JP2009049896A - Communication path encoding method, communication path encoding system, hierarchical communication path encoding program, and hierarchical communication path decoding program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide high-quality hierarchically encoded data without damaging a scalability function by uniformly adding an arbitrary error-correcting code to the hierarchically encoded data. <P>SOLUTION: According to the present invention, at a transmitting side (encoding device), a hierarchical coarse matrix constituted of a plurality of coarse matrices generated based on a random number is generated, parity data is obtained by performing exclusive OR operation between hierarchical data generated by dividing target data and the hierarchical coarse matrix, and the hierarchical data, the parity data and the hierarchical coarse matrix are transmitted to a receiving side (decoding device). At the receiving side, error-correcting decoding is performed by solving simultaneous equations using the received data. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a channel coding method, a channel coding system, a hierarchical channel coding program, and a hierarchical channel decoding program, and more particularly to generation of linear codes for related hierarchical data. Channel coding method and channel coding for hierarchically coded data for efficiently protecting from data errors by changing the configuration of the Tanner graph of matrix and check matrix The present invention relates to a system, a hierarchical channel encoding program, and a hierarchical channel decoding program.

  Currently, error correction techniques are widely used in various communication systems such as satellite digital broadcasting, communication on the Internet, and communication on mobile terminals. In particular, with the development of broadband environments in recent years, video distribution services using the Internet are expected, and error correction technology for the Internet network has become important. The following is an example of the Internet network.

  When a communication channel is viewed from the side that provides services using the Internet, the Internet network can be regarded as a lost communication channel (PEC). This is a binary erasure channel as shown in FIG. 9 grouped as a unit of packet, and the channel output is output as correct information or output as unknown "#" due to some trouble in the line. It is a communication path in either case. In this communication system, error correction codes called FEC (Forward Error Correction) and ARQ (Automatic Recovery Quotiet) are used. In general Internet services, the TCP / IP protocol is used, and only error detection is performed on the decoder side, and when an error occurs, a method of correcting the error by retransmission control or the like is used (ARQ method). However, when a large-scale service such as multicast distribution is assumed, it is required to correct errors based on data received on the receiving side, and a method that can correct errors only on the receiving side without requiring a feedback channel Is used (FEC method). Since this FEC method does not perform retransmission control or the like, the FEC method is also used in a video conference system in which real-time performance is important with little delay. In the following, we will talk about the FEC method.

  As the FEC method, Reed-Solomon code (RS code) is widely used in digital broadcasting and the like. In Japanese digital broadcasting, it is defined as a code length of 204 bytes, and by adding a parity byte of about 10 (%) to the original data of 188 bytes, tolerance is given to errors. However, it is generally known that performance is improved if a code having a long code length is used as an error correction code. However, there is a known disadvantage that decoding becomes complicated and the amount of calculation becomes enormous as the code length increases. ing. For this reason, RS codes are usually assumed to have a maximum code length of 255 bytes.

On the other hand, decoding methods based on the belief propagation algorithm (Belif propagation algorithm) are known to have excellent decoding characteristics with a practical amount of computation when the code length is long, and are defined by a very sparse graph. A low density parity check code (LDPC code) (for example, see Non-Patent Document 1), which is a linear code to be used, is attracting attention as a practical error correction method approaching the channel capacity defined by shannon. Has been. Moreover, as an erasure correction code based on a sparse graph, LT code (for example, refer nonpatent literature 2) and Raptor code (for example, refer nonpatent literature 3) of Digital Fountain are known, and encoding efficiency is increased. It is known that a code characteristic that can be decoded only by receiving arbitrary code data without deterioration is achieved with a practical amount of computation. This property is widely used in layered multicast communications in order to be compatible with ALC (Asynchronous Layered Coding) which is an Internet multicast protocol.
RGGallager, "Low density parity check codes," in Research Monograph series. Cambridge, MIT Press, 1963 M. Luby, "LT Codes", The 49rd Annual IEEE Symposium on Foundations of Computer Science, 2002. Shokrollahi, A, "Raptor codes," Information Theory, IEEE Transactions on Volume 52, Issue 6, 2006

