Information processing device, magnetic recording and reproducing device, and magnetic recording and reproducing system

Abstract

To provide an information processing device capable of improving processing accuracy, a magnetic recording / reproducing device, and a magnetic recording / reproducing system.SOLUTION: An information processing device includes an acquisition unit and a processing unit. The acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including recording media. The reproduction signal includes a first signal corresponding to information recorded on the recording media. The processing unit is capable of deriving a first output obtained by processing first information including the first signal by a first processing model and a second output obtained by processing the first information by a second processing model. The processing unit is capable of outputting processing results the first information based on a third output obtained based on the first output, the second output, and the first information.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing device, a magnetic recording and reproducing device, and a magnetic recording and reproducing system. [Background technology]

[0002] For example, information recorded on a magnetic recording medium or the like is reproduced by an information processing device, etc. For example, by improving the processing accuracy of the information processing device, the magnetic recording density can be improved. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] US Patent Application Publication No. 2020 / 0389188 Summary of the Invention [Problem to be solved by the invention]

[0004] The embodiments of the present invention provide an information processing device, a magnetic recording and reproducing device, and a magnetic recording and reproducing system that are capable of improving processing accuracy. [Means for solving the problem]

[0005] According to an embodiment of the present invention, an information processing device includes an acquisition unit and a processing unit. The acquisition unit is capable of acquiring a playback signal obtained from a recording unit including a recording medium. The playback signal includes a first signal corresponding to information recorded on the recording medium. The processing unit is capable of deriving a first output obtained by processing first information including the first signal using a first processing model, and a second output obtained by processing the first information using a second processing model. The processing unit is capable of outputting a result of processing the first information based on the first output, the second output, and a third output obtained based on the first information. [Brief explanation of the drawings]

[0006] [Figure 1]FIG. 1 is a schematic diagram illustrating an information processing device and an information recording / reproducing device according to the first embodiment. [Figure 2] FIG. 2 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 3] FIG. 3 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 4] FIG. 4 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 5] FIG. 5 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 6] FIG. 6 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 7] FIG. 7 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 8] FIG. 8 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 9] FIG. 9 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 10] 10(a) to 10(c) are schematic views illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. [Figure 11] FIG. 11 is a schematic view illustrating the operations of the information processing device and the information recording and reproducing device according to the first embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0007] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The drawings are schematic or conceptual, and the relationship between the thickness and width of each part, the size ratio between parts, etc. are not necessarily the same as those in reality. Even when the same part is shown, the dimensions and ratios may be different depending on the drawing. In this specification and in each drawing, elements similar to those previously described with reference to the previous drawings are designated by the same reference numerals, and detailed descriptions thereof will be omitted where appropriate.

[0008] (First embodiment) FIG. 1 is a schematic diagram illustrating an information processing device and an information recording / reproducing device according to the first embodiment. 1, the information recording / reproducing device 210 according to the embodiment includes an information processing device 70 and a recording unit 80D. The recording unit 80D includes a recording medium 80. The recording medium 80 may include, for example, a magnetic recording medium (such as a magnetic disk (HDD: Hard Disk Drive)). The recording unit 80D may include, for example, an SSD (Solid State Drive).

[0009] The recording unit 80D may include, for example, a reproducing unit 80R. The reproducing unit 80R includes, for example, a magnetic head 80H. The reproducing unit 80R (magnetic head 80H) reproduces information I00 recorded on the recording medium 80. A reproduction signal Sr1 corresponding to the reproduced information is output from the reproducing unit 80R. The reproduction signal Sr1 is an electrical signal. Information may be recorded on the recording medium 80 by the magnetic head 80H.

[0010] The information processing device 70 includes an acquisition unit 71 and a processing unit 72. The acquisition unit 71 is capable of acquiring a reproduction signal Sr1 obtained from a recording unit 80D including a recording medium 80. As described above, the reproduction signal Sr1 includes a first signal S1 corresponding to information I00 recorded on the recording medium 80.

[0011] In one example, the reproduction signal Sr1 may include information (attribute P0) related to the first signal S1. In this case, the attribute P0 is supplied from the recording unit 80D to the acquisition unit 71 (and the processing unit 72). In another example, the attribute P0 may not be included in the reproduction signal Sr1. For example, the attribute P0 may be predetermined in correspondence with the first signal S1. For example, the attribute P0 may be stored in another storage unit or the like, and provided to the acquisition unit 71 (and the processing unit 72) from the other storage unit. This other storage unit may be included in the information processing device 70. This other storage unit may be provided separately from the information processing device 70. In one example, the attribute P0 is, for example, recording density (BPI: bits per inch).

