KR20100040444A - Method of distributively recording and reading data on and from optical disc using error correction codes - Google Patents

Abstract

PURPOSE: A distributed data recording method for an optical medium using ECC is provided to restore damaged data with further stored codeword if unrecoverable data is damaged. CONSTITUTION: A distributed data recording method for an optical medium comprises the following steps. The information on recoverable damage levels are input by a user. Codewords are generated applying ECC(error correction code) to the original data to be stored in an optical disc. When the generated codewords are actually written to the disc, the sequence of blocks is changed in order to enable blocks dependent on the same codeword to be distributed to a wide area and not consecutively arranged.

Description

Method of distributively recording and reading data on and from optical disc using error correction codes}

The present invention relates to a method for providing the reliability required for securely storing data on physically susceptible optical discs (e.g., CD-ROM, DVD, Blu-ray, etc.), and more particularly on optical media. The present invention relates to a distributed recording and reproducing method of data using an error correcting code.

Unlike a hard disk, an optical disk has a disk on which data is stored directly exposed to the outside. Therefore, internal data can be easily lost due to scratches, external shocks or contamination. To prevent this, the optical disc uses an error-correcting code (ECC) on its own.

When data is recorded in the optical disk, an ECC block is generated through an error correction code algorithm in units of one sector or several sectors. In this block, parity data is added in addition to the original data. This parity data is used to recover if some of the original data is lost or changed. When the ECC blocks are actually written on the optical disc, they rotate from the inside of the disc and write the ECC blocks one by one in the helical direction. Unlike hard disks that record data in the form of concentric circles, data is continuously arranged in a spiral form. Therefore, unlike the definition of a track used in a hard disk (one concentric circle for recording data), a track in an optical disk defines a 360-degree rotation in a spiral as a track.

Optical discs differ in the degree of data loss depending on the direction of the scratch. Scratches in the direction of the track and in the vertical direction cause damage across several ECC blocks, but damage to each ECC block is as much as the thickness of the scratch, which causes relatively little data loss and is recovered by the disk's own ECC algorithm. . However, scratches that coincide with the track's direction of movement lose a large proportion of data in each ECC block, making it impossible to recover by the disc's own ECC algorithm, which actually causes data loss. As the thickness of the scratch increases, it affects multiple tracks, resulting in a greater amount of data loss. On the other hand, since the outer portion of the straight scratch over a certain length is staggered with the track direction, the amount of data loss due to the length does not show much correlation.

In other words, it is strong against scratches that do not coincide with the track direction, but vulnerable to scratches in the track direction.Scratch in the track direction is about 2.4 mm for CD, about 6 mm for DVD, and about 7 mm for Blu-ray Disc. Data can only be restored if surface damage occurs. In the case of more surface damage, data loss can occur.

The present invention uses an error correction code (ECC) in the original data to generate codewords containing redundant data, and to distribute them throughout the optical disc, thereby storing them by ECC using the optical disc itself as a standard. It provides distributed recording and reproducing method of data using error correction code on optical media that enables restoring the remaining data as part of additionally stored codeword in case of impossible data damage. The distribution form of redundant data can be arranged as widely as possible with regularity.

In order to achieve the above technical problem, the present invention provides a distributed recording method of data using an error correcting code on an optical medium.

A coding step (S1) of inputting an amount of damage to be repaired by a user and generating a codeword using an error correction code on original data to be stored on an optical disc; And

When the generated codewords are actually written to the disk, the dispersion of the data on the optical medium includes a dispersing step (S2) to change the order of the blocks so that the blocks dependent on the same codeword can be distributed over a wide area without successive arrangements. A distributed recording method is provided.

Preferably, one of various error correction codes such as Reed-Solomon Codes, Tornado Codes, and LDPC Codes may be used as the error correction code.

Preferably, prior to executing the coding step and the distribution step, including determining how many blocks of codewords are to be generated and how many parity blocks are to be generated (S2.5). The thickness and the number of scratches that can withstand are determined are characterized in that it can satisfy the reliability required by the user.

Preferably the step of determining how many blocks the codeword consists of and how many parity blocks to generate (S2.5) comprises the steps of: a user inputting a repairable damage thickness of the optical media (S39); Calculating a size of a codeword (S40); Inputting a number of recoverable damages by a user (S41); Calculating a number of parity blocks to be generated (S42); Determining whether the number of parity blocks is larger than a codeword size (S43); Calculating a size of an image after coding to be generated when the calculated number of parity blocks is applied when the number of parity blocks is smaller than a codeword size (S44); Determining whether the size of the image after coding is greater than the capacity of the disk, and ending if small (S45); And if the number of parity blocks is larger than the codeword size in step S43 or the size of the image after coding in step S45 is larger than the capacity of the disk, the process proceeds to step S41 after the coding from step S41 in which the user inputs the number of repairable damages. Determining whether the size of the image is larger than the capacity of the disk, and if the size is small, repeating the step (S45).

