Data transmission method and system for protecting device identity anonymity based on route hopping

Abstract

The present application relates to the field of network security, in particular to a data transmission method and system for device identity anonymous protection based on route hopping, the method comprising: generating a series of virtual network topologies based on real network nodes and false nodes; evaluating the nodes in the virtual network topologies to obtain a topology security level of the virtual network topologies, and taking the virtual network topologies with a topology security level greater than a preset threshold as a hopping pool; obtaining a hopping path satisfying a preset constraint condition in the hopping pool by using a backtracking method; screening the hopping path by using an evolutionary game model to obtain an optimal hopping strategy; and transmitting data based on the optimal hopping strategy. The present application firstly constructs multiple virtual network topologies to cause attack confusion, and simultaneously generates and selects a hopping path for the virtual network topologies to obtain a data transmission path with the highest security, thereby significantly improving the defense capability against network attacks.

Description

Technical Field

[0001] This invention relates to the field of network security, and specifically to a data transmission method and system for anonymizing device identity based on routing changes. Background Technology

[0002] With the rapid advancement of the Internet and the Internet of Things (IoT) technologies, the number and types of network devices have exploded, greatly enriching people's daily lives and work, but also bringing unprecedented challenges to device identity protection and network security. In this highly interconnected and information-sharing era, the leakage of any network device's identity or a network security incident can have serious consequences, including but not limited to data breaches, service interruptions, and even system crashes. Traditional network security measures, such as firewalls, intrusion detection systems, and static routing policies, while playing an important role in the past, are no longer sufficient to defend against today's increasingly complex and ever-changing network attacks. These traditional methods are often based on fixed rules and policies, making it difficult to deal with attackers employing advanced techniques and strategies. Existing device identity protection methods mainly rely on static security policies, which are significantly less effective against Advanced Persistent Threats (APTs) and Distributed Denial-of-Service (DDoS) attacks. APT attackers typically possess extremely high technical skills, capable of conducting extensive reconnaissance, deeply researching and exploiting known vulnerabilities and fixed network structures to gradually obtain the identity information of devices within the network, and then launching targeted attacks. DDoS attacks, on the other hand, overwhelm a target server with a massive influx of requests, preventing it from processing legitimate requests and thus paralyzing the service. Therefore, effectively preventing malicious attacks on network devices has become a pressing issue in the field of network security. Summary of the Invention

[0003] To address the limitations of existing technologies in coping with constantly changing network environments and increasingly complex network attacks, this invention, in a first aspect, proposes a data transmission method for anonymizing device identities based on route hopping, comprising:

[0004] A series of virtual network topologies are dynamically generated based on real network nodes and fake nodes;

[0005] The nodes in the virtual network topology are evaluated to obtain the topology security level of the virtual network topology, and the virtual network topology with a topology security level greater than a preset threshold is used as the jump pool.

[0006] The backtracking method is used to obtain the jump path that meets the preset constraints in the jump pool;

[0007] An evolutionary game theory model is used to screen the transition paths to obtain the optimal transition strategy;

[0008] Data is transmitted based on the optimal jump strategy described above.

[0009] Optionally, the step of evaluating the nodes in the virtual network topology to obtain the topology security level of the virtual network topology, and using virtual network topologies with a topology security level greater than a preset threshold as a transition pool, includes:

[0010] A general vulnerability scoring system is used to assess the vulnerabilities of nodes in a virtual network as the probability of node infection;

[0011] The topology security level of the virtual network topology is calculated based on the node infection probability using a topology security level assessment method.

[0012] Virtual network topologies with a topology security level greater than a preset threshold are used as the jump pool.

[0013] Optionally, the step of using virtual network topologies with a topology security level greater than a preset threshold as a jump pool includes:

[0014] The topology security level is compared with a preset threshold. If the topology security level is greater than the preset threshold, the corresponding virtual network topology is used as a jump pool, and the remaining virtual network topologies are discarded. If a preset proportion of virtual network topologies in a series have a security level less than or equal to the preset threshold, the virtual network topology is regenerated until the topology security level of the newly generated virtual network topology is greater than the preset threshold.

