Combat system modeling method based on multilayer network

Abstract

The invention discloses a combat system modeling method based on a multilayer network, and belongs to the technical field of combat systems. The method specifically comprises the following steps: determining five types of equipment nodes forming a multi-layer combat network model according to a combat process of a generalized OODA ring; based on different interaction relationships and connection probabilities among the equipment nodes, determining edges among the equipment nodes in the multi-layer combat network model; according to the type of equipment nodes, a sensing subnet, a communication subnet, a command and control subnet and a strike subnet are divided, so that a combat system is modeled into a multi-layer combat network model comprising five types of nodes and four subnets. According to the method, the multi-layer combat network with multiple links, high flexibility and high survivability can be constructed.

Description

Technical Field

[0001] The present invention belongs to the technical field of combat systems, and in particular relates to a combat system modeling method based on a multi-layer network. Background Art

[0002] With the development of information technology and the introduction of new combat concepts, modern combat systems have gradually evolved from traditional linear structures to more complex mosaic warfare, decision-making center warfare, and joint global operations. This transformation has led to the evolution of the kill chain to a more complex, systematic, and networked kill network. First of all, it is necessary to consider the ability of the combat network to maintain basic functions and combat capabilities when attacked or destroyed. Therefore, how to build a highly resistant combat network is of great significance to the mission success rate of the combat system.

[0003] At present, complex network theory is the most widely used framework in the field of network modeling research. However, with the increasing complexity of offensive and defensive confrontations in wars, it is difficult to fully describe the combat confrontation process by relying solely on complex network theory. Summary of the invention

[0004] In view of this, the present invention proposes a combat system modeling method based on a multi-layer network, which can construct a multi-layer combat network with multiple links, high flexibility and high anti-destruction ability.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] A combat system modeling method based on a multi-layer network includes the following steps:

[0007] S1, according to the combat process of the generalized OODA loop, determine the five types of equipment nodes that constitute the multi-layer combat network model;

[0008] S2, based on the different interaction relationships and connection probabilities between equipment nodes, determine the edges between equipment nodes in the multi-layer combat network model;

[0009] S3, according to the type of equipment nodes, divide the sensor subnet, communication subnet, command and control subnet and strike subnet, so that the combat system is modeled as a multi-layer combat network model containing 5 types of nodes and 4 subnets.

[0010] Furthermore, the five types of equipment nodes are:

[0011] Enemy target node T: the enemy target to be attacked in the combat mission;

[0012] Reconnaissance sensor node S: responsible for target detection and analysis;

[0013] Command and control node D: responsible for issuing commands and controlling battlefield situation;

[0014] Strike weapon node A: Carry out a devastating strike on enemy targets;

[0015] Communication network node C: responsible for information transmission and transfer.

[0016] Further, step S2 includes the following steps:

[0017] S201: Based on the combat process of the generalized OODA loop, the interaction relationships between equipment nodes are divided into 7 types, corresponding to the 7 types of edges in the multi-layer combat network model:

[0018] E T→S ,E S→S ,E S→C→D ,E D→D ,E D→C→S ,E D→A→T ,E A→T

[0019] Among them, E represents the edge, and the subscripts of E represent:

[0020] T→S represents the sensing and detection activities of the reconnaissance node on the target node;

[0021] S→C means that the communication node receives the information transmitted by the reconnaissance node, stores and forwards it;

[0022] C→D indicates that the command and control node receives information transmitted by the communication node;

[0023] D→A means that the command and control node issues combat orders to the strike node;

[0024] A→T means the strike node performs the mission and conducts a military strike on the target;

[0025] S→S represents information sharing between reconnaissance nodes;

[0026] D→D represents the level command and coordinated command activities between the command and control nodes;

[0027] D→C means the command node feeds back the decision information to the communication node and stores it;

[0028] C→S means that the communication node feeds back the cached decision information to the scout node;

[0029] S202: Based on the 7 types of edges, 7 combat kill chains are formed:

[0030] T→S→C→D→A→T is a typical kill chain;

[0031] T→S→S→C→D→A→T is a kill chain with information sharing;

[0032] T→S→C→D→D→A→T is a kill chain with collaborative decision-making;

[0033] T→S→S→C→D→D→A→T is a kill chain with both information sharing and collaborative decision-making;

[0034] T→S→C→D→C→S→C→D→A→T is a kill chain with information feedback;

[0035] T→S→S→C→D→C→S→C→D→A→T is a kill chain with information sharing and feedback;

[0036] T→S→C→D→D→C→S→C→D→A→T is a kill chain with collaborative decision-making and information feedback;

[0037] S203: Connect equipment nodes according to the combat kill chain, and set the connection probability between the two connected equipment nodes i and j according to the link attributes

[0038] S204: Setting connection probability threshold P min , the probability of disconnection Less than P min Set an edge between two equipment nodes with a connection and set the connection probability as the edge weight.

