JPS6077531A - Method for preventing erroneous detection and mis-correction - Google Patents

Abstract

PURPOSE:To prevent generation of mis-correction at overlooked CRCC by using three error pointers so as to attain correction. CONSTITUTION:Arrangement of signal is applied to a signal train obtained by converting an analog signal into a digital signal so as to divide the signals in the form of matrix. P parity and Q parity are generated in the column direction in one set of divided signal groups. Then an error generated at recording or reproduction is detected by the CRCC at each frame. In Fig., marks ''X'' indicate words of the frame where an error is detected and marks ''0'' show words of the frame where no error is detected. The 1st error pointer 1 is set to the words of the marks ''X''. Then the error detection using the Q parity is applied, and when a word is regarded as an error words, the 2nd error pointer 2 is set to all the words of the series. Moreover, when a word is regarded as an error word in the Q parity, the 3rd error pointer 3 is set to all the words where the 1st error pointer is not set. When the 3rd pointer 3 is set to words in the frame where no pointer 2 is set, the state is restored.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Use The present invention relates to a method for error detection and correction prevention in an apparatus that converts an analog signal into a digital signal and records and reproduces it on a recording medium.

Configuration of conventional example and its problems After converting an analog signal into a digital signal, this digital signal series D1. D2. D3...mark is divided into %xN words, and the signals are rearranged (interleaved) independently for each word, and further formed into an MxN matrix as shown in FIG. Here, according to a certain rule, one word signal is taken out from each row to preform P parity and added to the bottom row. Furthermore, one word signal is taken out from each row including the P parity according to a different rule from the one used when generating the P parity, and Q parity is generated and added below the P parity row. To simplify the explanation, we will take P parity sequences at regular intervals as shown in Figure 2 to generate P parity, and the truncated Q parity sequences will have a constant length L (K%L). ) to generate Q parity. Therefore, (M+2) signals in the vertical direction are made into a set of frames, a synchronization signal is added to the beginning, and a shortened cyclic code (CRCC) is generated from the signals in these frames and added to the end of the frame. , frames are recorded on the recording medium in the order of the first column frame, the second column frame, . . . , the N-th column frame. A signal reproduced from a recording medium contains errors due to scratches, dust, etc. on the recording medium, and therefore errors are detected in frame units by CRCC. The polynomial used in CRCC is 1+)C×
There is a 16th order of +X, which is 1 in the frame.
All errors of up to 6 bits can be detected, and errors of 17 bits or more can also be detected except for specific error patterns. In a frame in which an error is detected, error pointers are set in all words No. 4B, and this error information is sent to the sub-pressure circuit, where correction is repeatedly performed using two series of parities.

If an error of 17 bits or more occurs in the frame, C
The probability that an error detection error will occur by RCC is 1/216
≠ 1 + 6 × 10, and the error pointer does not hit the word where the error pointer should go, so it is considered correct. In the correction circuit, this word may correct other error words, and the two series of parity As a result of repeated corrections, the number of words that are incorrectly corrected increases. For example, if an audio signal is converted into a digital signal and recorded and reproduced, abnormal noises such as clicks appear in the reproduced signal, which becomes an extremely important problem.

Purpose of the Invention The present invention solves the above-mentioned problems by performing error detection using parity after error detection using CRCC and before starting a correction operation, and creating a new error pointer based on the detection result. By generating and sending this error information to the correction circuit, error detection errors in CRCC are prevented.
Error detection and
This document provides information on how to prevent erroneous corrections.

Structure of the Invention The present invention converts an analog signal into a digital signal, adds a blank series error correction code (P parity, Q parity),
A shortened cyclic code (C
This is a device that generates a synchronization signal (RCC) at the beginning of a frame and adds a CRCC to the end of the frame and records and reproduces it on a recording medium.During reproduction, it detects errors in each frame using the CRCC and collects the first error pointer information. and means for generating a second error pointer and a third error pointer using the first parity and the first error pointer information, By performing correction using the second and third error pointers, it is possible to detect missed error detection by CRCC and prevent erroneous correction in the correction circuit.

Description of the Embodiment First, as in the conventional example, an 8-bit signal sequence D1, D2, . D3.
The signals are rearranged (interleaved) and further divided into an M x N matrix as shown in FIG. Within this divided set of signals +711 signals, as shown in Figure 2, P is transmitted at regular intervals in opposite directions.
A parity sequence is taken, a P parity is generated, a Q parity sequence is taken at regular intervals in the layer direction, and the Q parity, including the P parity words, is made into a (M+2)×N matrix. Here, for each column, a CRCC generated from the signals of the entire column including the synchronization signal at the beginning and the synchronization signal at the end is added to form a frame, which is sequentially recorded on the recording medium frame by frame. An embodiment in which the error detection and error correction prevention method of the present invention is applied to a signal configured in the above format will be described below.

