JP2005223683A - Transmission/reception system, transmitter and transmitting method, receiver and receiving method, transmitter/receiver and transmitting/receiving method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To more easily perform more accurate communication. <P>SOLUTION: A transmitter 11 divides transmission data to generate packet data, performs EOR operation for the packet data by using a coefficient matrix in which the existence available region of a valid coefficient is limited, converts the packet data into EOR conversion data to be actually transmitted, and then, packetizes the converted data using a predetermined protocol such as UDP and transmit the data. The receiver 13 receives the packet supplied from the receiver 11 and extracts the EOR conversion data, and restores the original transmission data using the received EOR conversion data. In this case, the receiver 13 receives the EOR conversion data whose number is at least equal to or larger than the number of data required for restoring the original transmission data from the EOR data transmitted from the transmitter 11. This is applicable to a transmission system, for example. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a transmission / reception system, a transmission apparatus and method, a reception apparatus and method, a transmission / reception apparatus and method, and a program, and in particular, a transmission / reception system capable of performing transmission / reception of data more accurately and easily, The present invention relates to a transmission device and method, a reception device and method, a transmission and reception device and method, and a program.

  In recent years, with the development of broadband communication networks and the like, many streaming distributions and real-time broadcasts of data such as audio and video have been performed through networks typified by the Internet.

  Conventionally, in data distribution via such a network, for example, when data (packets) transmitted by the transmission device cannot be received by the reception device, such as data erase, There is a method of requesting the transmission apparatus to retransmit data for which the reception apparatus is insufficient when reception is started in the middle of the transmission process.

  Further, the transmission device converts the transmission data into a plurality of independent data using, for example, a Reed-Solomon code, and transmits the converted plurality of data, so that the reception device transmits the plurality of data. Among them, a method has been proposed in which a predetermined number or more of arbitrary data is received and the original transmission data is restored from the received data.

  For example, the encoder of the transmitting device uses the key I generated in the key generator to convert the output of the input symbol generator to generate an output symbol B (I) that is a Reed-Solomon value function, There is a method in which an output symbol is supplied to a receiving device, and a decoder of the receiving device converts the output symbol using a received key and recovers the input symbol (see, for example, Patent Document 1).

  Further, for example, as shown in FIG. 1, the transmission-side device repeatedly transmits the same data, and the reception-side device loses a packet, and then the data (packet group) to be transmitted next time. There is also a method using a carousel system that receives lost packets and restores data.

  FIG. 1 is a diagram illustrating an example of how a receiving device receives data. Since the transmission side device repeatedly transmits the same data 1, the reception side device repeatedly receives the same data, such as data 1A, data 1B, data 1C, and data 1D. Therefore, for example, even when the packet loss 2A occurs when the data 1A is received, the receiving side apparatus corresponds to the lost packet (packet loss 2A) of the data 1B to be received next. If the packet 2B is received, the data 1 can be restored using the packet 2B. Further, even if the packet 2B is lost when the data 1B is received, the receiving device can receive the packet 2C corresponding to the packet loss 2A of the data 1C to be received next. The data 1 can be restored using the packet 2C. Further, even when the packet 2C is lost, the receiving device only needs to receive the packet 2D. In this way, the receiving-side apparatus can receive each packet in at least one of the repeated transmissions, and if all packets can be received as a result, restores data 1. be able to.

JP 2001-189665 A

  However, in the case of the method for making a retransmission request as described above, for example, in the case of communication for transmitting data to a plurality of devices such as multicast, each device on the receiving side includes each device on the receiving side. Since the retransmission requests from the receivers are concentrated, there is a problem that the processing load on the transmission side apparatus increases, and there is a possibility that the transmission processing cannot be performed normally.

  In the case of a method using a Reed-Solomon code or the like, the data encoding process and decoding process require operations on the Galois field in order to prevent overflow, which increases the number of complex operations and increases the transmission apparatus and reception. There is a problem that the load on the apparatus is increased, and the processing performance thereof must be made high in order to transmit and receive data normally, which increases the manufacturing cost.

  Therefore, a method of reducing the amount of data to be converted can be considered in order to reduce the calculation processing on the Galois field so that the manufacturing cost is reduced, but if the amount of data transmitted by the transmitting device is reduced, the receiving device Data restoration performance may be reduced.

  Also, in the case of the carousel communication as described above, as shown in FIG. 1, when a packet loss 2A occurs when receiving data 1A, the receiving-side apparatus transmits data 1B as the next transmission. It is necessary to wait until the packet 2B corresponding to the packet loss 2A is received. That is, in this case, the reception processing time in the receiving device becomes longer by the time required to receive the data 1, and the load of the receiving processing in the receiving device increases. there were. That is, as shown in FIG. 1, when a packet loss 2A occurs, the receiving device receives the packet 2B to be transmitted next (the time required for the receiving device to receive data 1). ) It is necessary to wait and reception processing time increases. Further, when the packet 2B is lost, the receiving device needs to further extend the standby time and wait until the next packet 2C is received.

  The present invention has been made in view of such a situation, and makes it possible to transmit and receive data accurately and more easily.

  The transmission / reception system of the present invention is a transmission / reception system including a transmission device that transmits transmission data and a reception device that receives transmission data transmitted by the transmission device, and the transmission device divides the transmission data. A plurality of first data is generated by the above, a set including all the generated first data as elements is set as a whole set, and a subset consisting of any continuous element group that is a part of the whole set from the whole set , And arbitrarily select a plurality of first data from the set subset, perform an exclusive OR operation on the selected plurality of first data for each bit, and select a plurality of second data as a result of the operation. By converting the plurality of first data into a plurality of second data, packetizing each of the plurality of second data obtained, and transmitting the packet to the receiving device. The plurality of second data transmitted by the transmitting device is received by a number equal to or more than the number of first data corresponding to the second data, and a second number equal to or greater than the number of the first data corresponding to the received second data. Data is stored, a plurality of stored second data is used to generate a determinant corresponding to the second data, and the first data is restored by solving the determinant.

  The transmission apparatus according to the present invention includes a data generation unit that generates a plurality of first data by dividing transmission data, and a set that includes all the first data generated by the data generation unit as a whole set. And setting a subset of any continuous element group that is a part of the entire set from the entire set, arbitrarily selecting a plurality of first data from the set subset, and selecting the plurality of selected first data Conversion means for converting a plurality of first data into a plurality of second data by performing an exclusive OR operation on each other for each bit and obtaining second data as a result of operation for a plurality of subsets; Transmitting means for packetizing a plurality of second data obtained by converting a plurality of first data by the converting means and transmitting the packets to a receiving device is provided.

  The converting means includes a determinant generating means for generating a determinant for converting the first data into the second data using a coefficient matrix for selecting the first data to be subjected to the exclusive OR operation, and a matrix The determinant generated by the expression generation means can be provided with an operation means for obtaining the second data by calculating using an exclusive OR operation.

  The coefficient matrix is composed of an effective coefficient having a value of “1” and an invalid coefficient having a value of “0”, and the effective coefficient is set as a subset for each row of the coefficient matrix. This is a component arbitrarily selected within the range of columns in which effective coefficients can exist, and the range of each row is set so that the number of columns in the range is the same and the first column of the range is different for each row Can be done.

  A range setting unit that sets the range for each row, a determination unit that determines a value of each component of the coefficient matrix based on the range set by the range setting unit, and a coefficient matrix based on a determination result by the determination unit Coefficient matrix holding means for holding a coefficient matrix by holding all component values, and the determinant generating means generates a determinant using the coefficient matrix held by the coefficient matrix holding means. Can be.

  The converting means may convert a plurality of first data into second data larger than the number of first data.

  The second data obtained by converting the first data by the converting means further includes information adding means for adding information about the range, and the transmitting means includes a plurality of second data to which information is added by the information adding means. The two data can be packetized and transmitted to the receiving device.

  The data generation means can divide the transmission data into blocks and further divide the blocks.

  The transmission method of the present invention includes a data generation step for generating a plurality of first data by dividing transmission data, and a set including all the first data generated by the processing of the data generation step as elements. A subset consisting of any continuous element group that is a part of the entire set is set from the entire set, a plurality of first data are arbitrarily selected from the set subset, and the selected first Conversion step of converting a plurality of first data into a plurality of second data by performing an exclusive OR operation on each other bit by bit and obtaining a second data as an operation result for a plurality of subsets And a plurality of second data obtained by converting the plurality of first data by the process of the conversion step are packetized and transmitted to the receiving device. Tsu, characterized in that it comprises a flop.

  The first program of the present invention includes a data generation step for generating a plurality of first data by dividing transmission data and all the first data generated by the processing of the data generation step as elements. Set a set as a whole set, set a subset consisting of any continuous element group that is a part of the whole set from the whole set, arbitrarily select a plurality of first data from the set subset, The plurality of first data is converted into a plurality of second data by performing an exclusive OR operation on the first data for each bit and obtaining the second data as the operation result for a plurality of subsets. A conversion step and a transmission for controlling the plurality of second data obtained by converting the plurality of first data by the processing of the conversion step to be packetized and transmitted to the receiving device. To execute a control step to the computer.

  The receiving apparatus of the present invention is data transmitted by the transmitting apparatus, and each of the sets includes all first data generated by dividing the transmission data as an entire set, and the entire set from the entire set Set a subset consisting of any continuous element group that is a part of, select any number of first data from the set subset, and mutually exclusive the selected first data for each bit Receiving means for performing an OR operation and receiving a plurality of second data obtained for a plurality of subsets as an operation result of the exclusive OR operation, the number of which is equal to or more than the number of the first data corresponding to the second data; Using the holding means for holding the second data that is equal to or more than the number of the first data corresponding to the second data received by the receiving means, and using the plurality of second data held by the holding means, It corresponds to the data of 2 Generates a matrix expression, by solving the determinant, characterized in that it comprises a restoring means for restoring the first data.

  The restoring means includes a determinant generating means for generating a determinant for converting the first data into the second data using a coefficient matrix for selecting the first data to be subjected to the exclusive OR operation, and a matrix The determinant generated by the expression generating means may be provided with an arithmetic means for solving the first data as a variable and obtaining the first data.

  The coefficient matrix is composed of an effective coefficient having a value of “1” and an invalid coefficient having a value of “0”, and the effective coefficient is set as a subset for each row of the coefficient matrix. This is a component arbitrarily selected within the range of columns in which effective coefficients can exist, and the range of each row is set so that the number of columns in the range is the same and the first column of the range is different for each row Can be done.

  Coefficient matrix generation means for generating a coefficient matrix based on information about a range for arbitrarily selecting a plurality of first data to be subjected to exclusive OR operation added to the second data received by the reception means The determinant generating means can generate the determinant using the coefficient matrix generated by the coefficient matrix generating means.

  Each component of the coefficient matrix is divided into a plurality of small regions for each predetermined number of rows and columns, and based on the value of the component of each small region, the determinant generated by the determinant generating means Computation determining means for determining whether or not to perform the calculation of the corresponding part is further provided, and the calculation means is only for the part of the determinant generated by the determinant generation means that is determined to be calculated by the calculation determination means. It is possible to obtain the first data by performing an operation and solving the first data as a variable.

  The reception method of the present invention is data transmitted by a transmission apparatus, and each of the sets includes all first data generated by dividing transmission data as an entire set, and the entire set is set from the entire set. Set a subset consisting of any continuous element group that is a part of, select any number of first data from the set subset, and mutually exclusive the selected first data for each bit A logical sum operation is performed, and control is performed so as to receive a plurality of second data obtained for a plurality of subsets as a result of the exclusive logical sum operation, in a number equal to or more than the number of first data corresponding to the second data. A reception control step, a holding step for receiving second data that is controlled by the processing of the reception control step and holding the second data that is equal to or more than the number of first data corresponding to the second data, and a process of the holding step. A determinant corresponding to the second data is generated using the plurality of second data held, and a restoration step of restoring the first data by solving the determinant is included. .

  The second program of the present invention is data transmitted by the transmission device, and each of the sets is composed of all the first data generated by dividing the transmission data as a whole set, and from the whole set A subset consisting of any continuous element group that is a part of the entire set is set, a plurality of first data is arbitrarily selected from the set subset, and the selected plurality of first data is mutually bit by bit An exclusive OR operation is performed, and a plurality of second data obtained for a plurality of subsets as a result of the exclusive OR operation are received in a number equal to or more than the number of first data corresponding to the second data. A receiving control step for controlling, a holding step for holding second data equal to or more than the number of first data corresponding to the second data, which is controlled and received by the processing of the receiving control step, Generating a determinant corresponding to the second data using a plurality of second data held by the processing, and causing the computer to execute a restoration step of restoring the first data by solving the determinant .

  The transmission / reception apparatus according to the present invention includes a data generation unit that generates a plurality of first data by dividing transmission data, and an entire set including all the first data generated by the data generation unit as elements. And setting a subset of any continuous element group that is a part of the entire set from the entire set, arbitrarily selecting a plurality of first data from the set subset, and selecting the plurality of selected first data Conversion means for converting a plurality of first data into a plurality of second data by performing an exclusive OR operation on each other for each bit and obtaining second data as a result of operation for a plurality of subsets; A plurality of second data obtained by converting a plurality of first data by the converting means is packetized and transmitted to another transmitting / receiving apparatus, and a plurality of second data transmitted by other transmitting / receiving apparatuses. Receiving means for receiving data equal to or greater than the number of first data corresponding to the second data; and second data equal to or greater than the number of first data corresponding to the second data received by the receiving means. Using the holding means to hold and a plurality of second data held by the holding means, a determinant corresponding to the second data is generated, and the restoration is performed to restore the first data by solving the determinant Means.

  The transmission / reception method of the present invention includes a data generation step for generating a plurality of first data by dividing transmission data, and a set including all the first data generated by the processing of the data generation step as elements. A subset consisting of any continuous element group that is a part of the entire set is set from the entire set, a plurality of first data are arbitrarily selected from the set subset, and the selected first Conversion step of converting a plurality of first data into a plurality of second data by performing an exclusive OR operation on each other bit by bit and obtaining a second data as an operation result for a plurality of subsets Then, a plurality of second data obtained by converting the plurality of first data by the conversion step process are respectively packetized and transmitted so as to be transmitted to other transmission / reception devices. A control step, a reception control step for controlling a plurality of second data transmitted by other transmission / reception devices so as to receive more than the number of first data corresponding to the second data, and a process of the reception control step Using a holding step for holding second data that is equal to or more than the number of first data to which the second data corresponds, and a plurality of second data held by the processing of the holding step. And a restoration step of restoring the first data by generating a determinant corresponding to the second data and solving the determinant.

  The third program according to the present invention includes a data generation step of generating a plurality of first data by dividing the transmission data and all the first data generated by the processing of the data generation step. Set a set as a whole set, set a subset consisting of any continuous element group that is a part of the whole set from the whole set, arbitrarily select a plurality of first data from the set subset, The plurality of first data is converted into a plurality of second data by performing an exclusive OR operation on the first data for each bit and obtaining the second data as the operation result for a plurality of subsets. Control is performed so that the conversion step and the plurality of second data obtained by converting the plurality of first data by the processing of the conversion step are respectively packetized and transmitted to other transmission / reception devices. A transmission control step, a reception control step for controlling a plurality of second data transmitted by other transmission / reception devices so as to receive more than the number of first data corresponding to the second data, and a reception control step A holding step for holding second data that is equal to or more than the number of first data to which the second data corresponds is received by being controlled by the processing, and a plurality of second data held by the processing of the holding step And generating a determinant corresponding to the second data and solving the determinant to cause the computer to execute a restoration step of restoring the first data.

  The transmission / reception system of the present invention includes a transmission device that transmits transmission data and a reception device that receives transmission data transmitted by the transmission device. In the transmission device, the transmission data is divided. A plurality of first data is generated, and a set including all the generated first data as elements is set as a whole set, and a subset including any continuous element group that is a part of the whole set from the whole set. The plurality of first data are arbitrarily selected from the set subset, and the selected plurality of first data are subjected to exclusive OR operation for each bit, and the second data as the operation result is obtained. By obtaining a plurality of subsets, a plurality of first data is converted into a plurality of second data, and the plurality of second data obtained are respectively packetized and received by a receiving device. In the receiving device, the data transmitted by the transmitting device is set as a whole set, and each set is a whole set of all first data elements generated by dividing the transmission data. A subset consisting of any continuous element group that is a part of is set, the first data is arbitrarily selected from the set subset, and the selected first data is mutually exclusive for each bit A plurality of pieces of second data calculated for the plurality of subsets as a result of the exclusive OR operation are received and received by the number of the first data corresponding to the second data. Second data equal to or more than the number of first data corresponding to the second data is held, and a determinant corresponding to the second data is generated using the plurality of held second data, Solve determinant A result, the first data is restored.

  In the transmission apparatus and method and the first program of the present invention, a plurality of first data is generated by dividing the transmission data, and the entire set including all the generated first data as elements As a set, a subset composed of an arbitrary continuous element group that is a part of the entire set is set from the entire set, a plurality of first data are arbitrarily selected from the set subset, and a plurality of selected first data The first data is subjected to an exclusive OR operation for each bit, and the second data as the operation result is obtained for a plurality of subsets, whereby the plurality of first data is converted into the plurality of second data. The plurality of second data obtained by converting the plurality of first data is packetized and transmitted to the receiving device.