[0012] The processing unit 72 processes the first signal S1 (e.g., a waveform) based on first information I01 including the first signal S1. The processing unit 72, for example, processes the first signal S1 to determine whether the first signal S1 is a first value or a second value. The second value is different from the first value. The first value is, for example, one of "0" and "1." The second value is the other of "0" and "1." The processing unit 72 performs a "0 / 1 determination." The processing unit 72 can output a result R1 obtained by the processing performed by the processing unit 72. The result R1 includes the result of the "0 / 1 determination." The result R1 includes information regarding whether the first signal S1 is the first value or the second value.

[0013] In an embodiment, the processing in the processing unit 72 is performed based on one or more processing models, including, for example, neural networks.

[0014] 1, the processing unit 72 includes, for example, a plurality of neural networks (a plurality of processing models), such as NN1 (first neural network), NN2 (second neural network), NN3 (third neural network), and NN4 (fourth neural network).

[0015] Hereinafter, several examples of the operation of the processing unit 72 of the information processing device 70 according to the embodiment will be described.

[0016] FIG. 2 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. FIG. 2 illustrates a first operation OP1 performed by the information processing device 70 (or the information recording / reproducing device 210).

[0017] 2, the processing unit 72 acquires first information 101 including the first signal S1 (step S110). As described above, the acquisition unit 71 acquires the reproduced signal Sr1. The first information 101 including the first signal S1 included in the reproduced signal Sr1 is provided to the processing unit 72.

[0018] The processing unit 72 processes the first information I01 including the first signal S1 using a first processing model (first processing: step S141). The processing unit 72 calculates a first likelihood obtained by the first processing (step S161). The processing unit 72 processes the first information I01 using a second processing model (second processing: step S142). The processing unit 72 calculates a second likelihood obtained by the second processing (step S162).

[0019] In step S141, for example, processing is performed using a first neural network NN1 as a first processing model. As a result, a first likelihood is obtained. In step S142, for example, processing is performed using a second neural network NN2 as a second processing model. As a result, a second likelihood is obtained. In this way, the processing unit 72 can derive a first output (for example, a first likelihood) obtained by processing the first information I01 including the first signal S1 using the first processing model, and a second output (for example, a second likelihood) obtained by processing the first information I01 using the second processing model.

[0020] 2, the processing unit 72 derives a third output (e.g., a third likelihood) based on the first output (e.g., a first likelihood) and the second output (e.g., a second likelihood) (third processing: step S163). For example, the third likelihood is obtained by processing using a third processing model (third neural network NN3).

[0021] 2, the processing unit 72 outputs a result R1 (third output) of processing the first information I01 based on the third likelihood (step S164). After step S164, the process may return to step S110. In one example, the third output is, for example, the third likelihood. In one example, the third output may include the result of the "0 / 1 decision."

[0022] FIG. 3 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. 3 illustrates the first operation OP1. As shown in FIG. 3, a plurality of processing models 40 are provided, such as "Model A," "Model B," "Model C," "Model D," "Model E," etc. "Model A," "Model B," "Model C," "Model D," and "Model E" correspond to, for example, "BPI-A," "BPI-B," "BPI-C," "BPI-D," and "BPI-E," respectively.

[0023] The first information I01 including the first signal S1 is processed by each of the multiple processing models 40, and multiple likelihoods 45 are calculated. The multiple likelihoods 45 include, for example, "likelihood A," "likelihood B," "likelihood C," "likelihood D," "likelihood E," etc. For example, a processing model (e.g., "model X") is derived from the multiple likelihoods 45 and the first information I01 including the first signal S1 (step S163). In one example, the "model X" may be one of "model A," "model B," "model C," "model D," "model E," etc. The "model X" may also be another model derived from "model A," "model B," "model C," "model D," "model E," etc. For example, the multiple likelihoods 45 and the first information I01 including the first signal S1 are processed by the "model X," and a "likelihood Z" (third likelihood) is output using the output information (step S164). For example, a "0 / 1 decision" is made using "likelihood Z" (third likelihood).

[0024] For example, a third likelihood (e.g., likelihood Z) may be output as result R1 from the processing unit 72. For example, the result (third output) of the "0 / 1 determination" may be output as result R1 from the processing unit 72.

[0025] In the first operation OP1, first information I01 including the first signal S1 is processed by a plurality of processing models 40 (a plurality of neural networks), and a plurality of likelihoods 45 are derived based on the results. Processing is performed based on the plurality of likelihoods 45 and a third output (third likelihood, i.e., "likelihood Z") derived based on the first information I01, and a third output is output. In the third processing using the third likelihood, for example, "0 / 1 determination" can be performed with higher accuracy than the result of processing using a likelihood corresponding to the attribute P0 of the first signal S1 (e.g., recording density (BPI: bits per inch)). According to the embodiment, it is possible to provide an information processing device capable of improving processing accuracy.

