JP2003304285A - Method for estimating degradation in quality of distribution service and apparatus therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for estimating a degradation in quality of distribution service and apparatus therefor in which a degradation in service distributed to receiving terminals from a distribution server is easily estimated. <P>SOLUTION: The apparatus 6 for estimating the degradation in the quality of the distribution service integrates the number of packets on distributing streams sent to the receiving terminals 3 from the distribution server 1 for every unit time and calculates ä(integrated value of number of packets < number of times reaching lower limit threshold PL1)+(integrated value of number of packets > number of times reaching upper limit threshold PU1)}×(t1/t3) for every degradation determining time t3 as variation R1. When the variation R1 exceeds the upper limit value of a threshold overrun time rate Q1, estimation is made that the quality of the distribution service is degraded. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a quality deterioration estimation method and apparatus for a distribution service in which contents including media such as video and audio are packetized from a distribution server using a communication network and distributed to a receiving terminal. Is.

[0002]

2. Description of the Related Art In recent years, information transmission technology using a communication network represented by a network such as the Internet has rapidly developed, and information such as images and music is distributed in real time from a distribution server to a receiving terminal through the communication network. It has become possible to do. When carrying out such a distribution service, it is important that the quality of the distribution information at the receiving terminal satisfies the user.

For example, as a method for a video distributor to detect the deterioration of the video quality that the user actually sees at the receiving terminal, a method of going to the user's house and viewing at the video receiving terminal used by the user, or a network There is a method of estimating the subjective evaluation value from the packet loss rate by measuring the state of packet loss that occurs, or a method of estimating the delivery service quality on the user side by the access log of the user remaining in the delivery server.

[0004]

However, when the video distribution company wants to see the video quality status of the user, even if he or she wants to see the actual video at the video receiving terminal used by the user, the video is viewed at all user homes. Is impossible. In addition, it is impossible to estimate that the network information is not available to measure the packet loss that occurs in the network. In addition, when using the access log of the video application that remains on the distribution server, the quality information written in the log depends on the implementation of the application and cannot be used for general purposes unless the definition of the log information is published. There is.

In view of the above problems, an object of the present invention is to provide a delivery service quality deterioration estimation method and apparatus capable of easily estimating the deterioration of the service delivered from the delivery server to the receiving terminal.

[0006]

In order to achieve the above object, the present invention provides a method of packetizing contents including media such as video and audio from a distribution server using a communication network and distributing the contents to a receiving terminal. The computer device monitors a variation amount of at least one of the number of packets or transmission throughput of a delivery stream sent from the delivery server to a predetermined receiving terminal, and the computer device has the variation amount exceeding a predetermined threshold value. At the same time, we propose a delivery service quality deterioration estimation method that estimates that the quality of the delivery service has deteriorated.

According to the delivery service quality deterioration estimation method, when the number of packets of the delivery stream or the variation amount of the transmission throughput becomes large, the computer device automatically detects it and the quality degradation of the delivery service becomes large. Presumed to be

Further, according to the present invention, as a method of monitoring the variation amount of the number of packets, the computer device integrates the number of packets transmitted from the distribution server to a predetermined receiving terminal every unit time (t1). A method is proposed for calculating the packet number integrated value and monitoring the variation amount of the transmission packet number integrated value as the variation amount at each predetermined deterioration determination time (t3) that is larger than the unit time (t1). .

Further, according to the present invention, as a method of monitoring the fluctuation amount of the transmission throughput, the packet length of a packet transmitted from the distribution server to a predetermined receiving terminal by the computer device every unit time (t1). A method of calculating a transmission throughput value by integrating the above, and monitoring the fluctuation amount of the transmission throughput value as the fluctuation amount at each predetermined deterioration determination time (t3) larger than the unit time (t1). suggest.

Further, the present invention proposes the following three methods as a method for estimating the quality deterioration of the distribution service from the variation amount of the number of packets.

As a first method, the computer device sets the upper limit threshold value (PU1), the lower limit threshold value (PL1) and the threshold excess time rate upper limit value (Q1) of the packet number integrated value to the attribute of the delivery stream. For each of the deterioration determination times (t3) with respect to the packet number integrated value, {(number of times packet number integrated value <PL1) + (packet number integrated value> PU1). Number of times)}
X (t1 / t3) is calculated as the fluctuation amount (R1),
The fluctuation amount (R1) is the upper limit value (Q
A method of estimating that the quality of the delivery service has deteriorated when the value exceeds 1).

[0012] As a second method, the computer device causes a preset threshold excess time rate upper limit value (Q11).
Is stored for each attribute of the delivery stream, and {standard deviation of accumulated packet number} / {average of accumulated packet number} is calculated for each deterioration determination time (t3) as the variation amount (R11). ), The fluctuation amount (R11)
Exceeds the threshold excess time rate upper limit value (Q11), it is estimated that the quality of the delivery service has deteriorated.

As a third method, the computer apparatus has a preset threshold excess time rate upper limit value (Q12).
Is stored for each attribute of the delivery stream, and the {P% value of the distribution of the integrated packet number} / {the average value of the integrated packet number} is changed for each deterioration determination time (t3). Calculated as the amount (R12), and the variation (R1
The method of estimating that the quality of the delivery service has deteriorated when 2) exceeds the upper limit value (Q12) of the threshold excess time rate.

Further, the present invention proposes the following three methods as a method of estimating the quality deterioration of the delivery service from the variation of the transmission throughput.

As a first method, the computer device sets an upper limit threshold value (PU2), a lower limit threshold value (PL2) and a threshold excess time rate upper limit value (Q2) of the transmission throughput value for each attribute of the delivery stream. The change is stored and stored, and for each deterioration determination time (t3), {(number of times transmission throughput value <PL2) + (number of times transmission throughput value> PU2)} × (t1 / t3) is changed. A method of estimating that the quality of the delivery service has deteriorated when the fluctuation amount (R2) exceeds the threshold excess time rate upper limit value (Q2).

