A time synchronization method for distributed simulation systems based on step length

Abstract

The present invention discloses a time synchronization method for a distributed simulation system based on a step length, comprising: determining a synchronization step length between different simulation nodes; determining a local time synchronization step length t between each simulation node on each computer; determining a distributed simulation system time synchronization step length T; requiring T to be an integer multiple of t; on each computer, each simulation node performs synchronous operation according to the local time synchronization step length t, and each simulation node runs to a local synchronization baseline; when all simulation nodes reach the local synchronization baseline for the nth time and satisfy the condition n*t=T, a signal is sent to other computers in the network indicating that the computer has run to the next distributed synchronization baseline, and a signal is checked to determine whether all other computers have received the signal that they have run to the next distributed synchronization baseline. The present invention can reduce the impact of network delay, reduce the time synchronization overhead between virtual simulation nodes, and thus improve the performance of the distributed simulation system.

Description

Technical Field

[0001] The invention belongs to the field of computer technology, and in particular relates to a time synchronization method for a distributed simulation system based on step length. Background Art

[0002] With the rapid development of computer technology, network technology, virtualization technology and other technologies, multiple simulation devices distributed in space are connected together through communication networks to build a distributed simulation system, which can coordinate and complete complex simulation tasks with distributed characteristics.

[0003] A fundamental challenge facing distributed simulation systems is ensuring that the order of events processed does not deviate from the order of events in the physical system while performing parallel simulations on multiple computers in a network. This requires time synchronization between virtual simulation nodes in a distributed simulation system to ensure the correct timing of software tasks.

[0004] In view of the strong real-time requirements of embedded software in fields such as aerospace, the time synchronization of distributed simulation systems must reach the millisecond or even microsecond level. Therefore, controlling the time synchronization overhead caused by network delays is a key factor in ensuring the performance of distributed simulation systems.

[0005] One of the most common synchronization strategies in distributed simulation systems is the conservative time advancement mechanism. It ensures the logical correctness of events and the logical order of messages sent during the simulation, ensuring that events are processed and executed in chronological order. Currently, various time synchronization methods based on the conservative time advancement mechanism have been developed to control synchronization overhead.

[0006] Patent No. CN201410785390.8 applied for by Huazhong University of Science and Technology discloses a team simulation time synchronization method based on a time source. The method discloses a team simulation time synchronization method, which is characterized by: specifically: a manager entity joins the federation, first initializes the simulation start time, time advancement step and simulation start step number, and then starts to advance time; a common simulation entity joins the federation, receives time synchronization parameters from the manager entity, the time synchronization parameters include the simulation start time, time advancement step and simulation start step number, calculates its own current simulation time based on the time synchronization parameters, and then starts to advance time; the manager entity changes the time advancement step, advances time according to the changed time advancement step, updates the time synchronization parameters, that is, updates the simulation start step number to the simulation step number corresponding to the change of the time advancement step, and updates the simulation start time to the current simulation time of the manager entity; at the same time, the updated time synchronization parameters are group-sent to common simulation entities, and the common simulation entities calculate the current time and advance time based on the time synchronization parameters.

[0007] The patent applied for by Huazhong University of Science and Technology adopts a unified time advancement step for all simulation entities in the entire system. For strong real-time embedded software in fields such as aerospace, the time synchronization requirements of distributed simulation systems must reach one millisecond, hundreds of microseconds, or even tens of microseconds. However, in the current gigabit network, reliable communication transmission between two computers on the network sometimes takes hundreds of microseconds, which will cause network communication delay to become a key factor restricting the performance of distributed simulation systems.

[0008] Patent applicant Yu Yongli filed patent application number CN201210147868.5, disclosing a small-step, long-time advancement algorithm based on a conservative mechanism. This algorithm, based on the conservative time advancement mechanism in the high-level architecture, employs a method of inserting empty events timestamped with small steps when a federation member cannot determine the occurrence time of an event during time advancement. Once the federation member can determine the event's occurrence time, the small-step, long-time advancement is canceled and normal time advancement resumes. This algorithm addresses the issue of federation members violating the predictability principle of the conservative time advancement mechanism during federation simulation due to federation members being affected by the average guaranteed delay.

