JP6834642B2 - Time correction method - Google Patents

Description

The present invention relates to a time correction method in the IEEE 1588 standard "Precision Time Protocol" (hereinafter referred to as "PTP").

In the time synchronization by "PTP", the time of each device is synchronized by distributing the time between the devices. It is possible to improve the accuracy of the measurement by performing the transmission / reception measurement in this "PTP" by hardware.

Further, by reciprocating the "PTP" packet between the devices, the time difference between the reference clock such as the grand master clock (hereinafter referred to as "GM") on the network and the clock of each device is calculated and calculated. Correct the time based on the time difference.

The time synchronization of "PTP" is used, for example, for time synchronization between robots and time synchronization of relays. Explaining Patent Document 1 as an example, "PTP" is used for time synchronization when tracking a target with a plurality of robots. Specifically, first, each master device of the plurality of robot groups is time-synchronized with "GM", and then the slave device of each robot group is time-synchronized with the master device after the time synchronization.

JP 2014-211851

As described above, at the time of time synchronization of "PTP", the device on the network executes communication for reciprocating the "GM" and "PTP" packets. The correct time is calculated based on this round-trip time difference and the time is corrected, but this is premised on the same transmission time during the round-trip.

Here, if a communication device compatible with "PTP" is used in the network, the fluctuation of the transmission time is corrected, but in the communication device not supporting PTP, "1/2" of the delayed fluctuation of the transmission time is a measurement error. It is said that. Since this error appears in the direction of time advance and the direction of time delay, the width of the error is equal to the difference in transmission time. Therefore, if the error is corrected by a PTP-compatible communication device, the time will be accurate.

However, it is not always possible to construct a network only with communication devices compatible with "PTP", and the clock is usually adjusted assuming the existence of communication devices not compatible with "PTP". Servo control (“PID” control) is often used as a method of adjusting this watch.

That is, since the control including an error is performed by adjusting the clock, it fluctuates greatly. Therefore, the "P (proportional)" of the "PID" control by the servo control is reduced to prevent a large disturbance. This is the same as averaging over time, and the disadvantage is that it takes time to adjust the clock. Further, if the distribution of the transmission time is not uniform between the outward path and the return path of the "PTP" packet, the time deviates from the correct time.

Therefore, in servo control, it is difficult to calculate the accuracy of the current clock, and even if the servo control parameters are appropriate, the error from the correct time is not set to "1/2" of the maximum fluctuation. Then, from the viewpoint of setting the correct time, simple servo control is not possible, and a problem may occur especially in a system requiring high accuracy.

The present invention has been made to solve such a conventional problem, and an object of the present invention is to improve the accuracy of time correction in a system assuming a communication device that does not support "PTP".

The present invention relates to a method of reciprocating a packet between a master-side device and a slave-side device and correcting the time of the slave-side device based on the round-trip time.

This method records the measurement error calculation step of calculating the measurement error based on the time difference between the minimum round trip time of the packet and the actual round trip time of the packet, and the recording of the adjustment amount with respect to the clock of the internal clock of the slave side device. It has an estimation accuracy calculation step of acquiring the maximum adjustment amount based on the above and calculating the estimation accuracy based on the acquired maximum adjustment amount and the elapsed time after the previous adjustment. If the measurement error calculated in the measurement error calculation step is larger than the estimation accuracy calculated in the estimation accuracy calculation step, the measured value of the round trip time is not used for the time correction of the slave side device.

As one aspect of the measurement error calculation step, either (1) calculation of the minimum round trip time based on the distance of the transmission line and the number of stages of the relay device, or (2) acquisition of the minimum value based on the measurement record of the past round trip time. Can be obtained by

As one aspect of the measurement error calculation step, the measurement error can be calculated by {[measurement error] = ([round trip time]-[minimum round trip time]) / 2}. Further, as one aspect of the estimation accuracy calculation step, the estimation accuracy can be calculated by {[estimation accuracy] = [maximum adjustment amount] × [elapsed time after the previous adjustment]}.

The present invention may further include an output step that outputs an accuracy abnormality if the estimated accuracy is worse than the preset target accuracy. Further, in the estimation accuracy calculation step, it is preferable to discard the recording of the adjustment amount after a certain period of time has elapsed. Sa

According to the present invention, it is possible to improve the accuracy of time correction in a system assuming a communication device that does not support "PTP".

