JP2020039021A - Communication device, communication method and program - Google Patents

Abstract

To provide a communication device capable of preventing data delay in a jitter buffer caused by a clock skew between devices.SOLUTION: A communication device comprises a reception unit, a buffer controller, a transmission unit, and an estimation unit. The reception unit receives a packet from a counter device. The buffer controller controls writing and reading of the reception packet to and from a buffer. The transmission unit transmits the reception packet read out from the buffer to a transfer destination device. The estimation unit estimates a clock deviation reflection value to which clock deviation between the counter device and the own device is reflected, on the basis of a reception time of the reception packet and a time stamp included in the reception packet. The buffer controller decides a time when the reception packet is taken out from the buffer by using the clock deviation reflection value.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a communication device, a communication method, and a program.

  In the future, POI (Point Of Interface) will be converted to IP (Internet Protocol). In an IP network, it is ideal that jitter can be corrected by End-To-End, and it is expected that the amount of jitter of packets transmitted and received at network boundaries is within an allowable range in connection between carrier networks.

  Patent Literature 1 discloses a technique in which the number of packets stored in a jitter buffer is counted, and if the counter value is larger than a predetermined reference value, the packet stored in the jitter buffer is deleted.

  Patent Literature 2 discloses that a router constituting a group of routers is provided with a jitter buffer for storing voice packets transmitted on a communication network, and the router transmits the voice packets stored in the jitter buffer of the router to a reference clock signal. A technique for synchronizing and outputting an image is disclosed.

  Patent Document 3 discloses that in the case of communication using a mobile network, transmission by Delay Packing may be performed depending on the specifications of a base station (eNodeB) (for example, see FIG. 7 of Patent Document 3). ).

  Patent Document 3 discloses an operation in which a plurality of packets are transmitted in a burst. Normally, when a plurality of packets are not transmitted in a burst, the packets are transmitted and received as shown in FIG. On the other hand, when a plurality of packets are transmitted in bursts, packets are transmitted and received as shown in FIGS. FIG. 22 shows an example of transmission and reception of two packet bursts, and FIG. 23 shows an example of transmission and reception of three packet bursts.

JP 2011-101342 A JP 2001-345840 A JP 2013-046256 A

  The disclosures of the above prior art documents are incorporated herein by reference. The following analysis has been made by the present inventors.

  In recent years, research and development of NFV (Network Function Virtualization), which is a technology for virtualizing a network function (communication device), has been actively pursued. With the NFV, the network node of the communication carrier was once provided by a dedicated device, but the need to mount it on a virtualized general-purpose server has increased.

  Even in a network to which NFV is applied, suppression of jitter is required. However, the measures using the existing jitter buffer have the following problems: if the receiving speed is lower than the dequeuing speed due to clock skew, the delay will gradually increase until the packet is deleted, and the problem of packet deletion leading to voice degradation. is there.

  In the jitter countermeasure (measure using a jitter buffer) disclosed in Patent Literature 1, when the reception speed is lower than the dequeue speed due to clock skew (clock deviation), the problem that the delay gradually increases until the packet is deleted or the packet deletion is performed. However, there is a problem that leads to voice deterioration.

  Note that the clock skew indicates, for example, a deviation in the accuracy of a clock signal unique to a device (oscillator) in a device expressed in units of ppm (Parts Per Million). For example, the above problem will be described assuming that the clock accuracy is ± 100 ppm.

  If the clock accuracy of the transmitting device is +100 ppm and the clock accuracy of the receiving device is -100 ppm, the clock of the receiving device will advance 200 ppm ahead of the transmitting device. Therefore, when the receiving device transmits a packet at time intervals according to the RTP (Real Time Transport Protocol) time stamp value given to the received packet, the data accumulated in the jitter buffer in the receiving device is gradually depleted.

  On the other hand, when the clock accuracy of the transmitting device is −100 ppm and the clock accuracy of the receiving device is +100 ppm, the clock of the receiving device advances 200 ppm later than the transmitting device. When the receiving device transmits a packet at a time interval according to the RTP time stamp value given to the received packet, data stays in the jitter buffer in the receiving device.

  It is conceivable to use the technology disclosed in Patent Document 2 for a network to which NFV is applied. However, in order to realize the correction of the jitter at the network boundary by the NFV, the external synchronization using the dedicated hardware performed in the POI network as disclosed in Patent Document 2 cannot be used. That is, the method disclosed in Patent Document 2 has a configuration that requires an external clock, and requires dedicated hardware. Therefore, the technology disclosed in Patent Document 2 cannot be used for a network to which NFV is applied.

  As described above, in order to implement the jitter countermeasure according to Patent Document 2 with NFV, external synchronization using dedicated hardware, which has been performed in the existing POI network, cannot be used. Jitter countermeasures that can be realized).

  A main object of the present invention is to provide a communication device, a communication method, and a program that contribute to preventing data delay inside a jitter buffer due to clock skew between devices.

  According to a first aspect of the present invention or a disclosure, a receiving unit that receives a packet from an opposite device, controls writing and reading of a received packet to and from a buffer, a buffer control unit, and a packet that is read from the buffer A transmitting unit that transmits a received packet to a transfer destination device, and based on a reception time of the received packet and a time stamp included in the received packet, a clock deviation reflection value reflecting a clock deviation between the opposite device and the own device. An estimating unit for estimating, wherein the buffer control unit determines a time at which the received packet is extracted from the buffer using the clock shift reflection value.

  According to a second aspect of the present invention or a disclosure, a receiving unit that receives a packet from a counterpart device, a buffer in which the received packet is written, and transmits the received packet read from the buffer to a transfer destination device, A communication unit comprising: a transmitting unit, based on a reception time of the received packet and a time stamp included in the received packet, estimating a clock deviation reflection value reflecting a clock deviation between the opposite device and the own device; And determining a time at which the received packet is extracted from the buffer using the clock shift reflection value.

According to a third aspect of the present invention or a disclosure, a receiving unit that receives a packet from an opposite device, a buffer in which the received packet is written, and transmits the received packet read from the buffer to a transfer destination device, And a transmission unit, based on a reception time of the received packet and a time stamp included in the received packet, a clock deviation reflection value reflecting a clock deviation between the opposite device and the own device. And a process of determining a time at which the received packet is extracted from the buffer using the clock shift reflection value.
This program can be recorded on a computer-readable storage medium. The storage medium can be non-transient, such as a semiconductor memory, hard disk, magnetic recording medium, optical recording medium, and the like. The present invention can be embodied as a computer program product.