[0016] As a second method, the computer device causes a preset threshold excess time rate upper limit value (Q21).
Is stored for each attribute of the delivery stream, and {standard deviation of transmission throughput value} / {average value of transmission throughput value} is set as the variation amount (R21) for each deterioration determination time (t3). Calculate and calculate the variation (R2
A method of estimating that the quality of the delivery service is deteriorated when 1) exceeds the upper limit value (Q21) of the threshold excess time rate.

As a third method, the computer device causes a preset threshold excess time rate upper limit value (Q22).
Is stored for each attribute of the delivery stream, and the {P% value of distribution of transmission throughput value} / {average of transmission throughput value} for each deterioration determination time (t3) is the variation amount ( R22), and estimates that the quality of the delivery service has deteriorated when the variation amount (R22) exceeds the threshold excess time rate upper limit value (Q22).

Further, according to the present invention, the correspondence between the subjective evaluation value, which is the user's subjective evaluation value of the distribution service quality, and the threshold value is stored in advance in the storage means, and the deterioration of the quality of the distribution service is estimated. Prior to this, a method is proposed in which a threshold value corresponding to the subjective evaluation value of the delivery service quality that is the provision target is acquired from the storage means and the threshold value is set as the predetermined threshold value.

Further, according to the present invention, as the distribution service quality deterioration estimating device for carrying out the distribution service quality deterioration estimating method, at least either the number of packets or the transmission throughput of the distribution stream transmitted from the distribution server to a predetermined receiving terminal. An integrating unit that integrates and outputs one of them for each unit time (t1), and a distribution service of the distribution service when the variation amount of the integrated value per unit time (t1) output from the integrating unit exceeds a predetermined threshold value. A delivery service quality deterioration estimation device is proposed, which comprises an estimation means for estimating that the quality has deteriorated and outputting the estimation result.

In the delivery service quality deterioration estimation device,
At least one of the number of packets or transmission throughput of the distribution stream transmitted from the distribution server to the predetermined receiving terminal by the accumulating unit is a unit time (t
1) is integrated and output for each, and by the estimation means,
When the variation amount of the integrated value per unit time (t1) output from the integrating means exceeds a predetermined threshold value, it is estimated that the quality of the delivery service is deteriorated and the estimation result is output.

Further, according to the present invention, in the configuration of the delivery service quality deterioration estimating device as a device for realizing the method for monitoring the variation in the number of transmitted packets, the delivery server determines a predetermined time every unit time (t1). The packet number accumulating means for accumulating the number of packets transmitted to the receiving terminal and outputting the accumulated result as a transmission packet number integrated value is provided in the accumulating means, and the predetermined deterioration larger than the unit time (t1) is performed. The amount of change in the cumulative value of the number of transmitted packets is calculated for each determination time (t3), and it is estimated that the quality of the delivery service has deteriorated when the amount of change in the cumulative value of the transmitted packet exceeds a predetermined packet number variation threshold value. The estimating means is provided with a packet number threshold excess determining means.

Further, according to the present invention, in the configuration of the delivery service quality deterioration estimating device as a device for realizing the method for monitoring the variation amount of the transmission throughput, the delivery server determines a predetermined time every unit time (t1). The integrating means is provided with a transmission throughput value calculating means for integrating the packet lengths of the packets sent to the receiving terminals and outputting the integrated result as a transmission throughput value.
Predetermined deterioration determination time (t3) greater than 1)
The throughput value threshold excess determining means for calculating the variation amount of the transmission throughput value for each and estimating that the quality of the delivery service is deteriorated when the variation amount of the transmission throughput value exceeds a predetermined throughput value variation threshold value. Prepared for estimation.

Further, in order to realize the above-mentioned three methods of estimating the quality deterioration by monitoring the variation of the number of transmitted packets, the present invention has the following device configuration.

That is, in the case of implementing the first method, in the configuration of the delivery service quality deterioration estimating device, the packet number threshold excess determining means has an upper threshold value preset for the integrated value of the transmission packet number. (PU1), a lower limit threshold (PL1), and a storage unit that stores a threshold excess time rate upper limit (Q1) as the packet number variation threshold for each attribute of the delivery stream, and the deterioration determination time (t.
For each 3), {(number of times packet number integrated value <PL1) + (number of times packet number integrated value> PU1)} ×
A means for calculating (t1 / t3) as the variation amount (R1); and a quality of the delivery service deteriorated when the variation amount (R1) exceeds the threshold excess time rate upper limit value (Q1). And means for estimating.

Further, in the case of implementing the second method, in the configuration of the delivery service quality deterioration estimation device, the packet number threshold excess determining means is set to exceed a preset threshold value with respect to the transmission packet number integrated value. Time rate upper limit value (Q
11) is stored as the packet number variation threshold value for each attribute of the distribution stream, and for each deterioration determination time (t3), {standard deviation of packet number integrated value} /
The {average value of the integrated value of the number of packets} is the fluctuation amount (R11)
And a means for estimating that the quality of the delivery service has deteriorated when the fluctuation amount (R11) exceeds the threshold excess time rate upper limit value (Q11).

Further, in the case of implementing the third method, in the configuration of the delivery service quality deterioration estimating device, the packet number threshold excess determination means is set to exceed the threshold value preset for the integrated value of the transmission packet numbers. Time rate upper limit value (Q
12) is stored as the packet number variation threshold value for each attribute of the delivery stream, and for each deterioration determination time (t3), {P% of the distribution of the packet number integrated value.
Value} / {average of packet count integrated value} is the fluctuation amount (R
12) and a means for estimating that the quality of the delivery service has deteriorated when the fluctuation amount (R12) exceeds the threshold excess time rate upper limit value (Q12).

Further, in order to realize the above three methods of estimating the quality deterioration by monitoring the variation of the transmission throughput, the present invention has the following device configuration.

