JP2007074152A - Quality measuring method and instrument, and code error correcting method, system, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To eliminate unnecessary delay and congestion during data transfer. <P>SOLUTION: A data receiving device side performs quality measurement processing by sectioning data to be received into quality measurement blocks by the number of packets, representing a quality measurement block generation frequency for a packet loss number in a quality measurement block as a distribution, finding a deviation of the distribution from a reference distribution as a value of burst characteristics of a packet loss, and the ratio of a packet loss number in a time interval to the number of all packets to be received as a packet loss rate as quality information. Consequently, an encoding parameter during error correction encoding on a data transmitting device side is optimized, based upon the quality information measured by the data receiving device side. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a quality measuring method and apparatus, a code error correcting method, a system and a program, and relates to a quality measuring method and apparatus and a code error correcting method for improving data transfer quality in a network of best effort service including the Internet. And systems and programs.

  In recent years, the use of real-time video communication such as two-way video communication has increased with the broadbandization of communication networks. In real-time video communication, the quality of the communication network directly affects the quality of experience, so a small amount of packet loss leads to degradation of viewing quality.

  As a function to prevent quality degradation due to packet loss, there are a method of retransmitting a lost packet and a code error correction method of transmitting a redundant packet in advance and correcting the error on the receiving side. A code error correction method that does not involve retransmission of a packet is expected to be applied to a real-time application because it only generates a delay about the reception time of a coded block for decoding.

  In general, code error correction uses s bits as 1 symbol, performs error correction coding on k symbol information to be transmitted, and configures a code with a code length of n symbols (hereinafter referred to as a coding block) for transmission. However, even if an error occurs in the transmission process, the original k symbol information can be restored by performing error correction decoding.

  There are various coding methods for code error correction, and they are properly used depending on the field to be applied and the target error. The present invention targets burst packet loss in the field of wired communication best effort networks, and mainly uses a Reed-Solomon code or LDPC (Low Density Parity Check) code for symbol error correction. Therefore, hereinafter, error correction in units of packets using Reed-Solomon codes will be described.

  FIG. 12 shows an implementation example of error correction in units of packets using Reed-Solomon codes.

In the error correction coding capabilities of the transmitting side, the transmitting side performs a syndrome calculation in the Galois field GF (2 ^s) for the data sequence of k symbols (1 symbol = s bits), 2 to the original data sequence ^s -Send a redundant data sequence of k symbols. In the error correction decoding function on the receiving side, the same syndrome calculation is performed, the error position and amount in symbol units are specified from the result, and error correction is performed (for example, see Non-Patent Document 1).

  When performing error correction in packet units, correction is generally performed by executing p / s processes in units of s bits in parallel on p-bit packets of length. If a sequence number is assigned to each packet, it can be transferred arbitrarily by adding (nk) redundant packets to k information packets (hereinafter referred to as RS (n, k) encoding). (Nk) packet losses can be completely corrected.

  FIG. 13 shows an implementation example of error correction encoding processing combining Reed-Solomon code and packet level crossing method.

  The transmission side performs error correction in the column direction, arranges N blocks of encoded blocks composed of k information packets and (n−k) redundant packets, and transmits them in the row direction as shown in FIG. This method is called a crossing method, and the number N of encoded blocks is called the number of crossings. Thereby, the continuous packet loss which generate | occur | produced in the network can be disperse | distributed to each encoding block, and can be corrected.

In addition, for the problem that adding excessive redundant data causes further congestion in the network, the encoding parameters are adapted according to the error frequency and delay of the communication path, the buffer link time of the received data, etc. A method of changing the frequency is proposed (for example, see Patent Document 1).
Toshinobu Kaneko, Yoshizumi Eto, et al., “Error correction codes and their applications”, Ohm Publishing Bureau, published December 25, 1996, pp.107-115 JP 2004-215224 A

  However, as described above, the conventional method increases the tolerance against packet loss due to the application of excessive redundant data, but may cause further congestion in the network. Therefore, it is necessary to adaptively select a coding parameter that can minimize redundancy and completely repair loss according to the quality of the network. As a method of selecting an optimal encoding parameter, there is a method of acquiring network quality information in real time and adaptively changing the encoding parameter in accordance with the quality information. One effective quality information for determining the encoding parameter is a packet loss rate. By measuring the packet loss rate and adopting the appropriate redundancy, it is possible to suppress congestion because unnecessary redundant packets need not be sent.

  On the other hand, in an actual Internet network, random loss and burst loss are mixed, so if the encoding parameter is determined only by the packet loss rate, the loss may not be completely repaired, or unnecessary delay may occur. is there. For example, if there is no change in the packet loss rate but a lot of burst loss occurs, the redundancy need not be changed, but the loss may not be repaired unless the code length and the number of crossings are increased. high.

  On the other hand, if the loss is random, it is highly possible that the loss can be repaired even if the code length and the number of crossings are reduced to some extent, but if they remain large, the encoding that occurs on the receiving side There is a problem that time for receiving the packet of the block (hereinafter referred to as error correction delay) is wasted.

  The present invention has been made in view of the above points, and measures the burstiness and packet loss rate of loss that occurs in the network. From these two pieces of quality information, the repair capability is high, and unnecessary delay and congestion are not generated. It is an object of the present invention to provide a quality measurement method and apparatus, a code error correction method, a system, and a program for determining an optimal encoding parameter and performing code error correction while adaptively changing the encoding parameter.

