Data re-read method, electronic device and storage medium

Abstract

Provided in the present application are a data re-read method, an electronic device and a storage medium. The method comprises: recording information from multiple previous re-reads; on the basis of a plurality of re-read reference voltages and a plurality of pieces of column height information from multiple re-reads conducted at each read level, estimating state wave information of each read level; and on the basis of the estimated state wave information, determining a next re-read reference voltage of each read level, so as to carry out the next re-read. On the basis of the information from previous re-reads, quantity information of difference data between two consecutive re-reads is obtained, such that, on the basis of the quantity information, a read offset direction for reducing the data re-read error rate can be determined; thus, on the basis of the read offset direction and the re-read reference voltages from previous re-reads, the next re-read reference voltage can be determined, so as to carry out the next re-read, enabling the error rate of the next re-read to be reduced; that is, the success rate of data read is improved, and the time consumed by data read is thereby reduced.

Description

Data rereading method, electronic device and storage medium Technical Field

[0001] The disclosed embodiments of the present application relate to the field of radio frequency communication technology, and more specifically, to a data rereading method, an electronic device, and a storage medium. Background Art

[0002] Currently, mainstream NAND Flash generally uses 3D TLC / QLC. TLC / QLC's higher storage density and lower costs compared to SLC / MLC make it the preferred choice for most products. However, due to its higher storage density, it is more susceptible to the effects of noise from age, environmental factors, and other factors, exposing NAND Flash to increased threats to data reliability. To ensure data reliability, a reread list can be used to perform repeated read attempts when data errors occur. This is a crucial method for data error correction and recovery.

[0003] However, the current rereading process causes a lot of delay, and each rereading is performed independently, without fully utilizing the relevant information generated in the previous rereading process.

[0004] Therefore, how to reduce the time consumption of data reading becomes an urgent problem to be solved.

[0005] Summary of the Invention

[0006] According to an embodiment of the present application, the present invention provides a data rereading method to reduce the time consumption of data reading.

[0007] According to one aspect of the present application, an exemplary data reread method applied to a memory is disclosed, comprising: recording information of multiple previous rereads, wherein the information of the multiple previous rereads includes multiple reread reference voltages and multiple column height information for multiple rereads of each read level of the same read page, the column height information representing the number of difference data between the read data of two adjacent rereads; estimating state wave information of each read level based on the multiple reread reference voltages and multiple column height information for multiple rereads of each read level; determining the next reread reference voltage for each read level based on the estimated state wave information, and performing the next reread.

[0008] According to a second aspect of the present application, an electronic device is provided, comprising a memory and a processor coupled to each other, wherein the processor is configured to execute program instructions stored in the memory to implement the data rereading method of the first aspect.

[0009] According to a third aspect of the present application, a non-volatile computer-readable storage medium is provided, on which program instructions are stored. When the program instructions are executed by a processor, the data rereading method of the first aspect described above is implemented.

[0010] The above scheme obtains the number information of the difference data read in two adjacent rereads based on the information of the previous reread. The number information can be used to determine the reading offset direction that can reduce the error rate of the reread data. Therefore, the next reread reference voltage can be determined based on the reading offset direction and the reread reference voltage of the previous reread process. The next reread can reduce the error rate of the next reread, that is, improve the success rate of data reading, and thus reduce the time consumption of data reading. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The present application will be further described below with reference to the accompanying drawings and implementation methods, in which:

[0012] FIG1 is a schematic diagram of rereading after a threshold voltage shift occurs;

[0013] FIG2 is a flow chart of an embodiment of a data rereading method of the present application;

[0014] FIG3 is a schematic diagram of reading using different reading gears within a reading page;

[0015] FIG4 is a schematic diagram showing different intervals between different re-read reference voltages of the same read level at different gears;

[0016] FIG5 is a schematic diagram of a state wave fitting curve corresponding to a certain reading level;

[0017] FIG6 is a schematic diagram of a peak occurring during the curve fitting process;

[0018] FIG7 is a flow chart of an embodiment of a sub-step 1 of step S240;

[0019] FIG8 is a schematic diagram of a framework of an electronic device according to an embodiment of the present application;

[0020] FIG9 is a schematic diagram of a framework of an embodiment of a non-volatile computer-readable storage medium of the present application. DETAILED DESCRIPTION

[0021] To help those skilled in the art better understand the technical solutions of this application, the technical solutions of this application are further described below in conjunction with the accompanying drawings and specific implementation plans. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of this application without making any creative efforts are within the scope of protection of this application.

