A time-triggered network schedule generation method based on sequential relationship constraints

Abstract

The application relates to the technical field of computer communication, and discloses a time-triggered network schedule table generation method based on sequence relationship constraints. According to a time-triggered Ethernet time-triggered traffic scheduling mechanism, a constraint relationship between data frame scheduling time points is constructed. According to a message transceiving mechanism of a task, a message sequence relationship constraint is established for a message. According to a time point at which the task sends the message, a message generation time constraint is established. The message sequence relationship constraint and the message generation time constraint are added to a schedule table generation process. The generated schedule table meets the requirement that, under a specific application scenario, after a task receives a message, the task needs to process data and send a message to complete a response within a limited time, or after the task sends a request, the task needs to obtain a response within a limited time. In addition, the schedule points of the generated schedule table can be matched with the time points at which the task sends a message sending request, so that the task running performance is improved.

Description

Technical Field

[0001] This invention relates to the field of computer communication technology and discloses a method for generating time-triggered network scheduling tables based on sequential relationship constraints. Background Technology

[0002] Message: A message refers to the process of data transmission from a source node to a destination node at the network layer. The data in a message is generated at the source node and received and processed at the destination node.

[0003] Data frame: A data frame is a data unit transmitted from the sending node to the receiving node in the data link layer. Multiple data frames are relayed in the data link layer to form a message transmission process.

[0004] Time-triggered Ethernet, based on global clock synchronization, uses a scheduling table to control the scheduling of time-triggered traffic, achieving highly real-time and reliable network communication. The performance of the scheduling table determines the scheduling performance of time-triggered network traffic.

[0005] In specific application scenarios, after receiving a message, a task needs to process the data and send a response within a limited time, or a task needs to obtain a response within a limited time after issuing a request. However, current scheduling table generation algorithms only consider the message scheduling process in the network, without considering the task's message sending and receiving mechanism or the task's message sending time, which may lead to a significant deviation between the task message sending time and the message scheduling time. Summary of the Invention

[0006] The purpose of this invention is to provide a time-triggered network scheduling table generation method based on sequential relationship constraints, which can meet the needs of tasks in specific application scenarios to process data and send messages to complete the response within a limited time after receiving a message, or to obtain a response within a limited time after a task issues a request. Furthermore, it can match the scheduling points of the generated scheduling table with the time point when the task makes a message sending request, thereby improving task running performance.

[0007] To achieve the above-mentioned technical effects, the technical solution adopted by the present invention is as follows:

[0008] A method for generating a time-triggered network scheduling table based on sequential relationship constraints, comprising:

[0009] Parse the input data to obtain the link delay of the input data, as well as the sending delay, receiving delay, and forwarding delay of the nodes; the input data includes time-triggered Ethernet node data, time-triggered Ethernet network link data, time-triggered message data, data frame data, time-triggered message sequence relationship, and message generation time;

[0010] Based on the generation requirements of the time-triggered network scheduling table, a functional relationship between data frame scheduling time points is established as a solution constraint. The solution constraints include periodic constraints, conflict-free constraints, switch forwarding constraints, and total delay constraints.

[0011] Based on the time-triggered message order relationship and message generation time, generate message order relationship constraints and message generation time constraints;

[0012] By solving the constraints between data frame scheduling time points, as well as the constraints on message order and message generation time, the solution of the data frame scheduling time point variable group is obtained. Based on the solution of the message scheduling time point variable group, the scheduling time point of each data frame is calculated.

[0013] Furthermore, the periodic constraint is ,in: For the first The sending period of the message, frameSndPit i(j) is the message sending period of the i-th message. The first message The data frame scheduling time point for each data frame.

[0014] Furthermore, the conflict-free constraint is: for any two data frames A and B that traverse the same link, or ,in, For data frame A, the data frame scheduling time point. For data frame B, the data frame scheduling time point. For the transmission delay of data frame A, This is the transmission delay for data frame B.

[0015] Furthermore, the switch forwarding constraint is: frameSndPit i(j) > frameSndPit i(j - 1) + SndDelay i(j - 1) + WireDelay i(j - 1) + rcvDelay i(j - 1) + transDelay i(j - 1) + relayDelay i(j - 1, j) and frameSndPit i(j) < frameSndPit i(j - 1) + Delay max, where frameSndPit i(j) is the data frame scheduling time point of the j-th data frame of the i-th message, sndDelay i(j - 1) is the transmission delay of the (j - 1)-th data frame sending node of the i-th message; wireDelay i(j - 1) is the line delay of the link through which the (j - 1)-th data frame of the i-th message passes; rcvDelay i(j - 1) is the reception delay of the (j - 1)-th data frame receiving node of the i-th message; relayDelay i(j - 1, j) is the forwarding delay between the (j - 1)-th data frame and the j-th data frame of the i-th message, and Delay max is the maximum forwarding delay of the switch.