The present invention (Claim 1) is a quality measuring method in a data receiving apparatus,
In the data receiving device,
Store the number of packets, time interval, reference distribution in memory beforehand,
Data to be received is divided into quality measurement blocks for each number of packets acquired from the memory, and the number of occurrences of quality measurement blocks with respect to the number of packet losses in the quality measurement block is expressed as a distribution at the time interval acquired from the memory. Determining the degree of divergence between the distribution and the reference distribution obtained from the memory as a burstiness value of packet loss;
Obtaining a ratio of the number of packet losses in the time interval to the total number of packets to be received as a packet loss rate as quality information.

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

According to the present invention (Claim 2), one or more pieces of data are input as a continuous data sequence, error correction coding is performed, and a data transmission apparatus that transmits the data as an encoded block via a communication path, and the code received from the communication path. A code error correction method in a system including a data receiving apparatus that receives data of a generalized block, decodes and outputs the data,
The data receiving device
A quality measurement step (step 1) for measuring a burstiness value of packet loss and a packet loss rate at a predetermined time interval according to the method of claim 1;
A quality information notifying step (step 2) for periodically notifying the data transmitting apparatus of the burstiness value and the packet loss rate measured in the quality measuring step as quality information;
The data transmission device
When the burstiness value of quality information received from the data receiving apparatus is larger than a predetermined value, the code length as an encoding parameter is increased, and when it is smaller, the code length is decreased, and the packet loss rate is a predetermined value. A parameter determining step (step 3) for increasing the redundancy as an encoding parameter when it is larger than the value, and decreasing the redundancy when it is smaller;
An error correction coding step (step 4) for performing coding based on the code length and redundancy determined in the parameter determination step;
An information providing step of adding information for encoding and decoding including the parameter determined in the parameter determining step to the data of the encoded block encoded in the error correction encoding step and transmitting it to the data receiving device ( Step 5) and
The data receiving device
A decoding step (step 6) for decoding the data of the encoded block transmitted from the data transmitting apparatus using the information for encoding and decoding added to the data of the encoded block; Do.

The present invention (Claim 3) inputs one or more data as a continuous data sequence, performs error correction coding, and transmits the encoded block as a coded block via a communication path, and the code received from the communication path. A code error correction method in a system including a data receiving apparatus that receives data of a generalized block, decodes and outputs the data,
When using both error correction and crossing method,
The data receiving device
A quality measurement step of measuring a packet loss burstiness value and a packet loss rate at a predetermined time interval by the quality measurement method according to claim 1;
A quality information notifying step for periodically notifying the data transmitting apparatus of the burstiness value and the packet loss rate measured in the quality measuring step as quality information;
The data transmission device
When the burstiness value of the quality information received from the data receiving device is larger than a predetermined value, either the code length selected in advance or the number of crossings, or both are increased and decreased as the encoding parameter Reduces either the preselected code length or the number of crossings, or both, and increases the redundancy as an encoding parameter when the packet loss rate is larger than a predetermined value, and redundancy when the packet loss rate is smaller A parameter determination step to reduce the degree,
An error correction encoding step for performing encoding based on the code length and redundancy determined in the parameter determination step;
An information adding step of adding information for encoding and decoding including the parameter determined in the parameter determining step to the data of the encoded block encoded in the error correction encoding step and transmitting the data to the data receiving device; ,
The data receiving device
A decoding step of decoding the data of the encoded block transmitted from the data transmitting apparatus using information for encoding and decoding given to the data of the encoded block;
I do.

The present invention (Claim 4) is a quality information measuring device included in a data receiving device,
Memory that stores the number of input packets, time interval, and reference distribution in advance,
Data to be received is divided into quality measurement blocks for each number of packets acquired from the memory, and the number of occurrences of quality measurement blocks with respect to the number of packet losses in the quality measurement block is expressed as a distribution at the time interval acquired from the memory. Means for determining the degree of divergence between the distribution and the reference distribution acquired from the memory as a burstiness value of packet loss;
Means for obtaining a ratio of the number of packet losses in the time interval to the total number of packets to be received as a packet loss rate as quality information.

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

The present invention (Claim 5) inputs one or more pieces of data as a continuous data sequence, performs error correction coding, and transmits the data as an encoded block via a communication path, and the data transmission apparatus 100 received from the communication path. A code error correction system comprising a data receiving apparatus 200 that receives data of a coded block, decodes and outputs the data,
The data receiving device 200
5. A data transmission apparatus comprising the quality measuring apparatus according to claim 4, wherein the burst loss value of packet loss and a packet loss rate at a predetermined time interval are measured, and the burstiness value and the packet loss rate are used as quality information. Quality measurement means 210 for periodically informing
Decoding means 230 for decoding the data of the encoded block transmitted from the data transmitting apparatus using information for encoding and decoding attached to the data of the encoded block;
The data transmission device 100
When the burstiness value of the quality information received from the data receiving apparatus 200 is larger than a predetermined value, the code length as the encoding parameter is increased, and when it is smaller, the code length is decreased, and the packet loss rate is predetermined. Parameter determination means 110 for increasing the redundancy as an encoding parameter when the value is larger than the value, and decreasing the redundancy when the value is smaller;
Error correction coding means 120 for performing coding based on the code length and redundancy determined by the parameter determination means 110;
Information for encoding and decoding including the parameter determined by the parameter determination unit 110 is added to the data of the encoded block encoded by the error correction encoding unit 120 and transmitted to the data receiving apparatus Means 130.

