JP7547906B2 - Time synchronization method, time synchronization method, communication device - Google Patents

Description

The present invention relates to a technology for achieving time synchronization between TSN domains using time synchronization processing based on PTP (Precision Time Protocol).

(1) Point-to-point play
Among PTP-compatible relay devices (communication devices), there exists a type that can perform time synchronization in one step of "EtoE (between master and slave)". With this type of relay device, the transmission time Da (Meanpath) is calculated by formula (1) through the time synchronization processing procedure shown in FIG.
Equation (1): Transmission time Da = "(T4 - T3) + (T2 - T1)" / 2
In addition, time correction between the master and slave is performed based on the offset value calculated by Equation 2.
Formula (2): Offset = "(T2 - T1) - (T4 - T3)" / 2
It should be noted that Patent Document 1 is known as a technology for improving the accuracy of time synchronization by suppressing the influence of fluctuations in transmission time in PTP time synchronization.

(2) TSN
Time-Sensitive Networking (TSN) is a network technology that extends standard Ethernet (registered trademark) to enable industrial networks and IT networks to interoperate with each other, and is made up of multiple standards.

For example, according to the time synchronization of the "IEEE 802.1AS" standard (gPPT), the transmission time D is calculated by the formula (3) after going through the TSN time synchronization processing procedure shown in FIG. 2, that is, the "PtoP (adjacent devices)" TWO step.
Equation (3): Transmission time D = "(T4-T3) + (T2-T1)" / 2 = "(T4-T1) - (T3-T2)" / 2
3 shows the difference between the ONE step and the TWO step by taking the "SYNC" message as an example. That is, in the ONE step, the timestamp information of "T1" is set in the packet of the "SYNC" message and transmitted.

On the other hand, in the TWO step, "T1" is sent by setting it to a "Follow_UP" message packet rather than a "SYNC" message packet (the same goes for the "Pdelay_Resp_Follow_UP" message).

"gPTP" uses a TWO step method to calculate the elapsed time (t1) from "T1" using equation (4) that takes into account the delay in reception time (t3-T2) based on Figure 3, and performs time correction between the master and slave from the time difference between "t1" and "t3".
Equation (4): Elapsed time (t1) = T1 + D + (T3 - T4)

Patent Publication 2019-053633

TSN is a new technology, and according to the "IEEE802.1AS" standard, relay devices (switching hubs) that do not support TSN "gPTP" time synchronization cannot be used.

Therefore, when adding a relay device to an existing network, the area in which TSN is used is limited. Even when building a new network, it is difficult to make all relay devices TSN-compatible, so use is similarly limited.

On the other hand, there are many problems that can be solved by PTP time synchronization, such as when you want to unify management between distant TSN areas (TSN networks) or when you want to process at the same timing. In such cases, it is possible to consider preparing dedicated lines to directly connect TSN networks, but it is difficult to immediately address this issue using only current relay devices equipped with TSN.

The present invention was made to solve these conventional problems, and aims to make it possible to synchronize time between TSN networks through areas that do not support PTP (NON-PTP networks).

(1) One aspect of the present invention is a method for performing time synchronization between communication devices arranged in each TSN (Time-Sensitive Networking) area on a network via a PTP (Precision Time Protocol) non-compatible area,
a PTP-compatible communication device is connected to a boundary between each TSN area and the PTP-incompatible area;
A communication device connected to the boundary
When a TSN time synchronization packet is received from a communication device adjacent to the TSN area side, the TSN time synchronization packet is converted into a PTP packet, and the converted PTP packet is transmitted to a communication device adjacent to the TSN non-compatible area side,
When the converted PTP packet is received from a communication device adjacent to the TSN non-compatible area side, the PTP packet is converted into a TSN packet, and the converted TSN packet is transmitted to a communication device adjacent to the TSN compatible area side.