[0026] In the first operation OP1, the result R1 includes information regarding whether the first signal S1 corresponds to a first value or whether the first signal S1 corresponds to a second value. For example, processing the first information I01 based on a third likelihood (e.g., "likelihood Z") may include determining whether the first signal S1 corresponds to the first value or whether the first signal S1 corresponds to the second value based on the third likelihood (a "0 / 1 decision").

[0027] The first processing model includes, for example, a first neural network NN1 trained by machine learning based on a plurality of first training data including record information recorded with a first attribute. The second processing model includes, for example, a second neural network NN2 trained by machine learning based on a plurality of second training data including record information recorded with a second attribute. For example, these attributes (such as the first attribute and the second attribute) may relate to the recording density (e.g., BPI) of the recording medium 80.

[0028] FIG. 4 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. FIG. 4 illustrates a second operation OP2 performed by the information processing device 70 (or the information recording / reproducing device 210).

[0029] 4, the processing unit 72 acquires first information 101 including the first signal S1 (step S110). As described above, the acquisition unit 71 acquires the reproduced signal Sr1. The first information 101 including the first signal S1 included in the reproduced signal Sr1 is provided to the processing unit 72.

[0030] The processing unit 72 selects a first processing model from the plurality of processing models based on the first information 101 (step S120). For example, one neural network included in the plurality of neural networks is selected as the first processing model. In one example, this selection process may be performed using one of the plurality of neural networks (for example, the fourth neural network NN4).

[0031] The processing unit 72 processes the first information I01 based on the selected processing model (step S140).

[0032] For example, if the first processing model (for example, the first neural network NN1) is selected in step S120, the processing unit 72 processes the first information I01 based on the first processing model (first processing: step S141).

[0033] For example, if the second processing model (e.g., the second neural network NN2) is selected in step S120, the processing unit 72 processes the first information I01 based on the second processing model (second processing: step S142). In this example, the first processing model is selected in step S120, and in step S140, step S141 is performed but step S142 is not performed.

[0034] As shown in FIG. 4, the processing unit 72 outputs the result R1 of processing the first information I01 based on the selected processing model (the first processing model in this example) (step S150).

[0035] In this embodiment, an appropriate processing model is selected based on the first information I01 including the acquired first signal S1, thereby enabling more appropriate processing to be performed.

[0036] For example, there is an information processing device that processes signals using a neural network. In one example, signals with various attributes P0 are processed by one type of neural network in the information processing device. According to the inventor's investigation, it has been found that when signals with different signal characteristics (for example, attributes P0 such as recording density) are processed by one neural network, highly accurate processing results are not necessarily obtained.

[0037] In the embodiment, an appropriate processing model is selected from a plurality of processing models (a plurality of neural networks). For example, signals with different characteristics (attributes P0) can be processed more appropriately. According to the embodiment, processing results with higher accuracy can be obtained compared to the reference example in which processing is performed using a single neural network.

[0038] As shown in Fig. 4, after step S150, the process may return to step S110. The above operations may be repeated.

[0039] FIG. 5 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. 5 illustrates the above-described step S120 in the second operation OP2. For example, a processing model is selected based on the first information I01 including the first signal S1 (step S121). For example, the first processing model is selected as the processing model corresponding to the first signal S1, as described above. In this example, the first processing model is "model D" (step S122). "Model D" may be associated with, for example, attribute information estimated for the first signal S1. The attribute information may correspond to, for example, a recording density (e.g., BPI).

[0040] FIG. 6 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. 6 illustrates the above-mentioned step S140 in the second operation OP2. As shown in FIG. 6, multiple processing models 40 are provided, such as "Model A," "Model B," "Model C," "Model D," "Model E," etc. "Model A," "Model B," "Model C," "Model D," and "Model E" correspond to, for example, "BPI-A," "BPI-B," "BPI-C," "BPI-D," and "BPI-E," respectively.

[0041] In this example, as described above, a "model D" is selected for the first information I01 including the first signal S1. In this case, the first information I01 is processed by the selected "model D." For example, a "likelihood D" is obtained by the processing. The "likelihood D" becomes the "likelihood Z" used in the "0 / 1 determination" of the first signal S1. That is, the "0 / 1 determination" of the first signal S1 is performed using the selected "likelihood D" (step S145).