  As described above, an error correction code based on a sparse graph and a probability propagation algorithm achieves coding characteristics that could not be achieved conventionally. However, in the conventional technique, the data to be protected and the data to be protected (parity data) are limited to a one-to-one relationship. In particular, in recent years, data to be protected has scalability, such as being encoded in stages, but in the conventional technology, error correction that protects data from errors uniformly for this hierarchically encoded data. A sign cannot be added.

  The present invention has been made in view of the above points, and adds an arbitrary error correction code to the hierarchically encoded data in a unified manner so that high-quality hierarchically encoded data can be obtained without impairing the scalability function. It is an object of the present invention to provide a channel coding method, a channel coding system, a hierarchical channel coding program, and a hierarchical channel decoding program for hierarchically coded data that can provide the same.

  FIG. 1 is a diagram for explaining the principle of the present invention.

The present invention (Claim 1) is a channel coding method for protecting hierarchically coded data from errors in a system comprising a hierarchical channel coding apparatus and a hierarchical channel decoding apparatus,
In the hierarchical channel coding apparatus,
A hierarchical sparse matrix generation step (step 1) for generating a hierarchical sparse matrix composed of a plurality of sparse matrices;
A hierarchical encoding step (step 2) for obtaining parity data using a hierarchical sparse matrix for hierarchical data obtained by dividing the target data for each scalability function;
Transmitting the hierarchical data, parity data and hierarchical sparse matrix to the hierarchical channel decoding device (step 3),
In the hierarchical communication path decoding device,
A hierarchical decoding step (step 4) of decoding data transmitted from the hierarchical channel encoder using a hierarchical sparse matrix is performed.

The present invention (Claim 2) provides a hierarchical sparse matrix generation step (Step 1).
Performing a step of generating a hierarchical sparse matrix so that one piece of parity data is generated for one layer of data;
In the hierarchical decoding step (step 4):
A step of decoding the hierarchical data and the parity data on a one-to-one basis using a hierarchical sparse matrix is performed.

In the hierarchical sparse matrix generation step (step 1), the present invention (claim 3)
Generating a hierarchical sparse matrix using a matrix defined by one random sparse graph for a part or all of the hierarchical sparse matrix;
In the hierarchical decoding step (step 4):
Parity data is decoded using a hierarchical sparse matrix defined by one random sparse graph, part or all of the parity data.

The present invention (Claim 4) provides a hierarchical decoding step (Step 4):
Received data is selectively decoded.

  FIG. 2 is a principle configuration diagram of the present invention.

The present invention (Claim 5) is a system comprising a hierarchical channel encoder 20 and a hierarchical channel decoder 30 for protecting hierarchically encoded data from errors,
The hierarchical channel encoding device 20
Hierarchical sparse matrix generation means 104 for generating a hierarchical sparse matrix composed of a plurality of sparse matrices;
Hierarchical coding means 102 for obtaining parity data using a hierarchical sparse matrix for hierarchical data obtained by dividing the target data for each scalability function;
Transmission means for transmitting the hierarchical data, parity data and hierarchical sparse matrix to the hierarchical channel decoding device 30,
The hierarchical channel decoding device 30
There is a hierarchical decoding means for decoding data transmitted from the hierarchical channel encoding apparatus 20 using a hierarchical sparse matrix.

Further, the present invention (Claim 6) is provided in the hierarchical sparse matrix generation means 104 of the hierarchical channel encoding device 20,
Means for generating a hierarchical sparse matrix so that one piece of parity data is generated for one layer of data;
In the hierarchical decoding means 108 of the hierarchical communication path decoding device 30,
Means for decoding the hierarchical data and the parity data on a one-to-one basis using a hierarchical sparse matrix.

