Network convergence method

Abstract

A network fusion method is disclosed to fuse an EtherCAT (Ethernet for Control Automation Technology) network into an Ethernet network, and includes the following steps: connecting an EtherCAT device to a switch, wherein the switch has at least one port (including a normal port, an EtherCAT endpoint port, and an EtherCAT edge port) used to transmit at least one packet (including a normal packet and an EtherCAT packet); using port rules to specify these ports, and using a jump port setting to determine the destination port to which an EtherCAT packet is to be transmitted; when there is a transmission rate difference between an EtherCAT endpoint port and an EtherCAT edge port in transmission, compensating for the transmission rate difference to enable the switch to support cut-through; and when these ports transmit the normal packet and the EtherCAT packet, preferentially transmitting the EtherCAT packet.

Description

Technical Field

[0001] This application relates to a network convergence method, and more particularly to a network convergence method for integrating an EtherCAT network into an Ethernet network. Background Technology

[0002] Ethernet for Control Automation Technology EtherCAT (Electronic Automation Technology) is an open architecture and a real-time industrial Ethernet technology. The EtherCAT protocol, disclosed in the IEC standard IEC 61158, is applicable to the real-time hardware and software requirements in automation, testing, measurement, and other applications. EtherCAT is widely used in industrial automation, such as motion control and robotics, where real-time performance requirements are typically extremely high. In other words, EtherCAT networks are characterized by very low latency. Although EtherCAT networks are based on Ethernet technology, they are generally closed systems. That is, while EtherCAT networks share the lowest physical layer with Ethernet, their upper layers (including the data link layer) are closed. (data link layer), network layer, transport layer The network architecture (such as the Transport Layer) is custom-designed. Therefore, EtherCAT networks are typically built independently using dedicated hardware that supports EtherCAT, and are not combined with Ethernet networks. Summary of the Invention

[0003] The purpose of this application is to provide a network integration method for integrating an EtherCAT network into an Ethernet network, enabling the simultaneous transmission of ordinary packets and EtherCAT packets, while also meeting the low-latency real-time requirements of the EtherCAT network.

[0004] One aspect of this application is to provide a network convergence method for integrating Ethernet for Control Automation technology. Technology, EtherCAT network convergence into Ethernet network, this network convergence method includes the following steps: connecting EtherCAT devices to a switch. (switch), wherein the switch has at least one port for transmitting at least one packet. (port), and these ports include ordinary ports and EtherCAT endpoint ports. (Endpoint port) and EtherCAT edge port, wherein these packets contain both regular packets and EtherCAT packets; using port rules (port The rule specifies the normal ports, EtherCAT endpoint ports, and EtherCAT edge ports, and uses the next-hop port setting to determine the destination port to which the EtherCAT packet is to be transmitted; when there is a transmission rate difference between the EtherCAT endpoint port and the EtherCAT edge port during transmission, the transmission rate difference is compensated to enable the switch to support cut-through switching; and when these ports transmit both normal packets and EtherCAT packets, EtherCAT packets are transmitted first.

[0005] In some embodiments, the port rules specify that ordinary ports are used only for transmitting ordinary packets and not for transmitting EtherCAT packets.

[0006] In some embodiments, the port rules specify that the EtherCAT endpoint port is used only for transmitting EtherCAT packets and not for transmitting ordinary packets.

[0007] In some embodiments, the port rules specify that EtherCAT edge ports are used to transmit both regular packets and EtherCAT packets.

[0008] In some embodiments, the transmission rate of the EtherCAT edge port is greater than... Transmission rate of EtherCAT endpoint ports.

[0009] In some embodiments, the network fusion method further includes the following step: compensating for the transmission rate difference between the EtherCAT endpoint port and the EtherCAT edge port through packet replication operations.

[0010] In some embodiments, the replication ratio of the packet replication operation is determined based on the difference in transmission rates.

[0011] In some embodiments, the network fusion method further includes the following step: compensating for the transmission rate difference between EtherCAT endpoint ports and EtherCAT edge ports through packet sampling operations.

[0012] In some embodiments, the sampling ratio of the packet sampling operation is determined based on the transmission rate difference.

