JP2863168B2 - Error detection method - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an error detection method for decoding error-correction-coded data, and particularly to a method in which a burst error or the like occurs during data recording / reproduction or transmission. The present invention relates to an error detection method for preventing an omission of error detection in a case.

[Summary of the Invention]

The present invention relates to an error detection method for decoding digital data encoded by two or more error correction code sequences, according to a state of an error flag obtained at the time of decoding based on one error correction code sequence. Therefore, at the time of decoding based on the next other error correction code sequence, by making the data of the other error correction code sequence an error, leakage of error detection is greatly reduced, and adverse effects due to detection leakage are prevented. is there.

[Conventional technology]

Digital signal processing for performing error correction using two or more error correction code sequences is known in so-called CDs (compact disks), DATs (digital audio tape recorders), and the like. For example, in the case of a CD, as a CIRC (Cross-Interleaved Reed-Solomon code), a C1 sequence and a C2 sequence, which are a fourth-order Reed-Solomon code sequence, are combined via an interleave and are used. First encodes the C2 code, performs an interleave process, and then encodes the C1 code. In more detail, the interleaving process is performed shallowly before the C2 encoding and after the C1 encoding, respectively, but the description is omitted in this specification for the purpose of simplifying the description. In addition, the decoding process is performed in the order of the decoding of the C1 code, the de-interleaving, and the decoding of the C2 code, contrary to the order of the above-described encoding.

In an error correction process using two error correction code sequences of C1 and C2, an error flag (a flag indicating a possibility of an error) formed by decoding the C1 code is usually used. Error when decoding the next C2 code
Used as a pointer. Depending on the number of error flags at this time, one of the error correction, erasure correction, and uncorrectable processing modes is selected when decoding the C2 code. In such a case, the error flag is left as it is. Then, even if the decoding of the C2 code is completed, data for which the error flag is set is restored as uncorrectable data by interpolation processing or the like. This is because, if an erroneous correction is made, the reproduced signal is adversely affected by pulse noise or the like, so that error detection is more important than error correction.

[Problems to be solved by the invention]

By the way, when a so-called burst error occurs, in which many errors occur continuously due to a defect of a recording medium, adhesion of dust, or various fluctuations in a recording / reproducing process, etc.
Even if the above-described error detection is weighted, there remains a possibility that error detection cannot be performed effectively.

That is, as shown in FIG. 3, the decoding of each check word P of the C1 sequence is first performed, so that all check words P _{1 to} P _{n of} the burst error generator BE may have an error possibility. Should be set (to "1"), but one check word _Pk in these sequences indicates a predetermined error pattern and is erroneously recognized as no error. In this case, no error flag F is set for this series of check words P _k (“0”).
Becomes). At the time of decoding the C2 sequence shown in FIG. 3, the error correction flag F is often set to "1", so that the above-mentioned uncorrectable mode is selected, and the decoding process is performed in such a manner that the flag F remains for each data. Is completed, and the data D ^{* at} the intersection of the _Pk series and the C2 series is treated as correct data.

Here, considering the C1 sequence, which is the fourth-order Reed-Solomon code sequence in the case of the CD, when all the symbols (one symbol is 8-bit data) are in error due to the occurrence of a burst error, Error based on 4 symbols of C1 sequence check word (P parity)
The probability that the syndrome will be mistaken for no error by showing a predetermined error pattern is 1 / (2 ⁸ ) ⁴ = 1/2 ³² ≒ 2.33 × 10 ^-10 . Similarly, the probability that one-symbol error correction (erroneous correction) occurs is 32 · (2 ⁸ −1) / 2 ³² ≒ 1.90 × 10 ⁻⁶ , and the probability that two-symbol error correction (erroneous correction) occurs. increases the ^{_{32 C 2 · (2 8 -1}} ) 2/2 32 ≒ 7.51 × 10 -3.

The present invention has been made in view of such circumstances, and prevents a large number of errors from occurring when a burst error occurs and prevents the data from being erroneously recognized as correct data despite being erroneous. It is an object of the present invention to provide an error detection method that can greatly reduce the omission of detection of the error.

[Means for solving the problem]

An error detection method according to the present invention is directed to an error detection method for decoding digital data encoded by two or more error correction code sequences in order to solve the above-described problem. Setting an error flag indicating a possibility of an error at the time of decoding based on one of the sequences, and setting an error flag at the time of decoding based on the one sequence at the time of decoding based on another sequence of the error correction code sequence. A step of determining whether or not a burst error has occurred based on the flag; and, if it is determined that the burst error has occurred, the range of the other series in which the burst error has been determined to have occurred Setting an error flag for the data for which the error flag is not set.