Further, the present invention (Claim 7) is provided in the hierarchical sparse matrix generation means 104 of the hierarchical communication path encoding device 20,
Means for generating a hierarchical sparse matrix using a matrix defined by one random sparse graph for part or all of the hierarchical sparse matrix;
In the hierarchical decoding means 108 of the hierarchical communication path decoding device 30,
Means for decoding part or all of the parity data using a matrix of a hierarchical sparse matrix defined by one random sparse graph;

  The present invention (Claim 8) includes means for selectively decoding received data in the hierarchical decoding means 108 of the hierarchical communication path decoding apparatus 30.

  The present invention (Claim 9) is a hierarchical channel coding program for causing a computer to function as each means constituting the hierarchical channel coding apparatus of the channel coding system according to claim 6 or 7. .

  The present invention (Claim 10) is a hierarchical channel decoding program for causing a computer to function as means for constituting the hierarchical channel decoding apparatus of the channel encoding system according to any one of claims 6 to 8. .

  According to the present invention, it is possible to implement an error correction code considering an arbitrary relationship with respect to hierarchically encoded data, and impair the scalability function of hierarchically encoded data in accordance with the request of the receiving user. Error correction codes can be added uniformly.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 3 shows an error correction function for hierarchically encoded data that can be realized by the present invention. Although the present invention can be implemented without depending on input data, a case where the present invention is applied to image coding will be described as an example.

  When an image is an input signal, the information source coding apparatus 10 uses JPEG2000 (ISO / IEC 15444-2. “JPEG2000 image coding system,” 2000) or MPEG SVC (Scalable video coding) in hierarchical information source coding means. (J. Reichel, H. Schwarz and M. Wien: "Joint Scalable Video Model JSVM-6," ISO / IEC JTC1 / SC29 / Wg11, N8015, Montreux, April, 2006) The compression is performed, and the hierarchical data is output to the channel encoder 20. According to the present invention, in the channel coding apparatus 20, arbitrary error correction data can be added to this hierarchical data based on the relationship between the data layers, and any required condition (for example, decoding processing) Hierarchical data can be protected / decrypted according to the amount and bandwidth. That is, error correction with scalability that can add hierarchical error correction data according to an arbitrary request to compressed data with scalability encoded in the hierarchy is realized.

  FIG. 4 shows the configuration of a hierarchical channel coding system according to an embodiment of the present invention.

  The hierarchical channel coding system 100 shown in FIG. 1 includes a hierarchical channel coding apparatus 20 including a hierarchical data dividing unit 101, a hierarchical coding unit 102, a random number signal generating unit 103, and a hierarchical sparse matrix generating unit 104. From the hierarchical data synthesizing device 50 having the transmission module 105, the channel 106, the receiving module 107, the hierarchical channel decoding device 30 having the hierarchical decoding unit 108, the data rearrangement unit 109 and the header rewriting unit 110. Composed.

  The hierarchical channel encoding apparatus 20 includes a hierarchical data dividing unit 101, a hierarchical encoding unit 102, a random number signal generating unit 103, and a hierarchical sparse matrix generating unit 104.

  The hierarchical data dividing unit 101 of the hierarchical channel encoding device 20 divides the input hierarchical encoded data for each arbitrary hierarchy. The data divided for each hierarchy is temporarily stored in an internal memory (not shown), storage means (not shown) connected to the hierarchy data dividing unit 101, or the like.

  The random number signal generation unit 103 generates a random signal and outputs it to the hierarchical sparse matrix generation unit 104.

  The hierarchical sparse matrix generation unit 104 generates a sparse matrix that enables hierarchical encoding using the random signal generated by the random number signal generation unit 103. The generated sparse matrix (hierarchical sparse matrix) is temporarily stored in an internal memory (not shown) or storage means (not shown) connected to the hierarchical sparse matrix generation unit 104. Shall.