[0016] Furthermore, the total delay constraint: For the sending node of data frame C and the corresponding message source node being the same, the receiving node of data frame D and the corresponding message destination node being the same, and the IDs of the corresponding messages of data frame C and data frame D being the same, then frameSndPitD - frameSndPitC < tDelay, where frameSndPitC is the data frame scheduling time point of data frame C, frameSndPitD is the data frame scheduling time point of data frame D, and tDelay is the total delay of the corresponding messages of data frame C and data frame D.

[0017] Furthermore, the message order constraint is:

[0018] For two messages Mi and Mj with a send-receive order relationship on the same node, where Mi is the message to be sent and Mj is the message to be received, there is minGap(Mi,Mj) < frameSndPit(Mj,last) - frameSndPit(Mi,first) ≤ maxGap(Mi,Mj), where frameSndPit(Mj,last) is the scheduling time point of the last data frame of message Mj, frameSndPit(Mi,first) is the scheduling time point of the first data frame of message Mi, minGap(Mi,Mj) is the minimum interval in the order relationship between message Mi and message Mj, and maxGap(Mi,Mj) is the maximum interval in the order relationship between message Mi and message Mj;

[0019] For two messages Mk and Ml that have a receive-send order relationship on the same node, where Mk is the message to be received and Ml is the message to be sent, we have minGap (Mk,Ml) < frameSndPit (Ml,first) - frameSndPit(Mk,last) ≤ maxGap (Mk,Ml), where frameSndPit (Ml,first) is the scheduling time point of the first data frame of message Ml, frameSndPit (Mk,last) is the scheduling time point of the last data frame of message Mk, minGap (Mk,Ml) is the minimum interval in the order relationship between messages Mk and Ml, and maxGap (Mk,Ml) is the maximum interval in the order relationship between messages Mk and Ml.

[0020] Furthermore, the message generation time constraint is as follows: for the first data frame A0 of any message, we have: ;

[0021] Where frameSndPitA0 is the data frame scheduling time point of data frame A0, and creatTime is the message generation time. This is the preset upper limit difference.

[0022] Furthermore, during the analysis of the solution to the variable set at the data frame scheduling time point, if the solution fails, adjustments are made. The value is adjusted as follows: ,in This is the adjusted preset upper limit difference. For increments greater than 1, ceil indicates rounding up.

[0023] Furthermore, the constraint relationships between message scheduling time points are written into the SMT solver. The solution of the message scheduling time point variable group is obtained by analyzing the SMT solver. The scheduling time point of each data frame is calculated based on the solution of the message scheduling time point variable group, and the results are recorded and output.

[0024] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention constructs constraint relationships between data frame scheduling points based on the time-triggered Ethernet time-triggered traffic scheduling mechanism, establishes message order relationship constraints for messages based on the task's message sending and receiving mechanism, and establishes message generation time constraints based on the task's message sending time. These message order relationship constraints and message generation time constraints are incorporated into the scheduling table generation process. The generated scheduling table satisfies the requirements of specific application scenarios where tasks need to process data and send messages within a limited time after receiving messages to complete the response, or where tasks need to obtain a response within a limited time after issuing a request. This avoids the problem of significant deviations between the task message sending time and the message scheduling time. Furthermore, it enables the scheduling points of the generated scheduling table to match the time when the task makes a message sending request, improving task performance. Attached Figure Description

[0025] Figure 1 This is a flowchart of the time-triggered network scheduling table generation method based on sequential relationship constraints in Example 1 or 2; Detailed Implementation

[0026] The present invention will now be described in further detail with reference to the embodiments and accompanying drawings. However, this should not be construed as limiting the scope of the above-described subject matter of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.

[0027] Example 1

[0028] See Figure 1 A method for generating a time-triggered network scheduling table based on sequential relationship constraints, comprising:

[0029] Parse the input data to obtain the link delay of the input data, as well as the sending delay, receiving delay, and forwarding delay of the nodes; the input data includes time-triggered Ethernet node data, time-triggered Ethernet network link data, time-triggered message data, data frame data, time-triggered message sequence relationship, and message generation time;

[0030] Based on the generation requirements of the time-triggered network scheduling table, a functional relationship between data frame scheduling time points is established as a solution constraint. The solution constraints include periodic constraints, conflict-free constraints, switch forwarding constraints, and total delay constraints.