Another feature of the error detection method according to the present invention is that in the error detection method for decoding digital data encoded by two or more error correction code sequences, Setting an error flag indicating the possibility of an error at the time of decoding based on, and at the time of decoding based on another sequence of the error correction code sequence, the error flag set at the time of decoding based on the one sequence is continuously The step of determining that a burst error has occurred when a predetermined number or more has occurred, and the step of determining that the burst error has occurred when determining that the burst error has occurred, the error flag is determined to be continuous. Setting an error flag for the data before and after the data group in the set range.

[Operation] When a burst error has occurred, at the time of decoding one error correction code sequence prior to the decoding of the other error correction code sequence, for example, despite the occurrence of an error, a no error pattern may be generated. Even if an error cannot be detected by accident, the possibility of a burst error is estimated by estimating the possibility of a burst error based on the number of error flags and the flag pattern obtained when decoding the one sequence. An error flag can be set for the leaked data, and it is possible to effectively prevent the adverse effect of the error detection omission.

〔Example〕

Hereinafter, as a preferred embodiment of the error detection method according to the present invention, error detection at the time of decoding an error correction code when reproducing a so-called CD (compact disk) will be described with reference to the drawings. In the decoding process in this case, as described above, decoding is first performed using four symbols (one symbol is 8-bit data) serving as a C1 sequence check word (P parity), and then the C2 sequence is decoded. The decoding is performed using four symbols of the check word (Q parity).

First, FIG. 1 is a flowchart for explaining an embodiment of an error detection method according to the present invention. When decoding the C2 sequence, a flag indicating the possibility of an error obtained by decoding the C1 sequence ( estimating whether a burst error has occurred in accordance with the number N _F of the so-called error flag) is set, it is made different decoding process. C2 in FIG.
Prior to the description of the series decoding process, referring to FIG.
The schematic configuration and operation of the reproduction system of the CD player will be described.

In FIG. 2, an optical disk 1 which is a so-called CD
The recording signal is read by the optical pickup 3 while being rotated by the spindle motor 2. The reproduced signal from the optical pickup 3 is sent to the EFM demodulator 5 via the amplifier and the PLL circuit unit 4 and demodulated. The reproduced signal is first transmitted to the C1 sequence decoder 6 and then to the C2 sequence decoder 7.
To be decoded in this order. Here, a flag is set at the time of decoding each code sequence, and each of the flag setting circuits 8 and 9 is set.
The flag from the circuit 8 stored in the memory or the like and set at the time of decoding the C1 sequence is sent to the decoder 7 of the C2 sequence.
The decoded data from the C2 sequence decoder 7 and the flag from the flag setting circuit 9 are provided by a de-interleave and interpolation circuit unit.
The data is then sent to 10 where the data is deinterleaved and interpolated. The data to be interpolated at this time is determined by the flag. The digital audio data signal from the de-interleave and interpolation circuit unit 10 is sent to a D / A converter 11 to be converted into an analog signal, and is taken out via an audio amplifier 12.

Also, a control signal from the system controller 13 in FIG. 2, a clock signal from the amplifier and the PLL circuit section 4 and the like are sent to the drive control circuit 14, and the drive control circuit 14 controls the rotation of the spindle motor 2 and the like. Focus / tracking control of the optical pickup 3 is performed.

Next, a specific example of the decoding process in the C2 sequence decoder 7 in FIG. 2 will be described with reference to FIG.

First, the start of C2 decoding in FIG. 1 refers to the C2 sequence check word (four symbol symbols) for a data group that has been decoded using the C1 code sequence check word (four symbols of P parity). (Q parity). In step S1 after the start of C2 decoding, a so-called syndrome calculation using the Q parity of the four symbols is performed, and an error (error) on the C2 sequence is calculated.
Seeking the number N _S of the symbol. And by varying the following process in response to the error symbol number N _S, the steps
In S2, S3 and S4, N _s = 0, N _s = 1 and N _{s respectively}
= 2 is determined.

N _S = 0 at step S2, i.e. the process proceeds to step S5 when it is judged no error, the number N _F of the error flag is discriminated whether larger than the predetermined value a ₀ (a _0, for example 8) When YES (N _F > a ₀ ), error flags are set for all symbols on the C2 sequence, and the decoding process is terminated. This corresponds to the assumption that a so-called burst error has occurred as described above,
In FIG. 1, all errors are described. When NO is determined in step S5, the error flags for all symbols on the C2 sequence are cleared. This is called flag clear in FIG.

