A network anti-damage performance optimization method based on directed h-degree
By using directed h-degree optimization methods in the device guarantee network, selecting enhanced anti-destruction nodes and designing redundant links, the problem of insufficient network anti-destruction and connectivity in the existing technology is solved, and the security of the network is significantly improved. and information dissemination capabilities.
Patent Information
- Application Number
- CN202110910861.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-08-10
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2041-08-10
AI Technical Summary
The existing technology cannot simultaneously consider the selection of enhanced anti-destruction nodes in the network and improve connectivity reliability, resulting in the network being unable to recover in time and resist the risk of interruption in the face of failures, and lacking effective information dissemination and topology optimization solutions.
Adopt the network damage resistance performance optimization method based on directed h degree, and ensure the network through analysis equipment, calculate the directed h degree, select enhanced anti-destruction nodes, and design redundant links to improve the network's destruction resistance and connectivity reliability. Establish mathematical models to optimize network topology.
Directed h-degree better reflects the importance of network nodes, improves the overall damage resistance of the network, significantly improves the security and reliability of the network, reduces the possibility of network sharding, and ensures connectivity of information dissemination.
Smart Images

Figure CN113673064B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of network equipment reliability, and in particular to a network anti-damage performance optimization method based on directional h-degree. Background Art
[0002] The equipment security network is composed of multiple structures that coexist and change, and its topology will consciously or unconsciously present the characteristics of "scale-free". In a scale-free network, the failure of distributed nodes and key subnets may cause the entire network to collapse, leading to serious consequences. Therefore, studying the anti-vulnerability strategy of distributed nodes and exploring network topology optimization solutions not only provide important means for effectively preventing network attacks and improving anti-destruction capabilities, but also are an important guarantee for network security.
[0003] Networks are divided into weighted information networks and unweighted information networks according to whether there are weights between the links. Weighted information networks are further divided into directed weighted networks and undirected weighted networks. Traditional network structure research mainly focuses on undirected networks and weighted networks. Although most shared information networks are undirected and weighted networks, h-degree can be used for quantitative analysis, but many networks in military operations are directed networks. In such networks, the two directional connections of the same node cannot be combined for h-degree calculation. Therefore, for directed weighted information networks, the original h-degree still needs to be improved. This can better reflect the importance and function of network nodes and provide a theoretical basis for improving the anti-damage performance of the network.
[0004] Therefore, it is necessary to study the network anti-damage performance optimization method for directed weighted networks, optimize from two aspects: enhancing the selection of anti-damage nodes and improving connectivity reliability, and establish a mathematical model of the above contents to optimize the network anti-damage performance.
[0005] Therefore, it is necessary to provide a network anti-damage performance optimization method based on directed h-degree to solve the above technical problems. Summary of the invention
[0006] The present invention provides a network anti-damage performance optimization method based on directed h-degree, which solves the problems that the existing network equipment cannot simultaneously consider the selection of enhanced anti-damage nodes of the equipment protection network and improve the connection reliability, cannot obtain the best network topology solution, cannot ensure the timely dissemination of information, and lacks the ability to resist interruption risks.
[0007] In order to solve the above technical problems, the present invention provides a network anti-damage performance optimization method based on directional h-degree, which comprises the following steps:
[0008] S1. Analyze the equipment support network and abstract the equipment support system into a network structure based on the subject and structure of the equipment support system;
[0009] S2. Calculate the directed h-degree of the equipment support network nodes and select enhanced anti-destruction nodes. Use the directed h-degree to perform quantitative calculations to conduct quantitative analysis on the equipment support system and select the nodes that need to enhance anti-destruction performance.
[0010] S3, equipment security network h optimal centralization model;
[0011] S4. Design redundant links to improve the network's anti-destruction capability. Design a ring link to connect the adjacent networks of distributed nodes. By setting up several adjacent networks, there will be many ring links. The average shortest path between nodes in the network is used to measure whether the optimized topology is optimal.
[0012] S5. Comprehensively consider and obtain the network topology optimization plan.
[0013] Preferably, in step S1, the equipment protection network adopts a tree-like equipment protection network structure, that is, equipment protection is provided from high-level nodes to low-level nodes.