  According to each aspect of the present invention or the disclosure, a communication device, a communication method, and a program that contribute to preventing data delay inside a jitter buffer due to clock skew between devices are provided.

It is a figure for explaining an outline of one embodiment. FIG. 1 is a diagram illustrating an example of a schematic configuration of a communication system according to a first embodiment. FIG. 4 is a diagram for explaining an operation of the TrGW according to the first embodiment. FIG. 4 is a diagram for explaining an operation of the TrGW according to the first embodiment. FIG. 4 is a diagram illustrating an example of an operation of the TrGW according to the first embodiment. FIG. 3 is a diagram for explaining the advance or delay of a packet. FIG. 9 is a diagram for explaining an operation of determining the last packet of a burst packet by the TrGW. FIG. 4 is a diagram illustrating an example of connection information managed by a connection management unit. 5 is a flowchart illustrating an example of the operation of the receiving unit. It is a flowchart which shows an example of the receiving process of a receiving part. 6 is a flowchart illustrating an example of an operation of an enqueue unit. 9 is a flowchart illustrating an example of an enqueue process of an enqueue unit. 5 is a flowchart illustrating an example of the operation of the dequeue unit. 5 is a flowchart illustrating an example of an operation of a transmission unit. 9 is a flowchart illustrating an example of an operation of an estimation unit. 9 is a flowchart illustrating an example of an operation of DP (Delay Packing) determination by an estimation unit. It is a flowchart which shows an example of the operation | movement of inclination a1 and an adjustment amount by an estimation part. FIG. 3 is a diagram illustrating an example of a hardware configuration of a TrGW according to the first embodiment. FIG. 4 is a diagram for explaining control of the TrGW according to the first embodiment. It is a flowchart which shows an example of TS calculation processing for telephone-event correspondence. FIG. 4 is a diagram for explaining transmission and reception of a normal packet. FIG. 4 is a diagram illustrating an example of transmission and reception of two packet bursts. It is a figure showing an example of transmission and reception of a three packet burst.

  First, an outline of an embodiment will be described. It should be noted that the reference numerals in the drawings attached to this outline are added for convenience of each element as an example for facilitating understanding, and the description of this outline is not intended to limit the invention in any way. Further, connection lines between blocks in each drawing include both bidirectional and unidirectional. The one-way arrow schematically indicates the flow of a main signal (data), and does not exclude bidirectionality. Further, in a circuit diagram, a block diagram, an internal configuration diagram, a connection diagram, and the like shown in the disclosure of the present application, although not explicitly shown, an input port and an output port exist at an input terminal and an output terminal of each connection line. The same applies to the input / output interface.

  The communication device 100 according to one embodiment includes a reception unit 101, a buffer control unit 102, a transmission unit 103, and an estimation unit 104 (see FIG. 1). The receiving unit 101 receives a packet from a partner device. The buffer control unit 102 controls writing and reading of a received packet to and from a buffer. The transmitting unit 103 transmits the received packet read from the buffer to the transfer destination device. The estimating unit 104 estimates a clock shift reflection value reflecting a clock shift between the opposing device and the own device based on the reception time of the received packet and the time stamp included in the received packet. The buffer control unit 102 determines the time at which the received packet is extracted from the buffer using the clock shift reflection value.

  The communication device 100 estimates a clock shift reflection value reflecting a clock shift between the opposite device and the own device based on the packet reception time and the time stamp included in the packet. The TrGW 10 realizes jitter correction at an interval close to the transmission interval of the opposing device by using the value reflecting the clock shift for buffer control.

  Hereinafter, specific embodiments will be described in more detail with reference to the drawings. In each embodiment, the same components are denoted by the same reference numerals, and description thereof will be omitted.

[First Embodiment]
The first embodiment will be described in more detail with reference to the drawings.

  FIG. 2 is a diagram illustrating an example of a schematic configuration of the communication system according to the first embodiment. Referring to FIG. 2, the communication system includes a TrGW (Translation Gateway) 10 and an IBCF (Interconnect Border Control Function) 20.

  The communication system shown in FIG. 2 provides a service related to an IP telephone. In FIG. 2, a service relating to the IP telephone is provided to the terminals 30-1 and 30-2 belonging to different networks (networks).

  The TrGW 10 and the IBCF 20 provide a connection between different networks. The IBCF 20 provides a signaling function, and the TrGW 10 performs a Network Address Port Translation (NAPT) conversion or the like to provide an inter-network connection function.

  When the terminal 30-1 and the terminal 30-2 communicate, a session management function (CSCF; Call Session Control Function) transmits a SIP (Session Initiation Protocol) message to the IBCF 20. The TrGW 10 and the IBCF 20 communicate with each other through an interface defined by MEGACO (Media Gateway Control Protocol), and establish a session between the terminal 30-1 and the terminal 30-2.

  In the first embodiment, a case will be described in which the TrGW 10, which is an inter-network gateway, receives an RTP packet from the terminal 30-1, and transfers the received packet to the terminal 30-2.

  In an IP telephone, an audio signal is sampled at a predetermined cycle (for example, 8 kHz), a plurality of sample results are combined into one packet, and transmitted at predetermined intervals (for example, 20 ms). The TrGW 10 receives RTP packets transmitted from the terminal 30-1 at predetermined intervals, and transfers the RTP packets received at the predetermined intervals to the terminal 30-2, which is the transfer destination device of the terminal 30-1. At this time, the TrGW 10 absorbs the fluctuation (jitter) of the RTP packet transmitted from the terminal 30-1, which is the opposite device, and controls the RTP packet to be transmitted to the terminal 30-2 at a predetermined interval (FIG. 3). reference).

  At this time, the TrGW 10 controls the packet transfer so that the internal jitter buffer does not delay due to the clock skew between the terminal 30-1 as the opposing device and the own device. More specifically, the TrGW 10 brings the dequeuing speed of the jitter buffer closer to the RTP packet transmitted from the terminal 30-1, which is the opposite device, and the RTP packet stays in the jitter buffer at a timing after a predetermined period has elapsed. Adjust the time. That is, the TrGW 10 keeps the RTP packets in the jitter buffer such that the time (dequeuing speed) from when the packets are stored in the jitter buffer to when the packets are read approaches the RTP packet transmission interval of the terminal 30-1 which is the opposite device. Control the time you do.