[0009] The aforementioned patent addresses the impact of excessively small step sizes on system performance by employing a variable step size. The goal is to use small step sizes only at specific moments, thus minimizing the impact of prolonged small step synchronization on system performance. However, for highly real-time embedded software in fields such as aerospace, this cannot meet the requirements of certain system events that may require constant small step synchronization. Summary of the Invention

[0010] The technical problem solved by the present invention is: to overcome the shortcomings of the existing technology and provide a step-based distributed simulation system time synchronization method, which can reduce the impact of network delay and reduce the time synchronization overhead between virtual simulation nodes, thereby improving the performance of the distributed simulation system.

[0011] The object of the present invention is achieved through the following technical solutions: A time synchronization method for a distributed simulation system based on step length, comprising: (1) determining the synchronization step length between different simulation nodes; (2) according to the synchronization step length between different simulation nodes, preferentially deploying simulation nodes with smaller time synchronization step lengths on one or more computers, and deploying other nodes according to the computer load, and determining the local time synchronization step length t between each simulation node on each computer during deployment; (3) determining the time synchronization step length T of the distributed simulation system; (4) adjusting the time synchronization step length T of the distributed simulation system and the local time synchronization step length t between each simulation node on each computer, requiring T to be an integer multiple of t; (5) on each computer, each The simulation nodes are synchronized according to the local time synchronization step t, and each simulation node runs to the local synchronization baseline; when all simulation nodes reach the local synchronization baseline for the nth time and meet the condition n*t=T, a signal is sent to other computers in the network that the machine has run to the next distributed synchronization baseline, and a check is made to see whether a signal has been received from all other computers that the machine has run to the next distributed synchronization baseline; (6) if a computer has received a signal from all other computers that the machine has run to the next distributed synchronization baseline, the computer returns to step (5) and continues to run; otherwise, the computer suspends the run until a signal is received from all other computers that the computer has run to the next distributed synchronization baseline, and then the computer returns to step (5) and continues to run.

[0012] In the above-mentioned step-based distributed simulation system time synchronization method, in step (1), in the distributed simulation system, there are a total of 11 single machines that need to be simulated, and the time synchronization step between the simulation node of single machine 1, the simulation node of single machine 2, the simulation node of single machine 3 and the simulation node of single machine 4 is 0.5 milliseconds; the time synchronization step between the simulation node of single machine 5, the simulation node of single machine 6 and the simulation node of single machine 7 is 1.0 millisecond; the time synchronization step between the simulation node of single machine 8 and the simulation node of single machine 9 is 1.5 milliseconds, and the time synchronization step between the simulation node of single machine 10 and the simulation node of single machine 11 is 3.2 milliseconds.

[0013] In the above-mentioned step-based distributed simulation system time synchronization method, in step (2), 11 single-machine simulation nodes are deployed on four physical computers; wherein, the single-machine 1 simulation node, the single-machine 2 simulation node, the single-machine 3 simulation node and the single-machine 4 simulation node are deployed on computer A, and the local time synchronization step t between the single-machine 1 simulation node, the single-machine 2 simulation node, the single-machine 3 simulation node and the single-machine 4 simulation node is determined to be t1, and t1 is equal to 0.5 milliseconds; the single-machine 5 simulation node, the single-machine 6 simulation node and the single-machine 7 simulation node are deployed on computer B, and the local time synchronization step t between the single-machine 5 simulation node, the single-machine 6 simulation node and the single-machine 7 simulation node is determined to be t2, and t2 is equal to 1.0 milliseconds ; Deploy the single-machine 8 simulation node and the single-machine 9 simulation node on computer C, and determine that the local time synchronization step t between the single-machine 8 simulation node and the single-machine 9 simulation node is t3, and t3 is equal to 1.5 milliseconds; deploy the single-machine 10 simulation node on computer D; deploy the single-machine 11 simulation node on computer C. At this time, there are three nodes on computer C, namely the single-machine 8 simulation node, the single-machine 9 simulation node and the single-machine 11 simulation node. Although the time synchronization step between the single-machine 11 simulation node and the other two nodes is 3.2 milliseconds, since the time synchronization step between the single-machine 8 simulation node and the single-machine 9 simulation node is 1.5 milliseconds, the local time synchronization step t3 of the three simulation nodes deployed on computer C is still 1.5 milliseconds.

[0014] In the above-mentioned distributed simulation system time synchronization method based on step length, in step (3), the distributed simulation system time synchronization step length T is 3.2 milliseconds.