  The hierarchical encoding unit 102 generates hierarchical error correction data using the sparse matrix generated by the hierarchical sparse matrix generation unit 104 for the layers divided by the hierarchical data dividing unit 101. It is assumed that the generated error correction data is temporarily stored in an internal memory (not shown) or a storage means (not shown) connected to the hierarchical encoding unit 102.

  The transmission module 105 transmits the data output from the hierarchical encoding unit 102 and the hierarchical sparse matrix generation unit 104.

  The channel 106 is a channel that is a line such as the Internet.

  The reception module 107 receives data transmitted by the transmission module 105.

  The hierarchical decoding unit 108 of the hierarchical communication path decoding device 30 decodes error correction codes hierarchically. It is assumed that the decoded data is stored in an internal memory (not shown) or a storage means (not shown) connected to the hierarchical decoding unit 108.

  The hierarchical data synthesis device 50 includes a data rearrangement unit 109 and a header rewrite unit 110.

  The data rearrangement unit 109 converts the data decoded by the hierarchical decoding unit 108 into decodable data.

  The header rewriting unit 110 rewrites the header of the data rearranged by the data rearranging unit 109.

  The system shown in FIG. 4 is a hierarchical channel coding system capable of performing error correction not only on one layer but also on any layer from the input layer coded data. Therefore, in the following, a hierarchical channel coding method for hierarchically coded data will be described assuming video distribution over the Internet. However, the input data can be not only a video signal but also any data that is hierarchically encoded.

  It is assumed that many of the input images are encoded with the international standard JPEG2000 having a scalability function. JPEG2000 data can have various layers according to the progression order. FIG. 5 shows a JPEG2000 data structure having a progression order called RLCP.

  FIG. 6 is a flowchart of an outline operation in one embodiment of the present invention.

  Step 10) The random number signal generation unit 103 of the hierarchical channel encoding device 20 generates a random number (random signal) and outputs the random number to the hierarchical sparse matrix generation unit 104.

  Step 20) The hierarchical sparse matrix generation unit 104 generates a sparse matrix P based on the random signal generated by the random number signal generation unit 103, and generates a hierarchical sparse matrix G by combining a plurality of sparse matrices P. Or the like stored in a storage means (not shown).

  First, a method for creating the sparse matrix P will be described.

  Various methods for creating a sparse matrix have been proposed. Among them, Non-Patent Document 1, Reference 1 “David JC MacKay,“ Good Error-Correcting Codes Based on Very Sparse Matrices ”. IEEE Transaction on Information Theory , Vol. 45, No. 2, March, 1999 ”can easily and efficiently create a sparse matrix. Below, the method described in the literature 1 is demonstrated.

  FIG. 7 is a flowchart of processing for creating the sparse matrix P of the hierarchical sparse matrix generation unit according to the embodiment of the present invention. However, the following m shows the number of rows of the sparse matrix P, and k shows the number of columns.

  Step 201) An all-zero matrix P (m × k) is created.

  Step 202) It is assumed that the count i = 0 and j = 0.

  Step 203) Each column bit is inverted based on the random number input from the random number signal generation unit 103. The inversion probability in the random number signal generation unit 103 is set such that each column is inverted by about t bits.

  Step 204) i = i + 1.

  Step 205) If m <i, go to Step 206, otherwise go back to Step 203.

  Step 206) j = j + 1.

  Step 207) If k <j, go to Step 108, otherwise go back to Step 203.

  Step 208) The following processing is performed on the sparse matrix P generated above.

  For the sparse matrix P configured by the above processing, the number of standing “1” s in all the columns (m) is t as much as possible, and the row (k) in which all the “1” s are standing as much as possible. The numbers are also equal (the specific average number of bits depends on the size of the matrix, but if there is a relationship of 2m = k, the number of bits in a row is 2t). Here, “as much as possible” means being stochastically unstable based on random numbers. As a result, a loop as short as possible does not occur in the configured matrix P.

  The above method is a method for creating one sparse matrix P, and the hierarchical sparse matrix G used for encoding can be variously combined with the sparse matrix P.