(2) Another aspect of the present invention is a method executed by a communication device connected to a boundary between a time-sensitive networking (TSN) area on a network and a precision time protocol (PTP) non-compatible area, the method comprising:
a step of converting a TSN time synchronization packet into a PTP packet when receiving the TSN time synchronization packet from a communication device adjacent to the TSN area side, and transmitting the converted PTP packet to the communication device adjacent to the TSN non-compatible area side;
The method is characterized by having a step of converting the PTP packet into a TSN packet when the converted PTP packet is received from a communication device adjacent to the TSN non-compatible area side, and transmitting the converted TSN packet to a communication device adjacent to the TSN compatible area side.

(3) Yet another aspect of the present invention is a communication device disposed on a boundary between a time-sensitive networking (TSN) area and a precision time protocol (PTP) non-compatible area on a network, the communication device comprising:
When a TSN time synchronization packet is received from a communication device adjacent to the TSN area side, the TSN time synchronization packet is converted into a PTP packet, and the converted PTP packet is transmitted to a communication device adjacent to the TSN non-compatible area side,
When the converted PTP packet is received from a communication device adjacent to the TSN non-compatible area side, the PTP packet is converted into a TSN packet, and the converted TSN packet is transmitted to a communication device adjacent to the TSN compatible area side.

According to the present invention, time synchronization between TSN networks becomes possible via a "NON-PTP network."

FIG. 1 is a schematic diagram of the PTP time synchronization processing procedure (ONE step). FIG. 2 is a schematic diagram of a TSN time synchronization process; FIG. 2 is a schematic diagram of a processing procedure showing an example of the TWO step. A network configuration diagram in which a NON-PTP network and a TSN network are mixed. FIG. 1 is a network configuration diagram of a time synchronization system according to an embodiment of the present invention. 13 is a flowchart showing a main process of TSN packet transfer in the first embodiment; 13 is a flowchart showing a TSN packet forwarding determination process. 13 is a flowchart showing a TSN packet forwarding process. 11 is a flowchart showing a process for a "SYNC message." 11 is a flowchart showing the processing for a "Follow_UP message", an "ANNOUNCE message", and an "ELSE message;" 11 is a flowchart showing a PTP packet transmission process. 13 is a flowchart showing a TSN packet conversion process. 1 is a schematic diagram showing conversion between TSN packets and PTP packets.

The following describes a time synchronization method according to an embodiment of the present invention. This time synchronization method is intended for the network configuration shown in Figure 4, for example.

Here, there is a "NON-PTP" network C that is made up of switching hubs (hereafter referred to as hubs) that do not have a PTP time synchronization function and serves as a trunk line connecting factories, etc. TSN networks A and B, made up of TSN hubs, are adjacent to this network C, and the TSN networks A and B are independent areas.

The details will be explained based on Figure 5. 1-1 to 1-4 in Figure 5 indicate normal hubs 1 that do not have a PTP time synchronization function. These normal hubs 1-1 to 1-4 are placed within the "NON-PTP" network C.

3-1 to 3-3 indicate TSN hubs 3 equipped with "IEEE802.1AS (gPTP)" time synchronization and "IEEE802.1Qbv (traffic scheduling)." These TSN hubs 3-1 to 3-3 are located in TSN networks A and B, respectively. Here, TSN hub 3-1 is the master, and TSN hubs 3-2 and 3-3 are the slaves.

2-1 and 2-2 indicate PTP hub 2 located on the boundary between "NON-PTP" network C and TSN networks A and B. That is, PTP hub 2-1 is located between networks A and C, and is connected to normal hub 1-2 and TSN hub 3-2. On the other hand, PTP hub 2-2 is located between networks B and C, and is connected to normal hub 1-3 and TSN hub 3-3.

Here, PTP hub 2 is configured with processing settings (see Patent Document 1) that differ from normal PTP processing settings in order to maintain the accuracy of PTP time synchronization even when passing through normal hub 1. In addition, between PTP hub 2 and TSN hub 3 (for example, between PTP hub 2-1 and TSN hub 3-2, between PTP hub 2-2 and TSN hub 3-3, etc.), TSN "P to P" processing is possible by specifying a TSN port.

In this way, by connecting a PTP hub 2 to the boundary between the "NON-PTP" network C and the TSN networks A and B, time synchronization between the TSN networks A and B is possible via the "NON-PTP" network C.