[0015] In the above-mentioned distributed simulation system time synchronization method based on step size, in step (4), when the integer multiple relationship cannot be satisfied, T and t are adjusted downward so that T is an integer multiple of t.

[0016] In the above-mentioned distributed simulation system time synchronization method based on step size, in step (4), the distributed simulation system time synchronization step size T is adjusted to 3 milliseconds, and at this time T=6*t1=3*t2=2*t3.

[0017] In the above-mentioned step-based distributed simulation system time synchronization method, in step (5), on computer A, the single-machine 1 simulation node, the single-machine 2 simulation node, the single-machine 3 simulation node and the single-machine 4 simulation node are synchronized according to the local time synchronization step t1. Each simulation node reaches the local synchronization baseline at time t1. When all simulation nodes reach the local synchronization baseline for the sixth time, the condition 6*t1=T is satisfied. Computer A sends a signal to other computers in the network that the machine has run to the next distributed synchronization baseline, and checks whether it has received a signal from all other computers that they have run to the next distributed synchronization baseline.

[0018] In the above-mentioned step-based distributed simulation system time synchronization method, in step (5), on computer B, the single-machine 5 simulation node, the single-machine 6 simulation node and the single-machine 7 simulation node are synchronized according to the local time synchronization step t2, and each simulation node reaches the local synchronization baseline at time t2. When all simulation nodes reach the local synchronization baseline for the third time, the condition 3*t2=T is satisfied, and computer B sends a signal to other computers in the network that the machine has run to the next distributed synchronization baseline, and checks whether it has received a signal from all other computers that they have run to the next distributed synchronization baseline.

[0019] In the above-mentioned step-based distributed simulation system time synchronization method, in step (5), on computer C, the single-machine 8 simulation node, the single-machine 9 simulation node and the single-machine 11 simulation node are synchronized according to the local time synchronization step t3. Each simulation node reaches the local synchronization baseline at time t3. When all simulation nodes reach the local synchronization baseline for the second time, the condition 2*t3=T is satisfied. Computer C sends a signal to other computers in the network that the computer has run to the next distributed synchronization baseline, and checks whether it has received a signal from all other computers that they have run to the next distributed synchronization baseline.

[0020] In the above-mentioned step-based distributed simulation system time synchronization method, in step (5), there is only a single machine 10 simulation node on computer D, which directly runs to the distributed synchronization baseline, sends a signal to other computers in the network that the machine has run to the next distributed synchronization baseline, and checks whether it has received a signal that all other computers have run to the next distributed synchronization baseline.

[0021] Compared with the prior art, the present invention has the following beneficial effects:

[0022] The present invention can reduce the impact of network delays and lower the time synchronization overhead between virtual simulation nodes, thereby improving the performance of distributed simulation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0024] Figure 1 This is a time synchronization process diagram of a step-based distributed simulation system time synchronization method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0025] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, unless there is a conflict, the embodiments of the present disclosure and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0026] This embodiment provides a time synchronization method for a distributed simulation system based on a step size, the method comprising the following steps:

[0027] (1) Analyze the time synchronization requirements between simulation nodes and determine the synchronization step between different simulation nodes based on the synchronization requirements;

[0028] (2) Based on the time synchronization requirement analysis in step (1), simulation nodes with smaller time synchronization steps are preferentially deployed on one or more computers, and other nodes are deployed according to the computer load. While deploying, the local time synchronization step t between the simulation nodes on each computer is determined;

[0029] (3) According to the time synchronization requirement analysis in step (1), determine the time synchronization step size T of the distributed simulation system;

[0030] (4) Adjust the time synchronization step size T of the distributed simulation system and the local time synchronization step size t of the simulation nodes on each computer, requiring T to be an integer multiple of t; if the integer multiple relationship cannot be met, adjust T and t downward so that T is an integer multiple of t;

[0031] (5) Initialize n = 0. On each computer, each simulation node runs synchronously according to the local time synchronization step t. When each simulation node runs to the local synchronization baseline, it must wait until all other simulation nodes on the same computer have run to the same baseline before continuing to run. When all simulation nodes reach the local synchronization baseline for the nth time and meet the condition: n*t = T, they send a signal to other computers in the network that the computer has run to the next distributed synchronization baseline, and check whether they have received the signal that all other computers have run to the next distributed synchronization baseline.