[0042] Thus, in the second operation OP2, the result R1 of processing the first information I01 by the first processing model includes information on whether the first signal S1 corresponds to a first value (one of 0 and 1) or whether the first signal S1 corresponds to a second value (the other of 0 and 1). The result R1 may include, for example, the derived likelihood Z. The result R1 may include, for example, a "0 / 1 decision" result. The result R1 is output.

[0043] The multiple processing models 40 (see FIG. 6) may include an m-th processing model, where "m" is an integer between 1 and N, inclusive. "N" is an integer greater than or equal to 2. The m-th processing model includes, for example, an m-th neural network. The m-th neural network is machine-learned based on, for example, multiple training data sets including recording information recorded with the m-th attribute. The m-th attribute relates, for example, to the recording density (e.g., BPI) of the recording medium 80.

[0044] FIG. 7 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. FIG. 7 illustrates a third operation OP3 performed by the information processing device 70 (or the information recording / reproducing device 210).

[0045] In the third operation OP3, the reproduction signal Sr1 (see FIG. 1) includes the first signal S1 corresponding to the information I00 recorded on the recording medium 80.

[0046] 7, the processing unit 72 acquires first information 101 including the first signal S1 and an attribute P0 (step S111). As already described, the attribute P0 may be included in the reproduced signal Sr1. Alternatively, the attribute P0 may be predetermined in association with the first signal S1. For example, the attribute P0 may be provided from another storage unit.

[0047] 7, the processing unit 72 determines whether the attribute P0 is the first attribute (step S131). If the attribute P0 is the first attribute, the processing unit 72 processes the first information I01 using a first processing model corresponding to the first attribute (first processing: step S141).

[0048] If the attribute P0 is not the first attribute, the processing unit 72 processes the first information I01 using another processing model (e.g., a second processing model) corresponding to the attribute P0 (second processing: step S142). The first processing includes, for example, processing using the first neural network NN1. The second processing includes, for example, processing using the second neural network NN2.

[0049] The processing unit 72 outputs the result of the first process or the second process (step S150). After step S150, the process may return to step S111.

[0050] Thus, in the third operation OP3, the processing unit 72 outputs a result R1 (see FIG. 1) of processing the first information I01 including the first signal S1 (step S150). In the third operation OP3, if the attribute P0 is the first attribute, the result R1 is obtained by processing the first information I01 using a first processing model corresponding to the first attribute. If the attribute P0 is a second attribute different from the first attribute, the result R1 is obtained by processing the first information I01 using a second processing model corresponding to the second attribute. The second processing model is different from the first processing model. The processing model may correspond to the attribute P0 (e.g., BPI).

[0051] In the third operation OP3, the first information I01 including the first signal S1 is processed using a processing model according to the attribute P0 of the first signal S1, thereby obtaining a highly accurate processing result.

[0052] In the third operation OP3, for example, if the attribute P0 is the first attribute, the result R1 includes the result of a "0 / 1 decision" made based on a first output (e.g., a first likelihood) corresponding to the first attribute. For example, if the attribute P0 is the second attribute, the result R1 includes the result of a "0 / 1 decision" made based on a second output (e.g., a second likelihood) corresponding to the second attribute.

[0053] In the third operation OP3, the first processing model includes, for example, a first neural network NN1 trained by machine learning based on a plurality of first teacher data including record information recorded with a first attribute, and the second processing model includes a second neural network NN2 trained by machine learning based on a plurality of second teacher data including record information recorded with a second attribute.

[0054] FIG. 8 is a flowchart illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. FIG. 8 illustrates a fourth operation OP4 performed by the information processing device 70 (or the information recording / reproducing device 210).

[0055] In the fourth operation OP4, the reproduction signal Sr1 (see FIG. 1) includes a first signal S1 corresponding to the information I00 recorded on the recording medium 80 and an attribute P0 related to the first signal S1.

[0056] As shown in FIG. 8, the processing unit 72 acquires the first information I01 including the first signal S1 and the attribute P0 (step S111).

[0057] The processing unit 72 processes the first information I01 including the first signal S1 using a first processing model corresponding to the first attribute (first processing: step S141). The processing unit 72 processes the first information I01 using another processing model (second processing model) corresponding to another attribute (e.g., a second attribute) (second processing: step S142). The second processing model is different from the first processing model. The first processing includes, for example, processing using a first neural network NN1. The second processing includes, for example, processing using a second neural network NN2.

[0058] The processing unit 72 determines whether the acquired attribute P0 is the first attribute or not (step S131). If the attribute P0 is the first attribute, the processing unit 72 outputs a first processing result (step S151). If the attribute P0 is not the first attribute, the processing unit 72 outputs a second processing result (step S152). After step S151 or step S152, the processing may return to step S111.