[0039] Furthermore, the sensor subnet is composed of reconnaissance sensor nodes S, the communication subnet is composed of communication network nodes C, the command and control subnet is composed of command and control nodes D, and the strike subnet is composed of strike weapon nodes A.

[0040] The beneficial effects of the present invention are:

[0041] The present invention is based on a multi-layer network combat system modeling method, which decomposes the combat node composition and the relationship between nodes based on a generalized OODA loop, defines the node's own attributes based on the positive and negative effects of equipment nodes, and determines that the combat network can be modeled as 5 types of nodes, 7 types of edges, and 4 layers of subnets, thereby establishing a multi-layer combat network model. The network model established according to the present invention can efficiently formulate combat plans and facilitate guiding combat practices. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the implementation of the present invention or the existing technical solutions, the following briefly introduces the drawings required to be used in the embodiments or the description of the prior art.

[0043] Figure 1 It is a flow chart of a combat system modeling method based on a multi-layer network in an embodiment of the present invention.

[0044] Figure 2 is a topological diagram of a multi-layer combat network in an embodiment of the present invention;

[0045] Figure 3 is a comparison diagram of different invulnerability indicators under random attacks in an embodiment of the present invention;

[0046] Figure 4 is a comparison chart of different invulnerability indicators under moderate attack in an embodiment of the present invention;

[0047] Figure 5 4 is a comparison chart of different invulnerability indicators under betweenness attack in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0049] A combat system modeling method based on multi-layer networks, such as Figure 1 As shown, the following steps are included:

[0050] S1: Define the equipment nodes of the multi-layer combat network model based on the generalized OODA loop, see Table 1 below:

[0051] Table 1 Description of combat node types

[0052]

[0053] S2: Based on the different interaction relationships and connection probabilities between equipment nodes, the edges in the network are constructed; the specific method is as follows:

[0054] S201: Based on the combat process of the generalized OODA loop, the interaction relationships between nodes are divided into 7 types, corresponding to the 7 types of edges in the multi-layer combat network model:

[0055] E type ={E T→S ,E S→S ,E S→C→D ,E D→D ,E D→C→S ,E D→A→T ,E A→T} (1)

[0056] The connections between different weapon equipment nodes have different practical meanings. Table 2 lists the specific meanings of the existing network connections in the combat system:

[0057] Table 2 Meaning of the links in the combat network

[0058]

[0059] S202: According to the differences in the transmission and decision-making of reconnaissance information by different equipment, the specific connotation of the kill chain is also different. According to the connection relationship between the equipment in the equipment system combat network, 7 generalized kill chains and their meanings are listed in Table 3:

[0060] Table 3 Meaning of different kill chains in combat networks

[0061]

[0062] S203: Connect equipment nodes according to the combat kill chain, and set the connection probability between the two connected equipment nodes i and j according to the link attributes

[0063] S204: Setting connection probability threshold P min , the probability of disconnection Less than P min Set an edge between two equipment nodes with a connection and set the connection probability as the edge weight.

[0064] S3: Divide the sensor subnet according to the attributes of the nodes (G S ), communication subnet (G C ), the charge subnet (G D ) and attack subnet (G A ), thus modeling the combat system as a multi-layer combat network model containing 5 types of nodes and 4 subnets. Among them:

[0065] The sensor subnet is composed of reconnaissance sensor nodes S, which conduct detection and reconnaissance organization and command according to combat missions, realize full-dimensional battlefield situation awareness, and provide intelligence support for the command and control network and the strike network;

[0066] The command and control subnet is composed of the command and control node D, which obtains battlefield situation from the sensor subnet and communication network, makes intelligent decision-making and command, generates combat plans and firepower coordinated strike plans, and evaluates the strike effect;

[0067] The strike subnet is composed of strike weapon node A. According to the coordinated strike plan, it implements cross-link control between the command and control platform and the weapon platform, as well as between the weapon platforms, to achieve composite tracking of multiple weapon platform targets and coordinated firepower strikes.

[0068] The communication subnet is composed of communication network nodes C, which serve as the connecting link between the command and control network and the sensor subnet. A small-world network is formed inside the communication network, which improves the information flow of the overall network and also improves the anti-destruction performance.

[0069] The four subnets are intertwined, interact with each other, and are inseparable, supporting diverse and complex networked combat operations. By adding edges between different types of combat nodes based on information relationships and node attributes, the four different types of subnets can be mapped and integrated into a multi-layer combat network model.

