JP5229054B2 - Packet transmission / reception system - Google Patents

Description

  The present invention relates to a system for transmitting and receiving packets and a packet loss cause estimation method.

  The broad Internet (including the narrow Internet (The Internet), corporate LAN, NGN, etc.) takes a form in which a plurality of networks are interconnected by routers. There are a wide variety of network types here. For example, there are Ethernet (registered trademark) and optical cables as wired networks, and 802.11 and wide area mobile networks (3G / 3.5G / etc.) Exist as wireless networks. Routers interconnecting these various networks have a function of transferring packets between networks, and have a finite buffer for temporarily storing packets to be transferred.

  The operation when transmitting and receiving packets between terminal devices connected to the Internet is as follows. The sending terminal sends a packet to the network to which it is connected. The packet is transmitted over the network and arrives at the router (hereinafter referred to as network transmission). The packet is temporarily stored in the router buffer and then transferred to another network connected by the router (hereinafter referred to as router transfer). By repeating the above network transmission and router transfer, a packet transmitted from a transmission terminal arrives at a network to which a reception terminal is connected via a plurality of networks and routers. Then, the packet transmitted on the network arrives at the receiving terminal. However, there is a case where a packet transmitted from the transmission terminal does not arrive at the reception terminal and is discarded on the Internet (hereinafter referred to as packet loss). This packet loss is caused by either failure of router transfer or network transmission.

  An example of a method for estimating whether a packet loss is caused by a failure of router transfer or network transmission is described in Patent Document 1. In Patent Document 1, the cause of packet loss is estimated by the following procedure.

Suppose that the Internet is configured by (transmitting terminal + wired network group + wireless network group + receiving terminal).
A gateway is installed between the wired network group (packet loss due to router transfer failure) and the wireless network group (packet loss due to network transmission failure).
-After measuring the packet loss rate at each of the gateway and the receiving terminal, feed them back to the transmitting terminal.
By using the two packet loss rates that are fed back, the receiving terminal side determines whether the packet loss occurs in the wired network group or the wireless network group. That is, it is determined that the packet loss in the wired network group is caused by failure in router transfer, and the packet loss in the wireless network is determined as caused by failure in network transmission.

Japanese Patent No. 3708950

  However, the above-described method according to Patent Document 1 needs to measure the packet loss rate up to the middle of the communication path from the transmission terminal to the reception terminal. For this reason, it is necessary to install a dedicated device for measuring the packet loss rate in the middle of the communication path between the transmitting terminal and the receiving terminal. For example, when using the Internet in a narrow sense as the Internet, it is necessary to arrange a dedicated device in the Internet. Therefore, when the configuration of the Internet cannot be changed, the method of Patent Document 1 cannot be applied. Further, even when the configuration of the Internet can be changed, for example, when an in-house LAN is used as the Internet, the technique disclosed in Patent Document 1 is complicated because it is necessary to arrange a plurality of dedicated devices in the Internet.

  An object of the present invention is to determine whether a packet loss is caused by a failure of router forwarding or network transmission only in processing of a transmitting terminal and a receiving terminal in a system that transmits and receives packets via a communication path. There is to be able to do it.

  A first packet transmission / reception system according to the present invention is a packet transmission / reception system that transmits and receives packets between a transmission terminal and a reception terminal via a communication path composed of a plurality of networks and routers, and is a unit in the transmission terminal. A first measuring means for measuring a transition of the number of transmitted packets per time; a second measuring means for measuring a transition of the number of packet losses per unit time at the receiving terminal; and the measured transmission per unit time And an estimation means for calculating a correlation between the transition of the number of packets and the transition of the number of packet losses, and estimating the cause of the packet loss from the calculated correlation.

  The first packet loss cause estimation method of the present invention is a packet loss cause estimation method in a packet transmission / reception system that transmits and receives packets between a transmission terminal and a reception terminal via a communication path composed of a plurality of networks and routers. A) measuring a transition of the number of transmitted packets per unit time in the transmitting terminal; b) measuring a transition of the number of packet losses per unit time in the receiving terminal; c) measuring the measured Calculating the degree of correlation between the transition of the number of transmitted packets per unit time and the transition of the number of packet loss, and estimating the cause of the packet loss from the calculated degree of correlation.