[0014] Preferably, in step S2, the directed h-degree is divided into an in-h-degree and an out-h-degree, and the in-h-degree and the out-h-degree represent the in-connection and the out-connection of the node respectively.
[0015] Preferably, in step S2, the essence of the in-h centrality and out-h centrality is the normalization of the in-h degree and out-h degree, which will not change the level and is used to compare nodes from different networks or dynamic networks.
[0016] Preferably, in step S2, after calculating the directional h-degree of each node, they are sorted from large to small. Points with large values are nodes that have a greater impact on the network and need to be determined as enhanced anti-destruction nodes, and improved by redundancy or changing the network structure.
[0017] Preferably, in step S3, the h-degree algorithm emphasizes that the central node should meet two conditions at the same time: first, it should be connected to many nodes; second, it should be closely connected with other nodes.
[0018] Preferably, in step S4, the segmented networks are connected by a link so that there is at least one link between any two nodes in the entire network, thereby achieving communication between any two nodes.
[0019] Preferably, in step S5, it is necessary to comprehensively consider two aspects: enhancing the survivability of nodes and improving the connectivity reliability of the network.
[0020] Preferably, in step S5, the enhanced indestructible nodes may be improved by means of redundancy or changing the network structure, and the h-center value of the entire network may be optimized by changing the edge weights.
[0021] Preferably, in step S5, a network topology optimization strategy based on redundant links may be used to optimize the damaged network.
[0022] Compared with the related art, the network anti-damage performance optimization method based on directional h-degree provided by the present invention has the following beneficial effects:
[0023] The present invention provides a network anti-damage performance optimization method based on directed h-degree, optimizes the anti-damage performance of the equipment guarantee network, studies the optimization method oriented to network reliability, and establishes a mathematical model and solution algorithm for optimizing the network anti-damage performance while considering the selection of enhanced anti-damage nodes of the equipment guarantee network and the improvement of connectivity reliability.
[0024] 1. Directed h-degree can better reflect the importance of nodes in a directed network than h-degree.
[0025] 2. The network center h-degree can better reflect the anti-damage performance of the entire network and provide new objective functions and indicators for network structure optimization.
[0026] 3. Adopt network topology optimization based on redundant links. (1) Do not destroy the existing network structure. Node-based optimization reduces network risks by changing the distribution of nodes. On the other hand, the optimization scheme based on redundant links does not change the connection relationship between existing nodes. It only improves the security of the network by adding links and increasing redundancy. (2) Significantly improve network security. The network optimized by layered processing method is more inclined to tree shape. Once attacked, the network is likely to be divided into multiple fragments. The link-based optimization scheme reduces the possibility of this situation. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1 A schematic diagram of the equipment security network optimization process of the present invention;
[0028] Figure 2 It is the structure diagram of the equipment support system;
[0029] Figure 3 This is an example of calculating the input and output h values of point A;
[0030] Figure 4 The change of the h value of the entire network center after deleting a node;
[0031] Figure 5 The central nodes of each network slice are connected to form a ring-shaped interchange structure;
[0032] Figure 6 To compare the pictures before and after optimization; DETAILED DESCRIPTION
[0033] The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.
[0034] Please refer to Figure 1-Figure 6 ,in, Figure 1 A schematic diagram of the equipment security network optimization process of the present invention; Figure 2 It is the structure diagram of the equipment support system; Figure 3 This is an example of calculating the input h value and the output h value of point A. The network anti-damage performance optimization method based on directed h degree includes the following steps:
[0035] S1. Analyze the equipment support network and abstract the equipment support system into a network structure based on the subject and structure of the equipment support system;
[0036] S2. Calculate the directed h-degree of the equipment support network nodes and select enhanced anti-destruction nodes. Use the directed h-degree to perform quantitative calculations to conduct quantitative analysis on the equipment support system and select the nodes that need to enhance anti-destruction performance.
[0037] S3, equipment security network h optimal centralization model;
[0038] S4. Design redundant links to improve the network's anti-destruction capability. Design a ring link to connect the adjacent networks of distributed nodes. By setting up several adjacent networks, there will be many ring links. The average shortest path between nodes in the network is used to measure whether the optimized topology is optimal.
