A routing and scheduling based time sensitive network topology design method

Abstract

The application discloses a routing and scheduling based time sensitive network topology design method, and relates to the field of industrial networks.The application carries out mathematical modeling on a TSN topology-routing-scheduling joint problem, designs an optimization target, and solves the optimization target of the TSN topology-routing-scheduling joint problem based on schedulability analysis.The application can not only effectively reduce the number of switches and cables, but also significantly reduce deployment cost and maintenance cost;by simplifying the modeling of the scheduling part and converting complex nonlinear constraints into linear / quadratic modeling, the TSN routing and scheduling problem can be directly input into a commercial solver for efficient solving;the TSN network can be quickly deployed, real-time requirements of an industrial site are met, and the TSN network is adapted to rapid changes of a flexible production environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of industrial networks, and in particular to a time sensitive network topology design method based on routing and scheduling. BACKGROUND

[0002] With a large number of perception, execution, and computing terminals entering the industrial field, industrial networks are facing the problems of high demand for traffic, high real-time requirements, and gradually increasing costs. Time Sensitive Networking (TSN) provides protection for on-demand transmission of various types of traffic with a fine scheduling mechanism, and is widely concerned by the industry with an open standard system, and is expected to become the core of the next generation of industrial networks. However, as the number of field devices increases, the cost of network equipment (including the material cost of switches and cables and the labor cost required for deployment and maintenance) also increases. Too simple network topology cannot meet the traffic transmission requirements, and too complex network topology will dramatically increase network costs. Therefore, how to optimize the TSN network topology to reduce the number of switches and cables to meet the deterministic transmission requirements of traffic with the least cost has become a key problem for TSN development.

[0003] To design a TSN network topology that meets the transmission requirements, the behavior of the traffic, i.e., the routing and scheduling scheme, needs to be considered while designing the network topology structure. However, TSN routing and scheduling itself is a complex problem. A large number of academic papers point out that TSN scheduling is an NP-hard problem, which is difficult to solve within a polynomial time. If routing and scheduling are considered together when designing the topology, not only is the coupling relationship between the three difficult to describe, but the complexity of the problem is extremely high, and it is difficult to obtain a feasible joint problem solution using a brute-force solving method. Therefore, it is crucial to establish the coupling relationship between TSN topology-routing-scheduling and design an efficient solving algorithm for the joint problem to promote the application of TSN.

[0004] Therefore, the skilled person in the art is committed to developing a time sensitive network topology design method based on routing and scheduling. By studying the coupling relationship between topology and routing and scheduling, a TSN topology-routing-scheduling joint design problem is established, and an actually usable TSN network topology is designed and optimized. SUMMARY

[0005] In view of the above defects of the prior art, the technical problem to be solved by the present application is the topology design problem considering the routing and scheduling scheme in the TSN network.

[0006] To achieve the above-mentioned purpose, the present application provides a time sensitive network topology design method based on routing and scheduling, comprising the following steps:

[0007] Step 1, mathematically model the TSN topology-routing-scheduling joint problem, and design an optimization objective;

[0008] Step 2, based on the schedulability analysis, solve the optimization objective of the TSN topology-routing-scheduling joint problem.

[0009] Further, the step 1, establishes the parameter index of the switch and the cable, and combines the parameter in the TSN scheduling model to establish the mathematical model of the topology-routing-scheduling joint problem.

[0010] Further, the step 1, describes the coupling relationship through the constraints between the TSN topology-routing-scheduling variables.

[0011] Further, the step 1, takes the minimum network construction cost as the optimization objective of the joint problem.

[0012] Further, the step 1, models the variable of the scheduling part as a one-hot vector.

[0013] Further, the step 1, models the time slot occupation in the TSN scheduling process in the form of a one-hot vector, and represents the scheduling constraint by a linear quadratic expression.

[0014] Further, the step 1, includes the following steps:

[0015] Step 1.1, establish the relationship between the nodes and edges in the network topology in the joint problem;

[0016] Step 1.2, establish the relationship between the traffic routing and the network topology in the joint problem;

[0017] Step 1.3, establish the relationship between the traffic hop count and the routing in the joint problem;

[0018] Step 1.4, establish the device port number constraint;

[0019] Step 1.5, establish the variable constraint of the TSN traffic scheduling in the joint problem;

[0020] Step 1.6, establish the TSN delay constraint in the joint problem;

[0021] Step 1.7, establish the TSN time slot mapping constraint in the joint problem;

[0022] Step 1.8, establish the TSN time slot compression constraint in the joint problem;

[0023] Step 1.9, establish the TSN link capacity constraint in the joint problem;

[0024] Step 1.10, establish the optimization objective of the joint problem.