FIG. 5 is a system configuration diagram of "PTP" of the IEEE1588 standard for executing the time correction method according to the embodiment of the present invention. The explanatory view which shows the calculation of the time difference at the time of the round trip of the "PTP" packet. The explanatory view which shows the adjustment amount of the clock for the same 1 second. (A) is a graph showing the round-trip time and elapsed time of the "PTP" packet of the transmission line condition example 1, and (b) is a graph showing the time difference of (a). (A) is a graph showing the round-trip time and elapsed time of the "PTP" packet of the transmission line situation example 2, and (b) is a graph showing the time difference of (a). (A) is a graph showing the round-trip time and elapsed time of the "PTP" packet of the transmission line condition example 3, and (b) is a graph showing the time difference of (a). (A) is a graph showing the round-trip time and elapsed time of the "PTP" packet when the present invention is applied to the transmission line condition example of FIG. 5, (b) is a graph showing the time difference of (a), (c). Is a graph showing the state after the passage of time in (b).

Hereinafter, the time correction method according to the embodiment of the present invention will be described. This time correction method is applied to the "PTP" time synchronization system. According to this "PTP", the device on the network can obtain a time difference from the reference clock such as "GM", and if the time difference is accurate, it can be applied to the adjustment of the internal clock in the device.

However, in reality, when a "PTP" packet reciprocates on a transmission path, it is affected by fluctuations in the transmission time, and it is difficult to obtain an accurate time difference, and the measured value often includes an error. Conventionally, this point has been dealt with by PID control as servo control.

Originally, the servo control is a control in which when an accurate control input for obtaining a target output result cannot be calculated, the difference in the output result is obtained and returned to the control input for correction. However, since the target output result in "PTP" is the time, it is only necessary to adjust the time to advance or delay it, and an accurate control input can be calculated.

Then, feedback control such as servo control becomes unnecessary, but the time difference obtained by "PTP" includes an error due to the fluctuation of the transmission time, so that it reacts to the output. Therefore, it is necessary to loosen the responsiveness by servo control. At this time, instead of adjusting the output (time), the servo control is applied to the error when the output (time) is measured.

There are various possible causes for the transmission time of the "PTP" packet to fluctuate, but when a communication device that does not support "PTP" (for example, a switching hub) is used, the received "PTP" packet is an internal buffer. Since it is stored in, there may be a delay in the time until it is sent.

For example, if 10 "1000 bytes" packets are waiting to be transmitted on the "1 Gbps" transmission line, a delay of "80 μsec" occurs. At this time, the measurement error of the time difference in "PTP" is half of that, "40 μsec" (however, it varies depending on the transmission load condition).

Since it is not preferable that the time fluctuates due to such a measurement error, the fluctuation is suppressed by servo control (“PID” control) as described above. For example, if the time difference "1/10" is applied proportionally, the effect of the instantaneous transmission delay is suppressed to "1/10" and the time fluctuation becomes small (however, to match the measured time). It takes time.)

Even with the conventional servo control, the numerical value that stands out in the measured time difference is discarded. Alternatively, processing such as averaging is performed. However, since these methods are processes that do not consider accuracy, the problem cannot be solved even if a momentary abnormality can be dealt with.

For example, the transmission delay of a switching hub with a high and stable transmission load is also stable, and if the transmission load is high only in a specific direction in the round trip of the "PTP" packet, the time difference becomes stable in a biased direction. This has a problem in that the measurement error is not taken into consideration. The time correction method deals with such a problem.

≪System configuration example≫
A configuration example of the time synchronization system will be described with reference to FIG. Reference numeral 1 in FIG. 1 indicates a master-side device (hereinafter referred to as “GM”) selected as “GM”. Further, 2a and 2b indicate two slave-side devices (hereinafter, referred to as slave devices) connected to the “GM” 1 via a relay device (for example, a switching hub, a router, etc.) (not shown). There is.

According to the time synchronization system, a "PTP" packet is delivered from the "GM" 1 to the slave devices 2a and 2b, and the slave devices 2a and 2b execute time synchronization. To explain the outline based on FIG. 2, first, the “GM” 1 transmits the “Sync message (synchronous message)” of the arrow A to the devices 2a and 2b on the slave side, and stores the time t1 of the transmission. At this time, the “GM” 1 further transmits a “Follow_up message (synchronous message)” (not shown) including the time t1 to the devices 2a and 2b on the slave side.

