Network management device, transfer device, network system, and transfer schedule setting method
By calculating transfer schedules based on the time update width, the network management device addresses timing discrepancies in network devices, ensuring consistent communication performance and efficient bandwidth utilization.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2024-03-11
- Publication Date
- 2026-03-13
AI Technical Summary
Existing methods for setting transfer schedules in network devices fail to account for timing differences between devices, leading to potential mismatches in communication frame transfers and suboptimal bandwidth utilization.
A network management device calculates a transfer schedule using the time update width, which accounts for the difference between synchronization time and internal time, to ensure accurate timing of communication stream transfers.
This approach ensures consistent communication bandwidth and delay performance by adjusting transfer schedules to accommodate timing discrepancies, improving bandwidth utilization and reducing labor costs through automated scheduling.
Smart Images

Figure 0007829827000001 
Figure 0007829827000002 
Figure 0007829827000003
Abstract
Description
Technical Field
[0001] The present disclosure relates to a technique for setting a transfer schedule that determines a period assigned to each communication stream when transferring a plurality of communication streams while separating them temporally.
Background Art
[0002] In IEEE802.1Q-2018 (see Non-Patent Document 1), TAS is defined as a specification of a network device for ensuring the transmission quality of each communication stream in a network where communication streams with different quality-of-service levels coexist. TAS is an abbreviation for Time-Aware Shaper. In TAS, communication streams are stored in queues for each priority class by a network device, and read control of the queues scheduled based on the synchronization time in the network is performed. As a result, each communication stream is transferred while being separated temporally. A set value related to the read control of the scheduled queues is called a transfer schedule.
[0003] In Patent Document 1, a technique for setting a transfer schedule aimed at improving the accuracy of the transfer schedule is described. In Patent Document 1, network devices such as a network switch and a PLC acquire the propagation delay between the devices, and a network management device calculates a transfer schedule based on the propagation delay. Then, the network management device distributes the calculated transfer schedule to the network devices. PLC is an abbreviation for Programmable Logic Controller.
Prior Art Documents
Patent Documents
[0004]
Patent Document 1
Non-Patent Documents
[0005] [Non-Patent Document 1] IEEE 802.1Q-2018 - IEEE Standard for Local and Metropolitan Area Networks--Bridges and Bridged Networks. [Overview of the Initiative] [Problems that the invention aims to solve]
[0006] Patent Document 1 uses the propagation delay between devices to calculate the transfer schedule. However, due to timing differences in when network devices transmit communication streams, communication frames may not be transferred in the expected time slot, potentially resulting in an inability to obtain the expected communication bandwidth or delay. This disclosure aims to make it easier to achieve the expected communication bandwidth or delay even when there is a timing difference in the transmission of communication streams by network devices. [Means for solving the problem]
[0007] The network management device related to this disclosure is A network information acquisition unit acquires network information indicating the time update width, which is the difference between the synchronization time and the internal time managed by the transfer device before time synchronization by the synchronization time, from a transfer device that performs time synchronization by calculating the synchronization time based on a communication frame for time synchronization. A transfer schedule calculation unit calculates a transfer schedule that determines the period during which the transfer device performs data transfer for each priority of communication frames, using the time update width indicated by the network information acquired by the network information acquisition unit. A schedule distribution unit distributes the transfer schedule calculated by the transfer schedule calculation unit to the transfer device. It is equipped with. [Effects of the Invention]
[0008] In this disclosure, the transfer schedule is calculated using the time update width. The time update width is a parameter that affects the timing difference in which network devices transmit communication streams. Therefore, by calculating the transfer schedule using the time update width, it becomes easier to achieve the expected communication bandwidth or delay even if there is a timing difference in which the transfer device 10 transmits communication streams. [Brief explanation of the drawing]