[0039] S5. Comprehensively consider and obtain the network topology optimization plan.
[0040] See also Figure 1 , Figure 1 The middle node represents the warehouse or equipment user unit, the line represents the supply relationship, and the value on the line represents the supply quantity.
[0041] In the equipment support system network, the network nodes represent the entities involved in equipment support, and the connecting lines in the network represent the spare parts flow, information flow, capital flow, etc. According to the supply relationship, we can get Figure 1 Equipment protection network structure of equipment protection system.
[0042] Since the equipment support system is a directed weighted network, it cannot be quantitatively analyzed using traditional h-degree. Therefore, we introduce directed h-degree for quantitative calculation to screen out nodes that need to enhance anti-destruction performance. In addition, different network topologies will also have a great impact on the network's anti-destruction performance. When building a network, how to select enhanced anti-destruction nodes and improve the network's connectivity reliability is of great significance to improving the network's anti-destruction performance.
[0043] In the step S1, the equipment protection network adopts a tree-like equipment protection network structure, that is, equipment protection is provided from high-level nodes to low-level nodes.
[0044] In step S2, the directed h-degree is divided into an in-h-degree and an out-h-degree, and the in-h-degree and the out-h-degree represent the in-connection and the out-connection of the node respectively.
[0045] Indestructibility, also known as survivability, refers to the ability of a system to still complete its functions after being damaged. In 1999, Albert proposed a scale-free network model by studying networks such as the World Wide Web and cellular networks. He also placed the random network model and the scale-free network model in two types of attacks: random attacks and deliberate attacks. He conducted a lot of research on the network indestructibility of static topology, and Ellison et al. proposed a framework for analyzing the indestructibility of information systems.
[0046] Ellison proposed the concept of information system survivability and gave the SNA evaluation method. Survivability refers to the ability of a system to complete its tasks in the face of attacks and failures within a certain period of time. The SNA method is a method for analyzing the information system survivability framework.
[0047] Nodes and edges are the basis of the network, and h-degree can effectively measure nodes and their related connections. We use directed h-degree to measure the enhanced nodes of the directed weighted network.
[0048] Directed h-degree is divided into in-h-degree and out-h-degree and is defined as.
[0049] Definition 1. In-h-degree (hi). In a directed weighted network, the in-h-degree hi(n) of node n means that this node maintains in-connections greater than or equal to hi(n) with at most other hi(n) nodes.
[0050] Definition 2. Out-h-degree (ho): In a directed weighted network, the out-h-degree ho(n) of node n means that this node maintains out-links greater than or equal to h0(n) with at most other h0(n) nodes.
[0051] By definition, in-h-degree and out-h-degree characterize the in-connections and out-connections of a node, respectively. A high in-h-degree (or out-h-degree) indicates that the node not only links to many other nodes, but also maintains relatively high strength of in-links (or out-links) with these linked nodes. This directed h-degree can be quickly calculated by a program.
[0052] In step S2, the essence of the in-h centrality and out-h centrality is the normalization of the in-h degree and out-h degree, which will not change the level and is used to compare nodes from different networks or dynamic networks.
[0053] In step S2, after calculating the directional h-degree of each node, they are sorted from large to small. Points with large values are nodes that have a greater impact on the network and need to be determined as enhanced anti-destruction nodes, and improved by redundancy or changing the network structure.
[0054] Please refer again Figure 3 and Figure 4 , Figure 3 The data in the table are the results calculated according to the program. Note: The link weights of A are (6, 4, 4, 2, 2) from large to small. Therefore, the link weight of A is (6, 4, 4, 2, 2). The strength of A is 6+4+4+2+2=18, and the in-h degree is 2 (the strength of 3 in-links is greater than or equal to 2). In addition, the out-link weight of A is (4, 4, 2), so its out-degree is 4+4+2=10, and its out-h degree is 2 (the strength of 2 out-links is greater than or equal to 2).