  The TrGW 10 uses a high-resolution clock such as a Time Stamp Counter (hereinafter, referred to as TSC), which is a counter that is integrated for each operation clock of a general-purpose CPU (Central Processing Unit) at the time of receiving an RTP packet. (Hereinafter, referred to as an RT value). That is, the TrGW 10 measures the reception time (RT value) of the RTP packet with the accuracy of a high-resolution clock (for example, 4 GHz) built in the own device.

  The TrGW 10 estimates a clock deviation reflection value reflecting a clock deviation between the opposing device and the own device based on the packet reception time RT and the time stamp included in the packet. Specifically, the TrGW 10 calculates a slope of a regression line between the measured packet reception time RT and an RTP time stamp value (hereinafter, referred to as a TS value) as a clock shift reflection value. The TrGW 10 controls the dequeue time using the calculated slope (clock shift reflection value). Specifically, the TrGW 10 sets the interval for enqueuing and dequeuing the data packet in the jitter buffer to “cycle number difference from previous packet × slope”, so that the interval is close to the transmission interval of the terminal 30-1. Realizing jitter correction.

  For example, if the RTP time stamp value (TS value) and the packet reception time (RT value) are set on the X axis and the Y axis, respectively, a straight line as shown in FIG. 4 is obtained. The slope of the straight line is obtained by calculating (RT2-RT1) / (TS2-TS1). The TrGW 10 obtains the slope and calculates “cycle number difference from previous packet × slope”. If the calculation timing is the case of TS2 in FIG. 4, the cycle number difference from the previous packet is the number of clocks corresponding to (TS2−TS1). -RT1) The number of clocks is obtained. The TrGW 10 uses the number of clocks (RT2-RT1) as the dequeue time of the jitter buffer.

  As described above, the TrGW 10 determines the first received packet based on the time obtained by multiplying the difference between the corresponding time stamp of each of the first received packet and the second received packet received immediately before by the slope of the regression line. Is determined from the jitter buffer.

  Further, the TrGW 10 determines the delay packing reception in a situation where the delay packing can be received. The TrGW 10 accurately estimates the packet transmission interval of the terminal 30-1 by not using the packet delayed by Delay Packing for the estimation of the inclination based on the determination result. That is, when a plurality of packets are burst-received, the TrGW 10 does not use buckets other than the last received packet among the plurality of burst-received packets for calculating the slope of the regression line. The TrGW 10 does not include the delayed data in the sample data for estimating the slope of the regression line in an environment where burst reception such as Delay Packing is performed, so that the transmission interval of the opposing device (terminal 30-1) is reduced. Jitter correction at intervals close to

  First, a theoretical explanation for realizing the control of the TrGW 10 will be given.

<Adjustment of dequeue speed and delay time in TrGW 10>
A mechanism for adjusting the dequeuing speed of the TrGW 10 to the transmission speed (transmission interval) of the terminal 30-1, which is the opposite device, and adjusting the delay inside the TrGW 10 to operate at the jitter buffer depth set value will be described. Note that the buffer depth setting value is a TSC counter value corresponding to a delay time in the own device for absorbing fluctuation. Assuming that the packetization period is 20 ms, a TSC counter value corresponding to 20 ms is set as the buffer depth setting value. Further, assuming an environment in which burst transmission of three packets is performed, a TSC counter value corresponding to 60 ms (20 × 3 ms) is set as the buffer depth setting value.

  As described above, since the transmission interval of the packet by the TrGW 10 is proportional to the difference between the TS values, the slope (coefficient) of a straight line obtained from the TS value and the RT value is derived, and the coefficient (clock) derived as the difference between the TS values is derived. By multiplying by (shift reflection value), the packet can be transferred at the same transmission interval as that of the terminal 30-1.

  The relationship between the reception time RT of the RTP packet and the TS value given to the received RTP packet is as shown in the following equation (1) for each RTP session. Here, assuming that there is no large deviation in fluctuation, the slope of the regression line can be estimated by least squares estimation. Alternatively, it can be estimated by the slope between two samples as shown in the following equation (2).

[Equation 1]
RT (k) = a1 × TS (k) + a0 + fluctuation

  In Equation (1), RT (k) is the reception time of the k-th packet, TS (k) is the TS value assigned to the k-th packet, and a1 is the TS value 1 transmitted by the terminal 30-1 which is the opposite terminal. This is the time on the own device after the elapse of the clock (one clock of the TS value). A0 is a fixed component derived from an initial value such as an average delay time and a TS value randomly added.

[Equation 2]
a1 = {RT (ne) -RT (nb)} / {TS (ne) -TS (nb)}

  In equation (2), ne is a sample number after a lapse of a predetermined time, and nb is a reference sample number. Here, the sample number is a serial number of a packet in which “1” is set in the first received packet, which is unique within a session equivalent to the RTP sequence number.

  The time stamp described (included) in the RTP packet is a value given by the terminal 30-1 which is the opposite device. On the other hand, the reception time of the received RTP packet is a value given by the TrGW 10. In this way, the slope (gradient) a1 of the above equation (1) or (2) calculated from the time given by two different devices is a value reflecting the clock shift between the two devices.

  Hereinafter, the estimation result is determined at regular intervals, and the estimation result in each period is indicated by a1 (g). Note that g included in the above estimation result indicates the number of generations, and g = 0 indicates a value before estimation. a1 (0) is an initial value and indicates a rate converted into the number of clocks of the TrGW 10 (high-resolution clock, TSC).

  When the estimation result a1 (g) is significantly different from the initial value a1 (0), the TrGW 10 overwrites the estimation result a1 (g) with the initial value a1 (0). Specifically, the TrGW 10 employs the estimation result a1 (g) when the estimation condition represented by the following equation (3) is satisfied.

[Equation 3]
| A1 (g) -a1 (0) | / a1 (0) <α

  In Expression (3), α is a clock error range (for example, 0.1%) that is considered normal.

  The TrGW 10 sets a scheduled dequeue time for each packet data when enqueuing an internal jitter buffer. The TrGW 10 dequeues the packet from the jitter buffer and transmits the packet data to the outside when the scheduled dequeue time is reached for the packet data enqueued in the jitter buffer.

  The TrGW 10 reproduces the transmission interval between packets by dequeuing with a time corresponding to the codec-specific sampling frequency represented by the TS value difference obtained using the immediately preceding dequeued packet. Since there is an individual difference between the device clocks mounted on the TrGW 10 and the terminal 30-1, there is an error at first, and the TrGW 10 cannot accurately reproduce the packet transmission interval of the terminal 30-1.

