KR100583360B1 - Encoding method of multiword information, word decoding method, encoding device of information decoding by word-type interleaving and error protection, information recording medium - Google Patents

Abstract

The multiword information is encoded based on multibit symbols relative to the record carrier, while providing word interleaving and word error protection code facilities. This can provide clues that indicate error locations across multiple word groups that occur in high protection clue words and point to low protection target words. Moreover, the clue words described above have a uniform first size and are interspersed in a first uniform manner. The plurality of target words has a uniform second size and is interspersed in a second uniform manner. In particular, such a structure can be applied for use in an optical memory device.

Multiple Word, Encoding, Decoding, Clue Word, Protection Code, Record Media

Description

METHOD FOR ENCODING MULTIWORD INFORMATION BY WORDWISE INTERLEAVING AND ERROR PROTECTION, WITH ERROR LOCATIVE CLUES DERIVED FROM HIGH PROTECTIVITY WORDS AND DIRECTED TO LOW PROTECTIVITY WORDS, A METHOD FOR DECODING SUCH INFORMATION, A DEVICE FOR ENCODING AND / OR DECODING SUCH INFORMATION, AND A CARRIER PROVIDED WITH SUCH INFORMATION}             

(Background of invention)

The invention relates to a method as described in the preamble of claim 1. US Pat. No. 4,559,625 to Berlekamp et al. And US Pat. No. 5,299,208 to Blaum et al. Disclose a method of decoding interleaved and error-protected information words, where the error pattern found in the first word is of the same group of multiple words. It may provide clues to locate errors in other words. The document uses a standardized format and a failure model with error bursts of multiple symbols over multiple words. The occurrence of an error present in a particular word provides a high probability of an error occurring at a symbol location corresponding to that indicated in the next word or a plurality of words. Often, this process increases the number of corrected errors. The inventors have recognized a problem with this principle, which is that the clue is only embodied when the clue word is completely corrected.

(Summary of invention)

After all, an object of the present invention is, above all, to provide a coding format that can decode clue words accurately with greater certainty than target words. Thus, according to one aspect of the invention, the invention has features in accordance with the features of claim 1. The found clue may generate or point to an erase symbol. Using these instructions, error correction proceeds more strongly. In fact, many codes correct up to t errors if the indication of the location of any error is not known. Given an erase position, a larger number of erases can generally be corrected, e> t. In addition, protection against a combination of bursts and random errors is improved. In contrast, providing an erase location requires the use of a very small number of syndrome symbols, thus simplifying the operation. In principle, the present invention can be used in a transmission environment in addition to a storage environment.

In addition, the present invention provides a recording medium having a method for decoding such encoded information, an encoding and / or decoding apparatus for use in the above method, and information for connecting to such encoding and / or decoding via an interface. It is about. Further preferred configurations of the invention are described in the dependent claims.

Such inventions and advantages of the present invention will be described in more detail below with reference to the preferred embodiments, in particular with reference to the following accompanying drawings:

1 shows a system with an encoder, a record carrier and a decoder,

2 illustrates the code format principle,

3 shows a product code format,

4 shows a Long Distance Code with burst detection,

5 shows a picket code and a burst indicator subcode,

6 illustrates a burst indicator subcode format,

7 shows a picket code and its product symbol,

8 shows another configuration of the picket code and the product symbol,

9 shows another format,

10 shows details of interleaving.

1 shows an overall system according to the invention with an encoder, a record carrier and a decoder. This embodiment is used to encode, store, and finally decode a sequence consisting of a plurality of samples or multi-bit symbols derived from an audio or video signal or from data. Terminal 20 receives a symbol stream, for example 8 bits in size. Splitter 22 repeatedly and cyclically sends the first symbol intended for a plurality of clue words to encoder 24. Moreover, splitter 22 sends all other symbols to encoder 26. Inside the encoder 24, a number of clue words are generated by encoding the relevant data into a plurality of code words of the first multisymbol error correction code. As such a code, a Reed-Solomon code, a product code, an interleaved code, or a combination of these codes may be used. Inside the encoder 26, a target word is formed by encoding the code word of the second multisymbol error correction code. In this embodiment, all codewords have a uniform length, but this is not a strict requirement. Preferably, these two codes are Reed-Solomon codes with subcodes of the first code and the second code. As will be more apparent from FIG. 2, clue words generally have much higher error protection and include a relatively small number of non-free symbols.