[0059] Thus, in the fourth operation OP4, the processing unit 72 outputs the result of processing using the first processing model when the attribute P0 is the first attribute, and outputs the result of processing using the second processing model (result R1: see Figure 1) when the attribute P0 is the second attribute.

[0060] In the fourth operation OP4, the first information 101 including the first signal S1 is processed using a processing model according to the attribute P0 of the first signal S1, and the result is output. A highly accurate processing result is obtained.

[0061] In the fourth operation OP4, for example, when the attribute P0 is the first attribute, the output result R1 includes the result of the "0 / 1 determination" made based on the first output (first likelihood) corresponding to the first attribute. For example, when the attribute P0 is the second attribute, the output result R1 includes the result of the determination made based on the second output (e.g., second likelihood) corresponding to the second attribute.

[0062] In the fourth operation OP4, the first processing model includes a first neural network NN1 trained by machine learning based on a plurality of first training data including record information recorded with a first attribute. The second processing model includes a second neural network NN2 trained by machine learning based on a plurality of second training data including record information recorded with a second attribute. In the fourth operation OP4, the attribute P0 relates to, for example, the recording density (e.g., BPI) of the recording medium 80.

[0063] The attribute P0 may correspond to multiple ranges of BPI values. For example, the first attribute may be a first data group corresponding to a first BPI value or more and less than a second BPI value. For example, the second attribute may be a second data group corresponding to a third BPI value or more and less than a fourth BPI value. Some of the data included in the first data group may be the same as at least some of the data included in the second data group. At least some of the data included in the first data group may be different from at least some of the data included in the second data group.

[0064] In an embodiment, for example, information I00 recorded on a recording medium 80 is reproduced by a magnetic head 80H. The reproduced reproduction signal Sr1 is, for example, AD converted. Waveform equalization may be performed on the reproduction signal Sr1. The signal obtained by AD conversion (or the signal obtained by waveform equalization) is decoded. In the decoding, a "0 / 1 determination" is made from the waveform of the digital data. After this, error correction processing is performed using, for example, LDPC (low-density parity check). Furthermore, conversion of the bit string according to the encoding method is performed. As a result, the recorded information I00 is reproduced.

[0065] The above-described first to fourth operations OP1 to OP4 according to the embodiment may be applied to the above-described decoding.

[0066] FIG. 9 is a schematic view illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. 9, for example, the first signal S1 (for example, a waveform) may be the output of an equalizer 31. In this way, the first signal S1 may be obtained by equalizing at least a part of the reproduced signal Sr1.

[0067] For example, the first signal S1 is used to decode one bit. The first signal S1 may include, for example, a signal corresponding to the bit to be decoded and a signal that may have an effect such as interference on the signal. For example, the first signal S1 may include the bit to be decoded and signals corresponding to X bits adjacent to and before and after the bit to be decoded. In this case, the first signal S1 is a signal corresponding to "2X+1" bits.

[0068] The first information I01 including the equalized first signal S1 is processed by, for example, a neural network (NN). The neural network (NN) is, for example, a "0 / 1 classifier." The output of the "0 / 1 classifier" may be subjected to processing (e.g., LDPC) by, for example, an error corrector 32. The error correction may include processing based on at least one of an ECC (Error-Correcting Code) and an LDPC. The processing based on the LDPC is an example of error correction processing. For example, the processing unit 72 may be capable of outputting information obtained by performing error correction processing on the result R1.

[0069] As shown in FIG. 9, at least a part of the results of the processing by the error corrector 32 may be input to a "0 / 1 classifier" (NN).

[0070] 9, the first information I01 may further include second information I02 obtained by performing error correction processing on the result R1 in addition to the first signal S1. The second information I02 may include, for example, a likelihood obtained by the error correction processing.

[0071] In one example, the second information I02 may be, for example, information about 0 / 1. In one example, the second information I02 may be likelihood.

[0072] For example, the first decoding is performed using the target bit and a signal corresponding to the adjacent "X" bits (a signal corresponding to "2X+1" bits). In this case, the inputs for the second and subsequent decodings may be the first signal S1 acquired from the second time onwards and the output of the error corrector 32 corresponding to the past "2X+1" bits. The output of the error corrector 32 corresponding to the past "2X+1" bits is, for example, "2X+1" likelihoods corresponding to the "2X+1" bits. In this case, the model of the "0 / 1 classifier" (NN) has, for example, (2X+1) × 2 (= 4X+2) inputs.

[0073] An example of such processing (decoding) will be described below. 10(a) to 10(c) and 11 are schematic views illustrating the operations of the information processing device and the information recording / reproducing device according to the first embodiment. 10(a) to 10(c) and 11 illustrate the case where the above "X" is 1. In FIG. 10(a), the first determination process for the "i-1" bit is illustrated. [i] is an integer. In FIG. 10(b), the first determination process for the "i" bit is illustrated. In FIG. 10(c), the first determination process for the "i+1" bit is illustrated.