[0022] The terms "first," "second," and the like in this application are used to distinguish between different objects, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements, but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.

[0023] NAND flash memory is a mainstream storage medium for high-capacity storage devices. NAND flash memory consists of a basic memory cell, a MOS transistor similar to an NMOS transistor. The structure used to store charge is called a floating gate, composed of a semiconductor that allows unidirectional current conduction between the source and drain electrodes. The floating gate is surrounded by a silicon oxide film insulator to limit the free diffusion of electrons. A select / control gate sits above it, controlling the current flow between the source and drain electrodes. As a non-volatile memory, NAND flash memory retains the charge within the floating gate even after power is removed.

[0024] NAND flash memory stores data by limiting the diffusion of electrons through an insulating oxide layer. However, during the life cycle of flash memory, as the number of erase and write cycles to the flash memory cell increases, the insulating oxide layer gradually degrades, causing charge to be easily lost and data retention to be shortened. At the same time, in order to provide acceptable large-capacity storage devices, flash memory manufacturers have proposed MLC (Multi-Level Cell), TLC (Triple-Level Cell), and even QLC (Quad-Level Cell) based on SLC (Single-Level Cell). This means that a single storage cell will be used to store multiple bits of information, corresponding to multiple voltage states at the same time. The more bits of information each cell stores, the more complex the control voltage required, and the smaller the threshold voltage range corresponding to each state will be.

[0025] In the data retention scenario, the charge stored in the NAND flash memory storage layer gradually loses and diffuses before being erased. The longer the data in the flash memory is retained, the higher the reliability risk and the more serious the error, which will cause the storage cell threshold voltage to shift toward a low voltage. In the read disturbance scenario, as the number of read operations gradually accumulates, the effect of charge miswriting on the storage layer will gradually amplify, causing the threshold voltage distribution of the NAND flash memory to shift toward a high voltage.

[0026] Due to the threshold voltage shift, adjacent voltage state waves overlap, making it easy to misjudge bits when reading data using the original reread reference voltage, i.e., read errors. Therefore, a reread (read retry) is required to retrieve the correct data. Please refer to Figure 1, which illustrates a reread after a threshold voltage shift. As shown in Figure 1, due to the threshold voltage shift, two adjacent state waves overlap in the middle. The optimal reread reference voltage is the voltage near the trough of the two state waves. The dotted line represents the reread reference voltage (Vref) used for each reread. Five rereads were performed, and the reread reference voltage for the third (3rd) reread was closest to the trough, meaning it was at the optimal position. Therefore, the reread reference voltage used for the third reread was superior to the other reread reference voltages. During a reread, a reread list can typically be used to select a reread reference voltage from the list. However, a greater number of rereads often results in a longer data read latency. The disclosed embodiments of the present application can minimize the number of rereads to reduce the delay in data reading.

[0027] According to one aspect of the present application, an exemplary data rereading method for a memory is disclosed. Referring to FIG. 2 , FIG. 2 is a flow chart illustrating an embodiment of the data rereading method of the present application. Specifically, the data rereading method includes:

[0028] Step S220: Recording information of previous multiple rereads, wherein the information of previous multiple rereads includes multiple reread reference voltages and multiple column height information for each read level of the same read page, wherein the column height information represents the number of difference data between the read data of two adjacent rereads;

[0029] Step S240 : estimating state wave information of each reading level based on a plurality of reread reference voltages and a plurality of column height information obtained by rereading multiple times at each reading level;

[0030] Step S260 : Based on the estimated state wave information, determine the next re-read reference voltage for each read level and perform the next re-read.

[0031] The above scheme obtains the number information of the difference data read in two adjacent rereads based on the information of the previous reread. The number information can be used to determine the reading offset direction that can reduce the error rate of the reread data. Therefore, the next reread reference voltage can be determined based on the reading offset direction and the reread reference voltage of the previous reread process. The next reread can reduce the error rate of the next reread, that is, improve the success rate of data reading, and thus reduce the time consumption of data reading.