In block 28, the code words thus formed are delivered to one or more outputs in which any number is indicated so that the distribution on the recording medium described later is uniform. Block 30 symbolizes the record carrier itself receiving the encoded data. This may actually involve writing directly to the appropriate combination of recording mechanism and recording medium. Alternatively, the recording medium may be implemented as a copy from a master encoded recording medium such as a stamp. Preferably, the memory device is optical and completely in series, but other structures may be used. In block 32, a number of words are read back from the record carrier, and the clue words of the first code are then sent to the decoder 34 and decoded based on its inherent margin. Moreover, as can be seen from the description with reference to FIG. 2 below, such decoding provides a clue as to the location of the error inside other than these clue words. Box 35 includes a program for receiving these clues and using one or more different strategies to translate these clues into an erase location. The target word is decoded inside the decoder 36. Under the control of such an erase position, the error prevention of the target word rises to a level that can be tolerated. Finally, all decoded words are demultiplexed into output 40 using component 38 according to the original format. For simplicity, the mechanical configuration to interface the various subsystems with each other is omitted.

2 illustrates a relatively simple code format. As shown, the coded information is conceptually placed within a block of 16 rows and 32 columns, ie a block of 512 symbols. The storage on the record carrier proceeds row by row starting from the upper left column. The hatched area includes a plurality of check symbols, and words 0, 4, 8, and 12 each have eight check symbols and constitute clue words. The other words each contain four check symbols and constitute a target word. The entire block contains 432 information symbols and 80 check symbols. The latter check symbols can be placed in a more distributed manner compared to their respective words. Some of the information symbols may be dummy symbols. The Reed-Solomon code can correct errors up to four symbol errors inside each clue word. The actual symbol error is shown with a cross. After all, every clue word can be decoded correctly, unless it has more than four errors. In particular, however, words 2 and 3 cannot be decoded based solely on their own spare symbols. In Fig. 2, all errors except 62, 66 and 68 represent error strings. However, strings 52 and 58 that intersect at least three consecutive clue words are considered to be error bursts, so at least get the erase flag at all intermediate symbol positions. In addition, the target word before the first clue word of the burst and the target word immediately after the last clue symbol of the burst may take an erase flag at that location depending on the target that follows. String 54 is too short to think of as a burst.

As a result, two errors in word 4 create an erase flag inside the associated both columns. This configuration makes words 2 and 3 correctable using a single error symbol and two erase symbols, respectively. However, random errors 62, 68 rather than string 54 constitute clues for words 5, 6, and 7 because each of them contains only one clue word. In certain circumstances, erasure may result in a zero error pattern, since any error in an 8 bit symbol has a 1/256 probability of generating the correct symbol again. Likewise, long bursts that cross a particular clue word can produce the correct symbol therein. By the strategy of linking between preceding and subsequent clue symbols of the same burst, such exact symbols are integrated inside the burst, and likewise the clue symbol with errors is translated into the erase value for the associated target symbol. The above solution can be modified according to the decoding method which can be further controlled by other parameters.

(Explanation of actual format)

The actual format will be described below. 3 symbolizes the product code format. Words exist in the horizontal and vertical directions and hatched parity. Figure 4 symbolizes a so-called long range code with special bus detection on several upper words with more parity. The present invention provides a so-called picket cord that can be configured as a combination of the principles of FIGS. 3 and 4. As always, the recording process proceeds sequentially along the arrows shown in Figs.