N _S = 1 when it is judged (1 symbol errors) and in step S3, the process proceeds to step S6. This step
In S6, the positions of the error symbols obtained based on the syndromes of the C2 sequence are compared with the positions of the symbols with the error flags obtained at the time of the C1 decoding, so that these positions match. The number N _EE of symbols is obtained, and it is determined whether or not the number N _EE is 1. N at the time of _EE = 1 proceeds to step S7, whether the number N _F of the error flag is larger than the predetermined value a ₁ (a ₁
Performs the determination of 4), for example. If it is determined in step S7 that YES (N _F > a ₁ ), it means that the above-mentioned burst error has been estimated, and the C2
An error flag is set for all symbols in the sequence (all errors). If NO, the error flags for all the symbols on the C2 sequence are cleared (flag clear). In this case, one symbol correction is performed by C2 decoding. When it is judged NO at step S6 (N _EE ≠ 1), the process proceeds to step S8, the number N _F of a predetermined value a ₂ (e.g., 3) of the error flag is larger than (N _F>
a ₂ ) It is determined whether or not the error flag is set, and an error flag is set for all the symbols with YES (that is, all errors are determined by estimating a burst error), and an error flag is cleared with NO (flag clear). doing. This flag
At the time of clearing, one symbol correction is performed by C2 decoding.

Next, when the N _S = is judged 2 (2 symbol errors) in the step S4, the process proceeds to step S10 and subsequent steps. In steps S10 and S11, the number N _EE of symbols in which the position of the error symbol obtained during the C2 decoding and the position of the symbol with the error flag obtained at the time of the C1 decoding match are determined. Has been determined,
In step S10, it is determined whether or not N _EE = 0, and in step S11, N _{EE is determined.}
= 1 or not.

When it is judged YES (N _EE = 0) at step S10 advances to step S12, the number N _F of the error flag is a predetermined value a ₃ (eg 2) or less (N _F ≦ a ₃₎ whether the If the determination is NO, the process proceeds to step S13, and if the determination is YES, the process proceeds to step S14. Step S13, the N _F is performed above atmospheric (N _F> a ₄₎ whether determination predetermined value a ₄ (eg 4), and the all-error YES, the for all the symbols on the C2 sequence in NO The error flag is used as it is (flag copy). Step In S14 the N _F performs a predetermined value a ₁₀ (for example, 0) or less (N _F ≦ a ₁₀₎ whether the discrimination, and the all-error when NO, the
In the case of YES, the flag is cleared. When this flag is cleared, two-symbol error correction by C2 decoding is made valid.

If NO is determined in step S10, the process proceeds to step S11. If YES ( _NEE = 1) is determined in step S11, the process proceeds to step S15. In this step S15, the above N _F
There when YES, the process proceeds to a predetermined value a ₅ (e.g. 3) or less (N _F ≦ a ₅₎ whether the performed determination, step S16 when NO is proceeding to step S17. Step S16 In the N _F is performed above atmospheric (N _F> a ₆₎ determination of whether or not a predetermined value a ₆ (e.g. 4), and the all-error when YES, the case of NO above
The error flags for all symbols on the C2 sequence are used as they are (flag copy). In step S17 the N _F performs a predetermined value a ₁₁ (for example, 0) or less (N _F ≦ a ₁₁₎ whether the discrimination, and the all-error when the NO, YES
In the case of, the flag is cleared (two-symbol error correction).

If N _EE = 2, NO is determined in the step S11, and the process proceeds to a step S18. In the step S18, the N _F is performed above atmospheric (N _F> a ₇₎ judged whether or not a predetermined value a ₇ (eg 4), and the all-error when YES, the aforementioned flag copy when NO doing.

Next, when the number of error symbols _NS is three, the determination in step S4 is NO, and the process proceeds to step S20. Step S
In 20, the number of error flags N _F is the predetermined value a ₈ (e.g. 2) or less performed (N _F ≦ a ₈₎ whether the discrimination, if NO the process proceeds to step S21, the C2 sequence in case of YES Error flags are set for all the above symbols (all errors). In step S21 the N _F is the predetermined value a ₉ (e.g. 4)
It is determined whether the value is larger (N _F > a ₉ ). If YES, the above-mentioned all errors are detected. If NO, the error flags for all the symbols on the C2 sequence are used as they are (flag copy). .

Each of the predetermined values a _{1 to} a ₁₁ in the above-described embodiment may be set to an arbitrary value in consideration of a burst error occurrence state and the like. The main processing of C2 decoding is step
It may be performed at the time of S1, but alternatively, only the syndrome calculation may be performed at the time of step S1, and the subsequent calculations may be performed when necessary according to the conditions.