That is, in the case of implementing the first method, in the configuration of the delivery service quality deterioration estimating device, the throughput value threshold excess determining means has an upper limit threshold value (PU2) preset for the transmission throughput value. ), A lower limit threshold value (PL2), and a threshold value excess time rate upper limit value (Q2) as the throughput value fluctuation threshold value for each attribute of the delivery stream, and for each deterioration determination time (t3). A means for calculating {(the number of transmission throughput values <PL2) + (the number of transmission throughput values> PU2) × (t1 / t3) as the variation amount (R2), and the variation amount (R2) Is above the threshold excess time rate upper limit value (Q2), it is estimated that the quality of the delivery service has deteriorated.

Further, in the case of implementing the second method, in the configuration of the delivery service quality deterioration estimating device, the throughput value threshold excess determining unit is configured to have a threshold excess time ratio preset for the transmission throughput value. A storage unit that stores an upper limit value (Q21) as the throughput value variation threshold value for each attribute of the delivery stream, and {standard deviation of transmission throughput value} / {transmission throughput value for each deterioration determination time (t3). Means for calculating the fluctuation amount (R21) as an average value of the fluctuation amount (R2)
1) exceeds the threshold excess time rate upper limit value (Q21), the means for estimating that the quality of the delivery service has deteriorated is provided.

Further, in the case of implementing the third method, in the configuration of the prestige service deterioration estimating device, the throughput value threshold excess determining unit is configured to have a threshold excess time ratio preset for the transmission throughput value. Upper limit (Q
22) is stored as the throughput value variation threshold value for each attribute of the delivery stream, and for each deterioration determination time (t3), {P% value of distribution of transmission throughput value} / {transmission throughput value Means for calculating the fluctuation amount (R22), and a fluctuation amount (R2).
2) exceeds the threshold excess time rate upper limit value (Q22), the means for estimating that the quality of the delivery service has deteriorated.

Further, according to the present invention, the storage means for storing the correspondence between the threshold value and the subjective evaluation value which is the user's subjective evaluation value for the distribution service quality, and the subjective evaluation of the distribution service quality to be provided. It is proposed to include a threshold value setting unit that acquires a threshold value corresponding to a value from the storage unit and sets the acquired threshold value as the threshold value of the estimation unit.

[0032]

DETAILED DESCRIPTION OF THE INVENTION An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing a connection example of a delivery service quality deterioration estimating apparatus according to an embodiment of the present invention, and FIG. 2 is a configuration diagram showing a delivery service quality deterioration estimating apparatus according to the first embodiment of the present invention. .

In FIG. 1, reference numeral 1 is a distribution server, which packetizes contents including media such as video and audio and packetizes the contents via a communication network 2 such as the Internet to receive terminals 3.
To deliver to. The receiving terminal 3 receives the packet distributed from the distribution server 1, converts the packet into video and audio, and outputs the video and audio.

Reference numeral 4 is a connection line connecting the distribution server 1 and the communication network 2, and reference numeral 5 is a distribution stream which is a packet flow between the distribution server 1 and the receiving terminal 3.

Reference numeral 6 denotes a distribution service quality deterioration estimating device (hereinafter, simply referred to as quality deterioration estimating device), which is a well-known C.
It is composed of a computer device mainly composed of PU, monitors the packet of the distribution stream 5 flowing through the connection line 4 via the monitoring line 7, estimates the quality deterioration of the distribution service, and outputs the estimation result. .

As shown in FIG. 2, the quality deterioration estimating device 6 includes a packet number monitoring device 61, a transmission throughput monitoring device 62, a media attribute designation section 63, and an output device 64.
It consists of and.

The packet number monitoring device 61 is a unit time packet number integrator (hereinafter, simply referred to as a packet number integrator).
611 and a packet number threshold excess determination device 612.

The packet number integrator 611 monitors the packets of the delivery stream 5 sent from the delivery server 1 to the receiving terminal 3 via the monitoring line 7, and delivers the delivery stream 5
For each attribute, the number of packets is integrated every unit time t1, and the integrated value is output to the packet number threshold excess determination device 612.

The packet count threshold excess determination device 612 stores a storage unit 61 for storing a threshold packet count table 6121 for each media attribute.
22 and a packet number threshold excess judgment unit 6123, and the packet number accumulator 61 calculates the packet number integrated value of the attribute designated by the media attribute designation unit 63 by the operator's operation.
Input from 1 and use this packet number integrated value to estimate whether or not the quality of the delivery service has deteriorated at each preset deterioration determination time t3 (> t1), and output the estimation result. Notify the operator via device 64.

The table 6121 for threshold value for the number of packets classified by media attribute shows the upper limit threshold value PU1 and the lower limit threshold value PL1 of the integrated value of the number of packets per unit time t1 The excess time rate upper limit value Q1 is stored in advance.

The packet number threshold excess judgment unit 6123 acquires the packet number integrated value for each unit time t1 corresponding to the attribute designated through the media attribute designation unit 63 by the operator's operation from the packet number integrator 611. , Upper limit threshold value PU1 and lower limit threshold value PL1 and upper limit value Q1 of threshold excess time ratio set in the table 6121 for threshold packet number by media attribute
Is used to estimate the quality deterioration of the delivery service.

When estimating this quality deterioration, the packet number threshold excess judgment unit 6123 determines that the packet number integrated value per unit time t1 exceeds the upper limit threshold PU1 within the deterioration judgment time t3, as shown in equation (1). Then, the sum of the number of times of occurrence and the number of times of falling below the lower limit threshold PL1 is obtained, and a value obtained by multiplying this by (t1 / t3) is calculated as the variation amount R1.

Variation amount R1 = {(number of times packet count integrated value <PL1) + (number of times packet count integrated value> PU1)} × (t1 / t3) (1) Further, the packet number threshold is exceeded The determination unit 6123 determines that the fluctuation amount R1
And the threshold excess time rate upper limit value Q1 are compared, and the fluctuation amount R1
It is presumed that the quality of service is deteriorated when exceeds the upper limit value Q1 of the threshold excess time rate (R1> Q1).