  In the receiving apparatus and method and the second program of the present invention, each of the data transmitted by the transmitting apparatus is a set including all the first data generated by dividing the transmission data. A subset consisting of any continuous element group that is a part of the entire set is set from the entire set, a plurality of first data are arbitrarily selected from the set subset, and the selected first Are obtained by performing an exclusive OR operation on each data bit by bit, and a plurality of second data obtained for a plurality of subsets as a result of the exclusive OR operation is the first data corresponding to the second data The second data having the number equal to or greater than the number of the received first data corresponding to the second data is retained, and the second data is stored using the plurality of retained second data. Matrix corresponding to There is produced, by solving the matrix equation, first data is restored.

  In the transmission / reception device and method and the third program of the present invention, a plurality of first data is generated by dividing the transmission data, and the entire set including all the generated first data as elements Set a subset of any continuous element group that is a part of the whole set from the whole set, arbitrarily select a plurality of first data from the set subset, and select the plurality of first By performing an exclusive OR operation on the data bit by bit and obtaining the second data as the operation result for a plurality of subsets, the plurality of first data is converted into the plurality of second data, The plurality of second data obtained by converting the first data is packetized and transmitted to other transmission / reception devices, and the plurality of second data transmitted by the other transmission / reception devices is the second data More than the number of corresponding first data is received, and the received second data is more than the number of first data corresponding to the second data, and a plurality of second data held is used. Thus, a determinant corresponding to the second data is generated, and the first data is restored by solving the determinant.

  According to the present invention, data can be transmitted and received. In particular, data can be transmitted and received more accurately and easily.

  Embodiments of the present invention will be described below. Correspondences between constituent elements described in the claims and specific examples in the embodiments of the present invention are exemplified as follows. This description is to confirm that specific examples supporting the invention described in the claims are described in the embodiments of the invention. Therefore, even if there are specific examples that are described in the embodiment of the invention but are not described here as corresponding to the configuration requirements, the specific examples are not included in the configuration. It does not mean that it does not correspond to a requirement. On the contrary, even if a specific example is described here as corresponding to a configuration requirement, this means that the specific example does not correspond to a configuration requirement other than the configuration requirement. not.

  Further, this description does not mean that all the inventions corresponding to the specific examples described in the embodiments of the invention are described in the claims. In other words, this description is an invention corresponding to the specific example described in the embodiment of the invention, and the existence of an invention not described in the claims of this application, that is, in the future, a divisional application will be made. Nor does it deny the existence of an invention added by amendment.

  In the present invention, a transmission device (for example, the transmission device 11 in FIG. 2) that transmits data for transmission (for example, transmission data 41 in FIG. 4) and a reception device (for example, the transmission device 11 in FIG. 2) that receives transmission data transmitted by the transmission device. For example, a transmission / reception system (for example, the communication system 10 in FIG. 2) including the receiving device 13 in FIG. 2 is provided. The transmission device generates a plurality of first data (for example, the original data 43 in FIG. 4) by dividing the transmission data (for example, step S2 in FIG. 17), and all the generated first data. The set with the element as a whole set is set, and a subset (for example, the possible range of the effective coefficient in FIG. 8) consisting of any continuous element group that is a part of the whole set is set from the whole set. By arbitrarily selecting a plurality of first data from the set, performing an exclusive OR operation on the selected plurality of first data for each bit, and obtaining the second data as the operation result for a plurality of subsets The plurality of first data is converted into a plurality of second data (for example, EOR converted data 44 in FIG. 4) (for example, step S6 in FIG. 17), and the obtained plurality of second data is respectively packetized. And receiving equipment (For example, step S8 in FIG. 17), and the receiving apparatus receives a plurality of pieces of second data transmitted by the transmitting apparatus by the number of first data corresponding to the second data (for example, FIG. 17). 20 step S81), the second data that is equal to or more than the number of the first data corresponding to the received second data is held (for example, step S83 in FIG. 20), and the plurality of held second data is used. Then, a determinant corresponding to the second data is generated, and the first data is restored by solving the determinant (for example, step S85 in FIG. 20).

  In the present invention, a transmission device (for example, the transmission device 11 in FIG. 2) for transmitting transmission data (for example, the transmission data 41 in FIG. 4) to a reception device (for example, the reception device 13 in FIG. 2) is provided. . The transmission apparatus includes a data generation unit (for example, the block division unit 22 in FIG. 3) that generates a plurality of first data (for example, the original data 43 in FIG. 4) by dividing the transmission data, and data generation A set consisting of all the first data generated by the means as an element is defined as an entire set, and a subset (for example, the effective coefficient of FIG. A plurality of first data is arbitrarily selected from the set subset, and the selected plurality of first data is subjected to an exclusive OR operation for each bit, and a second result is obtained. 4 (for example, conversion data 44 in FIG. 4) is obtained for a plurality of subsets, thereby converting a plurality of first data into a plurality of second data (for example, EOR conversion data generation in FIG. 3). Part 2 ) And a plurality of second data obtained by converting the plurality of first data by the conversion unit, respectively, and packetizing and transmitting to the receiving device (for example, the UDP packetizing unit 27 to 27 in FIG. 3) Ethernet (R) processing unit 31).

  The converting means converts the first data into the second data by using a coefficient matrix (for example, the matrix of the first term on the right side of Expression (3)) for selecting the first data to be subjected to the exclusive OR operation. A determinant generating unit (for example, the determinant generating unit 61 in FIG. 6) that generates a determinant (for example, Equation (3)) to be converted into a determinant and the determinant generated by the determinant generating unit are exclusive ORed. An arithmetic means (for example, a determinant arithmetic unit 62 in FIG. 6) that obtains the second data by performing an arithmetic operation can be provided.

  The coefficient matrix is composed of an effective coefficient having a value of “1” and an invalid coefficient having a value of “0”, and the effective coefficient is set as a subset for each row of the coefficient matrix. 8 is a component arbitrarily selected within the range of columns in which effective coefficients can exist (for example, the possible range of effective coefficients in FIG. 8), and each row range has the same number of columns in the range. Can be set so that the first column of each is different for each row.

  Range setting means (for example, the range control unit 84 in FIG. 6) that sets the range for each row, and determination means (for example, that determines the value of each component of the coefficient matrix based on the range set by the range setting means) 6, coefficient determination unit 85) and coefficient matrix holding means for holding the coefficient matrix by holding the values of all components of the coefficient matrix based on the determination result by the determination means (for example, coefficient matrix holding in FIG. 6) The determinant generating means can generate a determinant using the coefficient matrix held by the coefficient matrix holding means (for example, step S39 in FIG. 18).

  Information adding means (for example, the head information adding unit in FIG. 9) for adding information about the range (for example, head information 122 in FIG. 10) to the second data obtained by converting the first data by the converting means. 102), and the transmission means may packet each of the plurality of second data to which the information is added by the information addition means, and transmit the packetized data to the reception apparatus.

  The data generation means can divide the transmission data into blocks (for example, the block 42 in FIG. 4) and further divide the blocks.

  In the present invention, there is a transmission method of a transmission apparatus (for example, the transmission apparatus 11 in FIG. 2) that transmits transmission data (for example, the transmission data 41 in FIG. 4) to a reception apparatus (for example, the reception apparatus 13 in FIG. 2). Provided. This transmission method includes a data generation step (for example, step S2 in FIG. 17) for generating a plurality of first data (for example, the original data 43 in FIG. 4) by dividing the transmission data, and a data generation step. A set including all the first data generated by the process as an element is defined as an entire set, and a subset (for example, the effective coefficient of FIG. A plurality of first data is arbitrarily selected from the set subset, and the selected plurality of first data is subjected to an exclusive OR operation for each bit, and a second result is obtained. (For example, EOR conversion data 44 in FIG. 4) for a plurality of subsets, thereby converting a plurality of first data into a plurality of second data (for example, a step in FIG. 17). S6) and a transmission control step (for example, control is performed so that the plurality of second data obtained by converting the plurality of first data by the conversion step processing are packetized and transmitted to the receiving device, respectively) , And step S8) of FIG.

  In the present invention, transmission processing for transmitting transmission data (for example, transmission data 41 in FIG. 4) to a reception device (for example, reception device 13 in FIG. 2) is executed in a computer (for example, transmission device 11 in FIG. 2). A first program is provided. The first program includes a data generation step (for example, step S2 in FIG. 17) for generating a plurality of first data (for example, the original data 43 in FIG. 4) by dividing the transmission data, and data generation A set including all the first data generated by the processing of the step as an element is defined as an entire set, and a subset including any continuous element group that is a part of the entire set from the entire set (for example, the effective set of FIG. 8 The coefficient existence range) is set, a plurality of first data are arbitrarily selected from the set subset, and the selected plurality of first data are subjected to an exclusive OR operation for each bit and obtained as a calculation result. A conversion step (for example, FIG. 4) for converting a plurality of first data into a plurality of second data by obtaining second data (for example, EOR conversion data 44 in FIG. 4) for a plurality of subsets. 7 and step S6), and a transmission control step for controlling the plurality of pieces of second data obtained by converting the plurality of pieces of first data into packets and transmitting them to the receiving device (step S6). For example, it includes step S8) of FIG.

  In the present invention, a receiving device (for example, the receiving device 13 in FIG. 2) that receives data transmitted by a transmitting device (for example, the transmitting device 11 in FIG. 2) is provided. This receiving device is data transmitted by the transmitting device, and each of them is all the first data (for example, the original data in FIG. 4) generated by dividing the transmission data (for example, the transmitting data 41 in FIG. 4). A set having the data 43) as an element is set as an entire set, and a subset (for example, a possible range of effective coefficients in FIG. 8) including any continuous element group that is a part of the entire set is set from the entire set. A plurality of first data are arbitrarily selected from the set subset, and the selected plurality of first data are subjected to an exclusive OR operation for each bit, and a plurality of subsets are obtained as an operation result of the exclusive OR operation. Receiving means (for example, Ethernet (R) processing in FIG. 11) that receives a plurality of obtained second data (for example, converted data 44 in FIG. 4) more than the number of first data corresponding to the second data. Part 131 to a Application data extraction unit 133), and holding unit (for example, buffer unit 134 in FIG. 11) that holds the second data that is received by the receiving unit and that is equal to or more than the number of first data corresponding to the second data. , Using a plurality of second data held by the holding means, a determinant corresponding to the second data is generated, and the determinant is solved to restore the first data (for example, FIG. 13 packet data restoring units 135).

  The restoration means converts the first data to the second data using a coefficient matrix (for example, the matrix of the first term on the right side of Expression (4)) for selecting the first data to be subjected to the exclusive OR operation. The determinant generating means (for example, the determinant creating unit 153 in FIG. 13) that generates the determinant (for example, the expression (4)) to be converted into the determinant generating means, Calculation means (for example, a determinant calculation unit 157 in FIG. 13) that solves as a variable and obtains first data may be provided.

  The coefficient matrix is composed of an effective coefficient having a value of “1” and an invalid coefficient having a value of “0”, and the effective coefficient is set as a subset for each row of the coefficient matrix. 8 is a component arbitrarily selected within the range of columns in which effective coefficients can exist (for example, the possible range of effective coefficients in FIG. 8), and each row range has the same number of columns in the range. Can be set so that the first column of each is different for each row.

  Information relating to a range (for example, a possible range of effective coefficients in FIG. 8) arbitrarily selected from a plurality of first data to be subjected to exclusive OR operation added to the second data received by the receiving means. Based on (for example, the head information 122 in FIG. 10), the apparatus further includes coefficient matrix generation means (for example, the coefficient matrix generation unit 152 in FIG. 13) for generating a coefficient matrix. A determinant can be generated using the generated coefficient matrix.

  Each component of the coefficient matrix is divided into a plurality of small regions (for example, tile 163 in FIG. 16) for each predetermined number of rows and columns, and generated by a determinant generation unit based on the component values of each small region. The apparatus further includes calculation determination means (for example, the tile processing unit 156 in FIG. 13) for determining whether or not to perform the calculation of the portion corresponding to the small area of the determinant, and the calculation means is generated by the determinant generation means The first data can be obtained by calculating only the portion of the determinant determined to be calculated by the calculation determination means and solving the first data as a variable.

  In the present invention, a receiving method of a receiving device (for example, the receiving device 13 in FIG. 2) that receives data transmitted by a transmitting device (for example, the transmitting device 11 in FIG. 2) is provided. This reception method is data transmitted by the transmission device, and each of them is all the first data (for example, the original data in FIG. 4) generated by dividing the transmission data (for example, the transmission data 41 in FIG. 4). A set having the data 43) as an element is set as an entire set, and a subset (for example, a possible range of effective coefficients in FIG. 8) including any continuous element group that is a part of the entire set is set from the entire set. A plurality of first data are arbitrarily selected from the set subset, and the selected plurality of first data are subjected to an exclusive OR operation for each bit, and a plurality of subsets are obtained as an operation result of the exclusive OR operation. A reception control step (for example, FIG. 20) that controls the plurality of obtained second data (for example, converted data 44 in FIG. 4) to receive more than the number of first data corresponding to the second data. Step S 1) and a holding step for holding the second data that is controlled by the processing of the reception control step and is equal to or more than the number of the first data corresponding to the second data (for example, step S83 in FIG. 20). And generating a determinant (for example, Expression (4)) corresponding to the second data using the plurality of second data held by the process of the holding step, and solving the determinant to obtain the first A restoring step (e.g., step S85 in FIG. 20).

  In the present invention, there is provided a second program for causing a computer (for example, the receiving device 13 in FIG. 2) to execute a receiving process for receiving data transmitted by a transmitting device (for example, the transmitting device 11 in FIG. 2). The second program is data transmitted by the transmission device, and each of the first programs is all the first data (for example, FIG. 4) generated by dividing the transmission data (for example, transmission data 41 in FIG. 4). A set consisting of the original data 43) as an element is set as an entire set, and a subset consisting of any continuous element group that is a part of the entire set (for example, the range in which the effective coefficient can exist in FIG. 8) is set. Then, a plurality of first data are arbitrarily selected from the set subset, and the plurality of selected first data are subjected to an exclusive OR operation for each bit, and a plurality of portions are obtained as an operation result of the exclusive OR operation. A reception control step (for example, FIG. 4) for controlling a plurality of second data (for example, converted data 44 in FIG. 4) obtained for the set so as to receive more than the number of first data corresponding to the second data. 20 Su S81) and a holding step (for example, the step of FIG. 20) that holds the second data that is controlled by the processing of the reception control step and that is equal to or more than the number of the first data corresponding to the second data. S83) and a plurality of second data held by the holding step process are used to generate a determinant (eg, equation (4)) corresponding to the second data, and by solving the determinant, A restoring step (eg, step S85 in FIG. 20) for restoring the first data.

  In the present invention, a transmission / reception apparatus (for example, the transmission / reception apparatus 411 or the transmission / reception apparatus 413 in FIG. 31) for transmitting / receiving transmission data (for example, transmission data 41 in FIG. 4) is provided. The transmission / reception apparatus includes a data generation unit (for example, the block division unit 22 in FIG. 3) that generates a plurality of first data (for example, the original data 43 in FIG. 4) by dividing the transmission data, and data generation A set consisting of all the first data generated by the means as an element is defined as an entire set, and a subset (for example, the effective coefficient of FIG. A plurality of first data is arbitrarily selected from the set subset, and the selected plurality of first data is subjected to an exclusive OR operation for each bit, and a second result is obtained. 4 (for example, conversion data 44 in FIG. 4) is obtained for a plurality of subsets, thereby converting a plurality of first data into a plurality of second data (for example, EOR conversion data generation in FIG. 3). Part 4) and transmission means for packetizing a plurality of second data obtained by converting a plurality of first data by the conversion means and transmitting them to other transmission / reception devices (for example, the UDP packetization unit in FIG. 3) 27 to the Ethernet (R) processing unit 31) and a plurality of second data transmitted by other transmission / reception apparatuses, the receiving means (for example, FIG. 11) receiving more than the number of first data corresponding to the second data. Ethernet (R) processing unit 131 to application data extraction unit 133) and receiving means (for example, holding means for holding second data equal to or more than the number of first data corresponding to the second data received by the receiving means) 11) and a plurality of second data held by the holding means, a determinant corresponding to the second data is generated, and the first data is solved by solving the determinant. Recover Restoring means for (e.g., packet data restoring unit 135 of FIG. 13) and a.

  In the present invention, a transmission / reception method of a transmission / reception apparatus (for example, the transmission / reception apparatus 411 or the transmission / reception apparatus 413 of FIG. 31) for transmitting / receiving transmission data (for example, transmission data 41 of FIG. 4) is provided. This transmission / reception method includes a data generation step (for example, step S2 in FIG. 17) for generating a plurality of first data (for example, original data 43 in FIG. 4) by dividing the transmission data, and a data generation step. A set including all the first data generated by the process as an element is defined as an entire set, and a subset (for example, the effective coefficient of FIG. A plurality of first data is arbitrarily selected from the set subset, and the selected plurality of first data is subjected to an exclusive OR operation for each bit, and a second result is obtained. (For example, EOR conversion data 44 in FIG. 4) for a plurality of subsets, thereby converting a plurality of first data into a plurality of second data (for example, in FIG. 17). Step S6) and a transmission control step (control for packetizing the plurality of second data obtained by converting the plurality of first data by the process of the conversion step and transmitting them to other transmission / reception devices) For example, step S8 in FIG. 17 and a reception control step (for example, control is performed so as to receive a plurality of second data transmitted by other transmission / reception apparatuses so as to receive more than the number of first data corresponding to the second data) , Step S81 in FIG. 20, and a holding step (for example, FIG. 20) that holds the second data that is controlled by the process of the reception control step and is equal to or more than the number of the first data corresponding to the second data. 20 using step S83) and a plurality of second data held by the holding step, a determinant (eg, equation (4)) corresponding to the second data is generated, By solving the equation, including restoration step for restoring the first data (e.g., step S85 of FIG. 20) and.