  However, the TrGW 10 uses the received information of the RTP packet to derive the least-squares estimation of the equation (1) or the slope of the straight line by the equation (2), and obtains the packet transmission rate of the terminal 30-1 obtained from the slope ( (Transmission interval) is reflected in the dequeue interval of the jitter buffer. As a result, the jitter correction at an interval close to the transmission interval of the terminal 30-1 is realized.

  Next, the scheduled dequeue time and the transmission interval by the TrGW 10 will be described.

  Immediately after the two devices start communication (g = 0), the TrGW 10 has not been able to derive the transmission rate (transmission interval) of the terminal 30-1 which is the opposite device. Thus, the TrGW 10 determines the transmission interval by counting the time corresponding to the TS value difference with the clock of the own device for the dequeue interval of the jitter buffer.

  The TrGW 10 sets the scheduled dequeue time DQT according to the following equation (4).

[Equation 4]
First received packet: DQT (1) = RT (1) + jitter buffer depth set value Second packet and thereafter:
Update cycle packet: DQT (nu (g)) = {TS (k) −TS (k−1)} × a1 (g) + DQT (k−1) + adjustment amount (g)
After the non-updated periodic packet: DQT (k) = {TS (k) -TS (k−1)} × a1 (g) + DQT (k−1)

  Note that, in equation (4), nu (g) is the first sample number used for calculating the estimated dequeue time using the estimation result a1 (g). The initial value of nu (g) is 1 (nu (0) = 1).

  According to Expression (4), at the time of receiving the first packet, the TrGW 10 delays the received packet with the set value of the jitter buffer depth (for example, 20 ms) from the time (RT (1)) at which the first packet was received to the transfer destination device. Forward.

  Except at the time of receiving the first packet, the TrGW 10 delays and transfers the packet to the transfer destination device by a delay time determined by the slope a1 calculated from Expression (1) or Expression (2) and the adjustment amount (g).

  Note that the adjustment amount (g) is a value for eliminating a time difference (advance of time, delay of time) between the two devices accumulated due to a device clock shift (clock skew). If the advance is adjusted every time an RTP packet is received, it becomes a cause of fluctuation, so the TrGW 10 determines the adjustment amount (g) only when updating the number of generations.

  In equation (4), a packet to be dequeued immediately after the generation update during the enqueue process is referred to as an “update periodic packet”, and the other packets are referred to as “non-update periodic packets”.

  According to Expression (4), at the beginning of communication, the first RTP packet is transmitted to the transfer destination after a lapse of the jitter buffer depth setting value (for example, 20 ms) from the time RT (1) when the TrGW 10 first receives the RTP packet. . After that, the TrGW 10 calculates the slope a1 calculated from the packet reception time of the two RTP packets and the time stamp value TS described in the two packets. The TrGW 10 grasps the packet transfer interval of the terminal 30-1, which is the opposing device, by multiplying the calculated a1 by the difference between the time stamp values described in the two packets. The TrGW 10 reflects the obtained packet transfer interval in the previous scheduled dequeue time DQT (k-1), and realizes the same packet transfer interval as the counterpart device.

  Next, updating of the estimation result of the gradient a1 to the scheduled dequeue time calculation will be described.

  The TrGW 10 calculates the average value of the delay from the immediately preceding calculation of the slope a1 to the above timing at the timing of calculating the slope a1 according to the equation (1) or (2) according to the following equation (5). That is, the TrGW 10 calculates, for each generation, an average value of a period during which a packet is stored in the jitter buffer 202 in a predetermined period.

[Equation 5]

  In Equation (5), nb (g) is the first sample number of data used for estimating the estimation result a1 (g). Also, nb (0) = 1. In equation (5), ne (g) is the last sample number of data used for estimating a1 (g), and Na is the number of samples to be calculated. In Expression (5), some packets of Delay Packing are excluded from the average value calculation.

  In addition, when | jitter buffer depth−average delay | is larger than the threshold value and does not satisfy the condition of Expression (3), the TrGW 10 adjusts the adjustment amount (g) at the update timing (timing for calculating the gradient a1). ) Is initialized (set to 0). That is, the TrGW 10 updates the adjustment amount according to the following equation (6).

[Equation 6]
| Jitter buffer depth−average delay |>β; adjustment amount (g + 1) = − average delay (g)
| Jitter buffer depth−average delay | ≦ β; adjustment amount (g + 1) = 0

  In Expression (6), β is a threshold. For example, if the packetization period is 20 ms, β is set to about 5 ms. As described above, the TrGW 10 calculates the adjustment value based on the average value (average delay) calculated by Expression (5).

  The operation of the TrGW 10 is schematically illustrated in FIG. As shown in FIG. 5, the TrGW 10 absorbs the fluctuation (jitter) of the RTP packet transmitted from the opposite device for each session by the jitter buffer. At that time, the TrGW 10 estimates the transmission rate (transmission interval) of the opposing device, and reflects the estimated value in the control of the jitter buffer. As a result, the TrGW 10 can transfer the packet to the transfer destination device at the same transmission interval as that of the opposite device.

<Outlier determination in specific packet reception case>
As described above, a plurality of packets may be transmitted in bursts depending on the specifications of the base station (see FIGS. 22 and 23).

  When the TrGW 10 receives packets at packet intervals as shown in FIG. 22 and FIG. 23, the fluctuations are large, which may affect the estimation of the slope by the linear regression shown in Expression (1). Therefore, the TrGW 10 uses the last packet in a burst reception unit having a small delay at the base station in estimating the gradient a1.

  The TrGW 10 checks a reception interval between a packet received immediately before and a received packet. The TrGW 10 checks the reception interval by the following equation (7).

[Equation 7]
DEL (k) = {RT (k) −RT (k−1)} − {TS (k) −TS (k−1)} × a1 (g)

  In Expression (7), if the value of the reception interval DEL (k) is a negative value, the reception interval is shorter than expected. If the value is 0, the reception interval matches the assumption. Indicates that it is longer than expected. The TrGW 10 performs a threshold process on the reception interval to determine how many of the packetization periods the packet is advanced or delayed (see FIG. 6).

  The TrGW 10 calculates, as E1 (k), how many periods the received packet is advanced or delayed in terms of the packetization period. E1 (k) and E2 (k) and E3 (k) described later are integer values, respectively. For example, when the packetization cycle is 20 ms and the reception interval is 40 ms, E1 (k) is 2 since the packetization cycle is advanced by two packetization cycles.

  Also, the TrGW 10 calculates E2 (k) and E3 (k) according to the following equation (8) using the calculated E1 (k).

