain a transfer probability matrix.

[0059] First, the probability that the memory cell falls within the multiple intervals corresponding to the test offset values ​​is determined.

[0060] See also Figure 4 , Figure 4 yes Figure 3 A schematic diagram of a channel model for an implementation of unit probability in an embodiment.

[0061] In this channel model, the probability that unit A falls into region 1, region 2, region 3, and region 4 is: p 11 、p 12 、p 13 、p 14 , the probability of unit B falling into region1, region2, region3, and region4 are: p 21 、p 22 、p 23 、p 24 .

[0062] The step of determining whether each unit falls into a corresponding different interval range includes: determining multiple read voltage values ​​corresponding to the test offset value based on the test offset value and the reference voltage of the memory; and soft-reading the memory based on the multiple read voltage values ​​to obtain a soft read result.

[0063] In this step, the memory is soft-read based on each test offset value to obtain the probability of each interval range corresponding to each test offset value.

[0064] In a specific application scenario, assume that there are C1, C2, C3, ..., C10, a total of 10 test offset values, and each test offset value corresponds to 4 intervals. Then, based on the read voltage values ​​corresponding to the 10 test offset values, the memory is soft-read 10 times, and each soft read obtains soft read data in 4 intervals, that is, the obtained soft read result contains 40 soft read data in total.

[0065] The probability of each unit falling into a plurality of intervals corresponding to each test offset value is determined based on the soft read result. Specifically, the probability of the corresponding interval is determined based on the number of storage units in each interval in the soft read result; wherein the probability is proportional to the number of corresponding storage units.

[0066] See also Figure 5 , Figure 5 Schematic diagram of an implementation of a soft read result.

[0067] pA1 is the ratio of the number of storage units of stateA that fall into region1 to the total number of storage units in stateA, pA2 is the ratio of the number of storage units of stateA that fall into region2 to the total number of storage units in stateA, and the subsequent probabilities are similar.

[0068] Among them, the number of storage units in each area can be mapped on the graph to the area size of the closed shape formed between the corresponding area, the wave and the horizontal axis. The number of storage units with a large area is large, and the number of storage units with a small area is small.

[0069] Based on the probability of each unit falling into the multiple intervals corresponding to each test offset value in the soft read result, the probability of each unit falling into the multiple intervals corresponding to each test offset value is combined to obtain a transfer probability matrix.

[0070] In a specific application scenario, for a multi-level cell Nand Flash, if the probability of each cell falling into each interval range after being divided into intervals is known, it can be converted into a transfer probability matrix. The probability transfer matrix is ​​shown in the following table:

[0071]

[0072] Among them, x1...xi are the units of the memory, and y1...yi are the interval range corresponding to the current test offset value. When there are multiple test offset values, there are also multiple corresponding probability transfer matrices. The mutual information is calculated and selected from the multiple probability transfer matrices.

[0073] The generation of the transfer probability matrix, for SLC NAND, there are two units. For the area range division of N4 soft decoding, a 2*4 transfer probability matrix is ​​generated. For MLC NAND, there are four units. For the softdecode division of N4, it will be divided into 10 region ranges, and the generated transfer probability matrix is ​​a 4*10 transfer probability matrix, and so on.

[0074] Step S23: Based on the transition probability matrix, respectively calculate the mutual information between the memory unit and a plurality of interval ranges corresponding to each test offset value.

[0075] The mutual information corresponding to the probability is calculated using a preset formula based on the probability in the transfer probability matrix, the number of columns of the probability in the transfer probability matrix, the number of rows of the probability in the transfer probability matrix, the order of the memory, and the information entropy function, wherein the preset formula includes

[0076]

[0077] The N in the above mutual information calculation method represents the order of NAND memory, for example, N of SLC NAND is 1, N of MLC NAND is 2, N of TLC NAND is 3, and so on. I is mutual information, H is information entropy function, X is unit, Y is interval range, P nm is the probability, n is P nm The number of rows in the transition probability matrix, m is P nm In the number of columns of the transfer probability matrix, i is the total number of rows of the transfer probability matrix, and j is the total number of columns of the transfer probability matrix.