  In the present invention, a third program for causing a computer (for example, the transmission / reception apparatus 211 or 213 of FIG. 18) to execute transmission / reception processing for transmitting / receiving transmission data (for example, transmission data 41 of FIG. 4) is provided. . The third program includes a data generation step (for example, step S2 in FIG. 17) for generating a plurality of first data (for example, the original data 43 in FIG. 4) by dividing the transmission data, and data generation A set including all the first data generated by the processing of the step as an element is defined as an entire set, and a subset including any continuous element group that is a part of the entire set from the entire set (for example, the effective set of FIG. 8 The coefficient existence range) is set, a plurality of first data are arbitrarily selected from the set subset, and the selected plurality of first data are subjected to an exclusive OR operation for each bit and obtained as a calculation result. A conversion step (for example, FIG. 4) for converting a plurality of first data into a plurality of second data by obtaining second data (for example, EOR conversion data 44 in FIG. 4) for a plurality of subsets. 7 and step S6), and transmission control for controlling the plurality of second data obtained by converting the plurality of first data by the processing of the conversion step into packets and transmitting them to other transmission / reception devices. Step (for example, step S8 in FIG. 17) and a reception control step for controlling a plurality of second data transmitted by other transmission / reception apparatuses so as to receive more than the number of first data corresponding to the second data. (For example, step S81 in FIG. 20) and a holding step (for example, holding second data equal to or more than the number of first data corresponding to the second data, which is controlled and received by the processing of the reception control step) , A determinant (for example, Expression (4)) corresponding to the second data is generated by using a plurality of second data held by the process of the holding step (Step S83 in FIG. 20). By solving the matrix equation, including restoration step for restoring the first data (e.g., step S85 of FIG. 20) and.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 2 shows a configuration example of a communication system to which the present invention is applied.

  The communication system 10 in FIG. 2 is a network represented by a transmission device 11 that transmits data, a LAN (Local Area Network), the Internet, and the like. The communication system 10 is connected to the network 12 to which the transmission device 11 is connected, and the network 12. The reception device 13 receives data transmitted from the transmission device 11.

  When the transmission device 11 acquires content data for transmission (transmission data) such as download data supplied from the outside, the transmission device 11 divides the data for packet transmission (packet data) when the transmission data is divided and transmitted. Generate and combine multiple packet data arbitrarily extracted from those packet data by bit-wise EOR operation (Exclusive OR) to create EOR-converted data for actual transmission of packet data Convert. The transmitter 11 generates a plurality of EOR converted data obtained by converting the packet data in such a manner, and packetizes and transmits them using a predetermined protocol such as UDP (User Datagram Protocol).

  Upon receiving the packet (EOR conversion data) supplied from the transmission device 11, the reception device 13 extracts EOR conversion data from the received packet, and restores the original transmission data using the received EOR conversion data. To do. At this time, the transmission device 11 transmits more EOR conversion data than the number required for the reception device 13 to restore the original transmission data to the reception device 13, and the reception device 13 transmits the transmitted EOR conversion data. At least the number of EOR conversion data necessary for restoring the original transmission data is received. As will be described later, since the EOR conversion data transmitted by the transmission device 11 is not guaranteed to be independent from each other, the reception device 13 can actually restore the original transmission data. As described above, the EOR conversion data supplied from the transmission device 11 is received more than the necessary number (a margin is ensured). When the receiving device 13 receives the EOR conversion data in this way, the EOR conversion data is generated using the EOR conversion data, and the original transmission data is restored by obtaining a solution of the determinant. To do. Details of a method for generating EOR conversion data in the transmission device 11 and a method for restoring original transmission data in the reception device 13 will be described later.

  By doing so, the reception device 13 restores the transmission data transmitted by the transmission device 11 without making a retransmission request to the transmission device 11 even if the reception device 13 does not receive all the packets transmitted by the transmission device 11. be able to.

  FIG. 3 is a block diagram illustrating a detailed configuration example of the transmission device 11. A change in transmission data processed by the transmission apparatus 11 will be described with reference to FIG.

  In FIG. 3, the data division unit 21 of the transmission device 11 divides transmission data such as content data supplied from the outside of the transmission device 11 into blocks having a predetermined data size, and supplies the blocks to the block division unit 22. That is, the data dividing unit 21 accumulates the supplied transmission data in a buffer (not shown), and each time data of a predetermined data size is accumulated, it is used as a block in the block dividing unit 22. Supply.

  For example, in the case of FIG. 4, the transmission data 41 supplied to the transmission device 11 is divided into three blocks 42 of blocks A to C by the data dividing unit 21.

  Returning to FIG. 3, similarly to the data dividing unit 21, the block dividing unit 22 further divides the transmission data in block units supplied from the data dividing unit 21 into a predetermined data size. That is, the block division unit 22 divides the transmission data in units of blocks into a data size (packet size) for packetization in the subsequent stage, and generates data for the packet (packet data). The block dividing unit 22 that has generated the packet data supplies it to the packet data holding unit 23.

For example, in FIG. 4, each block 42 is divided by the block dividing unit 22 into original data 43 that is packet data having a data size when packetized. In the case of FIG. 4, each block is divided into three original data 43. For example, the block A is divided into original data A _{1 to} A ₃ , the block B is divided into original data B _{1 to} B ₃ , and the block C is divided into original data C _{1 to} C ₃ .

  The packet data holding unit 23 sequentially holds the supplied packet data (original data), and manages each original data based on a block corresponding to each original data. Then, under the control of the data conversion control unit 26, the packet data holding unit 23 supplies the held original data to the EOR conversion data generation unit 24 in units of blocks at a predetermined timing.

  When the EOR conversion data generation unit 24 acquires the original data supplied for each block from the packet data holding unit 23, the EOR conversion data generation unit 24 performs an EOR operation using them and converts the original data to generate EOR conversion data. Specifically, in each block, the EOR conversion data generation unit 24 selects a plurality of original data from a group of original data corresponding to the block by a predetermined method, synthesizes them by EOR operation for each bit, EOR conversion data. The EOR conversion data generation unit 24 repeats such conversion processing, and generates a plurality (more than the number of original data) of EOR conversion data. A more specific generation method will be described later.

For example, in the case of FIG. 4, the EOR conversion data generation unit 24 converts the original data A _{1 to} A ₃ corresponding to the block A into six EOR conversion data A _{11 to} A ₁₆ by performing an EOR operation for each bit, The original data B _{1 to} B ₃ corresponding to the block B is subjected to EOR operation for each bit to convert it into six EOR converted data B _{11 to} B ₁₆ , and the original data C _{1 to} C ₃ corresponding to the block C are converted into bits. Every EOR operation is performed to convert the data into six EOR conversion data C _{11 to} C ₁₆ .

At this time, the EOR conversion data generation unit 24 first converts the original data A _{1 to} A ₃ corresponding to the block A to generate EOR conversion data A _11, and the original data B _{1 to} B ₃ corresponding to the block B B ₃ is converted to generate EOR conversion data B _11, and original data C _{1 to} C ₃ corresponding to block C are converted to generate EOR conversion data C ₁₁ . Next, the EOR conversion data generation unit 24 proceeds to the second process for each block, converts the original data A _{1 to} A ₃ corresponding to the block A to generate EOR conversion data A _12, and generates the block B EOR converted data B ₁₂ is generated by converting the original data B _{1 to} B ₃ corresponding to, and EOR converted data C ₁₂ is generated by converting the original data C _{1 to} C ₃ corresponding to the block C. Then, c repeats the conversion process of each block in order, for each block, the third conversion process, the fourth conversion process,...

Therefore, in the case of FIG. 4, when the EOR conversion data 44 is arranged in the order of generation, A ₁₁ , B ₁₁ , C ₁₁ , A ₁₂ , B ₁₂ , C ₁₂ , A ₁₃ , B ₁₃ , C ₁₃ , A ₁₄ , B ₁₄ , the _{C 14, a 15, B 15} , C 15, a 16, B 16, C 16.

  In general, when a packet is transmitted in a network or the like, a receiving side sometimes loses a plurality of consecutive packets in a burst manner (cannot be received). Therefore, when the transmission device 11 sequentially outputs the EOR conversion data, the packet loss may be concentrated on one block due to such bursty packet loss. As will be described later, in order to restore the original data, it is necessary to receive a certain number of packets (the EOR converted data). Therefore, if the packet loss is biased in this way, a specific block cannot be restored. There is a fear.

  Therefore, as described above, the transmission device 11 sequentially executes the conversion process for each block once to transmit the EOR conversion data of each block in the mixed order, thereby transmitting the packet. Even when bursty packet loss occurs, it is possible to suppress the loss from being concentrated on a specific block, and it is possible to increase resistance to bursty packet loss. That is, the transmission device 11 can transmit data to the reception device 13 more accurately. Therefore, the communication system 10 can realize more accurate data transmission / reception.

  The EOR conversion data generation unit 24 supplies EOR conversion data obtained by converting the original data to the application data generation unit 25. The EOR conversion data generation unit 24 is information related to this conversion processing, and supplies additional information added to the EOR conversion data to the application data generation unit 25 in association with the EOR conversion data.

  The application data generation unit 25 adds a header including additional information supplied from the EOR conversion data generation unit 24 to the EOR conversion data supplied from the EOR conversion data generation unit 24 to generate application data. Details of the contents of the header will be described later. The application data generation unit 25 supplies the generated application data to the UDP packetization unit 27. At this time, the application data generation unit 25 supplies output information indicating that the application data has been generated and output to the data conversion control unit 26.

  The data conversion control unit 26 determines whether or not all packets have been generated based on the output information supplied from the application data generation unit 25, and generation of packets (EOR conversion data) has not yet been completed. Is determined, the packet data holding unit 23 and the EOR conversion data generation unit 24 are controlled to generate new EOR conversion data.

  The processes so far are the application layer, presentation layer, and session layer in the OSI layer model (Open Systems Interconnection layer model), which is a model of the seven-layer network protocol structure established by ISO (International Organization for Standardization). It is processing of.

  The UDP packetization unit 27 supplied with the application data packetizes it based on the UDP protocol, and supplies the generated UDP packet to the buffer 28.

  As shown in FIG. 5, the UDP packet 51 generated in the UDP packetization unit 27 is composed of a 64-bit UDP header 52 and application data 53. The UDP header 52 includes information such as a transmission source port number, a transmission destination port number, a data length of the application data 53, and a checksum. The data size of the application data is a predetermined value that is arbitrarily determined, but is usually about 1 Kbyte.

  Returning to FIG. 3, the buffer 28 is configured by a semiconductor memory such as a RAM (Random Access Memory), accumulates UDP packets supplied from the UDP packetizing unit 27, and based on the control of the output control unit 29, Output accumulated UDP packets.

  The output control unit 29 sequentially acquires the UDP packets stored in the buffer 28 at a predetermined timing, and supplies them to an IP (Internet Protocol) processing unit 30. The processes by the UDP packetizing unit 27 to the output control unit 29 are the processes of the transport layer in the OSI hierarchical model.

  The IP processing unit 30 performs processing of the network layer in the OSI hierarchical model, and converts each UDP packet supplied from the output control unit 29 into an IP packet based on the IP protocol, and Ethernet (R) (Ethernet (R)) This is supplied to the processing unit 31. The Ethernet (R) processing unit 31 performs data link layer and physical layer processing based on the Ethernet (R) standard, further packetizes the IP packet supplied from the IP processing unit 30, and transmits the packet to the receiving device 13. As such, it is output to the network 12 via the cable 32.

  That is, the EOR conversion data 44 generated in the EOR conversion data generation unit 24 is packetized in the order of generation as shown in FIG. 4 and is output to the network 12 as a transmission packet 45. The packet output from the Ethernet (R) processing unit 29 is supplied to the receiving device 13 via the network 12.

  In this way, since the original data obtained by dividing the transmission data is not transmitted as it is, it is converted into EOR conversion data that includes the contents of multiple original data (part of the original data in the block) and then transmitted. The apparatus 11 enables the receiving apparatus 13 to restore the original data (transmission data) if the receiving apparatus 13 acquires a predetermined number or more of a plurality of packets transmitted from the transmitting apparatus 11. Can do. That is, the transmission device 11 can increase the tolerance of the reception device 13 against packet loss, and can transmit data so that data transmission / reception can be performed more accurately and easily.

  Note that the number of blocks 42, original data 43, EOR conversion data 44, and the like for one transmission data 41 shown in FIG. 4 is an example, and the number of pieces of generated data is the number shown in FIG. Of course it may be other than. Further, the transmission order of the transmission packets 45 is also an example, and the transmission packets 45 may be transmitted in an order other than that shown in FIG.

Next, details of the EOR conversion data generation unit 24 of FIG. 3 will be described. In the following, the transmission device 11 generates _N original data D (D _{1 to} D _N ) from one block of transmission data, converts the N original data D, and generates M pieces of original data D. A case where EOR conversion data S (S _{1 to} S _M ) is generated will be described. FIG. 6 is a block diagram illustrating a detailed configuration example of the EOR conversion data generation unit 24.

  In FIG. 6, the original data D for one block supplied from the packet data holding unit 23 in FIG. 3 is supplied to the determinant generation unit 61. The determinant generation unit 61 that has acquired the original data D for one block requests the coefficient matrix holding unit 71 for a coefficient matrix used to generate a determinant for generating EOR conversion data.

  The coefficient matrix is a matrix of coefficients of original data having M rows and N columns with all components being “1” or “0”, and is a matrix for selecting original data to be subjected to EOR calculation. As will be described later, the EOR conversion data is obtained by integrating a coefficient matrix and a matrix having each original data as a component. Accordingly, each row of the coefficient matrix corresponds to different EOR conversion data, and each component of each row is integrated with the original data as a coefficient of different original data. That is, each EOR conversion data is an EOR calculation result for each bit of the original data integrated with the component (coefficient) whose coefficient matrix value is “1”. In the following, a component (coefficient) having a value of “1” in the coefficient matrix is referred to as an effective coefficient, and a component (coefficient) having a value of “0” is referred to as an invalid coefficient.

  In addition, the coefficient matrix has a limited range in which effective coefficients can exist in each row. The range is different from one another for each row, and the effective coefficient is a component arbitrarily selected within the range. In other words, each EOR conversion data in the block is a combination of a plurality of original data arbitrarily extracted from different ranges of all of the aligned original data by a bitwise EOR operation. It is.

  In the coefficient matrix holding unit 71, such a coefficient matrix is generated and held in advance.

  The generation of this coefficient matrix will be described. As described above, in this coefficient matrix, in each row, possible ranges of effective coefficients are different for each row, and effective coefficients are components arbitrarily selected within the range, and other components are invalid. It is a coefficient. Therefore, in order to generate such a coefficient matrix, the EOR conversion data generation unit 24 first sets the possible range of effective coefficients in each row, and further arbitrarily selects a component within the range, As an effective coefficient, its value is set to “1”, and other components are set as invalid coefficients and its value is set to “0”. That is, the generation of the coefficient matrix is performed in units of rows.

  Specifically, the pseudo-random number generation unit 81 of the EOR conversion data generation unit 24 first performs an operation as shown in Expression (1) on the first row, and uses the linear congruence method described above. A pseudo-random number for determining the start position of the valid coefficient existence range is generated, and the value is supplied to the start position determination unit 82.

  In Expression (1), the constant A in the first term on the right side and the constant C in the third term on the right side are predetermined natural numbers, the variable rand in the second term on the right side is the pseudorandom number obtained last time, and the first term on the left side. The variable rand is a pseudo-random number to be obtained this time. The initial value of the variable rand is “0”, and the pseudo random number (the value of the variable rand in the first term on the left side) obtained this time is the remainder when the calculation result of the expression on the right side is divided by “N−bandwidth” It is. The bandwidth is a predetermined constant and is a value indicating the size (number of columns) of the range in which the effective coefficient can exist. This bandwidth is stored in advance in a ROM or RAM (both not shown) built in the pseudorandom number generator 81. The constant N is the number of original data of each block, and the information is also stored in advance in the ROM or RAM built in the pseudorandom number generator 81.

Therefore, the pseudo-random number (the value of the variable rand in the first term on the left side) obtained by Expression (1) is a value that is randomly determined between the value “0” and the value “N−1−bandwidth”. The value of this pseudo random number, that is, the start position (column) of the valid coefficient existence range in the row to be processed by the coefficient matrix, that is, the column of which the valid coefficient existence range is the column number “0”. This indicates from which column (first column) to column number “N-1-band width” (the (N-band width) column) starts.
When such a pseudo random number is generated, the pseudo random number generation unit 81 supplies it to the head position determination unit 82. The head position determination unit 82 stores the pseudo random number as the head position of the corresponding row of the coefficient matrix to be generated. As described above, the pseudo-random number generation unit 81 generates pseudo-random numbers for the second and subsequent rows of the coefficient matrix to be generated, and the head position determination unit 82 holds these pseudo-random numbers corresponding to the respective rows. .