The upper limit threshold PU1 and the lower limit threshold PL1 are
For example, the average value of the integrated value of the number of packets per unit time t1 that obtains the optimum service quality is obtained, and the value obtained by adding (2 × standard deviation) to this average value is the upper threshold PU1.
Then, a value obtained by subtracting (2 × standard deviation) from the average value can be set as the lower limit threshold PL1.

A specific example of setting the upper limit threshold PU1 and the lower limit threshold PL1 is, for example, 3 for the distribution stream 5.
Monitoring is performed for a minute (deterioration determination time t3), and the number of transmission packets per second (unit time t1) is integrated in a state where the service quality is not deteriorated. As a result, 180 pieces of data of the integrated value of the number of packets are obtained in 3 minutes. This 18
An average value and a standard deviation are obtained based on 0 data, and an upper limit threshold value PU1 and a lower limit threshold value PL1 are set. Further, it is preferable that the threshold excess time rate upper limit value Q1 is appropriately set according to the attribute of the distribution stream.

Here, the total number of packets is the lower threshold PL.
When it is less than 1, the quality of service deteriorates due to lack of information. Also, the quality deteriorates when the integrated value of the number of packets exceeds the upper limit threshold PU1. This is the receiving terminal 3
Depending on the specifications of the application program operating in, there is a function having a function of detecting a packet loss in the network (communication network 2) or the like and transmitting a retransmission request to the distribution server 1 when the packet loss occurs. To do. When packet loss occurs, the receiving terminal 3 lacks information and the quality of service deteriorates. Further, in this case, when a packet loss occurs, a retransmission packet is sent from the distribution server 1 to the receiving terminal 3, so that more packets are sent than in normal times. Therefore, the quality is deteriorated even when the integrated value of the number of packets exceeds the upper limit threshold PU1. Even in such a case, in the present embodiment, it is possible to detect deterioration of service quality.

The transmission throughput monitoring device 62 includes a unit time throughput value accumulator (hereinafter simply referred to as a throughput value accumulator) 621 and a throughput value threshold excess judging device 6.
It is composed of 22 and.

The throughput value integrator 621 monitors the packet of the delivery stream 5 sent from the delivery server 1 to the receiving terminal 3 via the monitoring line 7, and the unit of the packet length is for each attribute of the delivery stream 5. The transmission throughput value is calculated by integrating at each time t1, and the transmission throughput value is output to the throughput value threshold excess determination device 622.

The throughput value threshold excess determination device 622 is
It is composed of a storage unit 6222 that stores a table 6221 for threshold throughput values by media attributes, and a throughput value threshold excess determination unit 6223, and a transmission throughput value corresponding to a packet of an attribute designated by the media attribute designation unit 63 by an operator's operation. Input from the throughput value integrator 611,
Using this transmission throughput value, it is estimated whether or not the quality of the delivery service is deteriorated at every preset deterioration determination time t3 (> t1), and the estimation result is output to the operator via the output device 64. To notify.

Table 62 for Threshold Throughput Value by Media Attribute
The upper limit threshold value PU2, the lower limit threshold value PL2, and the threshold excess time rate upper limit value Q2 of the transmission throughput value per unit time t1 are stored in advance in 21 for each attribute of the delivery stream 5 such as the encoding speed and the media classification. .

The throughput value threshold excess determination unit 6223 acquires from the throughput value integrator 621 the transmission throughput value for each unit time t1 corresponding to the attribute designated by the operator through the media attribute designation unit 63,
The quality degradation of the distribution service is estimated using the upper limit threshold PU2, the lower limit threshold PL2, and the threshold excess time rate upper limit value Q2 set in the threshold throughput value table 6221 for each media attribute.

When estimating this quality deterioration, the throughput value threshold excess judgment unit 6223 has the transmission throughput value per unit time t1 within the deterioration judgment time t3, as shown in equation (2), exceeding the upper threshold PU2. The sum of the number of times and the number of times of falling below the lower limit threshold PL2 is obtained, and a value obtained by multiplying this by (t1 / t3) is calculated as the variation amount R2.

Variation amount R2 = {(number of transmission throughput values <PL2) + (number of transmission throughput values> PU2)} × (t1 / t3) (2) Further, the throughput value threshold excess determination unit 6223 compares the variation amount R2 with the threshold excess time rate upper limit value Q2, and the variation amount R2 exceeds the threshold excess time rate upper limit value Q2 (R2> Q).
It is presumed that the service quality has deteriorated.

As in the case of the number of packets described above,
Upper threshold PU2 and lower threshold PL for throughput values
2 is, for example, an average value of throughput values per unit time t1 at which the optimum service quality is obtained, and a value obtained by adding (2 × standard deviation) to the average value is used as the upper limit threshold P.
U2 and a value obtained by subtracting (2 × standard deviation) from the average value can be set as the lower limit threshold PL2. Further, it is preferable that the threshold excess time rate upper limit value Q2 is appropriately set according to the attribute of the distribution stream.

Here, the throughput value is the lower limit threshold PL2.
When it falls below, the quality of service deteriorates due to lack of information. Also, when the throughput value exceeds the upper limit threshold value PU2, the service quality deteriorates for the same reason as above. In this embodiment, it is possible to detect the deterioration of the service quality even when the throughput value exceeds the upper threshold PU2 in this way.

As described above, according to the present embodiment, the distribution business operator directly sees or listens to the user's receiving terminal 3 for the quality condition of the video or audio distributed from the distribution server 1 to the user's receiving terminal 3. Even in situations where it is not possible to obtain network information.
By monitoring the number of packets and the state of the throughput value at the time of transmission of the distribution server 1, it is possible to easily estimate the quality deterioration of the distribution service.

