JP6045950B2 - Communication control device and communication system - Google Patents

Description

  The present invention relates to a communication control device and a communication system, and is suitable for application to a communication control device and a communication system that synchronize time between communication devices.

  In a control system configured in a distributed manner, it may be required to synchronize the time between connected terminals in accordance with an application to be realized. In the case of a distributed control system connected via a network, the time can be synchronized by transmitting and receiving a time synchronization packet via the network.

  For example, in the case of an industrial manufacturing apparatus, a controlled object such as a robot arm, a chip mounter, or a machine tool table is operated by a plurality of servo motors. In order to make these controlled objects perform desired operations, it is necessary to control a plurality of controlled servo motors in synchronization. Other applications of time synchronization include the fields of measurement and measurement, base stations for multimedia and wireless communication, and the field of power systems.

  As a time synchronization method using a network, NTP (Network Time Protocol), SNTP (Simple Network Time Protocol), IEC 61158, IEC 61784-2 Communication Profile Family 12 (hereinafter referred to as EtherCAT (Registered Trademark) E) 15, etc. A synchronization protocol is mentioned.

  By using the above synchronization protocol, time information packets are transmitted and received on the network, and the communication delay between the communication devices is measured, thereby synchronizing the time between the communication devices. When synchronizing the time, a communication device that provides reference time information (hereinafter referred to as a time master) is determined on the network, and the remaining communication devices (hereinafter referred to as time slaves) are designated as time masters. Synchronize. Note that the time master and the time slave are not fixed functions but play a role in time synchronization, so that each communication device can dynamically become either a time master or a time slave by a time synchronization protocol.

  In a highly reliable control system that requires high availability, it is required to prevent the system from being stopped due to a time master failure. As a general measure, there is multiplexing such as duplication of time masters. For example, in IEEE 1588, another communication device can substitute for a time master failure. However, a predetermined time-out period must be waited until a time master failure is detected, and synchronization processing between communication devices is not executed during that time. Furthermore, even if the time master is made redundant, a transient error may occur due to the time slave initializing control parameters for time synchronization.

  As described above, when the time master is made redundant for high reliability of the control system, there is a problem that the synchronization accuracy is deteriorated when the time master is switched. Therefore, in Patent Document 1, a plurality of time masters function simultaneously, and in a time slave, time information from a plurality of time masters and communication delays to the respective time masters are combined in a predetermined method and synchronized. A technique for reducing the influence of a failure of two time masters is disclosed.

JP 2012-23654 A

  However, since the plurality of time masters function simultaneously in Patent Document 1, there is a problem in that the amount of communication due to time synchronization packets increases and the effective communication bandwidth in the control system is reduced. Furthermore, in a general time synchronization protocol, such an operation is not defined, and it is difficult to configure a system with a terminal device that complies with these standards.

  The present invention has been made in view of the above points, and proposes a communication control device and a communication system that can make a time master function redundant by using an existing synchronization protocol and make the time synchronization function highly reliable. It is something to try.

In order to solve such a problem, in the present invention, a communication control device that transmits and receives a synchronization packet to and synchronizes time with one or more communication devices connected via a network, As a parameter obtained by observing one synchronization packet, when a parameter storage unit that stores a sequence number of the first synchronization packet and an abnormality in the first communication device are detected , A packet generation unit that generates a second synchronization packet based on a sequence number of the first synchronization packet and time information of the communication control device itself; and a second synchronization packet other than the first communication device. There is provided a communication control device comprising a communication unit that transmits to another communication device.

According to such a configuration, the communication control device saves the first first sequence number of the synchronization packet is a parameter obtained by observing the synchronization packets of a first communication device acting as a time master, the When an abnormality of the communication device is detected, a second synchronization packet is generated based on the stored sequence number of the first synchronization packet and the time information of the communication control device itself, and is transmitted to another communication device. . As a result, the master function for providing time information can be made redundant, and a transient decrease in synchronization accuracy during master switching can be prevented.

  According to the present invention, the time master function can be made redundant by using an existing synchronization protocol, and the time synchronization function can be made highly reliable.

It is a block diagram which shows the structure of the communication system concerning the 1st Embodiment of this invention. It is a block diagram which shows the structure of the communication control apparatus concerning the embodiment. It is a block diagram which shows the structure of the communication control part concerning the embodiment. It is a conceptual diagram explaining the communication delay between the communication apparatuses concerning the embodiment. It is a conceptual diagram which shows the packet structure of EtherCAT concerning the embodiment. It is a block diagram which shows the function structure of the communication control apparatus concerning the embodiment. It is a flowchart which shows operation | movement of the communication control apparatus concerning the embodiment. It is a conceptual diagram which shows the change of the communication path concerning the embodiment. It is a conceptual diagram which shows the change of the communication path concerning the embodiment. It is a block diagram which shows the structure of the communication system concerning the 2nd Embodiment of this invention. It is a block diagram which shows the structure of the communication control apparatus concerning the embodiment. It is a block diagram which shows the function structure of the communication control apparatus concerning the embodiment. It is a block diagram which shows the execution procedure of the time synchronization protocol concerning the embodiment. It is a block diagram which shows message transmission / reception between the master and slave concerning the embodiment. It is a flowchart which shows operation | movement of the communication control apparatus concerning the embodiment. It is a conceptual diagram which shows the header format of IEEE1588 concerning the embodiment. It is a conceptual diagram explaining packet transmission of the communication control apparatus concerning the embodiment. It is a block diagram which shows the structure of the communication system concerning the 3rd Embodiment of this invention. It is a block diagram which shows the structure of BC concerning the embodiment. It is a block diagram which shows the function structure of BC concerning the embodiment.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

(1) First Embodiment (1-1) Hardware Configuration (1-1-1) Configuration of Communication System First, the configuration of a communication system will be described with reference to FIG. As shown in FIG. 1, the communication system includes a control computer 120 and a plurality of communication devices 121a, 121b, and 121c (hereinafter, communication devices 121a to 121c may be collectively referred to as communication device 121). And the communication control device 123. The communication control device, the control computer 120 and the communication device 121 are communicably connected via a network 122.

  In this embodiment, the communication control device 123, the control computer 120, and the communication device 121 are described as communicating using EtherCAT. The effects of the present invention are not limited to EtherCAT, but are effective for similar communication methods. In the present embodiment, the effects of the invention will be described using EtherCAT as an example.

  The control computer 120 is an EtherCAT master device, and transmits and receives control command values, obtains measurement values, and performs various settings by transmitting and receiving communication packets to and from the communication control device 123 and the communication device 121.

  The communication device 121 is an EtherCAT slave device, and executes various processes based on the control command of the control computer 120 and updates the measurement value according to the execution result.