[0032] It should be noted that the “multiple times” mentioned in the disclosed embodiments of this application refers to two times or more, such as 2 times, 3 times, etc., and this application does not limit this.

[0033] In this application, TLC is used as an example. In TLC, a basic memory cell can confine eight different magnitudes of electrons in its charge well, corresponding to eight different data values ​​(000 to 111). Furthermore, the basic memory cells corresponding to the eight different magnitudes of electrons stored in their charge wells have different threshold voltages, requiring seven different reread reference voltages to read the stored data. Data is typically read in units of read pages. In TLC, a read page consists of three parts: a low read page, a middle read page, and an upper read page. In MLC, a read page can include a low read page and an upper read page. In some embodiments of this application, the low read page and the upper read page each include two read levels, while the middle read page includes three read levels. That is, two different reread reference voltages are used to read the data in the low read page or the upper read page, while three different reread reference voltages are used to read the data in the middle read page. Of course, other reading methods can also be used for the three parts of the read page, for example, the low read page includes 1 read level, the medium read page includes 2 read levels, and the high read page includes 4 read levels, etc. This application is not limited to this. In this application, one read level corresponds to one reread reference voltage. In addition, the data reread method of this application can be used not only for rereading data of TLC, but also for rereading data of MLC, QLC, etc.

[0034] In some embodiments, the estimated state wave information includes an estimated adjustment offset direction pointing to the state wave trough; wherein, the memory stores a reread list (retry table), the reread list records multiple reread gears, each reread gear is used to record a corresponding reread reference voltage under each read level of the same read page, and the memory performs a reread operation with the corresponding reread reference voltage under each read level corresponding to the selected reread gear.

[0035] For example, for TLC, a reread gear can have seven different reread reference voltages corresponding to the same read page, each corresponding to seven different read levels. When the threshold voltage shifts, although the state wave shifts differently for different threshold voltages, an effective reread gear cannot ensure that each reread reference voltage is located at the optimal position (i.e., near the trough of the state wave after the shift, see Figure 1 for details). However, for a preferred reread gear, the reread reference voltage corresponding to the read level for the same read page (e.g., low read page, medium read page, high read page) will be closer to the trough of the state wave than the reread reference voltages for other gears. See Figure 3, which is a schematic diagram of reading using different read gears within a read page. As shown in Figure 3, taking the reading of a low-read page as an example, the left and right sides are state waves corresponding to two different read levels of the low-read page, and the dotted lines represent the reread reference voltages (Vref) corresponding to the reread gears used in the first (1st), second (2nd) and third (3rd) rereads, respectively. When reading using different reread gears, the number of difference data between the read data obtained between different reread gears is represented by the column heights between different reread reference voltages (i.e., the gray filled portion in Figure 3). The higher the column height, the greater the number of difference data. It can be seen that the column height between the reread reference voltages corresponding to the two read levels corresponding to the reread gear used in the third reread is the lowest, and thus it can be determined that the reread gear used in the third reread is superior to the other gears. Therefore, the direction of decreasing column height, that is, the direction in which the reread reference voltage becomes better, adjust the reread reference voltage in the direction of decreasing column height (i.e., toward the state wave valley), can improve the success rate of the next reread, thereby reducing the number of data rereads and reducing the data reading delay.

[0036] In some embodiments of the present application, the reread reference voltages corresponding to different read levels within the same read page can be adjusted with the same change trend or with different change trends; for example, for the two read levels of the low read page, the reread reference voltages can be increased and decreased respectively, or increased or decreased at the same time, and the present application does not limit this.

[0037] In some embodiments of the present application, recording information about previous rereads may include: storing first data read from a read page during a current reread in a first buffer unit; storing second data read from a read page during a previous reread in a second buffer unit; and storing difference data between the first data and the second data in a third buffer unit to count the number of difference data between the first data and the second data. In other words, the difference data between the first data and the second data can be obtained by storing the read data obtained from the two rereads in different buffer units and calculating the difference between the two buffer units.