  A first terminal device according to the present invention is a packet transmission source terminal device that transmits and receives packets to and from a packet transmission destination terminal device via a communication path constituted by a plurality of networks and routers, and has a unit time A first measuring means for measuring a transition of the number of transmitted packets per unit, a transition of the measured number of transmitted packets per unit time, and a packet loss number per unit time received from the terminal device of the packet destination An estimation unit that calculates a degree of correlation with the transition and estimates a cause of packet loss from the calculated degree of correlation;

  A second terminal device of the present invention is a packet transmission destination terminal device that transmits and receives packets to and from a packet transmission source terminal device via a communication path composed of a plurality of networks and routers, and has a unit time A first measuring means for measuring a transition of the number of transmitted packets per unit, a second measuring means for measuring a transition of the number of packet losses per unit time, and a transition of the measured number of transmitted packets per unit time, An estimation means for calculating a degree of correlation with the transition of the packet loss number and estimating a cause of the packet loss from the calculated degree of correlation;

  According to the present invention, in a system that transmits and receives packets via a communication path, it is estimated only by processing of a transmitting terminal and a receiving terminal that a packet loss is caused by a failure of router transfer or network transmission. Can do.

The block diagram of the 1st Embodiment of this invention. The figure which shows the transition of the number of transmission packets per unit time, and the number of packet losses. Explanatory drawing of the correlation derivation | leading-out by a single correlation coefficient. The block diagram of the 2nd Embodiment of this invention.

[First Embodiment]
FIG. 1 is a block diagram showing a configuration of a packet transmission / reception system according to the first embodiment of the present invention. With reference to FIG. 1, the system configuration of the present embodiment will be described in detail.

  The packet transmission / reception system of the present embodiment includes a transmission terminal 1 that is a packet transmission source and a reception terminal 2 that is a packet transmission destination. The transmission terminal 1 and the reception terminal 2 are configured by an information processing apparatus such as a personal computer having a communication function. The transmission terminal 1 and the reception terminal 2 are connected via a communication path 3. The communication path 3 includes a plurality of networks and routers. There are many types of networks. For example, there are Ethernet and optical cables as wired networks, and 802.11 and wide area mobile networks (3G / 3.5G / etc.) As wireless networks.

  The transmission terminal 1 includes a packet generation unit 1-1, a packet transmission unit 1-2, a packet loss cause determination unit 1-3, and a control unit 1-4. The packet generator 1-1 functions as a means for generating a packet. The packet transmitter 1-2 functions as a means for transmitting packets while adjusting the number of transmitted packets per unit time. The packet loss cause determination unit 1-3 determines whether the cause of packet loss is a failure of router transfer or network transmission based on the degree of correlation between the variation in the number of transmitted packets per unit time and the variation in the number of packet losses per unit time. It functions as a means for determining. The control unit 1-4 functions as a unit that performs control for taking measures against packet loss according to the cause of packet loss.

  The receiving terminal 2 includes a packet receiving unit 2-1 and a packet loss number returning unit 2-2. The packet receiver 2-1 functions as means for receiving a packet and measuring the number of packet losses. The packet loss number return unit 2-2 functions as means for feeding back the packet loss number to the transmission terminal 1.

  Next, the operation of the first embodiment of the present invention will be described in detail with reference to FIG.

  The packet generator 1-1 of the transmission terminal 1 generates a packet storing data. Any data can be used. For example, “streaming video”, “VOIP audio”, and the like can be considered. Note that when generating a packet, processing for packet loss countermeasures may be performed according to an instruction from the control unit 1-4 (details will be described later). The packet generator 1-1 supplies the generated packet to the packet transmitter 1-2 through the connection 1-a.

  The packet transmitting unit 1-2 transmits the supplied packet to the receiving terminal 2 through the connection 3-a. At this time, the packet transmission unit 1-2 varies the number of transmission packets per unit time. As a method of variation, there are a method (A) that uses variation in the number of generated packets per unit time of the packet generation unit 1-1, and a method in which the packet transmission unit 1-2 intentionally varies the number of transmitted packets (B There are two types. For example, when storing “streaming video” as data in a packet, the method (A) can be used. In this case, the number of packets generated per unit time by the packet generator 1-1 varies depending on the characteristics of the codec used for compressing the data. Therefore, the packet transmission unit 1-2 can vary the number of transmission packets per unit time by transmitting the packet every time a packet is supplied from the packet generation unit 1-1. For example, when “VOIP voice” is stored as data in a packet, the method (B) can be used. In this case, the number of packets generated by the packet generator 1-1 per unit time is often constant due to the characteristics of the codec used for compressing the data. Therefore, the packet transmission unit 1-2 changes the number of transmission packets per unit time by temporarily storing the packets supplied from the packet generation unit 1-1 and transmitting the packets while adjusting the transmission timing. . The packet transmitting unit 1-2 transmits the number of transmitted packets per unit time to the packet loss cause determining unit 1-3 through the connection 1-b.