According to the present invention (Claim 6), a data transmission apparatus that inputs one or more data as a continuous data sequence, performs error correction coding, and transmits the data as an encoded block via a communication path, and the code received from the communication path. A code error correction system comprising a data receiving device that receives, decodes, and outputs data of an encoded block,
When using both error correction and crossing method,
The data receiving device
5. A data transmission apparatus comprising the quality measuring apparatus according to claim 4, wherein the burst loss value of packet loss and a packet loss rate at a predetermined time interval are measured, and the burstiness value and the packet loss rate are used as quality information. Quality information measuring means for periodically informing,
Decoding means for decoding the data of the encoded block transmitted from the data transmitting apparatus using information for encoding and decoding attached to the data of the encoded block;
The data transmission device
When the burstiness value of the quality information received from the data receiving device is larger than a predetermined value, either the code length selected in advance or the number of crossings, or both are increased and decreased as the encoding parameter Reduces either the preselected code length or the number of crossings, or both, and increases the redundancy as an encoding parameter when the packet loss rate is larger than a predetermined value, and redundancy when the packet loss rate is smaller Parameter determining means for reducing the degree,
Error correction encoding means for performing encoding based on the code length and redundancy determined by the parameter determination means;
Information adding means for adding information for encoding and decoding including the parameter determined by the parameter determining means to the data of the encoded block encoded in the error correction encoding step and transmitting the data to the data receiving device; Have.

The present invention (Claim 7) is a quality measurement program,
Computer
Data to be received is divided into quality measurement blocks for each number of packets acquired from the memory, and the number of occurrences of quality measurement blocks with respect to the number of packet losses in the quality measurement block is expressed as a distribution at the time interval acquired from the memory. Means for determining the degree of divergence between the distribution and the reference distribution acquired from the memory as a burstiness value of packet loss;
A program for functioning as a means for obtaining a ratio of the number of packet losses in the time interval to the total number of packets to be received as a packet loss rate as quality information.

The present invention (Claim 8) is a code error correction program,
Computer
Data to be received is divided into quality measurement blocks for each number of packets stored in the memory in advance, and the number of times the quality measurement block is generated with respect to the number of packet losses in the quality measurement block in the time interval stored in advance in the memory As a distribution, and the degree of divergence between the distribution and a reference distribution stored in advance in the memory as a packet loss burstiness value, and the ratio of the number of packet losses in the time interval to the total number of packets to be received. Quality measurement means for notifying a data transmission apparatus connected via a communication path as a packet loss rate as quality information,
Program for functioning as decoding means for decoding encoded block data transmitted from a data transmitting apparatus using information for encoding and decoding added to the encoded block data It is.

The present invention (Claim 9) is a code error correction program,
Computer
When the burstiness value of quality information received from the data receiving apparatus is larger than a predetermined value, the code length as an encoding parameter is increased, and when it is smaller, the code length is decreased, and the packet loss rate is a predetermined value. Parameter determination means for increasing the redundancy as an encoding parameter if it is larger than, and decreasing the redundancy if it is smaller;
Error correction encoding means for performing encoding based on the code length and redundancy determined by the parameter determination means;
Information for encoding and decoding including parameters determined by the parameter determination unit is added to the data of the encoded block encoded by the error correction encoding unit, and is transmitted as an information adding unit to the data receiving device It is a program that makes it work.

The present invention (Claim 10) is a code error correction program,
Computer
When the burstiness value of the quality information received from the data receiving device is larger than a predetermined value, either the code length selected in advance or the number of crossings, or both are increased and decreased as the encoding parameter Reduces either the preselected code length or the number of crossings, or both, and increases the redundancy as an encoding parameter when the packet loss rate is larger than a predetermined value, and redundancy when the packet loss rate is smaller Parameter determining means for reducing the degree,
Error correction encoding means for performing encoding based on the code length and redundancy determined by the parameter determination means;
Information for encoding and decoding including parameters determined by the parameter determination unit is added to the data of the encoded block encoded in the error correction encoding step and transmitted to the data receiving device. It is a program that makes it work.

  As described above, according to the present invention, by determining the optimum parameters according to the burstiness and packet loss rate of packet loss occurring in the communication path, the repair capability is high, and unnecessary delay and congestion occur. Code error correction can be performed.

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

  FIG. 3 shows a basic configuration of a code error correction system according to an embodiment of the present invention.

  The system shown in FIG. 1 performs error correction coding on input data and transmits the data as a coding block to the communication path 300, and receives the coding block from the communication path 300, decodes the code, and outputs the output data. Is provided.

  The data transmission device 100 and the data reception device 200 are connected via a communication path 300. The data transmission device 100 and the data reception device 200 are arranged at a terminal, a communication network edge, or the like.

  The data transmission apparatus 100 includes a parameter determination unit 110, an error correction coding unit 120, and an information addition unit 130. Further, the data receiving apparatus 200 includes a quality measuring unit 210, an information monitoring unit 220, and an error correction decoding unit 230.

  The parameter determination unit 110 determines a parameter for encoding suitable for the quality of the communication network based on the quality information notified from the quality measurement unit 210 of the data receiving apparatus 200. Here, the parameters for encoding include at least one of code length, information data amount, redundancy, and number of crossings.