[0009] [Figure 1] Configuration diagram of the network system 100 according to Embodiment 1. [Figure 2] A functional configuration diagram of the transfer device 10 according to Embodiment 1. [Figure 3] Hardware configuration diagram of the transfer device 10 according to Embodiment 1. [Figure 4] Functional configuration diagram of the network management device 20 according to Embodiment 1. [Figure 5] Hardware configuration diagram of the network management device 20 according to Embodiment 1. [Figure 6] A schematic diagram illustrating the operation of the network system 100 according to Embodiment 1. [Figure 7] A flowchart illustrating the process from when the transfer device 10 according to Embodiment 1 receives a communication frame for time synchronization containing time information, until it notifies the network management device 20 of the time update width, etc. [Figure 8] A diagram illustrating the relationship between network information 51 and the transfer schedule 52. [Figure 9] A diagram illustrating the relationship between synchronization error and time slot margin. [Figure 10] A flowchart illustrating the process from when the network management device 20 according to Embodiment 1 receives a communication frame containing network information 51 until it distributes the transfer schedule 52 to the transfer device 10. [Modes for carrying out the invention]
[0010] Embodiment 1. ***Explanation of the structure*** Referring to FIG. 1, the configuration of the network system 100 according to Embodiment 1 will be described. The network system 100 includes a plurality of transfer devices 10, a network management device 20, a time distribution device 30, and a plurality of terminals 40. The plurality of transfer devices 10, the network management device 20, the time distribution device 30, and the plurality of terminals 40 are connected via a transmission line 90. Each transfer device 10 is a device that transfers a communication frame received from an adjacent device to another adjacent device. As a specific example, each transfer device 10 is a network device such as a network switch, a router, a hub, a bridge, a gateway, or a PLC. The network management device 20 is a computer that manages the operation of the network system 100. The time distribution device 30 is a computer that distributes communication frames for time synchronization. Each terminal 40 is a computer or the like used by a user.
[0011] Referring to FIG. 2, the functional configuration of the transfer device 10 according to Embodiment 1 will be described. The transfer device 10 includes a communication port 11, a timestamp acquisition unit 12, a communication control unit 13, and a setting unit 14. The communication control unit 13 includes a transfer unit 131, a time synchronization unit 132, an in-device delay calculation unit 133, and a notification unit 134. The time synchronization unit 132 includes a propagation delay calculation unit 1321, a synchronization time calculation unit 1322, and a time update width calculation unit 1323.
[0012] Referring to FIG. 3, the hardware configuration of the transfer device 10 according to Embodiment 1 will be described The transfer device 10 includes a processor 101, a memory 102, and data communication hardware 103. The processor 101, the memory 102, and the data communication hardware 103 are connected via a bus 104.
[0013] The communication port 11 is realized by the data communication hardware 103. The data processing units, comprising the timestamp acquisition unit 12, the communication control unit 13, and the setting unit 14, are implemented by hardware, software, firmware, or a combination thereof. The software and firmware are written as programs and stored in memory 102. When the data processing units are implemented by software or firmware, the data processing units are implemented by the processor 101 reading and executing programs from memory 102 for each of the data processing units to operate.
[0014] Referring to Figure 4, the functional configuration of the network management device 20 according to Embodiment 1 will be described. The network management device 20 includes a communication port 21, a network information acquisition unit 22, a synchronization error estimation unit 23, a setting unit 24, a transfer schedule calculation unit 25, and a transfer schedule distribution unit 26.
[0015] Referring to Figure 5, the hardware configuration of the network management device 20 according to Embodiment 1 will be described. The network management device 20 comprises a processor 201, memory 202, and data communication hardware 203. The processor 201, memory 202, and data communication hardware 203 are connected via a bus 204.
[0016] Communication port 21 is implemented by data communication hardware 203. The data processing units, comprising the network information acquisition unit 22, the synchronization error estimation unit 23, the setting unit 24, the transfer schedule calculation unit 25, and the transfer schedule distribution unit 26, are implemented by hardware, software, firmware, or a combination thereof. The software and firmware are written as programs and stored in memory 202. When the data processing units are implemented by software or firmware, the data processing units are implemented by the processor 201 reading and executing programs from memory 202 for each of the data processing units to operate.
[0017] Processors 101 and 201 are CPUs or system LSIs, etc. CPU stands for Central Processing Unit. LSI stands for Large Scale Integration. Processors 101 and 201 can also be referred to as processing units, arithmetic units, processors, microprocessors, microcomputers, or Digital Signal Processors.