  When the pseudo random numbers (leading positions) for all the rows of the coefficient matrix to be generated are held, the range control unit 84 requests a bandwidth value from the width setting unit 83. The width setting unit 83 holds a bandwidth value (number of columns) that is the width (number of columns) of the effective coefficient existence range described above in a built-in ROM or RAM (both not shown). Based on the request of the range control unit 84, the bandwidth value is supplied to the range control unit 84.

  The range control unit 84 that has acquired the bandwidth value next requests a pseudo-random number of each row, that is, information on the start position of the possible range of effective coefficients, to the start position determination unit 82. The head position determination unit 82 supplies the held pseudorandom number to the range control unit 84 based on the request. The range control unit 84 that has acquired the pseudorandom number determines the possible range of the effective coefficient in each row of the coefficient matrix to be generated based on the pseudorandom number and the bandwidth value acquired from the width setting unit 83.

  Next, the pseudo random number generation unit 81 performs an operation as shown in Expression (2), and uses a linear congruential modulo operation to set the value of the coefficient (each component of the coefficient matrix). , And supplies the value to the range controller 84.

  In equation (2), the variable L in the first term on the right side and the third term on the right side is predetermined seed data. For example, this kind of data may be common to all rows, or may be different from each other in each row. The variable rand in the second term on the right side is a pseudo random number obtained last time, and the variable rand in the first term on the left side is a pseudo random number obtained this time. The initial value of the variable rand is “0”, and the pseudorandom number (the value of the variable rand in the first term on the left side) obtained this time is, for example, a modulo arithmetic result of the expression on the right side performed by 32-bit processing.

  The pseudo random number generation unit 81 performs the calculation of the above formula (2) and generates a pseudo random number, and supplies the pseudo random number to the range control unit 84. When the range control unit 84 acquires the pseudo random number, based on the pre-determined effective coefficient existence range, whether or not the position of the coefficient corresponding to the pseudo random number is within the effective coefficient existence range. The pseudo-random number is supplied to the coefficient determination unit 85 only when it is within the range. When the value is out of the range, the value “0” is supplied to the coefficient determination unit 85 instead of the pseudo random number.

  The coefficient determination unit 85 determines the value of each component corresponding to the pseudo random number based on the acquired value of the pseudo random number. At this time, for example, the coefficient determination unit 85 provides a threshold value for the acquired pseudo random number value, and if the acquired pseudo random number value is equal to or greater than the threshold value, the value of the component to which the pseudo random number corresponds. Is smaller than the threshold value, the value of the component is determined as “0”, and the determination result is supplied to the coefficient matrix holding unit 71 and held.

  At that time, the coefficient determination unit 85 acquires the coefficient generation probability p held in the coefficient generation probability holding unit 86, and determines a threshold value based on the value of the coefficient generation probability p. That is, the coefficient determination unit 85 sets the threshold value so that the ratio of the effective coefficients becomes the value of the coefficient generation probability p within the possible range of the effective coefficients of each row described above.

  The coefficient generation probability p is a predetermined constant greater than “0” and smaller than “1”, and the ratio of the effective coefficient is within the possible range of the effective coefficient in each row. It is a value that specifies whether or not. That is, for example, when the value of the coefficient generation probability p is “0.5”, the coefficient determination unit 94 determines that the number of valid coefficients and the number of invalid coefficients are approximately within the possible range of valid coefficients in each row. The threshold value is set so as to be one-to-one (approximate). For example, when it is assumed that the value of the pseudo random number is distributed over the entire 32 bits with a uniform probability, the coefficient determination unit 85 sets the threshold value to “2147483648”.

  As described above, the coefficient matrix holding unit 71 is supplied with the determination result (“0” or “1”) determined by the coefficient determination unit 85. Each unit performs the above processing for each coefficient, determines the value of the coefficient one by one, and the coefficient matrix holding unit 71 holds these determination results as each component of the coefficient matrix. Holds all components of the coefficient matrix.

  As described above, the determinant generation unit 61 that has acquired the original data D for one block requests the coefficient matrix holding unit 71 for the coefficient matrix thus generated and held in the coefficient matrix holding unit 71. . In response to this request, the coefficient matrix holding unit 71 supplies the held coefficient matrix to the determinant generation unit 61.

  When the determinant generation unit 61 acquires the coefficient matrix, the determinant generation unit 61 generates a determinant for generating EOR conversion data for the block using the coefficient matrix and the original data for one block. Specifically, the matrix generation unit 61 generates a determinant as shown in the following expression (3).

In Equation (3), the first term on the right side is a coefficient matrix of M rows and N columns, which is a coefficient matrix supplied from the coefficient matrix supply unit 71. The second term on the right side is a matrix of the original data D (D _{1 to} D _N ), and the matrix on the left side is EOR conversion data S (S _{1 to} S _M ) calculated by this determinant. In Equation (3), bitwise EOR operation is performed instead of normal addition. The EOR operation will be described later.

  When such a determinant is generated, the determinant generator 61 supplies this determinant to the determinant calculator 62. The determinant is the same as a so-called simultaneous equation. The same applies to the following.

  Further, at this time, the determinant generation unit 61 confirms the valid coefficient existence range in each row of the coefficient matrix, associates the row number with the leading position (column) of the valid coefficient existence range for each row, To the determinant operation unit 62 together with the determinant. The coefficient information and the head information are supplied to the receiving device 13 together with the EOR conversion data, and are used to generate a coefficient matrix when restoring the original data from the EOR conversion data.

The determinant calculation unit 62 calculates a determinant as shown in Expression (3). That is, the determinant calculating unit 62 replaces each column component of the row in each row of the coefficient matrix of the first term on the right side with each row component (original data D _{1 to} D _N ) in the second term on the right side to which they correspond. The respective integration results are used, and the EOR calculation is performed for each bit corresponding to each other, thereby calculating one EOR calculation result for each row.

  Here, the calculation procedure in the integration of two matrices in the case of a normal decimal number will be described. First, each row of the first term matrix is integrated with each column of the second term matrix. Specifically, each row of the first term matrix and each column of the second term matrix are integrated for each corresponding component. For example, describing the first row of the first term, the component in the i-th column of that row is integrated with the component in the i-th row of each column in the second term. When integration is performed in this way, all integration results of each component are then added for each integration. For example, if the matrix of the first term is an N-column matrix and the matrix of the second term is an N-row matrix, in each multiplication of each row of the matrix of the first term and each column of the matrix of the second term Since the integration results of N components are respectively obtained, the N integration results are added, and the addition result is the integration result of the matrix row of the first term and the matrix column of the second term. It is said.

  In the determinant of Equation (3) described above, the calculation is basically performed according to the same calculation procedure as that described above. However, since the original data D and the EOR conversion data S are binary bit strings, instead of the normal addition process of the integration result of components in the calculation procedure in the case of the decimal number, the EOR calculation for each bit is performed. Is used.

In the order of the calculation procedure, the determinant calculation unit 62 converts each row of the coefficient matrix of the first term on the right side of Equation (3) to the matrix (original data of the second term on the right side) as in the case of the decimal number described above. D _{1 to DN} are integrated with the columns of the matrix. Specifically, each column component of each row of the coefficient matrix of the first term on the right side is integrated with each component (original data D _{1 to} D _N ) of the matrix of the second term on the right side. Since the value of each component of the coefficient matrix of the first term on the right side is “1” or “0”, does the integration result of each component become each component (original data D _{1 to} D _N ) of the second term on the right side? Or “0” (specifically, the number of bits is the same as that of the original data D, and the value of each bit is “0”).

  Next, the determinant operation unit 62 combines the integration results of these components by EOR operation for each bit corresponding to each other as described above. Specifically, the EOR operation will be described. The value of the EOR operation result of the value “0” and the value “0” is “0”, and the value of the EOR operation result of the value “0” and the value “1”. Is “1”, the value of the EOR operation result of the value “1” and the value “0” is “1”, and the value of the EOR operation result of the value “1” and the value “1” is “0”. It is. Therefore, for example, the EOR operation result of the bit string “11110000” and the bit string “10101010” is “01011010”. The determinant calculation unit 62 repeats such an EOR calculation and synthesizes all the integration results of the components. The EOR calculation result is used as one EOR conversion data S.

The above calculation is repeated for each row of the coefficient matrix of the first term on the right side of Equation (3), and EOR conversion data S _{1 to} S _M shown as the components of the matrix of the first term on all the left sides are obtained. .

As described above, since the value of each component of the coefficient matrix is “0” or “1”, the integration result of each component and the original data D _{1 to} D _N is the value of the second term on the right side. It becomes a component (original data D _{1 to} D _N ) or “0”. Therefore, in other words for the integration, determinant calculation unit 62, the integration of its components and the original data D ₁ to D _N, select the original data D used for generating the EOR conversion data S (extraction) and ing.

  Further, as described above, since the integration results of the respective components are synthesized by the EOR operation for each bit corresponding to each other, the value of each bit of the synthesis result (EOR operation result) is subjected to the EOR operation corresponding to the bit. When the number of integration results with a value “1” is an odd number, the value is “1”. When the number of integration results with a value “1” is an even number, “0” "become. Therefore, in other words, for the EOR operation, the determinant operation unit 62 adds up the values of the respective bits of the integration results for each component described above for each bit corresponding to each other, and the number of integration results whose value is “1”. Is an odd number, the value of the bit of the EOR conversion data S corresponding to that bit is “1”, and when the number of integration results with the value “1” is an even number, the EOR conversion data corresponding to that bit The value of the bit of S is “0”.

That is, the determinant calculation unit 62, from the original data D ₁ to D _N, select the original data D by a predetermined method, for each original data D that selected, the bits corresponding to each other, the value is "1 ”Is an odd number, the value of the bit of the EOR conversion data S corresponding to that bit is“ 1 ”, and when the number of the value“ 1 ”is an even number, the EOR conversion data S corresponding to that bit is Each EOR conversion data S is obtained with the value of the bit of “0” as “0”.

The determinant calculation unit 62 performs the calculation as described above, obtains EOR conversion data S (S _{1 to} S _M ), and supplies the solution to the solution output unit 63. At this time, the determinant arithmetic unit 62 supplies the above-described coefficient information and head information to the solution output unit 63 together with the obtained EOR conversion data.

  As described above, the EOR conversion data generation unit 24 converts the original data to generate EOR conversion data as shown in FIG.

In FIG. 7, the original data 43 (original data D _{1 to} D _N ) of one block is packet size data like the packet data 91-1 to 91-N, and as described above, EOR conversion data generation The unit 24 converts the data into EOR conversion data 44. The EOR conversion data 44 is composed of packet data 92-1 to 92-M. The packet data 92-1 to 92-M are data generated using the original data D _{1 to} D _N (packet data 91-1 to 91-N) and the coefficient matrix described above, respectively, and the expression (3) Thus, the result of EOR calculation for each bit of the plurality of original data 43 extracted by the above-described method from the original data D _{1 to} D _N (packet data 91-1 to 91-N).

  The solution output unit 63 supplies the solution calculated as described above to the application data generation unit 25 in FIG. 3 as EOR conversion data S at a predetermined timing. The solution output unit 63 supplies the above-described coefficient information and head information together with the EOR conversion data to the application data generation unit 25 in FIG.

  Next, the possible range of effective coefficients will be described.

  In the communication system 10, the receiving device 13 restores the original data by using the same coefficient matrix as the coefficient matrix used for generating the EOR transformed data in the transmitting device 11, as will be described later. If there is a bias in the original data used to generate the data and the bias can be specified, the receiving device 13 clearly has an effective coefficient in the coefficient matrix when restoring the original data. The calculation corresponding to the range not to be performed can be omitted.

Therefore, the transmitting apparatus 11, when selecting the source data used to generate each EOR conversion data, sets the predetermined range, among the original data D ₁ to D _N, is selected from the original data within the range Like that. That is, in each row of the coefficient matrix, a column range in which an effective coefficient exists is set, and the effective coefficient is a coefficient of a column arbitrarily selected within the column range.

  However, in order to improve the tolerance of the receiving device 13 against packet loss (loss of EOR converted data transmitted by the transmitting device 11), the restoration performance is not affected by the lost EOR converted data (which EOR converted data is lost). However, it is desirable that the restoration performance does not change. Therefore, when generating the EOR conversion data in the transmission device 11, the position of the effective coefficient in the coefficient matrix is not biased (so that each EOR conversion data is generated using the arbitrarily selected original data). There is a need to. More specifically, in the coefficient matrix, the effective coefficients are arranged uniformly in each column, and the arrangement of the effective coefficients is independent from each other for each row (no regularity). Desired.

  Therefore, for example, as illustrated in FIG. 8, the transmission device 11 randomly sets the position of the valid coefficient existence range for each row. FIG. 8 is a diagram schematically showing the position of the effective coefficient existence range in the equation (3).

8, the vertical axis represents the generated showed EOR conversion data S ₁ to S _M, the horizontal axis represents the original data D ₁ to D _N. In addition, the possible range of effective coefficients is indicated by black lines. That, EOR converted data S ₁ to S _M of the vertical axis, respectively, of the original data D ₁ through D _N on the abscissa, were arbitrarily extracted from the original data set within the range indicated by the black line ( This is the EOR operation result for each bit of the original data (corresponding to the effective coefficient).

As described above, since the width (band width) of the effective coefficient existence range is a predetermined value, the widths of the black lines in FIG. 8 are all the same length. Further, as described above, the start position of the effective coefficient existence range is randomly determined between the first column to the (N-bandwidth) column of the coefficient matrix, so the black line in FIG. The first position (the leftmost position) of each is the original data randomly determined from the original data D _{1 to the} original data D _{(N-bandwidth)} . Therefore, when the EOR conversion data S _{1 to} S _M are aligned as shown in FIG. 8, the positions of the effective coefficient existence ranges (black lines) corresponding to the respective EOR conversion data are uniformly arranged without any deviation. Will be.

  That is, as shown in FIG. 8, in the coefficient matrix used when generating the EOR conversion data, the effective coefficients are arranged uniformly in each column, and the effective coefficients are arranged for each row. Independent (no regularity).

However, in FIG. 8, when the EOR conversion data S _{1 to} S _M are aligned based on the start position of the effective coefficient existence range, the effective coefficient existence range is set. It will be biased near the diagonal component. That is, since the coefficient matrix has characteristics such as a band matrix, the receiving device 13 uses such a coefficient matrix when restoring the original data, so that there is clearly no effective coefficient. And an operation corresponding to the range can be omitted.

In other words, the EOR conversion data generation unit 24 defines a set including all original data D (original data D _{1 to} D _N ) as an entire set, and any continuous element that is a part of the entire set from the entire set. An effective coefficient existence range is set as a subset consisting of a group, and a plurality of original data D is arbitrarily selected from the set effective coefficient existence range, and the selected plurality of original data D is mutually exclusive for each bit. and sum operation, operation presence range results EOR converted data S a plurality of effective coefficients is (i.e., each row of the coefficient matrix) by obtaining the original data D ₁ to D _N and EOR conversion data S ₁ to S _M Convert to

  Next, application data will be described.

  FIG. 9 is a block diagram illustrating a detailed configuration example of the application data generation unit 25 of FIG.

  Initially, the block counter 104 is initialized before the packet data is supplied.

  The EOR conversion data and coefficient information supplied from the EOR conversion data generation unit 24 are supplied to the coefficient information addition unit 101 of the application data generation unit 25. The coefficient information adding unit 101 adds the acquired coefficient information to the acquired EOR conversion data and supplies it to the head information adding unit 102. Further, the top information supplied from the EOR conversion data generation unit 24 is added to the top information adding unit 102. The head information adding unit 102 adds head information to the EOR conversion data to which coefficient information is added, and supplies it to the block number adding unit 103. The block number adding unit 103 acquires the count value of the block counter 104 and adds it to the EOR conversion data to which coefficient information and head information are added as a block number. When the block counter 104 supplies the count value to the block number adding unit 103, the block counter 104 counts up and updates the count value. The block counter 104 is a maximum value (N-1) counter, and increments the count value by "1" every time it counts up. When counting up when the count value is the maximum value (N−1), the block counter 104 updates the count value to “0”. The block number adding unit 103 that adds the block number to the EOR conversion data supplies it to the application data output unit 105. The application data output unit 105 outputs the EOR conversion data added with the coefficient information, the head information, and the block number supplied from the block number adding unit 103 to the UDP packetizing unit 27 in FIG. 3 as application data. .

  FIG. 10 is a diagram illustrating a configuration example of application data output from the application data output unit 105.

  In FIG. 10, application data 110 includes a header 111 and EOR conversion data 112. The header 111 includes a block number 121, head information 122, and coefficient information 123. That is, the application data is obtained by adding the coefficient information 123, the head information 122, and the block number 121 as the header 111 to the EOR conversion data 112.