  The error correction encoding unit 120 performs encoding using an error correction code such as a Reed-Solomon code or an LDPC code, based on the parameter information for encoding determined by the parameter determination unit 110.

  The information assigning unit 130 assigns information for encoding and decoding including the parameters for encoding determined by the parameter determining unit 110 to, for example, a header of a packet constituting the encoded block. Here, the information for encoding and decoding refers to, for example, the code length, the amount of information data, the number of crossings, the encoding block number, and the sequence number.

  The quality measuring unit 210 of the data receiving apparatus 200 measures the quality information of the communication path 300 and periodically notifies the data transmitting apparatus 100 of the quality information. Here, the quality information refers to the burstiness value of packet loss occurring in the communication path 300 and the packet loss rate, and the data to be received is defined as a quality measurement block for each predetermined number of packets. The number of occurrences of quality measurement blocks with respect to the number of packet losses in the quality measurement block is expressed as a distribution at a predetermined time interval. The value of burstiness is assumed, and the ratio of the number of packet loss at a predetermined time interval to the total number of packets to be received is defined as a packet loss rate. It is assumed that the predetermined number of packets, time interval, and reference distribution are stored in a memory (not shown) of the data receiving apparatus 200.

  The information monitoring unit 220 acquires information for encoding and decoding for each encoding block from the header of the received encoding block packet.

  The error correction decoding unit 230 performs decoding by error correction calculation based on the information acquired by the information monitoring unit 220, and repairs packet loss that has occurred in the communication path.

  An outline of the operation of the parameter determination unit of the data transmission apparatus having the above configuration will be described.

  FIG. 4 is an operation flowchart (part 1) of the parameter determination unit of the data transmission apparatus according to the embodiment of the present invention.

  The parameter determination unit 110 in the data transmission device 100 periodically receives the packet loss burstiness value as quality information from the quality measurement unit 210 of the data reception device 200 (step 101), and the packet loss burstiness value is determined. When the value is larger than a predetermined value (step 102), the code length as an encoding parameter is increased (step 103), and the burstiness value of packet loss is smaller than a preset value (Step 102), the code length is reduced (step 104).

  The parameter determination unit 110 of the data receiving apparatus 100 periodically receives a packet loss rate as quality information from the quality measuring unit 210 of the data receiving apparatus 200 (step 105), and the packet loss rate is a predetermined value. Is larger (step 106), the redundancy as an encoding parameter is increased (step 107), and when the packet loss rate is smaller than a preset value, the redundancy is decreased (step 108). .

  In step 101 and step 105, it is described that the packet loss burstiness value and the packet loss rate are received. However, the packet loss rate may be periodically received simultaneously.

  As described above, when both the burstiness value and the packet loss rate as the quality information are different from the preset values, both the code length and the redundancy are changed. As a result, for example, when only the burstiness value of packet loss increases, the repair capability can be improved without adding unnecessary redundant packets by increasing the code length without changing the redundancy. When only the packet loss rate increases, the redundancy can be increased without changing the code length, so that the repair capability can be enhanced without causing unnecessary delay.

  Next, the case where the code error correction and the crossing method are used together will be described.

  FIG. 5 is a flowchart (part 2) of the operation of the parameter determination unit of the data transmission device according to the embodiment of the present invention.

  When the code error correction and the crossing method are used together, the parameter determination unit 110 of the data transmission device 100 periodically sends the packet loss burstiness value as quality information from the quality measurement unit 210 of the data reception device 200. (Step 201), and if the burst loss value of the packet loss is larger than a predetermined value (step 202), increase either or both of the code length and the number of crossings as an encoding parameter. However, if the burstiness value of packet loss is smaller than a predetermined value, either the code length or the number of crossings or both are reduced (step 204). In step 203, it is assumed that either or both of the code length and the number of crossings are changed in advance.

  The parameter determination unit 110 of the data transmission device 100 periodically receives the packet loss rate as quality information from the quality measurement unit 210 of the data reception device (step 205), and the packet loss rate is determined based on a predetermined value. Is larger (step 206), the redundancy as an encoding parameter is increased (step 207), and when the packet loss rate is smaller than a preset value (step 206), the redundancy is decreased (step 206). Step 208).

  In step 201 and step 205, it has been described that the packet loss burstiness value and the packet loss rate are received, respectively, but they may be received simultaneously periodically.

  As described above, when both the burstiness value and the packet loss rate as the quality information are different from the preset values, both the code length, the number of crossings, and the redundancy are changed. As a result, for example, when only the burstiness value of packet loss increases, by increasing either the code length and the number of crossings, or both, without changing the redundancy, useless redundancy Without adding packets, the repair capability can be increased, and when only the packet loss rate increases, the redundancy is increased without changing the code length and the number of crossings, so that unnecessary delay is not generated. In addition, the repair ability can be enhanced.

  The quality measurement function measures the burstiness value and packet loss rate of packet loss occurring on the communication path, and the parameter determination function can adaptively change the encoding parameters according to the measurement results. Therefore, it is possible to perform code error correction that is always suitable for the state of the communication path and has a high repair capability and does not cause unnecessary delay or congestion.

  Below, it demonstrates concretely with reference to drawings.

  The unit of the data amount is the number of packets, and when the packet size of the input data varies depending on the packet, padding was performed so that the number of packets when the data was divided into a specific packet size or a specific packet size Is the number of packets at the time.

