Method for handling a loop within a communication network, and communication system
Patent Information
- Application Number
- PCT/EP2026/054695
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-03-25
- Filing Date
- 2026-02-20
- Publication Date
- 2026-10-01
Smart Images
Figure EP2026054695_01102026_PF_FP_ABST
Abstract
Description
[0001] 202502358 Foreign version
[0002] 1
[0003] Description
[0004] Methods for handling a loop within a communication network and communication system
[0005] The present invention relates to a method for treating a loop within a communication network and a communication system suitable for carrying out the method, in particular for an industrial automation system.
[0006] An industrial automation system typically comprises a large number of automation devices interconnected via an industrial communication network and serves to control or regulate plants, machines, or equipment within the context of manufacturing or process automation. Due to time-critical conditions in industrial automation systems, real-time communication protocols such as PROFINET, PROFIBUS, Real-Time Ethernet, or Time-Sensitive Networking (TSN) are predominantly used for communication between automation devices.
[0007] Due to their use for often highly diverse applications, problems can arise in Ethernet-based communication networks, for example, when network resources for transmitting data streams or data frames with real-time requirements are competing for the transmission of data frames with large payloads without specific quality of service requirements. This can lead to data streams or data frames with real-time requirements not being transmitted according to the requested or required quality of service.
[0008] Numerous redundancy protocols have been developed, particularly for industrial communication systems, to compensate for line or component failures or malfunctions. It is important to distinguish between interruptible redundancy methods, such as MRP (Media Redundancy Protocol) or RSTP / MSTP (Rapid / Multiple Spanning Tree Protocol), which switch from primary to backup paths during reconfiguration, and seamless redundancy methods. Seamless redundancy methods, such as HSR (High-availability Seamless Redundancy), PRP (Parallel Redundancy Protocol), or MRPD (Media Redundancy for Planned Duplication), generally do not require reconfiguration of the communication system, thus avoiding even brief communication interruptions.
[0009] 2
[0010] EP 2 838220 B1 relates to a method for redundant message transmission in a communication network with an arbitrarily meshed network topology. Several communication devices each comprise a first and a second transmitting and receiving unit. Each transmitting and receiving unit is assigned a signal processing unit, which duplicates messages to be sent redundantly by the communication device and transmits them to the receiving units. Messages redundantly received by the transmitting and receiving units at the communication device are detected by the receiving units. For a communication link that is at least partially redundant between two network nodes, independent paths are determined within the communication network. Messages to be transmitted redundantly are exchanged within the communication network according to forwarding rules corresponding to the determined paths.
[0011] From EP 4425875 A1, a method for highly available data transmission within a communication system is known. The system comprises a first and a second network segment, which are redundantly connected to each other via at least one first and a second coupling link. Within the first network segment, messages with user data are transmitted according to a spanning tree protocol, and messages with topology information are forwarded, also via the coupling links and within the second network segment. In a fault-free state, user data transmission between the two network segments occurs only via a first coupling device of the first network segment and the first coupling link. If no topology information is received via the second coupling link, a coupling port of a second coupling device is switched to an edge port mode.Devices in the second network segment delete their respective MAC address tables upon receiving topology information forwarded via the coupling port of the second coupling device.
[0012] The earlier international patent application PCT / EP2025 / 081197 describes a method for transmitting datagrams according to a redundancy protocol, in which the datagrams are initially transmitted according to a first redundancy protocol used for redundant transmission of the datagrams within a first network segment. Between a first endpoint and a second endpoint, the datagrams are transmitted via a second network segment using a second redundancy protocol. The two endpoints connect the first and second network segments. Upon detection of a topology change in the first network segment by a detecting network device, the detecting network device sends [202502358 Foreign version]
[0013] 3
[0014] A topology change notification is sent to other network devices in the first network segment. This causes the network devices in the first network segment to update their respective source address tables. Upon receiving the topology change notification at a first endpoint device encompassing the first endpoint, the first endpoint device deletes its source address table immediately or after waiting a predetermined initial duration and begins learning source addresses from other network devices. Upon receiving the topology change notification at a second endpoint device encompassing the second endpoint, the second endpoint device waits a predetermined second duration before deleting its source address table and beginning to learn source addresses from other network devices. The second duration is longer than the first.
[0015] US 7911 938 B2 discloses a method for preventing network loops that addresses a discrepancy between the control plane and the data plane. In some networks, control packets and data packets can be handled differently, which can cause a disturbance in the control plane without affecting the data plane. To prevent a port from being erroneously unblocked when a control loop is no longer detected due to a control plane disturbance, but a data loop still exists, a special data packet is sent. This special data packet is treated as a normal data packet by all other network devices and is processed in a special way only by the sending device. If the special data packet returns to the sending device within a timeout period, it is determined that the data loop still exists, and the port remains blocked. Otherwise, the port is unblocked.This method is solely for passive verification by the sending device itself and does not include active notification of other network devices about topology or status changes.