  The coefficient information 123 is information for the reception device 13 to generate each component (coefficient) of the row corresponding to the EOR conversion data 112 of the coefficient matrix. For example, the coefficient information 123 is a random number at the time of generating a coefficient in the pseudo-random number generation unit 81. It is composed of generated seed data (for example, 32 bits = 4 bytes). In addition, the coefficient information 123 may be association information that indirectly specifies a coefficient, such as a line number or a packet number (for example, 2 bytes when generating 60,000 types), or an EOR conversion thereof. The data 112 may be a bit string of the corresponding coefficient itself (for example, 50 bytes when the bandwidth is 400), or a bit string obtained by compressing the bit string with a run length or the like. The receiving device 13 generates a row component of a coefficient matrix corresponding to each EOR conversion data based on such coefficient information.

  The head information 122 is information for the reception device 13 to generate each component (coefficient) of the row corresponding to the EOR conversion data 112 of the coefficient matrix. The head information 122 indicates the head position of the valid coefficient existence range in the coefficient matrix. It is information to show. The receiving device 13 recognizes the position of the range where the effective coefficient can exist in the generated coefficient matrix based on such head information, and determines a coefficient whose calculation is omitted based on the position.

  Returning to FIG. 9, when the application data as described above is output, the application data output unit 105 supplies output information indicating that the application data has been output to the data conversion control unit 26.

  Based on the output information, the data conversion control unit 26 determines whether all application data corresponding to the transmission data has been output. If it is determined that all application data has not been output, The packet data holding unit 23 and the EOR conversion data generation unit 24 are controlled so as to generate conversion data.

  As described above, instead of transmitting a plurality of original data obtained by dividing transmission data, the original data is converted into EOR-converted data and transmitted. It is possible to restore the original data by receiving a predetermined number or more of the converted data. Therefore, the transmission device 11 can transmit data more easily and more accurately.

  Further, since the EOR conversion data generated for each block is transmitted in the order in which the EOR conversion data of each block is mixed, the transmission device 11 can improve the tolerance of the reception device 13 against bursty packet loss.

  Furthermore, since the range in which the effective coefficient can exist can be limited in each row of the coefficient matrix, the reception device 13 can reduce the calculation amount of the original data restoration process.

  Furthermore, since the transmission device 11 transmits information such as coefficient information, head information, and block information, the reception device 13 can easily perform the original data restoration process.

  Next, the data receiving side will be described.

  FIG. 11 is a block diagram illustrating a detailed configuration example of the reception device 13 of FIG.

  A packet transmitted to the receiving device 13 via the network 12 is supplied to the Ethernet (R) processing unit 131 via the cable 130. The Ethernet (R) processing unit 131 performs data link layer and physical layer processing based on the Ethernet (R) standard, extracts an IP packet from the supplied packet, and supplies the IP packet to the IP processing unit 132.

  The IP processing unit 132 performs network layer processing based on the IP protocol, extracts a UDP packet from the supplied IP packet, and supplies the UDP packet to the application data extraction unit 133. The application data extraction unit 133 performs transport layer processing based on the UDP protocol, extracts application data from the UDP packet, and supplies it to the buffer unit 134.

  The buffer unit 134 distributes and holds the application data for each corresponding block based on the block number added to the application data, and supplies the application data to the packet data restoration unit 135.

  The packet data restoration unit 135 restores the original data in units of blocks based on the conversion data included in the supplied application data, and supplies it to the block data restoration unit 136. The block data restoration unit 136 restores block data using the original data (packet data) supplied from the packet data restoration unit 135 and supplies the block data to the reproduction output unit 137. The reception data output unit 137 restores reception data (original transmission data) using the block data, and outputs it to the outside of the reception device 13.

  A packet transmitted from the transmission device 11 is transmitted toward the reception device 13, but may not be received by the reception device 13 due to various causes. For example, as illustrated in FIG. 12, when nine transmission packets 141 are transmitted from the transmission packet 141 transmitted by the transmission device 11, the reception device 13 restores the original data and restores one block data. Suppose that the receiving device 13 lost the packets of the numbers 3, 7, and 8 of the transmission packet 141. That is, in the received packet 142, packets with packet numbers 3, 7, and 8 are lost. When three arbitrary packets are lost from the transmission packet 141 in this way, the receiving apparatus 13 receives three packets corresponding to the same block as the lost packet in addition to the packets with the packet numbers 1 to 9 described above (see FIG. In the case of 12, if the packet numbers 10 to 12) are received, the original data can be restored. That is, if the receiving device 13 can receive a necessary number of packets from among the packets transmitted by the transmitting device 11, the receiving device 13 can be any packet as long as the packet corresponds to the same block. The data can be restored. Details of the restoration processing will be described later.

  Note that the processing executed by the modules after the buffer unit 134 described above is processing of the application layer, the presentation layer, and the session layer in the OSI hierarchical model.

  As described above, the receiving device 13 can restore the original data (data for transmission) by acquiring a predetermined number or more of arbitrary packets among the packets transmitted from the transmitting device 11. That is, the receiving device 13 can improve the resistance against packet loss by restoring the original data using the converted data transmitted by the transmitting device 11, more accurately transmitting and receiving data, and Data can be received to make it easier.

  FIG. 13 is a block diagram illustrating a detailed configuration example of the packet data restoration unit 135.

  In FIG. 13, the buffer unit 134 accumulates application data for each block, and when a block that accumulates a sufficient number of application data to restore the original data occurs, supplies the application data to the restoration control unit 151. To do.

  The restoration processing unit 151 supplies the application data to the coefficient matrix creation unit 152 and creates a coefficient matrix for the block. When the coefficient matrix creating unit 152 creates the coefficient matrix, the coefficient matrix creating unit 152 supplies the coefficient matrix to the determinant creating unit 153. The restoration control unit 151 also supplies application data to the determinant creation unit 153 to create a determinant for restoring the original data for the block.

  The determinant creation unit 153 creates Formula (4), which is a determinant for restoring the original data, based on the application data supplied from the restoration control unit 151 and the coefficient matrix supplied from the coefficient matrix creation unit 152. To do.

In Equation (4), EOR calculation is performed instead of normal addition processing. Each component Q _{1 to} Q _K of the matrix of the first term on the right side of Equation (4) represents K (M>K> N) EOR conversion data received by the receiving device 13. Each component P _{1 to} P _N of the matrix of the second term on the left side indicates N original data to be obtained. The matrix of the first term on the left side of Equation (4) is the coefficient matrix of M rows and N columns held in advance in the coefficient matrix supply unit 123, that is, the coefficient matrix shown in the first term on the right side of Equation (2) and each matrix This is a coefficient matrix cut out from a coefficient matrix having the same component value (the same coefficient matrix as the coefficient matrix used when the transmission apparatus 11 generates EOR conversion data). However, the row components of the coefficient matrix shown in Expression (4) are configured only by the row components corresponding to the packets (EOR converted data Q _{1 to} Q _K ) received by the receiving device 13.

Therefore, the receiving device 13 can obtain the solutions of the original data P _{1 to} P _N by solving the determinant (simultaneous equations) shown in the equation (4).

  When the determinant creating unit 153 generates the determinant represented by the equation (4), the determinant creating unit 153 supplies the determinant to the determinant holding unit 154 and the determinant aligning unit 155. The determinant aligning unit 155 that has acquired the determinant replaces the rows of each term of the determinant based on the possible range of the effective coefficient of the coefficient matrix of the determinant and aligns them as shown in FIG.

  FIG. 14 is a diagram schematically illustrating the positions of possible ranges of effective coefficients in Expression (4) arranged by replacing rows.

In FIG. 14, as in FIG. 8, the vertical axis represents the received EOR conversion data Q _{1 to} Q _K , and the horizontal axis represents the original data P _{1 to} P _N to be obtained. In addition, the possible range of effective coefficients is indicated by black lines. That is, the EOR conversion data Q on the vertical axis is arbitrarily extracted from the original data group within the range indicated by the black line among the original data P _{1 to} P _{N on} the horizontal axis (corresponding to the effective coefficient). This is the EOR operation result for each bit of the original data.

  The determinant aligning unit 155 replaces the row component of each term of the determinant created by the determinant creating unit 153, and as shown in FIG. The head position is made to go from the first column to the Nth column in order from the first row to the Kth row.

  That is, the positions of effective coefficients in the coefficient matrix of the first term on the left side of Equation (4) at this time are concentrated near the diagonal components of the coefficient matrix as shown in FIG. 15A, for example, and the coefficient matrix is a band matrix. become that way. In practice, the size (number of rows and columns) of this coefficient matrix is very large, making the characteristics of the band matrix stronger as shown in FIG. 15B. In FIG. 15B, the range indicated by hatching is the range where the effective coefficient can exist, and the range indicated by double-headed arrow is the bandwidth 161. Note that the effective coefficients are randomly arranged within the possible range of the effective coefficients. By aligning the determinants in this way, the position of the effective coefficient is biased in the coefficient matrix. Therefore, the receiving apparatus 13 can omit the calculation in the area outside the possible range of the effective coefficient (the white area in FIG. 15B) as indicated by the hatch in FIG. 15B.

  When the determinants are aligned in this way, the determinant aligning unit 155 holds the aligned determinant in the determinant holding unit 154 and supplies the information to the tile processing unit 156. Based on the information, the tile processing unit 156 requests the determinant holding unit 154 for the aligned determinant and acquires it.

  When the tile processing unit 156 acquires the aligned determinant, the tile processing unit 156 sets a tile which is a unit of an area for controlling the arithmetic processing for the coefficient matrix of the acquired determinant, and restores the original data for each tile. It is determined whether or not to perform the operation. That is, the tile processing unit 156 sets a plurality of small regions (tiles) obtained by dividing the region of all the components of the coefficient matrix in the row direction and the column direction by a predetermined unit. FIG. 16 shows an example of a tile. In FIG. 16, a region 162 indicates a region of all components of the coefficient matrix. As described with reference to FIG. 15, in this area 162, an area where the effective coefficient of the bandwidth 161 can exist (area sandwiched between two oblique straight lines) is set, and the area 162 is within this area. Effective coefficients are distributed.

  The tile processing unit 156 divides the area 162 into small areas having a predetermined size as shown by the tile 163. In the case of the example in FIG. 16, the tile processing unit 156 sets the size of the region 162 to 2048 columns × 2112 rows and the size of the bandwidth 161 (the number of columns) to 400. A small area (for example, the number of columns 128 × the number of rows 132) whose one side is about a fraction of the bandwidth 161 (for example, 128) is set as a tile.

  Note that the size of the tile is set in advance and held in the tile processing unit 156. When such a tile is set, the tile processing unit 156 determines whether or not to calculate each tile. The presence / absence of the calculation is determined by the position of the tile. For example, when the tile exists in the vicinity of the valid coefficient existence range, the tile processing unit 156 determines that the calculation is performed on the tile, and the tile processing unit 156 A partial determinant corresponding to the tile component in the matrix is extracted and supplied to the determinant operation unit 157. Note that the tile processing unit 156 determines the next tile without processing the tile for a tile that is determined not to be calculated.

  In FIG. 16, tiles indicated by hatches are tiles set so as to perform calculations, and are set for 7 tiles in the horizontal direction. That is, the tile processing unit 156 sets a tile having a slightly larger range than the bandwidth 161 of the range in which the effective coefficient can exist as a tile to be calculated. The other tiles shown in white in FIG. 16 are set as tiles for which no operation is performed.

  That is, the tile in the area 162 is a processing unit of arithmetic processing in which the determinant calculating unit 157 calculates the determinant. At this time, since restoration of the original data requires more conversion data than the original data, the size of each tile is set so that the number of rows is larger than the number of columns of each tile.

  By setting the tile processing unit 156 as described above, the determinant calculating unit 157 calculates only the tiles indicated by hatches including the areas where the effective coefficients can exist as shown in FIG. It is possible to reduce the amount of calculation compared to performing calculation processing on the entire coefficient matrix in 162.

  As described above, when the masking process is performed on the determinant and the region to be operated is limited, the tile processing unit 156 supplies the partial determinant to the determinant calculating unit 157.

Determinant calculation unit 157 obtains the type of the partial matrix equation shown in equation (4), the determinant, for example, using the Gaussian elimination to obtain the solution of the original data D ₁ to D _N, The obtained original data D _{1 to} D _N are supplied to the solution output unit 158. A specific example of the Gaussian elimination method will be described later.

  The solution output unit 158 supplies the solution as original data to the block data restoration unit 136 of FIG.

  As described above, when the packet data restoration unit 135 performs the restoration process, the reception device 13 can easily restore the original data without making a retransmission request to the transmission device 11. In addition, as described with reference to FIG. 12, the receiving device 13 can restore the original data when obtaining a predetermined number or more of the converted data among the converted data transmitted by the transmitting device 11. So, for example, even if packets are lost, it is only necessary to arbitrarily receive more packets than the number of lost packets, and restoration processing can be performed at a higher speed than when packets are transmitted by the carousel method. The reception process can be executed at a higher speed. As a result, even if the receiving device 13 starts receiving from the packet in the middle of the packet group transmitted by the transmitting device 13, the original data is restored if the conversion data of a predetermined number or more is acquired. can do. Furthermore, since the receiving device 13 can restore the original data by EOR calculation, for example, the original data can be restored without performing a complicated calculation such as a calculation on a Galois field in the case of Reed-Solomon code. In addition, the restoration process can be performed at high speed, and the manufacturing cost of the receiving device 13 can be reduced.

  Next, the process flow of each unit described above in the transmission apparatus 11 will be described.

  When the supply of transmission data from the outside is started, the transmission device 11 starts data transmission processing. A data transmission process by the transmission apparatus 11 of FIG. 3 will be described with reference to a flowchart of FIG.

  First, when transmission data is supplied from the outside, the data dividing unit 21 of the transmission device 11 generates block data by dividing the transmission data into blocks in step S1. In step S2, the block dividing unit 22 further generates original data by further dividing the generated block data into packet data.

  The packet data holding unit 23 holds all packet data (original data) corresponding to one transmission data generated in the block dividing unit 22 in step S3.

  In step S4, the data conversion control unit 26 initializes the value of the variable j as the packet number of the EOR conversion data to be generated, and sets it to “0”, for example. In step S5, the data conversion control unit 26 initializes the value of the variable i as the block number of the EOR conversion data to be generated, and sets it to “0”, for example.

  In step S6, the data conversion control unit 26 controls the packet data holding unit 23 and the EOR conversion data generation unit 24 to generate EOR conversion data corresponding to the current packet number and block number values. Then, the generation processing of the EOR conversion data of the block number i and the packet number j using the original data group of the block corresponding to the block number i held by the packet data holding unit 23 is started. The EOR conversion data generation unit 24 is controlled by the data conversion control unit 26, performs EOR conversion data generation processing, performs EOR calculation using the supplied original data group, and assigns the specified block number i and packet number j. Generate EOR conversion data. Details of the EOR conversion data generation processing will be described later.

  When the EOR conversion data generation unit 24 generates EOR conversion data of the designated block number i and packet number j, the application data generation unit 25 performs application data generation processing in step S7, and generates the generated EOR conversion data. Are added with headers such as the coefficient information, the head information and the block number described above to generate application data. Details of the application data generation process will be described later.

  In step S8, the UDP packetization unit 27 converts the generated application data into a UDP packet, the buffer 28 holds the UDP packet, and the output control unit 29 reads the UDP packet held at a predetermined timing. The IP processing unit 30 converts the read UDP packet into an IP packet, and the Ethernet (R) processing unit 31 further converts the IP packet into a packet for transmission and transmits it to the network 12 via the cable 32 (network 12). To the receiving device 13 connected via

  In addition, when the data conversion control unit 26 confirms that the EOR conversion data is output based on the output information generated by the application data generation unit 25, the data conversion control unit 26 sets “1” to the value of the block number i in step S9. In step S10, it is determined whether or not the processing has been completed for all blocks. When the value of the block number i is equal to or less than the number of blocks divided by the block dividing unit 22 and it is determined that the processing has not been completed for all blocks, the data conversion control unit 26 returns the process to step S6, and next Generate EOR conversion data for the block.

  As described above, the processes of Steps S6 to S10 are repeated, and if it is determined in Step S10 that the process has been completed for all blocks, the data conversion control unit 26 proceeds to Step S11 and sets the value of the packet number j. “1” is added, and in step S12, it is determined whether or not processing has been completed for all packets. As shown in FIG. 4, the data conversion control unit 26 performs control so that EOR conversion data corresponding to each block is generated one by one. When the processing of all the blocks is completed, the EOR corresponding to each block is again performed. Control is performed to generate conversion data one by one. The data conversion control unit 26 repeats such control, and generates a predetermined number of EOR conversion data corresponding to each block.

  That is, in step S12, the data conversion control unit 26 determines whether or not the processing has been completed for all the packets by determining whether or not the value of the packet number j is larger than the predetermined number. judge. If it is determined that the processing has not been completed for all packets, the data conversion control unit 26 returns the processing to step S5 and repeats the subsequent processing, thereby allowing each unit to perform EOR conversion corresponding to the first block. Repeat from data generation. As described above, the processing from step S5 to step S12 is repeated, and if it is determined in step S12 that the processing has been completed for all packets, the data conversion control unit 26 ends the data transmission processing.

  The data transmission process is executed as described above. As a result, the transmission device 11 can generate redundant data at high speed without performing heavy processing such as calculation of Galois field, and can transmit data easily and accurately. Further, since the transmission apparatus 11 rearranges the transmission order of the packet data in units of blocks as described above, even if a bursty packet loss occurs, the lost packets are concentrated on a specific block. It can be suppressed and the resistance to bursty packet loss can be increased. Furthermore, when the transmission device 11 performs the transmission process in this manner, the reception device 13 that receives the packet transmitted by the transmission device 11 increases the manufacturing cost by performing a high-speed restoration process with high restoration performance. It can be realized without. That is, the communication system 10 can perform more accurate communication more easily.