[0031] Based on the time-triggered message order relationship and message generation time, generate message order relationship constraints and message generation time constraints;

[0032] By solving the constraints between data frame scheduling time points, as well as the constraints on message order and message generation time, the solution of the data frame scheduling time point variable group is obtained. Based on the solution of the message scheduling time point variable group, the scheduling time point of each data frame is calculated.

[0033] In this embodiment, a constraint relationship between data frame scheduling time points is constructed based on the time-triggered Ethernet time-triggered traffic scheduling mechanism. A message order relationship constraint is established for messages based on the task's message sending and receiving mechanism. A message generation time constraint is established based on the time point when the task sends a message. The message order relationship constraint and the message generation time constraint are added to the scheduling table generation process. The generated scheduling table meets the needs of specific application scenarios where the task needs to process the data and send a message to complete the response within a limited time after receiving the message, or the task needs to obtain a response within a limited time after issuing a request. In addition, it can match the scheduling points of the generated scheduling table with the time point when the task makes a message sending request, thereby improving the task running performance.

[0034] Example 2

[0035] See Figure 1 A method for generating a time-triggered network scheduling table based on sequential relationship constraints, comprising:

[0036] Step 1: Parse the input data to obtain the link delay, transmission delay, reception delay, and forwarding delay of the node; the input data includes time-triggered Ethernet node data, time-triggered Ethernet network link data, time-triggered message data, data frame data, time-triggered message sequence relationship, and message generation time;

[0037] This embodiment uses files as input, including time-triggered Ethernet node data, time-triggered Ethernet network link data, time-triggered message data, data frame data, time-triggered message sequence relationships, and message generation times. By parsing the imported data, a list of time-triggered Ethernet node attributes (as shown in Table 1), a list of time-triggered Ethernet link attributes (as shown in Table 2), a list of time-triggered message attributes (as shown in Table 3), a list of data frame attributes (as shown in Table 4), a list of message sequence relationships (as shown in Table 5), and time-triggered message generation time data (as shown in Table 6) are established.

[0038] Table 1 List of Time-Triggered Ethernet Nodes

[0039]

[0040] Table 2 List of Time-Triggered Ethernet Links

[0041]

[0042] Table 3. List of Time-Triggered Message Attributes

[0043]

[0044] Table 4 Data Frame Attribute List

[0045]

[0046] Table 5 Message Order Relationship List

[0047]

[0048] Table 6 Message Generation Time List

[0049]

[0050] The relational schema includes send-receive order relations and receive-send order relations. Send-receive order relations are used when a task receives a response within a required time after sending a message; receive-send order relations are used when a task receives data, processes it for a period of time, and then responds within a required time.

[0051] Step 2: Based on the generation requirements of the time-triggered network scheduling table, establish the functional relationship between data frame scheduling time points as the solution constraint. The solution constraint includes periodic constraint, conflict-free constraint, switch forwarding constraint, and total delay constraint.

[0052] In this embodiment:

[0053] The periodic constraint is ,in: For the first The sending period of the message, frameSndPit i(j) is the message sending period of the i-th message. The first message The scheduling time points for each data frame are guaranteed to be within the legal value range (within one cycle of the data frame).

[0054] The conflict-free constraint is: for any two data frames A and B that pass through the same link, or ,in, For data frame A, the data frame scheduling time point. For data frame B, the data frame scheduling time point. For the transmission delay of data frame A, This is the transmission delay for data frame B. It ensures that there are no scheduling conflicts between data frames on each link.

[0055] The switch forwarding constraint is: frameSndPit i(j) > frameSndPit i(j - 1) + SndDelay i(j - 1) + WireDelay i(j - 1) + rcvDelay i(j - 1) + transDelay i(j - 1) + relayDelay i(j - 1, j) and frameSndPit i(j) < frameSndPit i(j - 1) + Delay max, where frameSndPit i(j) is the data frame scheduling time point of the j-th data frame of the i-th message, sndDelay i(j - 1) is the transmission delay of the j - 1-th data frame of the i-th message at the sending node; wireDelay i(j - 1) is the line delay of the link through which the j - 1-th data frame of the i-th message passes; rcvDelay i(j - 1) is the reception delay of the j - 1-th data frame of the i-th message at the receiving node; relayDelay i(j - 1, j) is the forwarding delay between the j - 1-th data frame and the j-th data frame of the i-th message, and Delay max is the maximum forwarding delay of the switch. When a message is forwarded by the switch, the scheduling point of the data frame entering the switch should be before the scheduling point of the data frame sent out by the switch, and a certain interval should be ensured to meet the requirements of the transmission delay, reception delay, line delay, and switch forwarding delay of the node.