For example, assume that TSN hub 3-1 (master) located in TSN network A sends TSN time synchronization packets (gPTP: SYNC message, Follow_UP message, ANNOUNCE message).

In this case, the TSN time synchronization packet is converted to a PTP packet by PTP hub 2-1. The PTP packet converted here is then reconverted to a TSN time synchronization packet (gPTP) by PTP hub 2-2 via normal hubs 1-2 and 1-3, and is then sent to TSN hub 3-3 (slave) on TSN network B.

This enables time synchronization between TSN hubs 3-1 and 3-3 in TSN networks A and B. By correcting the transit time between PTP hubs 2-1 and 2-2, accurate time synchronization is possible between TSN hubs 3-1 and 3-3.

The specific processing contents of the PTP hub 2 are explained below. Note that TSN time synchronization (gPTP) time synchronization packets are at the L2 layer and differ from PTP in that only the "PtoP" transparency function can be used.

(1) TSN Packet Transfer Main Processing The TSN packet transfer main processing, that is, the overall processing content of the transfer processing of the PTP hub 2 arranged on the boundary between the "NON-PTP" network C and the TSN networks A and B, will be described with reference to FIG.

S01: To start processing, data required for initial setup is input from outside into the PTP hub 2 in advance. Specifically, the initial setup values for the number of TSN domains D1, TSN forwarding flag D2, TSN port D3, TSN destination MAC address D4, and TSN source MAC address D5 are input. The data input here is written as the initial setup values into the initial setup file F1 inside the PTP hub 2 and each is saved.

S02: Check whether initialization is required when the PTP hub 2 is started (during startup). If it is required, proceed to S03, and if not, proceed to S08.

S03, S04: Initialize each setting value with the initial setting values D1 to D5 saved in S01 (S03). Set these initialized setting values as internal data of the PTP hub 2 (S04).

S05: Check the TSN transfer flag D2. If the TSN transfer flag is valid, proceed to S06. If the TSN transfer flag is invalid, proceed to S07.

S06, S07: In S06, the device is set to a state waiting to receive a TSN packet from the TSN port. In S07, the device is set to a state waiting to receive a PTP packet from the PTP port.

S08, S09: In S08, a packet is received in the packet reception process, and the received packet P is stored. In S09, the packet forwarding decision process in FIG. 7 is executed, and the process ends after execution.

(2) TSN Packet Forwarding Determination Processing The packet forwarding determination processing in S09 will be described in detail with reference to FIG.

S11: First, when processing starts, it is checked whether the received packet P in S08 is a TSN packet (gPTP). If it is a TSN packet, the process proceeds to S12, and if it is not a TSN packet, the process proceeds to S13.

S12: Finish by going through the TSN transfer process in Figure 8.

S13: Check whether the received packet P is a PTP packet. If it is a PTP packet, proceed to S14, but if it is not a PTP packet, end the process.

S14: As shown by the arrow X1 in FIG. 13, the "Clock Identity" is obtained from the PTP information of the received packet P. Then, the source MAC address D6 is generated based on the obtained "Clock Identity" information.

S15: Check whether the source MAC address D6 generated in S14 is the same as the TSN source MAC address D5. If they are the same, proceed to S16, but if they are not the same, proceed to S17, where the conventional PTP process is executed, and the process ends.

S16: Execute the TSN packet conversion process shown in Figure 12, and then end the process.

(3) TSN packet forwarding process The TSN packet forwarding process in S12 will be described in detail with reference to Fig. 8. This TSN packet forwarding process is executed by the PTP hub 2-1 that receives the TSN packet in the network configuration of Fig. 5, for example. Here, according to S11, the received packet P becomes a TSN packet (gPTP).

S21: When processing begins, the TSN packet forwarding enable/disable status is confirmed by referencing the TSN forwarding enable/disable flag D2. If the result of the confirmation is that the TSN packet forwarding is enabled, the process proceeds to S22, whereas if the result is that the TSN packet forwarding is disabled, the process ends.