[0055] In a network with N nodes, Ni(N) represents the number of neighbor nodes of node n that maintain inlinks with node n, No(N) is the number of neighbor nodes of node n that maintain outlinks with node n, Di(N) and Do(N) represent the in-degree and out-degree of node n. Then it is easy to check the following inequality:
[0056] 0≤hI(n)≤NI(n) <N(1)
[0057] 0≤ho(n)≤No(n) <N(2)
[0058] dI(n)≥h2I(n)(3)
[0059] do(n)≥h2o(n)(4)
[0060] In a directed weighted network with N nodes (N>1), the in-h-centrality (CHI(N)) of node n is defined as:
[0061]
[0062] In a directed weighted network with N nodes (N>1), the out-h-centrality (CHo(N)) of node n is defined as:
[0063]
[0064] According to formula (1) and (2), the following formula can be derived.
[0065]
[0066]
[0067] The essence of in-h centrality and out-h centrality is the normalization of its in-h degree and out-h degree, which will not change its rank. Therefore, they may be used to compare nodes from different networks or dynamic networks.
[0068] After calculating the directed h-degree of each node, sort them from large to small. The nodes with large values have a greater impact on the network and need to be determined as enhanced anti-destruction nodes. They can be improved by redundancy or changing the network structure. The calculation and sorting results of the directed h-degree are shown in the table:
[0069]
[0070]
[0071] In step S3, the h-degree algorithm emphasizes that the central node should meet two conditions at the same time: first, it should be connected to many nodes; second, it should be closely connected with other nodes.
[0072] For directed h-degree, when the number of incoming connections (or outgoing connections) of a node is 0, the in-h-degree (or out-h-degree) is 0. When the number of incoming connections (or outgoing connections) of a node is large but the strength is 1, the in-h-degree (or out-h-degree) is only 1. When the number of incoming connections (or outgoing connections) of a node is 1, no matter how the strength of the connection is, the in-h-degree (or out-h-degree) will not be greater than l.
[0073] Objective function: The h-centrality of the overall out-degree and in-degree of the equipment guarantee network:
[0074] max
[0075] max
[0076] The model needs to satisfy the following constraints:
[0077] The h-degree algorithm emphasizes that the central node should meet two conditions at the same time: first, it is connected to many nodes; second, the connection with other nodes should be relatively close. Therefore, for directed h-degree, when the number of incoming connections (or outgoing connections) of a node is 0, the in-h-degree (or out-h-degree) is 0. When the number of incoming connections (or outgoing connections) of a node is large but the strength is 1, the in-h-degree (or out-h-degree) is only 1. When the number of incoming connections (or outgoing connections) of a node is 1, no matter how the strength of the connection is, the in-h-degree (or out-h-degree) will not be greater than 1, as shown in formulas (3) and (4).
[0078] The established model can be solved using the directed h-centrality algorithm. The solution results are shown in Figure 4 :
[0079] Figure 4This is the change of the h value of the entire network center after deleting a certain node. It can be seen from the figure that after deleting points 12, 13, and 14, the h value of the network center is almost zero, indicating that these three points are very important, which is consistent with the previous enhanced damage point analysis results, and once again verifies the effectiveness of the method
[0080] The calculation code and process of directed h-degree.
[0081] The calculation procedure of the directional h-degree is shown below.
[0082]
[0083]
[0084] Step 2: Code for the network center h value.
[0085] The calculation procedure of the network center h value is shown below.
[0086]
[0087]
[0088]
[0089]
[0090]
[0091]
[0092]
[0093] In step S4, the segmented networks are connected by a link so that there is at least one link between any two nodes in the entire network, thereby achieving communication between any two nodes.
[0094] Taking the topology of a ring network structure as an example, there is a distributed node in the network. When the node is destroyed, the edges connected to it will also fail. Since the nodes adjacent to the distributed node are not all connected to each other, after losing the aggregation of the distributed node, these nodes will be included in multiple network fragments. This article calls this network fragment the adjacency network of the distributed node.