[0016] At the data link layer, or in Layer 2 networks, loop detection protocols, such as the Simple Loop Prevention Protocol (SLPP), are used to detect and prevent network loops. For this purpose, a selected network device within a network regularly sends test messages via other network devices, and the selected device monitors their receipt. If a previously sent test message is received again at the selected network device, a loop exists within the network, which is then detected by the selected network device. In response to a detected network loop, a transmit or receive interface on the selected network device that was used to send the test messages or where the network loop was detected can be administratively shut down. In particular, a transmit or receive interface can be... [202502358 Foreign version]
[0017] 4
[0018] The receive port on the selected network device is placed in a blocked state, in which any further potentially looping data traffic is terminated. When the blocked state is reset, for example after a pre-configured timeout, the respective transmit or receive interface can be administratively reactivated.
[0019] Current loop detection protocols lack the means to actively respond to or signal state changes triggered by the loop detection protocol itself. In such cases, network reconfiguration typically occurs only after a regular update of the MAC source address tables of the network devices. It is common practice to perform this update every 300 seconds. For networks with real-time requirements, this is a relatively long time. Further problems can arise if a redundancy protocol is used alongside the loop detection protocol within the network.
[0020] The present invention is based on the objective of providing a method for dealing with a loop within a communication network that enables a fast and efficient response to loop formation or link interruption within a network, particularly with real-time requirements, and of providing a suitable implementation for carrying out the method.
[0021] This problem is solved according to the invention by a method with the features specified in claim 1 and by a communication system with the features specified in claim 10. Advantageous embodiments of the present invention are specified in the dependent claims.
[0022] According to the inventive method for handling a loop within a communication network, a selected communication device sends diagnostic messages to detect a link interruption within the communication network. The selected communication device comprises at least two ports, one of which is blocked from forwarding messages containing user data in an uninterrupted state of the communication network (no link interruption) to prevent a loop within the communication network. If a previously sent diagnostic message from the selected communication device is not received, the selected communication device detects a link interruption and sends a message about a link status change to other communication devices within the communication network that forward messages. Similarly, a detected [202502358 foreign version]
[0023] 5
[0024] A loop, for example after a previously broken link is reconnected, is signaled by a message about a link status change and handled analogously to the following explanations. Such a loop can be detected in particular by the repeated receipt of diagnostic messages.
[0025] The communication network can, for example, have a ring or mesh topology. Advantageously, the link status change message is sent via the same port of the selected communication device through which the diagnostic messages are sent or on which the link interruption was detected by the failure to receive the previously sent diagnostic message. According to a preferred embodiment of the present invention, the selected communication device cyclically sends data frames containing a current system time. In this case, the link status change message can be sent as an unscheduled data frame with a predefinable maximum value for the system time.
[0026] The additional communication devices each comprise a source address table and, according to the invention, delete this table after receiving the link status change message, thereby relearning the source address tables before the expiration of a validity period for stored address entries. The selected communication device only switches the blocked port to a forwarding state (messages with payload data are forwarded) after the link status change message has been sent. Preferably, the additional communication devices ignore further link status change messages arriving within a predetermined interval, and the respective source address table is not deleted within this predetermined interval.
[0027] The present invention enables very short recovery or reconfiguration times of significantly less than 1 second in the communication network after topology or status changes, compared to the standard aging of stored MAC source addresses (300 seconds). Thus, the inventive method for handling a loop within a communication network provides an excellent basis for implementing redundancy methods at the data link layer or for the redundant coupling of subnetworks.
[0028] According to a particularly preferred embodiment of the present invention, the link status change notification is converted into a native topology change notification of a selected protocol for highly available networks, if the selected 202502358 foreign version
[0029] 6
[0030] The protocol for high-availability networks is used by at least one communication device within the communication network. The selected protocol can be, for example, High-Availability Seamless Redundancy (HSR), Media Redundancy Protocol (MRP), Parallel Redundancy Protocol (PRP), or a spanning tree protocol, in particular Rapid Spanning Tree Protocol (RSTP). Advantageously, the link status change message is ignored if the respective communication device uses the selected high-availability network protocol. Specifically, preferably only the converted native topology change message is processed if the respective communication device uses the selected high-availability network protocol.According to a further advantageous embodiment, the message about the link status change is only forwarded via ports of the respective communication device that, according to the selected protocol for high-availability networks, are in a state in which messages with user data are being forwarded. Overall, with these further developments, the method according to the invention can be used efficiently and reliably in combination with existing redundancy protocols.