[Equation 8]
E1 (k) = E1 (k)
E2 (k) = E1 (k) + E1 (k-1)
E3 (k) = E2 (k) + E1 (k-2)

  Note that the TrGW 10 ends the calculation if any of the values is “0” when calculating E1 (k) to E3 (k).

  In the normal packet transmission / reception as shown in FIG. 21, no remarkable delay or advance occurs, so that E1 is "0". In the case of burst reception for every two packets as shown in FIG. 22, the reception returns to the normal timing for every two packets, so that E2 becomes “0”. When burst reception is performed for every three packets as shown in FIG. 23, reception returns to normal timing for every three packets, and thus E3 becomes “0”.

  The TrGW 10 determines the last packet of the burst packet using the values of E1 to E3. Specifically, the TrGW 10 determines the last packet of the burst packet using the relationship as shown in FIG. For example, referring to FIG. 7, if the calculation results of E1 to E3 are “−1”, “−2”, and “0”, respectively, the TrGW 10 indicates that burst reception is performed for every three packets. It can be understood that the th packet is the last packet. That is, the TrGW 10 can determine, using a combination of the above three variables, whether or not a periodic burst reception of three packets such as Delay Packing has occurred, and the last packet of the burst packet.

  Expression (8) can be extended to Ei (k) = E (i−1) (k) + E1 (ki−1). In equation (8), the presence / absence of periodic burst reception of packets is determined by three variables from i = 1 to 3, but burst transmission / reception with more packets can be determined by increasing the variables. .

[Device configuration]
Referring to FIG. 2, the TrGW 10 according to the first embodiment includes a reception unit 201, a jitter buffer 202, a buffer control unit 203, a transmission unit 204, an estimation unit 205, a TSC unit 206, a connection management unit 207. Further, the buffer control unit 203 includes an enqueue unit 211, a dequeue unit 212, and a storage unit for storing the scheduled dequeue time list 213.

  The receiving unit 201 is a unit that receives a packet from the terminal 30-1 which is the opposite device. The receiving unit 201 records the TSC counter value at the time of receiving the RTP packet, and stores a packet matching the session information in the receiving buffer.

  The jitter buffer 202 is a buffer for temporarily storing received packets (packet data) in order to absorb fluctuations.

  The buffer control unit 203 is a unit that controls writing and reading of a received packet to and from the jitter buffer 202. The buffer control unit 203 reads the data stored in the reception buffer and stores the data in the jitter buffer 202 (enqueue processing by the enqueue unit 211). The buffer control unit 203 writes, into the transmission buffer, data at the scheduled dequeue time among the data in the jitter buffer 202 (dequeue processing by the dequeue unit 212). At that time, the buffer control unit 203 determines a time (scheduled dequeue time) at which the received packet is extracted from the jitter buffer 202 using the clock deviation reflection value. In addition, the buffer control unit 203 transfers data to the estimation unit 205 at the time of enqueuing (copying the data).

  The scheduled dequeue time list 213 is a list in which scheduled dequeue times and packet data storage destinations of packet data are recorded for each session ID.

  The transmitting unit 204 is a unit that transmits the received packet read from the jitter buffer 202 to the terminal 30-2, which is the transfer destination device. The transmission unit 204 has a transmission buffer inside, and packetizes data stored in the transmission buffer and transmits it.

  The estimating unit 205 calculates the slope a1 that determines the transmission interval (transmission rate) of the opposing terminal and the delay adjustment amount based on the dequeued packet information, and notifies the buffer control unit 203 of the calculation result.

  The buffer control unit 203 calculates the scheduled dequeue time based on the obtained calculation result (slope a1, delay adjustment amount), and stores it in the scheduled dequeue time list.

  The TSC unit 206 is a high-resolution TSC (Time Stamp Counter) counter that is incremented by one every CPU cycle after the CPU mounted on the TrGW 10 is activated. The counter value by the TSC unit 206 is used for measuring the packet reception time in the reception unit 201 and confirming the transmission time by the dequeue process in the buffer control unit 203. Note that the clock that realizes the TSC unit 206 has higher frequency (resolution) and higher accuracy than at least the clock used for the operation of the reception unit 201 and the transmission unit 204 and the clock used for packet transmission and reception of the opposing device.

  The connection management unit 207 manages connection information in which transmission source session information and transfer destination session information are recorded. FIG. 8 is a diagram illustrating an example of the connection information managed by the connection management unit 207. The connection management unit 207 generates and manages connection information as shown in FIG. 8 using information obtained from the IBCF 20 or the like.

[Description of operation]
Next, the operation of the TrGW 10 according to the first embodiment will be described with reference to the drawings.

  FIG. 9 is a flowchart illustrating an example of the operation of the receiving unit 201.

  Upon receiving the packet (Step S01), the receiving unit 201 executes a receiving process (Step S02).

  FIG. 10 is a flowchart illustrating an example of the receiving process (step S02 in FIG. 9) of the receiving unit 201.

  The receiving unit 201 refers to the TSC counter value (Step S101).

  The receiving unit 201 checks whether or not connection information matching the received packet is included in the connection management information (step S102).

  If the matching information is included in the connection management information (Step S102, Yes branch), the receiving unit 201 stores the connection ID, the TSC counter value, and the packet data in the reception buffer (Step S103).

  If matching information is not included in the connection management information (step S102, No branch), the receiving unit 201 discards the received packet (packet data) (step S104).

  FIG. 11 is a flowchart illustrating an example of the operation of the enqueue unit 211 included in the buffer control unit 203.

  The enqueue unit 211 refers to the reception buffer (step S11) and checks whether or not unprocessed data exists (step S12).

  If there is unprocessed data (step S12, Yes branch), the enqueue unit 211 executes an enqueue process (step S13).

  If there is no unprocessed data (step S12, No branch), the enqueue unit 211 returns to step S11.

  FIG. 12 is a flowchart illustrating an example of the enqueue process (step S13 in FIG. 11) of the enqueue unit 211.

  The enqueue unit 211 increments (updates) cnt_eq, which is a sample number for enqueue (enqueue sample number) (step S201).

  The enqueue unit 211 determines an internal TS based on the updated enqueue sample number (Step S202).

  The enqueue unit 211 determines whether or not the number of generations has been updated based on the determined internal time stamp TS (step S203). Specifically, the enqueue unit 211 determines whether or not a generation has been updated based on whether or not a predetermined number of RTP packets have been received since the RTP packet was first received in each generation. For example, if the reception of 100 packets is determined to be one generation, the enqueue unit 211 has already received 100 packets in one generation, and when a new packet (the 101st packet) is received, determines that there is a generation update. I do.