[0070] In one embodiment, the topology diagram of the multi-layer combat network is constructed as follows: Figure 2 As shown in Figure 1, the base network structure of the combat network is distributed as follows: 1) 2 combat targets; 2) 20 combat warning sensor devices; 3) 10 communication relay devices; 4) 4 combat command centers; 5) 6 strike weapons and equipment. According to the connection probability between two nodes Determine the edges between nodes and set the connection probability threshold P min =0.7.

[0071] The following is an evaluation of the above multi-layer combat network. Figure 1 , including the following steps:

[0072] S4: Calculate the contribution of each device node in the multi-layer combat network model and calculate the contribution structural entropy; the specific method is as follows:

[0073] S401: Calculate the average out-degree weight ODE and the average in-degree weight IDE of the node:

[0074]

[0075] Among them, d + (v) represents the out-degree of node v, d - (v) represents the in-degree of node v, w i is the weight of the ith outgoing edge of node v, w j is the weight of the jth incoming edge of node v;

[0076] The average out-degree weight is defined as the average weight of all outgoing edges of node v's unidirectional connection to the outside, reflecting the node's influence on other nodes in the network or its information output capability. The average in-degree weight is defined as the average weight of all incoming edges received by node v, reflecting the amount of information or information input capability of the node in the network.

[0077] The functional attributes of the node are determined by the out-degree average weight ODE and the in-degree average weight IDE:

[0078] S i =ωODE+(1-ω)IDE

[0079] Among them, ω is the weight parameter;

[0080] The structural properties of a node are obtained by calculating the proportion of a node in all shortest paths:

[0081]

[0082] Among them, σ st is the number of all shortest paths between node s and node t, σ st (i) is the number of shortest paths between s and t that pass through node i;

[0083] S402: Calculate the contribution ND of node i i :

[0084]

[0085] Among them, T i is the type weight of node i, and the sum of the node type weights is 1; M F is the node structure attribute weight, M S is the node function attribute weight, and M F +M S =1; N is the number of nodes in the multi-layer combat network;

[0086] The importance of nodes in a multi-layer combat network depends on the contribution of two attributes to the network node and the type weight of the node.

[0087] S403: By extracting the functional attributes and structural attributes of the combat system as microscopic elements and bringing them into the theoretical formula of entropy, the contribution structural entropy E is obtained. ND :

[0088]

[0089] S5: Contribution structure entropy E ND As a measure of network survivability, the damage of the network is judged according to the change of the contribution structural entropy of the multi-layer combat network during network damage, so as to analyze and characterize the network's survivability.

[0090] At present, in the field of network invulnerability assessment, complex network theory still has certain advantages. Indicators such as network connectivity, node degree distribution, average path length, and network efficiency can all measure network invulnerability. However, a common problem with measurement indicators is that each node is treated indiscriminately for statistics, and these indicators focus on measuring invulnerability from a structural perspective, which is suitable for complex networks with indiscriminate nodes, but not for combat networks with heterogeneity. This embodiment uses contribution structural entropy as an invulnerability assessment indicator for combat networks, and its accuracy and sensitivity are significantly better than other network structural entropy invulnerability assessment indicators.

[0091] The following simulation is performed using Python 3.10:

[0092] Firstly, taking the joint assault operation under the background of system combat as an example, a multi-layer combat network model is constructed. Secondly, the network is attacked by random attack and deliberate attack strategies. The contribution degree structure entropy E(ND), degree distribution entropy E(DD), degree structure entropy E(SD), betweenness structure entropy E(BD), point-edge difference entropy E(CD) and relative benchmark degree difference entropy E(AD) network structure entropy are used to monitor the invulnerability of the combat network.

[0093] This embodiment assumes that the average out-degree weight is equal to the average in-degree weight, and sets ω = 0.5. In practice, functional attributes and structural attributes are often interdependent, and M is set F =M S =0.5.

[0094] The core of network survivability assessment is to evaluate the performance of the network after being attacked. Common attack methods can be divided into two categories: random attacks and deliberate attacks. Random attacks refer to randomly selecting a node in the network to attack in each step, while deliberate attacks are more selective, selecting the most important node evaluated according to certain criteria to attack in each step.

[0095] Under random attack mode, the indicators change as follows Figure 3 As shown in the figure, it can be observed that the changes of various indicators are relatively slow, and some invulnerability indicators may fluctuate due to changes in network structure, but the overall trend is downward. Among them, the contribution structural entropy decreases the fastest, and the betweenness structural entropy decreases rapidly after attacking 24 nodes. The remaining indicators show an accelerated downward trend only after attacking 35 nodes.