  Similar to the communication device 121, the communication control device 123 behaves as an EtherCAT slave device. On the other hand, it functions as a communication control device of the present invention and has a time master redundancy function. The time master redundancy function will be described in detail later.

(1-1-2) Hardware Configuration of Communication Control Device Next, a hardware configuration of the communication control device 123 will be described with reference to FIG. As shown in FIG. 2, the communication control device 123 includes a CPU 101, a PHY 104, a memory 108, a nonvolatile storage medium 109, a bus 110, a communication control unit 111, and the like.

  The CPU 101 functions as an arithmetic processing unit and a control unit, and transfers a program stored in the nonvolatile storage medium 109 to the memory 108 to execute various processes. Examples of the execution processing program include an operating system (hereinafter referred to as OS) and an application program operating on the OS. A program operating on the CPU 101 performs operation settings of the communication control unit 111 and acquires state information.

  The PHY 104 is a transceiver IC having a communication function with the network 122. As a communication standard provided by the PHY 104, an IEEE802.3 PHY (physical layer) chip is exemplified. In FIG. 2, since the PHY 104 and the communication control unit 111 are connected, processing of the IEEE 802.3 MAC (Media Access Control) layer is included in the communication control unit 111. However, an IC that provides the MAC function may be arranged between the communication control unit 111 and the PHY 104, or a communication IC that combines the IC that provides the MAC function and the PHY 104 may be connected to the communication control unit 111. The PHY 104 may be included in the communication control unit 111. In FIG. 2, four PHYs 104 are shown, but the number of PHYs 104 is the same as the number of communication ports supported by the communication control unit 111.

  The memory 108 is a temporary storage area for the CPU 101 to operate, and stores an OS, an application program, and the like transferred from the nonvolatile storage medium 109.

  The non-volatile storage medium 109 is a storage medium for storing various types of information, and stores an OS, an application, a device driver, etc., a program for operating the CPU 101, and a program execution result. . Examples of the nonvolatile storage medium 109 include a hard disk drive (HDD), a solid state drive (SSD), and a flash memory. Further, as an external storage medium that can be easily removed, a floppy (registered trademark) disk (FD), CD, DVD, Blu-ray (registered trademark), USB memory, compact flash (registered trademark), or the like may be used.

  The bus 110 connects the CPU 101, the communication control unit 111, the memory 108, and the nonvolatile storage medium 109 to each other. Examples of the bus 110 include a PCI bus, an ISA bus, a PCI Express bus, a system bus, or a memory bus.

  The communication control unit 111 is an IC that provides an EtherCAT communication function. The hardware configuration of the communication control unit 111 will be described with reference to FIG.

  As shown in FIG. 3, LANs 102 a to 102 d (hereinafter, LANs 102 a to d may be collectively referred to as LAN 102) are transceiver ICs for transmitting and receiving to and from a network 122. The LAN 102 transmits a packet received via the network 122 to an EPU (EtherCAT Processor UNIT) 103 or another LAN 102, or transmits a packet transmitted from the EPU 103 or another LAN 102 to the network 122.

  The LAN 102 includes IEEE 802.3 standard MAC (Media Access Control) chip, PHY (physical layer) chip, MAC and PHY composite chip, FPGA (Field Programmable Gate Array), CPLD (Complex Programmable Logic Device), ASIC (ASIC). Application Specific Integrated Circuit) and IC such as a gate array can be exemplified. Note that the LAN 102 may be included in the EPU 103. In FIG. 3, four LANs 102 are shown, but the present invention is not limited to this example, and a plurality of LANs 102 may be provided.

  The EPU 103 is an IC that executes communication processing in accordance with the EtherCAT specification. Specifically, the EPU 103 analyzes the received packet according to the EtherCAT standard, extracts necessary parameters, executes processing according to the parameters, and writes predetermined information to the packets. As the EPU 103, an IC such as a CPU, FPGA, CPLD, ASIC, or gate array is exemplified.

  The loopback function unit 112 sets the packet transfer destination to the network to which the LAN 102 is connected or an adjacent loopback function unit 112 according to the communication establishment status or setting in the LAN 102 to be connected.

(1-2) EtherCAT Time Synchronization Function Next, a mechanism of time synchronization in EtherCAT will be described with reference to FIG.

  The EtherCAT time synchronization function is composed of three stages: communication delay measurement between slaves, offset compensation of each slave's clock, and drift compensation of the clock progress of each slave. Among these, the measurement of the communication delay between each slave is based on the algorithm shown in FIG. In FIG. 4, it will be described below that the communication device 121a behaves as a slave A, the communication device 121b behaves as a slave B, and the communication device 121c behaves as a slave C.

  FIG. 4 shows the packet communication path and the reception time at each port of each slave in each of the communication devices 121 of EtherCAT slaves A, B, and C. As described above, the EPU 103 (denoted as EtherCAT Processor Unit in the figure) is an IC that executes communication processing in conformity with the EtherCAT specification.

  In FIG. 4, Ta, Tb, Tc, Td, and Te are reception times at predetermined ports for time synchronization (referred to as time synchronization packet 1), and tdiffAB, tdiffBC, tdiffCB, and tdiffBA are communication between slaves. Indicates delay. For example, tdiffAB indicates a communication delay between slave A and slave B. tp indicates the processing time of the EPU 103. TAP, TBP, and TCP are reception times at the EPU 103 of each of the slaves A, B, and C.

  Here, when paying attention to the slave A, Ta and Te are measured by the clock on the same communication apparatus 121a, but since the offset and drift of the clock are different from the communication apparatus 121b, they are directly compared with Tb and Td. It is not possible.

  When Ta to Te in FIG. 4 can be measured and tp is known, the communication delay tdiffBC between the slave C and the slave B can be calculated by Expression (1).

tdiffBC = (Td−Tb) ÷ 2 (1)

  However, it is assumed that tdiffBC and tdiffCB are equal. It is assumed that the EtherCAT slave processor unit uses the same type. If the processor units are different, it is necessary to consider the difference in tp.

  Further, the communication delay tdiffAB between the slave B and the slave A can be obtained by Expressions (2) and (3).

tdiffAB = ((Te−Ta) − (Td−Tb) −tp) ÷ 2 + tp (2)
tdiffBA = tdiffAB-tp (3)

  An EtherCAT slave that does not support the time synchronization function is regarded as a simple communication path. That is, it is necessary to transfer a packet with a preset fixed communication delay to an EtherCAT slave that does not support the time synchronization function.

  When realizing the time synchronization function in this way, it is necessary to circulate the packet for time synchronization on the ring network. That is, it is necessary to calculate the communication delay based on the reception time of the packet on the forward path and the return path in a certain slave.