  The packet receiving unit 2-1 of the receiving terminal 2 receives the packet transmitted from the transmitting terminal 1. At this time, not all the packets transmitted by the transmission terminal 1 arrive at the packet receiving unit 2-1, and some packets are lost in the communication path 3 due to network transmission or router transfer failure. The packet receiving unit 2-1 measures the number of packet losses per unit time and transmits the packet loss number to the packet loss number returning unit 2-2 through the connection 2-a. Here, any method can be used for packet loss detection in the packet receiving unit 2-1. For example, when RTP (Real-Time Transport Protocol) is used as a packet transmission protocol, it can be determined that a packet loss has occurred when a discontinuity occurs in the sequence number of the arriving RTP packet. In addition, if an arbitrary transmission protocol is used, the transmission terminal 1 inserts a sequence number in an arbitrary format in the transmission packet, and the packet reception unit 2-1 causes discontinuity of the sequence number. In such a case, it is possible to take a method for determining that a packet loss has occurred.

  The packet loss number return unit 2-2 periodically feeds back the transmitted packet loss number per unit time to the transmission terminal 1 through the connection 3-b. Here, any method can be used as the feedback method. For example, when RTP is used as a transmission protocol for packets transmitted from the data transmission unit 1-2, RTCP (RTP Control Protocol) is generally used for feedback from the packet loss number return unit 2-2. In this case, the number of packet losses can be periodically fed back from the packet loss number return unit 2-2 to the transmission terminal 1 by using RTCP RR (Receiver Report). Also, for example, a TCP (Transmission Control Protocol) connection is always established from the packet loss number return unit 2-2 to the transmission terminal 1, and the number of packet losses as a message such as HTTP (Hypertext Transfer Protocol) on the connection. It is also possible to take a feedback method.

  The packet loss cause determination unit 1-3 of the transmission terminal 1 includes a transition (a) of the number of transmission packets per unit time supplied from the packet transmission unit 1-2, and a packet per unit time fed back from the reception terminal 2. The degree of correlation between the two is obtained by comparing the transition (b) of the number of losses. When the degree of correlation between (a) and (b) is high, that is, when the number of transmitted packets per unit time increases, the number of packet losses increases, and when the number of transmitted packets decreases, the number of packet losses decreases. In this case, it is determined that the packet loss is caused by router transfer failure due to overflow of the finite buffer of the router. Conversely, when the degree of correlation between (a) and (b) is low, that is, when packet loss occurs regardless of the increase or decrease in the number of transmitted packets (for example, when packet loss always occurs at a constant rate). The packet loss is determined to be caused by network transmission failure. The reason for making such a determination is as follows.

(A) As will be described later, the router transfer failure is caused by a buffer overflow of the router. Therefore, “the number of packets per unit time received by the router” and “the probability of packet loss due to router transfer failure” are correlated.
(B) On the other hand, network transmission failure is not related to overflow of a finite buffer, so “the number of packets per unit time flowing into the network” and “probability of packet loss due to network transmission failure” do not correlate.
(C) Here, “the number of packets per unit time received by the router” and “the number of packets per unit time flowing into the network” are proportional to “the number of transmitted packets per unit time on the transmission side”. From (a) and (b), “If the degree of correlation between the number of transmitted packets per unit time and the number of packet losses per unit time is high, the cause of packet loss is router transfer failure. If the value is low, the cause of packet loss is network transmission failure ". As a method for determining whether the degree of correlation is high or low, for example, a threshold value is set in advance to discriminate between high and low, the correlation level is compared with the threshold value, and if the correlation level exceeds the threshold value, the correlation level Is determined to be high, and if the degree of correlation does not exceed the threshold, it can be determined that the degree of correlation is low.