Next, the slave devices 2a and 2b receive both of the above messages and store the time t2 at which the "Sync message" is received. Further, the time t1 is extracted from the "Follow_up message" and the time t1 is stored. Further, the slave devices 2a and 2b transmit the “Delay_req message (delay request message)” indicated by the arrow B to the “GM” 1 and record the transmission time t3.

Then, when the "GM" 1 receives the "Delay_req message", the reception time t4 is stored. Further, the “GM” 1 transmits the “Delay_Resp message (delayed response message)” of the arrow C including the time t4 to the slave devices 2a and 2b. This is one step of the "PTP" packet.

Here, if the transmission time "T1 (bound)" and the transmission time "T2 (return)" are the same, the time difference between the slave devices 2a and 2b and "GM" 1 is "time difference = {(t2-t1)-. (T4-t3)} / 2 ”is calculated. The time difference of the slave devices 2a and 2b is corrected using the calculated time difference.

However, according to the time correction method, it is determined whether or not the measured value of the round-trip time of the "PTP" packet is used for the time correction of the slave devices 2a and 2b (adjustment of the clocks 4b and 4c). Hereinafter, the specific processing contents of the correction method will be described.

<< First Embodiment >>
The time correction method of the first embodiment includes a measurement error calculation step of calculating [this measurement error] from a change in the round-trip transmission time of the “PTP” packet, a clock (time) accuracy, and a state of the “PTP” packet. It has an estimation accuracy calculation step for calculating [estimation accuracy] from the above, and a determination step for determining whether or not the [current measurement error] is larger than the [estimation accuracy].

(1) Measurement error calculation step The measurement error calculation step calculates [this measurement error] based on the time difference between the minimum round trip time of the "PTP" packet and the current transmission time of the actual "PTP" packet round trip. ..

To explain in detail, it can be said that the time when there is no transmission load and the minimum transmission time is a highly accurate measurement of the round-trip time of the "PTP" packet. On the other hand, it can be said that the longer it takes for the "PTP" packet to make a round trip, the longer the time difference includes an error. This [measurement error this time] is calculated by Equation 1.

Equation 1: [Measurement error this time] = ([Round trip time this time]-[Minimum round trip time]) / 2
The minimum round trip time in this equation 1 can be obtained by either (A) a calculation method based on the distance of the transmission line and the number of stages of the relay device, or (B) an acquisition method based on the measurement record of the past round trip time. ..

First, a calculation method based on the distance of the transmission line and the number of stages of the relay device will be described. Here, it is assumed that a switching hub is used as a relay device, and if the length of the transmission line and the number of stages of the switching hub are fixed, the minimum round trip time can be calculated by Equation 2.

Equation 2: [Minimum round-trip time] = {"distance" ÷ "speed of light" x 0.7)} + {("PTP" packet bit number "÷" transmission speed ") + α} x" HUB number "
(Α: Delay time in HUB packet transfer)
Generally, a switching hub receives a packet size and then transmits the packet size, so that it takes time for the packet size. Further, although the "PTP" packet is transferred at "100%" of the transmission speed, it may take time for a single transfer, so the adjustment is made with "α" in Equation 2. In the section to which the "PTP" compatible switching hub is applied, the delay is corrected, so only the time according to the "distance" is added. In addition, "speed of light" is the speed in vacuum, but copper wire and optical fiber are multiplied by "0.7" times the speed of "speed of light".

Next, an acquisition method based on the measurement record of the past round trip time will be described. That is, the storage device of the slave devices 2a and 2b records the past round-trip time (“T1” + “T2” in FIG. 2) of the “PTP” packet.

The past round-trip time during this recording is larger than the [minimum round-trip time] even if the clocks 4b and 4c are not time-synchronized with the clock 4a. Here, half of the minimum measured value (round trip time) in the specific period in the past, that is, the specific period in the recording is defined as the required time for one way, and the minimum measured value is defined as the minimum round trip time.

However, the acquisition method based on the measurement record of the past round trip time is not reliable, and it is not always the correct value in the case of a line where the transmission load is always high, so the specification is such that the minimum round trip time can be set manually. Is preferable.

(2) Estimated accuracy calculation step In the estimated accuracy calculation step, the maximum adjustment amount is acquired based on the recording of the adjustment amount for the clocks of the internal clocks 4b and 4c of the slave devices 2a and 2b, and the acquired maximum adjustment amount is adjusted to the acquired maximum adjustment amount after the previous adjustment. The electronic circuits of the internal clocks 4a to 4c for calculating the estimation accuracy by multiplying the elapsed time have a counter for adjusting the length of "1 second". Here, according to the "PTP" time correction, the internal clocks 4b and 4c of the slave devices 2a and 2b adjust the counter to increase or decrease the advancing speed of "1 second" to set the time inside the "GM" 1. It is set to the clock 4a.