[0018] Memory 102,202 is a non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, EPROM, or EEPROM. RAM stands for Random Access Memory. ROM stands for Read Only Memory. EPROM stands for Erasable Programmable Read Only Memory. EEPROM stands for Electrically Erasable Programmable Read-Only Memory. Memory 102,202 may also be a portable recording medium such as a magnetic disk, flexible disk, optical disk, compact disk, minidisc, or DVD. DVD stands for Digital Versatile Disc.
[0019] The data communication hardware 103 and 203 are network cards for Ethernet®, etc.
[0020] The data processing units of the transfer device 10 and the network management device 20 may be implemented using dedicated hardware. Examples of dedicated hardware include single circuits, complex circuits, programmed processors, parallel programmed processors, ASICs, FPGAs, or combinations thereof. ASIC stands for Application Specific Integrated Circuit. FPGA stands for Field Programmable Gate Array. A portion of the data processing unit may be implemented using dedicated hardware, while the remainder is implemented using the processors 101, 201 and memories 102, 202 described above.
[0021] ***Explanation of operation*** The operation of the network system 100 according to Embodiment 1 will be described with reference to Figures 6 to 10. The operation procedure of the network system 100 according to Embodiment 1 corresponds to the transfer schedule setting method according to Embodiment 1.
[0022] **Overview of Operation** Referring to Figure 6, an overview of the operation of the network system 100 according to Embodiment 1 will be described. In the network system 100, data transfer is controlled according to the TAS defined in IEEE 802.1Q-2018. Specifically, the network management device 20 sets the transfer schedule 52, which is a setting related to queue reading control. Each transfer device 10 performs data transfer according to the transfer schedule 52 by reading the queue according to the transfer schedule 52.
[0023] A time synchronization communication frame containing time information is distributed from the time distribution device 30. The time synchronization communication frame is distributed to each terminal 40 via the transfer device 10. Each transfer device 10 and each terminal 40 performs time synchronization based on the time information contained in the time synchronization communication frame.
[0024] When performing time synchronization, each transfer device 10 measures the propagation delay between itself and adjacent devices, as well as the internal delay within the transfer device 10. Each transfer device 10 also calculates the time update width, which is the difference between the synchronization time calculated from the time information and the internal time managed by the transfer device 10 before the synchronization. Each transfer device 10 transmits network information 51, indicating the propagation delay, internal delay, and time update width, to the network management device 20. The network management device 20 uses the network information 51 transmitted from each transfer device 10 to calculate a transfer schedule 52 that defines the period during which each transfer device 10 will perform data transfer. The network management device 20 then distributes the transfer schedule 52 to each transfer device 10. Then, each transfer device 10 performs data transfer during the period specified in the transfer schedule 52.
[0025] **Operation of Transfer Device 10** The operation of the transfer device 10 according to Embodiment 1 will be explained with reference to Figure 2. Communication port 11 forwards frames constituting the communication stream received from an external device to the timestamp acquisition unit 12. Communication port 11 also forwards frames constituting the communication stream received from the timestamp acquisition unit 12 to the external device. The timestamp acquisition unit 12 forwards the communication frame received from the communication port 11 to the communication control unit 13. The timestamp acquisition unit 12 also forwards the communication frame received from the communication control unit 13 to the communication port 11. At this time, the timestamp acquisition unit 12 acquires the reception time of the communication frame received from the communication port 11, or the transmission time of the communication frame to be sent to the communication port 11, and notifies the communication control unit 13.
[0026] In the communication control unit 13, the transfer unit 131 transfers the communication frame received from the timestamp acquisition unit 12 to the corresponding destination communication port. If the communication frame received from the timestamp acquisition unit 12 is a communication frame for time synchronization, the transfer unit 131 transfers the communication frame to the time synchronization unit 132.
