A method for collecting operation information of a time-sensitive network

Abstract

A time-sensitive network operation information collection method, the designated terminal device can periodically send a monitoring flow, traverse and collect the queue, address and link information and other data of each switch in the network, process and calculate the original data, update the routing of the monitoring flow, and feed back the operation information to the CNC configurator for adjustment. Realize the perception of TSN real-time operation information, improve the efficiency of network maintenance and fault perception, and reduce the additional overhead of monitoring flow on data plane.

Description

Technical Field

[0001] This application relates to the field of Time Sensitive Networks (TSN), specifically to a method for collecting operational information of a Time Sensitive Network. Background Technology

[0002] Network operational status information encompasses key data such as latency, routing, and link status. To maintain normal network operation, we need to extract robust and reliable real-time operational status information. Uncovering the intrinsic value of network-related information is of great guiding significance for maintaining normal network operation, expanding network functionality, and monitoring traffic. Currently, network operation and maintenance methods, including using ping commands, in-band telemetry, and Netflow to obtain network operational information, can be broadly categorized into two main approaches: active collection and passive collection.

[0003] Because TSN technology is a relatively new technology that has emerged over the past few decades, research on acquiring TSN operational information is relatively lacking. Existing methods mainly focus on real-time network monitoring and fault location solutions. One work proposed a TSN- Insight enables low-cost real-time monitoring of TSN networks by extending gPTP packets; another work focuses on collecting error information when network malfunctions occur, such as by sending data packets, and proposes a network fault information collection scheme that uses network topology analysis to construct probe packet paths to achieve deterministic fault information collection.

[0004] While these methods can acquire TSN network operational information, they lack the ability to dynamically adjust the monitoring flow configuration based on the network's real-time operational status. They also cannot collect real-time operational information with low latency and low overhead, incurring significant additional overhead despite achieving information collection. Furthermore, once a fault has occurred, although the error can be quickly identified, the consequences and losses are often irreversible. Therefore, a method for real-time monitoring of TSN switching and transmission equipment is urgently needed for fault detection and early warning. Summary of the Invention

[0005] The purpose of this application is to provide a method for collecting operational information of a time-sensitive network (TSN). This method involves sending additional monitoring traffic frames to traverse all switches in the network and collect operational information of switching devices and transmission links. This method can monitor various physical network elements of a TSN in real time and check whether the network is operating normally. At the same time, it can dynamically adjust the routing of the monitoring flow Probe_Message according to the real-time network operating status, thereby reducing the additional overhead of the monitoring flow on the network.

[0006] To achieve the above objectives, this application provides the following technical solution: a method for collecting operational information of time-sensitive networks, comprising the following steps:

[0007] Step 1: Obtain the global topology of the TSN network. The computing unit calculates the optimal route, performs frame attribute {Period, Len, Pri…} and route configuration, packages the relevant configuration and routing information, and sends it to the terminal device generating the monitoring stream in the form of control flow. The terminal device sends the initial monitoring stream data frame Probe_Message according to the specified sending time. (1) ;

[0008] Step 2: Under the calculated optimal route {H,S1,S2,S3,…,S1,H}, monitor the flow data frame Probe_Message. (1) After passing through all switches SW i The operating information of each hop switch is then added to the end of the frame to obtain a new monitoring stream data frame, Probe_Message. (i) The switch then sends the newly obtained monitoring stream data frame Probe_Message (i) Forward according to the routing order;

[0009] Step 3: After accessing all switches in the TSN network, the monitoring stream data frame containing the operating information of all switches is forwarded to the control plane; the control plane preprocesses and performs mathematical calculations on this raw data, updates the weight state matrix Δ, and calculates the new optimal route based on the new weight state matrix Δ.

[0010] Step 4: Send the results of preprocessing and mathematical calculations to the control center, which will then package the new routing and configuration information and send it to the terminal device that generates the monitoring stream to create a new monitoring stream.