It is also possible to create P _{m1, kl} in the following formula (1) by the above method, and the above method with P _{m1, kl} and P _{m2, k1} in formula (1) as P _{m1 + m2, k1} It is also possible to create it. Note that the coding strength changes by changing m, but the coding efficiency increases as the processing of the flowchart of FIG. 7 proceeds.

Now, it is assumed that the hierarchical data to be protected is (u _k1 , u _k2 ), and the hierarchical data (u _k2 ) is a signal that does not make sense if there is no hierarchical data (u _k1 ). This is often the case with hierarchically encoded data, such as the relationship between the low frequency component (LL) and the high frequency component (HH) of JPEG 2000 and the relationship between the quality component layers 1 and 2. In such a case, the present invention generates a hierarchical sparse matrix G so that error correction codes can be added uniformly based on the relationship of hierarchical data. That is, in the above example, the hierarchical sparse matrix G is

として生成する。但し、Ｐ_ｍ，ｋはｋの疎行列である。そうすることにより、階層的符号化部１０２では、次式の演算でパリティビット（ｃ_ｍ１、ｃ_ｍ２）が生成できる。

Generate as Here, P _{m, k} is a sparse matrix of k. By doing so, the hierarchical encoding unit 102 can generate parity bits (c _m1 , c _m2 ) by the following equation.

なお、ｃ_ｍ１は階層データｕ_ｋ１に対するパリティビットであり、ｃ_ｍ２は階層データｕ_ｋ１及びｕ_ｋ２に対するパリティビットである。また、式（２）はＧが疎行列であることと、exclusive OR(XOR)演算であることから、高速に少ない演算量でパリティビットを作成することができる。符号化は式（２）によってｃを生成することで行われる。

Note that c _m1 is a parity bit for the hierarchical data u _k1 , and c _m2 is a parity bit for the hierarchical data u _k1 and u _k2 . Also, since Equation (2) is that G is a sparse matrix and is an exclusive OR (XOR) operation, parity bits can be created at high speed with a small amount of calculation. Encoding is performed by generating c according to equation (2).

  Step 30), the data input to the hierarchical channel coding apparatus 100 is divided by the hierarchical data dividing unit 101 into units to be divided for each scalability function.

  FIG. 8 is a flowchart of the operation of the hierarchical data dividing unit in one embodiment of the present invention.

  Step 301) k shown below is the number of layers (number of layers), and k = 1 is set as an initial value. Note that the maximum number of hierarchies is L.

  Step 302) When dividing the hierarchically encoded data, initialization is performed as follows.

ステップ３０３） 階層データ分割部１０１は、入力された階層符号化データｕ_ｋを以下のように分割し、階層データ分割部１０１内のメモリ等の記憶手段（図示せず）に格納する。

Step 303) the hierarchical data division section 101, hierarchically encoded data u _k input divided as follows, and stores in the storage unit such as a memory in the hierarchy data division unit 101 (not shown).

ステップ３０４）

Step 304)

には、処理を終了し、それまではステップ３０５に移行する。

In this case, the process is terminated, and the process proceeds to step 305 until then.

  Step 305) Increments the value of k by 1 (k = k + 1) and returns to Step 303.

That is, in the case of JPEG2000, hierarchical data such as resolution, quality, color format, and position information is divided into arbitrary hierarchies. For convenience, it is assumed here that the data is divided into vector data of two layers (u _k1 , u _k2 ). k1 and k2 are the number of layers, respectively. The divided data is temporarily stored in a memory (not shown) and then sent to the hierarchical encoding unit 102.

  Step 40) The hierarchical encoding unit 102 adds an arbitrary error correction code to one layer and an arbitrary layer using the sparse matrix generated by the hierarchical sparse matrix generation unit 104 described later.

  FIG. 9 is a flowchart of the operation of the hierarchical encoding unit in an embodiment of the present invention.

  Step 401) As an initial value, the number m of parity data is set to m = 1.