According to another aspect of the present invention, in the reproduction method of data recorded by using an error correction code on optical media,

Converting the requested block number into a physical block number (S3);

Reading data corresponding to a physical block number (S3.5);

Determining whether a block read error occurs (S4);

If a block read error occurs, the block reading remaining valid blocks of the corresponding codeword (S5);

Recovering a segment using a recovery algorithm of an error correcting code (S6);

Returning the block in which the request is received from the recovered segment to a higher layer and ending (S7); And

If a block read error does not occur in step S4, a method of reproducing data is provided, which includes the step of returning the read block to a higher layer and ending (S8).

In the present invention, error correction codes are used to protect data. The error correction code recovers original data using parity data when some data is lost.

The original data and the parity data used when generating the parity data are called a codeword. Parity data generation or damaged data recovery is performed in units of codewords.

If the error correction code is used, the recovery algorithm can recover even if some data is lost in one codeword. The more parity data is generated when the codeword is generated, the more the data damage in the codeword occurs, the more restoring is possible by using the remaining original data and parity data.

By applying the method proposed in the present invention, it is possible to stably store data on the optical disc by providing additional reliability required by the user in addition to the reliability implemented in hardware.

Hereinafter, the detailed description of the present invention will be described with reference to the accompanying drawings.

1 is a diagram illustrating a framework of the present invention. The file system image created is largely stored in an optical disk in two steps.

First, a file system image is generated by receiving a file and a directory, and an encoded image is generated using an error correction code (Lead Solomon code, tornado code, LDPC code, etc.).

To this end, the first step is a coding step (S1) of generating a codeword by using an error correction code on the original data to be stored on the optical disk and receiving a degree of damage from the user. The second step is a distribution step of changing the order of the blocks so that the generated codewords can be distributed over a wide area without successively arranging the blocks that depend on the same codeword when they are actually written to the disk (S2).

Optical discs use the ECC block size of the disc as the basic data unit used in the coding and distribution stages because data recovery and loss occurs in units of ECC blocks. For example, an ECC block of a DVD records 32KB and a Blu-ray disc (Disc) records user data of 64KB, so that coding and distribution use 32KB and 64KB units, respectively.

In the coding step S1, the file system image is processed by an error correction code in units of segments composed of several blocks as shown in FIG. 3A. The coding layer reads the segments one by one and applies parity data to generate parity data. The generated parity data is pasted after the segment to generate a codeword (Fig. 3B).

In the present invention, various error correction codes such as Reed-Solomon Codes, Tornado Codes, and LDPC Codes may be applied instead of specific error correction codes.

FIG. 3C is a diagram illustrating a distributed image composed of several codewords in a distributed layer. Disk surface damage can affect the access of not only directly damaged blocks but also other blocks within the same track, resulting in data loss of the entire track.

Blocks belonging to one codeword are lost together as one scratch when placed on the same track or adjacent tracks. Therefore, blocks must be arranged at a predetermined track interval to prevent this. To this end, the distribution layer interleaves g codewords one block at a time. By interleaving the blocks of each codeword, the blocks belonging to one codeword are arranged at intervals of g blocks so that they are distributed to a wide area in several tracks without being biased in some areas when they are written to the disc. The finally generated distributed image is recorded on a disc using an optical disc recording tool.

FIG. 4 is a diagram illustrating a process of determining how many blocks a codeword consists of and how many parity blocks to generate before executing coding and distribution. This number determines the thickness and number of scratches that can be tolerated, so you must determine these numbers to meet the reliability you require.

The number of blocks that make up the codeword is related to how close the blocks are to each other. If the codeword is composed of a large number of blocks, the blocks are densely arranged in the disk, so that when the surface of the disk is damaged, several blocks belonging to one codeword may be lost. Therefore, the user receives an input of how much to withstand the damage (scratch) (S39) and determines the size of the codeword using this (S40). Blocks belonging to one codeword are arranged in a distributed image by spaced apart by a predetermined block by the distribution layer. However, when this is actually written on the disk, blocks are arranged in the form as shown in FIG. 6 by the structure of the disk which rotates from the inside and records data. Even with a certain block spacing, the perimeter of the track becomes longer as it goes to the outside, so that fewer tracks are written. That is, when comparing the distance between tracks d ₀ to d ₃ , the track in which the two outermost sectors are located has the distance between the nearest tracks. If the interval between blocks is set to meet the user's needs based on the interval between these two tracks, the blocks of the inner track are arranged to maintain more track intervals.

Inter-block gaps are used so that the gap between these two tracks is more than the thickness required by the user. Using the interval between blocks, the number of blocks constituting the codeword can be obtained by the following method. It calculates how many blocks can be written during recording while leaving a certain space over the total capacity of the disk (S40). The number of blocks constituting the codeword is determined by the following equation.