[0027] When forwarding a communication frame to a communication port, the forwarding unit 131 assigns a priority and an identifier to the communication frame based on the configuration information received from the configuration unit 14. The priority and identifier are determined by pattern matching between the received frame's attributes, such as the receiving port number, header information, payload information, and error information, and the configuration information received from the configuration unit 14. For example, suppose the configuration unit 14 is configured to assign a priority of 7 and an ID of 1 to a communication frame received on port number 1 if the destination MAC address is 12-34-56-AB-CD-EF and the EtherType is 0x0800 (IP communication). In this case, the forwarding unit 131 determines whether the communication frame matches this setting, and if it does, assigns a priority of 7 and an ID of 1 to the communication frame. The transfer unit 131 stores communication frames in queues corresponding to their priority. The transfer unit 131 then controls the reading of queues scheduled based on time, in accordance with the TAS defined in IEEE 802.1Q-2018. Here, the time in the transfer device 10 is synchronized with other devices by the time synchronization unit 132. This enables scheduled transfers, and each communication stream is transferred in a temporally separated manner.
[0028] Here, the transfer unit 131 receives a setting value related to the read control of the scheduled queue, i.e., the transfer schedule 52, from the setting unit 14. The setting unit 14 holds two transfer schedules 52: the currently running transfer schedule 52 and the transfer schedule 52 to be newly applied. When the time reaches the application time of the transfer schedule 52 to be newly applied, the transfer unit 131 adopts the transfer schedule 52 to be newly applied.
[0029] The time synchronization unit 132 performs time synchronization processing according to a time synchronization profile such as PTP or gPTP based on PTP as defined in IEEE 1588. PTP stands for Precision Time Protocol. gPTP stands for generalized PTP. The propagation delay calculation unit 1321 exchanges the reception time or transmission time of the time synchronization frame acquired by the timestamp acquisition unit 12 with an external device and calculates the propagation delay between it and the external time synchronization device. The propagation delay is the delay of the communication stream that occurs on the transmission path between devices. In the case of a wired network, the propagation delay is cable delay or fiber delay. The synchronization time calculation unit 1322 calculates the synchronization time based on the propagation delay calculated by the propagation delay calculation unit 1321 and the time information contained in the time synchronization communication frame distributed from the time distribution device 30. The time synchronization unit 132 synchronizes the internal time, which is the time managed by the transfer device 10, with the calculated synchronization time.
[0030] Here, errors occur in the calculated synchronization time due to factors such as the timestamp acquisition error and the clock frequency difference between the time distribution device 30 and the transfer device 10. As a result, a time jump occurs when synchronizing to the synchronization time calculated by the synchronization time calculation unit 1322. The time update width calculation unit 1323 calculates the time update width, which is the difference between the internal time, which is the time the transfer device 10 was running on its own before it synchronized with the time distribution device, and the synchronization time, which is the time it newly synchronizes.
[0031] The device delay calculation unit 133 calculates the delay of each communication stream within the transfer device 10. The delay of a communication stream is calculated by subtracting the reception time of the same communication stream from the transmission time of the communication stream acquired by the timestamp acquisition unit 12. The device delay calculation unit 133 calculates the device delay for each communication stream identifier assigned by the transfer unit 131.
[0032] The propagation delay calculated by the propagation delay calculation unit 1321, the time update width calculated by the time update width calculation unit 1323, and the in-device delay calculated by the in-device delay calculation unit 133 are transmitted to the notification unit 134. The propagation delay, in-device delay, and time update width may be the most recent measured values, or statistical values such as the minimum, maximum, average, or variance over a certain period may be notified. The notification unit 134 stores the received propagation delay, time update width, and in-device delay in a communication frame addressed to the network management device 20 and transmits it to an external device via the transfer unit 131, the timestamp acquisition unit 12, and the communication port 11. As a result, the propagation delay, time update width, and in-device delay are notified to the network management device 20.
[0033] The configuration unit 14 notifies the transfer unit 131 of the transfer schedule 52 it holds and the time at which the transfer schedule 52 will be applied. The configuration unit 14 acquires and holds the transfer schedule 52 distributed from the network management device 20 via the communication port 11, the timestamp acquisition unit 12, and the transfer unit 131. The configuration unit 14 may also be configured to hold the transfer schedule 52 set by the user.