  Next, the EOR conversion data generation process executed in step S6 of FIG. 17 will be described with reference to the flowchart of FIG.

  First, in step S31, the pseudo random number generation unit 81 of the EOR conversion data generation unit 24 generates a pseudo random number for setting the leading position of the valid coefficient existence range. The start position determination unit 82 determines the start position of the valid coefficient existence range based on the pseudo random number and holds the information. In step S32, the range control unit 84 can make an effective coefficient exist based on the information on the head position held by the head position determining unit 82 and the information on the width of the possible range of the effective coefficient held by the width setting unit 83. Determine the range.

  In step S33, the pseudo random number generation unit 81 generates a pseudo random number for coefficient setting. In step S34, the range control unit 84 determines whether or not the position of the coefficient (component) to which the generated pseudorandom number corresponds is within the possible range of the effective coefficient determined in step S32. If it is determined that it is within the possible range, the process proceeds to step S35. In step S35, the coefficient determination unit 85 sets the coefficient value to “0” based on the coefficient generation probability p held by the coefficient generation probability holding unit 86 and the pseudo random number for coefficient setting generated by the pseudo random number generation unit 81. ”Or“ 1 ”, the coefficient value is set to the determined value, and the process proceeds to step S37. If it is determined in step S34 that the position of the coefficient corresponding to the pseudo random number is outside the valid coefficient existence range, the range control unit 84 proceeds to step S36, and the coefficient matrix holding unit 71 Regardless of the value of the random number, the coefficient value is set to “0”, and the process proceeds to step S37.

  In step S37, the coefficient matrix holding unit 71 holds the set coefficient value, and in step S38, the coefficient matrix holding unit 71 determines whether the coefficient matrix is completed based on the held coefficient. If it is determined that the coefficient matrix has not been completed, the coefficient matrix holding unit 71 returns the process to step S33 and causes each unit to repeat the subsequent processes, thereby setting the value of the next coefficient.

  If it is determined in step S38 that the coefficient matrix has been completed, the coefficient matrix holding unit 71 advances the process to step S39. In step S39, the determinant generation unit 61 uses the coefficient matrix held by the coefficient matrix holding unit 71 to create a determinant for converting the original data. The determinant computing unit 62 computes the determinant in step S40. The solution output unit 63 outputs the determinant solution, which is the calculation result of the determinant operation unit 62, as EOR conversion data, ends the EOR conversion data generation process, and returns the process to step S6 in FIG. At this time, the solution output unit 63 also outputs coefficient information and head information together with the EOR conversion data as described above.

  As described above, the EOR operation is performed using the original data to generate the EOR conversion data, so that the EOR conversion data generation unit 24 can generate the EOR conversion data easily and at high speed. As a result, the transmission device 11 can generate EOR conversion data at high speed without performing heavy processing such as calculation of a Galois field, and can transmit data easily and accurately. In addition, when the transmission device 11 performs the transmission process in this manner, the reception device 13 that receives the packet transmitted by the transmission device 11 increases the manufacturing cost by performing a high-speed restoration process with high restoration performance. It can be realized without. That is, the communication system 10 can perform more accurate communication more easily.

  Next, details of the application data generation process executed in step S7 of FIG. 17 will be described with reference to the flowchart of FIG.

  First, in step S61, the coefficient information adding unit 101 adds coefficient information corresponding to the EOR conversion data to the EOR conversion data generated by the EOR conversion data generation unit 24 as a header. In step S62, the head information adding unit 102 adds head information corresponding to the EOR conversion data to the EOR conversion data to which the coefficient information is added as a header. Further, in step S63, the block number adding unit 103 adds the block number, which is the count value counted by the block counter 104, to the EOR conversion data to which coefficient information and head information are added as a header.

  When the block number is added to the EOR conversion data by the block number adding unit 103, the block counter 104 is controlled by the block number adding unit 103 and counts up the count value.

  In step S65, the application data output unit 105 outputs the EOR conversion data with the coefficient information, the head information, and the block number added as a header as application data, and generates output information for the application data in step S66. Output. When the process of step S66 ends, the application data output unit 105 ends the application data generation process, returns the process to step S7 of FIG. 16, and performs the subsequent processes.

  As described above, when the application data generation unit 25 adds the block number to the EOR conversion data, the reception device 13 can identify the block of the received EOR conversion data and can process each block. .

  In addition, when the application data generation unit 25 adds coefficient information or head information to the EOR conversion data, the reception device 13 generates a determinant so as to correspond to the case of the EOR conversion data generation processing in the transmission device 11, The original data can be restored by executing the restoration process of the original data using the determinant.

  Next, the flow of processing executed by the receiving device 13 will be described.

  First, data reception processing by the reception device 13 of FIG. 11 will be described with reference to the flowchart of FIG.

  When a packet is transmitted from the transmission device 11 to the reception device 13, and the packet is supplied to the Ethernet (R) processing unit 221 via the cable 210, the Ethernet (R) processing unit 131, in step S81, Receive the packet. The Ethernet processing unit 131 that has received the packet extracts an IP packet from the packet. The IP processing unit 132 extracts a UDP packet from the IP packet.

  The application data extraction unit 133 supplied with the UDP packet extracts application data from the UDP packet and supplies it to the buffer unit 134 in step S82. In step S83, the buffer unit 134 distributes and holds the supplied application data for each block. In step S <b> 84, the buffer unit 134 determines whether or not the original data can be restored in each block based on the number of pieces of application data stored in each block. When it is determined that the application data of each block is sufficient to restore the original data and the original data can be restored, the buffer unit 134 converts the application data into the packet data restoration unit 135 for each block. And the process proceeds to step S85. In step S85, the packet data restoration unit 135 performs original data restoration processing using the supplied application data, and restores the original data for each block. Details of the original data restoration process will be described later.

  When the original data restoration process ends, the packet data restoration unit 135 supplies the restored original data to the block data restoration unit 136, and the process proceeds to step S86. In step S <b> 86, the block data restoration unit 136 combines the supplied original data, restores the block data, and supplies the restored block data to the reception data generation unit 137. In step S87, the reception data generation unit 137 holds the block data supplied from the block data restoration unit 136, and when all the block data is held, combines them to generate reception data (original transmission data) ( (Restoration) and output it to the outside of the receiving device 13, and the data receiving process is terminated.

  In step S84, when it is determined that the application data of any block is not held in a sufficient number to restore the original data and the original data cannot be restored, the buffer unit 134 The process proceeds to step S88, and it is determined whether reception of the packet has been completed. If it is determined that the packet reception has not ended, the buffer unit 134 returns the process to step S81 and repeats the subsequent processes.

  If it is determined in step S88 that the packet reception process has ended, the buffer unit 134 proceeds to step S89, performs error processing, and ends the data reception process.

  By executing the data reception process as described above, the reception device 13 can realize a high-speed restoration process with a high restoration performance without increasing the manufacturing cost. Data can be received accurately.

  Next, details of the original data restoration processing executed for the original data of each block in step S85 in FIG. 20 will be described with reference to the flowchart in FIG.

  In step S101, the restoration control unit 151 of the packet data restoration unit 135 initializes the value of the block number i designating the block to be restored, and sets it to “0”, for example. Then, the restoration control unit 151 requests the application data group of the set block number i from the buffer unit 134 and acquires them. When the application data group of block number i is acquired, the restoration control unit 151 extracts a header including coefficient information, head information, and the like from these application data and supplies the header to the coefficient matrix creation unit 152. At the same time, the restoration control unit 151 supplies the acquired application data to the determinant creating unit 153 based on a request from the determinant creating unit 153.

  The coefficient matrix creation unit 152 supplied with the coefficient information, the head information, etc. creates a coefficient matrix for the block number i in step S102 and supplies it to the determinant creation unit 153. In step S103, the determinant creating unit 153 creates a determinant using the supplied application data and coefficient matrix, and supplies the determinant to the determinant holding unit 154 to hold it. Further, the determinant creation unit 154 supplies the head information included in the header of the application data to the determinant alignment unit 155. In step S104, the determinant sorting unit 155 replaces the determinant held in the determinant holding unit 154 with each row in each term and changes the row number based on the possible range of effective coefficients. Align expressions.

  When the determinants are aligned, the determinant aligning unit 155 supplies the information to the tile processing unit 156. In step S105, the tile processing unit 156 acquires the determinant held in the determinant holding unit 156, sets a tile for the coefficient matrix of the determinant, and based on the possible range of effective coefficients, Set whether to calculate each tile. When the tile is set, the tile processing unit 156 initializes and sets the value of the tile number k, which is a variable for selecting the tile to be processed, to “0” in step S106.

  Then, in step S107, the tile processing unit 156 determines whether or not to calculate the tile based on the setting of whether or not to calculate the tile for the tile with the tile number k. If it is determined that the tile operation is to be performed, the tile processing unit 156 proceeds with the process to step S108, extracts a partial determinant corresponding to the tile number k from the overall determinant, and supplies it to the determinant operation unit 157. To do. The determinant computing unit 157 computes the supplied partial determinant using a Gaussian elimination method or the like, and obtains a solution of each equation constituting the partial determinant. The determinant computing unit 157 supplies the obtained solution to the solution output unit 158. In step S109, the solution output unit 158 outputs the acquired partial determinant solution as original data, and the process proceeds to step S110.

  If it is determined in step S107 that the partial determinant operation corresponding to the tile to be processed is not performed, the tile processing unit 156 omits the processes in steps S108 and S109 and performs the process in step S110. Proceed.

  In step S110, the tile processing unit 156 adds “1” to the value of the tile number k, and proceeds to step S111. In step S111, the tile processing unit 156 determines whether or not processing has been completed for all tiles based on the value of the tile number k, and there are unprocessed tiles, and processing has been completed for all tiles. If it is determined that there is not, the process returns to step S107, and the subsequent processes of the block are repeated.

  If it is determined in step S111 that the processing has been completed for all tiles, the tile processing unit 156 supplies the information to the restoration control unit 151, and the process proceeds to step S112. In step S112, “1” is added to the value of block number i, and the process proceeds to step S113. If it is determined in step S113 that the processing has not been completed for all blocks, the restoration control unit 151 returns the processing to step S102 and repeats the subsequent processing. If it is determined in step S113 that the conversion process has been performed for all blocks, the restoration control unit 151 ends the original data restoration process and returns the process to S85 in FIG.

  By performing the restoration process as described above, the packet data restoration unit 135 can realize a restoration process with high speed and high restoration performance without increasing the manufacturing cost. Therefore, the receiving device 13 can receive data more easily and more accurately.

  Next, details of the solution processing executed in step S108 of FIG. 21 will be described with reference to the flowcharts of FIGS.

  In step S131, the determinant operation unit 157 supplied with the determinant for restoring the original data refers to the coefficient matrix included in the determinant, and whether or not there is a component having a value of “1” in all the columns. Determine whether or not. If at least one component having a value “1” does not exist in all the columns of the coefficient matrix, the lost original data corresponding to that column cannot be calculated. This determination process is performed.

  If it is determined in step S131 that at least one component having a value “1” exists in all the columns, the determinant computing unit 157 advances the process to step S132 and sets the value of the variable i to “1”. Start the forward erasure of the Gaussian elimination method. Forward erasure is a process of transforming a determinant by performing an EOR operation between rows of a coefficient matrix so as to reduce the number of components having a value of “1” in the determinant coefficient matrix. A specific example of the forward erasure process will be described later.

  The determinant computing unit 157 that has started the forward erasure process advances the process to step S133, and sequentially detects rows in which the component value of the i-th column is “1” from the first row of the target row. The target row is a row that is a target of forward erasure processing in the coefficient matrix. As will be described later, this target row changes each time the forward erasure process proceeds. In the initial state, the target rows are all rows.

  In step S134, the determinant calculation unit 157 determines whether or not a plurality of rows are detected by the process of step S133. If it is determined that a plurality of rows are detected, the first detected row in step S135 and 2 The EOR operation is performed on each component with the rows detected after the th. At this time, the EOR operation is performed so that the other terms of the determinant correspond to the EOR operation of the coefficient matrix.

  The determinant computing unit 157 that performed the EOR computation between the rows advances the process to step S136. Further, when it is determined in step S134 that only one row has been detected for the row whose component value in the i-th column is “1” (at least one component having a value “1” exists in each column). Therefore, one line is detected), the process of step S135 is omitted, and the process proceeds to step S136.

  In step S <b> 136, the determinant calculating unit 157 holds the detected row (remaining by the EOR calculation) whose component value in the i-th column is “1” for solution and excludes it from the target row. Then, the determinant operation unit 157 advances the process to step S137, adds “1” to the value of the variable i, and determines whether the value of the variable i is greater than K in step S138. When it is determined that the value of the variable i is equal to or less than K (that is, there is an unprocessed column), the determinant arithmetic unit 157 returns the process to step S133 and performs subsequent processing for the next column. repeat.

  By repeating the above processing, the determinant computing unit 157 advances the forward erasure processing, performs the processing on all the columns, and reduces the number of components whose coefficient matrix value is “1”.

  If it is determined in step S138 that the value of the variable i is greater than K, the determinant calculation unit 157 advances the processing to step S141 in FIG.

  In step S141 in FIG. 23, the determinant computing unit sets the value of the variable i to K, and starts backward substitution using the retained row. Backward substitution is a process in which a solution of the original data is obtained by a coefficient matrix in which a component having a value of “1” is reduced by the above-described forward elimination, and the obtained solution is sequentially substituted into the equation of another row. . A specific example of the backward substitution process will be described later.

  In step S142, the determinant computing unit 157 that has started backward substitution solves the equation of the i-th retained row by forward erasure. As will be described later, the determinant operation unit 157 solves the equation by substituting the previously obtained solution into the variable (original data) of the equation.

  The determinant computing unit 157 that has finished the process of step S142 proceeds to the process of step S143, subtracts “1” from the value of the variable i, and determines whether or not all solutions (original data) have been obtained in step S144. judge. When it is determined that there are unknown variables (original data) and not all solutions are obtained, the determinant computing unit 157 returns the processing to step S141 and repeats the subsequent processing until all the solutions are obtained. .

  If it is determined in step S144 that all the solutions have been obtained, the determinant computing unit 157 ends the solution finding process.

  If it is determined in step S131 in FIG. 22 that there is no component having a value of “1” in all columns, as described above, since all solutions cannot be obtained, the determinant computing unit 157 The processing proceeds to step S145 in FIG. 23, error processing is performed, a message indicating that the lost packet cannot be recovered is output, and the solution processing ends.

  As described above, the determinant operation unit 157 solves the determinant using the Gaussian elimination method, so that the packet data restoration unit 135 can easily restore the lost original data. Thereby, the receiving device 11 can realize high-speed restoration processing without increasing the manufacturing cost, and can receive data more easily and more accurately.

  Next, a specific example of the Gaussian elimination method described above will be described with reference to FIGS. Here, for simplification of description, a case will be described in which four original data D1 to D4 are obtained using five EOR conversion data X1 to X5. At this time, it is assumed that the determinant coefficient matrix is a 5 × 4 matrix of components as shown in FIG. 24 and the determinant creation unit 153 creates the determinant as shown in FIG.

  When the determinant shown in FIG. 24 is acquired, the determinant computing unit 157 first executes the process of step S131 in FIG. 22, and the value “in the first column of the left side in FIG. It is determined whether or not the component “1” exists. In the coefficient matrix of FIG. 24, since all the columns have components having a value of “1”, the original data D1 to D4 can be obtained.

  Next, the determinant computing unit 157 starts forward erasure and sets the first column (i = 1) as the processing target as the first processing. Then, the determinant computing unit 157 pays attention to the first column with all the rows as the target rows, and sequentially detects the rows in which the component having the value “1” exists from the first row of the target rows. In the case of FIG. 24, the row in which the component value in the first column indicated by the dotted line frame 171 is “1” is detected, and the first, third, and fifth rows indicated by the dotted line frames 181 to 183 are detected. They are detected in that order.

  The determinant computing unit 157 that detects the row in which the value of the component in the first column is “1” then performs an EOR operation between the components in the first detected row and the second and subsequent detected rows. . In the case of FIG. 24, the determinant computing unit 157 performs EOR computation on the first row indicated by the dotted line frame 181 and the third row indicated by the dotted line frame 182 for each component. Similarly, the determinant computing unit 157 performs EOR computation on the first row indicated by the dotted line frame 181 and the fifth row indicated by the dotted line frame 183 for each component. By this EOR operation, the value of each component of the coefficient matrix becomes as shown in the coefficient matrix (first term on the left side) shown in FIG.

  At this time, the determinant operation unit 157 performs EOR operation so that the matrix on the right side also corresponds to the coefficient matrix so that the determinant is established. In the case of FIG. 24, the determinant operation unit 157 performs the first row (X1) indicated by the dotted frame 191 and the third row (X3) indicated by the dotted frame 192 on the right-side EOR conversion data matrix. Is further EOR-calculated, and the first row (X1) indicated by the dotted frame 191 and the fifth row (X5) indicated by the dotted frame 193 are EOR-calculated. By this EOR operation, the value of each component of the matrix of EOR conversion data becomes like the matrix (right side) of the EOR conversion data shown in FIG.