S22, S23: The source MAC address D8 is obtained from the received packet P (S22). It is then confirmed whether the obtained source MAC address D8 is the same as the TSN source MAC address D5 (S23). If the confirmation result indicates that they are the same, the process proceeds to S24, whereas if they are not the same, the process ends.

S24, S25: Obtain "Message", i.e., type information D7, from the "gPTP" information of the received packet P (see FIG. 13) (S24), and confirm the processing type based on the type information D7 (S25).

If the confirmation reveals that the processing type is a "SYNC" message, proceed to S26. On the other hand, if the message is a "Follow_UP" message, an "ANNOUNCE" message, or an "ELSE" (other) message, proceed to S27.

S26: Processing for "SYNC" is executed.
S27: Processing for "Follow_UP", "ANNOUNCE" and "ELSE" is executed. After the processing of S26 and S27 is executed, the TSN packet forwarding processing is terminated.

(4) "SYNC" Processing The "SYNC" processing of S26 will be described in detail with reference to Fig. 9. Here, processing is shown for the case where the received packet P is a "SYNC" message.

S31: When processing begins, the timestamp of the time when the received packet P arrives at the PTP hub 2 is collected. The collected arrival time (reception time) is saved as "Ta".

S32: Execute the PTP packet transmission process shown in Figure 11.

S33: After the PTP packet transmission process of S32, the timestamp of the time when the PTP packet (received packet P) departs from the PTP hub 2 is collected. The collected departure time (transmission time) is saved as "Tb".

S34, S35: Calculate the residence time of the received packet P in PTP2 (residence time "Tc" = departure time "Tb" - arrival time "Ta") (S34). Check whether the residence time "Tc" calculated here is normal (not negative, etc.) (S35).

If the check result is normal, proceed to S36, but if not, end the process.

S36: Set the waiting flag for the packet to be processed. In this case, since the received packet P is a "SYNC" message, the target packet is the following "Follow_UP" message. Set the waiting flag D9 for this "Follow_UP" message and end the process.

(5) Processing for "Follow_UP", "ANNOUNCE", and "ELSE" Details of the processing shown in S27 will be explained with reference to Fig. 10. Here, details of the processing when the received packet P is a message other than a "SYNC" message will be shown.

S41: When processing begins, it is confirmed whether the received packet P is an "ANNOUNCE" message or not based on the processing type of the received packet P acquired in S25. If the confirmation result indicates that the received packet P is an "ANNOUNCE" message, the PTP packet transmission process of S45 is executed and this process ends. On the other hand, if the received packet P is not an "ANNOUNCE" message, the process proceeds to S42.

S42: Based on the processing type of the received packet P acquired in S25, it is confirmed whether the received packet P is a "Follow_UP" message. If the result of the confirmation is that the received packet P is not a "Follow_UP" message, the processing ends (ELSE processing). On the other hand, if the received packet P is a "Follow_UP" message, the processing proceeds to S43.

S43: Check the "Follow_UP" waiting flag to see if the "Follow_UP" message is waiting. If the check shows that the message is waiting, proceed to S44, but if the message is not waiting, end the process.

S44: Add the residence time of the "SYNC" message calculated in S34 to the "Correction" of the "PTP" information shown in FIG. 13 and proceed to S45.

S45: Clear the waiting flag of the target packet for processing. In other words, since the received packet P is the target packet's "Follow_UP" message, clear the "Follow_UP" waiting flag D9 and proceed to S46.

S46: Execute the PTP packet transmission process and end the process after execution.

(6) PTP Packet Transmission Processing The PTP packet transmission processing in S46 will be described in detail with reference to Fig. 11. Here, the processing when the received packet P is transferred to the normal hub 1 will be mainly shown.

S51: When processing begins, the MAC address of the PTP hub 2 is set as the source MAC address of the received packet P (TSN packet) (see Figure 13).

S52: Set the multicast address of the PTP MAC as the destination MAC address of the received packet P (see Figure 13).

