Time-triggered traffic gating list generation method

Abstract

The invention relates to a time-triggered traffic gating list generation method, and belongs to the technical field of communication networks. The solving process of the time-triggered traffic gating list in the time-sensitive network is mainly decomposed into two stages, namely a stage of determining a traffic scheduling sequence and a stage of determining initial transmission time stream by stream according to the determined sequence. The stage of determining the traffic scheduling sequence is based on the idea of harmonious search, and high-quality traffic scheduling sequences are obtained by utilizing according to a consideration rule and exploring according to a fine adjustment rule in a solution space; determining an initial transmission time stage stream by stream according to the determined sequence, distributing all traffic instances to corresponding links in a hypercycle period based on the initial transmission time, and generating a scheduling result; and finally, generating a sending time table of each end system and a gating list of each port of the switch according to a scheduling result. According to the invention, the problem of gating list generation of time-triggered traffic in a large-scale and high-load time-sensitive network is solved.

Description

Technical Field

[0001] The invention belongs to the technical field of communication networks, and in particular relates to a method for generating a time-triggered traffic gating list. Background Art

[0002] With the rapid development of emerging communication technologies, the application scenarios of communication technologies are becoming more and more abundant. Communication network systems have more stringent requirements on the real-time, deterministic and reliability of data traffic. Time-sensitive networks (TSNs) have become one of the main choices for high-standard hard real-time scenarios such as advanced driver assistance systems, aerospace, and industrial Internet. In time-sensitive networks (TSNs), time-triggered TT traffic is used in hard real-time scenarios with high standards for latency, latency jitter, and real-time performance, and has the highest scheduling priority in time-sensitive networks (TSNs). After the time-triggered TT traffic builds a unified time reference for all network devices, it is transmitted according to the gated list (GCL). It is this rigorous scheduling mechanism based on time triggering that effectively ensures the deterministic end-to-end latency and ultra-low jitter characteristics of the time-triggered TT flow.

[0003] Time-sensitive network (TSN) traffic scheduling technology directly affects the quality of service (QoS) of different network services, plays a decisive role in the real-time and deterministic nature of data transmission, and is one of the key technologies and key research directions in current time-sensitive network research. Time-triggered TT flows rely on the gated list (GCL) driver. The essence of the time-triggered TT traffic scheduling problem is the generation of the gated list (GCL). The gated list (GCL) is generated under the conditions of time slot constraints, delay constraints, routing constraints, conflict avoidance constraints, etc., and a feasible solution is obtained based on the characteristics of the traffic. The complexity of its solution is closely related to the scale of the scheduling problem and the number of constraints. The more complex the network and the more traffic, the more difficult it is to solve the gated list (GCL).

[0004] The existing technology currently has the following disadvantages:

[0005] First, there is a problem in scheduling planning where the complexity of the solution increases exponentially with the increase of network scale and network load.

[0006] Second, only using fixed rules for traffic sorting and scheduling results in low network bandwidth utilization and poor adaptability to network scenarios.

[0007] Third, in complex topologies and traffic scenarios, there are problems such as poor global search, complex iterative operations, and slow convergence rate. Summary of the invention

[0008] The technical problems to be solved by the present invention are:

[0009] In order to avoid the deficiencies of the prior art, the present invention provides a method for generating a gating list of time-triggered traffic, which is used to solve the problem of generating a gating list of time-triggered traffic in a large-scale and high-load time-sensitive network.

[0010] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0011] A method for generating a time-triggered traffic gating list, comprising:

[0012] Obtaining the periods of all time-triggered flows to be scheduled, and grouping all time-triggered flows to be scheduled based on the periods of all time-triggered flows to be scheduled;

[0013] Initialize the harmony memory matrix of each group, arrange the harmony memory matrix of each group from small to large according to the flow period, and form an initial total harmony memory matrix;

[0014] Calculate the harmonious fitness of each harmonious solution vector in the initial total harmonious memory matrix, sort each harmonious solution vector from best to worst according to the harmonious fitness, and take the first HMS harmonious solution vectors to form the initial total harmonious memory matrix; where HMS is the size of the initialized harmonious memory matrix;

[0015] Iterate based on the initial total harmonious memory matrix, and output the solution vector corresponding to the optimal fitness of the total harmonious memory matrix when the iteration ends, which is the optimized traffic scheduling order;

[0016] Determine the start transmission time for each flow based on the traffic scheduling order, allocate all traffic instances to corresponding links within the super cycle based on the start transmission time, and generate a scheduling result;

[0017] Based on the scheduling results, a sending schedule for each end system and a gating list for each port of the switch are generated.