[0038] For example, during the first reread, the second data obtained from the first reread can be stored in the first buffer unit. Since there is no previously reread read data at this time, no calculation is required. At the beginning of the second reread, or after the first reread, the second data stored in the first buffer unit is transferred to the second buffer unit, and the first data obtained from the second reread is stored in the first buffer unit. After the second reread, the number of difference data between the two buffer units is calculated and compared. Specifically, the data in the first and second buffer units can be XORed, and the data obtained after the XOR operation can be stored in the third buffer unit. The number of logical values ​​"1" in the third buffer unit can be counted to obtain the number of difference data between the first and second data. Then, the first data obtained from the second reread is transferred to the first buffer unit, and during the third reread, the read data obtained from the third reread is recorded and stored in the second buffer unit. Repeating the above steps can obtain the number of difference data between the two rereads, i.e., the column height information.

[0039] In some embodiments of the present application, estimating state wave information of each read level based on multiple reread reference voltages and multiple column height information of multiple rereads of each read level may include:

[0040] Step S242: In response to the level indication reading between the non-read levels of the memory, compare multiple column height information of multiple rereads of each read level, and estimate the adjustment offset direction of the next reread, wherein the adjustment offset direction is toward the column height decreasing direction.

[0041] Because a basic TLC memory cell can store eight different magnitudes of electrons in its charge well, reading the stored data requires seven different reread reference voltages. These seven different reread reference voltages correspond to seven different read levels. It is understood that the reread reference voltages corresponding to different read levels may have different voltage levels. For example, for one read level, the corresponding reread reference voltage is around 0.5V, while for another read level, the corresponding reread reference voltage may be around 5mV. Therefore, since the reread reference voltages corresponding to different read levels may have different voltage levels, the intervals between different reread reference voltages for different read levels will naturally vary. Without distinguishing between different read levels, if the intervals between reread reference voltages corresponding to the same read level in different read levels are different, it is impossible to convert them to equal intervals for statistical analysis. Therefore, when the memory does not perform level-indicating reading between read levels, a rough overall statistical analysis is performed.

[0042] In some embodiments of the present application, estimating state wave information of each read level based on multiple reread reference voltages and multiple column height information of multiple rereads of each read level may include:

[0043] Step S242: In response to the memory performing a level indication reading between read levels, determining reread reference voltages corresponding to reread gears for performing multiple rereads of each read level, and determining an interval between two adjacent reread reference voltages as an adjustment interval;

[0044] Step S244: comparing two adjacent adjustment intervals, and estimating the estimated column height information corresponding to the smaller adjustment interval based on the column height information corresponding to the larger adjustment interval and the two adjustment intervals;

[0045] Step S246: Compare the column height information of the record corresponding to the smaller adjustment interval with the estimated column height information, and estimate the adjustment offset direction for the next rereading.

[0046] Please refer to the above embodiment. In the case of not performing level indication reading, it is impossible to perform equal interval estimation for the case that the intervals between different re-read reference voltages of the same read level in different gears are different; in the case of performing level indication reading, that is, after distinguishing different read levels, it is impossible to perform equal interval estimation for the case that the intervals between the re-read reference voltages corresponding to the same read level in different gears are not equal. For example: for a certain read level, the re-read reference voltages of different gears are 5mv, 5.5mv and 7mv respectively (the voltage data is only for illustrative purposes and does not constitute a limitation to this application). The intervals of the re-read reference voltages in different gears are not equal, and equal interval estimation can be performed, that is, different voltage intervals can be converted to the same voltage interval.

[0047] The following is a specific example of equal interval estimation. Please refer to Figure 4, which is a schematic diagram of the different intervals between different reread reference voltages of the same read level at different gears; in Figure 4, only one state wave corresponding to the read level is shown. Determine the reread reference voltages corresponding to the reread gears of multiple rereads, for example, determine the reread reference voltages corresponding to the three reread gears as Vref1, Vref2, and Vref3 respectively; compare the intervals between adjacent reread reference voltages as adjustment intervals, for example, the adjustment interval between Vref1 and Vref2 is y1, and the adjustment interval between Vref2 and Vref3 is y2; since the two adjustment intervals are not equal, and y2 is greater than y1, they need to be converted to equal intervals. Obtain the column height information q2 corresponding to y2, and estimate the estimated column height information q corresponding to y1 after adjusting y1 to be equal to y2. The estimated column height information q can be calculated by the following formula: q = a t ,in, The base number a can be configured according to actual conditions. By comparing the estimated column height information q and the recorded column height information q1 corresponding to y1, the adjustment offset direction of the next reread is estimated. For example, when q is greater than q1, the next reread is adjusted in the direction of q1. In other embodiments, the estimated column height information corresponding to the larger adjustment interval can also be estimated based on the column height information corresponding to the smaller adjustment interval and the two adjustment intervals. This application does not limit this.