[First Embodiment]
FIG. 6 shows the configuration and initial setting values of the data receiving apparatus according to the first embodiment of the present invention.

  The data receiving apparatus 200 includes a quality measuring unit 210, an information monitoring unit 220, and an error correction decoding unit 230. The data receiving apparatus 200 has a unit time T, the number of packets M of a quality measurement block for measuring the burstiness of packet loss as an initial setting value in a memory (not shown), and a Poisson distribution with respect to the packet loss rate as a reference distribution. In addition, information on the occurrence rate per unit time of the quality measurement block in which the packet loss exceeding the 99.99% variance value of the Poisson distribution as the distribution divergence degree is stored.

  FIG. 7 shows an example of burstiness measurement in the first embodiment of the present invention.

  In the figure, an example of measurement of burstiness for one minute when the unit time is 1 minute and the number of packets of the quality measurement block is 100 packets is shown. The measurement result of the packet loss rate for 1 minute is 5%. The horizontal axis is the number of packet losses in the quality measurement block, and the vertical axis is the number of quality measurement block occurrences. The solid line in the graph in the figure is a Poisson distribution with respect to a packet loss rate of 5%, and the bar graph is a measured distribution. Since the packet loss rate in the quality measurement block exceeding the 99.99% variance value of the Poisson distribution with respect to the packet loss rate of 5% is 17 or more, the number of quality block occurrences having the number of packet losses in the quality measurement block of 17 or more is 1 The ratio of the total quality measurement block per minute is a burstiness value.

[Second Embodiment]
In the present embodiment, description will be made assuming that Reed-Solomon code is used as the encoding method.

  FIG. 8 shows the configuration and initial setting values of the data transmission apparatus according to the second embodiment of the present invention.

  The data transmission apparatus 100 includes a parameter determination unit 110, an error correction coding unit 120, and an information addition unit 130. The data transmitting apparatus 100 has parameters n = n0, k = k0, R = R0, burstiness tolerance B = B0, and packet loss rate tolerance as initial setting values in a memory (not shown). The value L = L0 is stored in advance. Note that n is the total number of packets (code length) in the coding block, k is the number of information packets in the coding block, and R is a redundancy indicating the ratio of the number of redundant packets to the total number of packets in the coding block.

  The operation will be described below. Note that the configuration and initial setting value of the data receiving apparatus 200 in the present embodiment are the same as the configuration and initial setting value of the first embodiment described above, and thus the description thereof is omitted here.

  FIG. 9 is a flowchart of the operation in the second embodiment of the present invention.

  In the data transmitting apparatus 100, the error correction encoding unit 120 uses RS (n0, k0) using initial values n = n0, k = k0 for encoding stored in a memory (not shown). And communication is started via the communication path 300 (step 301).

  The quality measuring unit 210 in the data receiving apparatus 200 measures a burstiness value and a packet loss rate as quality information every unit time T (step 302), and notifies the data transmitting apparatus 100 of the measured quality information (step 302). 303).

  The data transmitting apparatus 100 monitors the quality information notified from the quality measuring section 210 of the data receiving apparatus 200 in the parameter determining section 110 (step 304), and the burstiness value as the quality information is set to the burstiness allowable value B. If it exceeds, or if it exceeds W% (W is a predetermined value) of the allowable burst value B X times (X is a predetermined value) continuously (step 305), the code length n becomes large. Thus, the burst tolerance value B is increased (step 306). Conversely, when the burstiness value, which is quality information, is significantly lower than the burstiness tolerance B (for example, below a predetermined Z%), or Y% of the burstiness tolerance B continuously X times (Step 305), the code length n is changed so as to be reduced, and the allowable burst value B is lowered (step 307). The predetermined values W, Y, Z are W> Y> Z.

  When the packet loss rate, which is quality information, exceeds or is likely to exceed the packet loss rate allowable value L (X times (X is a predetermined value) continuously) If it exceeds (predetermined value) (step 308), the redundancy is changed so as to increase, and the packet loss rate allowable value L is increased (step 309). Conversely, when the packet loss rate, which is quality information, is significantly below the allowable packet loss value L (for example, a predetermined Z% or less), or Y% of the allowable packet loss rate L for X consecutive times. (When it becomes below (step 308), the redundancy R is changed so as to be reduced, and the packet loss rate allowable value L is lowered (step 310). > Y> Z.

  When only one of the burstiness value and the packet loss rate, which is quality information, has a difference between the allowable values, n and k are determined so that only one of the code length n and the redundancy R is changed. Only one of the burst tolerance value B and the packet loss rate tolerance value L is changed. When both the burstiness value, which is quality information, and the packet loss rate have a difference from the allowable value, n and k are determined so that both the code length n and the redundancy R are changed, and the burstiness is determined. Both the allowable value B and the packet loss rate allowable value L are changed.

  The error encoding unit 120 of the data transmitting apparatus 100 performs RS (n, k) encoding processing using the encoding parameters determined by the parameter determining unit 110 (step 311).

  The information adding unit 130 of the data transmitting apparatus 100 adds the following information (1) to (4) to the header of the packet in the encoded block (step 312). Here, (3) and (4) are information determined by the parameter determination unit 110.

(1) Coding block number (2) Sequence number (3) n (code length)
(4) k (number of information packets)
The data transmitting apparatus 100 transmits the information packet to which the information for encoding is added and the redundant packet to the data receiving apparatus 200 (step 313).