[0018] A further technical solution of the present invention: the initialization of the harmonious memory matrix of each group includes:

[0019] Calculate the number of flows n of the current group;

[0020] Use the Fisher-Yates shuffle algorithm to generate random permutations from 0 to n-1;

[0021] Execute the shuffle algorithm 2·HMS times, using the random permutation generated each time as a row vector to initialize the harmony memory matrix of the current group.

[0022] A further technical solution of the present invention: the quality criterion of the harmonious adaptability is:

[0023] Compare the number of scheduling failure flows. The smaller the number of scheduling failure flows, the better the harmony fitness.

[0024] When the number of scheduling failure flows is the same, the smaller the maximum flow completion time, the better the harmony fitness;

[0025] When the maximum flow completion time is the same, the harmonious fitness quality is the same, and the order is determined according to the original order of the harmonious memory matrix.

[0026] A further technical solution of the present invention is as follows: the iteration based on the initial total harmonious memory matrix and the output of the solution vector corresponding to the optimal fitness of the total harmonious memory matrix when the iteration is terminated include:

[0027] Determine the harmony memory consideration rate and adjustment rate for the current iteration number;

[0028] Based on the harmonious memory consideration rate and adjustment rate of the current iteration number, new solution vectors of the harmonious memory matrix of each group are composed according to the consideration rules and fine-tuning rules respectively, and the new solution vectors of each group are combined into a new solution vector of the total harmonious memory matrix. The fitness of the solution vector is calculated according to the harmonious fitness criterion.

[0029] A further technical solution of the present invention is as follows: the determination of the harmonious memory consideration rate and adjustment rate of the current number of iterations is specifically as follows:

[0030] Let the current iteration number be i, then

[0031] Among them, HMCR is the harmonious memory consideration rate of the current iteration number, AR is the adjustment rate, and MaxIter is the maximum number of iterations.

[0032] A further technical solution of the present invention is as follows: the new solution vectors of each group are combined into a new solution vector of the total harmonious memory matrix, including:

[0033] (3b-1) Initialize the new solution vector list of the total harmonious memory matrix to be empty, and take out the harmonious memory matrix of the first traffic group in the total harmonious memory matrix;

[0034] (3b-2) Calculate the number of flows n of the traffic group corresponding to the current harmonious memory matrix, and generate a list list from 0 to n-1;

[0035] (3b-3) Initialize the current group solution vector list to be empty; initialize the solution vector interception interval, which is a left-closed and right-open interval, with the left endpoint being 0 and the right endpoint being a random integer between the left endpoint plus 1 and n;

[0036] (3b-4) Generate a random floating point number in the range of [0,1) and determine whether the floating point number is less than or equal to the harmonious memory consideration rate. If it is less than or equal to the harmonious memory consideration rate, execute (3b-5); otherwise, execute (3b-7);

[0037] (3b-5) randomly select a solution vector of the current harmonious memory matrix, intercept the current solution vector according to the solution vector interception interval to generate a new solution vector, and determine one by one whether the elements in the current new solution vector exist in the list according to the list generated in (3b-2). If so, delete the element from the list; if not, randomly select an element in the list to replace the element of the current new solution vector, and delete the element from the list;

[0038] (3b-6) Generate a random floating point number in the range of [0,1) and determine whether the floating point number is less than or equal to the adjustment rate. If it is less than or equal to the adjustment rate, randomly swap the positions of the two elements in the current new solution vector; otherwise, do not perform any operation; execute (3b-8);

[0039] (3b-7) Randomly select an element from the list generated in (3b-2), generate a new solution vector of the intercept interval size, and delete the selected element from the list;

[0040] (3b-8) Add the current new solution vector to the current group solution vector list, and determine whether the current group solution vector list size is less than n. If it is less than n, update the left endpoint of the solution vector interception interval to the right endpoint, and update the right endpoint to a random integer between the left endpoint plus 1 and the list length, and continue to execute (3b-4); otherwise, the new solution vector of the harmonious memory matrix of the current traffic group is generated;

[0041] (3b-9) Add the new solution vector of the harmonious memory matrix of the current traffic group to the new solution vector list of the total harmonious memory matrix, and determine whether the harmonious memory matrix of the current traffic group is the last traffic group. If not, take out the harmonious memory matrix of the next traffic group from the total harmonious memory matrix and execute (3b-2); otherwise, output the new solution vector list of the total harmonious memory matrix as the new total harmonious memory matrix solution vector.