[0048] In some embodiments, determining a next re-read reference voltage for each read level based on the estimated state wave information and performing the next re-read includes:

[0049] Step S262 : Based on the estimated adjustment offset direction, a reference gear position for the next rereading is selected from the rereading gear positions of the multiple rereadings of each reading level.

[0050] For example, referring to FIG. 4 , assuming that the estimated adjustment offset direction is from Vref1 to Vref3 , the reread gear position corresponding to Vref3 may be used as the reference gear position for the next reread.

[0051] Step S264: selecting the reread selected gear position closest to the reference gear position from the reread selected gear positions as the next reread gear position, wherein the reread selected gear position is another reread gear position that meets the adjustment offset direction relative to the reference gear position.

[0052] For example, if the offset direction is adjusted from Vref1 to Vref3, that is, the direction of voltage increase, the reread reference voltage of a read page in the reread selection gear should be greater than the reread reference voltage of the read page corresponding to the reference gear.

[0053] In some embodiments, step S264 may include: traversing the distances of all reread selection gears relative to the reference gear in sequence, and selecting a reread selection gear with the closest distance as the next reread gear, wherein the distance of any reread selection gear relative to the reference gear is the square root of the sum of the squares of the differences between the corresponding reread reference voltage of the corresponding reread selection gear at each read level and the reread reference voltage of the reference gear at the corresponding read level.

[0054] Specifically, taking a mid-read page with three read levels as an example, the reread reference voltages corresponding to the three read levels in the base gear are [x1, x2, x3], and the reread reference voltages corresponding to the three read levels in the reread selection gear are [y i1 ,y i2 ,y i3 ], where i is a positive integer used to distinguish different re-reading selection gears; the distance D of the re-reading selection gear relative to the reference gear can be calculated by the following formula:

[0055] Step S266 : performing the next re-read based on the corresponding re-read reference voltage at each read level corresponding to the next re-read gear.

[0056] In some embodiments of the present application, the estimated state wave information may include an estimated state wave fitting curve. In this embodiment, in order to obtain the estimated state wave fitting curve, it is necessary to perform level indication reading to distinguish different reading levels, and then perform curve fitting on the column height information distribution of the state wave corresponding to different reading levels. Please refer to Figure 5, Figure 5 is a schematic diagram of the state wave fitting curve corresponding to a certain reading level; after distinguishing different reading levels, the state wave of a certain reading level is read using five reading gears, that is, the 5 re-read reference voltages are [v1, v2, v3, v4, v5], and the column height information between the five readings is statistically analyzed, and the obtained column heights are [d1, d2, d3, d4]; after obtaining the re-read reference voltage and the corresponding column height information, curve fitting can be performed. For example, if a quadratic function is used for curve fitting, then Substitute the quadratic function expression y=ax 2 +bx+c to solve a, b, c. After obtaining the quadratic fitting curve, the vertex formula or derivative can be used to solve the minimum point to obtain the re-read reference voltage corresponding to the minimum column height. The re-read reference voltage corresponding to the minimum column height can be used as the re-read reference voltage for the next re-reading.

[0057] It should be noted that the next reread reference voltage obtained by curve fitting is not necessarily the reread reference voltage in the reread list, that is, the next reread reference voltage may be a voltage other than the reread reference voltage recorded in the reread list.

[0058] It should be noted that curve fitting is not limited to quadratic function curve fitting, and various other curve functions can also be used for fitting. The quadratic function curve fitting herein is merely an example. In addition, if the intervals between the reread reference voltages of different read levels are different, the same interval estimation can be performed with reference to the above embodiment before performing curve fitting, which will not be further described here.

[0059] When using column height information and reread reference voltages for curve fitting, there may be an over-peaking problem. Please refer to Figure 6, which is a schematic diagram of an over-peaking problem during the curve fitting process. In the figure, v1 to v7 are the reread reference voltages used for multiple readings, and d1 to d6 are the column height information corresponding to the multiple readings. When using the column height information and reread reference voltages for quadratic curve fitting, a downward-opening curve will be fitted at points d1 to d2. In this case, the extreme point obtained by solving the vertex formula or derivative is the highest point, rather than the desired lowest point at the bottom of the state wave. Therefore, it is necessary to filter the column height information and use the filtered column height information for curve fitting.