  If the number of generations is updated (step S203, Yes branch), the enqueue unit 211 refers to the session ID, the number of generations, the slope a1, and the adjustment amount acquired from the estimation unit 205 (step S204).

  The enqueue unit 211 updates the inclination a1 (Step S205).

  The enqueue unit 211 determines the scheduled dequeue time DQT (cnt_eq (id)) using the updated slope a1 (step S206). Specifically, the enqueue unit 211 determines the scheduled dequeue time DQT using Expression (4).

  If the number of generations has not been updated (step S203, No branch), the enqueue unit 211 determines the scheduled dequeue time DQT using equation (4) (step S207).

  The enqueue unit 211 stores the enqueue sample number (cnt_eq (id)), the packet data, the connection information, the packet reception time RT (cnt_eq (id)), and the internal time stamp TS (cnt_eq (id)) in the jitter buffer 202 ( Step S208).

  The enqueue unit 211 registers the connection ID, the scheduled dequeue time DQT (cnt_eq (id)), and the storage location of the packet data in the scheduled dequeue list (step S209).

  The enqueue unit 211 writes the data to be enqueued into the jitter buffer 202 (Step S210). Further, the enqueue unit 211 delivers the data to be enqueued to the estimation unit 205.

  As described above, when the number of generations is updated, the TrGW 10 acquires the slope a1 and the adjustment amount, and determines the scheduled dequeue time using the acquired slope a1 and the adjustment amount. In other words, when the generation changes, the TrGW 10 reflects the adjustment value at the time when the packet is extracted from the jitter buffer 202 (scheduled dequeue time). In other words, the TrGW 10 reflects the adjustment value on the scheduled dequeue time, and after receiving a predetermined number of packets (after the generation is updated), reflects the adjustment value again on the scheduled dequeue time.

  FIG. 13 is a flowchart illustrating an example of the operation of the dequeue unit 212 included in the buffer control unit 203.

  The dequeue unit 212 refers to the TSC counter value (Step S301).

  The dequeue unit 212 refers to the scheduled dequeue time list 213 and determines the next dequeue target (step S302).

  The dequeue unit 212 checks whether or not the current TSC counter value is after the dequeue time (scheduled dequeue time) (step S303).

  If the current TSC counter value is equal to or later than the dequeue time (step S303, Yes branch), the dequeue unit 212 reads the packet data from the jitter buffer 202 and writes the data to the transmission buffer (step S304).

  If the current TSC counter value is not after the dequeue time (step S303, No branch), the dequeue unit 212 returns to step S301.

  The dequeue unit 212 increments (updates) cnt_dq (id), which is a dequeue sample number (dequeue sample number) (step S305).

  FIG. 14 is a flowchart illustrating an example of the operation of the transmission unit 204.

  The transmission unit 204 refers to the transmission buffer (Step S21) and checks whether or not unprocessed data exists (Step S22).

  If unprocessed data is included (step S22, Yes branch), the transmitting unit 204 packetizes the data (step S23).

  If unprocessed data is not included (step S22, No branch), the transmitting unit 204 returns to the process of step S21.

  The transmitting unit 204 transmits the packetized data (Step S24).

  FIG. 15 is a flowchart illustrating an example of the operation of the estimation unit 205.

  The estimating unit 205 receives a data packet for estimating the gradient a1 from the buffer control unit 203 (Step S31).

  The estimating unit 205 determines whether the data packet is the first reception packet in the generation g (Step S32).

  When the packet is the first reception packet (step S32, Yes branch), the estimation unit 205 calculates a difference (TS (k) -TS (nb (id))) from the first time stamp value TS (step S32). S33).

  If the packet is not the first received packet (step S32, No branch), the estimating unit 205 stores the first time stamp value TS (nb (id)) (step S34).

  The estimating unit 205 confirms the update cycle (Step S35). Specifically, the estimating unit 205 checks whether the difference calculated in step S33 is equal to or greater than a threshold.

  When the difference is equal to or larger than the threshold value (step S35, Yes branch), the estimating unit 205 determines the delay packing (DP determination; step S36).

  If the difference is not equal to or larger than the threshold (step S35, No branch), the estimating unit 205 temporarily stores the packet reception time RT, the time stamp value TS, and the scheduled dequeue time DQT of the generation g (step S37). .

  After the DP determination, the estimating unit 205 calculates the slope a1 and the adjustment amount g (Step S38).

  After that, the estimation unit 205 updates the number of generations (Step S39).

  FIG. 16 is a flowchart illustrating an example of the operation of the DP determination (step S36 in FIG. 15) by the estimation unit 205.

  The estimating unit 205 calculates E1 to E3 based on the above equation (8) (step S401).

  The estimating unit 205 checks the calculated value of E1 (step S402).

  If the value of E1 is “0”, the estimation unit 205 determines that the packet is an estimation target (step S405). If the value of E1 is “−1”, the estimating unit 205 executes Step S403. If the value of E1 is other than “0” or “−1”, the estimating unit 205 determines that the packet is out of the estimation target (step S406).

  The estimating unit 205 checks the calculated value of E2 (step S403). If the value of E2 is “0”, the estimating unit 205 determines that the packet is an estimation target (step S405). If the value of E2 is “−2”, the estimating unit 205 executes step S404. If the value of E2 is other than “0” or “−2”, the estimating unit 205 determines that the packet is out of the estimation target (step S406).

  The estimating unit 205 checks the calculated value of E3 (step S404). If the value of E3 is “0”, the estimation unit 205 determines that the packet is an estimation target (Step S405). If the value of E3 is other than “0”, the estimating unit 205 determines that the packet is out of the estimation target (step S406).

  FIG. 17 is a flowchart illustrating an example of the operation of calculating the inclination a1 and the adjustment amount (Step S38 in FIG. 15) by the estimation unit 205.

  The estimating unit 205 calculates the slope a1 using the equation (1) or (2) (step S501).

  The estimating unit 205 determines whether the calculated gradient a1 satisfies the condition of Expression (3) (confirms that a1 satisfies a predetermined condition; step S502).

  When the condition is satisfied (Yes in Step S502), the estimating unit 205 calculates an adjustment amount according to Expressions (5) and (6) (Step S503).

  When the condition is not satisfied (step S502, No branch), the estimating unit 205 initializes the slope a1 and the adjustment amount g (step S504). Specifically, a1 (g) = a1 (0) and the adjustment amount g = 0 are set.