  As described above, when the row in which the value of the first column component in the target row is “1” is set to only the first row, the determinant computing unit 157 proceeds to the second process. The second column of the coefficient matrix is the processing target (i = 2), the first row is retained, and the second to fifth rows are the target rows. Then, in the same manner as in the first time, the determinant calculation unit 157 sets the row in which the component having the value “1” exists for the second column of the coefficient matrix to the first row of the target row (in the case of FIG. 25, Detection is performed sequentially from the second row of the coefficient matrix. In the case of FIG. 25, the row in which the value of the component in the second column indicated by the dotted frame 201 is “1” is detected, and the second and fifth rows of the coefficient matrix indicated by the dotted frames 211 and 112 are detected. Eyes are detected in that order.

  The determinant computing unit 157 that detects the row in which the value of the component in the second column is “1” is the same as in the first time, and each of the first detected row and the second and subsequent rows is detected. EOR the components. In the case of FIG. 25, the determinant computing unit 157 performs EOR computation on the first row indicated by the dotted frame 211 and the fifth row indicated by the dotted frame 212 for each component. By this EOR operation, the value of each component of the coefficient matrix becomes as shown in the coefficient matrix (first term on the left side) shown in FIG.

  At this time, the determinant operation unit 157 performs the EOR operation so that the matrix on the right side also corresponds to the coefficient matrix so that the determinant is established, as in the first case. In the case of FIG. 25, the determinant operation unit 157 performs the second row (X2) indicated by the dotted frame 221 and the fifth row (X5 + X1) indicated by the dotted frame 222 on the right-side EOR conversion data matrix. ) And EOR. By this EOR operation, the value of each component of the matrix of EOR conversion data becomes like the matrix (right side) of the EOR operation data shown in FIG.

  As described above, when the value of the component in the second column in the target row is set to only the second row, the determinant calculating unit 157 proceeds to the third process. The third column of the coefficient matrix is the processing target (i = 3), the second row is further retained, and the third to fifth rows are the target rows. Then, in the same manner as in the first time, the determinant calculation unit 157 sets the row in which the component having the value “1” exists for the third column of the coefficient matrix to the first row of the target row (in the case of FIG. 26, Detection is performed sequentially from the third row of the coefficient matrix. In the case of FIG. 26, the row in which the value of the component in the third column indicated by the dotted line frame 231 is “1” is detected, and only the fourth row of the coefficient matrix indicated by the dotted line frame 241 is detected.

  Therefore, in this case, the determinant operation unit 157 omits the EOR operation. At this time, since the determinant computing unit 157 omits the EOR operation, the EOR operation is also omitted from the right-side matrix. In the case of FIG. 26, in the matrix of EOR conversion data on the right side, the fourth row (X4) indicated by the dotted frame 251 corresponds to the fourth row of the coefficient matrix, but the EOR operation is omitted. Therefore, the value of each component of the matrix of each term of the determinant does not change as shown in FIG. 26 by this third processing, and is the same as in FIG.

  Next, the determinant operation unit 157 proceeds to the fourth process, sets the fourth column of the coefficient matrix as a processing target (i = 4), further holds the fourth row detected last time, The target line is the fifth and fifth lines. Then, in the same manner as in the first time, the determinant calculation unit 157 sets the row in which the component having the value “1” exists for the fourth column of the coefficient matrix to the first row of the target row (in the case of FIG. 27, Detection is performed sequentially from the third row of the coefficient matrix. In the case of FIG. 27, the row in which the value of the component in the fourth column indicated by the dotted line frame 261 is “1” is detected, and only the third row of the coefficient matrix indicated by the dotted line frame 271 is detected ( The value of the component in the fifth row and the fourth column is “0”).

  Accordingly, in this case as well, the determinant operation unit 157 omits the EOR operation as in the third process. At this time, since the determinant computing unit 157 omits the EOR operation, the EOR operation is also omitted from the right-side matrix. In the case of FIG. 27, in the matrix of redundant data for calculation on the right side, the third row (X3 + X1) indicated by the dotted frame 281 corresponds to the third row of the coefficient matrix, but the EOR calculation is omitted. Therefore, the value of each term of the determinant does not change even by the fourth processing. The determinant computing unit 157 further holds the third row detected in the fourth process.

  As described above, when forward erasure is performed on all the columns, the determinant computing unit 157 ends forward erasure and then starts backward substitution. Due to the forward erasure, only the component having the value “1” in the last held row has one in the last column. The values of the components in the other columns are converted to “0” by the EOR operation so far. Similarly, the number of components having a value of “1” in the second row held from the last is two or less, and the number of rows held before that is three or less. As described above, the number of components having a value of “1” is Y or less in the Y-th row held from the end.

  Therefore, the determinant operation unit 157 obtains one solution from the last retained row by backward substitution, substitutes the solution in the reverse order of the retained order, and obtains all solutions.

  In this way, by using the Gaussian elimination method, the determinant operation unit 157 can solve the determinant by a simple operation such as an EOR operation, so the lost packet restoration unit 228 can easily recover the lost original data. And can be restored at high speed.

  FIG. 28 is a diagram illustrating an example of a relationship between occurrence of packet loss and content reception time.

  FIG. 28A shows an example of the relationship between the packet loss and the reception time in the case of the Rucell method described in FIG. For example, as described above, in the case of carousel transmission, since content is repeatedly transmitted as it is, such as content 291, content 292, and content 293, for example, content distributed for the first time Assuming that a packet in the period indicated by the packet loss occurrence period 291A is lost during reception of each of the packets 291, the receiving apparatus has a period 292A corresponding to the packet loss occurrence period 291A in the content 292 that is the second delivery. You have to wait until you receive That is, in this case, the reception time by the receiving device increases by one period of the content 291 due to packet loss.

  On the other hand, as described above, the transmission device 11 converts the content into the EOR conversion data 301 and then repeatedly transmits the content, so that the reception device 13 in the EOR conversion data 301 as illustrated in FIG. Even if the packet loss occurrence period 301A occurs, the original data (contents) can be restored by receiving the same number of packets as the number of lost packets. Therefore, the reception device 13 of the communication system 10 in FIG. 2 can suppress an increase in reception time when packet loss occurs compared to the case of the carousel method.

  FIG. 29 is a diagram for explaining the load of restoration processing under each condition.

  In FIG. 29A, as described above, table 311 shows the processing time of restoration processing by the receiving device 13 and the magnitude of the restoration processing load when an effective coefficient existable region in the coefficient matrix is provided. Yes. Here, in order to indicate the load of the restoration process, the restoration process is performed by the CPU, and the CPU occupation rate by the restoration process is shown. Table 311 shows the case where the block size is 1000 packets, 2000 packets, and 4000 packets. However, the packet size is 1 KB (kilobytes), and the generation probability of the effective coefficient in the area where the effective coefficient can exist is 1/4. In addition, the bandwidth of the effective coefficient existing area under each condition is 400 columns (400 packets).

  In FIG. 29B, a table 312 is a table corresponding to the table 311 in FIG. 29A, and the processing time of the restoration processing by the receiving device 13 and the load of the restoration processing when the transmitting device 11 does not provide an effective coefficient existence area. The size of is shown. Here, a case where the block size is 500 packets, 1000 packets, and 2000 packets is shown. Other various conditions are the same as in Table 311.

  As is apparent from a comparison between Table 311 in FIG. 29A and Table 312 in FIG. 29B, when the block size is the same, the processing time is longer when the effective coefficient existence area is set than when it is not set. It is about 1/10 or shorter and the CPU occupation rate is 1/10 or less. That is, it is possible to reduce the load of the restoration processing when the effective coefficient existence region is set in the coefficient matrix than when the effective matrix is not set.

  In the above description, the determinant holding unit 154 holds the determinant generated by the determinant creating unit 153 in the packet data restoring unit 135 of the receiving device 13 as described with reference to FIG. Although the determinant aligning unit 155 rearranges the rows and arranges the equations, and the tile processing unit 156 performs the processing on the determinant in which the rows are aligned, the tile processing unit performs the determinant processing. Can be processed in a state of being aligned, and determinants that are not actually aligned may be processed.

  For example, as illustrated in FIG. 30, the tile processing unit 321 includes a determinant generation unit 322 similar to the determinant alignment unit 155 described in FIG. 13, and application data supplied by the determinant generation unit 153 is included. You may make it acquire the head information contained. At this time, for example, the tile processing unit 321 aligns the rows of the determinant based on the head information by the determinant aligning unit 322 with respect to the determinant whose rows are not aligned obtained from the determinant holding unit 154. The generated information (alignment information) may be generated and processing may be performed based on the alignment information, or the determinant may be acquired for each row from the determinant holding unit 154, and the determinant alignment unit 322 may The order of acquisition may be controlled based on the generated alignment information.

  As described above, in the communication system 10 illustrated in FIG. 2, the transmission device 11 synthesizes and converts a plurality of original data generated by dividing transmission data into blocks using an EOR operation, A plurality of converted data (more than the number of original data) is generated for each block, and these are transmitted to the receiving device 13. The receiving device 13 receives the plurality of pieces of converted data transmitted in this way for each block at least as many as the number of original data, and restores the original data using the converted data that is equal to or greater than the number of received original data. By doing in this way, the transmitter 11 and the receiver 13 can transmit and receive data more easily and more accurately.

  Note that such converted data is made redundant data of the original data (since each converted data includes the contents of the original data), the transmitting device 11 transmits it together with the redundant data original data, and the receiving device 13 receives them. Only when the original data is lost, the lost original data may be restored using the received redundant data. In the case of this method, it is only necessary that the original data lost by the receiving device 13 can be restored. Therefore, the transmitting device 11 can reduce the number of redundant data (converted data) to be transmitted, and the receiving device 13 If the original data is not lost, the restoration process of the original data can be omitted.

  However, it is rare that packet loss (original data loss) does not occur in actual communication. Further, in the case of this method, for example, the transmission device 11 can mutually communicate in the case of original data and the case of redundant data. A header including different contents must be added to generate and transmit two types of application data, and the receiving device 13 determines whether the received packet is a packet of original data or a packet of redundant data. And processing each as separate data, determining whether or not the original data is lost, or restoring the lost original data using the received redundant data only when the original data is lost For example, it is necessary to create a determinant for the calculation and perform an operation to restore the original data. 13 manufacturing cost of some concern that increased.

  In the case of the communication method in the communication system 10 shown in FIG. 2 described above, it is possible to easily transmit and receive data without requiring such complicated processing, and the original data as described above. Can be restored from any conversion data, so that data can be transmitted and received more accurately.

  Furthermore, in the case of the communication method in the communication system 10 shown in FIG. 2, the receiving device 13 can restore the original data from arbitrary converted data, so that, for example, the transmitting device 11 transmits a packet group including the converted data. Even if packet reception is started in the middle of the transmission, the original data can be restored if more converted data than the number of original data can be received for all blocks. This is more effective in one-to-many communication such as multicast or broadcast.

  In the above description, the communication system 10 has been described as configured by the transmission device 11 that transmits data and the reception device 13 that receives data. However, the configuration of the communication system is not limited to this, and the communication system is, for example, As shown in FIG. 31, the transmission device 11 and the reception device 13 of FIG. 1 are each configured as a module, which is a transmission unit and a reception unit, and is configured by two transmission / reception devices including both of them. Good.

  In FIG. 31, the communication system 401 includes a transmission / reception device 411 and a transmission / reception device 413 connected to the network 412, similar to the network 12 of FIG. 2. The transmission / reception device 411 includes a transmission unit 421 and a reception unit 422, and the transmission / reception device 413 includes a transmission unit 431 and a reception unit 432. The transmission unit 421 and the transmission unit 431 have basically the same configuration as the transmission device 11 in FIG. 2 and operate in the same manner, and thus detailed description thereof is omitted.

  In FIG. 31, the receiving unit 422 and the receiving unit 432 have basically the same configuration as the receiving device 13 in FIG. 2 and operate in the same manner, and thus detailed description thereof is omitted.

  That is, the communication system 401 in FIG. 31 includes a transmission / reception device 411 and a transmission / reception device 413 each having a built-in transmission unit and reception unit, and the converted data obtained by converting the original data using the operations described above. By using this, communication processing is performed without retransmission control.

  In this way, even in bidirectional communication, the receiving unit can realize a restoration process that is high speed and has high restoration performance without increasing the manufacturing cost. Therefore, the transmission / reception device 411 and the transmission / reception device 413 can transmit and receive data more easily and accurately.

  In FIG. 2, the communication system 10 has been described as being configured by one transmission device 11 and one reception device 13, but not limited thereto, a plurality of transmission devices or a plurality of reception devices. An apparatus may be included.

  FIG. 32 is a diagram showing a configuration example of a content distribution system to which the present invention is applied. The server 511 of the content distribution system 501 is connected to a network 512 equivalent to the network of FIG. 2 and is configured by a plurality of terminal devices 513 and 514 that are also connected to the network 512.

  In FIG. 32, the server 511 has the same function as the transmission device 11, and transmits content data such as image data and audio data to each of the terminal device 513 and the terminal device 514 via the network 512. The data is divided, the above-described calculation is performed on the original data, and the converted data obtained by converting the original data is packetized and transmitted to the multicast.

  Each of the terminal device 513 and the terminal device 514 has the same function as the receiving device 13, receives each packet supplied from the server 511, restores the original data as described above, etc. Restore and play the data.

  Note that the server 511 may download and distribute the content data, or may distribute the content data.

  As described above, when the server 511 distributes a packet to a plurality of terminal devices 513 and 514, when the terminal device 513 and the terminal device 514 lose a packet, the retransmission request for the packet is set to the server 511. As a result, reproduction requests are concentrated on the server 511, the load on the server 511 increases, and there is a possibility of going down. Therefore, as described above, the server 511 generates the conversion data and distributes it, so that the terminal device 513 and the terminal device 514 receive the content data without requesting the reproduction even if the packet is lost. Can be restored. Therefore, an increase in load on the server 511 can be suppressed, and a stable content data distribution service can be provided.

  In the above description, the data sizes (block sizes) of the blocks obtained by dividing the transmission data have been described as being common to each other. However, the present invention is not limited to this, and the blocks may of course have different block sizes. . In this case, for example, the transmission device 11 adds the block size information of the block together with the block number to the header of the application data, and the reception device 13 receives the block based on the block size information. The number of packets may be determined.

  In the above description, the transmission data, which is content data, has been described as being cut out in order from the beginning to the end of the data. However, any block cutting method may be used. However, the block generation method is determined in advance in each device of the communication system 10 or the transmission device 11 supplies the reception device 13 with information on the method, so that the transmission device 11 and the reception device 13 can generate the block. It is necessary to share. Based on the information, the reception device 13 restores the transmission data from the block by a method corresponding to the block generation method by the transmission device 11.

  Since a block is a unit of processing, it is actually preferable that the transmission apparatus 11 divides content data according to the processing flow to generate a block. Therefore, the transmission device 11 may be extracted in a discrete manner from the content data according to a certain rule according to the order of processing, for example, to form one block.

  Of course, the above-described method of transmitting converted data and the method of using a retransmission request may be used in combination. In that case, if a packet loss occurs, the receiving device first tries to restore the original data using the received converted data, and if it cannot restore, it sends a retransmission request for the lost packet to the transmission source. To do. In this way, since the number of retransmission requests can be reduced, it is possible to reduce the processing load due to the retransmission request in both the transmission side and reception side devices. That is, the load on the entire system can be reduced.

  In the above description, it has been described that the transmission device 11 performs the conversion data generation processing and the data transmission processing at the same time. However, the present invention is not limited to this. For example, the transmission device 11 acquires the transmission data in advance and uses it as described above. Of course, the converted data may be converted and stored, and the stored converted data may be supplied based on the request of the receiving device 13.

  In the communication system 10 of FIG. 2, the network 12 to which the transmission device 11 and the reception device 13 are connected may be partly or wholly configured by wire or may be configured by radio. Good. The network 12 may be configured by a plurality of networks. That is, the transmission device 11 and the reception device 13 may be connected via a plurality of routers or the like.

  Furthermore, the transmission device 11 and the reception device 13 may include configurations other than those described above. For example, a display unit such as a monitor may be provided, or an input unit such as a keyboard and buttons may be provided.

  In the above, the transmission apparatus 11 has been described as adding coefficient information, head information, a block number, and the like to EOR conversion data and transmitting it as application data. However, the present invention is not limited to this. You may make it transmit as data different from conversion data. Further, the transmission apparatus 11 may supply the information to the reception apparatus 13 using a recording medium such as a magnetic disk or a writable optical disk. That is, the transmission device 11 may supply information such as coefficient information, head information, and block number to the reception device 13 by a method different from that for EOR conversion data.

  The series of processes described above can be executed by hardware or can be executed by software. In this case, for example, the transmission device 11 and the reception device 13 in FIG. 2 include a personal computer as shown in FIG.

  In FIG. 33, the CPU 611 of the personal computer 601 executes various processes according to a program stored in a ROM (Read Only Memory) 612 or a program loaded from the storage unit 623 to the RAM 613. The RAM 613 also appropriately stores data necessary for the CPU 611 to execute various processes.