[0074] For example, when a first determination is made on bit "i", a second waveform related to bit "i" is input to the first NN. The second waveform includes, for example, three pieces of information: information corresponding to bit "i", information corresponding to bit "i-1" which is an adjacent bit of bit "i", and information corresponding to bit "i+1" which is an adjacent bit of bit "i". The first NN is, for example, a three-input NN. Information on whether bit "i" is a first value or a second value (first output related to bit "i") is output.

[0075] Similarly, for example, when a first determination is made on the "i-1" bit, a first waveform related to the "i-1" bit is input to the first NN. The first waveform includes, for example, three pieces of information: information corresponding to the "i-1" bit, information corresponding to the "i-2" bit that is the adjacent bit of the "i-1" bit, and information corresponding to the "i" bit that is the adjacent bit of the "i-1" bit. The first NN is, for example, a three-input NN. Information regarding whether the "i-1" bit is a first value or a second value (first output related to the "i-1" bit) is output.

[0076] Similarly, for example, when a first determination is made on the "i+1" bit, a third waveform related to the "i+1" bit is input to the first NN. The third waveform is, for example, three pieces of information: information corresponding to the "i+1" bit, information corresponding to the "i" bit that is adjacent to the "i+1" bit, and information corresponding to the "i+2" bit that is adjacent to the "i+1" bit. The first NN is, for example, a three-input NN. Information regarding whether the "i+1" bit is the first value or the second value (first output related to the "i+1" bit) is output.

[0077] As shown in FIG. 11, for the "i" bit, the second waveform, the "k-1"th output for the "i-1" bit, the "k-1"th output for the "i" bit, and the "k-1"th output for the "i+1" bit are input to the second NN. The second NN is, for example, a 6-input NN. "k" is an integer equal to or greater than 2. When "k" is 2, the first outputs exemplified in FIG. 10(b) (the first output for the "i-1" bit, the first output for the "i" bit, and the first output for the "i+1" bit) are input.

[0078] The first output of the first determination process illustrated in FIG. 10 may be generated using the second NN illustrated in FIG. 11. For example, the first output may be generated by inputting set values as three values to be input to the second NN: the "k-1" output for the "i-1" bit, the "k-1" output for the "i" bit, and the "k-1" output for the "i+1" bit. The set value is, for example, an intermediate value between the lower limit and upper limit of the "k-1" output for the "i" bit. For example, if the "k-1" output for the "i" bit is a likelihood expressed as a value between 0 and 1, 0.5 is input as each of the three values: the "k-1" output for the "i-1" bit, the "k-1" output for the "i" bit, and the "k-1" output for the "i+1" bit.

[0079] For example, in the information processing device of the reference example, the output (waveform) of the equalizer 31 is processed by PRML (Partial Response Maximum Likelihood). For example, the waveform is processed based on a PR model. Furthermore, it is processed by a Viterbi algorithm (for example, SOVA: soft output Viterbi algorithm) to perform "0 / 1 decision."

[0080] In the embodiment, for example, "0 / 1 decision" based on a neural network is performed instead of PRML. By using an appropriate processing model (neural network), processing accuracy can be improved.

[0081] (Second embodiment) The second embodiment relates to an information recording and reproducing device 210 (see FIG. 1). The information recording and reproducing device 210 includes the information processing device according to the first embodiment and a recording unit 80D. The information recording and reproducing device 210 may include a reproducing unit 80R. The reproducing unit 80R is capable of reproducing information I00 recorded in the recording unit 80D. The reproducing unit 80R may be capable of reproducing the first signal S1 and attribute P0 recorded in the recording unit 80D. The reproducing unit 80R may include a magnetic head 80H.

[0082] A magnetic recording and reproducing system 310 according to the embodiment (see FIG. 1) includes the information processing device according to the first and second embodiments. The magnetic recording and reproducing system 310 may further include, for example, a recording unit 80D. Multiple elements included in the magnetic recording and reproducing system 310 (for example, an acquisition unit 71 and a processing unit 72) may be provided in different locations. Information may be transmitted and received using any communication method. For example, multiple parts included in the processing unit 72 (for example, multiple neural networks) may be provided in different locations.

[0083] The embodiment may include a program. The program causes a computer (information processing device 70) to perform the first to fourth operations OP1 to OP4. The embodiment may include a storage medium on which the program is stored.