[0013] In some embodiments, the network fusion method further includes the following steps: determining whether a packet in the EtherCAT network is an EtherCAT packet by the packet protocol number. Attached Figure Description

[0014] To make the above and other objects, features, advantages and embodiments of this application more apparent and understandable, the accompanying drawings are described below: Figure 1 This is a schematic diagram of the EtherCAT network architecture according to an embodiment of this application; Figure 2 This is a flowchart of a network fusion method according to an embodiment of this application; Figure 3 A schematic diagram of each port in a switch according to an embodiment of this application; and Figure 4 This is a schematic diagram of a switch and EtherCAT hybrid network according to an embodiment of this application. Detailed Implementation

[0015] The embodiments of this application are discussed in detail below. It will be understood that the embodiments provide many applicable concepts that can be implemented in a wide variety of specific situations. The discussed... The disclosed embodiments are for illustrative purposes only and are not intended to limit the scope of this application.

[0016] Ethernet for Control is a common technology for automation. An EtherCAT network architecture implemented by Automation Technology (hereinafter referred to as EtherCAT) typically has one EtherCAT master device and multiple EtherCAT slave devices. Packets are sent from the EtherCAT master device and transmitted one by one through each node (i.e., EtherCAT slave device). When a packet passes through each node, each EtherCAT slave device adds its own individual information to the packet in transmission and then continues to transmit the processed packet to the next node. When the packet has been transmitted to the last node, the packet stream... On the return path of the packet flow (i.e., during the process of transmitting packets back to the EtherCAT master device), although the packets in transit will still pass through each node, the EtherCAT slave devices will no longer process the packets in transit.

[0017] Figure 1This is a schematic diagram of an EtherCAT network architecture 100 according to an embodiment of this application. Figure 1 In one illustrated embodiment, the EtherCAT network has an EtherCAT master device 110 and multiple EtherCAT slave devices, wherein the multiple EtherCAT slave devices include, but are not limited to, a driver 120, a sensor 130, an analog input / output (AIO) component 140, and a digital input / output (DIO) component 150. EtherCAT packets are sent from the EtherCAT master device 110 and transmitted sequentially through each EtherCAT slave device. Specifically, as a packet is transmitted sequentially through the driver 120, sensor 130, analog input / output component 140, and DIO component 150, the driver 120, sensor 130, analog input / output component 140, and DIO component 150 will sequentially add their individual messages to the transmitted EtherCAT packet. After each EtherCAT slave device has completed processing of the EtherCAT packet, the EtherCAT packet will be directly transmitted back to the EtherCAT master device. The driver 120, sensor 130, analog input / output component 140 and digital input / output component 150 will no longer process the EtherCAT packets in transmission on the return path of the EtherCAT packet stream.

[0018] The advantage of EtherCAT networks is their high real-time performance, but their disadvantage is their bandwidth. The bandwidth is limited. In typical EtherCAT network applications, the transmission rate is megabits per second (Mbit / s), while in other specialized applications, the transmission rate can support up to gigabits per second (Gbps). (Gbit / s). In some emerging applications (such as robotics and autonomous driving), multiple sensors, including cameras and radar, are required. Because the amount of sensor data collected is enormous, the total bandwidth required exceeds the bandwidth limit of the EtherCAT network. Furthermore, issues such as cabling limitations may also exist. Traditional EtherCAT networks cannot meet the needs of these emerging applications. Compared to EtherCAT networks, Ethernet's supported bandwidth has gradually increased over time, evolving from early supported transmission rates of 10Gbit / s, 25Gbit / s, and 50Gbit / s to the current supported transmission rate of 100Gbit / s. Therefore, if EtherCAT networks can be integrated into Ethernet networks, the needs of some emerging applications (such as higher bandwidth) could potentially be met.

[0019] However, directly connecting an EtherCAT network to an Ethernet network may present three problems, as described below.

[0020] Issue 1: The packet streams of ordinary packets (i.e., Ethernet packets) and EtherCAT packets cannot be effectively isolated, which may cause EtherCAT devices to receive or send non-EtherCAT packets, potentially leading to malfunctions in the EtherCAT devices.

[0021] Question 2: Switches typically have store-and-forward functionality. Between store-and-forward and cut-through packet transmission methods, generally speaking, store-and-forward transmission has too much latency, making it difficult to meet the requirements of EtherCAT networks. In contrast, while cut-through transmission has the advantage of lower latency, the transmission rate of EtherCAT endpoint ports is limited. (Typically 100 Mbit / s) If it cannot match the transmission rate of the EtherCAT edge port (which may be 10 Gbit / s), the switch cannot perform pass-through switching transmission.