[0042] A further technical solution of the present invention is as follows: determining the start transmission time flow by flow based on the traffic scheduling order, and allocating all traffic instances to corresponding links within the super cycle based on the start transmission time, including:

[0043] (5a) Schedule the first traffic according to the determined traffic scheduling order, and initialize the starting sending time to 0;

[0044] (5b) Determine whether the start sending time of the current stream satisfies the stream period constraint condition. If not, execute (5h);

[0045] (5c) Try to schedule the traffic transmission on the first hop link of the current traffic according to the starting sending time. If a conflict occurs, it means that there is a conflict with the previously scheduled traffic, and execute (5e); if there is no conflict, turn to the next hop link;

[0046] (5d) Schedule the traffic on the current link. If a conflict occurs, it means that there is a conflict with the previously scheduled traffic. Execute (5e); otherwise, execute (5f);

[0047] (5e) Calculate the new start sending time of the current traffic according to the minimum cost conflict resolution criterion and execute (5b);

[0048] (5f) Determine whether the current link is the last hop link of the flow. If so, execute (5g); otherwise, turn to the next hop link and execute (5d);

[0049] (5g) According to the start sending time of the current flow, all instances of the current flow in the super cycle are scheduled to the corresponding links according to the cycle of the flow;

[0050] (5h) Determine whether the traffic scheduling sequence has been traversed. If so, end (5); otherwise, set the start sending time to 0, traverse the next traffic, and execute (5b).

[0051] A further technical solution of the present invention: Step (5g) also includes, during scheduling, if it is found that the time slot interval occupied by the current flow in the link and the time slot interval occupied by the previously scheduled flow in the link can be merged, then the intervals are merged.

[0052] A further technical solution of the present invention is as follows: the method of generating a transmission schedule of each end system and a gating list of each port of the switch based on the scheduling result includes:

[0053] According to the currently traversed sending link, initialize the gating list of the sending port corresponding to the sending link to be empty, select the first time slot interval of the current sending link, and determine whether the left endpoint of the time slot interval is 0. If the left endpoint of the time slot interval is 0, set the gating duration to the length of the time slot interval, and add the gating record "OCCCCCC" and the gating duration to the gating list; if not, first set the gating duration to the left endpoint of the time slot interval, add the gating record "COOOOOOO" and the gating duration to the gating list, set the gating duration to the length of the time slot interval, and add the gating record "OCCCCCC" and the gating duration to the gating list;

[0054] Set the gate duration to the difference between the left endpoint of the current time slot interval and the right endpoint of the previous time slot interval, add the gate record "COOOOOOO" and the gate duration to the gate list; set the gate duration to the length of the time slot interval, add the gate record "OCCCCCC" and the gate duration to the gate list;

[0055] Calculate the difference between the super cycle and the current time slot interval. If the difference is not 0, first set the gating duration to the difference, add the gating record "COOOOOOO" and the gating duration to the gating list; until the time slot interval is the last time slot interval of the current sending link.

[0056] The beneficial effects of the present invention are:

[0057] The present invention provides a method for generating a time-triggered traffic gating list. The method adopts the idea of ​​harmonious search, obtains high-quality traffic scheduling and sorting through global search and iterative optimization mechanism, has a fast solution speed, avoids the limitation of using only fixed rules for traffic sorting and scheduling, and designs the optimal generation mechanism of the first-generation harmonious memory matrix, the dynamic probability adjustment mechanism and the efficient generation mechanism of the new harmonious solution vector, which improves the problems of poor global search, complex iterative operation and slow convergence speed. The details are as follows:

[0058] First, since the present invention is based on the idea of ​​a meta-heuristic algorithm, it can significantly reduce the time complexity of scheduling and planning time-triggered TT traffic in large-scale and high-load time-sensitive networks.

[0059] Secondly, the present invention adopts the idea based on harmonious search to determine the traffic scheduling order, thereby improving the limitation of only using fixed rules to determine the traffic scheduling order.

[0060] Third, the present invention enhances the global searchability of the algorithm, reduces the complexity introduced by iterative operations and improves the convergence rate of the algorithm due to the preprocessing mechanism of traffic period grouping, the optimal generation mechanism of the initial total harmonious memory matrix, the dynamic probability adjustment mechanism and the efficient new harmonious solution vector generation strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The drawings are only for the purpose of illustrating particular embodiments and are not to be considered limiting of the present invention. Like reference symbols denote like components throughout the drawings.

[0062] Figure 1 It is a flow chart for implementing the present invention.

[0063] Figure 2 This is a schematic diagram of network topology modeling in step (1a) of the present invention.