[0032] (6) If a computer has received a signal from all other computers that they have reached the next distributed synchronization baseline, it returns to step (5) and continues running; otherwise, it pauses running until it receives a signal from all other computers that they have reached the next distributed synchronization baseline, and then returns to step (5) and continues running.

[0033] Figure 1 This is a time synchronization process diagram of a step-based distributed simulation system time synchronization method according to the present invention. The distributed simulation system shown in the figure consists of 11 simulation nodes deployed on four physical computers A, B, C, and D. The time synchronization period of the distributed simulation system is T, and:

[0034] The local time synchronization step of the four nodes deployed on computer A is t1, and 6*t1=T.

[0035] The local time synchronization step of the three nodes deployed on computer B is t2, and 3*t2=T.

[0036] The local time synchronization step of the three nodes deployed on computer C is t3, and 2*t3=T.

[0037] Only one node is deployed on computer D, and local time synchronization is not required. This node is synchronized with nodes deployed on other computers according to the synchronization simulation period T.

[0038] In this control subsystem, there are a total of 11 single machines that need to be simulated. When the time deviation between single machines 1 to 4 exceeds 0.5 milliseconds, the system will work abnormally; when the time deviation between single machines 5 to 7 exceeds 1.0 milliseconds, the system will work abnormally; when the time deviation between single machines 8 and 9 exceeds 1.5 milliseconds, the system will work abnormally; when the time deviation between other single machines exceeds 3.2 milliseconds, the system will work abnormally.

[0039] At the same time, it is known that the single machine 10 involves a large number of mathematical operations, and the simulation requires high computer performance. The simulation of other single machines does not require special computer performance. Currently, there are four physical computers available to build a distributed simulation system, and these four physical computers have exactly the same software and hardware configurations.

[0040] In view of the above situation, the deployment of the distributed simulation system is implemented according to the following steps:

[0041] (1) According to the requirements of the control subsystem, the time synchronization step between the simulation nodes of the single machine 1 to 4 is determined to be 0.5 milliseconds, the time synchronization step between the simulation nodes of the single machine 5 to 7 is determined to be 1.0 milliseconds, the time synchronization step between the simulation nodes of the single machine 8 and 9 is determined to be 1.5 milliseconds, and the time synchronization step between the simulation nodes of the other single machines is determined to be 3.2 milliseconds.

[0042] (2) Based on the time synchronization requirement analysis in step (1) and the computer load situation, 11 stand-alone simulation nodes are deployed on four physical computers:

[0043] Standalone simulation nodes 1 to 4 that require a smaller time synchronization step are deployed on computer A, and the local time synchronization step between these simulation nodes is determined to be t1, and t1 is equal to 0.5 milliseconds.

[0044] Deploy 5 to 7 stand-alone simulation nodes that require a smaller time synchronization step on computer B, and determine that the local time synchronization step between these nodes is t2, and t2 is equal to 1.0 millisecond.

[0045] The stand-alone simulation nodes 8 and 9 that require a smaller time synchronization step are deployed on computer C, and the local time synchronization step between the two nodes is determined to be t3, and t3 is equal to 1.5 milliseconds.

[0046] Since the simulation of the single machine 10 requires high computer performance and the time synchronization requirement between it and other simulation nodes is no more than 3.2 milliseconds, it is deployed on computer D.

[0047] Standalone machine 11 is deployed on computer C. This is because computer A already has four nodes deployed, computer B has three nodes deployed, and although computer D only has one simulation node, standalone machine 10, this simulation node requires higher computer performance. Therefore, standalone machine 11 simulation node is deployed on computer C. At this point, there are three nodes on computer C: standalone machines 8, 9, and 11. Although the time synchronization step between standalone machine 11 and the other two nodes can reach 3.2 milliseconds, the time synchronization step between standalone machines 8 and 9 is 1.5 milliseconds. Therefore, the local time synchronization step t3 of the three simulation nodes deployed on computer C remains 1.5 milliseconds.

[0048] (3) According to the time synchronization requirement analysis in step (1), the time synchronization step size T of the distributed simulation system is set to 3.2 milliseconds;

[0049] (4) The time synchronization step T of the distributed simulation system is 3.2 milliseconds. The local time synchronization steps t1, t2, and t3 of computers A, B, and C are 0.5 milliseconds, 1.0 milliseconds, and 1.5 milliseconds, respectively. Computer D has only one simulation node deployed, which synchronizes time with other computers directly according to the distributed simulation system time synchronization step T. To meet the integer multiple relationship, T is adjusted to 3 milliseconds. In this case, T = 6*t1 = 3*t2 = 2*t3.