[0025] Further, the step 2, by analyzing TSN schedulability, judges whether the scheme after iteration can meet the traffic transmission demand.

[0026] Further, the step 2 includes a feasible solution judgment module, an update judgment module, a route update module, a link update module and a node update module.

[0027] Further, the step 2 includes the following steps:

[0028] Step 2.1, generate an initial solution of TSN topology and routing, and enter the feasible solution judgment module; after entering the feasible solution judgment module, judge whether the current solution is schedulable, if schedulable, judge whether it is a feasible solution, if schedulable and a feasible solution, the algorithm finds a feasible solution and returns the solution; if schedulable but the solution is not feasible, it means that the schedulable judgment device makes a false judgment, then add this solution to the taboo table, and do not search this solution in the future, then enter the update judgment module; if not schedulable, enter the update judgment module;

[0029] Step 2.2, after the current solution enters the update judgment module, first judge whether the route is updatable; if the route is updatable, enter the route update module, otherwise judge whether the link is updatable; if the link is updatable, enter the link update module, otherwise judge whether the node is updatable; if the node is updatable, enter the node update module, otherwise return no feasible solution;

[0030] Step 2.3, after the current solution enters the route update module, first update the flow routing by using the simulated annealing algorithm, evaluate the congestion degree of the route, and the simulated annealing algorithm optimizes the route to the direction of the minimum congestion degree; after optimization, judge the schedulability and make the current solution enter the feasibility judgment module;

[0031] Step 2.4, after the current solution enters the link update module, first find the link with the maximum congestion degree in the network as the bottleneck link, find the node associated with the bottleneck link as the bottleneck node, add the link with the minimum cost among the unselected links associated with the bottleneck node to the current network, and judge the schedulability, then the current solution enters the feasibility judgment module;

[0032] Step 2.5, after the current solution enters the node update module, first add the node with the minimum cost among all the candidate nodes in the network, then judge the schedulability and enter the feasible solution judgment module.

[0033] The existing TSN scheduling scheme cannot solve the TSN network topology design problem and does not have the ability to optimize the TSN network topology. The present application models the TSN topology-routing-scheduling joint design problem mathematically, abstracts the coupling relationship between TSN topology, routing and scheduling, and designs the optimization problem, which converts the TSN topology-routing-scheduling design into a solvable mathematical problem. The present application establishes the related parameter indicators of switches and cables, and establishes the mathematical model of the topology-routing-scheduling joint problem in combination with the related parameters in the TSN scheduling model, and describes the coupling relationship through the constraints between variables, and takes the minimum network construction cost as the optimization target of the joint problem.

[0034] The existing TSN scheduling scheme contains complex nonlinear constraint parts when the model is established, which is difficult to solve in a commercial solver. In the modeling of the present application, the key variables of the scheduling part are modeled by one-hot vector, and the scheduling constraints are represented by a new modeling method, which is beneficial to model explicitness. The present application models the time slot occupation in the TSN scheduling process in the form of one-hot vector, introduces the processes of time slot mapping and time slot compression, and represents the complex scheduling constraints in the form of linear / quadratic.

[0035] The existing TSN network related research cannot efficiently solve the TSN topology-routing-scheduling joint design problem, and the solving time is too long. The present application proposes a TSN topology-routing-scheduling efficient solving algorithm based on schedulability analysis, which can effectively solve the TSN topology-routing-scheduling joint design problem. The present application quickly judges whether the scheme after iteration can meet the traffic transmission demand by analyzing the TSN schedulability, saves the time of solving complex coupled problems one by one, and greatly improves the problem solving efficiency.

[0036] The TSN network topology-routing-scheduling joint optimization scheme of the present application has significant technical advantages, especially in the aspect of topology optimization. By optimizing the network topology, the present application not only can effectively reduce the number of switches and cables, but also can significantly reduce the deployment cost and maintenance cost. This optimization measure not only meets the application traffic transmission demand, but also reduces the equipment investment and manual maintenance overhead, providing a more economical and efficient solution for TSN network deployment in the industrial field.

[0037] In addition, the present application simplifies the modeling of the scheduling part, converts the complex nonlinear constraints into linear / quadratic modeling, so that the TSN routing and scheduling problem can be directly input into a commercial solver for efficient solving. This method greatly improves the solving efficiency and significantly shortens the solving time. Compared with the traditional scheme, the improvement of the solving speed enables the TSN network to be quickly deployed, meets the real-time requirements of the industrial field, and adapts to the rapid changes of the flexible production environment.