[0031] The communication system according to the invention is designed for carrying out a method as described above and comprises a selected communication device for sending diagnostic messages to detect a link interruption within a communication network, as well as further communication devices for forwarding messages within the communication network. The selected communication device comprises at least two ports, one of which is blocked in an uninterrupted state of the communication network to prevent the forwarding of messages containing user data, in order to avoid a loop within the communication network.
[0032] According to the invention, the selected communication device is configured to detect a link interruption if a previously sent diagnostic message from the selected communication device is not received, and to send a message about a link status change to the other communication devices. The other communication devices each comprise a source address table and are configured to delete their respective source address tables after receiving the message about the link status change, thereby relearning the source address tables before the expiration of a validity period for stored address entries. Furthermore, the selected communication device is also configured to switch the blocked port to a forwarding state only after the message about the link status change has been sent. 202502358 Foreign version
[0033] 7
[0034] The present invention further relates to a selected communication device for the above communication system. The selected communication device comprises at least two ports, one of which is blocked in an uninterrupted state of the communication network to prevent the forwarding of messages containing user data in order to avoid a loop within the communication network, and is configured to detect a link interruption if a diagnostic message previously sent by the selected communication device is not received and to send a message about a link status change to the other communication devices.Furthermore, the selected communication device is also configured to switch the blocked port to a forwarding state only after sending the message about the link status change, especially after the other communication devices have deleted their respective source address tables after receiving the message about the detected loop.
[0035] The present invention is explained in more detail below using an exemplary embodiment with reference to the drawing. It shows
[0036] Figure 1 shows a communication network with several communication devices connected in a ring topology, one of which is selected to detect a link break or a loop within the communication network.
[0037] Figure 2 shows a communication network with several communication devices connected in a mesh topology, several of which use a redundancy protocol and one selected communication device is provided for detecting a link break or a loop within the communication network.
[0038] The communication network shown in Figure 1 comprises several communication devices 101-103, in particular switches, connected within a ring topology, each having two ring ports P11-P12, P21-P22, P31-P32. In the present embodiment, automation devices, such as a programmable logic controller 120 or an operator and monitoring station 130, are connected to non-ring ports P23, P33 of the communication devices 102, 103. Each of the communication devices 101-103 is configured to transmit data frames within the communication network to an adjacent communication device via links or point-to-point connections.
[0039] 8
[0040] to forward the data. For this purpose, each of the communication devices 101-103 has a forwarding database or source address table 111, 121, 131, in which the source addresses of received data frames are stored. In principle, the communication devices 101-103 can be assigned to a network controller, which controls the forwarding functions of the assigned communication devices.
[0041] A typical programmable logic controller (PLC) comprises a communication module, a central processing unit (CPU), and at least one internal or peripheral input / output (I / O) module. A communication module allows the PLC to be connected to a switch, router, or fieldbus system. I / O modules are used to exchange manipulated variables and measured values between the PLC and machines or devices controlled by the PLC. Interfaces between the PLC and machines or devices can be analog or digital. The CPU of a PLC is specifically designed and configured to determine manipulated variables from measured values and to generate control commands.Preferably, the above-mentioned components of a programmable logic controller are interconnected via a backplane bus system.
[0042] The communication module of a programmable logic controller (PLC) comprises a transmit and a receive unit, implemented using a PHY circuit, and a functional unit, implemented using a MAC circuit, which controls access to a communication medium. Specifically, the communication module accesses a physical communication network adapter via an interface driver as hardware abstraction elements.
[0043] The operator and monitoring station 120 is used to visualize process control data, manipulated variables, and measured values, which are processed or collected by programmable logic controllers (PLCs), I / O modules, and sensors. In particular, the operator and monitoring station 120 in this embodiment comprises at least a graphical user interface, an input device, a processor, and a communication module.
[0044] To handle loops within the communication network, a selected communication device sends 101 diagnostic messages to detect a link interruption within the communication network. In the present embodiment202502358 Foreign version
[0045] 9
[0046] The selected communication device comprises two ports, P11 and P12, one of which, P11, is blocked in an uninterrupted state of the communication network (i.e., without link interruption within the ring topology shown in Figure 1) to prevent the forwarding of messages containing user data, thus avoiding a network loop. This serves, in particular, to prevent the uncontrolled circulation of messages within the communication network as a result of a loop.