Number of blocks constituting codeword = total number of blocks / interblock interval

As described above, if the size of the codeword is determined, the user inputs the number of recoverable damages (S41) to determine how many of them to use as a parity block. Using many parity blocks allows recovery even if several blocks are damaged. To determine this number, we input how many damages (scratches) the user can endure, and create a number of parity blocks that guarantees the number of fault tolerances. For example, when using the Reed Solomon code, as many parity sectors as the number of fault tolerances are generated (S42).

If the size of the parity block is larger than the size of the codeword as a result of the calculation, it is calculated by receiving input from the user again (S43). In addition, the size of the coded image to be generated is calculated by applying the calculated number of parity blocks (S44). When the size exceeds the capacity of the disk, the user receives the thickness again (S45). The size of the coded image to be generated is calculated as follows.

Number of codewords to be generated = size of data to store / size of segment

Size of coded image to be generated = number of codewords to be generated x size of codeword

5 is a diagram illustrating a process of reading data from a disc. The data written on the disk is stored in a block order through the coding layer and the distribution layer. Therefore, the read request for the specific block number coming down from the upper layer should be converted to the physical block number in which the target block is actually stored (S3). When the request for the block number s is received, the conversion process is as follows.

CW = s / (m + k)

blk = s% (m + k)

Physical block number = (blk × (m + k)) + cw

Here, (m + k) represents the size of the codeword in the number of blocks. The physical block number is used to actually read the data from disk and pass the data to the upper layer. Read the data corresponding to the physical block number (S3.5). If the block cannot be read due to partial damage of the disk, recovery is performed (S4). For restoring, the remaining valid block of the codeword to which the corresponding block belongs is read into the buffer (S5). The damaged blocks of the codeword are recovered by using a recovery algorithm of an error correcting code used in storing the data (S6). Read the block from the buffer and pass the data to the upper layer.

1 is a view showing the overall structure of the present invention.

2 is a view showing a process in which data is recorded on a disc according to the present invention.

3A is a diagram showing the division of a segment used in the present invention.

3B is a diagram showing the structure of a codeword used in the present invention.

3C is a diagram illustrating a process of interleaving codewords serving as a distribution layer in the present invention.

4 is a diagram illustrating a process of determining the size of a codeword and the amount of parity blocks to be generated in order to provide reliability required by a user in the present invention.

5 is a view showing a process of reading data from the optical disk in the present invention.

6 is a diagram illustrating an example of arrangement of one codeword.

Claims (5)

In the distributed recording method of data using error correction code on optical media, A coding step (S1) of inputting an amount of damage to be repaired by a user and generating a codeword using an error correction code on original data to be stored on an optical disc; And When the generated codewords are actually written to the disk, the dispersion of the data on the optical medium includes a dispersing step (S2) to change the order of the blocks so that the blocks dependent on the same codeword can be distributed over a wide area without successive arrangements. Distributed recording method. The method of claim 1, wherein the error correction code is one of various error correction codes such as Reed-Solomon Codes, Tornado Codes, and LDPC Codes. Distributed recording method of data on an optical medium. The method according to claim 1 or 2, further comprising the step S2.5 of determining how many blocks of codewords and how many parity blocks are to be generated before executing the coding step and the distribution step. And the thickness and number of scratches that can be tolerated are determined according to the determined numerical value to satisfy the reliability required by the user. The method of claim 1 or 2, wherein the determining of how many blocks the codeword consists of and how many parity blocks to generate is performed by the user inputting a recoverable damage thickness of the optical media. Step S39; Calculating a size of a codeword (S40); Inputting a number of recoverable damages by a user (S41); Calculating a number of parity blocks to be generated (S42); Determining whether the number of parity blocks is larger than a codeword size (S43); Calculating a size of an image after coding to be generated when the calculated number of parity blocks is applied when the number of parity blocks is smaller than a codeword size (S44); Determining whether the size of the image after coding is greater than the capacity of the disk, and ending if small (S45); And if the number of parity blocks is larger than the codeword size in step S43 or the size of the image after coding in step S45 is larger than the capacity of the disk, the process proceeds to step S41 after the coding from step S41 in which the user inputs the number of repairable damages. Determining whether the size of the image is larger than the capacity of the disk, and repeating the step (S45) if the image size is smaller. In the reproduction method of data recorded by using error correction code on optical media, Converting the requested block number into a physical block number (S3); Reading data corresponding to a physical block number (S3.5); Determining whether a block read error occurs (S4); If a block read error occurs, the block reading remaining valid blocks of the corresponding codeword (S5); Recovering a segment using a recovery algorithm of an error correcting code (S6); Returning the block in which the request is received from the recovered segment to a higher layer and ending (S7); And If a block read error does not occur in step S4, returning the read block to a higher layer and terminating (S8).

Images