[0011] The monitoring flow is additional control plane overhead that collects link physical information through additional traffic frames.

[0012] The monitoring stream needs to perceive the switch status in real time, which is time-sensitive data. According to the TSN priority scheduling mechanism, the monitoring stream is assigned a priority of 7.

[0013] Initial monitoring stream Probe_Message (1) The frame is generated to meet the minimum Ethernet frame length requirement of 64 bytes. The frame header consists of 18 bytes, and the remaining 46 bytes carry identification and description of basic control flow configuration information. After each pass through a switch, the monitoring flow frame length is increased by n bytes of running information.

[0014] The operational information of each switch is stored in a dictionary format, in the following form: {[Q occupy :,A],[Q capacity :,B],[T:,C],[SR:,{S i ,S j Each dictionary entry occupies 4 bytes of memory.

[0015] The operating information for each switch includes the address. i Link X i Queue Q i and routing {S i S j}wait.

[0016] Each switch can identify the Probe_Message for monitoring traffic and the Traffic_Message for network traffic.

[0017] The update rules for the state matrix Δ are as follows:

[0018] At the start of monitoring, a weighted state matrix Δ of the running link edge is initialized based on the global topology, and this is done on the switch SW. i and switch SW j Weights that are connected are set to 0, and weights that are not connected are set to W. ij Setting it to ∞ yields the weight state matrix Δ of the running edge;

[0019] After each round of monitoring stream collection is completed, the weight state matrix Δ is updated, and the queue capacity of any port on the switch is Q. capacity The queue currently occupies a frame length of Q. occupy SW switch i and switch SW j The edge weights between them are calculated by the following formula:

[0020] W ij =Q occupy(i) / Q capacity(j) ;

[0021] Q occupy(i) and Q capacity(j) Refers to switch SW i Arrival at switch SW j The queue information for the corresponding port;

[0022] If the monitored flow passes through a switch multiple times during a single routing process, the weight state matrix Δ is updated according to the information obtained from the last pass. If a link failure is detected, the weight is set to ∞.

[0023] The new optimal route is calculated based on the new edge weight state matrix Δ, specifically as follows:

[0024] Define the path cost function of the monitoring flow based on the weight state matrix. The route is {H,S} i ,S j ,…,S i If H}, then the optimization objective to minimize the cost is as follows:

[0025]

[0026] cost represents the total cost function value of a single routing path, p represents the pair (x,y), (x,y) represents the path from switch x to switch y, and the sequence of routing paths L is {(a,b),(b,c),(c,d),……(z,a)}.

[0027] The optimization objective can be reduced to the Traveling Salesman Problem (TSP). The TSN network topology can be abstracted as a graph G(V,W), where V represents the switches and terminal nodes in the graph, and W represents the weights of the edges connecting different switches. G(V,W) is transformed into G'(V,W), where any edge W in G' represents the shortest distance between any two points in V. The shortest distance between any two points can be obtained using a breadth-first search algorithm. Two auxiliary points X and Y are added to G'(V,W) to form a new graph G''(V,W), where the weight from X to all points is 0, and the weight from Y to the fixed starting point is 0. The optimization objective problem is thus abstracted into the solution of the classic TSP.

[0028] In step 4, the results of preprocessing and mathematical calculation are sent to the CNC configurator. The CNC configurator calculates the order and offset of traffic scheduling in the next supercycle based on the current load of different port queues of each switch.

[0029] Compared with the prior art, the beneficial effects of the present invention are:

[0030] A method for obtaining real-time physical operation information of various TSN switching devices has been invented, filling a gap in the field and enhancing the real-time monitoring of TSN operation information. By using a single monitoring flow and dynamic routing calculation method, the impact of network monitoring traffic information on normal network traffic transmission is reduced, and the additional traffic generated by monitoring and maintenance is reduced. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart illustrating the specific process of collecting operational information in an embodiment of this application.