Further, measuring and monitoring the number of packets transmitted from the distribution server 1 and the throughput value at the time of transmission is versatile because it does not depend on the application. Further, by using the time rate at which the number of packets per unit time t1 and the throughput value exceed the upper thresholds PU1 and PU2,
It is also possible to correlate it with the packet loss occurrence time rate in the network (communication network 2), whereby the deterioration of service quality detected by the distribution business operator is detected as deterioration by the network business operator even if the scale is different. It is possible.

Next, a second embodiment of the present invention will be described.

In the second embodiment, the packet number threshold excess judgment unit 6123 is used in the same device configuration as the first embodiment.
And a method of estimating deterioration of service quality in the throughput value threshold excess determination unit 6223 is changed.

That is, in the second embodiment, the threshold excess time rate upper limit value Q11 for each attribute of the distribution stream 5 is set in advance in the threshold packet number table 6121 for each media attribute.

Further, in the second embodiment, the packet number threshold excess judgment unit 6123 packetizes the packet number integrated value for each unit time t1 corresponding to the attribute designated through the media attribute designation unit 63 by the operator's operation. Number integrator 611
Table 6121 for the threshold number of packets by media attribute
The quality deterioration of the delivery service is estimated using the threshold excess time rate upper limit value Q11 set to.

When estimating this quality deterioration, the packet number threshold excess judgment unit 6123 uses the standard deviation and the average value of the packet number integrated value for each unit time t1 within the deterioration judgment time t3 as shown in equation (3). And the fluctuation amount R1
1 is calculated as {standard deviation of integrated value of packet number} / {average value of integrated value of packet number}.

Variation amount R11 = {standard deviation of integrated value of packet number} / {average value of integrated value of packet number} (3) Furthermore, the packet number threshold excess determination unit 6123 determines the variation amount R1.
1 and the threshold excess time rate upper limit value Q11 are compared, and the fluctuation amount R11 exceeds the threshold excess time rate upper limit value Q11 (R1
1> Q11), it is presumed that the quality deterioration of the delivery service has occurred.

The threshold excess time rate upper limit value Q11 is preferably set appropriately according to the attribute of the distribution stream.

On the other hand, in the second embodiment, the threshold excess time rate upper limit value Q21 for each attribute of the distribution stream 5 is preset in the media attribute threshold throughput value table 6221.

Further, in the second embodiment, the throughput value threshold excess determination unit 6223 determines the transmission throughput value for each unit time t1 corresponding to the attribute designated by the operator through the media attribute designation unit 63 as the throughput value. The quality deterioration of the delivery service is estimated using the threshold excess time rate upper limit value Q21 acquired from the integrator 621 and set in the media attribute threshold throughput value table 6221.

When estimating the quality deterioration, the throughput value threshold excess judging unit 6223 calculates the standard deviation and the average value of the transmission throughput value for each unit time t1 within the deterioration judging time t3 as shown in the equation (4). And the variation R21 is {standard deviation of transmission throughput value} /
Calculate {average transmission throughput value}.

Variation amount R21 = {standard deviation of transmission throughput value} / {average value of transmission throughput value} (4) Further, the throughput value threshold excess determination unit 6223 further includes the variation amount R21 and the threshold excess time rate upper limit value Q21. By comparing the above, it is estimated that when the fluctuation amount R21 exceeds the threshold excess time rate upper limit value Q21 (R21> Q21), the quality of the delivery service deteriorates.

The threshold excess time rate upper limit value Q21 is preferably set appropriately according to the attribute of the distribution stream.

Next, a third embodiment of the present invention will be described.

The third embodiment has the same device configuration as the above-described first embodiment, but has a packet number threshold excess judgment unit 6123.
And a method of estimating deterioration of service quality in the throughput value threshold excess determination unit 6223 is changed.

In other words, in the third embodiment, the threshold excess time rate upper limit value Q12 for each attribute of the distribution stream 5 is set in advance in the threshold packet number table 6121 for each media attribute.

Further, in the third embodiment, the packet number threshold excess judgment unit 6123 packetizes the packet number integrated value for each unit time t1 corresponding to the attribute designated by the operator through the media attribute designation unit 63. Number integrator 611
Table 6121 for the threshold number of packets by media attribute
Deterioration of service quality is estimated using the threshold excess time rate upper limit value Q12 set to.

When estimating this quality deterioration, the packet number threshold excess judgment unit 6123 uses the P% value of the distribution of the packet number integrated value for each unit time t1 within the deterioration judgment time t3 as shown in equation (5). And the average value of the packet number integrated values, and also calculates {P% value of the distribution of the packet number integrated values} / {average value of the packet number integrated values} as the variation R12.

Variation amount R11 = {P% value of the distribution of the integrated packet number} / {average value of the integrated value of the packet} (5) Here, P% of the P% value of the distribution of the integrated packet number Indicates the probability. For example, the 95% value of the distribution indicates the value that includes 95% of the distribution. If there are 100 data, the 95th The value of the data is called 95% value. For example, in the case of a normal distribution, it is said that 95% of the total falls within the range of mean ± 2 × standard deviation.

Further, the packet number threshold excess judgment unit 6123
Compares the fluctuation amount R12 with the threshold excess time rate upper limit value Q12, and when the fluctuation amount R12 exceeds the threshold excess time rate upper limit value Q12 (R12> Q12), the quality of the delivery service deteriorates. Presumably.

It is preferable that the upper limit value Q12 of the threshold excess time ratio is appropriately set according to the attribute of the distribution stream.

On the other hand, in the third embodiment, the threshold excess time rate upper limit value Q22 for each attribute of the distribution stream 5 is set in advance in the media attribute threshold throughput value table 6221.

Further, in the third embodiment, the throughput value threshold excess determination unit 6223 determines the transmission throughput value for each unit time t1 corresponding to the attribute designated by the operator through the media attribute designation unit 63 as the throughput value. The quality deterioration of the delivery service is estimated using the threshold excess time rate upper limit value Q22 acquired from the integrator 621 and set in the media attribute-based threshold throughput value table 6221.