[0038] In the aspect of industrial application, the technical solution of the application has broad prospects in the field of intelligent manufacturing and industrial internet. By improving the adaptability and rapid adjustment capability of the network, the TSN network can efficiently support various application scenarios such as automated production lines, robot collaboration and real-time monitoring.

[0039] Compared with the prior art, the application has the following obvious essential features and significant advantages:

[0040] 1. The application gives the TSN network topology optimization capability, which can reduce the number of deployed switch costs and cable length on the basis of meeting the application traffic transmission demand, greatly reducing the network deployment cost and manual maintenance cost.

[0041] 2. The application linearly / secondarily models the scheduling part, which can directly input the TSN scheduling or TSN topology-routing-scheduling joint design problem into a commercial solver. It is beneficial to solve TSN scheduling related problems in a more standardized and rapid way, which can greatly improve the problem solving efficiency and further improve the deployment efficiency of the TSN network.

[0042] 3. The application can greatly improve the solving speed of the TSN topology-routing-scheduling joint problem, so that the TSN has the capability of rapidly adjusting the network topology and gives the TSN the potential to adapt to flexible production requirements.

[0043] The concept, specific structure and technical effects of the application will be further described below with reference to the accompanying drawings, so as to fully understand the purpose, features and effects of the application. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 is a state transition diagram of a preferred embodiment of the application. DETAILED DESCRIPTION

[0045] The following describes several preferred embodiments of the application with reference to the accompanying drawings, so as to make the technical content of the application clearer and easier to understand. The application can be embodied in many different forms, and the protection scope of the application is not limited to the embodiments mentioned herein.

[0046] In the drawings, the same components have the same reference numerals, and components with similar structures or functions have similar reference numerals. The size and thickness of each component shown in the drawings are arbitrarily shown, and the application does not limit the size and thickness of each component. In order to make the drawing clearer, the thickness of some components is appropriately exaggerated in some places.

[0047] The application aims to solve the topology design problem considering routing and scheduling scheme in TSN network. To this end, the application proposes a mathematical modeling method and an efficient solving algorithm, and the state transition diagram of the solving algorithm is as followsFigure 1 are shown.

[0048] In the modeling method, the TSN network is configured in a Cyclic Queuing and Forwarding (CQF) mode defined by the IEEE 802.1Qch protocol. The complete steps of modeling are as follows:

[0049] Step 1-1: Establish the relationship between nodes and edges in the network topology in the joint problem. Use n i to represent whether there is a node at position i (1 if there is, 0 if there is not), and use l e to represent whether there is a link on edge e = (i, j) (1 if there is, 0 if there is not). At this time, the link exists on the premise that the node exists, and the mathematical expression is:

[0050] l (i,j) ≤n i .

[0051] Step 1-2: Establish the relationship between traffic routing and network topology in the joint problem. Use to represent whether the routing of flow k passes through edge e (1 if it does, 0 if it does not), and the routing exists on the premise that the topology exists, that is:

[0052]

[0053] And there is a continuity constraint on the routing:

[0054]

[0055]

[0056] Step 1-3: Establish the relationship between traffic hop count and routing in the joint problem. Use to represent the hop count of flow k on edge e, and there is a constraint relationship between the flow hop count and the routing:

[0057]

[0058] And there is a continuity constraint on the hop count itself:

[0059]

[0060] Step 1-4: In order to ensure that the number of switch ports does not exceed the limit, establish the device port number constraint. Use N P ,N E to represent the maximum number of ports of the switch and the maximum number of ports of the terminal device, respectively, and there is the following constraint:

[0061]

[0062] Step 1-5: Establish the relevant variable constraints of TSN traffic scheduling in the joint problem. In order to better linearize / quadraticize the constraints, define whether flow k is occupied in the s-th time slot as Δ k (s), where 1 is occupied and 0 is not occupied. Obviously, each flow has only one component equal to 1 in a cycle. Let p k represent the cycle of flow k, then the constraint expression is:

[0063]

[0064] Step 1-6: Establish the TSN delay constraint in the joint problem. Since the upper bound of the delay in the CQF model is Let d k represent the maximum end-to-end delay of flow k, then the delay constraint can be expressed as:

[0065]

[0066] Step 1-7: Establish the TSN time slot mapping constraint in the joint problem. Since the CQF model requires that the traffic in each time slot in the super cycle does not overflow, the time slot mapping can be defined as follows: the occupation of the s-th time slot of flow k on edge e (0-p k is occupied in the range of 0-p