[0060] After obtaining the estimated state wave fitting curve, in some embodiments, when the estimated state wave information includes the estimated state wave fitting curve, please refer to FIG. 7 , which is a flowchart of an embodiment of a sub-step 1 of step S240. The above-mentioned step S240 may include:

[0061] Step S241: Determine the scene type based on the multiple column height information recorded by rereading multiple times at each reading level.

[0062] Step S243 : In response to the scene type being consistent with the first type of scene, performing quadratic function curve fitting based on a plurality of reread reference voltages and a plurality of column height information recorded by rereading multiple times at each read level.

[0063] In some embodiments, the judgment conditions of the scene type include: condition a, the smallest column height information in the previously recorded multiple-time reread information is in the non-edge part; condition b, the largest column height information in the previously recorded multiple-time reread information is in the non-edge part.

[0064] The first type of scenario includes a first sub-scenario and a second sub-scenario. If condition a holds and condition b fails, the scenario is considered the first sub-scenario, indicating that the trough of the state wave is in the middle of the previously repeated information. If condition a and condition b fail, the scenario is considered the second sub-scenario, indicating that neither the peak nor the trough of the state wave is in the middle of the previously repeated information. Therefore, in the first type of scenario, when using the collected column height information for curve fitting, a downward-opening curve will not be obtained. For example, as shown in Figure 5, the minimum column height is d3, which is located outside the edge, so condition a holds; the maximum column height is d1, which is located at the edge, so condition b fails, thus belonging to the first sub-scenario of the first type of scenario. That is, in the first type of scenario, the maximum column height must be located at the edge of the fitting area (e.g., the area containing d1 or d4 in Figure 5). The fitting area is the area between v1 and v5 in Figure 5. In the first type of scenario, the collected column height information and multiple re-read reference voltages can be directly used to perform quadratic function curve fitting without screening, because no over-peak will occur in this case.

[0065] Step S245 : In response to the scene type being consistent with the second type scene, screening multiple reread reference voltages and multiple pillar height information of each read level, and performing quadratic function curve fitting based on the screened parts of the multiple reread reference voltages and multiple pillar height information.

[0066] The second type of scenario includes sub-scenarios 3 and 4. If condition a fails but condition b does, the scenario is considered the third sub-scenario, with the peak of the state wave located in the middle of the previously reread data. If both conditions a and b are met, the scenario is considered the fourth sub-scenario, with both the peak and trough of the state wave located in the middle of the previously reread data. In other words, in the second type of scenario, the maximum column height information collected must be located in the non-edge portion of the fitting area. As shown in Figure 6, the minimum column height is d5, which is located in the non-edge portion; the maximum column height is d2, which is also located in the non-edge portion. Therefore, both conditions a and b hold, representing the fourth sub-scenario. In the second type of scenario, due to overpeaking, the multiple reread reference voltages and column height information collected must be screened to prevent the resulting quadratic function curve from opening downward after fitting, which would prevent the next reread reference voltage from being obtained.

[0067] In some embodiments, in response to the scenario type being consistent with the second type scenario, filtering a plurality of reread reference voltages and a plurality of pillar height information for each read level, and performing quadratic function curve fitting based on the filtered portion of the plurality of reread reference voltages and the plurality of pillar height information, including:

[0068] Step S245a: Based on the multiple column height information in the previously reread information, select the location of the largest column height information.

[0069] Step S245b: Based on the location of the maximum column height information, the multiple column height information in the previously reread information is divided into two groups, and the sum of the column height information in each group is counted.

[0070] Step S245c: The group with the smaller sum of column height information is selected as the part to be fitted with a quadratic function curve.

[0071] As shown in Figure 6, in the second type of scenario, the position of the maximum column height information is first determined. Taking the maximum column height information as the basis, the multiple column height information on both sides are divided into two groups, that is, d1 and d2 are one group, and d3 to d6 are one group. The sum of the two groups of column height information is counted, and the group with the smaller sum of column height information is used as fitting data to perform quadratic function curve fitting; specifically in Figure 6, since the sum of d3 to d6 is smaller than the sum of d1 and d2, d3 to d6 are used as the screened out part and used as fitting data to fit the quadratic function curve.