  Here, any method can be used as a method for obtaining the degree of correlation. For example, a single correlation coefficient may be obtained. For example, the transition (a) of the number of transmitted packets per unit time supplied from the packet transmitting unit 1-2 and the transition (b) of the number of packet losses per unit time fed back from the receiving terminal 2 are shown in FIG. In this case, the single correlation coefficient of (a) and (b) is obtained as shown in FIG. 3, and the calculation result is 0.85. Here, the closer the single correlation coefficient is to 1, the higher the degree of correlation.

  The packet loss cause determination unit 1-3 transmits the determined packet loss cause (router transfer failure or network transmission failure) to the control unit 1-4 through the connection 1-c. The control unit 1-4 determines an effective packet loss countermeasure process according to the transmitted packet loss factor, and then performs the process to the packet generation unit 1-1 through the connection 1-d. Communicate. Effective countermeasures against packet loss differ depending on whether it is caused by router transfer or network transmission failure. Various countermeasures can be taken for each packet loss countermeasure. Some examples are given below.

(1) Countermeasures against packet loss due to router transfer failure First, the router transfer failure will be described. The router that receives the packet stores the packet in its own finite buffer. Then, after the network transmission of the transfer destination network becomes possible, the packet stored in the finite buffer is taken out and sent to the transfer destination network. Here, when the next packet is received before the transfer destination network becomes capable of network transmission, the packet is also stored in the finite buffer. As described above, when a situation in which a packet cannot be transferred just by receiving a packet in the finite buffer continues, the finite buffer becomes full of packets, and a new received packet cannot be stored in the finite buffer. As a result, the received packet is discarded without being transferred. The packet loss due to the overflow of the finite buffer of the router described above is the packet loss due to the router transfer failure.

  In order to prevent packet loss due to this router transfer failure, it is necessary to reduce the number of transmitted packets per unit time on the transmitting terminal side so that the router does not use up the finite buffer due to frequent packet reception. That is, it is effective to reduce the number of transmission packets per unit time in the packet transmission unit 1-2. For this purpose, it is necessary to reduce the number of generated packets per unit time in the packet generator 1-1. This is because, when data without restriction on the transmission time from the start to the end of data is stored in a packet, only the number of packets per unit time generated by the packet generator 1-1 is reduced without changing the data contents. Can be realized. In this case, the time required to finish transmitting the entire series of data from the start to the end increases. Also, when the transmission time from the start to the end of a series of data is limited, such as when storing "streaming video" or "VOIP audio" as data, increase the compression rate when compressing with a codec. By reducing the total amount of data, the number of packets generated per unit time by the packet generator 1-1 is reduced. In this case, the time required to complete the transmission from the start to the end of the series of data does not change.

(2) Measures against packet loss due to network transmission failure First, network transmission failure will be described. As already mentioned, there are various types of networks, but network transmission is not always successful depending on the physical damage of the network medium and the surrounding environment. For example, in the case of a wireless network, a network link error may occur due to the deterioration of radio wave conditions, and several percent of all packets may be lost during network transmission. Further, even in a wired network, a network link error may occur due to physical damage of a network medium, and the success rate of network transmission may decrease, resulting in packet loss. The packet loss during network transmission resulting from the above network link error is a packet loss due to network transmission failure. To prevent packet loss due to this network transmission failure, copy the data as multiple packets so that the data can be restored on the receiving terminal side by using a packet different from the lost packet. It is necessary to perform control such as transmission to increase packet redundancy on the transmission terminal side. That is, when a packet loss occurs in the communication path 3, it is effective to be able to restore the data stored in the packet using another packet that has arrived at the receiving terminal 2. For this purpose, it is conceivable to use FEC (Forward Error Collection). The simplest type of FEC is to copy the same data into a plurality of packets and store them. Thereby, even when a packet loss occurs in the communication path 3, the receiving terminal 2 is more likely to be able to restore data using another received packet.

  With the above mechanism, it is possible to determine whether the packet loss is caused by the failure of router transfer or network transmission only by the processing of the transmission terminal 1 and the reception terminal 2. Further, according to the present embodiment, it is possible to implement further effective packet loss countermeasures.