When estimating this quality deterioration, the throughput value threshold excess determining unit 6223 determines the P% value of the distribution of the transmission throughput value for each unit time t1 within the deterioration determining time t3, as shown in equation (6). , And the average value of the transmission throughput values, and at the same time, calculates {P% value of the transmission throughput value} / {average value of the transmission throughput values} as the variation amount R22.

Variation amount R22 = {P% value of transmission throughput value} / {average value of transmission throughput value} (6) Further, the throughput value threshold excess determination unit 6223 determines the variation amount R22 and the threshold excess time rate upper limit value. It is estimated by comparing with Q22 that the quality of the delivery service is deteriorated when the fluctuation amount R22 exceeds the threshold excess time rate upper limit value Q22 (R22> Q22).

It is preferable that the threshold excess time rate upper limit value Q22 is appropriately set according to the attribute of the distribution stream.

Next, a fourth embodiment of the present invention will be described with reference to FIG. FIG. 3 is a configuration diagram of a delivery service quality deterioration estimation device in the fourth exemplary embodiment of the present invention.

As shown in FIG. 3, the quality deterioration estimating apparatus of the present embodiment differs from that of the above-described first to third embodiments in that it includes a subjective evaluation data storage section 65 and a table setting section 66. It is in the point of being prepared. Other components and operations are similar to those of the first to third embodiments. In the present embodiment, the subjective evaluation data storage unit 65 is added to the first embodiment.
And the table setting unit 66 is added. Only the differences between this embodiment and the first embodiment will be described below.

The subjective evaluation data storage section 65 is provided in the receiving terminal 3
Subjective evaluation value which is a value subjectively evaluated by the user regarding the quality of service, and the table 6121 for threshold packet number by media attribute and the table for threshold throughput value by media attribute
6221 thresholds PU1, PL1, Q1, PU2, PL2
The correspondence relationship with Q2 is stored. The data indicating the correspondence between the subjective evaluation value and each threshold is based on the data collected by the experiments described below.

In this experiment, the evaluation video is provided to the subject while changing the quality state of the network, the number of packets, and the throughput value, and the subject's scores for this evaluation video are collected. Examples of the evaluation type include items such as a degree of deterioration obstruction, a service satisfaction degree, and a service acceptance degree when the images are compared with each other by using images without deterioration as original images. For these items, as shown in FIG. 4, the scores are converted into numerical values, which are stored as data. Then, the average of the scores for all the subjects is used as the disturbance measure (DMOS: Degradation).
Mean Score), Service Satisfaction (MOS: Mean Opinion Sc
ore), and the acceptance rate. The subjective evaluation value means a generic name of these. For the evaluation conditions, see ITU-R Rec.50.
0-10, "Methodology for the subjective assessment o
f the quality of television pictures ", March 2000
Please refer to it because it is detailed in.

The table setting unit 66 acquires from the subjective evaluation data storage unit 65 each threshold value corresponding to the subjective evaluation value of the service quality that is the target of service specified by the operator's operation, and acquires each acquired threshold value as a media attribute. Table 6121 for different threshold number of packets and Table 6221 for throughput value by media attribute
Set to. This threshold is set by the media attribute designation unit 6
Performed for each media attribute specified in 3. Hereinafter, a specific example of acquisition of the threshold value in the table setting unit 66 will be described.

Here, a case where service satisfaction is used as a subjective evaluation value will be described. The service quality to be provided is managed so that 80% or more of the people voted for "satisfied", "satisfied", and "very satisfied" with regard to service satisfaction. . In this case, a service satisfaction MOS satisfying the above conditions (assuming it is 3.5) is obtained from the data in the subjective evaluation data storage unit 65, and the service satisfaction MOS satisfies 3.5. Each threshold value is extracted from the subjective evaluation data storage unit 65. Then, the extracted threshold values are set in the media attribute threshold packet count table 6121 and the media attribute throughput value table 6221. By setting each threshold value in this way, 80% of users can manage in an environment satisfying service quality.

In the present embodiment, the subjective deterioration data storage unit 65 and the table setting unit 66 are added to the quality deterioration estimating apparatus of the first embodiment, but the quality of the second or third embodiment is different. The subjective estimation data storage unit 65 is included in the deterioration estimation device.
Alternatively, the table setting unit 66 may be added.

In each of the above embodiments, the distribution service quality deterioration estimation device 6 is provided separately from the distribution server 1, but the distribution service quality deterioration estimation device 6 may be provided in the distribution server 1. In this case, the distribution service quality deterioration estimation process can be entirely implemented by software.

Further, the delivery service quality deterioration estimation device 6 of each of the above-mentioned embodiments includes the packet number monitoring device 61 and the transmission throughput monitoring device 62, but the device configuration including either one of them is the same as that of the above-described embodiment. It is possible to obtain almost the same effect.

Further, the threshold value excess determining units 6123 and 6223 may perform determination processing other than those described in the above embodiments.

[0094]

As described above, according to the present invention, a situation or network in which a delivery business operator cannot directly see or listen to the quality state of a service such as video or audio delivered to a user at a user's receiving terminal or network. Even if it is impossible to obtain the information of (1) or (2), it is possible to easily estimate the deterioration of the quality of the delivery service by monitoring the number of packets and the state of throughput at the time of transmission from the server. Furthermore, since the monitoring of the number of packets and the throughput value at the time of sending of the distribution server does not depend on the application, the method and apparatus of the present invention are versatile. It is also possible to correlate the excess time rate of the upper threshold of the number of packets per unit time with the time rate of packet loss occurrence in the network, and the network operator also measures the deterioration of the quality of the delivery service detected by the delivery operator. Even if they are different, it is possible to detect them as deterioration.

[Brief description of drawings]

FIG. 1 is a diagram showing a connection example of a distribution service quality deterioration estimation device according to an embodiment of the present invention.

FIG. 2 is a configuration diagram showing a delivery service quality deterioration estimation device according to the first embodiment of the present invention.

FIG. 3 is a configuration diagram showing a delivery service quality deterioration estimation device according to a fourth embodiment of the present invention.