  The estimating unit 205 notifies the buffer control unit 203 of the gradient a1 (g) and the adjustment amount (g) (step S505).

[Hardware configuration]
FIG. 18 is a diagram illustrating an example of a hardware configuration of the TrGW 10. The TrGW 10 has a configuration illustrated in FIG. For example, the TrGW 10 includes a CPU (Central Processing Unit) 11, a memory 12, a NIC (Network Interface Card) 13 as communication means, and the like, which are mutually connected by an internal bus.

  However, the configuration illustrated in FIG. 18 is not intended to limit the hardware configuration of the TrGW 10. The TrGW 10 may include hardware (not shown). The number of CPUs and the like included in the TrGW 10 is not limited to the example illustrated in FIG. 18. For example, a plurality of CPUs 11 may be included in the TrGW 10.

  The memory 12 is a random access memory (RAM), a read only memory (ROM), an auxiliary storage device (such as a hard disk), or the like.

  The function of the TrGW 10 is realized by the processing module described above. The processing module is realized, for example, by the CPU 11 executing a program stored in the memory 12. Also, the program can be downloaded via a network or updated using a storage medium storing the program. Further, the processing module may be realized by a semiconductor chip. That is, the function performed by the processing module may be realized by any hardware or software executed using the hardware.

  As described above, the TrGW 10 according to the first embodiment uses the RTP time stamp given to the RTP packet and the TSC (Time Stamp Counter) value at the time of receiving the RTP packet to determine the clock with the terminal 30-1 which is the opposite device. The shift is estimated for each session. The TrGW 10 uses the estimated value to synchronize the packet transfer interval when transferring the received RTP packet to the transfer destination device with the transmitting terminal, thereby reducing the delay inside the jitter buffer 202 due to a clock shift (clock skew). To prevent. As a result, when the receiving speed is lower than the dequeuing speed due to the clock skew, the delay of the packet is gradually increased until the packet is deleted, and the problem of the existing jitter buffer control in which the packet deletion leads to the voice deterioration is solved. That is, even in an environment where network synchronization cannot be used, the TrGW 10 can transfer the RTP packet to the transfer destination device at a transmission interval close to the RTP packet transmission interval of the opposing device and at a fixed delay time.

  For example, referring to FIG. 19, the delay (packet transfer interval) may be reduced by the amount of adjustment performed appropriately, regardless of whether the reception speed is faster or slower than the dequeue speed due to a clock shift. Converge.

  For example, consider a case where the clock accuracy of the opposing device (terminal 30-1) is -100 ppm, and the clock accuracy of the TrGW 10 is +100 ppm. In this case, the clock of the TrGW 10 advances 200 ppm later than the counterpart device. When the TrGW 10 transmits a packet at a predetermined time interval (for example, 20 ms of a packetization period) according to the RTP time stamp value (TS value) given to the RTP packet received by the TrGW 10, the data is stored in the jitter buffer 202 in the TrGW 10. Stays. However, in the TrGW 10 according to the first embodiment, even when a delay occurs due to a clock shift as in the above example, the jitter buffer 202 is controlled to cancel the shift.

  In the above example, the TrGW 10 perceives the time of 20 ms × (1-200 ppm) as 20 ms and performs packet transfer. As a result, if the above control is not performed, one packet stays in 100 seconds (= 20 ms / 200 ppm), but the TrGW 10 can eliminate the stay of the packet by the above control.

  In addition, in the clock shift estimation under the situation transmitted by Delay Packing, delay packing is determined from the periodicity of burst reception to determine whether or not Delay Packing is performed. Is not used for estimation.

[Modification]
It should be noted that the configuration and operation of the communication system described in the above embodiment are examples, and various modifications are possible. For example, the communication system may have the following configuration or operation.

  After performing the jitter absorption other than the voice packet, the TrGW 10 may adjust the transmission interval according to the RTP time stamp difference.

  Alternatively, the TrGW 10 may estimate the clock shift from the duration of RFC (Request for Comments) 4733.

  In order to transmit a DTMF (Dual-Tone Multi-Frequency) signal, a telephone-event signal described in RFC4733 may be used in an IP telephone. AMR-NB and G. The RTP timestamp value given to the voice packet such as 711 means the time of the head data of the packet, while the telephone-event signal means the time of the start of the same DTMF event. The duration of the DTMF is represented by a duration field.

  Normally, the packetization cycle is the same as the voice packet, and the duration of the DTMF event is often 40 ms or more, and packets having the same RTP time stamp and different duration values are continuously transmitted in a 20 ms cycle. is there. In addition, the last packet of the event is set to be transmitted with Ebit = 1, and transmission of the same packet three times is specified by RFC 4733.

  For the above reason, in order to reproduce the transmission interval of the telephone-event packet, it is not enough to refer to the RTP time stamp, and it is necessary to use a value converted to the RTP time stamp at the time of dequeue.

  FIG. 20 is a flowchart illustrating an example of a TS calculation process for a telephone-event.

  The estimating unit 205 determines whether the received packet is a telephone-event packet or a voice packet (step S601).

  If the packet is a voice packet, the estimating unit 205 sets the RTP time stamp value as an internal time stamp used for setting the dequeue time (step S602). Specifically, the estimation unit 205 sets the internal TS value (cnt_eq) = TS value (Lcnt_eq).

  In the case of a telephone-event, the estimating unit 205 uses an Ebit, Duration, or RTP sequence number to convert it to an RTP time stamp value equivalent to a voice packet. Specifically, the estimating unit 205 checks the value of Ebit (step S603).

  If not Ebit = 1 (final packet), the estimating unit 205 determines an internal TS value by the following equation (9) (step S604).

[Equation 9]
Internal TS value (cnt_eq) = TS value (cnt_eq) + Duration (cnt_eq)-sampling frequency x ptime

  If Ebit = 1 (final packet), the estimating unit 205 determines an internal TS value using the following equation (10) (step S605).

[Equation 10]
Internal TS value (cnt_eq) = {SN (cnt_eq) -SN (cnt_eq-1)} x sampling frequency x ptime + internal TS (cnt_eq-1)

  In addition, AMR-NB, G. The sampling frequency of 711 is 8 KHz, and ptime is a packetization period (eg, 20 ms).

  As described above, even when the packet related to the telephone-event is received, the TrGW 10 can estimate the clock shift between the devices and transfer the packet at the transmission rate (transmission interval) of the opposite device.

  From the above description, the industrial applicability of the present invention is clear. However, the present invention is suitably applicable to a voice RTP packet transfer device or the like that requires packet transmission with low delay and fluctuation absorption. . For example, use in a TrGW located at the boundary between the IP-POI network connecting different mobile carriers and the own carrier network is assumed.