[0078] In a specific application scenario, Figure 4 Taking the channel model of the implementation method as an example, the mutual information is calculated as follows:

[0079]

[0080] In other application scenarios, the calculation of mutual information is similar to this and will not be repeated here.

[0081] Step S24: Determine the test offset value corresponding to the maximum mutual information as the target offset value for soft decoding.

[0082] After obtaining multiple transition probability matrices of mutual information between the memory unit and multiple intervals corresponding to each test offset value in the above manner, the largest mutual information is selected, and the test offset value corresponding to the largest mutual information is determined as the target offset value for soft decoding. That is, the test offset value corresponding to the multiple read voltage values ​​based on which the largest mutual information is obtained is determined as the target offset value for soft decoding.

[0083] When the mutual information is the largest, it means that at this time, the read voltage values ​​divided into multiple intervals are optimal and the success rate of soft decoding is the highest. The test offset value corresponding to the maximum mutual information is determined as the target offset value of soft decoding, which can optimize the parameter configuration of LDPC soft decoding and improve the LDPC soft decoding capability.

[0084] Step S25: Determine the corresponding target intervals based on the target offset value; perform soft decoding on the memory to obtain the log-likelihood ratio of each target interval; calculate the bit error rate of the log-likelihood ratio of each target interval, and determine the log-likelihood ratio with the minimum bit error rate as the target log-likelihood ratio of soft decoding.

[0085] Based on the target offset value, multiple target read voltage values ​​are determined, and multiple corresponding target intervals are obtained by dividing the multiple target voltage values; based on the multiple target voltage values, the memory is soft-decoded to obtain the log likelihood ratio of each target interval. The log likelihood ratio can be calculated by using a log-map algorithm or a max-log-map algorithm, for example, LLR1 = ln[(P(0)) / (P(1))], etc.

[0086] Among them, the method of determining multiple target read voltage values ​​based on the target offset value and dividing them into corresponding multiple target intervals is the same as the method of determining multiple read voltage values ​​based on the test offset value and then dividing the interval range in step S11 of the aforementioned embodiment. Please refer to the above text and will not be repeated here.

[0087] After obtaining the log-likelihood ratio of each target interval, in order to minimize the quantization loss, the original log-likelihood ratio needs to be scaled and then quantized, and the log-likelihood ratio of each target interval is scaled and quantized using the scaling factor. In a specific application scenario, the scaling factor can be adjusted based on the number of LDPC iterations, the signal-to-noise ratio of the input signal, and other confidence-related information. In another specific application scenario, the scaling factor can also be determined based on the quantization error (including saturation error and rounding error) as a measure and the goal of minimizing the quantization error. The specific method for determining the scaling factor is not limited here.

[0088] Subsequently, the original bit error rate of the scaled quantized log-likelihood ratio of each target interval is calculated, and the scaled quantized log-likelihood ratio with the smallest original bit error rate (BERE) is determined as the target log-likelihood ratio for soft decoding. The smallest original bit error rate means that the corresponding log-likelihood ratio is optimal and the soft decoding success rate is the highest. Therefore, determining the scaled quantized log-likelihood ratio with the smallest original bit error rate as the target log-likelihood ratio for soft decoding can further optimize the parameter configuration of LDPC soft decoding and improve the LDPC soft decoding capability.

[0089] In a specific application scenario, in response to the existence of a log-likelihood ratio after scaling and quantization being greater than a preset threshold, the preset threshold can be directly determined as a target log-likelihood ratio for soft decoding. The preset threshold can be -15 to +15, and the value of the target log-likelihood ratio can be the maximum value, such as -15 or +15.

[0090] A soft decoding parameter of the memory is obtained based on the target offset value and the target log-likelihood ratio, and then the soft decoding parameter of the memory is configured based on the soft decoding parameter of the memory.