[Second Embodiment]
A form in which the cause of packet loss is determined by a receiving terminal that receives the packet is also conceivable. FIG. 4 is a block diagram showing this configuration. With reference to this figure, the configuration of the second exemplary embodiment of the present invention will be described in detail.

  The packet transmission / reception system of the present embodiment includes a transmission terminal 4 that is a packet transmission source and a reception terminal 5 that is a packet transmission destination. The transmission terminal 4 and the reception terminal 5 are configured by an information processing apparatus such as a personal computer having a communication function, for example. The transmission terminal 4 and the reception terminal 5 are connected via a communication path 6. Similar to the communication path 3, the communication path 6 is composed of a plurality of networks and routers.

  The transmission terminal 4 includes a packet generation unit 4-1, a packet transmission unit 4-2, and a control unit 4-3. The packet generator 4-1 functions as a means for generating a packet. The packet transmission unit 4-2 functions as a means for transmitting packets while adjusting the number of transmission packets per unit time. The control unit 4-3 functions as a unit that performs control for taking measures against packet loss according to the cause of packet loss.

  The receiving terminal 5 includes a packet receiving unit 5-1 and a packet loss cause determining unit 5-2. The packet receiver 5-1 functions as means for receiving a packet and measuring the number of transmitted packets and the number of packet losses. The packet loss cause determination unit 5-2 determines whether the cause of packet loss is a failure of router transfer or network transmission based on the degree of correlation between the variation in the number of transmitted packets per unit time and the variation in the number of packet losses per unit time. It functions as means for determining whether or not and feeding back the result to the receiving terminal 5.

  Next, the operation of the present embodiment will be described in detail with reference to FIG.

  The packet generation unit 4-1 of the transmission terminal 4 generates a packet storing data. As in the first embodiment, any data can be used. Note that when generating a packet, processing for countermeasures against packet loss may be performed according to an instruction from the control unit 4-3. The packet generator 4-1 supplies the generated packet to the packet transmitter 4-2 through the connection 4-a. The packet transmitter 4-2 transmits the supplied packet to the receiving terminal 5 through the connection 6-a. At this time, the packet transmission unit 4-2 varies the number of transmission packets per unit time. The method of variation is the same as the method described in the first embodiment.

  The packet receiver 5-1 of the receiving terminal 5 receives the packet transmitted from the transmitting terminal 4. At this time, not all packets transmitted by the transmission terminal 4 arrive at the packet receiving unit 5-1, and some packets are lost in the communication path 6 due to network transmission or router transfer failure. The packet receiving unit 5-1 measures the number of packet losses per unit time and transmits it to the packet loss cause determining unit 5-2 through the connection 5-a. Similar to the first embodiment, any packet loss detection method can be used. Further, the packet reception unit 5-1 obtains the number of packets transmitted by the transmission terminal 4 per unit time by adding the number of received packets per unit time and the number of packet losses. The packet receiving unit 5-1 transmits the obtained number of transmission packets per unit time to the packet loss determining unit 5-2 through the connection 5-a.

  The packet loss cause determination unit 5-2 compares the transition (a) of the number of transmitted packets per unit time and the transition (b) of the number of packet losses per unit time transmitted from the packet reception unit 5-1. Thus, the degree of correlation between the two is obtained. When the degree of correlation between (a) and (b) is high, that is, when the number of transmitted packets per unit time increases, the number of packet losses increases, and when the number of transmitted packets decreases, the number of packet losses decreases. In this case, it is determined that the packet loss is caused by router transfer failure due to depletion of the router's finite buffer. Conversely, when the degree of correlation between (a) and (b) is low, that is, when packet loss occurs regardless of the increase or decrease in the number of transmitted packets (for example, when packet loss always occurs at a constant rate). The packet loss is determined to be caused by network transmission failure. As described in the first embodiment, any method can be used as a method for obtaining the correlation degree and a method for determining the level of the obtained correlation degree.

  The packet loss cause determination unit 5-2 feeds back the determined packet loss cause (router transfer failure or network transmission failure) to the control unit 4-3 of the transmission terminal 4 through the connection 6-b. The control unit 4-3 determines processing for effective packet loss countermeasures according to the fed packet loss factor, and then performs the processing to the packet generation unit 4-1 through the connection 4-b. Communicate. As described in the first embodiment, various measures can be taken as countermeasures against packet loss.