S53: The following items are added to the received packet P:
(A) Add the IP address of the PTP packet as the source IP address. (B) Add the multicast IP address for PTP hub 2 as the destination IP address of the IP address. (C) Add the port number for PTP as the destination port and source port of UDP. S54: As shown in FIG. 13, add the number of TSN domains to the “SubdomainNumber” of the “gPTP” information of the received packet P.

S55: The received packet P is sent as a PTP packet from the PTP port of the PTP hub 2. For example, in the network configuration of FIG. 5, the PTP hub 2-1 sends the received packet P after the processing of S51 to S54 to the normal hub 1-2. This transmission ends the process.

(7) PTP Packet Conversion Processing The details of the TSN packet conversion processing in S16 will be described with reference to Fig. 12. For example, the PTP hub 2-2, which received a received packet P via normal hubs 1-2 and 1-3, executes the PTP packet conversion processing in S16 when transferring the received packet P to the TSN hub 3-3. Here, according to S13 to S15 in Fig. 7, the received packet P becomes a PTP packet.

S61: When processing begins, check whether a TSN port has been set. If the check shows that the port has been set, proceed to S62; if not, end processing.

S62: Delete the UDP/IP information (UDP source port in Figure 13, etc.) from the received packet P.

S63: As shown by the arrow X1 in FIG. 13, the "Clock Identity" is obtained from the PTP information of the received packet P. Then, the source MAC address D6 is generated based on the obtained "Clock Identity" information. This conversion method is the same as S14.

S64, S65: The MAC address D11 converted in S63 is set as the source MAC address of the received packet P to be forwarded (S64). Also, the multicast MAC address for TSN is set as the destination MAC address (TSN) (S65).

S66: Subtract the TSN domain number D1 from the "SubdomainNumber" in the PTP information of the received packet P.

S67: Check whether the processing type according to the "MessageId" in the PTP information of the received packet is a "Follow_UP" message. If the result of the check is a "Follow_UP" message, proceed to S68, and if it is not a "Follow_UP" message, proceed to S70.

S68, S69: The transmission delay "Meanpath" D10 calculated using normal PTP time synchronization is obtained (S68). The "Meanpath" D10 obtained here is added to the "correction" of the received packet P.

S70: The received packet P is sent as a TSN packet from the TSN port and the process ends.

(8) Effects and Benefits According to the time synchronization method, the following effects can be obtained.

(A) Achieving Time Synchronization As described above, the TSN time synchronization packet (gPTP) sent by the TSN hub 3-1 of the TSN network A reaches the PTP hub 2-1 via the TSN hub 3-2.

At this time, the received packet P at PTP hub 2-1 is converted into a PTP packet (S12) and then sent to normal hub 1-2 (S46). After that, the received packet P reaches PTP hub 2-2 via normal hubs 1-2 and 1-3.

The PTP hub 2-2 that receives the received packet P converts the received packet P to a TSN packet (gPTP) through a PTP packet conversion process (S16) and forwards it to the TSN hub 3-3 in the TSN network B.

This allows the slave (TSN hub 3-3) to receive the TSN time synchronization packet (gPTP) sent by the master (TSN hub 3-1) via the "NON-PTP" network C. Note that transmission from the slave (TSN hub 3-3) to the master (TSN hub 3-1) can also be done in a similar manner.

As a result, it is possible to achieve PTP time synchronization between TSN hubs 3-1 and 3-3 of TSN networks A and B, which exist independently and at a distance, via a "NON-PTP" network C without laying a dedicated line.

In this case, no changes are required on the TSN (gPTP) networks A and B side, and it is sufficient to simply connect the PTP hub 2 as the boundary between the networks A and B and the network C. In this respect, it is possible to reduce capital investment required for time synchronization between the TSN networks A and B.

(B) Prevention of deterioration of time synchronization accuracy According to the PTP hub 2, the residence time of the TSN packet of the "SYNC" message is calculated (S34). The residence time calculated here is added to the "Correction" of the "Follow_UP" message (the master transmission time of the "SYNC" message) (S44). The transmission time between the PTP hubs 2-1 and 2-2 is also added in the same way (S68).