[0015] Optionally, the formula for calculating the topology security level of the virtual network topology based on the node infection probability using the topology security level assessment method is as follows:

[0016]

[0017] Where TSL(·) represents the topological security level of the virtual network topology, and G is the directed graph of the virtual network topology. i Let be a directed graph of subgraphs in the virtual network topology, e be the number of directed edges in the virtual network topology, N be the number of nodes in each subnet, and W(a,b) be the probability that node a will infect node b.

[0018] Optionally, obtaining a transition path that satisfies preset constraints in the transition pool using a backtracking method includes:

[0019] The initial node is obtained based on the network topology information of the virtual network topology;

[0020] Randomly select adjacent nodes starting from the initial node;

[0021] Verify whether the path meets the preset constraints. If it does not meet the constraints, return to the previous node and select a new path. If it does meet the constraints, determine whether the node is the destination node or the source node. If it is the destination node or the source node, the generated path is a jump path and stored in the jump path space. If it is not the destination node or the source node, use the node as a new extension node to continue extending the path until the destination node or the source node is reached.

[0022] Optionally, obtaining the initial node based on the network topology information of the virtual network topology includes:

[0023] Based on the network topology information of the virtual network topology, the source node and the destination node are obtained, and the node with fewer adjacent nodes between the source node and the destination node is selected as the initial node.

[0024] Optionally, the preset constraints include: link capacity constraints, transmission delay constraints, reachability constraints, and non-overlapping constraints.

[0025] Optionally, the step of using an evolutionary game model to screen transition paths and obtain the optimal transition strategy includes:

[0026] An evolutionary game model is constructed using the aforementioned jump path as a defense strategy;

[0027] The evolutionary game model is solved using a replication dynamic equation based on Q-learning to obtain the optimal transition strategy.

[0028] Optionally, before generating a series of virtual network topologies based on real network nodes and dummy nodes, the method further includes:

[0029] The system determines the number of times a fake node is scanned within a preset time. If the number of scans exceeds the preset number or more than half of the fake nodes are detected, the source host of the scanned traffic is identified as a malicious host. The malicious host is then isolated and excluded from virtual network topology construction.

[0030] Furthermore, the present invention also provides a data transmission system for anonymizing device identity based on route hopping, comprising:

[0031] The generation module is used to dynamically generate a series of virtual network topologies based on real network nodes and fake nodes;

[0032] The evaluation module is used to evaluate the nodes in the virtual network topology, obtain the topology security level of the virtual network topology, and use the virtual network topology with a topology security level greater than a preset threshold as the jump pool.

[0033] The acquisition module is used to obtain a jump path that meets the preset constraints from the jump pool using a backtracking method;

[0034] The filtering module is used to filter the transition paths using an evolutionary game model to obtain the optimal transition strategy;

[0035] The transmission module is used to transmit data based on the optimal hopping strategy.

[0036] Optionally, the evaluation module evaluates the nodes in the virtual network topology to obtain the topology security level of the virtual network topology, and uses virtual network topologies with a topology security level greater than a preset threshold as a transition pool, including:

[0037] A general vulnerability scoring system is used to assess the vulnerabilities of nodes in a virtual network as the probability of node infection;

[0038] The topology security level of the virtual network topology is calculated based on the node infection probability using a topology security level assessment method.

[0039] Virtual network topologies with a topology security level greater than a preset threshold are used as the jump pool.

[0040] Optionally, the evaluation module uses virtual network topologies with a topology security level greater than a preset threshold as a jump pool, including:

[0041] The topology security level is compared with a preset threshold. If the topology security level is greater than the preset threshold, the corresponding virtual network topology is used as a jump pool, and the remaining virtual network topologies are discarded. If a preset proportion of virtual network topologies in a series have a security level less than or equal to the preset threshold, the virtual network topology is regenerated until the topology security level of the newly generated virtual network topology is greater than the preset threshold.