Explaining with reference to FIG. 3, X1 indicates "1 second" of the internal clock 4a of "GM" 1, and X2 indicates "1 second" of the internal clocks 4b and 4c of the slave devices 2a and 2b. Here, since the frequencies of the crystals 5b and 5c are higher than those of the crystal 5a, "X2 <X1" is established, and the internal clocks 4b and 4c are 1 second shorter than the internal clock 4a.

At this time, in the time synchronization of the arrow P, the count value set to "1 second" of the internal clocks 4b and 4c is increased in order to match the 1 second of the internal clock 4a. X3 in FIG. 2 indicates the time (adjustment amount) of the portion increased by the time synchronization. The past adjustment amount such as X3 shall be recorded in the storage devices of the slave devices 2a and 2b.

The latest [maximum adjustment amount] is acquired based on the fluctuation of the adjustment amount in the recorded data recorded here, and the acquired [maximum adjustment amount] is used for estimation accuracy, that is, estimation of clock accuracy. For example, adjust "1 second" to "300 nsec" shorter in the measurement 30 seconds ago, adjust "1 second" to "100 nsec" longer in the measurement 20 seconds ago, and then adjust "1 second" to "1 second" in the measurement 10 seconds ago. It is assumed that the adjustment is shortened by "200 nsec".

In this case, the fluctuation of the adjustment amount is a change of "400 nsec / 10 seconds" from 30 seconds to 20 seconds before, and a change of "300 nsec / 10 seconds" from 20 seconds to 10 seconds before. The maximum adjustment amount] is "40 nsec / sec", and the deviation is "40 nsec / sec" every second. At this time, the [estimation accuracy] is calculated by Equation 3.

Equation 3: [Estimation accuracy] = ([Maximum adjustment amount] x [Elapsed time after the previous PTP adjustment])
Here, in Equation 3, the [maximum adjustment amount] cannot be calculated before the "PTP" time correction, so the worst value based on the data sheet of the crystals 5b, 5c or the like may be adopted when the slave devices 2a, 2b are activated. .. Further, if the high-precision measurement can be performed twice, it can be interpreted that the speed at which the clock advances can be corrected with the accuracy at that time.

When this is repeated, the amount of adjustment to the crystals 5b and 5b is required to change. The maximum fluctuation can be set as the [maximum adjustment amount] as the slope of the estimation error. Here, since the accuracy of quartz generally depends on the environment such as temperature and voltage, calculation based on old data is not preferable.

Therefore, it is desirable to set a timeout at the time of measurement and discard the old data. According to Equation 3, the [estimation accuracy] at the moment when the time is corrected by the “PPT” time correction corresponds to the [current measurement error] at that time.

(3) Judgment step In the judgment step, [estimation accuracy] and [current measurement error] are compared, and it is determined whether or not [current measurement error] is larger than [estimation accuracy].

As a result of this determination, if [this measurement error] is larger than [estimation accuracy], the current measurement value (round trip time) is not used for "PTP" time correction. On the other hand, if [this measurement error] is not larger than [estimation accuracy], the current measurement value (round trip time) is used for "PTP" time correction.

For example, suppose that when a "PTP" packet is received while a "1 Gbps 1518 byte" packet is being transmitted by a switching hub, a delay of up to "12.1 μsec" occurs in the "PTP" packet.

At this time, the [current measurement error] is calculated as {[current measurement error] = ([current round trip time]-[minimum round trip time]) / 2 = 6.0 μsec}.

Further, the adjustment amount of the "PTP" correction up to this state is {[maximum adjustment amount] = 40 nsec / sec}. Then, when 1 second has passed from the previous time adjustment, the relationship of {[estimation accuracy] <[current measurement error]} is established, so the current measurement value is not used for "PTP" time correction.

That is, if it is used for "PTP" time correction at this point, the measurement error is large and only disturbs, and when 150 seconds have passed thereafter, the [measurement accuracy] exceeds "6.0 μsec". If this time ([current measurement error] = 6.0 μsec) is established, the measured value can be used for “PTP” time correction.

