Super real-time simulation method for digital twin network

Abstract

The invention discloses a super real-time simulation method based on digital twinning, designs a digital twinning network deduction simulation system, and designs a network simulation module, a service generation module and a simulation deduction module in the digital twinning network deduction simulation system. Through a data transmission zero-copy technology and an intelligent data distribution model, service data generation and distribution of a digital twin network are accelerated, a multi-speed network simulation deduction function based on a digital twin network scene is realized, a core service communication process of network deduction can be rapidly and accurately reflected, network super-real-time topology perspective and traffic holography are realized, and network performance is improved. And network new technology deployment verification is assisted, and network problems and future network ubiquitous interconnection challenges can be dealt with more efficiently.

Description

Technical Field

[0001] The present invention relates to network simulation technology, and in particular to an ultra-real-time simulation method for digital twin networks. Background Art

[0002] Network simulation is a network research method that has rapidly emerged in recent years. It mainly relies on computers, mathematical modeling, and statistical analysis to simulate network behavior and can analyze the network performance parameters required for network design or optimization.

[0003] In real-time network simulation, computer models and physical systems run at the same rate, ensuring that the simulated system maintains a high degree of time consistency with the actual physical scale. Real-time network simulation can make the simulation more accurate and allow for fine-tuning and debugging. However, to maintain the real-time nature of the simulation, a significant amount of time is consumed to simulate the network's performance over the same time period, significantly reducing simulation efficiency. Super real-time simulation technology, in which the system simulation time process is faster than the actual system time process, can effectively address the low simulation efficiency problem that exists in real-time network simulation. Simulation systems implemented using super real-time simulation technology are called super real-time simulation systems. Their main characteristic is that the time process of the system simulation model is faster than the actual system time.

[0004] A digital twin network is a network system consisting of a physical network entity and a virtual twin that can interact and map with each other in real time. Within this system, various network management and applications can leverage the virtual twins of the network built using digital twin technology to efficiently analyze, diagnose, simulate, and control the physical network based on data and models. A digital twin network has four core elements: data, models, mapping, and interaction. Given the increasing multi-dimensional, full-scenario ubiquitous connectivity of future networks, digital twin networks will become a new direction for future network planning, operation, and management, and a key tool for network intelligence and automation. However, digital twin networks reflect the actual operation of the network and primarily rely on real-time network simulation. This lack of support for rapid network simulation review and timely decision-making is problematic. Summary of the Invention

[0005] In response to the problems raised by the background technology, the present invention discloses an ultra-real-time simulation method for digital twin networks. The method designs a digital twin network deduction and simulation system, and designs a network simulation module, a service generation module, and a simulation deduction module in the digital twin network deduction and simulation system. The network simulation module realizes the topology planning, virtual body generation and virtual network construction functions of the digital twin network. The service generation module realizes the data service generation and data transmission zero-copy acceleration functions. The simulation deduction module realizes the intelligent data distribution model and deduction scheduling control functions. The digital twin network deduction and simulation system establishes an equivalent virtual network for the physical network, simulates the network networking and data communication of the physical network, and evaluates, analyzes and predicts the physical network by monitoring the impact of data communication on network networking and the effect of data communication. The system can provide support for intelligent decision-making of future ubiquitous networks.

[0006] The working steps of the digital twin network deduction simulation system to achieve ultra-real-time simulation method are as follows:

[0007] S101, the network simulation module plans the simulation network topology, the topology content includes device parameters, node composition, connection relationship between devices and node deployment location information;

[0008] S102, the network simulation module creates a digital simulation network according to the simulation network topology and generates virtual entity devices;

[0009] S103, the network simulation module constructs a connection of virtual entities according to the simulation network topology;

[0010] S104, the simulation deduction module plans and generates data services based on time series and sets a super real-time step size;

[0011] S105, the simulation deduction module generates ultra-real-time time series data services according to the step size and time series data services;

[0012] S106: The simulation module starts the simulation and sends the simulation ultra-real-time step size and ultra-real-time time series data service to the service generation module;

[0013] S106. The service generation module generates services according to ultra-real-time time series data services and sends data services to the digital twin network.