[0042] Optionally, the evaluation module calculates the topology security level of the virtual network topology based on the node infection probability using the topology security level evaluation method. The calculation formula is as follows:

[0043]

[0044] Where TSL(·) represents the topological security level of the virtual network topology, and G is the directed graph of the virtual network topology. i Let be a directed graph of subgraphs in the virtual network topology, e be the number of directed edges in the virtual network topology, N be the number of nodes in each subnet, and W(a,b) be the probability that node a will infect node b.

[0045] Optionally, the acquisition module uses a backtracking method to obtain a transition path that satisfies preset constraints in the transition pool, including:

[0046] The initial node is obtained based on the network topology information of the virtual network topology;

[0047] Randomly select adjacent nodes starting from the initial node;

[0048] Verify whether the path meets the preset constraints. If it does not meet the constraints, return to the previous node and select a new path. If it does meet the constraints, determine whether the node is the destination node or the source node. If it is the destination node or the source node, the generated path is a jump path and stored in the jump path space. If it is not the destination node or the source node, use the node as a new extension node to continue extending the path until the destination node or the source node is reached.

[0049] Optionally, the acquisition module acquires the initial node based on the network topology information of the virtual network topology, including:

[0050] Based on the network topology information of the virtual network topology, the source node and the destination node are obtained, and the node with fewer adjacent nodes between the source node and the destination node is selected as the initial node.

[0051] Optionally, the preset constraints in the acquisition module include: link capacity constraints, transmission delay constraints, reachability constraints, and non-overlapping constraints.

[0052] Optionally, the filtering module uses an evolutionary game model to filter the transition paths to obtain the optimal transition strategy, including:

[0053] An evolutionary game model is constructed using the aforementioned jump path as a defense strategy;

[0054] The evolutionary game model is solved using a replication dynamic equation based on Q-learning to obtain the optimal transition strategy.

[0055] Optionally, before generating a series of virtual network topologies based on real network nodes and dummy nodes, the generation module further includes:

[0056] The system determines the number of times a fake node is scanned within a preset time. If the number of scans exceeds the preset number or more than half of the fake nodes are detected, the source host of the scanned traffic is identified as a malicious host. The malicious host is then isolated and excluded from virtual network topology construction.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0058] This invention provides a data transmission method and system for anonymizing device identity based on route hopping. The method includes: dynamically generating a series of virtual network topologies based on real network nodes and fake nodes; evaluating the nodes in the virtual network topologies to obtain their topology security levels, and using virtual network topologies with security levels greater than a preset threshold as a hopping pool; obtaining hopping paths that meet preset constraints from the hopping pool using a backtracking method; using an evolutionary game model to filter the hopping paths to obtain the optimal hopping strategy; and transmitting data based on the optimal hopping strategy. This application first constructs multiple virtual network topologies to create attack deception, and simultaneously generates and selects hopping paths for the virtual network topologies to obtain the most secure data transmission path, significantly improving the defense capability against network attacks. Attached Figure Description

[0059] Figure 1 This is a flowchart of the device identity anonymity protection method based on route hopping proposed in this invention;

[0060] Figure 2 The present invention proposes Figure 1 A detailed flowchart of step S2 in the process;

[0061] Figure 3 The present invention proposes Figure 1 A detailed flowchart of step S3 in the process;

[0062] Figure 4 The present invention proposes Figure 1 A detailed flowchart of step S4 in the process;

[0063] Figure 5 This is a flowchart illustrating the overall concept of the device identity anonymity protection method based on route hopping proposed in this application.