Some or all of the above embodiments may be described as in the following supplementary notes, but are not limited thereto.
[Appendix 1]
This is the same as the communication device according to the first aspect described above.
[Appendix 2]
The estimating unit includes:
The communication device according to claim 1, wherein a slope of a regression line between the time stamp and the reception time of the received packet is calculated as the clock deviation reflection value.
[Appendix 3]
The estimating unit, when a plurality of packets are received in a burst, buckets other than the last received packet among the plurality of burst-received packets are not used for calculating the slope of the regression line, preferably. Is the communication device according to supplementary note 2.
[Appendix 4]
The buffer control unit includes:
The first reception is performed based on a time obtained by multiplying a difference between corresponding timestamps of a first reception packet and a second reception packet received immediately before the first reception packet by a slope of the regression line. The communication device according to claim 3, wherein a time at which a packet is read from the buffer is determined.
[Appendix 5]
The estimating unit includes:
5. The communication device according to claim 1, wherein the communication device calculates an adjustment value for eliminating a time difference between the opposing device and the own device that is accumulated due to a clock shift between the opposing device and the own device.
[Appendix 6]
The estimating unit includes:
The communication device according to claim 5, wherein, during a predetermined period, an average value of a period during which a received packet is stored in the buffer is calculated, and the adjustment value is calculated based on the calculated average value.
[Appendix 7]
The buffer control unit includes:
7. The communication device according to claim 5, wherein the adjustment value is reflected on a time at which the received packet is extracted from the buffer.
[Appendix 8]
The buffer control unit includes:
After reflecting the adjustment value at the time of extracting the received packet from the buffer and then receiving a predetermined number of packets, reflect the adjustment value again at the time of extracting the received packet from the buffer. The communication device as described.
[Appendix 9]
The clock used for measuring the reception time of the received packet is higher in frequency and accuracy than the clock used for the operation of the receiving unit and the transmitting unit, and preferably is any one of supplementary notes 1 to 8. Communication device.
[Appendix 10]
The estimating unit includes:
The communication device according to claim 1, wherein the clock shift reflection value is calculated from the two time stamps and the reception time of the received packet.
[Appendix 11]
This is the same as the communication method according to the second viewpoint described above.
[Supplementary Note 12]
This is as in the program according to the third aspect described above.
In addition, the form of Supplementary Note 11 and the form of Supplementary Note 12 can be expanded to the form of Supplementary Note 2 to the form of Supplementary Note 10 in the same manner as the form of Supplementary Note 1.

  The disclosures of the above cited patent documents and the like are incorporated herein by reference. Modifications and adjustments of the embodiments or examples are possible within the framework of the entire disclosure (including the claims) of the present invention and based on the basic technical concept thereof. In addition, various combinations or selections of various disclosed elements (including each element of each claim, each element of each embodiment or example, each element of each drawing, and the like) within the frame of the entire disclosure of the present invention. (Including partial deletions) are possible. That is, the present invention naturally includes various variations and modifications that can be made by those skilled in the art according to the entire disclosure including the claims and the technical idea. In particular, with respect to the numerical ranges described herein, any numerical values or small ranges included in the ranges should be construed as being specifically described even if not otherwise specified.

10 TrGW
11 CPU
12 Memory 13 NIC
20 IBCF
30-1, 30-2 Terminal 100 Communication device 101, 201 Receiver 102, 203 Buffer controller 103, 204 Transmitter 104, 205 Estimator 202 Jitter buffer 206 TSC unit 207 Connection manager 211 Enqueue unit 212 Dequeue unit 213 Dequeue Scheduled time list

Claims (10)

A receiving unit that receives a packet from the opposite device;
A buffer control unit that controls writing and reading of the received packet to and from the buffer,
Transmitting a received packet read from the buffer to a transfer destination device, a transmitting unit,
Based on the reception time of the received packet and a time stamp included in the received packet, an estimation unit that estimates a clock deviation reflection value in which a clock deviation between the opposing device and the own device is reflected,
With
The communication device, wherein the buffer control unit determines a time at which the received packet is extracted from the buffer using the clock shift reflection value. The estimating unit includes:
The communication device according to claim 1, wherein a slope of a regression line between the time stamp and the reception time of the reception packet is calculated as the clock deviation reflection value.   The estimating unit, when a plurality of packets are received in a burst, buckets other than the last received packet among the plurality of packets received in the burst are not used for calculating the slope of the regression line. Item 3. The communication device according to item 2. The buffer control unit includes:
The first reception is performed based on a time obtained by multiplying a difference between corresponding timestamps of a first reception packet and a second reception packet received immediately before the first reception packet by a slope of the regression line. The communication device according to claim 3, wherein a time at which a packet is read from the buffer is determined. The estimating unit includes:
The communication device according to claim 1, wherein the communication device calculates an adjustment value for eliminating a time difference between the opposing device and the own device that is accumulated due to a clock shift between the opposing device and the own device. The estimating unit includes:
The communication device according to claim 5, wherein, during a predetermined period, an average value of a period during which a received packet is stored in the buffer is calculated, and the adjustment value is calculated based on the calculated average value. The buffer control unit includes:
The communication device according to claim 5, wherein the adjustment value is reflected in a time at which the received packet is extracted from the buffer. The buffer control unit includes:
8. The method according to claim 7, wherein the adjustment value is reflected at a time when the received packet is extracted from the buffer, and after a predetermined number of packets are received, the adjustment value is reflected again at a time when the received packet is extracted from the buffer. Communication device. A receiving unit that receives a packet from the opposite device;
A buffer into which the received packet is written,
Transmitting a received packet read from the buffer to a transfer destination device, a transmitting unit,
In a communication device comprising:
Based on the reception time of the received packet and the time stamp included in the received packet, a step of estimating a clock deviation reflection value reflecting a clock deviation between the opposite device and the own device,
Determining a time to retrieve the received packet from the buffer using the clock shift reflection value;
Communication method including. A receiving unit that receives a packet from the opposite device;
A buffer into which the received packet is written,
Transmitting a received packet read from the buffer to a transfer destination device, a transmitting unit,
To a computer mounted on a communication device having
Based on the reception time of the received packet and the time stamp included in the received packet, a process of estimating a clock deviation reflection value reflecting a clock deviation between the opposite device and the own device,
A process of determining a time at which the received packet is extracted from the buffer using the clock shift reflection value;
A program that executes

Images