In this way, according to the time correction method, when adjusting the internal clocks 4b and 4c of the slave devices 2a and 2b, the [estimation accuracy] and the [current measurement error] are compared and used for the "PTP" time correction. Whether it is possible or not is determined (the same applies when applied to servo control). As a result, the following effects can be obtained by the time correction.

(A) If there is no delay fluctuation in the transmission line, the time can be synchronized at high speed. That is, in the servo control method, it takes time for the time to stabilize even if the proportion or integral is adjusted. On the other hand, according to the time correction method, since the measurement error is grasped and the "PTP" time correction is performed, it is difficult to receive disturbance even with the same servo control, and the time until it stabilizes can be shortened.

(B) As a result of comparison between [estimation accuracy] and [current measurement error], in other words, comparison between the current time error and the measurement error, the time is corrected only when the measurement error is small. This makes it less susceptible to disturbance even if the measured value has a large error. In addition, the time fluctuation is stable, and the proportional item of servo control can be set high, so that the responsiveness can be improved.

(C) Parameter setting becomes easy in servo control. That is, in the usual method, it was necessary to set the parameters according to the characteristics of the line. At this time, when there is a lot of disturbance due to delay fluctuation, it is necessary to reduce the proportion, but the responsiveness is poor.

On the other hand, if the proportion is increased, the responsiveness is good, but the fluctuation becomes large. Integral is to eliminate the offset that cannot be adjusted to the target value, but if it is applied strongly, the output will fluctuate.

In this way, it is not easy to determine the parameters according to the characteristics of the transmission line. Regarding this point, according to the time correction method, a measured value having a large error is excluded by determining the comparison result between the [estimation accuracy] and the [current measurement error]. As a result, it is less susceptible to disturbance than before, and parameters can be easily determined.

(D) According to the [estimation accuracy] calculation, the difference between the reference such as the internal clock 4a of "GM" 1 and the internal clock of the slave devices 2a and 2b can be estimated. That is, the servo control merely controls in the direction of the time that seems to be correct based on the measured time difference.

Therefore, the error in the measurement of "PTP" cannot be recognized, and the difference in the current time cannot be grasped. On the other hand, according to the time correction method, the fluctuation amount of the clock is extracted based on the recording of the round trip time of the "PTP" packet, and the time difference (maximum adjustment amount) is calculated. In this respect, it can contribute to the improvement of time accuracy, and it is possible to appropriately cope with the system in which the accuracy is limited.

<< Second Embodiment >>
In the time correction method of the second embodiment, an output step is added to output an accuracy abnormality if the [estimated accuracy] is worse than the [target accuracy].

That is, in a system in which the accuracy of time synchronization is a problem, it is necessary to output an abnormality and take measures against it when the accuracy cannot be maintained. Therefore, in the second embodiment, the [target accuracy] is determined, and when the [estimation accuracy] is inferior to that, a signal with abnormal accuracy is output. This [target accuracy] may be preset by the user as the accuracy required by the system.

Then, the judgment of the accuracy abnormality depends on the [target accuracy], and even if there is a change in the transmission time, if the [target accuracy] is set larger than the fluctuation and is within the set range, it is not judged as an abnormality. No accuracy error is output.

On the other hand, when the [target accuracy] becomes high, the fluctuation of the round-trip time of the “PTP” packet becomes unacceptable, and the time during which the [estimation accuracy] can maintain the [target accuracy] tends to be short. If an accuracy error is output here, the error processing according to the system specifications is executed.

Therefore, according to the time correction method of the second embodiment, the [estimation accuracy] is calculated from the adjustment amount of the internal clocks 4b and 4c of the slave devices 2a and 2b, which is worse than the [target accuracy] determined by the user. If so, the process as an accuracy error is executed.

≪Effect of fluctuation of transmission delay≫
Hereinafter, “PTP” time correction for fluctuations in transmission delay will be described with reference to FIGS. 4 to 7. Here, the vertical axis of FIGS. 4 (a), 5 (a), 6 (a), and 7 (a) shows the round-trip time of the "PTP" packet (T1 + T2 of FIG. 2), and the horizontal axis shows the elapsed time. Is shown. The arrow D in each figure indicates the round-trip time (transmission line time: time at a transmission speed of 70% of the speed of light) required for the transmission line.

4 (b), 5 (b), 6 (b), 7 (b) and (c) show the reference clock (such as the internal clock 4a of "GM" 1) and the internal clock of the slave side device. The difference between the two, that is, the fluctuation of the time error is shown.