  With the above mechanism, it is possible to determine whether the packet loss is caused by failure of router transfer or network transmission only by the processing of the receiving terminal 2. Further, according to the present embodiment, it is possible to implement further effective packet loss countermeasures.

[Other embodiments]
In the configuration of the first embodiment shown in FIG. 1, a third embodiment in which the packet loss cause determination unit 1-3 is moved from the transmission terminal 1 to the reception terminal 2 can be considered. In the third embodiment, the number of transmitted packets per unit time measured by the packet transmission unit 1-2 is transmitted from the transmission terminal 1 to the packet loss cause determination unit 1-3 of the reception terminal 2 through the communication path 3. The packet loss cause determination unit 1-3 includes a change in the number of transmitted packets per unit time received from the packet transmission unit 1-2, and a change in the number of packet losses per unit time notified from the packet reception unit 2-1. The cause of packet loss is estimated and the estimation result is fed back from the receiving terminal 2 to the transmitting terminal 1.

  Further, in the configuration of the second embodiment illustrated in FIG. 4, a fourth embodiment in which the packet loss cause determination unit 5-2 is moved from the reception terminal 5 to the transmission terminal 4 can be considered. In the fourth embodiment, the number of transmitted packets and the number of packet losses per unit time measured by the packet receiving unit 5-1 are returned to the packet loss cause determining unit 5-2 of the transmitting terminal 4 through the communication path 6. The

  In addition, the packet generating unit, the packet transmitting unit, the packet loss cause determining unit, and the control unit of the transmitting terminal in each of the above-described embodiments can implement the functions possessed by hardware, and the computer constituting the transmitting terminal And a program. The program is stored in the memory of the computer that constitutes the transmission terminal, and is read by the computer at the time of startup of the computer, and by controlling the operation of the computer, the packet generator, the packet transmission A packet loss cause determination unit and a control unit.

  In addition, the packet receiving unit, the packet loss number returning unit, and the packet loss cause determining unit of the receiving terminal in each of the above-described embodiments can be realized by hardware, and the computer constituting the receiving terminal Can be realized with a program. The program is stored in the memory of a computer constituting the receiving terminal, read by the computer when the computer is started up, etc., and by controlling the operation of the computer, the packet receiving unit, the packet loss A number return unit and a packet loss cause determination unit are realized.

DESCRIPTION OF SYMBOLS 1 ... Transmission terminal 1-1 ... Packet generation part 1-2 ... Packet transmission part 1-3 ... Packet loss cause determination part 1-4 ... Control part 2 ... Reception terminal 2-1 ... Packet reception part 2-2 ... Packet loss Number return unit 3 ... communication path 4 ... transmission terminal 4-1 ... packet generation unit 4-2 ... packet transmission unit 4-3 ... control unit 5 ... reception terminal 5-1 ... packet reception unit 5-2 ... packet loss cause determination Part 6 ... Communication path

Claims (11)