[0064] Figure 3This is a schematic diagram of the initial total harmonious memory matrix generation in step (2) of the present invention (HMS is 2).

[0065] Figure 4 This is a schematic diagram of generating a new solution vector of the harmonious memory matrix in step (3b) of the present invention. DETAILED DESCRIPTION

[0066] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0067] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0068] The overall idea of ​​the present invention is to decompose the process of solving the time-triggered traffic gating list in the time-sensitive network into two stages, namely, the stage of determining the traffic scheduling order and the stage of determining the starting transmission time for each flow according to the determined order. The stage of determining the traffic scheduling order is based on the idea of ​​harmonious search, and high-quality traffic scheduling sorting is obtained by utilizing and fine-tuning the rules in the solution space according to the consideration rules; the stage of determining the starting transmission time for each flow according to the determined order is based on the idea of ​​the earliest transmission of the flow and the minimum cost conflict decomposition, and a scheduling result with high bandwidth utilization is obtained by arranging the transmission of traffic as early as possible and allowing multiple traffic to be aggregated and transmitted as much as possible; finally, the sending schedule of each end system and the gating list of each port of the switch are generated according to the scheduling results.

[0069] In order to enable those skilled in the art to better understand the present invention, the present invention is described in detail below in conjunction with specific embodiments.

[0070] The flowchart for implementing the present invention is shown in the attached Figure 1 As shown, steps (2)-(4) are the stage of determining the traffic scheduling sequence, and step (5) is the stage of determining the starting transmission time for each flow according to the determined sequence.

[0071] (1) Building a Time-Sensitive Network (TSN) Scheduling Model:

[0072] (1a) Define the network topology as a directed graph and number all nodes and links in the network. The numbering method is as shown in the attached figure. Figure 2 As shown;

[0073] (1b) Initialization network switch processing delay d proc , link propagation delay d prop , link bandwidth bw;

[0074] (1c) Initialize the routing parameters of all time-triggered TT flows to be scheduled, and define the routes of all time-triggered TT flows according to the link numbers in (1a);

[0075] (1d) Obtain the period parameters of all time-triggered TT flows to be scheduled, and calculate the lowest common multiple of the periods of all time-triggered TT flows, which is recorded as the super cycle period HC;

[0076] (2) Harmony memory matrix initialization:

[0077] (2a) Initialize the harmonious memory matrix size HMS, the maximum number of iterations MaxIter and set the current number of iterations to 0;

[0078] (2b) grouping all the time-triggered flows to be scheduled according to the periods of all the time-triggered TT flows to be scheduled to form a group set;

[0079] (2c) Initializing the harmony memory matrix of each group according to the grouping situation according to twice the size of the harmony memory matrix HMS, and arranging the harmony memory matrices of each group from small to large according to the flow period to form an initial total harmony memory matrix;

[0080] Specifically, the implementation process of initializing the harmonious memory matrix of each group is as follows:

[0081] (2c-1) traverse the group set and calculate the number of flows n of the current group;

[0082] (2c-2) Use the Fisher-Yates shuffle algorithm to generate random permutations from 0 to n-1;

[0083] (2c-3) Execute (2c-2) 2·HMS times, and use the random permutation generated each time as the row vector to initialize the harmony memory matrix of the current group, denoted as H T , the subscript T represents the period of the current group.

[0084] (2c-4) After the group traversal is completed, the harmony memory matrices of each group are arranged from small to large according to the flow period to form the initial total harmony memory matrix, denoted as H init ,Right now And T1<T2<…<T N , where N is the grouping set size.

[0085] (2d) calculating the harmony fitness of each harmony solution vector in the initial total harmony memory matrix according to the harmony fitness criterion;

[0086] Specifically, the calculation method of the harmonious fitness of each harmonious solution vector in the initial total harmonious memory matrix is:

[0087] (2d-1) Initialize the maximum flow completion time to 0 and the number of scheduling failure flows to 0;

[0088] (2d-2) The harmony solution vector is the row vector of the initial total harmony memory matrix, and the first element of the harmony solution vector is taken out;

[0089] (2d-3) Determine which periodic group the current element belongs to, then select the stream of the periodic group according to the value of the element as the index, and initialize the start sending time sendTime of the stream to 0;

[0090] (2d-4) Determine whether the start sending time of the current stream meets the stream period constraint, that is,

[0091]

[0092] Where hop is the number of routing hops of the current flow, FrameLength is the frame length of the current flow, and T represents the period of the current flow. If it is not satisfied, the number of scheduling failure flows is increased by one, and (2d-10) is executed;