[0014] The network simulation module features a topology planning function. Within this simulated network, network elements (NEs) can be added, configured, and positioned using drag-and-drop within the software interface. This allows for easy network topology construction, with customized topology types and connection relationships to simulate complex network environments, including star, ring, and hierarchical topologies. The module intelligently adjusts the topology layout based on user layout preferences to ensure an elegant arrangement of NE connections. The module features a detailed parameter configuration interface tailored to the specific NE design. This interface includes fixed attributes and performance parameters, enabling more refined adjustments to network behavior and ultimately, a network planning solution.

[0015] The network simulation module has data storage capabilities and can save the planned network planning scheme through the database. Users can call the network planning scheme again when needed and perform secondary editing and planning of the topology and parameters in the network planning scheme.

[0016] The network simulation module's virtual network element generation function is designed to support multiple operating modes, including namespace NS (Namespace), Docker containers, virtual machines, and OVS (Open vSwitch) instances. The network simulation module configures and generates virtual network elements according to the planned network planning scheme. Users can modify the properties of each virtual network element, including operating mode, IP address, port mapping, resource allocation, and management port, to ensure that the generated network elements meet specific experiment or test requirements.

[0017] The network simulation module simulates the functional design of network construction. Using configuration commands and scripts, the module fine-tunes virtualized network elements, customizing each element's behavior, routing rules, security policies, and interconnections to meet network planning requirements. Connection configuration parameters include bandwidth, latency, and packet loss rate, as well as inter-radio radio link parameters, including key parameters such as channel, signal-to-noise ratio, and transmit power.

[0018] The construction process of the network simulation module is not limited to the logical configuration of the network itself, but also includes the configuration and calculation of geographic information and interference conditions. Users can adjust the geographic location information through the map before or after creation to further simulate the impact of specific geographic information on network performance.

[0019] The service generation module adopts a containerized design, supports distributed deployment, has data communication capabilities such as simulated service reception and service sending, and has data acceleration functions such as zero-copy data transmission.

[0020] The data acceleration design of the service generation module is primarily intended to address resource consumption issues associated with traditional data transmission. Traditional data transmission relies on the CPU to move data multiple times, first reading the data from disk, then copying the data from kernel space to user space, and then from user space to application memory. When the data volume is large, traditional data transmission consumes a significant amount of CPU resources and memory bandwidth. The service generation module has data acceleration capabilities, bypassing user space and conducting data exchange directly through kernel space. The main steps of data acceleration in the service generation module are as follows:

[0021] S201, the user data transmission process initiates an IO call to the operating system kernel through the mmap (memory map) method;

[0022] S202, the CPU copies the data from the hard disk to the kernel buffer;

[0023] S203, mmap returns the execution result to the user;

[0024] S204: The user process initiates a write IO call, and the CPU copies the data from the memory buffer to the socket buffer.

[0025] S205: The CPU copies the data from the socket buffer to the network card.

[0026] The business generation module simulates business reception, business sending and other data communication designs. The business generation module has independent deployment capabilities through containerized design and can be interconnected with any network virtual body of the digital twin network. In the digital twin network, it is used to replace the business access system of the actual network. When the business generation module receives the time series data business from the simulation deduction module, it will send data to the digital twin network according to the parameters of the time series business. The data communication of the business generation module is divided into two categories based on the sending form, one is one-way business and the other is two-way business. Among them, one-way business refers to the business generation module sending data directly to a specified IP. Two-way business refers to the process in which the business generation module and other business generation modules will have message responses. Taking the example of the first business generation module sending data to the second business generation module, the two-way business communication steps of the business generation module are:

[0027] S301: The first service generation module sends data to the second service generation module of the digital twin network. The data includes a service type identifier and a response requirement for the bidirectional service.

[0028] S302: The second service generation module receives the data and identifies the service type of the data. If it is a bidirectional service, the process proceeds to the next step; otherwise, the process ends.

[0029] S303: The second service generation module sends response data to the first service generation module. If a response from the first service generation module is required, the data includes a service type identifier and a response requirement for the new two-way service. If not, the data does not include the service type identifier and a response requirement.