Step 402) Parity data _Cm is generated based on the hierarchical sparse matrix G generated by the hierarchical sparse matrix generation unit 104, and storage means (not shown) such as a memory in the hierarchical encoding unit 102 To store.

  Step 403)

であればステップ４０４に移行し、そうでなければ処理を終了する。

If so, the process proceeds to step 404; otherwise, the process ends.

  Step 404) The count m is incremented by 1 (m = m + 1), and the routine proceeds to Step 402.

  Step 50) The hierarchical data u and the parity data c and the hierarchical sparse matrix G used for encoding are sent to the communication module 105 and distributed to the receiving module 107 through the channel 106.

  Step 60) The receiving module 107 transmits the data received from the channel 106 to the hierarchical communication path decoding device 30.

  Step 70) The distributed data is sent to the hierarchical communication path decoding device 30, temporarily stored in a storage means (not shown) such as a memory, and the hierarchical decoding unit 108 performs decoding according to scalability.

  FIG. 10 is a flowchart of the operation of the hierarchical decoding unit in one embodiment of the present invention.

  Step 701) The layer count is initialized (L = 1).

  Step 702) The number of received data (n, c) is initialized (d = 1).

  Step 703) Based on the hierarchical sparse matrix G and the received data u, c, the simultaneous equations of the above equation (2) are solved.

  Step 704)

であればステップ７０６に移行し、そうでなければステップ７０５に移行する。

If so, the process proceeds to step 706; otherwise, the process proceeds to step 705.

  Step 705) It is determined whether L <max (L). If so, the process proceeds to Step 707, and if not, the process ends.

  Step 706) d is incremented by 1 (d = d + 1), and the process proceeds to Step 703.

  Step 707) L is incremented by 1 (L = L + 1), and the process proceeds to the next layer process (Step 702).

That is, the hierarchical decoding unit 108 of the present invention can selectively perform error correction decoding in accordance with requirements such as decoder resolution and quality requirements and processing speed. In the above two-layer example, decoding is performed by solving simultaneous equations so as to satisfy Expression (2) using the hierarchical matrix G. The decoding procedure is executed by the same operation as the decoding of a general LDPC (Low Density Parity Check) erasure correction code, and is executed by repeatedly performing the operation of solving the linear simultaneous equations for each row. However, for example, when decoding only certain hierarchical data from the request of the receiving terminal, here, hierarchical data (u _k1 ), that is, partial decoding, it is only necessary to perform decoding by solving the following equation as a linear simultaneous equation: .

この機能により、従来技術では実現不可能であった階層符号化データの選択的誤り訂正復号が可能となり、プログレッション機能を損なうことなく誤り訂正符号を付加することが可能である。

This function enables selective error correction decoding of hierarchically encoded data that could not be realized with the prior art, and an error correction code can be added without impairing the progression function.

  Step 80) The selectively decoded data is temporarily stored in a memory (not shown), and the data rearrangement unit 109 rearranges the data.

  Step 90) The header rewriting unit 110 performs a header rewriting operation without any arithmetic processing, and converts it so that it can be decoded by a general decoder such as JPEG2000.

  As described above, hierarchical channel encoding and decoding are performed on hierarchically encoded data.

In the present invention, the case where the number of hierarchies is two has been described as an example. However, the present invention is easily expanded when there are many hierarchies. For example, hierarchical data (u _k1 , u _k2 , u _k3 ) and hierarchical data (u _k1 ) and hierarchical data (u _k2 ), and the hierarchical data (u _k1 ) and hierarchical data (u _k3 ) are strongly related, and the hierarchical data (u _k2 ) and hierarchical data (u _k3 ) are not related. Are encoded and decoded using the following sparse matrix.

  For the sake of explanation, consider the following sparse matrix.