  Thus, after obtaining the communication delay between the slaves, these communication delays can be integrated to obtain the communication delay between each slave and the slave having the reference time. For example, when the slave A is used as a reference in FIG. 4, the communication delay between the slave A and the slave B is tdiffAB, and the communication delay tdiffAC between the slave A and the slave C is obtained by Expression (4).

tdiffAC = tdiffAB + tdiffBC (4)

  Using these communication delays, the master notifies the slaves of the TAP of slave A that is the reference time. Each slave can calculate the time offset with respect to the slave A, which is the reference time, using the reception time of the EPU 103 of each slave. For example, the time offset OffsetB with respect to the slave A in the slave B is obtained by Expression (5).

OffsetB = TAP-TBP (5)

  Each slave can synchronize its own time with the reference time by using a communication delay with the slave A and a time offset.

  In EtherCAT, the time of the slave closest to the master (small number of hops) among the slaves corresponding to time synchronization is set as the reference time. In FIG. 1, the communication device 121a is a time master.

  Next, the EtherCAT packet configuration will be described with reference to FIG. The communication content of Ethernet CAT is stored in the data area 202 of the Ethernet (registered trademark) frame 200. In EtherCAT, the Type field of the Ethernet (registered trademark) header 201 is 0x88A4. Whether or not the data is correct is checked using an FCS (Frame Check Sequence) 203.

  The EtherCAT data structure includes an EtherCAT header 204 and one or a plurality of EtherCAT telegrams 205. The EtherCAT telegram 205 includes a telegram header 206, telegram data 207, and a working counter 208 (WKC). The working counter 208 is a field that is counted up by a predetermined number by the slave every time the telegram is correctly processed by the slave that should process the telegram.

  A command 220 indicates a command in the EtherCAT specification. The identifier 221 is a datagram identifier. The address 222 is an address in the EtherCAT specification. The data size 223 is the size of the data 207. C225 indicates that the frame is circulating. M226 indicates that there is a subsequent telegram. The IRQ 227 is used to notify the control computer 120 of the occurrence of a predetermined event from the slave.

(1-3) Functional Configuration of Communication Control Device Next, a functional configuration of the communication control device 123 will be described with reference to FIG. As illustrated in FIG. 6, the communication control device 123 includes a parameter storage unit 130, communication units 131a and 131b (hereinafter, may be collectively referred to as the communication unit 131), a time master failure determination unit 132, and a time distribution packet. An update unit 133 is provided.

  The parameter storage unit 130 has a function of storing parameters necessary for time synchronization processing. These parameters may be the data itself in the packet received by the communication unit 131, or may be parameters that are modified based on the data or calculated by a predetermined method. Examples of parameters stored by the parameter storage unit 130 include a communication delay with the time master and an error from the time of the time master. The parameter storage unit 130 includes any one or more of the EPU 103, the memory 108, the nonvolatile storage medium 109, and the OS operating on the CPU 101.

  The communication unit 131 has a function of transmitting and receiving packets. The communication unit 131 includes the LAN 102 or the EPU 103 when the communication function 131 is included in the EPU 103. In FIG. 6, two communication units 131 are shown. The communication unit 131a corresponds to the LAN 102a, and the communication unit 131b corresponds to the LAN 102b.

  The time master failure determination unit 132 has a function of determining a failure of the time master based on data included in a packet transmitted / received by the communication unit 131, transmission / reception timing, and the like. The time master failure determination unit 132 is configured by one or a plurality of software programs operating on the EPU 103 and the CPU 101.

  The time distribution packet update unit 133 has a function of substituting the time master by updating data in the time distribution packet when the time master failure determination unit 132 detects a failure of the time master. The time distribution packet update unit 133 is configured by one or a plurality of software programs operating on the EPU 103 and the CPU 101.

  Separately, an IC such as FPGA, CPLD, or ASIC may be provided on the bus 110 to configure one or more of the parameter storage unit 130, the time master failure determination unit 132, and the time distribution packet update unit 133. .

(1-4) Details of Operation of Communication Control Device Next, details of the operation of the communication control device 123 will be described with reference to FIG. As shown in FIG. 7, the communication control device 123 waits until a time information distribution packet is received (S001). Packets other than the time information distribution packet are processed as normal EtherCAT slaves in the same manner as the communication device 121.

  If it is determined in step S001 that the time information distribution packet has been received, the time master failure determination unit 132 determines whether the time master is abnormal (S002). As described above, the time master failure determination unit 132 determines whether the time master is a failure based on the data included in the distribution packet received in step S001 and the transmission / reception timing.

  If it is determined in step S002 that the time master is abnormal, it is determined whether the parameter storage unit 130 stores necessary parameters (S003). Specifically, the parameters stored by the parameter storage unit 130 are a communication delay with the time master and a time error with the time master. These parameters are set in the initial procedure by the control computer 120 in the EtherCAT time synchronization procedure. That is, when necessary parameters are stored in the parameter storage unit 130, the time of the communication device 121a as the time master and the communication control device 123 are synchronized, and the communication control device 123 replaces the communication device 121a. This means that it can act as a time master.

  The parameter storage unit 130 may determine that the necessary parameters are stored because these parameters are not zero. Further, in the packet from the control computer 120, if there is an access to a register that holds a delay time and a time error, it may be determined that the parameters are stored, assuming that these parameters have been updated. Good. Note that these parameters are set to 0 only for a communication device that behaves as a time slave.

  Next, when it is determined in S003 that the parameter storage unit 130 stores necessary parameters, the time distribution packet update unit 133 updates the time information to be distributed (S004). At this time, the time information to be updated is determined based on the time of the communication control device 123 itself and the parameters stored in the parameter storage unit 130. Specifically, it is obtained by Expression (6).

T = L−d + O (6)

  Here, T is the updated time information, L is the time of the communication control device 123 itself, d is a communication delay between the time master and the communication control device 123, and O is a time error between the time master and the communication control device 123.

  Moreover, you may calculate by Numerical formula (7) using the time S which the communication control apparatus 123 is synchronizing with the time master.

T = S−d (7)

  Note that the update time T may be the time L of the communication control device 123 itself. However, in order to replace the original time master and prevent a transient decrease in synchronization accuracy, the formula (6) or the formula (7 ) It is recommended to update.

  After executing step S004, the communication unit 131b is used to transfer the packet to the subsequent communication device 121 (S006).

  On the other hand, if it is determined in step S002 that the time master is not abnormal, that is, the time master has not failed, the packet is processed in the same manner as a normal slave (S005). If it is determined in S003 that the necessary parameters are not stored in the parameter storage unit 130, the packet is processed in the same manner as a normal slave (S005). Thereafter, the communication unit 131b is used to transfer the packet to the subsequent communication device 121 (S006).