According to such an embodiment of the present invention, at the time of a burst error or the like, there is a C1 sequence for which no error has been set in the C1 decoding and no error flag has been set even though an error has occurred. even if, in the C2 decoding, depending on the state of the error flag for all symbols of the C2 series is set to be currently decoded, in particular number N _F flag is standing is larger than a predetermined number (e.g. 4) At times, it is assumed that a burst error has occurred, and an error flag is set for all symbols of the C2 sequence (all errors). Therefore, it is possible to more reliably prevent the error detection from being omitted. When applied to error detection during recording and reproduction of computer data such as a so-called CD-ROM or the like, it can withstand use even in a bad environment.

Further, in the above embodiment, the number of the number N _F and, C2 sequence of error symbols standing of the error flag
Depending on the combination of the N _S, because then divided if finer C2 decoding process, the decoding process of the optimum is performed in each case.

It should be noted that the present invention is not limited to only the above embodiment. That is, for example, in the above-described embodiment, C2
During decoding, referring to the error flag which each symbol is obtained in C1 when decoding for each C1 sequence belonging on C2 sequence to be currently decoded, focusing only on the number N _F of the flag is set Although the C2 decoding is performed in this manner, the arrangement pattern of these error flags may be taken into consideration.
For example, when an error flag is set for symbols before and after one symbol on the C2 sequence to be decoded, an error plug for the one symbol may be set, or a predetermined number (for example, 4 If an error flag is set for a group of symbols that are continuous as described above, the error flags of the symbols before and after this group of symbols are set. As for the decoding processing according to the pattern of the error flag, similarly to the above embodiment,
C2 in combination with the error symbol number N _S series, may be more finely case analysis. Furthermore, if the occurrence of a burst error can be detected by lowering the level of the reproduced RF signal or the like, the correction capability does not deteriorate much. In addition, it goes without saying that various changes can be made without departing from the spirit of the present invention.

〔The invention's effect〕

According to the error detection method of the present invention, burst decoding is performed in accordance with the number of flag flags and the flag pattern obtained when decoding one error correction code sequence prior to decoding another error correction code sequence. Estimate the possibility that an error has occurred, and check the data for which the error flag is not set among all the data of the other error correction code sequence, or before and after the data group in the range where the continuous error occurs. The data is processed so that an error flag is set. For example, due to the occurrence of a burst error, no error pattern is accidentally generated despite the occurrence of an error and cannot be detected. In this case, it is possible to set an error flag for the data of the detection omission, and it is possible to effectively prevent the adverse effect of the error detection omission. That. In addition, even if it is determined that a burst error has occurred, an error flag is not set indefinitely, but is not limited to data having an error flag set within the range of the other error correction code sequence, or An error flag is set for the data before and after the data group in the range where a continuous error has occurred, preventing the problem that an error flag is set even for a part where no burst error has occurred. And an optimal decoding process can be performed.

In particular, when the present invention is applied to error detection for a computer data recording / reproducing apparatus, it can withstand use even in a bad environment.

[Brief description of the drawings]

FIG. 1 is a flowchart for explaining an embodiment of an error detection method according to the present invention, FIG. 2 is a block circuit diagram showing a schematic configuration of a signal reproducing system to which the embodiment is applied, and FIG. FIG. 4 is a schematic diagram for explaining an example of a correction method. 6 C1 sequence decoding circuit 7 C2 sequence decoding circuit 8, 9 Flag setting circuit 10 De-interleave and interpolation circuit section

Claims (2)

(57) [Claims] 1. An error detecting method for decoding digital data encoded by two or more error correction code sequences, the method comprising: indicating a possibility of an error when decoding based on one of the error correction code sequences. Setting an error flag, and, at the time of decoding based on another sequence of the error correction code sequence, determining whether a burst error has occurred based on the error flag set at the time of decoding based on the one sequence. When it is determined that the burst error has occurred,
Setting an error flag for data in which the error flag has not been set within the range of the other series determined to have the burst error. 2. An error detection method for decoding digital data encoded by two or more error correction code sequences, the method comprising: indicating a possibility of an error when decoding based on one of the error correction code sequences. Setting an error flag, and, when decoding based on another sequence of the error correction code sequence, a burst error occurs when a predetermined number or more of error flags set during decoding based on the one sequence occur continuously. Determining that a burst error has occurred; and determining that the burst error has occurred,
Setting an error flag for data before and after a data group in a range in which the error flag determined to have the burst error has been continuously set. Method.