[0033] Figure 2 This is an overall architecture diagram of the collection method in an embodiment of this application. Detailed Implementation

[0034] The technical solutions of the embodiments of this application will now be described with reference to the accompanying drawings. It should be noted that similar reference numerals and letters in the following drawings indicate similar items; therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0035] The terms “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0036] The following is combined Figure 1 This application introduces a method for collecting operational information of time-sensitive networks, including the following specific steps:

[0037] Step 1: Obtain the global network topology through relevant protocols such as LLDP. The control plane computing unit calculates the optimal route, performs traffic attribute {Period, Len, Pri…} and route configuration, packages the relevant routing and configuration information, and sends it as a control flow to the terminal device generating the monitoring traffic. The function of this control flow is to specify a particular terminal device (Host) to send monitoring streams according to a specified time or period. This monitoring stream passes through all switches in the TSN network once and then returns to the initial terminal device. The terminal sends monitoring stream data frames (Probe_Message) according to the specified period. (1)The monitoring stream meets the configuration requirements issued by the control center. Since this monitoring stream needs to perceive the switch status in real time and is time-sensitive data, it is assigned a priority of 7 according to the TSN priority scheduling mechanism. The initial monitoring stream, Probe_Message, is generated to meet the minimum Ethernet frame length requirement of 64 bytes. The frame header includes 18 bytes, and the remaining 46 bytes carry identification and description of basic control flow configuration information. After each switch hop, the monitoring stream frame length increases by n bytes for operational information. Since the maximum data segment length that an Ethernet frame can carry is 1500 bytes, if the operational information for each switch is 32 bytes, it can access a maximum of 46 switches at a time. After passing through the last hop switch, it is forwarded to the control plane.

[0038] Step 2: Based on the calculated optimal route {H, S1, S2, S... 3, Under the guidance of …,S1,H}, it will pass through all switches SW i Furthermore, each switch can identify the monitoring flow Probe_Message and network traffic Traffic_Message; it periodically records the address of each hop switch at the most recent sampling time. i Link X i Queues Q on different ports i Routing {S i S j The runtime information, such as}, is appended to the end of the frame to obtain a new monitoring stream data frame, Probe_Message. (i) Then, the monitoring stream is forwarded out according to the routing order; the operating information of each switch is stored in a dictionary format, in the following form: {[Q occupy :,A],[Q capacity :,B],[T:,C],[SR:,{S i ,S j Each dictionary entry occupies 4 bytes of memory.

[0039] Step 3: After accessing all switches within the topology, the monitoring flow will collect real-time operational information from each switch and forward it to the control plane for further processing. The control plane's computation unit will preprocess and perform mathematical calculations on the collected raw data, and then update the weight state matrix Δ. The update rules for the weight state matrix Δ are as follows:

[0040] At the start of monitoring, a weighted state matrix Δ of the running link edge is initialized based on the global topology, and this is done on the switch SW. i and switch SW j Edges that are connected are set to a weight of 0, while edges that are not connected are set to a weight of W.ij Set to ∞. Obtain the weight state matrix Δ of the running edge. The weight state matrix Δ is an N x N matrix, where N is the number of switches in the TSN network. The switches in the TSN network are numbered from 1 to N. The rows represent the starting point from switch number i, and the columns represent the destination from switch number j.

[0041] After each round of monitoring stream collection is completed, the weight state matrix Δ is updated. The queue capacity of any port on the switch is Q. capacity The queue currently occupies a frame length of Q. occupy Switch SW i and SW j The weights between them are calculated by the following formula:

[0042] W ij =Q occupy(i) / Q capacity(j) ;

[0043] Q occupy(i) and Q capacity(j) Refers to switch SW i Arrival at switch SW j The queue information for the corresponding port;

[0044] If a monitored flow passes through a switch multiple times during a single routing process, the state matrix is ​​updated based on the information obtained from the last pass. If a link failure is detected, the weight is set to ∞.