Therefore, the retention time can be taken into account when correcting the time using TSN (gPTP), which makes it possible to prevent a decrease in the accuracy of time synchronization. In addition, since the source MAC address and destination MAC address are checked (S15, S23, etc.) within the PTP hub 2, it is also possible to prevent the erroneous transmission of unnecessary packets.

1 (1-1 to 1-4)...Normal hub 2 (2-1, 2-2)...PTP hub 3 (3-1 to 3-3)...TSN hub

Claims (6)

Time synchronization between TSN-compatible communication devices installed in each TSN (Time-Sensitive Networking) network .
This is a method that is performed via a "NON-PTP (Precision Time Protocol)" network composed of communication devices that do not have a PTP time synchronization function ,
A communication device having a PTP time synchronization function is connected to the boundary between each of the TSN networks and the "NON-PTP" network ,
Each communication device connected to the boundary
When a TSN time synchronization packet is received from a communication device of the TSN network , the TSN time synchronization packet is converted into a PTP packet, and the converted PTP packet is transferred to the communication device of the "NON-PTP"network;
A time synchronization method characterized by: when the PTP packet is received from a communication device of the "NON-PTP" network, the PTP packet is reconverted into a TSN time synchronization packet, and the reconverted TSN time synchronization packet is transferred to the communication device of the TSN network . The boundary communication device that receives a TSN time synchronization packet from the communication device of the TSN network,
The MAC address of the device itself is set as the source MAC address in the received TSN time synchronization packet, the multicast address of the PTP MAC is set as the destination MAC address, the UDP (User Datagram Protocol) information and IP information of the PTP packet are added, and the number of TSN domains is added to the domain number of the "gPTP" information before forwarding the packet.
The boundary communication device that receives the PTP packet from the communication device of the "NON-PTP" network,
The UDP information and IP information are deleted from the received PTP packet, a MAC address is generated based on the PTP information of the received packet, the generated MAC address is set as a source MAC address, and the number of TSN domains is subtracted from the domain number of the PTP information before forwarding.
2. The time synchronization system according to claim 1. Each TSN (Time-Sensitive Networking) network;
A "NON-PTP (Precision Time Protocol)" network composed of communication devices that do not have a PTP time synchronization function;
A communication device having a PTP time synchronization function, which is connected to the boundary between each of the TSN networks and the "NON-PTP"network;
Equipped with
A method for performing time synchronization between TSN-compatible communication devices disposed in each of the TSN networks via the NON-PTP (Precision Time Protocol) network , the method being executed by a communication device connected to the boundary,
a step of converting a TSN time synchronization packet into a PTP packet when the TSN time synchronization packet is received from a communication device of the TSN network , and transferring the converted PTP packet to the communication device of the "NON-PTP"network;
a step of reconverting the PTP packet into a TSN time synchronization packet when the PTP packet is received from a communication device of the "NON-PTP" network , and transferring the reconverted TSN time synchronization packet to the communication device of the TSN network ;
A time synchronization method comprising: Each TSN (Time-Sensitive Networking) network;
A "NON-PTP (Precision Time Protocol)" network composed of communication devices that do not have a PTP time synchronization function;
and a communication device having a PTP time synchronization function,
When a TSN time synchronization packet is received from a communication device of the TSN network , the TSN time synchronization packet is converted into a PTP packet, and the converted PTP packet is transferred to the communication device of the "NON-PTP"network;
A communication device characterized by : when receiving the PTP packet from a communication device of the "NON-PTP" network , reconverting the PTP packet into a TSN time synchronization packet, and transferring the reconverted TSN time synchronization packet to the communication device of the TSN network . If the TSN time synchronization packet received from the communication device of the TSN network is a "SYNC" message,
Calculating a residual time from a time difference between a transmission time of the converted PTP packet and a reception time of the TSN time synchronization packet;
If the TSN time synchronization packet received from the communication device of the TSN network is a "Follow_UP" message,
5. The communication device according to claim 4, wherein the remaining time is added to time information described in the "Follow_UP" message . 6. The communication device according to claim 4, further comprising a switching hub.

Images