[0093] (2d-5) Try to schedule the flow transmission at the first hop link of the current flow according to the start sending time. If a conflict occurs, execute (2d-7); if there is no conflict, update sendTime to Turn to the next hop link;

[0094] (2d-6) Schedule the traffic on the current link according to sendTime. If a conflict occurs, execute (2d-7); otherwise, execute (2d-8);

[0095] (2d-7) Subtract the right endpoint of the previously scheduled traffic occupied time slot interval from the left endpoint of the currently scheduled traffic occupied time slot interval, recorded as Δ, update the start sending time of the current traffic to sendTime+Δ, and execute (2d-4);

[0096] (2d-8) Determine whether the current link is the last hop link of the flow. If so, execute (2d-9); otherwise, turn to the next hop link and execute (2d-5);

[0097] (2d-9) According to the start sending time of the current flow, all instances of the current flow in the super cycle period are scheduled to the corresponding link. If it is found during scheduling that the time slot interval occupied by the current flow in the link and the time slot interval occupied by the previously scheduled flow in the link can be merged, then the intervals are merged; the maximum completion time of the current flow is calculated using the following calculation method: If the value is greater than the current maximum flow completion time, the maximum flow completion time is updated to this value; otherwise, no operation is performed;

[0098] (2d-10) Determine whether the traversal of the harmonious solution vector elements is complete. If completed, output the maximum flow completion time and the number of scheduling failure flows; otherwise, take out the next element of the harmonious solution vector and execute (2d-3).

[0099] (2e) sorting the harmonious solution vectors in the initial total harmonious memory matrix from best to worst according to the harmonious fitness, and taking the first HMS harmonious solution vectors to form the initial total harmonious memory matrix;

[0100] Specifically, the criteria for harmonious fitness are:

[0101] (2e-1) Compare the number of scheduling failure flows. The smaller the number of scheduling failure flows, the better the harmony fitness.

[0102] (2e-2) When the number of scheduling failure flows is the same, the smaller the maximum flow completion time, the better the harmony fitness;

[0103] (2e-3) When the maximum flow completion time is the same, the harmony fitness quality is the same, and the order is determined according to the original order of the harmony memory matrix.

[0104] (3) Iterative optimization of the total harmonious memory matrix:

[0105] (3a) determining the harmony memory consideration rate HMCR and adjustment rate AR of the current iteration number;

[0106] Specifically, the method for determining the harmonious memory consideration rate HMCR and the adjustment rate AR of the current iteration number is as follows:

[0107] Let the current iteration number be i, then

[0108] (3b) forming new solution vectors of the harmonious memory matrix of each group according to the consideration rule and the fine-tuning rule, forming a new solution vector of the total harmonious memory matrix with the new solution vectors of each group, and calculating the fitness of the solution vector according to the criterion of the harmony fitness;

[0109] Specifically, the method for generating the new solution vector of each group harmony memory matrix and the new solution vector of the total harmony memory matrix is:

[0110] (3b-1) Initialize the new solution vector list of the total harmonious memory matrix to be empty, and take out the harmonious memory matrix of the first traffic group in the total harmonious memory matrix;

[0111] (3b-2) Calculate the number of flows n of the traffic group corresponding to the current harmonious memory matrix, and generate a list list from 0 to n-1;

[0112] (3b-3) Initialize the current group solution vector list to be empty; initialize the solution vector interception interval, which is a left-closed and right-open interval, with the left endpoint being 0 and the right endpoint being a random integer between the left endpoint plus 1 and n;

[0113] (3b-4) Generate a random floating point number in the range of [0,1) and determine whether the floating point number is less than or equal to HMCR. If it is less than or equal to HMCR, execute (3b-5); otherwise, execute (3b-7);

[0114] (3b-5) randomly select a solution vector of the current harmonious memory matrix, intercept the current solution vector according to the solution vector interception interval to generate a new solution vector, and determine one by one whether the elements in the current new solution vector exist in the list according to the list generated in (3b-2). If so, delete the element from the list; if not, randomly select an element in the list to replace the element of the current new solution vector, and delete the element from the list;

[0115] (3b-6) Generate a random floating point number in the range of [0,1) and determine whether the floating point number is less than or equal to AR. If it is less than or equal to AR, randomly swap the positions of the two elements in the current new solution vector; otherwise, do not perform any operation; execute (3b-8);

[0116] (3b-7) Randomly select an element from the list generated in (3b-2), generate a new solution vector of the intercept interval size, and delete the selected element from the list;