従って、Ａ，Ｇ，Ｄはｕ１（Layer1）に、Ｂ，Ｅ，Ｈはu2（Layer2）に、Ｃ，Ｆ，Ｉはｕ３（Layer3）にそれぞれ対応することがわかる。ここで、０行列の挿入位置は、どのパリティデータにどのレイヤを保護する機能を持たせるかによって決定される。以下の式（４）は、ｃ１にｕ１を保護するように、ｃ２にｕ１及びｕ２を保護するように、ｃ３にｕ１及びｕ３を保護するようにそれぞれデータを付加した例である。

Therefore, it can be seen that A, G, and D correspond to u1 (Layer1), B, E, and H correspond to u2 (Layer2), and C, F, and I correspond to u3 (Layer3). Here, the insertion position of the 0 matrix is determined by which parity data has a function to protect which layer. The following formula (4) is an example in which data is added so that u1 is protected to c1, u1 and u2 are protected to c2, and u1 and u3 are protected to c3.

（４）
上記のように、送信側の階層的通信路符号化装置２０では、乱数に基づき生成した複数の疎行列Ｐからなる式（１）、（４）等で示される階層的疎行列Ｇを生成し、対象データを分割して生成した階層データｕと階層的疎行列Ｇとの間で式（２）で表される排他的論理和演算を行うことにより、パリティデータｃを求め、階層データｕ、パリティデータｃ、及び階層的疎行列Ｇを送信する。

(4)
As described above, the hierarchical communication path encoding device 20 on the transmission side generates the hierarchical sparse matrix G represented by the equations (1), (4), etc. composed of a plurality of sparse matrices P generated based on random numbers. The parity data c is obtained by performing the exclusive OR operation represented by the equation (2) between the hierarchical data u generated by dividing the target data and the hierarchical sparse matrix G, and the hierarchical data u, Parity data c and hierarchical sparse matrix G are transmitted.

  The hierarchical channel decoding device 30 on the receiving side uses the received hierarchical data u, parity data c, and hierarchical sparse matrix G to solve the simultaneous equations so as to satisfy the equation (2), thereby correcting the error. Decrypt.

  When performing partial decoding of hierarchical data, the receiving side uses simultaneous parts of the received hierarchical data u, parity data c, and hierarchical sparse matrix G to satisfy equation (3). By solving the above, error correction and decoding are performed.

  As described above, the sparse matrix of the present invention has various variations depending on the hierarchical data, and a sparse matrix created by a method such as an LDPC code (Non-Patent Document 1) can be used as a method for creating the sparse matrix.

  The operation of the above-described hierarchical channel encoding device and hierarchical channel decoding device is constructed as a program, and is installed and executed on a computer used as the hierarchical channel encoding device and hierarchical channel decoding device. Or can be distributed via a network.

  The present invention is not limited to the above-described embodiment, and various modifications and applications can be made within the scope of the claims.

  The present invention is applicable to error correction techniques in satellite digital broadcasting, communication on the Internet, communication on mobile terminals, and the like.

It is a figure for demonstrating the principle of this invention. It is a principle block diagram of this invention. It is a figure which shows the error correction function with respect to the hierarchy coding data realizable by this invention. 1 is a basic configuration diagram of a hierarchical channel coding system according to an embodiment of the present invention. It is a figure which shows the JPEG2000 data structure of a progression order RLCP. It is a flowchart of an outline operation in an embodiment of the invention. It is a flowchart of the preparation process of the sparse matrix P of the hierarchical sparse matrix production | generation part in one embodiment of this invention. It is a flowchart of operation | movement of the hierarchy data division part in one embodiment of this invention. It is a flowchart of operation | movement of the hierarchical encoding part in one embodiment of this invention. It is a flowchart of operation | movement of the hierarchical decoding part in one embodiment of this invention. It is a figure which shows a binary erasure | elimination communication path.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Information source encoding apparatus 20 Hierarchical encoding apparatus 30 Hierarchical decoding apparatus 40 Information source decoding apparatus 50 Hierarchical data synthesizing apparatus 100 Hierarchical channel encoding system 101 Hierarchical data dividing unit 102 Hierarchical encoding means, hierarchical code Generator 103 random number signal generator 104 hierarchical sparse matrix generator, hierarchical encoder 105 transmitter, transmitter module 106 channel 107 receiver, receiver module 108 hierarchical decoder, hierarchical decoder 109 data rearranger 110 Header rewriting part