[0047] In particular, diagnostic messages 10 are forwarded by the two other communication devices 102 and 103 in the ring topology during uninterrupted communication. If a diagnostic message 10 previously sent by the selected communication device 101 is not received, the selected communication device 101 detects a link interruption and sends a message 11 about a link status change to the two other communication devices 102 and 103 within the communication network. Similarly, a detected loop, for example, after a previously interrupted link is reconnected, can also be signaled by means of a message 11 about a link status change and handled analogously to the following explanations. A loop can be detected, in particular, by the repeated receipt of diagnostic messages 10.
[0048] In the present embodiment, a link interruption 100 between port P12 of the selected communication device 101 and port P32 of the further communication device 103, to which the programmable logic controller 130 is connected, is considered as a fault scenario. During the uninterrupted state, the link between the blocked port P11 of the selected communication device 101 and port P21 of the further communication device 102, to which the operator and monitoring station 120 is connected, was inactive. Before the selected communication device 101 switches the blocked port P11 to a forwarding state, the selected communication device 101 sends a message 11 about a link status change within the communication network.
[0049] Message 11 regarding the link status change can generally be sent via the same port of the selected communication device 101 through which the diagnostic messages 10 are sent, or on which the link interruption 100 was detected due to the failure to receive the previously sent diagnostic message 10. The selected communication device 101 can, for example, cyclically send data frames, in particular the diagnostic messages 10, with a current system time reference.
[0050] 10
[0051] Message 11, reporting a link status change, is sent as an unscheduled data frame with a predefined maximum value for the system time. This maximum value can be decremented, for example, by repeatedly transmitting Message 11 link status changes.
[0052] Upon receiving message 11 about the link status change, the two other communication devices 102 and 103 each convert the message 11 into an internal flush command 12 and 13, thereby essentially deleting their source address tables 121 and 131 immediately. In this way, the source address tables 121 and 131 are relearned, particularly before the expiration of a validity period (time to live) of stored address entries. Only after sending message 11 about the link status change does the selected communication device 101 switch the previously blocked port P11 to the forwarding state. Preferably, the two other communication devices 102 and 103 ignore further messages 11 about link status changes arriving within a predefined interval (grace period), and no deletion of the respective source address tables 121 and 131 occurs within this predefined interval.
[0053] According to the embodiment of a communication network shown in Figure 2, with communication devices 201-209 connected in a mesh topology, communication devices 204-206 process only messages about link status changes, while communication devices 201-203 and 207-208 process both messages about link status changes and apply a selected protocol for high-availability networks or a redundancy protocol. Communication devices 201-207 form a first mesh, while communication devices 203-209 form a second mesh. In principle, High-Availability Seamless Redundancy (HSR), Media Redundancy Protocol (MRP), Parallel Redundancy Protocol (PRP), or a spanning tree protocol, in particular Rapid Spanning Tree Protocol (RSTP), can be used as the redundancy protocol.As in the embodiment shown in Figure 1, each of the communication devices 201-209 is configured to forward data frames within the communication network via links to a neighboring communication device.
[0054] If a redundancy protocol is used by at least one communication device within the communication network, the previously described link status change message is advantageously converted into a native topology change message of the respective redundancy protocol, which is sent or forwarded either in addition to or instead of the link status message. Link status message - 202502358 (foreign version)
[0055] 11
[0056] Status changes are preferably ignored if the respective communication device uses one of the aforementioned redundancy protocols. Thus, only the converted native topology change message is processed if the respective communication device uses a redundancy protocol. Furthermore, communication devices using a redundancy protocol only forward link status change messages via ports that, according to the respective redundancy protocol, are in a state where messages containing payload data are being forwarded.
[0057] In the present embodiment, communication devices 201-203 and 207-208 use MRP as the redundancy protocol, with communication device 201 operating as the MRP redundancy manager, while communication devices 202-203 and 207-208 have an MRP client role. Of the communication devices 204-206, which only process messages about link status changes, communication device 204 has a role as the selected communication device or sender of messages about link status changes, while communication devices 205 and 206 merely forward messages about link status changes.
[0058] In a fault-free topology of the communication network shown in Figure 2, i.e. without link interruption, the communication device 201 with role as MRP redundancy manager and the communication device 204 with role as sender of messages about link status changes deactivate the links 210, 220 terminating at them, by blocking the corresponding ports against the forwarding of messages with payload data.