[0045] The new optimal route is then calculated based on the new weight state matrix Δ. The calculation method for the new route is as follows:

[0046] The path cost function of the monitoring flow is defined based on the weight state matrix as follows: The route is {H,S} i ,S j ,…,S i The optimization objective we need to minimize is as follows:

[0047]

[0048] cost represents the total cost of a path, p represents a pair (x, y), and (x, y) represents a path from switch x to switch y. The set of routing paths, L, is {(a, b), (b, c), (c, d), ..., (z, a)}.

[0049] The optimization objective can be reduced to the Traveling Salesman Problem (TSP). The TSN network topology can be abstracted as a graph G(V,W), where V represents the switches and terminal nodes in the graph, and W represents the weights of the edges connecting different switches. G(V,W) is transformed into G'(V,W), where any edge W in G' represents the shortest distance between any two points in V, which can be obtained using breadth-first search. Two auxiliary points X and Y are added to G'(V,W) to form a new graph G''(V,W), where the weight from X to all points is 0, and the weight from Y to the fixed starting point is 0. The optimization objective problem is then abstracted into solving the classic TSP problem. Heuristic algorithms such as genetic algorithms and ant colony algorithms can be used to solve the classic TSP problem.

[0050] After the monitoring flow completes route collection and reaches its destination, it corrects and obtains a new weight state matrix Δ based on the collected operational information, then performs route calculation to find the route with the minimum cost, and the control plane sends this route to the terminal device that generated the monitoring traffic to update the route.

[0051] Step 4: The processing results from the control plane are distributed to the control center and the CNC configurator. The control center then distributes the results to the terminal devices that generate monitoring traffic, creating new monitoring flows. This information is also used as reference data for the CNC configurator to calculate a new window planning scheme for traffic scheduling, generating feedback adjustment. The CNC configurator calculates the order and offset of traffic scheduling in the next supercycle based on the current port queue load of each switch.

[0052] This invention enables access to the operational information of all switches through a single monitoring stream. The control plane selects a designated terminal device (Host) as the generator of the monitoring stream. This designated terminal device can send monitoring streams at specified times or periods. The monitoring stream passes through all switches in the TSN network once and then returns to the initial starting terminal device.

[0053] As the monitoring flow passes through each switch, the switch can append the queue and link information for each different port to the Probe_Message of the monitoring flow, and then forward the modified monitoring traffic frame according to the route. Under the pre-configured route in the control center, a single monitoring flow can access all switches, allowing repeated access to certain side links during the access process. The monitoring flow data sent to the control plane is handed over to the computing unit for verification, data processing, and route calculation. The processed results are then sent to the CNC configurator, providing feedback to the CNC's gating scheduling of traffic. Simultaneously, the processed results are also fed back to the control center, which issues new configurations to the terminals that generate monitoring traffic and updates the routes.

Claims

1. A method for collecting operational information of a time-sensitive network, characterized in that, Includes the following steps: Step 1: Obtain the global topology of the TSN network. The computing unit calculates the optimal route, performs frame attribute and route configuration, packages the relevant configuration and routing information, and sends it to the terminal device generating the monitoring stream in the form of control flow. The terminal device sends the initial monitoring stream data frame Probe_Message according to the specified sending time. (1) ; Step 2: Under the calculated optimal route, monitor the flow data frame Probe_Message. (1) After passing through all switches SW i The operating information of each hop switch is then added to the end of the frame to obtain a new monitoring stream data frame, Probe_Message. (i) The switch then sends the newly obtained monitoring stream data frame Probe_Message (i) Forward according to the routing order; Step 3: After accessing all switches in the TSN network, the monitoring stream data frame containing the operating information of all switches is forwarded to the control plane; the control plane preprocesses and performs mathematical calculations on this raw data, updates the weight state matrix Δ, and calculates the new optimal route based on the new weight state matrix Δ. Step 4: Send the results of preprocessing and mathematical calculations to the control center, which will then package the new routing and configuration information and send it to the terminal device that generates the monitoring stream to create a new monitoring stream.