FIG. 4 is a diagram illustrating a subjective evaluation value.

[Explanation of symbols]

1 ... Delivery server, 2 ... Communication network, 3 ... Receiving terminal, 4 ... Connection line, 5 ... Delivery stream, 6 ... Delivery service quality deterioration estimation device, 61 ... Packet number monitoring device, 611 ... Unit time packet number integrator, 612 ... Packet number threshold excess determination device, 6121 ... Threshold packet number table for each media attribute, 6122 ...
Storage unit, 6123 ... Packet number threshold excess determination unit, 62 ... Transmission throughput monitoring device, 621 ... Unit time throughput value accumulator, 622 ... Throughput value threshold excess determination device, 6221
... Threshold throughput value table for each media attribute, 6222 ... Storage unit, 6223 ... Throughput value threshold excess determination unit, 63 ... Media attribute designating unit, 64 ... Output device, 65 ... Subjective evaluation data storage unit, 66 ... Table setting unit.

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 5C061 BB07 CC03                 5K014 EA01 GA01                 5K030 HA08 HB01 HB02 MB09 MC08

Claims (20)

[Claims] 1. A method of estimating quality deterioration of a distribution service, wherein a content including media such as video and audio is packetized from a distribution server to a receiving terminal by using a communication network, and a predetermined reception from the distribution server is performed. The computer device monitors a variation amount of at least one of the number of packets or transmission throughput of the delivery stream sent to the terminal, and the computer device degrades the quality of the delivery service when the variation amount exceeds a predetermined threshold value. A method for estimating deterioration of delivery service quality, which is characterized in that 2. The unit of time for a computer (t)
For each 1), the total number of packets transmitted from the delivery server to the predetermined receiving terminal is calculated, and the predetermined deterioration determination time (t3) larger than the unit time (t1) is calculated. The delivery service quality deterioration estimation method according to claim 1, wherein the variation amount of the integrated value of the number of transmitted packets is monitored as the variation amount. 3. The unit of time (t)
For each 1), the transmission throughput value is calculated by accumulating the packet lengths of the packets sent from the distribution server to a predetermined receiving terminal, and the unit time (t1) is calculated.
The delivery service quality deterioration estimation method according to claim 1, wherein the fluctuation amount of the transmission throughput value is monitored as the fluctuation amount for each predetermined deterioration determination time (t3) larger than the above. 4. The computer device stores an upper limit threshold value (PU1), a lower limit threshold value (PL1) and a threshold excess time rate upper limit value (Q1) of the packet number integrated value for each attribute of the delivery stream. In addition, the deterioration determination time (t
For each 3), {(number of times packet number integrated value <PL1) + (number of times packet number integrated value> PU1)} ×
(T1 / t3) is calculated as the variation amount (R1), and the variation amount (R1) is the threshold excess time rate upper limit value (Q1).
The delivery service quality deterioration estimation method according to claim 2, wherein it is estimated that the quality of the delivery service has deteriorated when the delivery service quality is exceeded. 5. The computer device stores a preset threshold excess time rate upper limit value (Q11) for each attribute of the distribution stream, and stores a {packet for each deterioration determination time (t3). The standard deviation of the number integrated value} / {the average value of the packet number integrated value} is calculated as the fluctuation amount (R11), and the fluctuation amount (R11) is the upper limit value (Q) of the threshold excess time rate.
The delivery service quality deterioration estimation method according to claim 2, wherein it is estimated that the quality of the delivery service has deteriorated when the value exceeds 11). 6. The computer apparatus stores a preset threshold value excess time rate upper limit value (Q12) for each attribute of the delivery stream, and stores a {packet for each deterioration determination time (t3). The P% value of the distribution of the number integrated value} / {the average value of the packet number integrated value} is calculated as the fluctuation amount (R12), and the fluctuation amount (R12) is the upper limit value (Q) of the threshold excess time ratio.
The delivery service quality deterioration estimation method according to claim 2, wherein it is estimated that the quality of the delivery service has deteriorated when it exceeds 12). 7. The computer device comprises an upper limit threshold (PU2) and a lower limit threshold (PL2) of the transmission throughput value.
And the threshold excess time rate upper limit value (Q2) are stored and provided for each attribute of the delivery stream, and {(the number of transmission throughput values <PL2) + () for each deterioration determination time (t3). Transmission throughput value> P
U2)) × (t1 / t3) is calculated as the variation amount (R2), and the variation amount (R2) is the threshold excess time rate upper limit value (Q
The delivery service quality deterioration estimation method according to claim 3, wherein it is estimated that the quality of the delivery service has deteriorated when it exceeds 2). 8. The computer device stores a preset threshold excess time rate upper limit value (Q21) for each attribute of the delivery stream, and stores {transmission for each deterioration determination time (t3). The standard deviation of the throughput value} / {the average value of the transmission throughput values} is calculated as the fluctuation amount (R21), and the fluctuation amount (R21) is the upper limit value (Q) of the threshold excess time ratio.
21. The delivery service quality deterioration estimation method according to claim 3, wherein it is estimated that the quality of the delivery service has deteriorated when the value exceeds 21). 9. The computer device stores a preset threshold excess time rate upper limit value (Q22) for each attribute of the distribution stream, and stores {transmission for each deterioration determination time (t3). The P% value of the distribution of the throughput values} / {the average value of the transmission throughput values} is calculated as the fluctuation amount (R22), and the fluctuation amount (R22) is the upper limit value (Q) of the threshold excess time ratio.
22. The delivery service quality deterioration estimation method according to claim 3, wherein it is estimated that the quality of the delivery service has deteriorated when the value exceeds 22). 10. A correspondence relation between a subjective evaluation value, which is a user's subjective evaluation value for distribution service quality, and the threshold value is stored in advance in a storage unit, and is provided prior to estimation of quality deterioration of the distribution service. 10. The distribution service according to claim 2, wherein a threshold value corresponding to a target subjective evaluation value of the distribution service quality is acquired from the storage unit, and the threshold value is set as the predetermined threshold value. Quality deterioration estimation method. 11. A quality deterioration estimation device for a distribution service, which uses a communication network to packetize contents including media such as video and audio from a distribution server and distributes the packets to a receiving terminal, the predetermined reception from the distribution server. Integrating means for integrating and outputting at least one of the number of packets or transmission throughput of the delivery stream sent to the terminal every unit time (t1); and integrating per unit time (t1) output from the integrating means. A delivery service quality deterioration estimation device, comprising: an estimation unit that estimates that the quality of the delivery service has deteriorated when the variation amount of the value exceeds a predetermined threshold value and outputs the estimation result. 12. The unit for integrating the unit time (t