(1) In the transmission line condition example 1 of FIG. 4A, the round-trip time of the “PTP” packet is substantially equal to the transmission line time, and the accuracy is high. As a result, as shown by the arrow F in FIG. 4B, the time difference is close to "0" almost every time, and the accuracy is high.

However, the round trip time is long only once indicated by the arrow E, and the measurement error is large. Performing "PTP" time correction using this measurement causes large fluctuations. In this case, since the measured value is discarded by the conventional method, the fluctuation of the transmission line delay is small, but the error increases with the passage of time.

(2) In the transmission line condition example 2 of FIG. 5A, the transmission time increases after the round trip of the “PTP” packet of arrow E1, and the state of large [current measurement error] continues. At this time, the "PTP" time correction cannot be performed by the conventional method, and the time error increases as shown by the arrow F1 in FIG. 5 (b).

(3) For example, an increase in transmission time due to a communication device that does not support PTP occurs due to a transmission load. Therefore, as shown in FIGS. 4 and 5, the transmission time may increase, and as shown in FIG. 6, in rare cases, high-precision measurement is possible. In the case of FIG. 6, according to the conventional method, "PTP" time correction can be performed when the error is small.

(4) If the time can be corrected only when the transmission time is close to the [minimum round-trip time] as in the conventional method, the state shown in FIG. 5 is obtained. Therefore, in the time correction method, the time is corrected when the [current measurement error] becomes smaller than the [estimation accuracy]. Here, since the [estimation accuracy] increases as time elapses according to the equation 3, the time correction is executed when the value of the [measured value this time] is exceeded. This point is shown in FIG.

As shown in FIG. 7A, the transmission time has increased after the round trip of the “PTP” packet of arrow E3 as in FIG. 5, and the state of large [current measurement error] continues. At this time, {[measurement error (A / 2)] << [estimation accuracy (B)]} is established during the round trip of the “PTP” packet shown by the arrow E4 in FIG. 7 (b), and the time correction is performed. ..

Since the time has been corrected here, the [estimation accuracy (B)] is [current measurement accuracy (A / 2)], and the [estimation accuracy] is [current measurement error] as shown in FIG. 7 (c). (A / 2)] increases with the passage of time (“β” in the figure indicates the increase).

As described above, according to the time correction method, accuracy and stability can be obtained by determining whether or not to perform time correction based on the magnitude of the estimated accuracy of the measurement error in the "PTP" time correction. .. The accuracy of the time is determined by the situation of the delay time in the transmission line.

1 ... Master device (master device)
2a, 2b ... Slave side device (slave device)
4a-4c ... Internal clock 5a-5c ... Crystal

Claims (6)

It is a method of reciprocating a packet between a device on the master side and a device on the slave side, and correcting the time of the device on the slave side based on the time of the reciprocation.
A measurement error calculation step of calculating a measurement error based on the time difference between the minimum round-trip time of the packet and the actual round-trip time of the packet, and
An estimation accuracy calculation step of acquiring the maximum adjustment amount based on the recording of the adjustment amount with respect to the clock of the internal clock of the device on the slave side and calculating the estimation accuracy based on the acquired maximum adjustment amount and the elapsed time after the previous adjustment. Have,
If the measurement error calculated in the measurement error calculation step is larger than the estimation accuracy calculated in the estimation accuracy calculation step,
A time correction method characterized in that the measured value of the round trip time is not used for time correction of the slave side device. The minimum round trip time in the measurement error calculation step is
Calculation based on the distance of the transmission line and the number of stages of the relay device,
Acquisition of the minimum value based on the measurement record of the past round trip time,
The time correction method according to claim 1, wherein the time correction method is obtained by any of the above. In the measurement error calculation step, the measurement error is
[Measurement error] = ([Round trip time]-[Minimum round trip time]) / 2
The time correction method according to claim 1 or 2, wherein the time is calculated according to the above. Estimation accuracy In the calculation step, the estimation accuracy is
[Estimation accuracy] = [Maximum adjustment amount] x [Elapsed time after the previous adjustment]
The time correction method according to any one of claims 1 to 3, wherein the time is calculated according to the above. The time correction method according to any one of claims 1 to 4, further comprising an output step for outputting an accuracy abnormality if the estimated accuracy is worse than a preset target accuracy. In the estimation accuracy calculation step
The time correction method according to any one of claims 1 to 5, wherein the record of the adjustment amount is discarded after a lapse of a certain period of time.

Images