  In the data receiving apparatus 200, when the information monitoring unit 220 receives the information packet and the redundant packet to which information for encoding is added, the error correction decoding unit 230 performs decoding based on the information, The packet loss of the corresponding coding block number is repaired and output (step 314).

  Note that when the number of packet losses in the coding block exceeds the number of redundant packets, it is impossible to repair the packet loss. Therefore, the data reception device 200 outputs the received packet of the coding block as it is. .

[Third Embodiment]
In this embodiment, an example in which a Reed-Solomon code and a crossing method are used together as an encoding method will be described.

  FIG. 10 shows the configuration and initial setting values of the data transmitting apparatus according to the third embodiment of the present invention.

  The data transmission apparatus 100 includes a parameter determination unit 110, an error correction coding unit 120, and an information addition unit 130.

  The data transmitting apparatus 100 has, as initial setting values, parameters n = n0, k = k0, R = R0, N = N0, burstiness tolerance B = B0, in a memory (not shown). The packet loss rate allowable value L = L0 is stored in advance. Note that n is the total number of packets (code length) in the coding block, k is the number of information packets in the coding block, R is the redundancy indicating the ratio of the number of redundant packets to the total number of packets in the coding block, and N is The number of crossings.

  Since the configuration and initial setting value of the data receiving apparatus 200 in the present embodiment are the same as the configuration and initial setting value of the first embodiment described above, description thereof is omitted here.

  Next, the operation of the present embodiment will be described.

  FIG. 11 is a flowchart of the operation in the third embodiment of the present invention.

  Below, the example which changes the frequency | count of crossing with the value of burstiness is demonstrated. This is merely an example, and the code length or the number of crossings and the code length may be set in advance to be changed.

  In the data transmitting apparatus 100, the error correction encoding unit 120 uses the initial setting values n = n0, k = k0, and N = N0 for the parameters for encoding, and RS (n0, k0) at the number of crossings N0. Encode and start communication (step 401).

  The quality measuring unit 210 in the data receiving apparatus measures the burstiness value and the packet loss rate as quality information every unit time T (step 402), and sends it to the data transmitting apparatus 100 as quality information (step 403).

  The parameter determination unit 110 of the data transmission device 100 monitors the quality information notified from the quality measurement unit 210 of the data reception device 200 (step 404), and the burstiness value as the quality information exceeds the burstiness tolerance B. Or when the number of crossings N exceeds W% (W is a predetermined value) of the allowable burst value B (X is a predetermined value) continuously (step 405). Thus, the burst tolerance value B is increased (step 406). Conversely, when the burstiness value, which is quality information, is significantly lower than the burstiness tolerance B (for example, below a predetermined Z%), or Y% of the burstiness tolerance B continuously X times If (Y> Z) or less (step 405), the number of crossings N is changed so as to decrease, and the burstiness allowable value B is lowered (step 407).

  When the packet loss rate, which is quality information, exceeds or is likely to exceed the packet loss rate allowable value L (X times (X is a predetermined value) continuously) If it exceeds (predetermined value) (step 408), the redundancy R is changed so as to increase, and the packet loss rate allowable value L is increased (step 409). Conversely, when the packet loss rate, which is quality information, is significantly lower than the packet loss rate allowable value L (for example, when it is equal to or less than a predetermined Z%), or Y of the packet loss rate allowable value L consecutively X times. If it is less than or equal to% (Y> Z) (step 408), the redundancy R is changed to be small, and the packet loss rate allowable value L is lowered (step 410).

  As described above, when only one of the burstiness and the packet loss rate as quality information has a difference from the allowable value, at least one of the code length n and the number of crossings N and only one of the redundancy R are changed. Thus, n, k, and N are determined, and only one of the burstiness tolerance B and the packet loss rate tolerance L is changed. When both the burstiness and the packet loss rate, which are quality information, have a difference from the allowable value, the number of crossings N (if the code length is selected in advance, the code length n, the code length and the crossover number are selected in advance) If the code length n and the number of times of intersection N) and the redundancy R are changed, k and N (if the code length is selected in advance, the code length n and the length of the code are crossed in advance) If the number of times is selected, the code length n and the number of crossings N) are determined, and both the burstiness tolerance B and the packet loss rate tolerance L are changed.

  The error correction encoding unit 120 in the data transmitting apparatus 100 performs RS (n, k) encoding processing of the number of crossings N using the encoding parameters determined by the parameter determining unit 110 (step) 411).

  The information adding unit 130 of the data transmitting apparatus 100 adds the following information (1) to (5) to the header of each packet in the encoded block (step 412). Here, (3), (4), and (5) are information determined by the parameter determination unit 110.

(1) Coding block number (2) Sequence number (3) n (code length)
(4) k (number of information packets)
(5) N (number of crossings)
The data transmitting apparatus 100 transmits the information packet to which the information for encoding is added and the redundant packet to the data receiving apparatus 200 (step 413).

  In the data receiving apparatus 200, when the information monitoring unit 220 receives the information packet and the redundant packet to which the information for encoding is added, the error correction decoding unit 230 performs decoding based on the information, The packet loss of the corresponding coding block number is repaired and output (step 414).

  If the packet loss rate in the coding block exceeds the number of redundant packets, it is impossible to repair the packet loss. Therefore, the data reception device 200 outputs the received packet of the coding block as it is. .