2. The method for collecting operational information of a time-sensitive network according to claim 1, characterized in that: The monitoring flow is additional control plane overhead that collects link physical information through additional traffic frames.

3. A method for collecting operational information of a time-sensitive network according to claim 1 or 2, characterized in that: The monitoring stream needs to perceive the switch status in real time, which is time-sensitive data. According to the TSN priority scheduling mechanism, the monitoring stream is assigned a priority of 7.

4. The method for collecting operational information of a time-sensitive network according to claim 1, characterized in that: Initial monitoring stream Probe_Message (1) The frame is generated to meet the minimum Ethernet frame length requirement of 64 bytes. The frame header consists of 18 bytes, and the remaining 46 bytes carry identification and description of basic control flow configuration information. After each pass through a switch, the monitoring flow frame length is increased by n bytes of running information.

5. The method for collecting operational information of a time-sensitive network according to claim 2, characterized in that: The operational information of each switch is stored in a dictionary format, and the operational information of each switch includes the address. i Link X i Queue Q i and routing {S i S j }wait.

6. The method for collecting operational information of a time-sensitive network according to claim 1, characterized in that: Each switch can identify the Probe_Message for monitoring traffic and the Traffic_Message for network traffic.

7. The method for collecting operational information of a time-sensitive network according to claim 1, characterized in that: The update rules for the state matrix Δ are as follows: At the start of monitoring, a weighted state matrix Δ of the running link edge is initialized based on the global topology, and this is done on the switch SW. i and switch SW j Weights that are connected are set to 0, and weights that are not connected are set to W. ij Setting it to ∞ yields the weight state matrix Δ of the running edge; After each round of monitoring stream collection is completed, the weight state matrix Δ is updated, and the queue capacity of any port on the switch is Q. capacity The queue currently occupies a frame length of Q. occupy SW switch i and switch SW j The edge weights between them are calculated by the following formula: W ij = Q occupy(i) / Q capacity(j) ; Q occupy(i) and Q capacity(j) Refers to switch SW i Arrival at switch SW j The queue information for the corresponding port; If the monitored flow passes through a switch multiple times during a single routing process, the weight state matrix Δ is updated according to the information obtained from the last pass. If a link failure is detected, the weight is set to ∞.

8. A method for collecting operational information of a time-sensitive network according to claim 1 or 7, characterized in that: The new optimal route is calculated based on the new edge weight state matrix Δ, specifically as follows: The path cost function cost= is defined based on the weight state matrix. The route is {H,S} i ,S j ,…,S i If we have H, then the optimization objective for minimizing the cost is as follows: Minimize cost= ; cost represents the total cost function value of a single routing path, p represents the pair (x,y), (x,y) represents the path from switch x to switch y, and the sequence of routing paths L is {(a,b),(b,c),(c,d),……(z,a)}.

9. The method for collecting operational information of a time-sensitive network according to claim 8, characterized in that: The optimization objective can be reduced to the Traveling Salesman Problem (TSP). The TSN network topology can be abstracted as a graph G(V,W), where V represents the switches and terminal nodes in the graph, and W represents the weights of the edges connecting different switches. G(V,W) is transformed into G'(V,W), where any edge W in G' represents the shortest distance between any two points in V. The shortest distance between any two points can be obtained using a breadth-first search algorithm. Two auxiliary points X and Y are added to G'(V,W) to form a new graph G''(V,W), where the weight from X to all points is 0, and the weight from Y to the fixed starting point is 0. The optimization objective problem is thus abstracted into the solution of the classic TSP.

10. The method for collecting operational information of a time-sensitive network according to claim 1, characterized in that: In step 4, the results of preprocessing and mathematical calculation are sent to the CNC configurator. The CNC configurator calculates the order and offset of traffic scheduling in the next supercycle based on the current load of different port queues of each switch.

Images