[0064] Figure 6 This is a schematic diagram of the data transmission system for anonymizing device identity based on route switching proposed in this invention. Detailed Implementation

[0065] This invention proposes a data transmission method for anonymizing device identities based on route hopping. By dynamically adjusting network topology and path strategies, it enhances network defense capabilities and security. First, a virtual network topology is randomly generated based on collected real and fake nodes. Then, a topology security level assessment method is used to perform security analysis on each node in the fake network topology, evaluating its structural robustness and defense capabilities. Virtual network topologies that meet security level requirements are stored in a hopping pool of the random network topology. Next, a hopping path space is constructed using a backtracking method, ensuring that paths meet standards in terms of capacity, transmission delay, and reachability. Non-overlapping constraints are set to improve path diversity and security. Finally, an optimal hopping strategy is generated through an evolutionary game model, combined with Q-Learning and replicating dynamic equations to optimize the hopping strategy through trial and error. This method, by introducing a route hopping mechanism and game theory, and combining virtual network topology generation, security level assessment, hopping path construction, and dynamic adjustment strategies, achieves significant improvements in device identity anonymization and network security.

[0066] Example 1:

[0067] A method for anonymizing device identities based on route hopping mainly consists of three key parts: constructing a hopping pool for the virtual network topology, generating hopping paths, and selecting the optimal hopping strategy. Figure 1 As shown, it includes S1 to S5.

[0068] S1: Generate a series of virtual network topologies based on real network nodes and fake nodes.

[0069] The generation of fake network topologies is based on information about real nodes in the network (such as the IP addresses and operating system types of real hosts) and fake nodes (fake hosts), constructing a fake network topology structure to resist malicious scanning. By scanning and monitoring actual network nodes, information such as the IP address, operating system type, and open ports of each network node is collected. Using this collected information, a set of fake network nodes can be simulated by rearranging or replacing the hardware configurations and IP addresses of real nodes. These nodes are structurally similar to real network nodes. A series of virtual network topologies are generated by connecting real and fake nodes according to random connection rules and a preset topology structure. This process needs to consider the complexity and diversity of the topologies to increase the difficulty for attackers to identify the fake network. Simultaneously, the fake network topology needs to be updated regularly (e.g., periodically changing the IP addresses of nodes) to prevent attackers from identifying fake nodes through long-term observation. The specific steps are as follows.

[0070] Step S11: First, perform a deep scan and real-time monitoring of the actual nodes in the network to capture and record the core attributes of each node, including but not limited to detailed data such as IP address, operating system type, and open ports. This embodiment assumes that there are three different operating systems in the subnet: Windows_10, Mac_OS_X 10.10, and Ubuntu_14.

[0071] It is particularly important to note that in real networks, there may be instances where the same operating system is deployed within the same subnet. If one host is a malicious node, the attacker, with their prior experience, has a higher probability of successfully infecting the other host. To reduce the probability of infection and increase the security of the network topology, a fake host can be added to the virtual network topology to confuse attackers. For example, suppose there are two hosts running Windows 10 in a subnet, one of which is malicious and the other is the target host. In this scenario, if an additional fake host running the same Windows 10 operating system is deployed in the subnet, and if this fake host is sufficiently deceptive to prevent the attacker from distinguishing between the target host and the fake host, then the malicious host can only randomly choose one of the target host and the fake host to infect, thus reducing the probability of the target host being infected to half. Conversely, if the fake host fails to achieve sufficient deception, then the attacker will spend a significant amount of time identifying it.

[0072] Therefore, by constructing a certain number of virtual hosts, a group of fake network nodes can be simulated by rearranging or replacing the hardware configurations and IP addresses of real nodes. Since virtual hosts also participate in virtual IP hopping, considering the hopping space, one cannot arbitrarily add any number of virtual hosts in pursuit of absolute security. This approach may not only fail to achieve the desired security effect but may also introduce new security risks. Therefore, one should not blindly increase the number of virtual hosts in pursuit of absolute security. Specifically, when the cumulative number of times one or more virtual hosts are scanned reaches a preset threshold, or when more than half of the virtual hosts are scanned, the source host of the scan traffic will be considered malicious and added to a blacklist, thereby achieving effective isolation from other hosts.