[0022] Question 3: When ports in a switch transmit a mixture of EtherCAT packets and regular packets, EtherCAT packets cannot be guaranteed to be transmitted first, resulting in excessive delays.

[0023] Figure 2 This is a flowchart of a network fusion method 200 according to an embodiment of this application. The network fusion method 200 is used to fuse an EtherCAT network into an Ethernet network, and includes the following steps: Step S210 connects an EtherCAT device to a switch, wherein the switch has at least one port for transmitting at least one packet, and wherein these ports include ordinary ports, EtherCAT endpoint ports, and EtherCAT edge ports, and wherein these packets include ordinary packets and EtherCAT packets; Step S220 uses port rules to classify ordinary ports, EtherCAT endpoint ports and EtherCAT edge ports, and configure them using jump ports. (Next-hop port setting) determines the destination port to which the EtherCAT packet is intended to be transmitted; step S230 explains when the EtherCAT packet is being transmitted. When a transmission rate difference exists between the endpoint port and the EtherCAT edge port, the transmission rate difference is compensated to enable the switch to support cut-through switching; step S240 states that when each of these ports transmits both ordinary packets and EtherCAT packets, EtherCAT packets should be transmitted first. The following explains how steps S210-S240 of network fusion method 200 overcome the above three network fusion problems.

[0024] First, regarding the first problem (the inability to effectively isolate the packet streams of ordinary packets and EtherCAT packets, potentially causing malfunctions in EtherCAT devices), this application uses designed port rules to determine the packet transmission method for each port in the switch, achieving effective isolation between ordinary packets and EtherCAT packets. It should also be noted that the distinction between ordinary packets and EtherCAT packets can be made by identifying the packet protocol number within the packet. For example, when the packet protocol number in the packet is identified as... When the packet number is 0x88A4, it indicates that the packet is an EtherCAT packet. When the packet protocol number is not identified as 0x88A4, it indicates that the packet is a normal packet.

[0025] Figure 3 This is a schematic diagram of a switch 300 port according to an embodiment of this application. Figure 3 As shown, switch 300 has ports P0-P4, where ports P0 and P4 are EtherCAT edge ports (hereinafter referred to as EtherCAT edge ports), port P1 is a normal port, and ports P2 and P3 are EtherCAT endpoint ports (hereinafter referred to as EtherCAT endpoint ports). It should be noted that in an EtherCAT network, EtherCAT slave device 310 typically communicates with the two EtherCAT endpoint ports (i.e., ports P2 and P3) of switch 300; on the other hand, a personal computer (PC) typically communicates with the normal port P1 of switch 300. Figure 3 In one illustrated embodiment, this application uses designed port rules to regulate each port of switch 300 (including, but not limited to, EtherCAT edge ports P0 and P4, ordinary port P1, and EtherCAT endpoint ports P2 and P3). It should also be noted that... Figure 3 The document also indicates the common transmission rates for each port in the switch 300. For example, the transmission rate of EtherCAT endpoint ports P2 and P3 is generally 100 Mbit / s (expressed as 100 M), and the transmission rate of ordinary port P1 is generally 1000 Mbit / s (expressed as 1000 M). The transmission rate of EtherCAT edge ports P0 and P4 is typically 10 Gbit / s (based on 10 Gbit / s). (Indicated). Furthermore, in each embodiment, the number of switches 300 is not... Figure 3 The content shown is limited.

[0026] Specifically, the port rule specifies that the ordinary port P1 is only used for transmitting ordinary packets. (That is, no EtherCAT packets are transmitted); and the EtherCAT endpoint ports P2 and P3 are specified to be used only for transmitting EtherCAT packets (i.e., no ordinary packets are transmitted), and other packets other than EtherCAT packets are discarded; and the EtherCAT edge ports P0 and P4 are specified to be used for transmitting both ordinary packets and EtherCAT packets. In other words, the port rules specify that EtherCAT endpoint ports P2 and P3 and ordinary port P1 cannot transmit packets to each other, thus achieving effective isolation between EtherCAT endpoint ports P2 and P3 and ordinary port P1.

[0027] Furthermore, EtherCAT packets are transmitted between EtherCAT endpoint ports and between EtherCAT endpoint ports and EtherCAT edge ports according to the jump port settings. Specifically, after receiving an EtherCAT packet, EtherCAT endpoint ports P2 and P3 and EtherCAT edge ports P0 and P4, which can be used to transmit EtherCAT packets, can transmit the EtherCAT packet to the destination port according to the jump port settings. In addition, other packets can be transmitted between ordinary ports and between EtherCAT edge ports and ordinary ports according to the original port rules of the switch 300.