[0056] The total delay constraint is: for the sending node of data frame C being the same as the corresponding message source node, the receiving node of data frame D being the same as the corresponding message destination node, and the IDs of the corresponding messages of data frame C and data frame D being the same, then frameSndPitD - frameSndPitC < tDelay, where frameSndPitC is the data frame scheduling time point of data frame C, frameSndPitD is the data frame scheduling time point of data frame D, and tDelay is the total delay of the corresponding messages of data frame C and data frame D. It can be determined that the interval between the scheduling points of the first data frame and the last data frame of a message in the network is within the threshold to ensure the total delay link of the message in the network.

[0057] Step 3, generate the message sequence relationship constraint and the message generation time constraint according to the time-triggered message sequence relationship and the message generation time;

[0058] In this embodiment:

[0059] The message sequence constraint is:

[0060] For two messages Mi and Mj that have a send-receive order relationship on the same node, where Mi is the message to be sent and Mj is the message to be received, we have minGap(Mi,Mj)<frameSndPit(Mj,last)-frameSndPit(Mi,first)≤maxGap(Mi,Mj), where frameSndPit(Mj,last) is the scheduling time point of the last data frame of message Mj, frameSndPit(Mi,first) is the scheduling time point of the first data frame of message Mi, minGap(Mi,Mj) is the minimum interval in the order relationship between messages Mi and Mj, and maxGap(Mi,Mj) is the maximum interval in the order relationship between messages Mi and Mj.

[0061] For two messages Mk and Ml that have a receive-send order relationship on the same node, where Mk is the message to be received and Ml is the message to be sent, we have minGap (Mk,Ml) < frameSndPit (Ml,first) - frameSndPit(Mk,last) ≤ maxGap (Mk,Ml), where frameSndPit (Ml,first) is the scheduling time point of the first data frame of message Ml, frameSndPit (Mk,last) is the scheduling time point of the last data frame of message Mk, minGap (Mk,Ml) is the minimum interval in the order relationship between messages Mk and Ml, and maxGap (Mk,Ml) is the maximum interval in the order relationship between messages Mk and Ml.

[0062] The message generation time constraint is as follows: For the first data frame A0 of any message, we have: Where frameSndPitA0 is the data frame scheduling time point of data frame A0, and creatTime is the message generation time. The preset upper limit difference is a relatively small value, which can be taken as 1 in practical applications. During the analysis and obtaining of the solution for the variable set at the data frame scheduling time point, if the solution fails, adjustments are made. The value is adjusted as follows: ,in This is the adjusted preset upper limit difference. For increments greater than 1, ceil indicates rounding up.

[0063] Step 4: By solving the constraints between data frame scheduling time points, as well as the constraints on message order and message generation time, analyze and obtain the solution of the data frame scheduling time point variable group. Calculate the scheduling time point of each data frame based on the solution of the message scheduling time point variable group.

[0064] In this embodiment, the constraint relationship between message scheduling time points is written into the SMT solver. The solution of the message scheduling time point variable group is obtained by analyzing the SMT solver. The scheduling time point of each data frame is calculated based on the solution of the message scheduling time point variable group. The results are recorded and output into the scheduling table.

[0065] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for generating a time-triggered network scheduling table based on sequential relationship constraints, characterized in that, Including: Analyze the input data to obtain the link line delay of the input data, as well as the transmission delay, reception delay, and forwarding delay of the nodes; the input data includes time-triggered Ethernet node data, time-triggered Ethernet network link data, time-triggered message data, data frame data, time-triggered message sequence relationship, and message generation time; According to the generation requirements of the time-triggered network schedule, establish a functional relationship between data frame scheduling time points as a solution constraint, and the solution constraints include cycle constraint, conflict-free constraint, switch forwarding constraint, and total delay constraint; Generate message sequence relationship constraints and message generation time constraints according to the time-triggered message sequence relationship and message generation time; Through the solution constraints between data frame scheduling time points, as well as message sequence relationship constraints and message generation time constraints, analyze and obtain the solution of the data frame scheduling time point variable group, and calculate each data frame scheduling time point according to the solution of the message scheduling time point variable group.

2. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 1, characterized in that, The periodic constraint is ,in: For the first The sending period of the message, frameSndPit i(j) is the message sending period of the i-th message. The first message The data frame scheduling time point for each data frame.

3. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 1, characterized in that, The conflict-free constraint is: for any two data frames A and B that pass through the same link, or ,in, For data frame A, the data frame scheduling time point. For data frame B, the data frame scheduling time point. For the transmission delay of data frame A, This is the transmission delay for data frame B.

4. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 1, characterized in that, The switch forwarding constraint is: frameSndPit i(j) > frameSndPit i(j - 1) + SndDelay i(j - 1) + WireDelay i(j - 1) + rcvDelay i(j - 1) + transDelay i(j - 1) + relayDelay i(j - 1, j) and frameSndPit i(j) < frameSndPit i(j - 1) + Delay max, where frameSndPit i(j) is the data frame scheduling time point of the j-th data frame of the i-th message, sndDelay i(j - 1) is the transmission delay of the sending node of the (j - 1)-th data frame of the i-th message; wireDelay i(j - 1) is the link line delay of the (j - 1)-th data frame of the i-th message; rcvDelay i(j - 1) is the reception delay of the receiving node of the (j - 1)-th data frame of the i-th message; relayDelay i(j - 1, j) is the forwarding delay between the (j - 1)-th data frame and the j-th data frame of the i-th message, Delay max is the maximum forwarding delay of the switch; transDelay i(j - 1) is the transmission delay of the (j - 1)-th data frame of the i-th message.

5. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 1, characterized in that, The total delay constraint: For the sending node of data frame C and the corresponding message source node being the same, the receiving node of data frame D and the corresponding message destination node being the same, and the IDs of the corresponding messages of data frame C and data frame D being the same, then frameSndPitD - frameSndPitC < tDelay, where frameSndPitC is the data frame scheduling time point of data frame C, frameSndPitD is the data frame scheduling time point of data frame D, and tDelay is the total delay of the corresponding messages of data frame C and data frame D.

6. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 1, characterized in that, The message sequence constraint is: For two messages Mi and Mj that have a send-receive order relationship on the same node, where Mi is the message to be sent and Mj is the message to be received, we have minGap(Mi,Mj)<frameSndPit(Mj,last)-frameSndPit(Mi,first)≤maxGap(Mi,Mj), where frameSndPit(Mj,last) is the scheduling time point of the last data frame of message Mj, frameSndPit(Mi,first) is the scheduling time point of the first data frame of message Mi, minGap(Mi,Mj) is the minimum interval in the order relationship between messages Mi and Mj, and maxGap(Mi,Mj) is the maximum interval in the order relationship between messages Mi and Mj. For two messages Mk and Ml that have a receive-send order relationship on the same node, where Mk is the message to be received and Ml is the message to be sent, we have minGap (Mk,Ml) < frameSndPit (Ml,first) - frameSndPit (Mk,last) ≤ maxGap (Mk,Ml), where frameSndPit (Ml,first) is the scheduling time point of the first data frame of message Ml, frameSndPit (Mk,last) is the scheduling time point of the last data frame of message Mk, minGap (Mk,Ml) is the minimum interval in the order relationship between messages Mk and Ml, and maxGap (Mk,Ml) is the maximum interval in the order relationship between messages Mk and Ml.

7. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 1, characterized in that, The message generation time constraint is as follows: For the first data frame A0 of any message, we have: ; Where frameSndPitA0 is the data frame scheduling time point of data frame A0, and creatTime is the message generation time. This is the preset upper limit difference.

8. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 7, characterized in that, If the solution to the variable set at the time point of the data frame scheduling is not found during the analysis process, adjustments should be made. The value is adjusted as follows: ,in This is the adjusted preset upper limit difference. For increments greater than 1, ceil indicates rounding up.

9. The method for generating a time-triggered network scheduling table based on sequential relationship constraints according to claim 1, characterized in that, The constraints between message scheduling time points are written into the SMT solver. The solution of the message scheduling time point variable group is obtained by analyzing the SMT solver. The scheduling time point of each data frame is calculated based on the solution of the message scheduling time point variable group, and the results are recorded and output.

Images