  The present invention constructs each function (means) of the data receiving apparatus (quality measuring apparatus) and data transmitting apparatus shown in the above embodiment as a program and uses it as a data receiving apparatus (quality measuring apparatus) and data transmitting apparatus. It is also possible to install the program on a computer to be executed and distribute the program via a network.

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

  The present invention is applicable to a network of best effort service.

It is a figure for demonstrating the principle of this invention. It is a principle block diagram of this invention. 1 is a basic configuration diagram of a system in an embodiment of the present invention. It is a flowchart (the 1) of operation | movement of the parameter determination of the data transmitter in one embodiment of this invention. It is a flowchart (the 2) of operation | movement of the parameter determination part of the data transmitter in one embodiment of this invention. It is a figure which shows the structure and initial setting value of a data receiver in the 1st Embodiment of this invention. It is a measurement example of burstiness in the first embodiment of the present invention. It is a figure which shows the structure and initial setting value of a data transmission apparatus in the 2nd Embodiment of this invention. It is a flowchart of the operation | movement in the 2nd Embodiment of this invention. It is a figure which shows the structure and initial setting value of a data transmitter in the 3rd Embodiment of this invention. It is a flowchart of the operation | movement in the 3rd Embodiment of this invention. This is error correction in units of IP packets using Reed-Solomon codes. It is error correction combining Reed-Solomon code and crossing method.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Data transmission apparatus 110 Parameter determination means, parameter determination part 120 Error correction encoding means, Error correction encoding part 130 Information addition means, Information addition part 200 Data reception apparatus 210 Quality measurement means, Quality measurement part 220 Information monitoring part 230 Decoding Means, error correction decoding unit

Claims (10)