[0073] Step S12: By randomly combining these nodes, a set of structurally complete fake network topologies is constructed. This process not only requires accurate simulation of the actual network structure, but also needs to ensure the complexity and diversity of the fake topologies to effectively confuse potential attackers and improve the overall security of the network.

[0074] This method first determines the required types and quantities of nodes such as hosts, switches, and routers, and names them. Then, it generates the connection relationships between nodes according to random connection rules and the selected topology (such as star, tree, mesh, or hybrid topology). Next, it assigns unique IP addresses to nodes and randomly divides subnets and configures gateways. In addition, it assigns random bandwidth and latency values ​​to each connection to simulate the communication characteristics in a real network. Finally, it can generate a visual network topology map to clearly show the network topology structure.

[0075] S2: Evaluate the nodes in the virtual network topology to obtain the topology security level of the virtual network topology, and use the virtual network topology with a topology security level greater than a preset threshold as the jump pool.

[0076] Topological Security Level (TSL) is a security assessment method based on the Common Vulnerability Scoring System (CVSS) that scores the security of each node. Under the premise of quantitative analysis of vulnerabilities in different operating systems within the CVSS framework, i.e., knowing the infection probability of a node, TSL further performs security analysis on each node in the network topology.

[0077] In further optimized solutions, such as Figure 2 As shown, step S2 includes the following steps S21 and S22:

[0078] S21: Use a general vulnerability scoring system to evaluate the vulnerabilities of nodes in the virtual network as the probability of node infection.

[0079] First, vulnerabilities in different operating systems are assessed using the CVSS vulnerability quantification system. After comprehensively considering the vulnerability's impact on network integrity and confidentiality, a score between 0 and 10 is generated, intuitively representing the vulnerability's severity. The CVSS assessment framework consists of three core parts: a base score, a time score, and an environment score. The time score measures the risk of exploitation over a period of time. For example, when an official patch is released, the time score decreases accordingly, reflecting the reduced risk of exploitation over time. The environment score requires evaluation in a specific operating environment and is usually set by the user based on their actual usage environment. This score considers the impact of different environments on the likelihood of vulnerability exploitation. The base score reflects the severity based on the inherent characteristics of the vulnerability itself. It is unaffected by time changes and assumes the worst-case consequences that may occur in various deployment environments. Therefore, in the CVSS system, the base score is the primary basis for assessing vulnerability severity.

[0080] According to statistics from the National Information Security Vulnerability Sharing Platform, the expected base score for Mac OS X 10.10 vulnerabilities is approximately 6.6, for Windows 10 vulnerabilities it is approximately 5.4, and for Ubuntu 14 vulnerabilities it is approximately 4.4. This paper normalizes these three values, resulting in values ​​of 0.33, 0.27, and 0.22, respectively, which are used as the probability of infection for the three operating systems.

[0081] S22: The topology security level of the virtual network topology is calculated based on the node infection probability using the topology security level assessment method.

[0082] To quantitatively assess the security of virtual network topologies, this embodiment proposes an evaluation method called Topological Security Level (TSL). This method estimates the probability of a node being infected based on the operating system vulnerability security level of the nodes, thus characterizing the expected probability of infection risk for each node in the entire topology. The larger the TSL value, the higher the security coefficient of the virtual network topology, where TSL∈(0,1).

[0083] The formula for calculating the topology security level is:

[0084]

[0085] Where G is a directed graph of the virtual network topology, G i Let be a directed graph of subgraphs in the virtual network topology, e be the number of directed edges in the virtual network topology, N be the number of nodes in each subnet, and W(a,b) be the probability that node a will infect node b.

[0086] S23: Use virtual network topologies with a topology security level greater than a preset threshold as the jump pool.

[0087] The topology security level is compared with a preset threshold. If the topology security level is greater than the preset threshold, the corresponding virtual network topology is used as a jump pool, and the remaining virtual network topologies are discarded. If a preset proportion of virtual network topologies in a series have a security level less than or equal to the preset threshold, the virtual network topology is regenerated.

[0088] S3: Use the backtracking method to obtain a jump path that meets the preset constraints in the jump pool.