[0067]

[0068] Step 1-8: Establish the TSN time slot compression constraint in the joint problem. Since the flow in the TSN scheduling problem has periodicity, the total time of conflict determination can be compressed from the least common multiple of each flow cycle to each cycle. Define the time slot compression variable to represent the number of cycles that flow k passes through in the s-th time slot of edge e, then the time slot compression satisfies the following two constraints:

[0069]

[0070] Step 1-9: Establish the TSN link capacity constraint in the joint problem. After using time slot compression and time slot mapping to represent the periodicity of the flow and the relationship between the edge and the number of hops, the TSN link capacity constraint can be written in the form of not higher than quadratic. Assuming that the upper limit of the capacity of each time slot is Λ, then the TSN link capacity constraint can be expressed as:

[0071]

[0072] Step 1-10: Establish the optimization objective of the joint problem. After all the relevant constraints are constructed, the optimization objective is to minimize the cost required to construct the network topology. Set the cost of each switch as The unit length link cost is The optimization objective can be expressed as:

[0073]

[0074] In this way, the TSN network design mathematical model considering routing and scheduling is established, and linear / quadratic is realized in the model.

[0075] The TSN topology design solving algorithm includes five modules of feasible solution judgment, update judgment, route update, link update, and node update Figure 1 ), which specifically includes the following steps:

[0076] Step 2-1: Generate the initial solution of TSN topology and routing, and enter the feasible solution judgment module. After entering the feasible solution judgment module, the behavior includes the following process: judge whether the current solution is schedulable, if it is schedulable, judge whether it is a feasible solution, if it is schedulable and a feasible solution, the algorithm finds a feasible solution and returns this solution; if it is schedulable but the solution is not feasible, it means that the schedulable judgment device makes a false judgment, then add this solution to the taboo table (do not search this solution again in the future) and enter the update judgment module; if it is not schedulable, enter the update judgment module.

[0077] Step 2-2: After the current solution enters the update judgment module, first judge whether the routing is updatable. If the routing is updatable, enter the routing update module, otherwise judge whether the link is updatable. If the link is updatable, enter the link update module, otherwise judge whether the node is updatable. If the node is updatable, enter the node update module, otherwise return no feasible solution.

[0078] Step 2-3: After the current solution enters the routing update module, first update the flow routing by using the simulated annealing algorithm, evaluate the congestion degree of the routing, and the simulated annealing algorithm optimizes the routing to the direction of the minimum congestion degree. After the optimization is completed, judge the schedulability and make the current solution enter the feasibility judgment module.

[0079] Step 2-4: After the current solution enters the link update module, first find the link with the maximum congestion degree in the network as the bottleneck link, find the nodes associated with the bottleneck link as the bottleneck nodes, and add the link with the minimum cost among the unselected links associated with the bottleneck nodes to the current network, and judge the schedulability, then make the current solution enter the feasibility judgment module.

[0080] Step 2-5: After the current solution enters the node update module, the node with the minimum cost is first added in all the alternative nodes of the network, and then the schedulability is judged to enter the feasible solution judging module.

[0081] The preferred embodiments of the present application are described in detail above. It should be understood that modifications and variations can be made by those of ordinary skill in the art without departing from the spirit and scope of the present application. Accordingly, it is intended that all such modifications and variations be included within the scope of the following claims and their equivalents.