A quality measuring method in a data receiving device, comprising:
In the data receiving device,
Store the number of packets, time interval, reference distribution in memory beforehand,
Data to be received is divided into quality measurement blocks for each number of packets acquired from the memory, and the number of quality measurement block occurrences with respect to the number of packet losses in the quality measurement block is distributed as the distribution at the time interval acquired from the memory. Representing the degree of divergence between the distribution and the reference distribution acquired from the memory as a burstiness value of the packet loss;
Obtaining a ratio of the number of packet losses in the time interval to the total number of packets to be received as a packet loss rate as quality information;
A quality measurement method characterized by: One or more pieces of data are input as a continuous data sequence, error-correction-coded, a data transmission device that transmits the data as a coded block via a communication channel, and the data of the coded block received from the communication channel is received and decoded A code error correction method in a system consisting of a data receiving device that generates and outputs data,
The data receiving device is:
A quality measuring step for measuring a packet loss burstiness value and a packet loss rate at a predetermined time interval according to the method of claim 1;
A quality information notifying step of periodically notifying the data transmitting device as the burstiness value and the packet loss rate measured in the quality measuring step;
The data transmission device includes:
When the burstiness value of the quality information received from the data receiving apparatus is larger than a predetermined value, the code length as an encoding parameter is increased, and when it is smaller, the code length is decreased, and the packet loss rate Is greater than a predetermined value, the parameter determination step of increasing the redundancy as an encoding parameter, and decreasing the redundancy when it is small;
An error correction coding step for performing coding based on the code length and the redundancy determined in the parameter determination step;
Information to be transmitted to the data receiving apparatus by adding information for encoding and decoding including the parameter determined in the parameter determining step to the data of the encoded block encoded in the error correction encoding step Granting step;
The data receiving device is:
A decoding step of decoding the data of the encoded block transmitted from the data transmission device using the information for encoding and decoding added to the data of the encoded block;
The code error correction method characterized by performing. One or more pieces of data are input as a continuous data sequence, error correction coding is performed, and a data transmission apparatus that transmits the data as a coded block via a communication path, and the data of the coded block received from the communication path are received and decoded. A code error correction method in a system consisting of a data receiving device that generates and outputs data,
When using both error correction and crossing method,
The data receiving device is:
A quality measurement step of measuring a packet loss burstiness value and a packet loss rate at a predetermined time interval by the quality measurement method according to claim 1;
A quality information notifying step of periodically notifying the data transmitting device as the burstiness value and the packet loss rate measured in the quality measuring step;
The data transmission device includes:
When the burstiness value of the quality information received from the data receiving device is larger than a predetermined value, either a preselected code length or the number of crossings or both are increased as an encoding parameter. If the packet loss rate is smaller, either the preselected code length or the number of crossings or both are reduced, and if the packet loss rate is greater than a predetermined value, the redundancy as an encoding parameter is increased, If it is small, the parameter determining step to reduce the redundancy,
An error correction coding step for performing coding based on the code length and the redundancy determined in the parameter determination step;
Information to be transmitted to the data receiving apparatus by adding information for encoding and decoding including the parameter determined in the parameter determining step to the data of the encoded block encoded in the error correction encoding step Granting step;
The data receiving device is:
A decoding step of decoding the data of the encoded block transmitted from the data transmission device using the information for encoding and decoding added to the data of the encoded block;
The code error correction method characterized by performing. A quality information measuring device included in the data receiving device,
Memory that stores the number of input packets, time interval, and reference distribution in advance,
Data to be received is divided into quality measurement blocks for each number of packets acquired from the memory, and the number of quality measurement block occurrences with respect to the number of packet losses in the quality measurement block is distributed as the distribution at the time interval acquired from the memory. Means for determining the degree of divergence between the distribution and the reference distribution acquired from the memory as a burstiness value of the packet loss;
Means for determining a ratio of the number of packet losses in the time interval to the total number of packets to be received as a packet loss rate as quality information;
A quality measuring apparatus comprising: One or more pieces of data are input as a continuous data sequence, error correction coding is performed, and a data transmission apparatus that transmits the data as a coded block via a communication path, and the data of the coded block received from the communication path are received and decoded. A code error correction system comprising a data receiving device that outputs the converted data,
The data receiving device is:
5. A quality measuring apparatus according to claim 4, wherein a burstiness value of packet loss and a packet loss rate at a predetermined time interval are measured, and the data transmission is performed using the burstiness value and the packet loss rate as quality information. Quality measuring means for periodically informing the device;
Decoding means for decoding the data of the encoded block transmitted from the data transmitting apparatus using information for encoding and decoding attached to the data of the encoded block;
The data transmission device includes:
When the burstiness value of the quality information received from the data receiving apparatus is larger than a predetermined value, the code length as an encoding parameter is increased, and when it is smaller, the code length is decreased, and the packet loss rate Parameter determination means for increasing the redundancy as an encoding parameter when the value is larger than a predetermined value, and decreasing the redundancy when it is small;
Error correction coding means for performing coding based on the code length and the redundancy determined by the parameter determination means;
Information to be transmitted to the data receiving apparatus by adding information for encoding and decoding including the parameter determined by the parameter determining means to the data of the encoded block encoded by the error correction encoding means And a sign error correction system. One or more pieces of data are input as a continuous data sequence, error correction coding is performed, and a data transmission apparatus that transmits the data as a coded block via a communication path, and the data of the coded block received from the communication path are received and decoded. A code error correction system comprising a data receiving device that outputs the converted data,
When using both error correction and crossing method,
The data receiving device is:
5. A quality measuring apparatus according to claim 4, wherein a burstiness value of packet loss and a packet loss rate at a predetermined time interval are measured, and the data transmission is performed using the burstiness value and the packet loss rate as quality information. Quality information measuring means for periodically notifying the device;
Decoding means for decoding the data of the encoded block transmitted from the data transmitting apparatus, using information for encoding and decoding given to the data of the encoded block;
The data transmission device includes:
When the burstiness value of the quality information received from the data receiving device is larger than a predetermined value, either a preselected code length or the number of crossings or both are increased as an encoding parameter. If the packet loss rate is smaller, either the preselected code length or the number of crossings or both are reduced, and if the packet loss rate is greater than a predetermined value, the redundancy as an encoding parameter is increased, A parameter determining means for reducing the redundancy when it is small;
Error correction coding means for performing coding based on the code length and the redundancy determined by the parameter determination means;
Information to be transmitted to the data receiving apparatus by adding information for encoding and decoding including the parameter determined by the parameter determining means to the data of the encoded block encoded in the error correction encoding step Granting means;
A code error correction system comprising: A quality measurement program,
Computer
Data to be received is divided into quality measurement blocks for each number of packets acquired from the memory, and the number of occurrences of quality measurement blocks with respect to the number of packet losses in the quality measurement block is expressed as a distribution at the time interval acquired from the memory. Means for determining the degree of divergence between the distribution and the reference distribution acquired from the memory as a burstiness value of packet loss;
Means for determining a ratio of the number of packet losses in the time interval to the total number of packets to be received as a packet loss rate as quality information;
Quality measurement program characterized by functioning as Computer
Data to be received is divided into quality measurement blocks for each number of packets stored in the memory in advance, and the number of times the quality measurement block is generated with respect to the number of packet losses in the quality measurement block in the time interval stored in advance in the memory As a distribution, the degree of divergence between the distribution and the reference distribution stored in advance in the memory as the burstiness value of the packet loss, and the ratio of the number of packet losses in the time interval to the total number of packets to be received A quality measurement means for notifying a data transmission apparatus connected via a communication path, the packet loss rate as quality information,
The coding block data transmitted from the data transmitting apparatus is made to function as a decoding unit that performs decoding using information for coding and decoding given to the coding block data. A code error correction program characterized by the above. Computer
When the burstiness value of quality information received from the data receiving apparatus is larger than a predetermined value, the code length as an encoding parameter is increased, and when it is smaller, the code length is decreased, and the packet loss rate is a predetermined value. Parameter determination means for increasing the redundancy as an encoding parameter if it is larger than, and decreasing the redundancy if it is smaller;
Error correction coding means for performing coding based on the code length and the redundancy determined by the parameter determination means;
Information to be transmitted to the data receiving apparatus by adding information for encoding and decoding including the parameter determined by the parameter determining means to the data of the encoded block encoded by the error correction encoding means A code error correction program which functions as an adding means. Computer
When the burstiness value of the quality information received from the data receiving device is larger than a predetermined value, either the code length selected in advance or the number of crossings, or both are increased and decreased as the encoding parameter Reduces either the preselected code length or the number of crossings, or both, and increases the redundancy as an encoding parameter when the packet loss rate is larger than a predetermined value, and redundancy when the packet loss rate is smaller Parameter determining means for reducing the degree,
Error correction coding means for performing coding based on the code length and the redundancy determined by the parameter determination means;
Information to be transmitted to the data receiving apparatus by adding information for encoding and decoding including the parameter determined by the parameter determining means to the data of the encoded block encoded in the error correction encoding step A code error correction program which functions as an adding means.

Images