[0030] The intelligent data distribution model functional design of the simulation deduction module uses the Douglas Peucker algorithm to optimize time series data to achieve faster simulation model time processes than the actual system time processes. The Douglas Peucker algorithm is a classic algorithm for simplifying lines. It connects the first and last points of the time series data curve with a dotted line, calculates the distance from all points other than the first and last two points to this dotted line, finds the maximum distance, and compares it with a pre-set threshold. If the maximum distance is less than the threshold, all points on the line are removed. If the maximum distance is greater than or equal to the threshold, the coordinate corresponding to the maximum distance is retained, and the curve is divided into two parts using this point as the boundary. The method is repeated for both parts until the algorithm terminates.

[0031] The steps for optimizing time series data are as follows:

[0032] S401. First, some representative points (such as the first and last nodes, turning points, etc.) are determined on the time series data curve; then, the first and last nodes are connected by a dotted line; finally, a line segment is drawn perpendicular to the dotted line through the remaining nodes, and the length of the line segment (that is, the distance between the remaining nodes and the dotted line) is calculated;

[0033] S402, find the maximum distance calculated in the previous step and compare it with the threshold;

[0034] S403. If the maximum distance is less than the threshold, connect the first and last nodes with a line segment, and use the line segment to replace the remaining nodes; if the maximum distance is greater than the threshold, retain the node and use the node as the dividing point to divide the time series data curve into two curves;

[0035] S404, determine whether the process is complete: In all the divided curve parts, in addition to the first and last nodes, repeat steps S401, S402, S403, and S404 for the parts that still have other nodes; stop the process for the parts that only have the first and last nodes; the whole process ends when all the curve parts only have the first and last nodes. There are only two judgment conditions for determining whether the process is complete: the first is whether the maximum distance in the part is greater than the threshold; the second is whether the part only has the first and last nodes;

[0036] S405 . Connect the remaining nodes with line segments to obtain a simplified time series data curve.

[0037] In the present invention, in addition to being able to achieve super real-time simulation and simplify the simulation process, it is also possible to simulate super real-time at a certain multiple according to actual needs. Assuming that there are 60 nodes on the time-characteristic diagram of the system simulation, then if super real-time simulation is performed at one speed, the present invention will simulate 30 nodes (such nodes are called estimated nodes in the present invention); if super real-time simulation is performed at two times the speed, 15 nodes will be simulated. The rest are similar. However, by simplifying the curve using the Douglas Peucker algorithm, it is impossible to determine how many nodes are in the final result. To address this situation, the present invention adds an interpolation algorithm and a removal algorithm.

[0038] Case 1: When the number of remaining nodes is greater than the estimated number of nodes, a certain number of remaining nodes need to be removed. The calculation formula for the number of nodes to be removed is as follows:

[0039] n=α-β,α>β≥2

[0040] Where n is the number of nodes removed, α is the number of remaining nodes, and β is the estimated number of nodes. In this invention, the time-characteristic graph should contain at least two nodes: the first node and the last node. Therefore, α is greater than β, and β is greater than or equal to 2. Furthermore, by default, the first and last nodes are not removed.

[0041] After calculating the number n of nodes to be removed, we need to calculate the nodes that need to be deleted. The algorithm is as follows:

[0042] Assume that the vertical coordinate value of each node is y i Indicates that A i Represents the absolute value of the difference between the ordinate value of the i-th node and the absolute values ​​of the ordinate values ​​of the previous node and the next node. The expression is as follows:

[0043] A i =||y i -y i-1 |-|y i -y i+1 ||,2≤i<α-1

[0044] Then for all the calculated A i Sort in ascending order, and then delete the first n nodes. This is because in the time-characteristic graph, the smaller the gap between a node and the previous and next nodes, the smaller the change amplitude of the point, and the less important it is in the trend change process, so it should be removed first.

[0045] Case 2: When the number of remaining nodes is less than the estimated number of nodes, a certain number of nodes need to be inserted. The calculation formula for the number of inserted nodes is as follows:

[0046] c=β-α,β>α,β≥2

[0047] Among them, c is the number of nodes that need to be inserted, β is the estimated number of nodes, and α is the number of remaining nodes.