[0096] When a deliberate attack is used, the indicators change as follows Figure 4 , Figure 5 As shown in the figure. The destructiveness is significantly higher than that of random attacks, and the index decreases rapidly. Under the strategies of degree-first attack and betweenness-first attack, the degree structural entropy, betweenness structural entropy, and contribution structural entropy drop significantly when attacking the first 2-3 nodes, and the contribution structural entropy drops the most. When attacking around the 6th node, the contribution structural entropy drops significantly again. When the number of attacked nodes reaches 10-11, the contribution structural entropy continues to drop until the entropy value becomes 0.

[0097] In contrast, although the degree structural entropy experienced a significant decline in the early stage, it tended to be flat after the number of attack nodes exceeded two. Other indicators remained relatively stable in the early stage, but as the number of attack nodes increased, the entropy value gradually decreased slowly, indicating that these indicators were slow to respond when evaluating the invulnerability of the network structure. It can be seen that the evaluation indicators based on the contribution structural entropy are particularly sensitive to node failures and can quickly reflect major changes in the network structure, thereby more accurately evaluating the network's invulnerability.

[0098] In summary, the method of the present invention can construct a multi-layer combat network with multiple links, high flexibility and high anti-destruction ability, which is convenient for efficiently formulating combat plans and guiding combat practice.

Claims

1. A combat system modeling method based on a multi-layer network, characterized in that: The following steps are involved: S1, according to the combat process of the generalized OODA loop, determine the five types of equipment nodes that constitute the multi-layer combat network model; S2, based on the different interaction relationships and connection probabilities between equipment nodes, determine the edges between equipment nodes in the multi-layer combat network model; S3, according to the type of equipment nodes, divide the sensor subnet, communication subnet, command and control subnet and strike subnet, so that the combat system is modeled as a multi-layer combat network model containing 5 types of nodes and 4 subnets.

2. The combat system modeling method based on a multi-layer network according to claim 1 is characterized in that: The five types of equipment nodes are: Enemy target node T: the enemy target to be attacked in the combat mission; Reconnaissance sensor node S: responsible for target detection and analysis; Command and control node D: responsible for issuing commands and controlling battlefield situation; Strike weapon node A: Carry out a devastating strike on enemy targets; Communication network node C: responsible for information transmission and transfer.

3. The combat system modeling method based on a multi-layer network according to claim 1 is characterized in that: Step S2 includes the following steps: S201: Based on the combat process of the generalized OODA loop, the interaction relationships between equipment nodes are divided into 7 types, corresponding to the 7 types of edges in the multi-layer combat network model: AND T→s ,AND S→S ,AND S→C→D ,AND D→D ,AND D→C→s ,AND D→A→T ,AND A→T Among them, E represents the edge, and the subscripts of E represent: T→S represents the sensing and detection activities of the reconnaissance node on the target node; S→C means that the communication node receives the information transmitted by the reconnaissance node, stores and forwards it; C→D indicates that the command and control node receives information transmitted by the communication node; D→A means that the command and control node issues combat orders to the strike node; A→T means the strike node performs the mission and conducts a military strike on the target; S→S represents information sharing between reconnaissance nodes; D→D represents the level command and coordinated command activities between the command and control nodes; D→C means the command node feeds back the decision information to the communication node and stores it; C→S means that the communication node feeds back the cached decision information to the scout node; S202: Based on the 7 types of edges, 7 combat kill chains are formed: T→S→C→D→A→T is a typical kill chain; T→S→S→C→D→A→T is a kill chain with information sharing; T→S→C→D→D→A→T is a kill chain with collaborative decision-making; T→S→S→C→D→D→A→T is a kill chain with both information sharing and collaborative decision-making; T→S→C→D→C→S→C→D→A→T is a kill chain with information feedback; T→S→S→C→D→C→S→C→D→A→T is a kill chain with information sharing and feedback; T→S→C→D→D→C→S→C→D→A→T is a kill chain with collaborative decision-making and information feedback; S203: Connect equipment nodes according to the combat kill chain, and set the connection probability between the two connected equipment nodes i and j according to the link attributes S204: Setting connection probability threshold P min , the probability of disconnection Less than P min Set an edge between two equipment nodes with a connection and set the connection probability as the edge weight.

4. The combat system modeling method based on a multi-layer network according to claim 1 is characterized in that: The sensor subnet is composed of reconnaissance sensor nodes S, the communication subnet is composed of communication network nodes C, the command and control subnet is composed of command and control nodes D, and the strike subnet is composed of strike weapon nodes A.

Images