[0084] The embodiment may include the following configurations (for example, technical solutions). (Configuration 1) an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit is capable of deriving a first output obtained by processing first information including the first signal using a first processing model, and a second output obtained by processing the first information using a second processing model; The information processing device, wherein the processing unit is capable of outputting a result of processing the first information based on the first output, the second output, and a third output obtained based on the first information.

[0085] (Configuration 2) 2. The information processing device according to configuration 1, wherein the result includes information regarding whether the first signal corresponds to a first value or whether the first signal corresponds to a second value different from the first value.

[0086] (Configuration 3) the first processing model includes a first neural network that is machine-learned based on a plurality of first training data including record information recorded with a first attribute; The information processing device according to configuration 1 or 2, wherein the second processing model includes a second neural network machine-learned based on a plurality of second training data including recorded information recorded with a second attribute.

[0087] (Configuration 4) an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit is capable of selecting a first processing model from a plurality of processing models based on first information including the first signal; The information processing device, wherein the processing unit is capable of outputting a result of processing the first information based on the first processing model.

[0088] (Configuration 5) 5. The information processing device according to configuration 4, wherein the result includes information regarding whether the first signal corresponds to a first value or whether the first signal corresponds to a second value different from the first value.

[0089] (Configuration 6) the plurality of processing models includes an m-th processing model, The m is an integer of 1 or more and N or less, N is an integer of 2 or more, 6. The information processing device according to configuration 4 or 5, wherein the mth processing model includes an mth neural network that has been machine-trained based on a plurality of training data including recorded information recorded with an mth attribute.

[0090] (Configuration 7) 7. The information processing device according to configuration 6, wherein the mth attribute relates to a recording density of the recording medium.

[0091] (Configuration 8) an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit is capable of outputting a result of processing first information including the first signal; the result is obtained by processing the first information with a first processing model corresponding to the first attribute when the attribute related to the first signal is a first attribute; An information processing device wherein, when the attribute is a second attribute different from the first attribute, the result is obtained by processing the first information using a second processing model corresponding to the second attribute, and the second processing model is different from the first processing model.

[0092] (Configuration 9) if the attribute is the first attribute, the result includes information derived based on a first output corresponding to the first attribute regarding whether the first signal corresponds to a first value or whether the first signal corresponds to a second value different from the first value; 9. The information processing device of claim 8, wherein, when the attribute is the second attribute, the result includes information derived based on a second output corresponding to the second attribute regarding whether the first signal corresponds to the first value or the first signal corresponds to the different second value.

[0093] (Configuration 10) the first processing model includes a first neural network that is machine-learned based on a plurality of first training data including record information recorded with a first attribute; The information processing device of configuration 8 or 9, wherein the second processing model includes a second neural network machine-trained based on a plurality of second training data including recorded information recorded with a second attribute.

[0094] (Configuration 11) an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit processes first information including the first signal using a first processing model corresponding to a first attribute, and processes the first information using a second processing model corresponding to a second attribute, the second processing model being different from the first processing model; The information processing device, wherein the processing unit is capable of outputting the results of processing using the first processing model when the attribute related to the first signal is the first attribute, and outputting the results of processing using the second processing model when the attribute is the second attribute.

[0095] (Configuration 12) if the attribute is the first attribute, the result includes information derived based on a first output corresponding to the first attribute regarding whether the first signal corresponds to a first value or whether the first signal corresponds to a second value different from the first value; 12. The information processing device of claim 11, wherein, when the attribute is the second attribute, the result includes information derived based on a second output corresponding to the second attribute regarding whether the first signal corresponds to the first value or the first signal corresponds to the different second value.

[0096] (Configuration 13) the first processing model includes a first neural network that is machine-learned based on a plurality of first training data including record information recorded with a first attribute; An information processing device according to configuration 11 or 12, wherein the second processing model includes a second neural network machine-trained based on a plurality of second training data including recorded information recorded with a second attribute.

[0097] (Configuration 14) 14. The information processing device according to any one of configurations 8 to 13, wherein the attribute relates to a recording density of the recording medium.

[0098] (Configuration 15) 15. The information processing device according to any one of configurations 1 to 14, wherein the first signal is obtained by performing equalization processing on at least a part of the reproduced signal.

[0099] (Configuration 16) 16. The information processing device according to configuration 15, wherein the processing unit is capable of outputting second information obtained by performing error correction processing on the result.

[0100] (Configuration 17) 16. The information processing device according to configuration 15, wherein the first information further includes second information obtained by performing error correction processing on the result.

[0101] (Configuration 18) An information processing device according to any one of configurations 1 to 17; The recording unit; a reproducing unit capable of reproducing the information recorded in the recording unit; A magnetic recording and reproducing device comprising:

[0102] (Configuration 19) 19. The magnetic recording and reproducing apparatus according to configuration 18, further comprising a magnetic head including the reproducing section.