[0048] Before inserting a node, you need to determine the position of the node to be inserted. Generally speaking, the position of the inserted node should be after the first node and before the last node. Assume that the horizontal coordinate of the i-th node is x i , the interval between the i-th node and the i+1-th node is w i , whose expression is:

[0049] w i =x i+1 -x i ,i<x-1

[0050] For the calculated w i In this invention, it is considered that the node should be inserted into the first c wider intervals, and the horizontal coordinate x of the inserted node is i The expression is as follows:

[0051]

[0052] The present invention adopts Lagrange interpolation method for interpolation. Assume that the coordinates of the α nodes are:

[0053] (x1,y i ),(x2,y2),...,(x a ,y a )

[0054] Construct the Lagrange interpolation expression:

[0055]

[0056] When solving the Lagrange interpolation expression, we only need to substitute the horizontal coordinate of the interpolation value calculated above to calculate the corresponding vertical coordinate, thereby obtaining the specific coordinate and realizing the insertion function.

[0057] Case 3: When the number of remaining nodes is less than the estimated number of nodes, no additional operations are required and the results are output directly.

[0058] In summary, the curve optimization process in the present invention can be summarized into the following steps:

[0059] S501, obtaining a simulation time-characteristic diagram;

[0060] S502, setting a threshold D, and simplifying the time-characteristic graph obtained in the previous step using the Douglas Peucker algorithm;

[0061] S503. The user can independently select the multiple of the ultra-real-time simulation process;

[0062] S504. Compare the number of remaining nodes with the estimated number of nodes, and perform different operations according to the three different situations of the comparison results; in case 1, when the number of remaining nodes is greater than the estimated number of nodes, a node removal algorithm is used to delete redundant and less important nodes; in case 2, when the number of remaining nodes is less than the estimated number of nodes, Lagrange interpolation method is used to interpolate at places where the node spacing is relatively large; in case 3, when the number of remaining nodes is equal to the estimated number of nodes, no additional operation is required.

[0063] The simulation scheduling control function of the simulation module is designed. The simulation module starts the simulation and sends the simulation ultra-real-time step and ultra-real-time time series data business to the business generation module through the HTTP network interface and RESTFUL API interface specification. The sent content can be dynamically changed according to user operations, including control instructions such as end, pause, and restart.

[0064] The beneficial technical effect of the present invention is: a super real-time simulation method for digital twin networks, which accelerates the business data generation and distribution of digital twin networks through zero-copy data transmission technology and intelligent data distribution models, designs a digital twin network deduction and simulation system, and realizes multi-speed simulation and deduction functions in digital twin network scenarios. It can quickly and accurately reflect the core business communication process of network deduction, realize network super real-time topology perspective and traffic holography, assist in the deployment and verification of new network technologies, and can more efficiently deal with network problems and the challenges of ubiquitous interconnection of future networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The drawings described herein are used to provide further understanding of the present invention and constitute a part of this application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0066] Attachment Figure 1 、 one A module composition diagram of an ultra-real-time simulation method for digital twin networks;

[0067] Attachment Figure 2 、 one A working step diagram of an ultra-real-time simulation method for digital twin networks;

[0068] Attachment Figure 3 、 one Schematic diagram of the initial curve of an ultra-real-time simulation method for digital twin networks;

[0069] Attachment Figure 4 、 oneSchematic diagram of step 1 of the Douglas Peucker algorithm for an ultra-real-time simulation method for digital twin networks;

[0070] Attachment Figure 5 、 one Schematic diagram of step 2 of the Douglas Peucker algorithm for an ultra-real-time simulation method for digital twin networks;

[0071] Attachment Figure 6 、 one Schematic diagram of step 3 of the Douglas Peucker algorithm for an ultra-real-time simulation method for digital twin networks;

[0072] Attachment Figure 7 、 one A simplified diagram of an ultra-real-time simulation method for digital twin networks.