  Next, a time master failure determination method by the time master failure determination unit 132 in step S002 will be described.

  In EtherCAT, the time master stores its time information in the packet by accessing a predetermined address by the packet transmitted by the control computer 120. If the time master function is not operating normally, an abnormal value is written in the corresponding area of the packet received by the communication control device 123. An example of an abnormal value is 0. Even if the time master is normal, it may happen to be zero as time information. In this case, the time master failure determination unit 132 erroneously determines that the time master is abnormal. However, in step S004 in FIG. 7, the time master failure determination unit 132 updates to the time synchronized with the time master, so the synchronization accuracy is maintained.

  Alternatively, the time master packet abnormality may be determined by measuring the time interval of receiving the time distribution packet with the time of the communication control device 123 itself and comparing it with the difference in the time information stored in the time distribution packet. This determination criterion is expressed by Expression (8).

| (Li + 1−Li) − (Ti + 1−Ti) |> α (8)

  Here, Li is the time of the communication control device 123 when it is received i-th, Ti is the time information stored in the i-th received time distribution packet, and α is a threshold value for determining an abnormality. It is shown that α is determined based on an error in time advancement. For example, it is indicated that the error is greater than or equal to the maximum error that can be considered from the error in the progress of time.

  The above formula (8) is determined by using time distribution packets received continuously. However, when the condition of the formula (8) is continuously satisfied a predetermined number of times, it is determined that there is an abnormality, or the formula (8) ) (Li + 1−Li) − (Ti + 1−Ti) of a predetermined number of times, and a statistical value (average value, variance, integral value, total value, etc.) calculated by a predetermined number of times is compared with a predetermined threshold value. The time master may be determined to be abnormal.

  The causes of the abnormality that can be determined based on these reception intervals include a change in communication delay due to a change in the communication path as shown in FIG. 8 in addition to the time abnormality of the time master itself. In FIG. 8, it is assumed that the communication device 121a is a time master. Assume that an abnormality has occurred in the communication path between the communication device 121a and the communication control device 123, and that packets from the control computer 120 return to the control computer 120 after passing through the communication device 121a.

  As described above, the EtherCAT bucket is provided with a counter portion (working counter 208) to which a predetermined value is added when each slave processes. The control computer 120 detects an abnormality on the path when the counter on the packet is different from the expected value based on the total number of slaves. Alternatively, an abnormality on the route may be detected by receiving the packet at the communication port (communication port 1 in FIG. 8) that transmitted the packet.

  The control computer 120 that has determined an abnormality on the route by the above method transmits a packet from the alternative communication port (communication port 2 in FIG. 8). Note that the control computer 120 may transmit the same packet from both communication ports without determining the abnormality of the communication path.

  In this way, the time distribution packet is returned to the control computer 120 again after being received by the communication device 121a, and then transferred to the communication control device 123. In this case, since the communication delay from the reception by the communication device 121a to the reception by the communication control device 123 becomes longer than that at the normal time, the abnormality of the time information can be determined by the above formula (8) or the like.

  Further, when the time master failure determination unit 132 can detect an abnormality of the communication device 121a that is the time master, the time information can be corrected from the own time and the information of the time master (FIG. 7). Step S004). Thereby, it can prevent that the time precision of the other communication apparatus 121 falls.

  In addition, as shown in FIG. 9, there may be a case where the communication device 121a itself, which is the time master, goes down due to a failure. The failure of the communication device 121a itself includes an abnormality in the communication path at both ends of the communication device 121a. In this case, since the time distribution packet does not pass through the communication device 121a, the time information on the packet and the value of the working counter 208 are zero. Therefore, when the time information on the packet and the value of the working counter 208 are 0, it can be determined that the time master is abnormal.

  Also, the abnormality of the time master may be determined using the EtherCAT auto-increment address. The auto-increment address is an address system that increments the address value by 1 each time each slave processes a packet, and determines that the datagram is a packet for itself if the address value when received is 0. The datagram is a unit of EtherCAT packet and is the data 207 described above. In the time distribution packet, time information is distributed by a datagram using an auto increment address called ARMW command (Auto Increment Read Multiple Write).

  As shown in FIG. 9, when the time distribution packet does not pass through the communication device 121a, the auto increment address of the ARMW command is transferred to the communication control device 123 without incrementing 1. Therefore, the communication control device 123 can determine the abnormality of the time master by confirming the auto increment address described in the time distribution packet.

  If it is normal, when the communication control device 123 is a node adjacent to the communication device 121a, the auto-increment address of the received packet is 1 (the communication device 121a is 0). If the number of hops between the communication control device 123 and the communication device 121a is N (1 when adjacent) and the auto-increment address of the received packet is not N, it is determined that the time master is abnormal. The time master failure determination unit 132 obtains the hop count N between the communication device 121a as the time master and the communication control device 123 from the auto-increment address of the packet received in the initial procedure of time synchronization, and stores it in the parameter storage unit 130. keeping.

  When the auto increment address is not the expected value N, the time distribution packet update unit 133 updates the time information in the procedure of S004 in FIG. Further, the auto-increment address of the datagram is updated to the value obtained by Equation (9) and transferred to the next communication device 121.

AutoAddr = N + 1 (9)

  Here, AutoAddr is an auto increment address.

  In addition, when the time distribution packet is not received for a predetermined period, the time master failure determination unit 132 may determine that the time master is abnormal. The observation of the predetermined period may be started from the initial procedure of the time synchronization procedure, or may be started after receiving the time distribution packet. Alternatively, the observation of the time distribution packet reception interval may be started after receiving a predetermined packet instructing the start of observation of the time distribution packet reception interval. Alternatively, the communication control device 123 itself may be provided with predetermined input means (keyboard, switch, microphone, etc.), and observation of the reception interval of the time distribution packet may be started after inputting a predetermined signal.

  In addition, a fixed value may be set in advance as the reception interval threshold of the time distribution packet for determining abnormality, or a packet in a predetermined format may be received from the control computer 120 and set. Further, a threshold value may be set according to an input from an input unit provided in the communication control device 123. The threshold value of the reception interval may be determined by the accuracy of the time of the communication control device 123 itself determined by physical characteristics such as a crystal resonator and the synchronization accuracy required by the system. In other words, the execution cycle is determined so that the accuracy error that shifts in the period when the synchronization accuracy is not performed does not fall below the required synchronization accuracy.