[0103] (Configuration 20) 18. A magnetic recording and reproducing system comprising the information processing device according to any one of configurations 1 to 17.

[0104] (Configuration 21) 21. The magnetic recording and reproducing system according to configuration 20, further comprising the recording unit.

[0105] According to the embodiments, it is possible to provide an information processing device, a magnetic recording and reproducing device, and a magnetic recording and reproducing system that are capable of improving processing accuracy.

[0106] The embodiments of the present invention have been described above with reference to examples. However, the present invention is not limited to these examples. For example, the specific configurations of the control unit, processing unit, recording unit, and other elements included in the information processing device, magnetic recording and reproducing device, and magnetic recording and reproducing system are within the scope of the present invention as long as a person skilled in the art can implement the present invention in a similar manner and obtain similar effects by appropriately selecting them from within the known range.

[0107] Any combination of two or more elements of each example within the scope of technical feasibility is also included within the scope of the present invention as long as it encompasses the gist of the present invention.

[0108] All information processing devices, magnetic recording and reproducing devices, and magnetic recording and reproducing systems that can be implemented by a person skilled in the art by making appropriate design modifications based on the information processing devices, magnetic recording and reproducing devices, and magnetic recording and reproducing systems described above as embodiments of the present invention also fall within the scope of the present invention, as long as they include the gist of the present invention.

[0109] Within the scope of the concept of the present invention, a person skilled in the art may conceive of various modifications and alterations, and it is understood that these modifications and alterations also fall within the scope of the present invention.

[0110] Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are included within the scope and spirit of the invention, and are also included in the scope of the invention and its equivalents as defined in the claims. [Explanation of symbols]

[0111] 31...equalizer, 32...error corrector, 40...processing model, 45...likelihood, 70...information processing device, 71...acquisition unit, 72...processing unit, 80...recording medium, 80D...recording unit, 80H...magnetic head, 80R...reproduction unit, 210...information recording and reproduction device, 310...magnetic recording and reproduction system, I01, I02...first and second information, NN1 to NN4...first to fourth neural networks, OP1 to OP4...first to fourth operations, P0...attribute, R1...result, S1...first signal, Sr1...reproduction signal

Claims

1. an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit is capable of deriving a first output obtained by processing first information including the first signal using a first processing model, and a second output obtained by processing the first information using a second processing model; The information processing device, wherein the processing unit is capable of outputting a result of processing the first information based on the first output, the second output, and a third output obtained based on the first information.

2. The information processing device according to claim 1 , wherein the result includes information regarding whether the first signal corresponds to a first value or whether the first signal corresponds to a second value different from the first value.

3. the first processing model includes a first neural network that is machine-trained based on a plurality of first teacher data including record information recorded with a first attribute; The information processing device according to claim 1 , wherein the second processing model includes a second neural network that is machine-trained based on a plurality of second teacher data including recorded information recorded with a second attribute.

4. an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit is capable of outputting a result of processing first information including the first signal, the result is obtained by processing the first information with a first processing model corresponding to a first attribute when the attribute related to the first signal is a first attribute; the result is obtained by processing the first information with a second processing model corresponding to the second attribute when the attribute is a second attribute different from the first attribute, the second processing model being different from the first processing model; The information processing device, wherein the attribute relates to the recording density of the recording medium.

5. an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit processes first information including the first signal using a first processing model corresponding to a first attribute, and processes the first information using a second processing model corresponding to a second attribute, the second processing model being different from the first processing model; the processing unit is capable of outputting a result of processing using the first processing model when an attribute related to the first signal is the first attribute, and is capable of outputting a result of processing using the second processing model when the attribute is the second attribute; The information processing device, wherein the attribute relates to the recording density of the recording medium.

6. 6. The information processing device according to claim 1, wherein the first information further includes second information obtained by performing error correction processing on the result.

7. an acquisition unit; a processing unit; Equipped with the acquisition unit is capable of acquiring a reproduction signal obtained from a recording unit including a recording medium, the reproduction signal including a first signal corresponding to information recorded on the recording medium; the processing unit is capable of selecting a first processing model from a plurality of processing models based on first information including the first signal; the processing unit is capable of outputting a result of processing the first information based on the first processing model; The information processing device, wherein the first information further includes second information obtained by performing error correction processing on the result.

8. An information processing device according to any one of claims 1 to 7; The recording unit; a reproducing unit capable of reproducing the information recorded in the recording unit; A magnetic recording and reproducing device comprising:

9. A magnetic recording and reproducing system comprising the information processing device according to any one of claims 1 to 7.

Images