1) has a packet number accumulating means for accumulating the number of packets transmitted from the distribution server to a predetermined receiving terminal for each and outputting the accumulation result as a transmission packet number accumulating value; The variation amount of the integrated value of the number of transmitted packets is calculated for each predetermined deterioration determination time (t3) larger than (t1), and the variation amount of the integrated value of the transmitted packet number exceeds a predetermined packet number variation threshold value. The delivery service quality deterioration estimation device according to claim 11, further comprising a packet number threshold excess determination unit that estimates that the quality of the distribution service has deteriorated. 13. The unit for integrating the unit time (t
1) each has a transmission throughput value calculating means for accumulating the packet lengths of the packets transmitted from the distribution server to a predetermined receiving terminal and outputting the accumulation result as a transmission throughput value, wherein the estimating means comprises the unit The variation amount of the transmission throughput value is calculated for each predetermined degradation determination time (t3) that is greater than the time period (t1), and the distribution is performed when the variation amount of the transmission throughput value exceeds a predetermined throughput value variation threshold value. 12. A throughput value threshold excess judgment means for estimating that the quality of service has deteriorated is provided.
The delivery service quality deterioration estimation device described in. 14. The packet number threshold excess determining means is an upper threshold value (PU1), a lower threshold value (PL1), and a threshold value excess time rate as the packet number variation threshold value, which are set in advance with respect to the transmission packet number integrated value. A storage unit that stores an upper limit value (Q1) for each attribute of the delivery stream, and {(the number of packets accumulated value <the number of times PL1) + (the number of packets accumulated value) for each deterioration determination time (t3). > PU1
Number of times)} × (t1 / t3) is the fluctuation amount (R
1) as a means for calculating the fluctuation amount (R1) as the upper limit value (Q
The delivery service quality deterioration estimation device according to claim 12, further comprising means for estimating that the quality of the delivery service has deteriorated when the value exceeds 1). 15. The packet number threshold excess determination means uses the threshold excess time rate upper limit value (Q11) set in advance for the transmission packet number integrated value as the packet number variation threshold value for each attribute of the delivery stream. A storage unit that stores the information; and a unit that calculates {standard deviation of integrated value of packet number} / {average value of integrated value of packet number} as the variation amount (R11) for each deterioration determination time (t3). The fluctuation amount (R11) is the upper limit value (Q
The delivery service quality deterioration estimation device according to claim 12, further comprising means for estimating that the quality of the delivery service has deteriorated when the value exceeds 11). 16. The packet number threshold excess determination means sets, for each attribute of the delivery stream, a threshold excess time rate upper limit value (Q12) set in advance for the cumulative number of transmitted packets as the packet number variation threshold value. For each of the deterioration determination time (t3) stored, a {P% value of distribution of integrated packet number} / {average of integrated packet number} is calculated as the variation (R12). Means and the variation amount (R12) is the upper limit value (Q
The delivery service quality deterioration estimation device according to claim 12, further comprising means for estimating that the quality of the delivery service has deteriorated when the value exceeds 12). 17. The throughput value threshold excess determination means is configured to set an upper limit threshold value (PU2) and a lower limit threshold value (PL2) preset for the transmission throughput value, and a threshold value excess time rate upper limit value as the throughput value variation threshold value. A storage unit that stores (Q2) for each attribute of the delivery stream, and {(number of times transmission throughput value <PL2) + (transmission throughput value> P) for each deterioration determination time (t3).
U2)} × (t1 / t3) as the fluctuation amount (R2), and the fluctuation amount (R2) is the threshold excess time rate upper limit value (Q).
The delivery service quality deterioration estimation device according to claim 13, further comprising means for estimating that the quality of the delivery service has deteriorated when the value exceeds 2). 18. The throughput value threshold excess determination means stores a threshold excess time rate upper limit value (Q21) preset for the transmission throughput value as the throughput value variation threshold value for each attribute of the delivery stream. Storing means, means for calculating {standard deviation of transmission throughput value} / {average value of transmission throughput value} as the variation amount (R21) for each deterioration determination time (t3), and the variation amount (R21). R21) is the upper limit value (Q
21. The distribution service quality deterioration estimation device according to claim 13, further comprising means for estimating that the quality of the distribution service has deteriorated when the value exceeds 21). 19. The throughput value threshold excess determination means stores a preset threshold excess time rate upper limit value (Q22) for the transmission throughput value as the throughput value variation threshold value for each attribute of the delivery stream. A storage unit that stores, and a unit that calculates {P% value of distribution of transmission throughput value} / {average value of transmission throughput value} as the variation amount (R22) for each deterioration determination time (t3), The fluctuation amount (R22) is the upper limit value (Q
22. The distribution service quality deterioration estimation device according to claim 13, further comprising means for estimating that the quality of the distribution service has deteriorated when the value exceeds 22). 20. Corresponding to a subjective evaluation value of distribution service quality that is a provision target, a storage unit that stores a correspondence relationship between a subjective evaluation value that is a user's subjective evaluation value of distribution service quality and the threshold value. 20. A threshold value setting means for acquiring a threshold value from the storage means and setting the acquired threshold value as a threshold value for the estimating means.
The delivery service quality deterioration estimation device according to the item.