[0091] The above method uses the maximum mutual information value in the transfer probability matrix to find the optimal value for the memory voltage interval division. Based on the division interval value, the corresponding LLR value is calculated to provide a search method for the LLR sequence required for soft solution. At the same time, the value of the LLR sequence is optimized by scaling factors and bit error rates to improve the soft solution performance of LDPC and the reliability of data storage. At the same time, an application method in an actual NAND channel is provided.

[0092] Through the above steps, the method for determining the soft decoding parameters of the present embodiment first obtains multiple test offset values ​​of the memory and multiple interval ranges corresponding to each test offset value; determines the probability that the unit of the memory falls into the multiple interval ranges corresponding to each test offset value, obtains a transfer probability matrix, and calculates the mutual information between the unit of the memory and the multiple interval ranges corresponding to each test offset value based on the transfer probability matrix; determines the test offset value corresponding to the maximum mutual information as the target offset value of soft decoding, and finally determines the corresponding target intervals based on the target offset value; performs soft decoding on the memory to obtain the log likelihood ratio of each target interval; finally calculates the bit error rate of the log likelihood ratio of each target interval, and determines the log likelihood ratio with the minimum bit error rate as the target log likelihood ratio of soft decoding, so as to use the offset when the mutual information is maximum as the standard for dividing the interval range, and then selects the optimal target log likelihood ratio based on the offset at this time based on the scaling factor and the bit error rate, so as to double optimize the parameter configuration of LDPC soft decoding, improve the LDPC soft decoding capability, improve the success rate and efficiency of soft decoding, and ensure the reliability of data.

[0093] See also Figure 6 , Figure 6 6 is a schematic diagram of a framework of an embodiment of an electronic device of the present application. The electronic device 60 includes a memory 61 and a processor 62 coupled to each other, and the processor 62 is used to execute program instructions stored in the memory 61 to implement the steps of the above method embodiment. In a specific implementation scenario, the electronic device 60 may include but is not limited to: a microcomputer, a server, and in addition, the electronic device 60 may also include a laptop, a tablet computer, Nand Flash, etc., which are not limited here.

[0094] Specifically, the processor 62 is used to control itself and the memory 61 to implement the steps of any of the above method embodiments. The processor 62 can also be called a CPU (Central Processing Unit). The processor 62 may be an integrated circuit chip with signal processing capabilities. The processor 62 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 62 can be implemented by an integrated circuit chip.

[0095] The above scheme can optimize the parameter configuration of LDPC soft decoding, improve the LDPC soft decoding capability, improve the success rate and efficiency of soft decoding, and ensure the reliability of data.

[0096] See also Figure 7 , Figure 7 The computer-readable storage medium 70 stores program instructions 701 that can be executed by a processor, and the program instructions 701 are used to implement the steps of any of the above method embodiments.

[0097] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0098] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0099] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0100] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.

[0101] The above description is only an implementation mode of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for determining soft decoding parameters, characterized in that: The soft decoding parameters of the memory include a target offset value and a target log-likelihood ratio, and the method for determining the soft decoding parameters includes: Acquire multiple test offset values ​​of the memory and multiple interval ranges corresponding to each of the test offset values; Performing multiple soft reads on the memory using each of the test offset values ​​to determine mutual information between a unit of the memory and a plurality of interval ranges corresponding to each of the test offset values; Determine the test offset value corresponding to the maximum mutual information as the target offset value of soft decoding; The memory is soft-decoded using the target offset value to obtain a soft-decoded target log-likelihood ratio.

2. The method for determining soft decoding parameters according to claim 1, characterized in that: The step of performing multiple soft reads on the memory using the test offset values ​​to determine the mutual information between the memory unit and multiple intervals corresponding to the test offset values ​​comprises: Performing multiple soft reads on the memory using each of the test offset values ​​to obtain probabilities that the cells of the memory fall within multiple intervals corresponding to each of the test offset values, and obtaining a transfer probability matrix; Based on the transition probability matrix, mutual information between the units of the memory and a plurality of interval ranges corresponding to each of the test offset values ​​is calculated respectively.