The practical aspect of the present invention is brought about by a newer method for digital optical storage. A special feature is that in the case of a readout projecting onto a substrate, the upper transmissive layer has a thickness as thin as 100 microns. Since the channel bits can be approximately 0.14 microns in size, at a channel rate of 2/3, the data bytes can only be 1.7 microns long. The beam diameter at the top surface has a diameter of approximately 125 microns. Caddy or lid on the disk reduces the probability of large bursts. However, deviated particles smaller than 50 microns can cause short failures. Above all, the inventors used a failure model in which such failures through error propagation could result in bursts of 200 microns corresponding to about 120 bytes. In particular, we used an error model with a fixed size burst of 120B starting randomly with a probability of one burst per 2.6 * 10 ^-5 bytes, or on average, 32 kB blocks. Although the present invention has been made by the development of an optical disk storage device, the present invention as described herein may be applied to a configuration such as a multitrack tape and other techniques such as magnetic and magneto-optical.

5 shows a picket code and a burst indicator subcode. The picket code consists of two subcodes A and B. The burst indicator subcode (BIS) contains a clue word. By format, this corresponds to a very closely interleaved long distance code that allows it to be placed in the position of multiple burst errors. The error pattern thus found is signal processed to obtain the erase information for the target word composed of the product code (PS) in this embodiment. The product subcode described above corrects a combination of a plurality of bursts and random errors by using an erase flag obtained from the burst indicator subcode.

The following format is suggested:

A block of '32 kB 'contains 16 DVD compatible sectors.

Each such sector contains 2064 = 2048 + 16 bytes of data.

After ECC encoding each sector contains 2368 bytes.

Therefore, the coding rate is 0.872.

Inside the block, 256 sync blocks are formatted as follows.

Each sector contains 16 sync blocks.

Each sync block consists of four groups of 37Bs.

Each group of 37Bs contains a carefully coded interleaved burst indicator subcode of 1B and a product code of 36B.

As shown in Fig. 5, the rows are read sequentially from the disc starting from the preceding sync pattern. Each row contains four bytes of BIS, hatched and numbered consecutively, separated by 36 different bytes. Sixteen rows constitute one sector and 256 rows constitute one sync block.

FIG. 6 shows only the burst indicator subcode format of the same 64 numbered bytes per sector of FIG. 5, constructed as follows:

There are 16 rows with [64, 32, 33] RS codes, each with t = 16.

Multiple rows are derived sequentially from the disc as indicated by the arrow, so that a group of four columns is derived from a single sector for quick addressing.

The BIS can represent at least 16 bursts of 592B (~ 1mm) each.

The BIS contains 32B data per sector, with four rows of BIS in particular, including 16B DVD headers, 5B parity for fast address reading and 11B user data.

7 shows a picket word constructed from a target word and its product subcode. The bytes of the product subcodes described above are numbered in the order in which they are read from the disc, while ignoring BIS bytes.

8 shows various other configurations of this embodiment of the product subcode. In particular, this product subcode corresponds to the [256,228,29] * [144,143,2] product code of the Reed-Solomon code. The number of data bytes is 228 * 143 = 32604, i.e., 16 times user bytes of (2048 + 11) plus 12 spare bytes.                 

FIG. 9 illustrates an alternative format for FIG. 8, in which all Reed-Solomon codes in the horizontal direction are omitted. The horizontal block size is 36 bytes (1/4 of FIG. 7) and uses the [256,224,33] Reed-Solomon code. Each sector has 2368 bytes, and no dummy bytes are needed.

The code in the first column is formed in two steps. From each sector, 16 header bytes are first encoded inside the [20, 16, 5] code to allow for quick address lookup. The sum of the resulting 20 bytes per sector plus another 32 user bytes make up the data byte and is further encoded as a whole. Data symbols of one 2K sector may exist only in one physical sector as follows. Each column of [256,224,33] codes contains eight parity symbols per 2K sector. Moreover, each [256,208,49] code has 12 parity symbols per 2K sector and four parity symbols of [20,16,5] code to obtain a [256,208,49] code with 48 margin bytes. .