[0073] Attachment Figure 8 、 one Schematic diagram comparing the initial and simplified curves of an ultra-real-time simulation method for digital twin networks;

[0074] Attachment Figure 9 、 one Curve optimization flowchart of an ultra-real-time simulation method for digital twin networks. DETAILED DESCRIPTION

[0075] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0076] See attached Figure 1 As shown, this embodiment provides a network simulation method based on digital twins with the following working steps:

[0077] S101, the network simulation module plans the simulation network topology, the topology content includes device parameters, node composition, connection relationship between devices and node deployment location information;

[0078] S102, the network simulation module creates a digital simulation network according to the simulation network topology and generates virtual entity devices;

[0079] S103, the network simulation module constructs a connection of virtual entities according to the simulation network topology;

[0080] S104, the simulation deduction module plans and generates data services based on time series and sets a super real-time step size;

[0081] S105, the simulation deduction module generates ultra-real-time time series data services according to the step size and time series data services;

[0082] S106: The simulation module starts the simulation and sends the simulation ultra-real-time step size and ultra-real-time time series data service to the service generation module;

[0083] S106. The service generation module generates services according to ultra-real-time time series data services and sends data services to the digital twin network.

[0084] See attached Figure 2 As shown, this embodiment provides a schematic diagram of the composition of a network simulation method based on digital twins, designs a digital twin network deduction and simulation system, and designs a network simulation module, a service generation module, and a simulation deduction module in the digital twin network deduction and simulation system. The network simulation module implements the topology planning, virtual body generation, and virtual network construction functions of the digital twin network. The service generation module implements the data service generation and data transmission zero-copy acceleration functions. The simulation deduction module implements the intelligent data distribution model and deduction scheduling control functions. The digital twin network deduction and simulation system establishes an equivalent virtual network for the physical network, simulates the network networking and data communication of the physical network, and evaluates, analyzes, and predicts the physical network by monitoring the impact of data communication on network networking and the effect of data communication. The system can provide support for intelligent decision-making of future ubiquitous networks.

[0085] The network simulation module features a topology planning function. Within this simulated network, network elements (NEs) can be added, configured, and positioned using drag-and-drop within the software interface. This allows for easy network topology construction, with customized topology types and connection relationships to simulate complex network environments, including star, ring, and hierarchical topologies. The module intelligently adjusts the topology layout based on user layout preferences to ensure an elegant arrangement of NE connections. The module features a detailed parameter configuration interface tailored to the specific NE design. This interface includes fixed attributes and performance parameters, enabling more refined adjustments to network behavior and ultimately, a network planning solution.

[0086] The network simulation module has data storage capabilities and can save the planned network planning scheme through the database. Users can call the network planning scheme again when needed and perform secondary editing and planning of the topology and parameters in the network planning scheme.

[0087] The network simulation module's virtual network element generation function is designed to support multiple operating modes, including namespace NS (Namespace), Docker containers, virtual machines, and OVS (Open vSwitch) instances. The network simulation module configures and generates virtual network elements according to the planned network planning scheme. Users can modify the properties of each virtual network element, including operating mode, IP address, port mapping, resource allocation, and management port, to ensure that the generated network elements meet specific experiment or test requirements.

[0088] The network simulation module simulates the functional design of network construction. Using configuration commands and scripts, the module fine-tunes virtualized network elements, customizing each element's behavior, routing rules, security policies, and interconnections to meet network planning requirements. Connection configuration parameters include bandwidth, latency, and packet loss rate, as well as inter-radio radio link parameters, including key parameters such as channel, signal-to-noise ratio, and transmit power.

[0089] The construction process of the network simulation module is not limited to the logical configuration of the network itself, but also includes the configuration and calculation of geographic information and interference conditions. Users can adjust the geographic location information through the map before or after creation to further simulate the impact of specific geographic information on network performance.

[0090] The service generation module adopts a containerized design, supports distributed deployment, has data communication capabilities such as simulated service reception and service sending, and has data acceleration functions such as zero-copy data transmission.