[0117] (3b-8) Add the current new solution vector to the current group solution vector list, and determine whether the current group solution vector list size is less than n. If it is less than n, update the left endpoint of the solution vector interception interval to the right endpoint, and update the right endpoint to a random integer between the left endpoint plus 1 and the list length, and continue to execute (3b-4); otherwise, the new solution vector of the harmonious memory matrix of the current traffic group is generated;

[0118] (3b-9) Add the new solution vector of the harmonious memory matrix of the current traffic group to the new solution vector list of the total harmonious memory matrix, and determine whether the harmonious memory matrix of the current traffic group is the last traffic group. If not, take out the harmonious memory matrix of the next traffic group from the total harmonious memory matrix and execute (3b-2); otherwise, output the new solution vector list of the total harmonious memory matrix as the new total harmonious memory matrix solution vector.

[0119] (3c) updating the total harmonious memory matrix. If the fitness of the solution vector is better than the worst fitness of the current total harmonious memory matrix, the solution vector replaces the solution vector corresponding to the worst fitness of the current harmonious memory matrix; otherwise, the total harmonious memory matrix is ​​not updated.

[0120] (3d) Determine whether the current number of iterations has reached the maximum number of iterations MaxIter or whether the iteration stop condition has been reached. If not, increase the current number of iterations by one and execute (3a); otherwise, end the iterative optimization of the total harmonious memory matrix;

[0121] Specifically, the iteration stop condition is:

[0122] If the total harmonious memory matrix has the optimal harmonious fitness after iterating When there is still no change, the iteration stop condition is reached.

[0123] (4) Output the solution vector corresponding to the optimal fitness of the current total harmonious memory matrix, which is the optimized traffic scheduling order;

[0124] (5) Determine the start transmission time for each stream in the determined order:

[0125] (5a) Schedule the first traffic according to the determined traffic scheduling order, and initialize the starting sending time to 0;

[0126] (5b) Determine whether the start sending time of the current stream satisfies the stream period constraint condition. If not, execute (5h);

[0127] Specifically, the flow period constraint is the same as the flow period constraint in (2b-4).

[0128] (5c) Try to schedule the traffic transmission on the first hop link of the current traffic according to the starting sending time. If a conflict occurs, it means that there is a conflict with the previously scheduled traffic, and execute (5e); if there is no conflict, turn to the next hop link;

[0129] (5d) Schedule the traffic on the current link. If a conflict occurs, it means that there is a conflict with the previously scheduled traffic. Execute (5e); otherwise, execute (5f);

[0130] (5e) Calculate the new start sending time of the current traffic according to the minimum cost conflict resolution criterion and execute (5b);

[0131] Specifically, the method for calculating the new start sending time of the current traffic is the same as (2b-7).

[0132] (5f) Determine whether the current link is the last hop link of the flow. If so, execute (5g); otherwise, turn to the next hop link and execute (5d);

[0133] (5g) According to the start transmission time of the current flow, all instances of the current flow in the super cycle period HC are scheduled to the corresponding link according to the cycle of the flow. During scheduling, if it is found that the time slot interval occupied by the current flow in the link can be merged with the time slot interval occupied by the previously scheduled flow in the link, then the interval is merged;

[0134] (5h) Determine whether the traffic scheduling sequence has been traversed. If so, end (5); otherwise, set the start sending time to 0, traverse the next traffic, and execute (5b).

[0135] (6) Generate the sending schedule of each end system and the gating list of each port of the switch based on the scheduling results.

[0136] Specifically, the method for generating the sending schedule of each end system and the gating list of each port of the switch according to the scheduling result is:

[0137] (6a) Traverse the sending links of all end system ports and the sending links of all switch ports;

[0138] (6b) According to the currently traversed sending link, initialize the gating list of the sending port corresponding to the sending link to be empty, select the first time slot interval of the current sending link, and determine whether the left endpoint of the time slot interval is 0. If the left endpoint of the time slot interval is 0, set the gating duration to the length of the time slot interval, and add the gating record "OCCCCCC" (O means that the sending port queue is allowed to transmit, and C means that the sending port queue is allowed to transmit, from left to right, representing the queues of the sending port numbers 7 to 0) and the gating duration to the gating list; if not, first set the gating duration to the left endpoint of the time slot interval, add the gating record "COOOOOOO" and the gating duration to the gating list, and then set the gating duration to the length of the time slot interval, and add the gating record "OCCCCCC" and the gating duration to the gating list;

[0139] (6c) Determine whether the time slot interval is the last time slot interval of the current transmission link. If not, select the next time slot interval and execute 4); otherwise, execute 5);

[0140] (6d) First, set the gate duration to the difference between the left endpoint of the current time slot interval and the right endpoint of the previous time slot interval, add the gate record "COOOOOOO" and the gate duration to the gate list, then set the gate duration to the length of the time slot interval, add the gate record "OCCCCCC" and the gate duration to the gate list, and execute 3);

[0141] (6e) Calculate the difference between the super period HC and the current time slot interval. If the difference is not 0, first set the gate duration to the difference, and add the gate record "COOOOOOO" and the gate duration to the gate list;

[0142] (6f) The gating list of the current sending port is generated.