  For example, when using a communication device in which a synchronization accuracy error of F (time) is generated without executing the synchronization processing for 1 second and the required synchronization accuracy is P, the cycle Q at which the synchronization processing is to be executed is calculated from Equation (10). I want.

Q = 1 [unit: second] × P / F (10)

  For example, if F = 100 microseconds and P = 1 microseconds, then Q = 1 × 1/100 = 10 milliseconds.

  The period Q calculated by the equation (10) may be applied to the transmission interval of the time distribution packet of the control computer 120. This transmission interval needs to be matched to a node having the strictest conditions among the communication device 121 and the communication control device 123 on the network 122.

  As described above, when the time master failure determination unit 132 detects an abnormality of the time master, the time distribution packet cannot be received, and thus the procedure illustrated in FIG. 7 cannot be applied. In such a case, the communication control device 123 itself may generate and transmit a time distribution packet. At this time, the time information can be calculated by either the above formula (6) or formula (7).

  In addition, when the auto increment address is used, the auto increment address of the datagram to be transferred to the next communication device 121 can be calculated by Expression (9). The timing at which the communication control device 123 transmits the time distribution packet may be the timing at which the time master failure determination unit 132 determines that the time master is abnormal, or may be transmitted at a predetermined cycle. The period at this time may be determined based on Equation (10).

  In addition, a plurality of communication control devices 123 to which the present invention is applied may be used and configured in multiple stages. In this case, the time information on the packet is updated to the time information by the communication control device 123 that last processed the time distribution packet among the communication control devices 123 that the time master failure determination unit 132 determined to be abnormal. .

  Alternatively, the fact that the communication control device 123 has updated the time information may be recorded in a packet, and the communication control device 123 at the subsequent stage may control the time information update process by looking at the record.

  As a recording method of whether or not update processing is performed on the packet, an item on the header of the datagram of the EtherCAT packet (for example, IRQ: Interrupt Request) may be used, or a predetermined datagram indicating the information is added. May be.

  In addition, it is possible to indicate which communication control device 123 has executed the update process by the representation method of those records, and based on this, the control of the communication control device 123 at the subsequent stage can be changed. For example, when the IRQ of the datagram header is used, the bit position may be associated with the identification information of the communication control device 123.

  Further, if there is a communication control apparatus 123 that has already executed the update process, the communication control apparatus 123 at the subsequent stage may not execute the update process. Alternatively, priority may be set between the plurality of communication control devices 123, and whether or not to apply the update process may be determined based on the priority. For example, each communication control device 123 does not update its own time information when the communication control device 123 having a higher priority than itself updates the time information, and has a lower priority than itself. If 123 is updating the time information, its own time information may be updated.

  The priority may be determined based on the time accuracy and computer function of the communication control device 123. Examples of the computer function include the performance of the CPU 101, the capacity of the memory 108 and the nonvolatile storage medium 109, and the throughput. In addition, the priority is a statistical value (average value, minimum value, maximum value, variance, etc.) calculated based on the position on the network, communication bandwidth, communication volume per unit time, communication delay, communication medium, etc. You may decide based on.

(1-5) Effects of this Embodiment As described above, according to this embodiment, the communication control device 123 stores parameters obtained by observing the synchronization packet of the communication device 121 acting as a time master. When an abnormality is detected in the communication device 121, a synchronization packet is generated based on the saved parameters and the time information of the communication control device 123 itself, and is transmitted to another communication device. As a result, the master function for providing time information can be made redundant, and a transient decrease in synchronization accuracy during master switching can be prevented.

(2) Second Embodiment In the present embodiment, a case where IEEE 1588 is used for the time synchronization protocol will be described. The reference numerals used in the present embodiment mean that they are the same as the functions and elements described in the first embodiment unless otherwise specified. In the following, detailed description of the same configuration as that of the first embodiment will be omitted, and a configuration different from that of the first embodiment will be described in detail.

(2-1) Hardware Configuration (2-1-1) Configuration of Communication System FIG. 10 shows the configuration of the communication system according to the present embodiment. As shown in FIG. 10, the communication control device 140 has an IEEE 1588 function and communicates with another communication device 141 via the network 142.

  Similar to the communication control device 140, the communication device 141 has an IEEE 1588 function, and communicates with other communication devices 141 and the communication control device 140 via the network 142. Examples of the network 142 include IEEE 802.3 and various industrial networks.

  The network 142 is a broadcast domain, and broadcast packets and multicast packets transmitted from a certain communication control device 140 and communication device 141 are transferred to each communication control device 140 and communication device 141 connected to the network 142. The network 142 may include various network relay devices such as a network switch, a bridge, a router, an IEEE 1588 TC (Transparent Clock), and an OpenFlow switch.

(2-1-2) Hardware Configuration of Communication Control Device Next, a hardware configuration of the communication control device 140 will be described with reference to FIG. As shown in FIG. 11, the communication control device 140 includes a CPU 101, a LAN 102, a PHY 104, a memory 108, a nonvolatile storage medium 109, a bus 110, and the like. Since the CPU 101, the LAN 102, the PHY 104, the memory 108, the non-volatile storage medium 109, and the bus 110 are the same as those in the first embodiment, detailed description thereof is omitted.

(2-2) Functional Configuration of Communication Control Device Next, a functional configuration of the communication control device 140 will be described with reference to FIG. As illustrated in FIG. 12, the communication control device 140 includes a parameter storage unit 130, a communication unit 131, a time master abnormality determination unit 150, a time master parameter dynamic extraction unit 151, and a time master substitution unit 152. Since the parameter storage unit 130 and the communication unit 131 have the same functions as those in the first embodiment, detailed description thereof is omitted.

  The time master abnormality determination unit 150 has a function of determining abnormality of the time master based on data included in a packet transmitted / received by the communication unit 131, transmission / reception timing, and the like. The time master abnormality determination unit 150 is configured by a software program that operates on the CPU 101.

  The time master parameter dynamic extraction unit 151 has a function of extracting a parameter on a packet received from the communication device 141 acting as a time master and storing the parameter in the parameter storage unit 130 in IEEE 1588.

  The time master substitution unit 152 has a function of substituting the time master function when the time master abnormality determination unit 150 determines a time master abnormality.

(2-3) IEEE 1588 Time Synchronization Here, with reference to FIG. 13, an execution procedure of the IEEE 1588 time synchronization protocol will be described. In IEEE 1588, the communication device 141 takes a master-slave configuration. FIG. 14 shows message transmission / reception between a master and a slave in IEEE 1588.

  As shown in FIG. 13, first, the master transmits a Sync message to the slave (S060). At this time, the master records the transmission time t1 of the Sync message (S061). Then, when receiving the Sync message, the slave records the reception time t2 (S062).