[0091] The data acceleration design of the business generation module is mainly to solve the resource consumption problem during traditional data transmission. Traditional data transmission requires the CPU to move data multiple times. First, the data is read from the disk, then the data is copied from the kernel space to the user space, and then from the user space to the application memory. For example, in JAVA development, the main steps for sending data through Socket are to read the data through FileInputStream, build a buffer, and then send the data through DataOutputStream. When the amount of data is relatively large, the traditional data transmission method consumes a lot of CPU resources and memory bandwidth. The business generation module has data acceleration capabilities and can skip the user space and directly interact with data through the kernel space. The main steps of data acceleration in the business generation module are as follows:

[0092] S201, the user data transmission process initiates an IO call to the operating system kernel through the mmap (memory map) method;

[0093] S202, the CPU copies the data from the hard disk to the kernel buffer;

[0094] S203, mmap returns the execution result to the user;

[0095] S204: The user process initiates a write IO call, and the CPU copies the data from the memory buffer to the socket buffer.

[0096] S205: The CPU copies the data from the socket buffer to the network card.

[0097] The business generation module simulates business reception, business sending and other data communication designs. The business generation module has independent deployment capabilities through containerized design and can be interconnected with any network virtual body of the digital twin network. In the digital twin network, it is used to replace the business access system of the actual network. When the business generation module receives the time series data business from the simulation deduction module, it will send data to the digital twin network according to the parameters of the time series business. The data communication of the business generation module is divided into two categories based on the sending form, one is one-way business and the other is two-way business. Among them, one-way business refers to the business generation module sending data directly to a specified IP. Two-way business refers to the process in which the business generation module and other business generation modules will have message responses. Taking the example of the first business generation module sending data to the second business generation module, the two-way business communication steps of the business generation module are:

[0098] S301. The first service generation module sends data to the second service generation module of the digital twin network, where the data includes a service type identifier and a response requirement of the bidirectional service.

[0099] S302: The second service generation module receives the data and identifies the service type of the data. If it is a bidirectional service, the process proceeds to the next step; otherwise, the process ends.

[0100] S303: The second service generation module sends response data to the first service generation module. If a response from the first service generation module is required, the data includes a service type identifier and a response requirement for the new two-way service. If not, the data does not include the service type identifier and a response requirement.

[0101] The intelligent data distribution model functional design of the simulation deduction module uses the Douglas Peucker algorithm to optimize time series data to achieve faster simulation model time processes than the actual system time processes. The Douglas Peucker algorithm is a classic algorithm for simplifying lines. It connects the first and last points of the time series data curve with a dotted line, calculates the distance from all points other than the first and last two points to this dotted line, finds the maximum distance, and compares it with a pre-set threshold. If the maximum distance is less than the threshold, all points on the line are removed. If the maximum distance is greater than or equal to the threshold, the coordinate corresponding to the maximum distance is retained, and the curve is divided into two parts using this point as the boundary. The method is repeated for both parts until the algorithm terminates.

[0102] See attached Figure 3 As shown, this embodiment provides an initial curve schematic diagram of an ultra-real-time simulation method for digital twin networks.

[0103] In the attached Figure 3 Assume there are eight representative points, each labeled with a serial number. Let's use this curve as an example to simplify it using the Douglas Peucker algorithm. Connect the first and last nodes with a dashed line. Then, draw a line perpendicular to the dashed line for the remaining nodes and find their distance from the dashed line.

[0104] See attached Figure 4 As shown in FIG, a schematic diagram of step 1 of the Douglas Peucker algorithm of a super real-time simulation method for digital twin networks provided in this embodiment. Assuming that the selected threshold D is 8 cm, Figure 4 It can be seen that the maximum distance d max = is the distance from point 6 to the dotted line, which is 10 cm. Therefore, the next step is to divide the curve into two parts using point 6 as the dividing point, connect the first and last nodes of the two parts with a dotted line, and then draw a line segment perpendicular to the dotted line for all the points to find the distance.

[0105] See attached Figure 5 As shown in the figure, a schematic diagram of step 2 of the Douglas Peucker algorithm for a super real-time simulation method for digital twin networks provided in this embodiment clearly shows that in the left part, the distance from point 4 to the dotted line is the farthest and is greater than the threshold D. Therefore, this part is divided into two parts with point 4 as the boundary. In the right part, in addition to the first and last nodes, there is only one node, point 7, and the distance is equal to the threshold D. Therefore, the algorithm retains node 7 and uses it as the dividing line to re-divide the part into two parts. Since the two re-divided nodes do not have other nodes besides the first and last nodes, a line segment is used to connect all the points in this part, and the algorithm ends at this part.