A packet transmission / reception system that transmits and receives packets between a transmission terminal and a reception terminal via a communication path composed of a plurality of networks and routers,
First measuring means for measuring a transition of the number of transmitted packets per unit time in the transmitting terminal;
Second measuring means for measuring a transition in the number of packet losses per unit time in the receiving terminal;
The degree of correlation between the measured transition of the number of transmitted packets per unit time and the transition of the number of packet loss is calculated, the cause of packet loss is estimated from the calculated degree of correlation , and in the estimation, the calculated degree of correlation Estimating means for estimating that the cause of packet loss is due to router transfer failure, and otherwise estimating that the cause of packet loss is due to network transmission failure ;
Packet transmission and reception system comprising a. A packet transmission / reception system that transmits and receives packets between a transmission terminal and a reception terminal via a communication path composed of a plurality of networks and routers,
First measuring means for measuring a transition of the number of transmitted packets per unit time in the transmitting terminal;
Second measuring means for measuring a transition in the number of packet losses per unit time in the receiving terminal;
An estimation means for calculating a degree of correlation between the measured transition of the number of transmitted packets per unit time and a transition of the number of packet losses, and estimating the cause of packet loss from the calculated degree of correlation;
With
The transmission terminal is a means for varying the number of transmitted packets per unit time.
A packet transmission / reception system. The transmitting terminal Ru provided with a control means for controlling the execution of packet loss countermeasures process corresponding to the cause of the estimated packet loss by the estimating means
Packet transmission system according to 請 Motomeko 1 or 2. The first measuring unit and the estimating unit are included in the transmitting terminal, the second measuring unit is included in the receiving terminal, and the second measuring unit transmits a measurement result through the communication path to the estimating unit. you sent to
Packet transmission system according to 請 Motomeko 3. Has said estimating means and said first measuring means and the second measuring means to the receiving terminal, said estimating means that sends the estimated result to the control unit through the communication channel
Packet transmission system according to 請 Motomeko 3. A packet transmission source terminal device that transmits and receives packets to and from a packet transmission destination terminal device via a communication path composed of a plurality of networks and routers,
A first measuring means for measuring a transition of the number of transmitted packets per unit time;
The degree of correlation between the measured transition of the number of transmitted packets per unit time and the transition of the number of packet losses per unit time received from the terminal device of the packet transmission destination is calculated, and the packet loss is calculated from the calculated degree of correlation. In the estimation, if the calculated degree of correlation is higher than a predetermined threshold, it is estimated that the cause of packet loss is a router transfer failure, otherwise the cause of packet loss is the network An estimation means for estimating a transmission failure ;
A terminal device comprising: A packet transmission destination terminal device that transmits and receives packets to and from a packet transmission source terminal device via a communication path composed of a plurality of networks and routers,
A first measuring means for measuring a transition of the number of transmitted packets per unit time;
A second measuring means for measuring the transition of the number of packet loss per unit time;
Calculating the correlation between the transition of the number of transmitted packets per unit time measured and the transition of the number of packet loss, and estimating means for estimating the cause of the packet loss from the calculated correlation;
End end device Ru provided with a <br/> means for varying the number of outgoing packets per unit time. A computer that constitutes a packet transmission source terminal device that transmits and receives packets to and from a packet transmission destination terminal device via a communication path constituted by a plurality of networks and routers.
A first measuring means for measuring a transition of the number of transmitted packets per unit time;
The degree of correlation between the measured transition of the number of transmitted packets per unit time and the transition of the number of packet losses per unit time received from the terminal device of the packet transmission destination is calculated, and the packet loss is calculated from the calculated degree of correlation. In the estimation, if the calculated degree of correlation is higher than a predetermined threshold, it is estimated that the cause of packet loss is a router transfer failure, and otherwise, the cause of packet loss is An estimation means for estimating network transmission failure;
Program to make it function . A computer that constitutes a packet transmission destination terminal device that transmits and receives packets to and from a packet transmission source terminal device via a communication path composed of a plurality of networks and routers.
A first measuring means for measuring a transition of the number of transmitted packets per unit time;
A second measuring means for measuring the transition of the number of packet loss per unit time;
Calculating the correlation between the transition of the number of transmitted packets per unit time measured and the transition of the number of packet loss, and estimating means for estimating the cause of the packet loss from the calculated correlation;
A program for functioning as means for varying the number of packets sent per unit time . A packet loss cause estimation method in a packet transmission / reception system that transmits and receives packets between a transmission terminal and a reception terminal via a communication path composed of a plurality of networks and routers,
a) measuring the transition of the number of transmitted packets per unit time in the transmitting terminal;
b) measuring a transition of the number of packet losses per unit time in the receiving terminal;
c) the correlation between the measured units of transitions and the number of packets lost in sending number of packets per time transient was computed to estimate the cause of the packet loss from correlation was said calculated in the estimated, and the calculated It is higher than a threshold correlation degree predetermined cause of packet loss is estimated that the failure of the router forwarding, steps and the including of estimating a cause of packet loss otherwise is the failure of network transmission packet loss cause estimation method. A packet loss cause estimation method in a packet transmission / reception system that transmits and receives packets between a transmission terminal and a reception terminal via a communication path composed of a plurality of networks and routers,
a) measuring the transition of the number of transmitted packets per unit time in the transmitting terminal;
b) measuring a transition of the number of packet losses per unit time in the receiving terminal;
c) calculating the degree of correlation between the measured transition of the number of transmitted packets per unit time and the transition of the number of packet losses, and estimating the cause of packet loss from the calculated degree of correlation;
d) the transmitting terminal fluctuating the number of transmitted packets per unit time;
Packet loss cause estimation method including

Images