  Then, the master notifies the slave of t1 by one of the following means (S063). The first means is a method of putting t1 information in the Sync message. The second means is a method of putting t1 information in the Follow_Up message following the Sync message.

  Subsequently, the slave transmits a Delay_Req message to the master (S064). At this time, the slave records the transmission time t3 of the Delay_Req message (S065). When receiving the Delay_Req message, the master records the reception time t4 (S066).

  Then, the master puts the information of t4 on the Delay_Resp message and notifies the slave of t4 (S067). The slave that has received the Delay_Resp message calculates a communication delay and a time difference between the master and the slave from t1, t2, t3, and t4 (S068).

  The calculation of the communication delay is based on the premise that the communication delay between the master and the slave is equal in the round trip path. Therefore, the one-way communication delay td is calculated by Expression (11).

td = ((t4−t3) + (t2−t1)) / 2 (11)

  Further, the difference tdiff between the master and slave times is calculated by Equation (12).

tdiff = {(t4-t3)-(t2-t1)} / 2 (12)

  The master and the slave can synchronize time by using tdiff calculated by the above equation (12).

(2-4) Details of Operation of Communication Control Device Next, details of the operation of the communication control device 140 according to the present embodiment will be described with reference to FIG.

  In IEEE 1588, the roles of the master and slave described above are determined in each port of each communication control device 140 and communication device 141 by a BMC (Best Master Clock) algorithm or static setting. At this time, it is assumed that the communication control device 140 becomes a slave in order to achieve an alternative function of the time master when the time master fails. Hereinafter, for convenience of explanation, it is assumed that the communication device 141a is a master. The communication device 141a that becomes the master periodically multicasts Sync packets to the network 142.

  First, the time master abnormality determination unit 150 determines whether the time master is normal (S010). If it is determined in step S010 that the time master is normal, the communication control apparatus 140 waits for reception of a packet (S011), and determines whether the received packet is a sync (S012).

  If it is determined in step S012 that the received packet is a sync, the communication control device 140 records the identifier, sequence ID, and domain number of the communication device 141a recognized as the master (S013), and a normal slave Similarly, the reception time of the Sync packet is recorded (S014).

  Here, in order to describe the parameters recorded by the communication control apparatus 140, the header format of IEEE 1588 will be described with reference to FIG.

  As illustrated in FIG. 16, the identifier of the communication device 141 a is sourcePortIdentity 168, the sequence number is sequenceId 169, and the domain number is domainNumber 165.

  In the communication control device 140, every time a Sync packet is received, the time master parameter dynamic extraction unit 151 extracts these parameters and stores them in the parameter storage unit 130.

  In step S010 of FIG. 15, when the time master abnormality determination unit 150 detects an abnormality of the time master function (N in S010), the time master substitution unit 152 uses the parameters stored in the parameter storage unit 130 to perform synchronization. A packet is generated (S015), and thereafter, the master function is replaced (S016).

  Specifically, in step S015, a Sync packet is generated using the source PortIdentity 168 and domainNumber 165 stored in the parameter storage unit 130. Also, in IEEE 1588, sequenceId 169 must be a number incremented by one, so the value obtained by adding 1 to the value of sequenceId 169 last extracted by time master parameter dynamic extraction unit 151 and stored in parameter storage unit 130 is used. Generate a Sync packet.

  The process shown in FIG. 15 may be executed periodically, or may be executed at a timing when the time master abnormality determination unit 150 changes the determination state of the time master.

  Next, the determination criteria of the time master abnormality determination unit 150 will be described. One criterion by which the time master abnormality determination unit 150 determines that the time master is abnormal may be determined to be abnormal due to a timeout that is defined in IEEE 1588 and has not been received for a predetermined period. However, in this method, a timeout may occur in the other communication device 141 (slave) as well, and the synchronization control parameter may be initialized, resulting in a transient decrease in accuracy.

  Therefore, it may be determined that there is an abnormality when the Sync packet is not received for a predetermined period or a predetermined period, or when the number of receptions per predetermined period is equal to or less than a predetermined threshold instead of the timeout of the Announce message.

  Alternatively, the communication control device 140 communicates a packet such as Ping to the communication device 141a which is the time master, and when there is no reply, when there is no continuous reception for a predetermined number of times, or the number of conversions during a predetermined period is You may determine with abnormality when it is below a predetermined threshold value.

  Similarly to Equation (8) above, the abnormality of the time master may be determined based on the difference between the reception interval of the time distribution packet (Sync) and the time information on the time distribution packet. Formula (8) is determined using time distribution packets received continuously, but when the condition of Formula (8) is continuously satisfied a predetermined number of times, or (Li + 1−Li of Formula 8). )-(Ti + 1-Ti) is measured a predetermined number of times, and the time master is determined to be abnormal by comparing a statistical value (average value, variance, integral value, total value, etc.) calculated with a predetermined threshold value. May be.

  However, when determining based on Formula (8), the time of the communication control apparatus 140 itself may be abnormal. In order to prevent such a case, the communication control devices 140 are multiplexed, and the determination result of each communication control device 140 is broadcast to the network 142 or multicast to the multicast group including the group of the communication control devices 140. Also good. The broadcast or multicast packet includes at least the determination result, and may include the identifier of the determined communication control device 140.

  Further, priority (weighting) may be given to each communication control device 140, and the degree according to the priority may be scored and included in the packet. This priority may be determined based on the degree of reliability design of each communication control device 140 (a combination of software and hardware redundancy, failure rate, etc.). Further, it may be determined based on the computer performance of the communication control device 140 (the performance of the CPU 101, the capacity of the memory 108, the nonvolatile storage medium 109, the throughput, etc. Further, the position in the network 142 and the hop to the time master It may be determined based on statistical values (average value, minimum value, maximum value, variance, etc.) calculated based on the number, communication delay, communication band, communication medium, and the abnormality determined by the communication control device 140 The factor may be reflected in the score.

  The communication control apparatus 140 that has received the packet determines that the time master is abnormal when the above score is greater than a predetermined threshold value or when the predetermined communication control apparatus 140 determines that the abnormality is abnormal.

  In the case of the time master abnormality determination as described above, the communication device 141a can continuously transmit the synchronization packet. In order to prevent each communication device 141 and communication control device 140 from receiving a plurality of synchronization packets (Sync), the communication control device 140 may be configured to stop transmission of the communication device 141a. For example, a management message for stopping the transmission of the synchronization packet or a unique TLV (Tag Length Value) message is transmitted to the communication device 141a. These messages may include at least contents for stopping the transmission of the synchronization packet.