[0089] The core objective of hop path design is to ensure network performance stability during network path hops and reduce transient problems caused by path changes, such as network performance degradation, packet sorting disorder, and packet loss.

[0090] The node set and link set define the hop path space, with each path consisting of several forwarding nodes. During link construction, constraints on the links themselves need to be considered. This part constrains the links from three dimensions: forwarding path capacity, transmission delay, and reachability, to ensure that the generated paths meet preset standards in these methods.

[0091] Link capacity constraints are a mechanism that strictly constrains the generation process of route hop space within the available bandwidth of the communication path to prevent packet loss caused by link congestion. This mechanism ensures that the generated path has high reliability in data transmission services, thereby guaranteeing the stability and efficiency of data transmission. Therefore, link capacity constraints satisfy:

[0092] C(e i,j )≥C t

[0093] Among them, C(e) i,j ) for link e i,j Available forwarding capacity, C t This is the threshold for the link capacity required for communication.

[0094] In terms of communication quality assurance, to ensure that users' normal communication needs are efficiently met, constraints on the communication delay of the path need to be implemented, especially during the generation of the route hop space. Considering that changes in new communication delays are often closely related to the number of route hops, transmission delay constraints can be seen as an indirect control over the number of route hops, thus guaranteeing and limiting data delay during transmission. Therefore, transmission delay constraints satisfy:

[0095] D(v j )≤L max

[0096] Wherein, D(v) j To explore to the new node v j The depth of exploration at that time, L max This represents the maximum number of route hops allowed.

[0097] Reachability constraints aim to prevent forwarding loops by restricting the selection process of forwarding routing nodes. Specifically, define the i-th node v i The set of incoming edges for (1≤i≤a) is I. i The set of outgoing edges is O i 'a' represents the total number of nodes. This is to indicate the node v. i Whether a path is included in the forwarding path is determined by a Boolean variable b in this application. i When a node is selected, b i The value of b is 1; otherwise, b i The value is 0.

[0098]

[0099]

[0100]

[0101] Among them, e i The i-th edge, O S Let I be the set of outgoing edges of node S. S Let O be the set of incoming edges of node S. D Let I be the set of outgoing edges of node D. D Let D be the set of incoming edges.

[0102] While constraining the links, it's necessary to consider the degree of overlap between different links to ensure the unpredictability of the route hopping mechanism. Therefore, non-overlapping constraints are set for paths to ensure that the paths between the source and destination nodes are not repeated and that no two intermediate nodes are the same. This prevents attackers from intercepting complete communication content or blocking communication, enhancing the security of random route changes. Simultaneously, link capacity is constrained. When generating hopping paths, the available bandwidth of the communication path is constrained to prevent link congestion and ensure the reliability of data packet transmission.

[0103] The non-overlapping constraint aims to ensure the unpredictability of route hopping mechanisms, preventing them from being easily determined. To achieve this, special care must be taken to avoid the multiple occurrences of a single node (excluding source and destination nodes) when generating the route hopping space. Therefore, the degree of overlap between different paths should be fully considered and evaluated during the generation of the route hopping space, striving to reduce it to a minimum to ensure route diversity and security.

[0104]

[0105] Where L is the generated route hop space, which stores hop paths. Under the non-overlapping constraint, there are no other overlapping nodes in the hop space besides the source and destination nodes.

[0106] In a further preferred embodiment, the backtracking method is used to obtain a transition path that meets preset constraints in the transition pool, such as... Figure 3 As shown, it includes steps S31 to S33.

[0107] S31: Obtain the initial node based on the network topology information of the virtual network topology.

[0108] Obtain the network topology information of the virtual network topology, and obtain the source node S and the destination node D based on the network topology information. Then obtain the initial node, which can be either the source node or the destination node. The initial node is determined according to the condition ψ(S)≤ψ(D), that is, the node with fewer adjacent nodes between the source node and the destination node is taken as the initial node. ψ(S) is the number of adjacent nodes of the source node S, and ψ(D) is the number of adjacent nodes of the destination node D.