Claims

1. A time-sensitive network topology design method based on routing and scheduling, characterized in that, Includes the following steps: Step 1: Mathematically model the TSN topology-routing-scheduling joint problem and design an optimization objective; Step 1-1: Establish the relationships between nodes and edges within the network topology in the joint problem; using Indicates position Does a node exist? A value of 1 indicates it exists, and 0 indicates it does not. Representing an edge To determine if a link exists, a value of 1 indicates that the link exists, and 0 indicates that it does not exist. In this case, the existence of the link presupposes the existence of the node, and the expression is: Steps 1-2: Establish the relationship between traffic routing and network topology in the joint problem; using Represents a stream Does the route pass through the edge? If the value is 1, the edge passes through; if it is 0, the edge does not pass through. The existence of an upper route presupposes the existence of a topology, that is: Furthermore, the routing has continuity constraints: Steps 1-3: Establish the relationship between traffic hop count and routing in the joint problem; using Represents a stream On the side The number of hops in a flow indicates a constraint between the number of hops and the routing: Furthermore, the number of hops itself has a continuity constraint: Steps 1-4: To ensure the number of switch ports does not exceed the limit, establish device port number constraints; use These represent the maximum number of ports on the switch and the maximum number of ports on the terminal device, respectively, with the following constraints: Steps 1-5: Establish relevant variable constraints for TSN traffic scheduling in the joint problem; [The sentence is incomplete and requires more context to be translated accurately.] In the Whether a time slot is occupied is defined as The value of an element is 1 if it is occupied and 0 if it is not occupied; clearly, each flow has exactly one component that is 1 within a cycle; using Representative Stream The period, and the constraint expression are: Steps 1-6: Establish TSN delay constraints in the joint problem; since the upper bound of delay in the CQF model is... , record flow The maximum end-to-end delay is The time delay constraint is expressed as: Steps 1-7: Establish TSN slot mapping constraints in the joint problem; since the CQF model requires that the traffic in each slot does not overflow within the supercycle, this is achieved by defining slot mapping constraints. Flow On the side The Time slot occupancy status, 0- If the area is occupied, then the corresponding If the value is 1, and it is mapped to the first hop time slot, then the time slot mapping satisfies the following constraint: Steps 1-8: Establish TSN slot compression constraints in the joint problem; since the flows in the TSN scheduling problem are periodic, the total time for conflict determination can be compressed from the least common multiple of the flow cycles to within each cycle; define slot compression variables. Represents a stream On the side The The number of cycles a time slot passes through determines whether time slot compression satisfies the following two constraints: Steps 1-9: Establish TSN link capacity constraints in the joint problem; after representing the periodicity of the flow and the relationship between edges and hops using time slot compression and time slot mapping, write out the TSN link capacity constraints in a form no higher than quadratic; assume the upper limit of time slot capacity is... Then the TSN link capacity constraint is expressed as: Steps 1-10: Establish the optimization objective for the joint problem; After establishing all relevant constraints, the optimization objective is to minimize the cost required to construct the network topology; the cost of each switch is set to... The cost per unit length of the link is The optimization objective is then expressed as: In this way, a mathematical model of the TSN network that takes into account routing and scheduling is established; Step 2: Based on schedulability analysis, solve the optimization objective of the TSN topology-routing-scheduling joint problem; Step 2.1: Generate the initial solution for TSN topology and routing, and proceed to the feasible solution judgment module; After entering the feasible solution judgment module, it is determined whether the current solution is schedulable. If it is schedulable, it is determined whether it is a feasible solution. If it is both schedulable and feasible, the algorithm has found a feasible solution and returns this solution. If it is schedulable but the solution is not feasible, it means that the schedulability judge has made a wrong judgment. In this case, this solution is added to the tabu list and will not be searched for again. Then it enters the update judgment module. If it is not schedulable, proceed to the update judgment module; Step 2.2: After the current solution enters the update judgment module, it first determines whether the route is updatable; If the route is updatable, proceed to the route update module; otherwise, determine if the link is updatable. If the link is updatable, proceed to the link update module; otherwise, determine if the node is updatable. If a node can be updated, proceed to the node update module; otherwise, return that no feasible solution exists. Step 2.3: After the current solution enters the route update module, the flow route is first updated using the simulated annealing algorithm to evaluate the congestion level of the route. The simulated annealing algorithm optimizes the route in the direction of minimum congestion. After optimization, the schedulability is judged and the current solution enters the feasibility judgment module. Step 2.4: After the current solution enters the link update module, it first finds the link with the highest congestion in the network as the bottleneck link, identifies the nodes associated with the bottleneck link as bottleneck nodes, adds the lowest-cost link among the unselected links associated with the bottleneck nodes to the current network, and determines schedulability. After that, the current solution enters the feasibility judgment module. Step 2.5: After the current solution enters the node update module, the node with the lowest cost is first added to all the candidate nodes in the network. Then, the schedulability is determined, and the feasible solution judgment module is entered.

2. The time-sensitive network topology design method based on routing and scheduling as described in claim 1, characterized in that, Step 1 involves establishing parameter indicators for switches and cables, and combining these with parameters from the TSN scheduling model to create a mathematical model for the joint topology-routing-scheduling problem.

3. The time-sensitive network topology design method based on routing and scheduling as described in claim 1, characterized in that, Step 2 involves analyzing the schedulability of the TSN to determine whether the iterative solution can meet the traffic transmission requirements.

4. The time-sensitive network topology design method based on routing and scheduling as described in claim 1, characterized in that, Step 2 includes a feasible solution judgment module, an update judgment module, a route update module, a link update module, and a node update module.

Images