  In addition, the network relay device is set so that the communication device 141a cannot be transmitted using a network relay device (not shown) having functions such as SNMP (Simple Network Management Protocol), VLAN, IEEE802.1Q, etc. Also good. Alternatively, the network relay device may be used to filter only the time distribution packet (Sync) and not to transfer it to another communication device.

Similarly, when a plurality of communication control devices 140 detect a time master abnormality using a plurality of communication control devices 140, a plurality of synchronization packets (Sync) may be transmitted.
In order to prevent such a case, as described above, priority may be provided to the plurality of communication control devices 140, and only the communication control device 140 with the highest priority may transmit the synchronization packet. Alternatively, the communication control device 140 with the highest priority may set the network relay device, and other communication control devices 140 may be set so that the time distribution packet (Sync) cannot be transmitted. At this time, as a criterion for determining the priority, the time at which the abnormality is determined may be earlier, and the highest priority may be given to the communication control device 140 that has determined the abnormality of the time master earliest.

  Alternatively, a method of allocating a fictitious time slot to each communication control device 140 within a predetermined period based on the transmission period of the Sync or the transmission period of the Announce message is exemplified. Specifically, when each communication control device 140 determines that the time master is abnormal, another communication control device 140 may transmit a synchronization packet in a time slot before the time slot to which it is assigned. For example, it does not send a synchronization packet. Moreover, when the other communication control apparatus 140 is not transmitting the synchronous packet, you may make it transmit a synchronous packet itself.

  Next, with reference to FIG. 17, an operation when the communication control devices 140a to 140d are used will be described. When attention is paid to the communication control device 140d, since the communication control device 140b transmits a Sync packet in FIG. 17A, the communication control device 140d itself does not transmit a synchronization packet even if it determines that the time master is abnormal. .

  On the other hand, in the case of FIG. 17B, since the other communication control device 140 has not transmitted the Sync packet before the communication control device 140d, the communication control device 140d determines that the time master is abnormal. The control device 140d transmits a Sync packet.

  The priorities described above may be used for time slot allocation. In addition, since the determination of the priority includes a dynamic reference, the time slots shown in FIG. 17 may be dynamically configured, and the time slots assigned to the communication control devices 140 may be communicated. Such a time slot may be executed by one of the communication control devices 140 or may be executed by a dedicated communication device 141.

  Since each communication control device 140 is synchronized by a time synchronization protocol such as IEEE 1588, synchronization as shown in FIG. 17 can be started at an arbitrary timing. Therefore, the communication device that performs time slot allocation notifies each communication control device 140 of the start of the cycle or the length of the cycle.

  Further, the communication control apparatus 140 may always transmit the Sync packet at a predetermined cycle after storing necessary parameters of the communication apparatus 141a (time master) in step S013 of FIG. In such a case, the communication device 141 that has received two identical Sync packets may use both of the two packets, or may use only one (for example, late arrival discard).

  Further, after determining that the time master is abnormal, it may be determined that the time master is normal again. In such a case, the communication control device 140 may stop transmitting its own synchronization packet or may continue to transmit the synchronization packet. The time master abnormality determination unit 150 may synchronize with the time master while the time master is determined to be normal. With this configuration, even if the time master becomes abnormal and the communication control device 140 is substituted, it is possible to prevent a decrease in synchronization accuracy.

  The communication control device 140 has the functional configuration shown in FIG. 12, but in combination with the communication unit 131, any one of the time master abnormality determination unit 150, the time master parameter storage unit 151, the parameter storage unit 130, and the time master substitution unit 152. Or more than one may be configured as separate communication devices. In this case, the relationship between the functional units, the request, and the notification of information are realized as a network 142 and a packet on the network 142. For example, the determination result of the time master abnormality determination unit 150 becomes a predetermined packet format and is transferred to the communication device serving as the time master substitution unit 152.

  Constructed as described above, for example, by arranging a plurality of time master abnormality determination units 150, the abnormality detection rate and accuracy of the time master are improved, or a partial function is a dedicated high-performance device or a low-performance device. And can optimize performance and cost optimization.

  In addition, even in the case of a multi-level broadcast domain, the time master can be replaced across domains. In this case, the communication device representing each functional unit is managed for each domain.

  Moreover, the communication control apparatus 140 may transmit the Announce message distributed by the communication apparatus 141a (time master) instead of the time master. In this case, while the time master abnormality determination unit 150 determines that the time master is normal, the Announce message distributed from the communication device 141a may be analyzed and necessary parameters may be stored in the parameter storage unit 130. .

  Note that a GPS (Global Positioning System) may be connected to the communication control device 140 and used as a time reference for synchronizing the GPS time. The network 142 can be applied with any topology, and may be a line topology, a star topology, or a ring topology. In the case of a ring topology, application of IEC62439-3 HSR (High Availability Seamless Redundancy) is exemplified.

(2-5) Effects of the present embodiment As described above, according to the present embodiment, the time master function can be made redundant by using the communication control device 140 to make the time synchronization function highly reliable. it can. Furthermore, by detecting an abnormality of the time master and substituting the time master, it is possible to prevent a transient decrease in synchronization accuracy. Further, even when the time master is dynamically changed, the time master function can be made redundant, and the reliability of the alternative function can be improved by multiplexing the communication control devices 140.

(3) Third Embodiment In this embodiment, a case will be described in which the IEEE 1588 master and slaves span multiple hierarchies using the Boundary Clock (hereinafter referred to as BC) of the time synchronization protocol IEEE 1588. . Unless otherwise specified, the reference numerals used in the present embodiment mean the same functions and elements as those described in the first embodiment and the second embodiment described above.

(3-1) Hardware Configuration (3-1-1) Communication System Configuration FIG. 18 shows a system configuration using the BC 180 of the present embodiment. In FIG. 18, it is assumed that the communication device 141a is a grand master (an IEEE 1588 master, the highest master in the slave hierarchy).

  The BC 180 has an IEEE 1588 BC function, and determines the status of the IEEE 1588 master, slave, etc. to each communication port according to the BMC algorithm or the BMC algorithm, and communicates the time synchronization packet.

  The network relay device 181 is various network relay devices such as a network switch, a bridge, a router, an IEEE 1588 TC (Transparent Clock), an IEEE 1588 BC, and an OpenFlow switch. In addition, you may use BC180 concerning this Embodiment.

(3-1-2) Hardware Configuration of Boundary Clock Next, the hardware configuration of the BC 180 will be described with reference to FIG. As shown in FIG. 19, the BC 180 includes a CPU 101, a PHY 104, a memory 108, a nonvolatile storage medium 109, a bus 110, a path control IC 190, and the like. Since the CPU 101, the PHY 104, the memory 108, the nonvolatile storage medium 109, and the bus 110 are the same as those in the first embodiment, detailed description thereof is omitted.