FIG. 3 shows an example in which errors occurring during recording or reproduction due to scratches, dust, etc. on the recording medium are detected for each frame by CRCC in the case of M=7. 16 at CRCC
Use the following generator polynomial.

1'2' frames are constructed in the vertical direction (w'w..., w7.p, o), and in the figure, the X mark indicates the word of the frame in which an error was detected, and the ○ mark indicates that no error was detected. For simplicity, the frames of the Q parity sequence are shown horizontally, but in reality there are K(L-1) frames between each frame. A first error pointer is placed on the word marked with an X to distinguish the correct word. Next, before starting the correction operation, error detection is performed using Q parity. In Q parity error detection, only when all the words in the Q parity sequence are correct and are considered to be errors in the Q parity, it is assumed that a false detection error by CRCC has occurred, and all words in that sequence are A second error pointer is set, and when the first error pointer is set for at least one word in the Q parity series and the Q parity is considered to be an error, the A third error pointer is set for all words that do not have a first error pointer. In FIG. 3, if a sequence deemed to be erroneous based on the Q parity error detection result is indicated by a Δ mark, the error pointer will eventually become as shown in FIG. 4. In addition, in the figure, 1
, 2, 3 are the first, second . It shows the word on which the third error pointer is pointing. Since the second error pointer/unfilled frame has extremely low fjJ probability of detection error by CRCC, the third error pointer is
If the word to which the error pointer is pointing does not exist,
The third error pointer is restored. When the above processing is performed, FIG. 4 becomes as shown in FIG. 5. The final remaining 2nd. The word the third error pointer is pointing to is the CRC
This is sent to the correction circuit along with the error pointer of Chiku 1, as there is a high possibility that this is a detection error caused by C. In the correction circuit, correction using P parity and correction using Q parity are sequentially and repeatedly performed.

At this time, the word on which the second error pointer is pointing is P
After the parity becomes correctable and the value remains unchanged from the value before correction, it is determined that there is no CRCC detection error in this frame, and the error pointer in the other word in the frame is returned to its original value. When there are words with two or more box 2 error pointers in one frame, the pointers in other words are restored only when the values of both words remain unchanged even after correction. In this way, even during correction, frames for which the possibility of detection errors by CRCC is low are determined to be error-free, thereby increasing the ability to detect error detection errors by CRCC and preventing a decrease in correction ability.

In addition, in the present invention, instead of error detection using Q parity, P
It can also be applied to a method of error detection using parity. In this case, the correction ability will be higher if the correction circuit repeatedly performs the Q parity correction and then the P parity correction in this order. Furthermore, in the embodiment, the P parity series is
Although the Q parity series is configured with L and uniform intervals, it may be configured with irregular intervals.

Effects of the Invention In the present invention, in addition to error detection using CRCC, error detection is performed using parity to improve the detection ability for CRCC detection errors. By introducing two error pointers, we actively consider words for which there is a low possibility of a CRCC detection error as the correct answer, thereby increasing the possibility of a detection error even during normal operation. If it is considered to be low, the answer is correct and the correction ability is reduced by 1C.
It is possible to improve the ability to detect missing RCCs and to minimize the occurrence of erroneous corrections when missing CRCCs.

[Brief explanation of the drawing]

Figure 1 explains the formatting of digital signals/
Figure 2 is a diagram to show the parity series, Figure 3 is a diagram to show the parity series.
The figure shows an example of error detection using CRCC, FIG. 4 shows the process of detecting a CRCC detection error to explain an embodiment of the present invention, and FIG. It is a figure which shows the detection process of a detection error of CRCC similarly. Name of agent: Patent attorney Toshio Nakao (1st person)
Figure 2

Claims (1)

[Claims] After converting an analog signal to a digital signal and generating two sequences of error correction codes from multiple words, a frame is constructed from multiple words including the error correction code and different from the parity sequence, and a synchronization code is placed at the beginning of the frame. After adding a shortened cyclic code obtained from the signal within the frame to the end of the signal and frame, when recording and playing back on a recording medium, error detection is performed for each frame using the shortened cyclic code during playback, and errors are detected. means to set a first error pointer in every word in a frame if the first error pointer is set in every word in a parity sequence; means to set a second error pointer in every word of the parity series, and the word to which the second error pointer is set belongs to 71/-)-in/7''IMj
If the first error pointer is on at least one word in the parity series of the second word in word , and an error due to parity is confirmed, the third error pointer is applied only to the word in the frame. After the word that the second error pointer is pointing to is corrected, if the value remains the same as before the correction, the word that the second error pointer is pointing to in that frame is changed to another. If the second error pointer does not exist, all error pointers of each word in that frame are restored, and if there are words with two or more second error pointers in the same frame, after both are corrected, Restore all error pointers for each word in the frame only when the value remains the same as before correction! Error in assuming that it is a 1h characteristic (
How to prevent minor errors and incorrect corrections.