[0072] According to a second aspect of the present application, an electronic device is provided. Referring to FIG. 8 , FIG. 8 is a schematic diagram illustrating a framework of an embodiment of the electronic device of the present application. The electronic device includes a memory and a processor coupled to each other, the processor being configured to execute program instructions stored in the memory to implement the aforementioned data rereading method. In a specific implementation scenario, the electronic device 80 may include, but is not limited to, a microcomputer and a server. Furthermore, the electronic device 80 may also include a mobile device such as a laptop computer or a tablet computer, without limitation herein.

[0073] Specifically, the processor 82 is used to control itself and the memory 81 to implement the steps of the training method embodiment of any of the above-mentioned image detection models, or to implement the steps in the above-mentioned image detection method embodiment. The processor 82 can also be referred to as a CPU (Central Processing Unit). The processor 82 may be an integrated circuit chip with signal processing capabilities. The processor 82 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a 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 82 can be implemented by an integrated circuit chip.

[0074] Please refer to Figure 9, which is a schematic diagram of a framework of an embodiment of a non-volatile computer-readable storage medium of the present application. The non-volatile computer-readable storage medium 90 stores program instructions 901 that can be executed by a processor. When the program instructions 901 are executed by the processor, the steps of any of the above-mentioned data rereading method embodiments are implemented.

[0075] The above scheme obtains the number information of the difference data read in two adjacent rereads based on the information of the previous reread. The number information can be used to determine the reading offset direction that can reduce the error rate of the reread data. Therefore, the next reread reference voltage can be determined based on the reading offset direction and the reread reference voltage of the previous reread process. The next reread can reduce the error rate of the next reread, that is, improve the success rate of data reading, and thus reduce the time consumption of data reading.

[0076] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments. For the sake of brevity, it will not be repeated here.

[0077] The above description of the various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced with each other and will not be repeated herein for the sake of brevity.

[0078] In the several embodiments provided in this application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation methods described above are 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. For example, 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.

[0079] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0080] 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, 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, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0081] It is easy for a person skilled in the art to know that many modifications and variations can be made to the apparatus and method while maintaining the teaching content of the present application.Therefore, the above disclosure should be considered as being limited only by the scope of the appended claims.

Claims

1. A data rereading method, applied to a memory, characterized in that, it includes: Recording information of multiple prior rereadings, wherein the information of the multiple prior rereadings includes multiple rereading reference voltages and multiple column height information for each reading level of the same read page, and the column height information represents the number information of difference data between the read data of two adjacent rereadings; Based on the multiple rereading reference voltages and multiple column height information for each reading level during multiple rereadings, estimating the state wave information for each reading level; Based on the estimated state wave information, determining the next rereading reference voltage for each reading level and performing the next rereading.

2. The method according to claim 1, characterized in that, The estimated state wave information includes an estimated adjustment offset direction pointing to the trough of the state wave; wherein, the memory stores a rereading list, the rereading list records multiple rereading gears, each rereading gear is respectively used to record the rereading reference voltage at each reading level of the same read page, and the memory performs a rereading operation with the rereading reference voltage at each reading level corresponding to the selected rereading gear, and the information of the multiple prior rereadings is the information of multiple rereadings using multiple rereading gears in the rereading list.

3. The method according to claim 2, characterized in that, Based on the multiple rereading reference voltages and multiple column height information for each reading level during multiple rereadings, estimating the state wave information for each reading level includes: In response to the memory performing a level indication read between the reading levels, comparing the multiple column height information for each reading level during multiple rereadings, and estimating the adjustment offset direction of the next rereading, wherein the adjustment offset direction is towards the direction of decreasing column height.

4. The method according to claim 2, characterized in that, Based on the multiple rereading reference voltages and multiple column height information for each reading level during multiple rereadings, estimating the state wave information for each reading level includes: In response to the memory performing a level indication read between the reading levels, determining the rereading reference voltages corresponding to the rereading gears for each reading level during multiple rereadings respectively, and determining the interval between two adjacent rereading reference voltages as the adjustment interval; Comparing two adjacent adjustment intervals, and estimating the estimated column height information corresponding to the adjustment interval with a smaller value based on the column height information corresponding to the adjustment interval with a larger value and the two adjustment intervals; Comparing the recorded column height information corresponding to the adjustment interval with a smaller value and the estimated column height information, and estimating the adjustment offset direction of the next rereading.