  The route control IC 190 has a network packet transfer and relay function, and also has an IEEE 1588 BC function, and executes state determination for each communication port, transmission / reception of a time synchronization packet, acquisition of a time stamp, setting, and the like. Further, the number of PHYs 104 is required by the number of communication ports included in the BC 180.

(3-2) Functional Configuration of Boundary Clock Next, the functional configuration of the BC 180 will be described with reference to FIG. As shown in FIG. 20, the BC 180 includes a parameter storage unit 130, a communication unit 131, a time master abnormality determination unit 150, a time master parameter dynamic extraction unit 151, a time master substitution unit 152, and the like.

  Since the parameter storage unit 130 and the communication unit 131 have the same functions as those in the first embodiment, detailed description thereof is omitted. Note that the number of communication units 131 is the same as the number of PHYs 104 in FIG.

  In the BC 180, one communication port becomes a slave and the other communication port becomes a master by BMC or an algorithm equivalent to BMC. Of the communication units 131, the communication unit 131 connected to the grand master communication device 141a is a slave port.

  The time master parameter dynamic extraction unit 151 extracts necessary parameters from the master packet received from the communication unit 131 and depends on the parameter storage unit 130.

  When the time master abnormality determination unit 150 determines that the communication device 141a is abnormal, the grand master may become another communication device due to re-execution of the BMC algorithm, which may cause a transient deterioration in time accuracy. . In order to prevent such a case, the time master substitution unit 152 generates an Announce packet similar to that when the communication device 141a is normal and transmits it from the communication unit 131. Further, the master processing of each port is continued based on the own time synchronized with the communication device 141a.

(3-3) Effects of this Embodiment As described above, according to this embodiment, even when a plurality of BCs 180 are used and the IEEE 1588 network is hierarchized, the time master function is made redundant. High reliability can be achieved.

120 Computer for Control 121 Communication Device 123 Communication Control Device 101 CPU
104 PHY
DESCRIPTION OF SYMBOLS 108 Memory 109 Nonvolatile storage medium 110 Bus 111 Communication control part 130 Parameter preservation | save part 131 Communication part 132 Time master failure determination part 133 Time distribution packet update part

Claims (20)

A communication control device that transmits and receives a synchronization packet to and synchronizes time with one or more communication devices connected via a network,
A parameter storage unit that stores a sequence number of the first synchronization packet as a parameter obtained by observing the first synchronization packet of the first communication device;
Packet that generates a second synchronization packet based on the sequence number of the first synchronization packet as the parameter and the time information of the communication control device itself when an abnormality of the first communication device is detected A generator,
A communication control device comprising: a communication unit that transmits the second synchronization packet to a communication device other than the first communication device. The packet generator is
2. The communication control device according to claim 1, wherein the time information of the second synchronization packet is generated based on the synchronized time information of the first communication device. 3. A failure determination unit for determining whether a failure has occurred in the first communication device;
The failure determination unit
When the time information included in the first synchronization packet is 0, it is determined that a failure has occurred in the first communication device, and an abnormality of the first communication device is detected. The communication control device according to claim 1. A failure determination unit for determining whether a failure has occurred in the first communication device;
The failure determination unit
A reception interval of the first synchronization packet, wherein the first communication apparatus continuously receives, based on the interval of the first time information included in the synchronization packet, the first communication device The communication control device according to claim 1, wherein an abnormality is detected. A failure determination unit for determining whether a failure has occurred in the first communication device;
The failure determination unit
When the reception interval of said first synchronization packet for a first communication device continuously receives the predetermined value greater than the spacing of the time information included in the first synchronization packet, wherein the first The communication control device according to claim 1, wherein an abnormality of the communication device is detected. A failure determination unit for determining whether a failure has occurred in the first communication device;
The failure determination unit
The communication control device according to claim 1, wherein an abnormality of the first communication device is detected when the parameter included in the first synchronization packet is different from a predetermined expected value. A failure determination unit for determining whether a failure has occurred in the first communication device;
The failure determination unit
The communication control device according to claim 1, wherein an abnormality of the first communication device is detected when the first synchronization packet is not received for a predetermined period. A failure determination unit for determining whether a failure has occurred in the first communication device;
The failure determination unit
The communication control device according to claim 1, wherein an abnormality of another communication device is detected based on a result of communication with the first communication device by the communication unit. The failure determination unit
The communication result between the first communication device by the communication unit, when there is no reply from the first communication device, and detecting an abnormality of another communication device, according to claim 8 Communication control device. A failure determination unit for determining whether a failure has occurred in the first communication device;
When an abnormality of the first communication device is detected by the failure determination unit,
The communication control device according to claim 1, wherein the communication unit disables transmission of the first synchronization packet. A failure determination unit for determining whether a failure has occurred in the first communication device;
When it is determined that the first communication device in which an abnormality is detected by the failure determination unit is in a normal state,
The communication control apparatus according to claim 1, wherein the communication unit stops transmission of the second synchronization packet. A failure determination unit for determining whether a failure has occurred in the first communication device;
When an abnormality of the first communication device is detected by the failure determination unit,
The communication control according to claim 1, wherein the communication unit transmits time attribute information including a communication delay and a time error with the first communication device instead of the first communication device. apparatus. The communication unit is
The communication control apparatus according to claim 1, wherein the second synchronization packet is transmitted at a predetermined cycle. The packet generator is
The communication control device according to claim 1, wherein generation information indicating that the communication control device has generated the second synchronization packet is included in the second synchronization packet. The packet generator is
The communication control device according to claim 1, wherein identification information for identifying the communication control device is included in the second synchronization packet in the second synchronization packet. The communication control device is connected to a plurality of other communication control devices via a network,
The packet generator is
The second synchronization packet is not generated when any one of the plurality of other communication control devices generates the first synchronization packet. Communication control device. When priority is set for the other plurality of communication control devices,
The communication control apparatus according to claim 1, wherein the communication unit transmits the second synchronization packet when the set priority is the highest. The communication control device according to claim 17 , wherein the communication unit transmits the second synchronization packet in a time division manner based on the set priority. A network relay device,
The communication control device according to claim 1, wherein the communication unit sets the network relay device to disable communication of the first synchronization packet. A communication system in which at least two communication devices synchronize time via a network,
Storing a sequence number of the first synchronization packet as a parameter obtained by observing the first synchronization packet of one communication device;
When an abnormality of the one communication device is detected, another communication device performs second synchronization based on the sequence number of the first synchronization packet as the parameter and the time information of the other communication device. Generate packets,
The communication system, wherein the other communication device transmits the second synchronization packet.

Images