[0034] Referring to Figure 7, the process from when the transfer device 10 according to Embodiment 1 receives a communication frame for time synchronization containing time information to when it notifies the network management device 20 of the time update width, etc., will be explained. As described above, the transfer device 10 receives a communication frame for time synchronization. Then, the transfer unit 131 transfers the communication frame for time synchronization to the time synchronization unit 132.
[0035] (Step S11: Synchronization time calculation process) The synchronization time calculation unit 1322 calculates the synchronization time based on the propagation delay calculated by the propagation delay calculation unit 1321 and the time information included in the communication frame for time synchronization.
[0036] (Step S12: Initial calculation determination process) The synchronization time calculation unit 1322 determines whether the synchronization time calculation in step S11 is the first synchronization time calculation. The synchronization time calculation unit 1322 terminates processing if it is the first time calculating the synchronization time. On the other hand, if it is the second or subsequent time calculating the synchronization time, the synchronization time calculation unit 1322 proceeds to step S13. When calculating the synchronization time for the first time, it is expected that the difference between the calculated synchronization time and the internal time managed by the transfer device 10 will be larger compared to subsequent synchronization time calculations. Therefore, when calculating the synchronization time for the first time, the processing from step S13 onward is not performed, and the notification of the time update range at this timing is stopped.
[0037] (Step S13: Time update width calculation process) The time update width calculation unit 1323 calculates the time update width, which is the difference between the internal time and the synchronization time. In other words, the time update width D is expressed as D = TM - TS, where TM is the synchronization time and TS is the internal time.
[0038] (Step S14: Time update width notification process) The time update width calculation unit 1323 transmits the time update width calculated in step S13 to the notification unit 134. The notification unit 134 then transmits network information 51 to the network management device 20, which includes the time update width, the propagation delay transmitted from the propagation delay calculation unit 1321, and the in-device delay transmitted from the in-device delay calculation unit 133.
[0039] **Operation of Network Management Device 20** Referring to Figure 4, the operation of the network management device 20 according to Embodiment 1 will be explained. Communication port 21 forwards communication frames received from external devices to the network information acquisition unit 22 and the transfer schedule distribution unit 26. Communication port 21 also forwards communication frames received from the network information acquisition unit 22 and the transfer schedule distribution unit 26 to external devices.
[0040] The network information acquisition unit 22 acquires network information 51 from the communication frame transferred from the communication port 21. Once the network information acquisition unit 22 acquires the network information 51, it notifies the synchronization error estimation unit 23 of the time update width indicated by the network information 51, and notifies the transfer schedule calculation unit 25 of the propagation delay and in-device delay indicated by the network information 51.
[0041] The synchronization error estimation unit 23 estimates the statistical value of the time update range in the most recent base period as the synchronization error. Specifically, the synchronization error estimation unit 23 estimates the maximum value of the time update range in the most recent base period as the synchronization error. Alternatively, the synchronization error estimation unit 23 could calculate the standard deviation of the time update range in the most recent base period and estimate the range of 2σ as the synchronization error.
[0042] The configuration unit 24 holds system specification information that indicates the specifications of the network system 100. The configuration unit 24 notifies the transfer schedule calculation unit 25 of the system specification information. The system specification information includes topology information necessary for calculating the transfer schedule 52, as well as the transmission period, frame length, priority, and delay constraints of each communication stream.
[0043] The transfer schedule calculation unit 25 calculates the transfer schedule 52 using the synchronization error, propagation delay, and in-device delay for each transfer device 10, as well as the system specification information held by the setting unit 24. Specifically, the transfer schedule calculation unit 25 sets margins before and after each time slot using the synchronization error for each transfer device 10. A time slot is the period during which the transfer device 10 performs data transfer, as defined for each priority of the communication frame. In other words, the transfer schedule calculation unit 25 extends the period allocated to the transfer device 10 for each priority of the communication frame by the amount of the synchronization error. Then, while considering the margins, the transfer schedule calculation unit 25 calculates the transfer schedule 52 by searching for a combination of setting values that satisfies constraints such as the delay of each communication stream, based on propagation delay, in-device delay, and system specification information. The transfer schedule calculation unit 25 sends the transfer schedule 52, with the application time set, to the transfer schedule distribution unit 26.