10 illustrates such interleaving in detail. Where 'x' represents header byte, '□' represents parity of [20,16] code, and '●' represents 32 " other " data bytes of [256,208] code and parity byte of 12 .

Claims (22)

A method of encoding multiword information based on multiple bit symbols relatively adjacent to a recording medium while providing word-type interleaving and word-type error protection code facilities to provide clues of error locations across multiple word groups. To Generating a clue in the high protection clue word to be directed to the low protection target word. The method of claim 1, The clue word has a first uniform size, a second uniform size, and is interspersed in a first uniform manner with respect to a plurality of target words interspersed in a second uniform manner. Encoding Method. The method of claim 1, The encoding method is applied to storage on an optical record carrier. The method of claim 1, And the clue word includes header information of a sector related to the inside of the block including the code facility, the header information being given to the recording medium in correspondence with the position of each related sector. The method of claim 4, wherein And wherein said header information has additional error protection outside of said code facility for each sector. Word deinterleaving and error protection code, including the step of decoding received multiword information based on a given multi-bit symbol relatively adjacent to a record carrier, while evaluating clues indicating error location across multiple word groups In the decoding method for performing the decoding of the equipment, And deriving the clue from the high protection clue word towards the low protection target word. The method of claim 6, The decoding method comprises a plurality of clue words having a first uniform size and interspersed in a first uniform manner, and a plurality of target words having a second uniform size and interspersed in a second uniform manner. Decoding method characterized in that it is based. The method of claim 6, Said decoding method being applied to storage on an optical record carrier. The method of claim 6, The decoding method is characterized in that a corrected symbol inside the clue word provides each clue, and successive clues in the received information series collectively generate an erase flag for the intermediate symbol of the target word. The method of claim 9, And a conceptual clue is assigned to an unaltered clue word symbol in the middle of the series. The method of claim 6, And the clue word includes header information of a sector related to the inside of the block including the code facility and derives the header information from the recording medium corresponding to the position of each related sector in sequence. The method of claim 11, And the decoding method undertakes error protection per sector for header information outside of the code facility. Interleaving means for encoding multiword information based on multi-bit symbols relative to a recording medium and providing word-type interleaving; coding means for providing word-type error protection code facilities; An encoding apparatus having an allocation means for generating a clue indicating an error location, And said assigning means is configured to provide a clue for generating internally to a high degree of protection clue word and directing to a low degree of protection target word. The method of claim 13, The interleaving means is configured to interleave the clue words having a first uniform size interspersed in a first uniform manner with respect to a target word having a second uniform size interspersed in a second uniform manner. And an encoding device. Deinterleaving means for decoding received multiword information based on a given multibit symbol relatively adjacent to a recording medium, performing word-type deinterleaving, decoding means for decoding an error protection code facility, and multiword group A decoding apparatus having evaluation means for evaluating a clue indicating an error position over And said evaluating means is adapted to derive a clue from a high degree of protection clue word toward a low degree of protection word. The method of claim 15, The decoding apparatus is based on a plurality of clue words having a first uniform size and interspersed in a first uniform manner and a plurality of target words having a second uniform size and interspersed in a second uniform manner. Decoding apparatus characterized in that. A physical recording medium produced by carrying out the method of claim 1, And a plurality of interleaved clue words and an array of target words, wherein the clue words have better error protection for the target words. The method of claim 17, The clue word has a first uniform size, a second uniform size, and is interspersed in a first uniform manner with respect to a target word interspersed in a second uniform manner media. The method of claim 17, And the recording medium is based on optical memory. The method of claim 17, And the recording medium is configured to be used for a substrate incidence reading process. The method of claim 17, And the clue word includes header information of a sector related to the inside of the block including the code facility, the header information being given to the medium in correspondence with the position of each related sector. The method of claim 21, And the header information has error protection outside the code facility for each sector.

Images