[0050] (5) On computer A, stand-alone simulation nodes 1 to 4 run synchronously according to the local time synchronization step t1. When each simulation node reaches the local synchronization baseline at time t1, it waits until all other simulation nodes on the same machine have reached the same baseline before continuing to run. When all simulation nodes reach the local synchronization baseline for the sixth time, the condition 6*t1=T is satisfied. Computer A sends a signal to other computers in the network that it has reached the next distributed synchronization baseline and checks whether it has received the signal from all other computers that have reached the next distributed synchronization baseline. Similarly, the stand-alone simulation nodes on computers B and C also perform similar synchronization operations. Computer D has only one stand-alone simulation node, which runs directly to the distributed synchronization baseline and sends a signal to other computers in the network that it has reached the next distributed synchronization baseline and checks whether it has received the signal from all other computers that have reached the next distributed synchronization baseline.

[0051] (6) If computers A, B, C, and D have received a signal that all other computers have reached the next distributed synchronization baseline, they return to step (5) and continue running; otherwise, they pause running until they receive a signal that all other computers have reached the next distributed synchronization baseline, and then return to step (5) and continue running.

[0052] By preferentially deploying simulation nodes with smaller time synchronization steps on one or more computers, the present invention allows simulation nodes deployed on different computers in a distributed simulation system to synchronize using different time synchronization steps. Simulation nodes deployed on the same computer can directly synchronize using the communication mechanism provided by the computer's operating system, eliminating the need for time synchronization through a computer network with high latency. Because local communication speeds are much higher than network communication speeds, even if the time synchronization steps between several simulation nodes deployed on a single computer are very small, the impact of time synchronization on the performance of the distributed simulation system can be significantly reduced by properly deploying the simulation nodes of the distributed simulation system.

[0053] Although the present invention has been disclosed above in terms of preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solutions of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the scope of protection of the technical solutions of the present invention.

Claims

1. A time synchronization method for a distributed simulation system based on step length, characterized in that include: (1) Determine the synchronization step between different simulation nodes; (2) deploying different simulation nodes on multiple computers based on the synchronization step between different simulation nodes, and determining the local time synchronization step t between the simulation nodes on each computer during deployment; (3) Determine the time synchronization step T of the distributed simulation system; (4) Adjust the time synchronization step T of the distributed simulation system and the local time synchronization step t between each simulation node on each computer, requiring T to be an integer multiple of t; (5) On each computer, each simulation node runs synchronously according to the local time synchronization step t, so that each simulation node runs to the local synchronization baseline; when all simulation nodes reach the local synchronization baseline for the nth time and the condition n*t=T is satisfied, a signal is sent to other computers in the network indicating that the computer has run to the next distributed synchronization baseline, and the computer checks whether the signal has been received from all other computers that they have run to the next distributed synchronization baseline; (6) If a computer has received a signal from all other computers that they have reached the next distributed synchronization baseline, it returns to step (5) and continues running; otherwise, it pauses running until it receives a signal from all other computers that they have reached the next distributed synchronization baseline, and then returns to step (5) and continues running.

2. The time synchronization method for a distributed simulation system based on step size according to claim 1, characterized in that: In step (1), in the distributed simulation system, there are a total of 11 single machines that need to be simulated, and the time synchronization step between the simulation node of single machine 1, the simulation node of single machine 2, the simulation node of single machine 3 and the simulation node of single machine 4 is 0.5 milliseconds; the time synchronization step between the simulation node of single machine 5, the simulation node of single machine 6 and the simulation node of single machine 7 is 1.0 milliseconds; the time synchronization step between the simulation node of single machine 8 and the simulation node of single machine 9 is 1.5 milliseconds, and the time synchronization step between the simulation node of single machine 10 and the simulation node of single machine 11 is 3.2 milliseconds.