  The CPU 611, the ROM 612, and the RAM 613 are connected to each other via the bus 614. An input / output interface 620 is also connected to the bus 614.

  The input / output interface 620 includes an input unit 621 including a keyboard and a mouse, a display including a CRT and an LCD, an output unit 622 including a speaker, a storage unit 623 including a hard disk, a modem, and the like. A communication unit 624 is connected. The communication unit 624 performs communication processing via a network including the Internet.

  A drive 625 is connected to the input / output interface 620 as necessary, and a removable medium 626 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory is appropriately attached, and a computer program read from them is loaded. It is installed in the storage unit 623 as necessary.

  When the above-described series of processing is executed by software, a program constituting the software is installed from a network or a recording medium.

  For example, as shown in FIG. 33, the recording medium is distributed to distribute the program to the user separately from the apparatus main body, and includes a magnetic disk (including a flexible disk) on which the program is recorded, an optical disk ( Removable media 626 made of CD-ROM (compact disk-read only memory), DVD (including digital versatile disk), magneto-optical disk (including MD (mini-disk) (registered trademark)), or semiconductor memory In addition to being configured, it is configured by a ROM 612 in which a program is recorded and a hard disk included in the storage unit 623, which is distributed to the user in a state of being incorporated in the apparatus main body in advance.

  In the present specification, the step of describing the program recorded on the recording medium is not limited to the processing performed in chronological order according to the described order, but is not necessarily performed in chronological order. It also includes processes that are executed individually.

  Further, in this specification, the system represents the entire apparatus constituted by a plurality of apparatuses.

It is a figure explaining the data transmission method by the conventional carousel system. It is a figure which shows the structural example of the communication system to which this invention is applied. It is a block diagram which shows the structural example of the transmitter of FIG. It is a figure which shows the flow of a process of the data in the transmitter of FIG. It is a schematic diagram which shows the structural example of a UDP packet. FIG. 4 is a block diagram illustrating a detailed configuration example of an EOR conversion data generation unit in FIG. 3. It is a schematic diagram explaining the mode of generation | occurrence | production of EOR conversion data. It is a schematic diagram which shows the position of the range in which the effective coefficient can exist in the transmission apparatus of FIG. It is a block diagram which shows the detailed structural example of the application data generation part of FIG. It is a schematic diagram which shows the structural example of application data. FIG. 3 is a block diagram illustrating a configuration example of a receiving device in FIG. 2. It is a figure which shows the example of a mode that a packet is lost. It is a block diagram which shows the detailed structural example of the packet data decompression | restoration part of FIG. FIG. 3 is a schematic diagram illustrating positions of valid ranges of effective coefficients in the receiving apparatus of FIG. 2. It is a schematic diagram which shows the example of distribution of the effective coefficient in a coefficient matrix. It is a schematic diagram explaining the structural example of a tile. It is a flowchart explaining the example of a data transmission process. It is a flowchart explaining the example of an EOR conversion data generation process. It is a flowchart explaining the example of an application data generation process. It is a flowchart explaining the example of a data reception process. It is a flowchart explaining the example of an original data restoration process. It is a flowchart explaining the example of a solution process. It is a flowchart following FIG. 21 explaining the example of a solution process. It is a figure explaining the example of the procedure of the Gaussian elimination method. It is a figure explaining the example of the procedure of the Gaussian elimination method. It is a figure explaining the example of the procedure of the Gaussian elimination method. It is a figure explaining the example of the procedure of the Gaussian elimination method. It is a figure explaining the example of the relationship between a packet loss and reception processing time. It is a figure which shows the example of the processing time of a restoration process, and the magnitude | size of load. It is a block diagram which shows the other detailed structural example of the packet data decompression | restoration part of FIG. It is a figure which shows the other structural example of the communication system to which this invention is applied. It is a figure which shows the structural example of the content delivery system to which this invention is applied. And FIG. 16 is a block diagram illustrating a configuration example of a personal computer.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Communication system 11 Transmission apparatus 12 Network 13 Reception apparatus 21 Data division part 22 Block division part 23 Packet data holding part 24 EOR conversion data generation part 25 Application data generation part 26 Data conversion control part 27 UDP packetization unit, 28 buffer, 29 output control unit, 30 IP processing unit, 31 Ethernet (R) processing unit, 32 cable, 41 transmission data, 42 blocks, 43 original data, 44 EOR conversion data, 45 transmission packet, 61 Determinant generating unit, 62 Determinant calculating unit, 63 Solution output unit, 71 Coefficient matrix holding unit, 81 Pseudo random number generating unit, 82 Start position determining unit, 83 Width setting unit, 84 Range control unit, 85 Coefficient determining unit, 86 Coefficient generation probability holding unit, 101 Coefficient information Append section 101, 102 head information append section, 103 block number append section, 104 block counter, 105 application data output section, 110 application data, 111 header, 112 EOR conversion data, 121 block number, 122 head information, 123 coefficient information, 130 cable, 131 Ethernet processing unit, 132 IP processing unit, 133 application data extraction unit, 134 buffer unit, 135 packet data restoration unit, 136 block data restoration unit, 137 received data generation unit, 151 restoration control unit, 152 Coefficient matrix creation unit, 153 determinant creation unit, 154 determinant holding unit, 155 determinant alignment unit, 156 tile processing unit, 157 determinant calculation unit, 158 solution output unit, 321 File processing unit, 322 determinant sorting unit, 401 communication system, 411 transmission / reception device, 412 network, 413 transmission / reception device, 421 transmission unit, 422 reception unit, 431 transmission unit, 432 reception unit, 501 content distribution system, 511 server, 512 Network, 513 terminal device, 514 terminal device, 601 personal computer, 611 CPU, 612 ROM, 613 RAM, 614 bus, 620 I / O interface, 621 input unit, 622 output unit, 623 storage unit, 624 communication unit, 625 drive, 626 removable media

Claims (20)

A transmission / reception system comprising: a transmission device that transmits data for transmission; and a reception device that receives the transmission data transmitted by the transmission device,
The transmitting device generates a plurality of first data by dividing the transmission data, and sets a set including all the generated first data as an entire set, and the entire set from the entire set. A subset consisting of any continuous element group that is a part of the set is set, and a plurality of the first data are arbitrarily selected from the set subset, and the selected plurality of the first data are bit by bit Obtained by converting each of the plurality of first data into a plurality of the second data by performing an exclusive OR operation every time and obtaining the second data as a result of the operation for the plurality of the subsets. Packetizing each of the plurality of second data and transmitting to the receiving device,
The receiving device receives a plurality of the second data transmitted by the transmitting device by the number of the first data corresponding to the second data, and the received second data corresponds to the second data. The second data that is equal to or more than the number of first data is held, a determinant corresponding to the second data is generated using the plurality of held second data, and the determinant is solved The transmission / reception system characterized in that the first data is restored. A transmission device that transmits data for transmission to a reception device,
Data generating means for generating a plurality of first data by dividing the transmission data;
A set including all the first data generated by the data generation means as a whole set is set as a whole set, and a subset including any continuous element group that is a part of the whole set is set from the whole set. , Arbitrarily selecting a plurality of the first data from the set subset, performing an exclusive OR operation on the selected plurality of the first data for each bit, and a plurality of second data as the operation result Conversion means for converting a plurality of the first data into a plurality of the second data by obtaining the subset of
A transmission device comprising: a transmission unit configured to packetize each of the plurality of second data obtained by converting the plurality of first data by the conversion unit, and transmit the packetized data to the reception device. The converting means includes
Determinant generating means for generating a determinant for converting the first data into the second data using a coefficient matrix for selecting the first data to be subjected to the exclusive OR operation;
3. The computing device according to claim 2, further comprising: a computing unit that obtains the second data by computing the determinant generated by the determinant generating unit using the exclusive OR operation. Transmitter device. The coefficient matrix includes an effective coefficient having a value of “1” and an invalid coefficient having a value of “0”.
The effective coefficient is a component arbitrarily selected within a range of columns in which the effective coefficient can be set as the subset for each row of the coefficient matrix,
The transmission device according to claim 3, wherein the range of each row is set such that the number of columns in the range is the same, and the first column of the range is different for each row. Range setting means for setting the range for each row;
Determination means for determining the value of each component of the coefficient matrix based on the range set by the range setting means;
Coefficient matrix holding means for holding the coefficient matrix by holding values of all components of the coefficient matrix based on the determination result by the determination means, and
The transmission apparatus according to claim 4, wherein the determinant generating unit generates the determinant using the coefficient matrix held by the coefficient matrix holding unit. The transmission device according to claim 2, wherein the conversion unit converts a plurality of the first data into the second data larger than the number of the first data. Information adding means for adding information about the range to the second data obtained by converting the first data by the converting means;
The transmission device according to claim 2, wherein the transmission unit packetizes each of the plurality of second data to which the information is added by the information addition unit, and transmits the packetized data to the reception device. The transmission apparatus according to claim 2, wherein the data generation unit divides the transmission data into blocks, and further generates the plurality of first data by dividing the block. A transmission method of a transmission device for transmitting data for transmission to a reception device,
A data generation step of generating a plurality of first data by dividing the transmission data;
A set including all the first data generated by the processing of the data generation step as an entire set is defined as an entire set, and a subset including any continuous element group that is a part of the entire set from the entire set. Set, arbitrarily select a plurality of the first data from the set subset, perform an exclusive OR operation on the selected plurality of the first data for each bit, and obtain the second data as the operation result Converting a plurality of the first data into a plurality of the second data by obtaining a plurality of the subsets;
A transmission control step of controlling the plurality of second data obtained by converting the plurality of first data by the conversion step to be packetized and transmitted to the receiving device, respectively. A transmission method characterized by the above. In a program for causing a computer to perform transmission processing for transmitting data for transmission to a receiving device,
A data generation step of generating a plurality of first data by dividing the transmission data;
A set including all the first data generated by the processing of the data generation step as an entire set is defined as an entire set, and a subset including any continuous element group that is a part of the entire set from the entire set. Set, arbitrarily select a plurality of the first data from the set subset, perform an exclusive OR operation on the selected plurality of the first data for each bit, and obtain the second data as the operation result Converting a plurality of the first data into a plurality of the second data by obtaining a plurality of the subsets;
A transmission control step of controlling the plurality of second data obtained by converting the plurality of first data by the conversion step to be packetized and transmitted to the receiving device, respectively. A program characterized by A receiving device for receiving data transmitted by a transmitting device,
The data transmitted by the transmitting device, each of which is a set including all the first data generated by dividing transmission data as a whole set, and a part of the whole set from the whole set A subset consisting of any continuous element group is set, and a plurality of the first data are arbitrarily selected from the set subset, and the plurality of selected first data are mutually exclusive for each bit. Perform a logical OR operation, and receive a plurality of second data obtained for a plurality of the subsets as a result of the exclusive OR operation, by the number of the first data corresponding to the second data. Receiving means;
Holding means for holding the second data equal to or more than the number of the first data corresponding to the second data received by the receiving means;
Restoring means for generating a determinant corresponding to the second data using the plurality of second data held by the holding means, and for restoring the first data by solving the determinant And a receiving device. The restoration means includes
Determinant generating means for generating a determinant for converting the first data into the second data using a coefficient matrix for selecting the first data to be subjected to the exclusive OR operation;
The receiving apparatus according to claim 11, further comprising: an arithmetic unit that solves the determinant generated by the determinant generating unit using the first data as a variable and obtains the first data. The coefficient matrix includes an effective coefficient having a value of “1” and an invalid coefficient having a value of “0”.
The effective coefficient is a component arbitrarily selected within a range of columns in which the effective coefficient can be set as the subset for each row of the coefficient matrix,
The receiving apparatus according to claim 12, wherein the range of each row is set such that the number of columns in the range is the same, and the first column of the range is different for each row. The coefficient matrix is generated based on information about the range that is arbitrarily selected from the plurality of first data to be subjected to the exclusive OR operation, which is added to the second data received by the receiving unit. A coefficient matrix generating means for performing
The receiving apparatus according to claim 12, wherein the determinant generating unit generates the determinant using the coefficient matrix generated by the coefficient matrix generating unit. Each component of the coefficient matrix is divided into a plurality of small regions for each predetermined number of rows and columns, and the determinant generated by the determinant generating unit based on the value of the component of each small region Computation determining means for determining whether or not to perform the calculation of the portion corresponding to the small region,
The calculation means calculates only the portion of the determinant generated by the determinant generation means that is determined to be calculated by the calculation determination means, and solves the first data as a variable. The receiving apparatus according to claim 12, wherein the first data is obtained. A receiving method of a receiving device for receiving data transmitted by a transmitting device,
The data transmitted by the transmitting device, each of which is a set including all the first data generated by dividing transmission data as a whole set, and a part of the whole set from the whole set A subset consisting of any continuous element group is set, and a plurality of the first data are arbitrarily selected from the set subset, and the plurality of selected first data are mutually exclusive for each bit. Perform a logical OR operation, and receive a plurality of second data obtained for a plurality of the subsets as a result of the exclusive OR operation, by the number of the first data corresponding to the second data. A reception control step for controlling
A holding step of holding the second data equal to or more than the number of the first data corresponding to the second data, which is controlled and received by the processing of the reception control step;
The determinant corresponding to the second data is generated using a plurality of the second data held by the process of the holding step, and the first data is restored by solving the determinant A receiving method comprising: a restoring step. In a program for causing a computer to perform reception processing for receiving data transmitted by a transmission device,
The data transmitted by the transmitting device, each of which is a set including all the first data generated by dividing transmission data as a whole set, and a part of the whole set from the whole set A subset consisting of any continuous element group is set, and a plurality of the first data are arbitrarily selected from the set subset, and the plurality of selected first data are mutually exclusive for each bit. Perform a logical OR operation, and receive a plurality of second data obtained for a plurality of the subsets as a result of the exclusive OR operation, by the number of the first data corresponding to the second data. A reception control step for controlling
A holding step of holding the second data equal to or more than the number of the first data corresponding to the second data, which is controlled and received by the processing of the reception control step;
The determinant corresponding to the second data is generated using the plurality of second data held by the holding step, and the first data is restored by solving the determinant A program comprising: a restoration step. A transmission / reception device for transmitting / receiving data for transmission,
Data generating means for generating a plurality of first data by dividing the transmission data;
A set including all the first data generated by the data generation means as a whole set is set as a whole set, and a subset including any continuous element group that is a part of the whole set is set from the whole set. , Arbitrarily selecting a plurality of the first data from the set subset, performing an exclusive OR operation on the selected plurality of the first data for each bit, and a plurality of second data as the operation result Conversion means for converting a plurality of the first data into a plurality of the second data by obtaining the subset of
Transmitting means for packetizing a plurality of the second data obtained by converting the plurality of first data by the converting means, respectively, and transmitting the packet to other transmitting / receiving devices;
Receiving means for receiving a plurality of the second data transmitted by the other transmission / reception device by the number of the first data corresponding to the second data;
Holding means for holding the second data equal to or more than the number of the first data corresponding to the second data received by the receiving means;
Restoring means for generating a determinant corresponding to the second data using the plurality of second data held by the holding means, and for restoring the first data by solving the determinant A transmission / reception device comprising: A transmission / reception method of a transmission / reception apparatus for transmitting / receiving data for transmission,
A data generation step of generating a plurality of first data by dividing the transmission data;
A set including all the first data generated by the processing of the data generation step as an entire set is defined as an entire set, and a subset including any continuous element group that is a part of the entire set from the entire set. Set, arbitrarily select a plurality of the first data from the set subset, perform an exclusive OR operation on the selected plurality of the first data for each bit, and obtain the second data as the operation result Converting a plurality of the first data into a plurality of the second data by obtaining a plurality of the subsets, and
A transmission control step for controlling each of the plurality of second data obtained by converting a plurality of the first data by the processing of the conversion step into packets and transmitting them to other transmission / reception devices;
A reception control step of controlling a plurality of the second data transmitted by the other transmission / reception apparatus so as to receive more than the number of the first data corresponding to the second data;
A holding step for holding the second data equal to or more than the number of the first data corresponding to the second data, which is controlled and received by the processing of the reception control step;
The determinant corresponding to the second data is generated using the plurality of second data held by the holding step, and the first data is restored by solving the determinant A transmission / reception method comprising: a restoration step. In a program for causing a computer to perform transmission / reception processing for transmitting / receiving data for transmission,
A data generation step of generating a plurality of first data by dividing the transmission data;
A set including all the first data generated by the processing of the data generation step as an entire set is defined as an entire set, and a subset including any continuous element group that is a part of the entire set from the entire set. Set, arbitrarily select a plurality of the first data from the set subset, perform exclusive OR operation on the selected plurality of the first data for each bit, and obtain the second data as the operation result Converting a plurality of the first data into a plurality of the second data by obtaining a plurality of the subsets;
A transmission control step for controlling each of the plurality of second data obtained by converting a plurality of the first data by the processing of the conversion step into packets and transmitting them to other transmission / reception devices;
A reception control step of controlling a plurality of the second data transmitted by the other transmission / reception apparatus so as to receive more than the number of the first data corresponding to the second data;
A holding step of holding the second data equal to or more than the number of the first data corresponding to the second data, which is controlled and received by the processing of the reception control step;
The determinant corresponding to the second data is generated using the plurality of second data held by the holding step, and the first data is restored by solving the determinant A program comprising: a restoration step.

Images