[0044] Refer to Figure 8 to explain the relationship between network information 51 and transfer schedule 52. The transfer schedule 52 is calculated based on the propagation delay between the transfer devices 10 and the internal delay occurring within the transfer device 10, while taking into account a margin based on the synchronization error between the transfer devices. Specifically, transfer device #2 has a propagation delay X between it and transfer device #1. Transfer device #2 also has an internal delay Y. Therefore, the time slot allocated to transfer device #2 starts at a timing later than the time slot allocated to transfer device #1 by the propagation delay X + internal delay Y. From this timing, a period of time slot for transfer device #2 is allocated that includes a margin of synchronization error Z before and after the required period P determined based on the system specifications. The synchronization error between the transfer devices 10 affects the timing error of when the transfer device 10 transmits the communication frame. Therefore, as shown in Figure 8, it is desirable to set a margin of the synchronization error before and after the time slot.
[0045] Refer to Figure 9 to explain the relationship between synchronization error and time slot margin. The transfer schedule calculation unit 25 sets the maximum synchronization error during the transfer process as the time slot margin for each transfer path of the communication stream. For example, for transfer device #6, since the transfer path passes through transfer devices #1, #2, and #5, the maximum synchronization error among these, ±30 ns (nano second), is set as the margin.
[0046] The transfer schedule distribution unit 26 distributes the transfer schedule 52 transmitted by the transfer schedule calculation unit 25 to each transfer device 10 via the communication port 21. Then, in each transfer device 10, the distributed transfer schedule 52 is set in the setting unit 14, and when the applicable time arrives, it is adopted by the transfer unit 131.
[0047] Referring to Figure 10, the process from when the network management device 20 according to Embodiment 1 receives a communication frame containing network information 51 to when it delivers the transfer schedule 52 to the transfer device 10 will be explained. As described above, the network management device 20 receives a communication frame containing network information 51. The network information acquisition unit 22 then acquires the network information 51, notifies the synchronization error estimation unit 23 of the time update interval, and notifies the transfer schedule calculation unit 25 of the propagation delay and in-device delay.
[0048] (Step S21: Synchronization error estimation process) The synchronization error estimation unit 23 estimates the statistical value of the time update range in the most recent reference period as the synchronization error. The synchronization error estimation unit 23 notifies the transfer schedule calculation unit 25 of the estimated synchronization error.
[0049] (Step S22: Information determination process) The transfer schedule calculation unit 25 determines whether the information necessary for calculating the transfer schedule 52 has been collected. The information necessary for calculating the transfer schedule 52 includes the synchronization error, propagation delay, and in-device delay for each transfer device 10, as well as the system specification information held by the setting unit 24. The transfer schedule calculation unit 25 proceeds to step S23 if it has collected the necessary information. On the other hand, if the transfer schedule calculation unit 25 has not collected the necessary information, it terminates the process.
[0050] (Step S23: Margin setting process) The transfer schedule calculation unit 25 sets a margin before and after each time slot using the estimated synchronization error for each transfer device 10.
[0051] (Step S24: Transfer schedule calculation process) The transfer schedule calculation unit 25 searches for a combination of setting values that satisfies the constraints such as the delay of each communication stream, based on the propagation delay, in-device delay, and system specification information, while taking into account the margin set in step S23. Based on this, the transfer schedule calculation unit 25 calculates the transfer schedule 52.
[0052] (Step S25: Transfer schedule distribution process) The transfer schedule distribution unit 26 distributes the transfer schedule 52 calculated in step S24 to each transfer device 10.
[0053] ***Effects of Embodiment 1*** As described above, in the network system 100 according to Embodiment 1, the transfer device 10 periodically notifies the network management device 20 of the time update interval with the time distribution device 30. The network management device 20 then sets the transfer schedule 52 using the synchronization error obtained from the time update interval. This makes it easier to achieve the expected communication bandwidth or delay even if there is a timing difference in when the transfer device 10 transmits the communication stream.
[0054] By estimating the maximum time update range in the most recent reference period as the synchronization error, it becomes easier to achieve the expected communication bandwidth or delay even if there is a timing difference in the transmission of the transfer device 10. However, if the synchronization error is too large, the bandwidth utilization efficiency will deteriorate. By calculating the standard deviation of the time update range in the most recent reference period and estimating the range of 2σ as the synchronization error, it is possible to improve the average bandwidth utilization efficiency while achieving the expected communication bandwidth or delay.