[0095] For the entire network, if a distributed node is lost, many pairs of nodes will not be able to communicate with each other. When the network is attacked, in order to ensure communication between the remaining nodes, corresponding measures must be taken to ensure connectivity between the remaining nodes. Therefore, a segmented network is used to connect them with a link, so that there is at least one link between any two nodes in the entire network, and communication between any two nodes can be achieved.
[0096] The algorithm for selecting the average shortest path between nodes in the network to measure whether the optimized topology is optimal is as follows:
[0097] For network A, its topology optimization algorithm is as follows:
[0098] Step 1: Find the node Nodei with the maximum degree in network A that is not involved in the optimization;
[0099] Step 2: Construct the network Ai′ which includes all nodes except Nodei;
[0100] Step 3: Ai′ is composed of multiple network fragments, which are recorded as Ai1′, Ai2′, …, Ain′;
[0101] Step 4: Count the number of nodes in Ai1′, Ai2′, …, Ain′, recorded as Ni1′, Ni2′, …, Nin′;
[0102] Step 5: Find the geometric center nodes in the Ai 1′, Ai2′, …, Ain′ network fragments, denoted as Nodei1′, Nodei2′, …, Nodein′, and each center node satisfies the shortest total distance to other nodes in the network.
[0103] Step 6: Assume that Ni 1′>Ni 2′>…>Nin′, and then connect the central nodes of each network slice to form a ring interchange structure. Figure 5 .
[0104] Step 7: Mark Nodei as an optimized node;
[0105] Step 8: Count the number of nodes after optimization to determine whether the optimization ratio required by the solution is met. If successful, terminate the optimization program, otherwise repeat the first step.
[0106] We use 10 and 100-800 network nodes to perform damage simulation and optimize them using the proposed method. The results are shown in Figure 6 shown.
[0107] in accordance with Figure 6 It can be seen that the results after optimization are better than those before optimization, indicating that the optimization method is feasible and effective.
[0108] In step S5, it is necessary to comprehensively consider two aspects: enhancing the survivability of nodes and improving the connectivity reliability of the network.
[0109] By adding enhanced resistant nodes and integrating the network topology optimization strategy based on redundant links, an optimization plan for the topology structure of the equipment protection network is formulated.
[0110] In step S5, the enhanced indestructible nodes may be improved by means of redundancy or changing the network structure, and the h-center value of the entire network may be optimized by changing the edge weights.
[0111] In step S5, the damaged network may be optimized using a network topology optimization strategy based on redundant links.
[0112] Compared with the related art, the network anti-damage performance optimization method based on directional h-degree provided by the present invention has the following beneficial effects:
[0113] In order to optimize the anti-damage performance of the equipment support network, an optimization method oriented to network reliability was studied. While considering the selection of enhanced anti-damage nodes of the equipment support network and the improvement of connectivity reliability, a mathematical model and solution algorithm for optimizing the network anti-damage performance were established.
[0114] 1. Directed h-degree can better reflect the importance of nodes in a directed network than h-degree.
[0115] 2. The network center h-degree can better reflect the anti-damage performance of the entire network and provide new objective functions and indicators for network structure optimization.
[0116] 3. Use network topology optimization based on redundant links. (1) Do not destroy the existing network structure. Node-based optimization reduces network risks by changing the distribution of nodes. On the other hand, the optimization scheme based on redundant links does not change the connection relationship between existing nodes. It only improves the security of the network by adding links and increasing redundancy. (2) Significantly improve network security. The network optimized by the hierarchical processing method is more inclined to tree shape. Once attacked, the network is likely to be divided into multiple fragments. The link-based optimization scheme reduces the possibility of this situation.