[0143] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should be included in the protection scope of the present invention.

Claims

1. A method for generating a time-triggered traffic gating list, characterized in that: include: Obtaining the periods of all time-triggered flows to be scheduled, and grouping all time-triggered flows to be scheduled based on the periods of all time-triggered flows to be scheduled; Initialize the harmony memory matrix of each group, arrange the harmony memory matrix of each group from small to large according to the flow period, and form an initial total harmony memory matrix; Calculate the harmonious fitness of each harmonious solution vector in the initial total harmonious memory matrix, sort each harmonious solution vector from best to worst according to the harmonious fitness, and take the first HMS harmonious solution vectors to form the initial total harmonious memory matrix; where HMS is the size of the initialized harmonious memory matrix; Iterate based on the initial total harmonious memory matrix, and output the solution vector corresponding to the optimal fitness of the total harmonious memory matrix when the iteration ends, which is the optimized traffic scheduling order; Determine the start transmission time for each flow based on the traffic scheduling order, allocate all traffic instances to corresponding links within the super cycle based on the start transmission time, and generate a scheduling result; Based on the scheduling results, a sending schedule for each end system and a gating list for each port of the switch are generated.

2. According to claim 1, a method for generating a time-triggered traffic gating list is characterized in that: The initializing the harmonious memory matrix of each group includes: Calculate the number of flows n of the current group; Use the Fisher-Yates shuffle algorithm to generate random permutations from 0 to n-1; Execute the shuffle algorithm 2·HMS times, using the random permutation generated each time as a row vector to initialize the harmony memory matrix of the current group.

3. According to claim 1, a method for generating a time-triggered traffic gating list is characterized in that: The quality criteria of the harmonious adaptability are: Compare the number of scheduling failure flows. The smaller the number of scheduling failure flows, the better the harmony fitness. When the number of scheduling failure flows is the same, the smaller the maximum flow completion time, the better the harmony fitness; When the maximum flow completion time is the same, the harmonious fitness quality is the same, and the order is determined according to the original order of the harmonious memory matrix.

4. According to claim 1, a method for generating a time-triggered traffic gating list is characterized in that: The iteration based on the initial total harmonious memory matrix and outputting the solution vector corresponding to the optimal fitness of the total harmonious memory matrix when the iteration is terminated include: Determine the harmony memory consideration rate and adjustment rate for the current iteration number; Based on the harmonious memory consideration rate and adjustment rate of the current iteration number, new solution vectors of the harmonious memory matrix of each group are composed according to the consideration rules and fine-tuning rules respectively, and the new solution vectors of each group are combined into a new solution vector of the total harmonious memory matrix. The fitness of the solution vector is calculated according to the harmonious fitness criterion.

5. According to claim 4, a method for generating a time-triggered traffic gating list is characterized in that: The determination of the harmonious memory consideration rate and adjustment rate of the current number of iterations is specifically as follows: Let the current iteration number be i, then Among them, HMCR is the harmonious memory consideration rate of the current iteration number, AR is the adjustment rate, and MaxIter is the maximum number of iterations.