[0059] When communication device 204 detects a topology change, it sends Loop Detection State Change Frames 21, causing communication devices 204-206 to delete and retrain their respective forwarding databases. In the case of a topology change detected by communication device 201, this MRP sends Topology Change Frames 22, causing communication devices 202-203 and 207-208 to delete and retrain their respective forwarding databases. If an existing or subsequently integrated MRP communication device is unable to process the Loop Detection State Change Frames 21, MRP communication devices 202-203 and 207-208 advantageously forward the Loop Detection State Change Frames 21 to the MRP Redundancy Manager 201 via their ring ports.As soon as the MRP Redundancy Manager 201 receives a Loop Detection State Change Frame 21, it sends a native MRP Topology Change Frame 22 in response to the communication devices 202-203, 207-208 with MRP Client role. 202502358 Foreign version.
[0060] 12
[0061] The above conversion allows any MRP communication device to be integrated into the communication network shown in Figure 2, even if it does not support processing of Loop Detection State Change Frames 21. Preferably, all MRP communication devices suspend processing of Loop Detection State Change Frames 21 to avoid a duplicate response to topology changes on Loop Detection State Change Frames 21 and MRP Topology Change Frames 22.
Claims
202502358 Foreign version 13 Patent claims 1. Method for handling a loop within a communication network, wherein a selected communication device (101, 204) sends diagnostic messages (10) to detect a link break within the communication network, the selected communication device comprising at least 2 ports (P11, P12), one of which (P11) is blocked against forwarding messages with payload data in an uninterrupted state of the communication network to prevent a loop within the communication network. - the selected communication device (101, 204) detects a link interruption (100) if a diagnostic message (10) previously sent by the selected communication device is not received and sends a message (11, 21) about a link status change to further communication devices (102-103, 201-203, 205-209) within the communication network that forward messages, - the other communication devices each include a source address table (121, 131) and delete this after receiving the message about the link status change, whereby the source address tables are relearned before the expiry of a validity period of stored address entries, - the selected communication device (101, 204) switches the blocked port to a forwarding state only after sending the message about the link status change.
2. Method according to claim 1, where the link status change message (11, 21) is converted into a native topology change message (22) of a selected high-availability network protocol if the selected high-availability network protocol is used by at least one communication device (201-203, 207-209) within the communication network.
3. Method according to claim 2, where the message (11, 21) about the link status change is ignored if the selected protocol for highly available networks is applied by the respective communication device (201-203, 207-209).
4. Procedure according to claim 3,202502358 Foreign version 14 where only the converted native topology change message (22) is processed if the selected protocol for high-availability networks is applied by the respective communication device (201-203, 207-209).
5. Method according to one of claims 3 or 4, where the message (11 , 21) about the link status change is only forwarded via ports of the respective communication device (201-203, 207-209) which, according to the selected protocol for high-availability networks, are in a state in which messages with payload are forwarded.
6. Method according to any one of claims 2 to 5, where the selected protocol is High-availability Seamless Redundancy, HSR, Media Redundancy Protocol, MRP, Parallel Redundancy Protocol, PRP, or a spanning tree protocol.
7. Method according to any one of claims 1 to 6, where the communication network has a ring or mesh topology and where the message about the link status change (11, 21) is sent via the same port of the selected communication device, through which the diagnostic messages are sent or where the link interruption has been detected by failure to receive the previously sent diagnostic message.
8. Method according to any one of claims 1 to 7, where the selected communication device (11, 21) cyclically sends data frames with a current system time indication and where the message about the link status change is sent in the form of an unscheduled data frame with a predefinable maximum value for the system time indication.
9. Method according to any one of claims 1 to 8, where the other communication devices (102-103, 201-203, 205-209) ignore further messages about a link status change arriving within a specified interval and no deletion of the respective source address table takes place within the specified interval.
10. Communication system for carrying out a method according to one of claims 1 to 9, wherein the communication system 202502358 Foreign version 15 - a selected communication device (101, 204) for sending diagnostic messages (10) for detecting a link interruption within a communication network and further communication devices (102-103, 201-203, 205-209) for forwarding messages within the communication network, - wherein the selected communication device includes at least 2 ports (P11, P12), one of which (P11) is blocked in an uninterrupted state of the communication network to prevent a loop within the communication network from forwarding messages with payload data, - the selected communication device (101, 204) is configured to detect a link interruption (100) and send a message (11, 21) about a link status change to the other communication devices (102-103, 201-203, 205-209) if a diagnostic message (10) previously sent by the selected communication device is not received, - the other communication devices (102-103, 201-203, 205-209) each include a source address table (121, 131) and are configured to delete their respective source address table after receiving the message about the link status change, whereby the source address tables are relearned before the expiry of a validity period of stored address entries, - the selected communication device (101, 204) is further configured to switch the blocked port to a forwarding state only after sending the message about the link status change.