[0117] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A network anti-damage performance optimization method based on directed h-degree, It is characterized in that The following steps are involved: S1. Analyze the equipment support network and abstract the equipment support system into a network structure based on the subject and structure of the equipment support system; S2. Calculate the directed h-degree of the equipment support network nodes and select enhanced anti-destruction nodes. Use the directed h-degree to perform quantitative calculations to conduct quantitative analysis on the equipment support system and select nodes that need to enhance anti-destruction performance. S3, equipment security network h optimal centralization model; S4. Design redundant links to improve the network's anti-destruction capability. Design a ring link to connect the adjacent networks of distributed nodes. By setting up several adjacent networks, there will be many ring links. The average shortest path between nodes in the network is used to measure whether the optimized topology is optimal. S5. Comprehensively consider and obtain the network topology optimization scheme; In a network with N nodes, Ni(n) represents the number of neighbor nodes of node n that maintain an incoming link with node n, No(n) is the number of neighbor nodes of node n that maintain an outgoing link with node n, di(n) and do(n) represent the in-degree and out-degree of node n; check the following inequality: 0≤hi(n)≤Ni(n) <N(1) 0≤ho(n)≤No(n) <N(2) di(n)≥h 2 i(n)(3) do(n)≥h 2 o(n)(4) In a directed weighted network with N nodes, N>1, the in-h-centrality (C hi(n) ) is defined as: In a directed weighted network with N nodes, N>1, the out-h-centrality (C ho(n) ) is defined as: According to formula (1) and (2), the following formula can be derived: The essence of in-h centrality and out-h centrality is the normalization of its in-h degree and out-h degree, which will not change its level. They are used to compare nodes from different networks or dynamic networks. After calculating the directed h-degree of each node, sort them from large to small. The nodes with large values have a greater impact on the network and need to be determined as enhanced anti-destruction nodes; Objective function: The h-centrality of the overall out-degree and in-degree of the equipment guarantee network: The model needs to satisfy the following constraints: The h-degree algorithm emphasizes that the central node should meet two conditions at the same time: first, it is connected to many nodes; second, the connection with other nodes should be relatively close. For directed h-degree, when the number of incoming connections or outgoing connections of a node is 0, the in-h-degree or out-h-degree is 0. When the number of incoming connections or outgoing connections of a node is large but the strength is 1, the in-h-degree or out-h-degree is only 1. When the number of incoming connections or outgoing connections of a node is 1, no matter what the strength of the connection is, the in-h-degree or out-h-degree will not be greater than 1.
2. According to the method for optimizing network anti-damage performance based on directional h-degree according to claim 1, It is characterized in that In the step S1, the equipment protection network adopts a tree-like equipment protection network structure, that is, equipment protection is provided from high-level nodes to low-level nodes.
3. According to the method for optimizing network anti-damage performance based on directed h-degree in claim 1, It is characterized in that In step S2, the directed h-degree is divided into an in-h-degree and an out-h-degree, and the in-h-degree and the out-h-degree represent the in-connection and the out-connection of the node respectively.
4. According to the method for optimizing network anti-damage performance based on directed h-degree in claim 1, It is characterized in that In step S2, the essence of the in-h centrality and out-h centrality is the normalization of the in-h degree and out-h degree, which will not change the level and is used to compare nodes from different networks or dynamic networks.
5. According to the method for optimizing network anti-damage performance based on directed h-degree in claim 1, It is characterized in that In step S2, after calculating the directional h-degree of each node, they are sorted from large to small. Points with large values are nodes that have a greater impact on the network and need to be determined as enhanced anti-destruction nodes, and improved by redundancy or changing the network structure.
6. The network anti-damage performance optimization method based on directed h-degree according to claim 1, It is characterized in that In step S4, the segmented networks are connected by a link so that there is at least one link between any two nodes in the entire network, thereby achieving communication between any two nodes.
7. The network anti-damage performance optimization method based on directed h-degree according to claim 1, It is characterized in that In step S5, it is necessary to comprehensively consider two aspects: enhancing the survivability of nodes and improving the connectivity reliability of the network.
8. The network anti-damage performance optimization method based on directed h-degree according to claim 7, It is characterized in that In step S5, the enhanced indestructible nodes are improved by redundancy or changing the network structure, and the h-center value of the entire network is optimized by changing the edge weights.
9. The network anti-damage performance optimization method based on directed h-degree according to claim 8, It is characterized in that In step S5, the damaged network is optimized using a network topology optimization strategy based on redundant links.
Citation Information
Patent Citations
Software-definition survivable network construction method
CN105406988A
Lossless path restoration method of software defined destroy-resistant network
CN107733719A
Design of a meta-mesh of chain sub-networks
US20020071392A1