6. According to claim 4, a method for generating a time-triggered traffic gating list is characterized in that: The new solution vectors of each group are combined into a new solution vector of the total harmonious memory matrix, including: (3b-1) Initialize the new solution vector list of the total harmonious memory matrix to be empty, and take out the harmonious memory matrix of the first traffic group in the total harmonious memory matrix; (3b-2) Calculate the number of flows n of the traffic group corresponding to the current harmonious memory matrix, and generate a list list from 0 to n-1; (3b-3) Initialize the current group solution vector list to be empty; initialize the solution vector interception interval, which is a left-closed and right-open interval, with the left endpoint being 0 and the right endpoint being a random integer between the left endpoint plus 1 and n; (3b-4) Generate a random floating point number in the range of [0,1) and determine whether the floating point number is less than or equal to the harmonious memory consideration rate. If it is less than or equal to the harmonious memory consideration rate, execute (3b-5); otherwise, execute (3b-7); (3b-5) randomly select a solution vector of the current harmonious memory matrix, intercept the current solution vector according to the solution vector interception interval to generate a new solution vector, and determine one by one whether the elements in the current new solution vector exist in the list according to the list generated in (3b-2). If so, delete the element from the list; if not, randomly select an element in the list to replace the element of the current new solution vector, and delete the element from the list; (3b-6) Generate a random floating point number in the range of [0,1) and determine whether the floating point number is less than or equal to the adjustment rate. If it is less than or equal to the adjustment rate, randomly swap the positions of the two elements in the current new solution vector; otherwise, do not perform any operation; execute (3b-8); (3b-7) Randomly select an element from the list generated in (3b-2), generate a new solution vector of the intercept interval size, and delete the selected element from the list; (3b-8) Add the current new solution vector to the current group solution vector list, and determine whether the current group solution vector list size is less than n. If it is less than n, update the left endpoint of the solution vector interception interval to the right endpoint, and update the right endpoint to a random integer between the left endpoint plus 1 and the list length, and continue to execute (3b-4); otherwise, the new solution vector of the harmonious memory matrix of the current traffic group is generated; (3b-9) Add the new solution vector of the harmonious memory matrix of the current traffic group to the new solution vector list of the total harmonious memory matrix, and determine whether the harmonious memory matrix of the current traffic group is the last traffic group. If not, take out the harmonious memory matrix of the next traffic group from the total harmonious memory matrix and execute (3b-2); otherwise, output the new solution vector list of the total harmonious memory matrix as the new total harmonious memory matrix solution vector.

7. According to claim 1, a method for generating a time-triggered traffic gating list is characterized in that: The method of determining the start transmission time flow by flow based on the traffic scheduling order, and allocating all traffic instances to corresponding links within the super cycle period based on the start transmission time, includes: (5a) Schedule the first traffic according to the determined traffic scheduling order, and initialize the starting sending time to 0; (5b) Determine whether the start sending time of the current stream satisfies the stream period constraint condition. If not, execute (5h); (5c) Try to schedule the traffic transmission on the first hop link of the current traffic according to the starting sending time. If a conflict occurs, it means that there is a conflict with the previously scheduled traffic, and execute (5e); if there is no conflict, turn to the next hop link; (5d) Schedule the traffic on the current link. If a conflict occurs, it means that there is a conflict with the previously scheduled traffic. Execute (5e); otherwise, execute (5f); (5e) Calculate the new start sending time of the current traffic according to the minimum cost conflict resolution criterion and execute (5b); (5f) Determine whether the current link is the last hop link of the flow. If so, execute (5g); otherwise, turn to the next hop link and execute (5d); (5g) According to the start sending time of the current flow, all instances of the current flow in the super cycle are scheduled to the corresponding links according to the cycle of the flow; (5h) Determine whether the traffic scheduling sequence has been traversed. If so, end (5); otherwise, set the start sending time to 0, traverse the next traffic, and execute (5b).

8. A method for generating a time-triggered traffic gating list according to claim 7, characterized in that: Step (5g) also includes, during scheduling, if it is found that the time slot interval occupied by the current flow in the link and the time slot interval occupied by the previously scheduled flow in the link can be merged, then merging the intervals.

9. The method for generating a time-triggered traffic gating list according to claim 1, characterized in that: The method of generating a sending schedule of each end system and a gating list of each port of the switch based on the scheduling result includes: According to the currently traversed sending link, initialize the gating list of the sending port corresponding to the sending link to be empty, select the first time slot interval of the current sending link, and determine whether the left endpoint of the time slot interval is 0. If the left endpoint of the time slot interval is 0, set the gating duration to the length of the time slot interval, and add the gating record "OCCCCCC" and the gating duration to the gating list; if not, first set the gating duration to the left endpoint of the time slot interval, add the gating record "COOOOOOO" and the gating duration to the gating list, set the gating duration to the length of the time slot interval, and add the gating record "OCCCCCC" and the gating duration to the gating list; Set the gate duration to the difference between the left endpoint of the current time slot interval and the right endpoint of the previous time slot interval, add the gate record "COOOOOOO" and the gate duration to the gate list; set the gate duration to the length of the time slot interval, add the gate record "OCCCCCC" and the gate duration to the gate list; Calculate the difference between the super cycle and the current time slot interval. If the difference is not 0, first set the gating duration to the difference, add the gating record "COOOOOOO" and the gating duration to the gating list; until the time slot interval is the last time slot interval of the current sending link.

Images