[0028] In detail, when the packets received by EtherCAT endpoint ports P2 and P3 are not EtherCAT packets (i.e., ordinary packets), EtherCAT endpoint ports P2 and P3 will discard the ordinary packets or transmit them according to other special rules designed by the switch 300; when the packets received by EtherCAT endpoint ports P2 and P3 are EtherCAT packets, EtherCAT endpoint ports P2 and P3 will transmit the EtherCAT packets according to the jump port settings.

[0029] When the packets received by EtherCAT edge ports P0 and P4 are not EtherCAT packets (i.e., ordinary packets), EtherCAT edge ports P0 and P4 will transmit these ordinary packets according to the normal rule designed by the switch 300; when the packets received by EtherCAT edge ports P0 and P4 are EtherCAT packets, EtherCAT edge ports P0 and P4 will transmit the EtherCAT packets according to the jump port settings.

[0030] When a packet received by ordinary port P1 is not an EtherCAT packet (i.e., a normal packet), ordinary port P1 will transmit the normal packet according to the normal rules designed by switch 300; when a packet received by ordinary port P1 is an EtherCAT packet, ordinary port P1 will discard the EtherCAT packet or transmit the EtherCAT packet according to other special rules designed by switch 300.

[0031] Next, regarding the second problem (the transmission rates of EtherCAT endpoint ports and EtherCAT edge ports are not equal, causing the switch to be unable to perform cut-through switching transmission), this application compensates for the transmission rate difference between the high-speed ports (e.g., EtherCAT edge ports) and low-speed ports (e.g., EtherCAT endpoint ports) of the switch, enabling the switch to support cut-through switching transmission mode, avoiding the use of store-and-forward transmission mode accompanied by high latency, and thus meeting the low latency requirement of EtherCAT.

[0032] Figure 4 This is a schematic diagram of an EtherCAT hybrid network 400 switch according to an embodiment of this application. By observation... Figure 4 As can be seen, the EtherCAT hybrid network 400 consists of one EtherCAT master device 420, two EtherCAT slave devices (driver 430 and sensor 440), and separately connected switches 410. Figure 3 What they have in common is that the common transmission rates of each port in each switch 410 are already indicated. Figure 4 Obviously, The transmission rates of the EtherCAT endpoint ports (100 Mbit / s) and the EtherCAT edge ports (10 Gbit / s) are not equal. To achieve pass-through switching, this inequality must be addressed (generally, the transmission rate of the EtherCAT edge ports is much higher than that of the EtherCAT endpoint ports). Therefore, this application proposes to compensate for the transmission rate difference between the high-speed and low-speed ports of switch 410 through packet copying and packet sampling operations.

[0033] Specifically, in Figure 4 In one illustrated embodiment, when a low-speed port B (e.g., an EtherCAT endpoint port) receives an EtherCAT packet transmitted from a high-speed port A (e.g., an EtherCAT edge port), a packet copying operation can be performed to copy the original EtherCAT packet at a certain copy ratio, generating a copied EtherCAT packet. This compensates for the transmission rate difference between the high-speed port A and the low-speed port B. It is worth noting that the copy ratio used in the packet copying operation is determined based on the transmission rate difference between the high-speed port A and the low-speed port B. For example, if the transmission rate of the high-speed port A is 10 times that of the low-speed port B, the length of the EtherCAT packet after the packet copying operation will be increased to 10 times the length of the original packet, thus compensating for the transmission rate difference between the high-speed and low-speed ports and meeting the conditions for pass-through switching transmission in the switch 410.

[0034] exist Figure 4 In another embodiment illustrated, when high-speed port A (EtherCAT edge port) receives an EtherCAT packet transmitted from low-speed port B (EtherCAT endpoint port), a packet sampling operation can be performed to sample the copied EtherCAT packet at a certain sampling ratio to restore the original EtherCAT packet (i.e., to recover the valid information from the copied EtherCAT packet; in some embodiments, the packet may also include special information such as the source port transmission rate and packet type), thereby compensating for the transmission rate difference between high-speed port A and low-speed port B. It is worth noting that the sampling ratio used in the packet sampling operation is determined based on the transmission rate difference between high-speed port A and low-speed port B. For example, if the transmission rate of low-speed port B is 0.1 times that of high-speed port A, the length of the EtherCAT packet after the packet sampling operation will be truncated to 0.1 times the original packet length. This is to compensate for the difference in transmission rate between high-speed and low-speed ports, thus meeting the conditions for the switch 410 to perform pass-through switching transmission.