3. The method for determining soft decoding parameters according to claim 2, characterized in that: The step of performing multiple soft reads on the memory using each of the test offset values ​​to obtain the probability that the cells of the memory fall into multiple intervals corresponding to each of the test offset values, and obtaining a transfer probability matrix comprises: Based on the test offset value of the memory and a reference voltage, determining a plurality of read voltage values ​​corresponding to the test offset value; Performing a soft read on the memory based on the plurality of read voltage values ​​to obtain a soft read result; Determine, based on the soft read result, the probability that each of the cells of the memory falls within a plurality of intervals corresponding to each test offset value; The transfer probability matrix is ​​obtained by synthesizing the probabilities of each of the units falling into a plurality of intervals corresponding to each of the test offset values.

4. The method for determining soft decoding parameters according to claim 3, characterized in that: The step of determining the probability that each of the cells of the memory respectively falls within a plurality of intervals corresponding to each test offset value based on the soft read result comprises: Determine the probability of the corresponding interval range based on the number of storage units in each interval range in the soft read result; The probability is proportional to the number of corresponding storage units.

5. The method for determining soft decoding parameters according to claim 2, characterized in that: The step of respectively calculating the mutual information between the units of the memory and the multiple interval ranges corresponding to the test offset values ​​based on the transition probability matrix comprises: Calculate the mutual information corresponding to the probability using a preset formula based on the probability in the transfer probability matrix, the number of columns of the probability in the transfer probability matrix, the number of rows of the probability in the transfer probability matrix, the order of the memory, and the information entropy function; The preset formula includes Among them, I is the mutual information, H is the information entropy function, X is the unit, Y is the interval range, P nm is the probability, n is P nm The number of rows in the transition probability matrix, m is P nm In the number of columns of the transfer probability matrix, i is the total number of rows of the transfer probability matrix, and j is the total number of columns of the transfer probability matrix.

6. The method for determining soft decoding parameters according to claim 1, characterized in that: The step of soft-decoding the memory by using the target offset value to obtain a soft-decoded target log-likelihood ratio comprises: Determine a plurality of target read voltage values ​​based on the target offset value, and divide the plurality of target voltage values ​​to obtain a plurality of corresponding target intervals; performing soft decoding on the memory based on the plurality of target voltage values ​​to obtain a log likelihood ratio of each target interval; The bit error rate of the log-likelihood ratio of each target interval is calculated, and the log-likelihood ratio with the minimum bit error rate is determined as the target log-likelihood ratio of soft decoding.

7. The method for determining soft decoding parameters according to claim 6, characterized in that: The step of calculating the bit error rate of the log-likelihood ratio of each target interval and determining the log-likelihood ratio with the minimum bit error rate as the target log-likelihood ratio of soft decoding comprises: Using a scaling factor to scale the log-likelihood ratio of each target interval; The original bit error rate of the log-likelihood ratio of each target interval after scaling and quantization is calculated, and the scaled and quantized log-likelihood ratio with the smallest original bit error rate is determined as the target log-likelihood ratio of soft decoding.

8. The method for determining soft decoding parameters according to claim 7, characterized in that: The step of calculating the original bit error rate of the log-likelihood ratio of each target interval after scaling and quantization, and determining the scaled and quantized log-likelihood ratio with the smallest original bit error rate as the target log-likelihood ratio of soft decoding also includes: In response to the existence of a log likelihood ratio after scaling and quantization being greater than a preset threshold, the preset threshold is determined as a target log likelihood ratio for the soft decoding.

9. An electronic device, characterized in that: It comprises a memory and a processor coupled to each other, wherein the processor is used to execute program instructions stored in the memory to implement the method for determining soft decoding parameters according to any one of claims 1 to 8.

10. A computer-readable storage medium having program instructions stored thereon, characterized in that: When the program instructions are executed by a processor, the method for determining soft decoding parameters according to any one of claims 1 to 8 is implemented.

Images