5. The method according to claim 2, characterized in that, Based on the estimated state wave information, determining the next rereading reference voltage for each reading level and performing the next rereading includes: Based on the estimated adjustment offset direction, selecting the next rereading reference gear from the multiple rereading gears for each reading level; Select the reread selection gear closest to the reference gear in the reread selection gears as the next reread gear, where the reread selection gears are other reread gears that conform to the adjustment offset direction relative to the reference gear; Perform the next reread based on the corresponding reread reference voltages at each of the read levels corresponding to the next reread gear.

6. The method according to claim 5, wherein, Selecting the reread selection gear closest to the reference gear in the reread selection gears as the next reread gear includes: Traverse the distances of all the reread selection gears relative to the reference gear in sequence, and select the reread selection gear with the closest distance as the next reread gear, where the distance of any reread selection gear relative to the reference gear is the square root of the sum of the squares of the differences between the corresponding reread reference voltages of the corresponding reread selection gear at each read level and the reread reference voltage of the reference gear at the corresponding read level.

7. The method according to claim 1, wherein, The estimated state wave information includes an estimated state wave fitting curve.

8. The method according to claim 7, wherein, Estimating the state wave information for each read level based on the multiple reread reference voltages and the multiple column height information obtained by performing multiple rereads at each read level includes: Judging the scene type based on the multiple column height information recorded by performing multiple rereads at each read level; In response to the scene type conforming to the first type of scene, perform quadratic function curve fitting based on the multiple reread reference voltages and the multiple column height information recorded by performing multiple rereads at each read level; In response to the scene type conforming to the second type of scene, screen the multiple reread reference voltages and the multiple column height information for each read level, and perform quadratic function curve fitting based on the selected part of the multiple reread reference voltages and the multiple column height information.

9. The method according to claim 8, wherein, The judgment conditions for the scene type include: Condition a, the smallest column height information in the previously recorded information of multiple rereads is in the non-edge part; Condition b, the largest column height information in the previously recorded information of multiple rereads is in the non-edge part; The first type of scene includes a first sub-scene and a second sub-scene, where if Condition a is satisfied and Condition b is not satisfied, it is determined as the first sub-scene, indicating that the trough of the state wave is in the middle part of the previously recorded information of multiple rereads; if Condition a is not satisfied and Condition b is not satisfied, it is determined as the second sub-scene, indicating that neither the peak nor the trough of the state wave is in the middle part of the previously recorded information of multiple rereads; The second type of scenario includes a third sub-scenario and a fourth sub-scenario. Among them, if condition a does not hold and condition b holds, it is determined as the third sub-scenario, which represents that the peak of the state wave is in the middle part of the information read multiple times previously; if condition a holds and condition b holds, it is determined as the fourth sub-scenario, which represents that both the peak and the trough of the state wave are in the middle part of the information read multiple times previously.

10. According to the method described in claim 8, wherein, in response to the scenario type conforming to the second type of scenario, screening the multiple reread reference voltages and the multiple column height information of each read level, and performing quadratic function curve fitting based on the selected part of the multiple reread reference voltages and the multiple column height information, including: Based on the multiple column height information in the information read multiple times previously, selecting the position where the maximum column height information is located; Based on the position where the maximum column height information is located, dividing the multiple column height information in the information read multiple times previously into two groups, and statistically calculating the sum of the column height information in each group; The group with the smaller sum of the column height information is used as the selected part for quadratic function curve fitting.

11. According to the method described in claim 1, wherein, recording the information read multiple times previously includes: storing the first data read from the read page during the current reread in the first buffer unit; storing the second data read from the read page during the previous reread in the second buffer unit; storing the difference data between the first data and the second data in the third buffer unit to statistically calculate the number of difference data between the first data and the second data.

12. An electronic device, wherein, it includes a memory and a processor coupled to each other, and the processor is configured to execute program instructions stored in the memory to implement the data reread method described in any one of claims 1 to 11.

13. A non-volatile computer-readable storage medium, on which program instructions are stored, wherein, when the program instructions are executed by a processor, the data reread method described in any one of claims 1 to 11 is implemented.

Images