[0035] Next, addressing issue three (when EtherCAT packets and regular packets are transmitted together on a switch port, EtherCAT packets cannot be guaranteed to be transmitted first, leading to excessive latency), this application employs a preemptive packet transmission method to ensure that EtherCAT packets are transmitted first. Specifically, after the switch determines the destination port to which the EtherCAT packet is to be transmitted using its jump port settings, it must interrupt the currently transmitting regular packets and prioritize the transmission of EtherCAT packets. The interrupted regular packets will only resume transmission after the EtherCAT packet transmission is complete. In other words, if the switch waits for the regular packets to complete before transmitting the EtherCAT packets, excessive latency will occur, failing to meet the real-time requirements of EtherCAT networks. Therefore, the switch must also support preemptive transmission to minimize latency.

[0036] In summary, the network fusion method of this application establishes a transmission link for a virtual EtherCAT network by designing port rules for the switch, compensating for the transmission rate difference between ports, and supporting preemptive packet transmission on the ports. This integrates the EtherCAT network into Ethernet, enabling the simultaneous transmission of ordinary packets and EtherCAT packets, while also meeting the low-latency real-time requirements of the EtherCAT network.

[0037] Although this application has been disclosed above with reference to embodiments, it is not intended to limit the scope of this application. Anyone skilled in the art can make various changes, substitutions and modifications without departing from the spirit and scope of this application. Therefore, the scope of protection of this application shall be determined by the appended claims. [Symbol Explanation] 100: EtherCAT Network Architecture 200: Network Convergence Methods 300, 410: Switches 310: EtherCAT from device 400: Switch EtherCAT Hybrid Network 420: EtherCAT Master Device 430: Driver 440: Sensor A: High-speed port B: Low-speed port P0, P4: EtherCAT endpoint ports P1: Normal Port P2, P3: EtherCAT edge ports PC: Personal Computer S210, S220, S230, S240: Steps

Claims

1. A network integration method for integrating an Ethernet network control automation technology EtherCAT network into an Ethernet network, characterized in that, This network fusion method includes: Connect an EtherCAT device to a switch, wherein the switch has at least one port for transmitting at least one packet, and wherein the at least one port includes a normal port, an EtherCAT endpoint port and an EtherCAT edge port, and wherein the at least one packet includes a normal packet and an EtherCAT packet. The EtherCAT endpoint port, EtherCAT edge port, and EtherCAT endpoint port are defined using a port rule, and a jump port setting is used to determine the destination port to which the EtherCAT packet is to be transmitted. When there is a transmission rate difference between the EtherCAT endpoint port and the EtherCAT edge port during transmission, the transmission rate difference is compensated to enable the switch to support pass-through switching. as well as When at least one port transmits both the normal packet and the EtherCAT packet, the EtherCAT packet is transmitted first.

2. The network fusion method as described in claim 1, wherein the port rule specifies that the ordinary port is used only for transmitting the ordinary packet and not for transmitting the EtherCAT packet.

3. The network fusion method as described in claim 1, wherein the port rule specifies that the EtherCAT endpoint port is used only for transmitting the EtherCAT packet and not for transmitting the ordinary packet.

4. The network convergence method as described in claim 1, wherein the port rule specifies the EtherCAT edge port for transmitting the ordinary packet and the EtherCAT packet.

5. The network fusion method as described in claim 1, wherein the transmission rate of the EtherCAT edge port is greater than the transmission rate of the EtherCAT endpoint port.

6. The network fusion method as described in claim 1, further comprising: The difference in transmission rate between the EtherCAT endpoint port and the EtherCAT edge port is compensated by a packet replication operation.

7. The network fusion method of claim 6, wherein a replication ratio of the packet replication operation is determined based on the transmission rate difference.

8. The network fusion method as described in claim 1, further comprising: The transmission rate difference between the EtherCAT endpoint port and the EtherCAT edge port is compensated by a packet sampling operation.

9. The network fusion method as described in claim 8, wherein a sampling ratio of the packet sampling operation is determined based on the transmission rate difference.

10. The network fusion method as described in claim 1, characterized in that, Also includes: The packet protocol number is used to determine whether a packet in the EtherCAT network is an EtherCAT packet.

Images