[0055] Furthermore, in the network system 100 according to Embodiment 1, the process from the transmission of network information 51 by each transfer device 10 to the distribution of the transfer schedule 52 by the network management device 20 is automated. This reduces the labor required to set the transfer schedule 52, thereby lowering the costs of system implementation and operation.
[0056] ***Other configurations*** <Example 1> In Embodiment 1, the network management device 20 calculated the synchronization error from the time update interval. However, each transfer device 10 may calculate the synchronization error from the time update interval and notify the network management device 20. In this case, the time update interval calculation unit 1323 of each transfer device 10 calculates the synchronization error from the time update interval using the same method as the synchronization error estimation unit 23. The time update interval calculation unit 1323 then notifies the notification unit 134 of the synchronization error instead of the time update interval.
[0057] <Modification 2> In Embodiment 1, as described with reference to Figure 10, the transfer schedule calculation unit 25 recalculated and distributed the transfer schedule 52 upon receiving network information 51. However, the transfer schedule calculation unit 25 may also recalculate and distribute the transfer schedule 52 at at least one of the following times: at periodically determined timings, when there is a change in the network information 51 exceeding a certain threshold, or when data transfer occurs in violation of the period defined in the transfer schedule 52.
[0058] When a periodically defined timing is adopted, the network management device 20 is equipped with a periodic timer. When the periodic timer expires, the transfer schedule calculation unit 25 recalculates the transfer schedule 52 and distributes it.
[0059] If the system adopts the case where there has been a change in network information 51 that exceeds a certain threshold, the transfer schedule calculation unit 25 determines whether the network information 51 acquired by the network information acquisition unit 22 has changed by a certain threshold compared to previously acquired network information 51. If the change exceeds the threshold, the transfer schedule calculation unit 25 recalculates the transfer schedule 52 and delivers the information.
[0060] If the system adopts the case where data transfer occurs in violation of the period specified in the transfer schedule 52, the notification unit 134 of each transfer device 10 notifies the network management device 20 that data transfer has occurred in violation of the period specified in the transfer schedule 52. When this notification is received, the transfer schedule calculation unit 25 recalculates the transfer schedule 52 and distributes it. Here, data transfer occurring in violation of the transfer schedule 52 means that a communication stream for which transfer is not permitted at the current time is stored in the queue of the transfer device 10, or that the gate closes while the communication stream is being output from the transfer device 10 (when the time for which output is not permitted has arrived). The degree of this violation may be determined by considering the number of violations and the number of bytes exceeding the time slot of the violation frame, etc.
[0061] By adopting the approach of using the following criteria: when there is a change in network information 51 exceeding a certain threshold, or when data transfer occurs in violation of the period specified in the transfer schedule 52, it is possible to reduce the processing load on the network management device 20 and the transfer device 10, and to reduce the network bandwidth used.
[0062] In addition, the term "part" in the above explanation may be replaced with "circuit," "process," "procedure," "processing," or "processing circuit."
[0063] The embodiments and variations of this disclosure have been described above. Some of these embodiments and variations may be implemented in combination. Alternatively, some or all of them may be implemented in part. However, this disclosure is not limited to the embodiments and variations described above, and various modifications are possible as needed. [Explanation of symbols]
[0064] 100 Network system, 10 Transfer device, 11 Communication port, 12 Timestamp acquisition unit, 13 Communication control unit, 14 Setting unit, 131 Transfer unit, 132 Time synchronization unit, 133 In-device delay calculation unit, 134 Notification unit, 1321 Propagation delay calculation unit, 1322 Synchronization time calculation unit, 1323 Time update width calculation unit, 101 Processor, 102 Memory, 103 Data communication hardware, 104 Bus, 20 Network management device, 21 Communication port, 22 Network information acquisition unit, 23 Synchronization error estimation unit, 24 Setting unit, 25 Transfer schedule calculation unit, 26 Transfer schedule distribution unit, 201 Processor, 202 Memory, 203 Data communication hardware, 204 Bus, 30 Time distribution device, 40 Terminal, 51 Network information, 52 Transfer schedule, 90 Transmission path.