[0106] See attached Figure 6As shown, the present embodiment provides a schematic diagram of step three of the Douglas Peucker algorithm of a super real-time simulation method for digital twin networks. Obviously, according to the principle of the Douglas Peucker algorithm, the distance from any point to the dotted line is greater than the threshold D, so the dotted line is used instead of the curve to achieve the purpose of simplifying the curve.

[0107] See attached Figure 7 As shown, this embodiment provides a simplified curve diagram of an ultra-real-time simulation method for digital twin networks.

[0108] From the attached Figure 7 As can be seen, the original curve has 8 points. After simplification by the Douglas Peucker algorithm, only 5 relatively important and representative nodes are retained (such nodes are called remaining nodes in the present invention), thereby achieving the purpose of simplifying the curve.

[0109] See attached Figure 8 As shown, this embodiment provides a schematic diagram comparing the initial curve and simplified curve of the ultra-real-time simulation method for digital twin networks. It can be seen that the simplified result graph using the Douglas Peucker algorithm retains the more representative points, and the general change trend is also consistent with the original curve.

[0110] In a real-time simulation system, simulating the entire time-characteristics of the simulation system's time process (characteristics can be a variety of time-varying characteristics, such as traffic flow, bandwidth, and packet loss rate) requires a significant amount of computational time, increasing both computational difficulty and time cost. To address these issues, the present invention employs the Douglas Peucker algorithm to implement ultra-real-time simulation, removing less important time nodes while retaining representative ones, thereby simplifying the simulation process.

[0111] See attached Figure 9 As shown, this embodiment provides a curve optimization flow chart of an ultra-real-time simulation method for digital twin networks. The specific steps are:

[0112] Step 1: Obtain simulation time-characteristic diagram;

[0113] Step 2: Set the threshold D and use the Douglas Peucker algorithm to simplify the time-characteristic graph obtained in the previous step;

[0114] Step 3: The user can select the multiple of the ultra-real-time simulation process independently;

[0115] Step 4: Determine the size of the remaining number of nodes and the estimated number of nodes, and perform different operations according to the three different comparison results. In case 1, when the number of remaining nodes is greater than the estimated number of nodes, a node removal algorithm is used to delete redundant and less important nodes. In case 2, when the number of remaining nodes is less than the estimated number of nodes, Lagrange interpolation is used to interpolate at locations with large node spacing. In case 3, when the number of remaining nodes is equal to the estimated number of nodes, no additional operations are required.

Claims

1. A super real-time simulation method for digital twin networks, characterized by: A digital twin network simulation system was designed. It includes a network simulation module, a service generation module, and a simulation simulation module. The network simulation module implements the topology planning, virtual body generation, and virtual network construction functions of the digital twin network. The service generation module implements data service generation and zero-copy acceleration of data transmission. The simulation simulation module implements an intelligent data distribution model and simulation scheduling control functions. The digital twin network simulation system establishes an equivalent virtual network for the physical network, simulating the physical network's network networking and data communication. By monitoring the impact of data communication on network networking and the effectiveness of data communication, it evaluates, analyzes, and predicts the physical network. The working steps of the digital twin network simulation system to implement the ultra-real-time simulation method are as follows: S101, the network simulation module plans the simulation network topology, the topology content includes device parameters, node composition, connection relationship between devices and node deployment location information; S102, the network simulation module creates a digital simulation network according to the simulation network topology, generating virtual entity devices; S103, the network simulation module constructs a connection of virtual entities according to the simulation network topology; S104, the simulation deduction module plans and generates data services based on time series and sets a super real-time step size; S105, the simulation deduction module generates ultra-real-time time series data services according to the step size and time series data services; S106: The simulation module starts the simulation and sends the simulation ultra-real-time step size and ultra-real-time time series data service to the service generation module; S106. The service generation module generates services according to ultra-real-time time series data services and sends data services to the digital twin network.