3. The time synchronization method for a distributed simulation system based on step size according to claim 2, characterized in that: In step (2), 11 stand-alone simulation nodes are deployed on four physical computers; Deploy the stand-alone 1 simulation node, the stand-alone 2 simulation node, the stand-alone 3 simulation node, and the stand-alone 4 simulation node on computer A, and determine that the local time synchronization step t between the stand-alone 1 simulation node, the stand-alone 2 simulation node, the stand-alone 3 simulation node, and the stand-alone 4 simulation node is t1, and t1 is equal to 0.5 milliseconds; Deploy the stand-alone 5 simulation node, the stand-alone 6 simulation node, and the stand-alone 7 simulation node on computer B, and determine that the local time synchronization step t between the stand-alone 5 simulation node, the stand-alone 6 simulation node, and the stand-alone 7 simulation node is t2, and t2 is equal to 1.0 millisecond; Deploy the single-machine 8 simulation node and the single-machine 9 simulation node on computer C, and determine that the local time synchronization step t between the single-machine 8 simulation node and the single-machine 9 simulation node is t3, and t3 is equal to 1.5 milliseconds; Deploy the single-machine 10 simulation node on computer D; The stand-alone 11 simulation node is deployed on computer C. At this time, there are three nodes on computer C: the stand-alone 8 simulation node, the stand-alone 9 simulation node, and the stand-alone 11 simulation node. Although the time synchronization step between the stand-alone 11 simulation node and the other two nodes is 3.2 milliseconds, since the time synchronization step between the stand-alone 8 simulation node and the stand-alone 9 simulation node is 1.5 milliseconds, the local time synchronization step t3 of the three simulation nodes deployed on computer C is still 1.5 milliseconds.

4. The time synchronization method for a distributed simulation system based on step length according to claim 3, characterized in that: In step (3), the time synchronization step T of the distributed simulation system is 3.2 milliseconds.

5. The time synchronization method for a distributed simulation system based on step size according to claim 1, characterized in that: In step (4), when the integer multiple relationship cannot be satisfied, T and t are adjusted downward so that T is an integer multiple of t.

6. The time synchronization method for a distributed simulation system based on step size according to claim 4, characterized in that: In step (4), the time synchronization step T of the distributed simulation system is adjusted to 3 milliseconds, and at this time T=6*t1=3*t2=2*t3.

7. The time synchronization method for a distributed simulation system based on step length according to claim 6, characterized in that: In step (5), on computer A, the single-machine 1 simulation node, the single-machine 2 simulation node, the single-machine 3 simulation node and the single-machine 4 simulation node are synchronized according to the local time synchronization step t1. Each simulation node reaches the local synchronization baseline at time t1. When all simulation nodes reach the local synchronization baseline for the sixth time, the condition 6*t1=T is satisfied. Computer A sends a signal to other computers in the network that the machine has run to the next distributed synchronization baseline, and checks whether it has received a signal from all other computers that they have run to the next distributed synchronization baseline.

8. The time synchronization method for a distributed simulation system based on step length according to claim 7, characterized in that: In step (5), on computer B, the single-machine 5 simulation node, the single-machine 6 simulation node and the single-machine 7 simulation node run synchronously according to the local time synchronization step t2. Each simulation node reaches the local synchronization baseline at time t2. When all simulation nodes reach the local synchronization baseline for the third time, the condition 3*t2=T is satisfied. Computer B sends a signal to other computers in the network that the machine has run to the next distributed synchronization baseline, and checks whether it has received a signal from all other computers that they have run to the next distributed synchronization baseline.

9. The time synchronization method for a distributed simulation system based on step length according to claim 8, characterized in that: In step (5), on computer C, the single-machine 8 simulation node, the single-machine 9 simulation node and the single-machine 11 simulation node are synchronized according to the local time synchronization step t3. Each simulation node reaches the local synchronization baseline at time t3. When all simulation nodes reach the local synchronization baseline for the second time, the condition 2*t3=T is satisfied. Computer C sends a signal to other computers in the network that the machine has run to the next distributed synchronization baseline, and checks whether it has received a signal from all other computers that they have run to the next distributed synchronization baseline.

10. The time synchronization method for a distributed simulation system based on step length according to claim 9, characterized in that: In step (5), there is only a single machine 10 simulation node on computer D, which runs directly to the distributed synchronization baseline, sends a signal to other computers in the network that the machine has run to the next distributed synchronization baseline, and checks whether it has received a signal from all other computers that they have run to the next distributed synchronization baseline.

Images