Claims
1. A network information acquisition unit acquires network information indicating the time update width, which is the difference between the synchronization time and the internal time managed by the transfer device before time synchronization by the synchronization time, from a transfer device that performs time synchronization by calculating the synchronization time based on a communication frame for time synchronization. A transfer schedule calculation unit calculates a transfer schedule that determines the period during which the transfer device performs data transfer for each priority of communication frames, using the time update width indicated by the network information acquired by the network information acquisition unit. A schedule distribution unit distributes the transfer schedule calculated by the transfer schedule calculation unit to the transfer device. A network management device equipped with the following features.
2. The aforementioned network management device further, Synchronization error estimation unit estimates the statistical value of the time update width indicated by the network information during the reference period acquired by the network information acquisition unit as the synchronization error. Equipped with, The transfer schedule calculation unit calculates the transfer schedule using the synchronization error estimated by the synchronization error estimation unit. The network management device according to claim 1.
3. The transfer schedule calculation unit calculates the transfer schedule such that it extends the period allocated to the transfer device for each priority of communication frames by the amount of the synchronization error. The network management device according to claim 2.
4. The network information indicates the in-device delay for the transfer device. The transfer schedule calculation unit calculates the transfer schedule using the internal delay of the device. The network management device according to claim 1.
5. The transfer schedule calculation unit recalculates the transfer schedule at at least one of the following times: at periodically determined timings, when there is a change in the network information exceeding a certain threshold, or when data transfer occurs in violation of the period specified in the transfer schedule. The network management device according to claim 1.
6. This is a transfer device that performs data transfer according to the period allocated by the transfer schedule distributed from the network management device. A synchronization time calculation unit that calculates the synchronization time based on a communication frame for time synchronization, A time update width calculation unit calculates the time update width, which is the difference between the synchronization time calculated by the synchronization time calculation unit and the internal time managed by the device. A notification unit that notifies the network management device of the time update width calculated by the update width calculation unit. A transporter device is provided.
7. The transfer device further, Synchronization error estimation unit estimates the statistical value of the time update range during the reference period as the synchronization error. Equipped with, The notification unit notifies the network management device of the synchronization error estimated by the synchronization error estimation unit. The transfer device according to claim 6.
8. The notification unit notifies the network management device of the internal delay of the transfer device. The transfer device according to claim 6.
9. The notification unit notifies the network management device that data transfer has been performed in violation of the period specified in the transfer schedule. The transfer device according to claim 6.
10. The network management device calculates the transfer schedule based on the time update interval notified by the notification unit. The transfer device according to claim 6.
11. A network system comprising a transfer device and a network management device, The transfer device, A synchronization time calculation unit that calculates the synchronization time based on a communication frame for time synchronization, A time update width calculation unit calculates the time update width, which is the difference between the synchronization time calculated by the synchronization time calculation unit and the internal time managed by the device. A notification unit that notifies the network management device of the time update width calculated by the update width calculation unit. Equipped with, The aforementioned network management device is A transfer schedule calculation unit calculates a transfer schedule that determines the period during which the transfer device performs data transfer for each priority of communication frames, using the time update width notified by the notification unit. A schedule distribution unit distributes the transfer schedule calculated by the transfer schedule calculation unit to the transfer device. A network system equipped with these features.
12. The transfer device calculates the synchronization time based on the communication frame for time synchronization. The transfer device calculates the time update width, which is the difference between the synchronization time and the internal time managed by the device. The transfer device notifies the network management device of the time update interval, The network management device calculates a transfer schedule that defines the period during which the transfer device performs data transfer for each priority of communication frames, using the notified time update interval. The network management device distributes the transfer schedule to the transfer device. How to set up a transfer schedule.
Citation Information
Patent Citations
Webcast method and system for time-synchronizing multiple independent media streams
JP2004525545A
Network system and communication device
JP2013106216A
Time synchronization client and protection system
JP2017022526A
Control system, setting method and network controller
JP2022136722A