2. The ultra-real-time simulation method for digital twin networks according to claim 1, characterized in that: The network simulation module's topology planning function allows users to add, configure, and locate network elements (NEs) by dragging and dropping them in the software interface. This allows users to easily build network topologies and customize the planned topology types and connection relationships to simulate complex network environments, including star, ring, or hierarchical topologies. The network simulation module will intelligently adjust the topology layout based on user layout habits to ensure an elegant arrangement of NE connections. The network simulation module provides a detailed parameter configuration interface differentiated by the different designs of NEs, including fixed attributes, performance parameters, etc., to enable more refined adjustments to network behavior and ultimately form a network planning solution.

3. The ultra-real-time simulation method for digital twin networks according to claim 1, characterized in that: Two-way business refers to the process in which the business generation module and other business generation modules will have message responses. The two-way business communication steps of the business generation module are as follows: Step 1: The first service generation module sends data to the second service generation module of the digital twin network. The data includes a service type identifier and a response requirement for the bidirectional service. Step 2: The second service generation module receives the data and identifies the service type of the data. If it is a bidirectional service, it proceeds to the next step; otherwise, it ends. Step 3: The second service generation module sends response data to the first service generation module. If a response from the first service generation module is required, the data includes a service type identifier and response requirements for the new two-way service. If not, the data does not include the service type identifier and response requirements.

4. The ultra-real-time simulation method for digital twin networks according to claim 1, characterized in that: The steps for optimizing time series data are: Step 1: First, identify some representative points (the first and last nodes, turning points, etc.) on the time series data curve; then connect the first and last nodes with a dotted line; finally, draw a line segment perpendicular to the dotted line through the remaining nodes, and calculate the length of the line segment (that is, the distance between the remaining nodes and the dotted line); Step 2: Find the maximum distance calculated in the previous step and compare it with the threshold; Step 3: If the maximum distance is less than the threshold, connect the first and last nodes with a line segment, and use the line segment to replace the remaining nodes; If the maximum distance is greater than the threshold, the node is retained and the time series data curve is divided into two curves with the node as the dividing point; Step 4: Determine whether the process is complete: Repeat steps S401, S402, S403, and S404 for all the divided curve parts that have nodes other than the first and last nodes. Stop the process for parts that have only the first and last nodes. The process ends when all curve parts have only the first and last nodes. There are only two conditions for determining whether the process is complete: the first is whether the maximum distance in the part is greater than a threshold; the second is whether the part has only the first and last nodes. Step 5: Use line segments to connect the remaining nodes to obtain a simplified time series data curve.

5. The ultra-real-time simulation method for digital twin networks according to claim 1, characterized in that: The curve optimization process steps are as follows: Step 1: Obtain simulation time-characteristic diagram; Step 2: Set the threshold , use the Douglas Peucker algorithm to simplify the time-characteristic graph obtained in the previous step; Step 3: The user can select the multiple of the ultra-real-time simulation process independently; Step 4: Compare the number of remaining nodes with the estimated number of nodes, and perform different operations according to the three different comparison results. In case 1, when the number of remaining nodes is greater than the estimated number of nodes, a node removal algorithm is used to delete redundant and less important nodes. In case 2, when the number of remaining nodes is less than the estimated number of nodes, Lagrange interpolation is used to interpolate at locations with large node spacing. In case 3, when the number of remaining nodes is equal to the estimated number of nodes, no additional operations are required.

6. The ultra-real-time simulation method for digital twin networks according to claim 1, characterized in that: The simulation and deduction module has a deduction scheduling control function. The simulation and deduction module starts the deduction simulation and sends the simulation ultra-real-time step and ultra-real-time time series data business to the business generation module through the HTTP network interface and RESTFUL API interface specification. The sent content can be dynamically changed according to user operations, including control instructions such as end, pause, and restart.

7. The ultra-real-time simulation method for digital twin networks according to claim 1, characterized in that: The network simulation module fine-tunes virtualized network elements through configuration commands and scripts, customizing each element's behavior, routing rules, security policies, and interconnections to meet network planning requirements. Connection configuration parameters include bandwidth, latency, and packet loss rate, as well as inter-radio radio link parameters, including key parameters like channel, signal-to-noise ratio, and transmit power.

Images