Claims (10)

In a system composed of a hierarchical channel encoding device and a hierarchical channel decoding device, a channel encoding method for protecting hierarchically encoded data from errors,
In the hierarchical channel encoder,
A hierarchical sparse matrix generation step for generating a hierarchical sparse matrix composed of a plurality of sparse matrices;
A hierarchical encoding step for obtaining parity data using the hierarchical sparse matrix for hierarchical data obtained by dividing the target data for each scalability function;
Transmitting the hierarchical data, the parity data, and the hierarchical sparse matrix to the hierarchical channel decoding device, and
In the hierarchical channel decoding apparatus,
A hierarchical decoding step of decoding data transmitted from the hierarchical channel encoder using the hierarchical sparse matrix;
A communication path encoding method characterized by: In the hierarchical sparse matrix generation step,
Performing the step of generating the hierarchical sparse matrix so that one parity data is generated for one hierarchical data;
In the hierarchical decoding step,
2. The channel coding method according to claim 1, wherein the step of decoding the hierarchical data and the parity data on a one-to-one basis using the hierarchical sparse matrix. In the hierarchical sparse matrix generation step,
Generating the hierarchical sparse matrix using one or all of the hierarchical sparse matrix defined by a matrix defined by one random sparse graph;
In the hierarchical decoding step,
The channel coding method according to claim 1 or 2, wherein a part or all of the parity data is decoded using a matrix of the hierarchical sparse matrix defined by one random sparse graph. In the hierarchical decoding step,
The channel coding method according to claim 2, wherein the received data is selectively decoded. A system comprising a hierarchical channel encoding device and a hierarchical channel decoding device for protecting hierarchically encoded data from errors,
The hierarchical channel encoder is
A hierarchical sparse matrix generating means for generating a hierarchical sparse matrix composed of a plurality of sparse matrices;
Hierarchical encoding means for obtaining parity data using the hierarchical sparse matrix for hierarchical data obtained by dividing the target data for each scalability function;
Transmitting means for transmitting the hierarchical data, the parity data and the hierarchical sparse matrix to the hierarchical channel decoding device;
The hierarchical communication path decoding device comprises:
A hierarchical channel coding system comprising hierarchical decoding means for decoding data transmitted from the hierarchical channel coding apparatus using the hierarchical sparse matrix. The hierarchical sparse matrix generating means of the hierarchical channel encoding device comprises:
Means for generating the hierarchical sparse matrix so that one parity data is generated for one hierarchical data;
The hierarchical decoding means of the hierarchical channel decoding device comprises:
6. The channel coding system according to claim 5, further comprising means for decoding the hierarchical data and the parity data on a one-to-one basis using the hierarchical sparse matrix. The hierarchical sparse matrix generating means of the hierarchical channel encoding device comprises:
Means for generating the hierarchical sparse matrix using a matrix defined by one random sparse graph for part or all of the hierarchical sparse matrix;
The hierarchical decoding means of the hierarchical channel decoding device comprises:
7. The channel coding system according to claim 5, further comprising means for decoding part or all of the parity data using the matrix of the hierarchical sparse matrix defined by one random sparse graph. The hierarchical decoding means of the hierarchical channel decoding device comprises:
The channel coding system according to claim 6, further comprising means for selectively decoding the received data.   A hierarchical channel encoding program that causes a computer to function as each means constituting the hierarchical channel encoding apparatus of the channel encoding system according to claim 6 or 7.   9. A hierarchical channel decoding program that causes a computer to function as means for constituting the hierarchical channel decoding device of the channel encoding system according to claim 6.

Images