[0109] S32: Randomly select adjacent nodes starting from the initial node.

[0110] S33: Verify whether the path meets the preset constraints. If it does not meet the constraints, return to the previous node and select a new path. If it does meet the constraints, determine whether the node is the destination node or the source node. If it is the destination node or the source node, the generated path will be a jump path and stored in the jump path space. If it is not the destination node or the source node, the node will be used as a new extension node to continue extending the path until the destination node or the source node is reached.

[0111] Through a series of orderly steps and judgment conditions, a diverse jump path space that meets the preset conditions is gradually constructed.

[0112] Alternatively, we can first model the network based on the selected secure virtual network topology. Let the network structure be an undirected graph G = (V, E), where V is the set of forwarding nodes containing a nodes (v1, v2, ..., v...). a E is a set of links containing b links (e1, e2, ..., e...). b ), with node v i The set of adjacent nodes is ψ(v) i ), connecting node v i and v j The link is e i,j Node v i With v j The jump space between them is The node set and link set define the hop path space, and each path consists of several forwarding nodes.

[0113] S4: Use an evolutionary game model to screen the transition paths and obtain the optimal transition strategy.

[0114] In a further optimized scheme, this step, based on an evolutionary game model using Q-Learning and replicating dynamic equations, finds the optimal transition strategy through repeated attack and defense confrontations via a mechanism of continuous trial and error and exploration. Figure 4 As shown, step S4 includes steps S41 and S42:

[0115] S41: Construct an evolutionary game model using the aforementioned jump path as a defense strategy.

[0116] S411: Initialize the evolutionary game model (N,T,S,θ,Q,τ), N=(N A N D N represents the set of players in the attacking and defending game. A As the attacker, N D For the defender, T represents the stage number of the attack-defense evolutionary game based on the Q-Learning replicating dynamic equation, and S = (S A ,S D S represents the strategy space of an evolutionary game of offense and defense. A For the attack strategy set, S D For the set of defense strategies, θ = (P A ,P D P represents the belief set of an evolutionary game model of attack and defense. A For the attacker's belief set, P D For the defender's belief set, Q = (Q A Q D Q represents the set of payoff functions in an evolutionary game of attack and defense. A Let Q be the attack payoff function. D Let τ be the defense payoff function, and τ be the exploration factor, representing the degree of exploration of game information by both the attacker and the defender.

[0117] S412: Constructing an attack strategy set and its belief set P A ={p A1 ,p A2 ,…,p Am The attacker uses probability p i ∈P A Select and execute the attack strategy; the defense strategy set is the path transition space. and its belief set P D ={p D1 ,p D2 ,…,p Dm The defender uses probability q j ∈P D Select and execute the attack strategy.

[0118] S413: Combination of computational strategies Attack Profit Matrix M A and defense benefit matrix M D .

[0119] Attack Profit Matrix M A By the attacker in the combination of attack and defense strategies (S Aj ,S Dj The matrix consists of attack profit values ​​generated under the condition that the attack profit value a is generated. The matrix has (m×n) elements, where the attack profit value a is a. ijIt can be represented as:

[0120] a ij =Q A (S Ai ,S Dj ) = R A -C A

[0121] Among them, S Ai For the i-th attack strategy, S Dj For the j-th defense strategy, R A In return for the attack, C A Cost of attack.

[0122] Attack Profit Matrix M A The following can be represented as:

[0123]

[0124] Defense Benefit Matrix M D By the attacker in the combination of attack and defense strategies (S Ai ,S Dj The matrix consists of attack benefit values ​​generated under the conditions of (m×n) and has (m×n) elements, where the defense benefit value d ij It can be represented as:

[0125] d ij =Q D (S Ai ,S Dj ) = R D -C D

[0126] Among them, R D For defensive returns, C D For defense costs.

[0127] Defense Benefit Matrix M D The following is represented as:

[0128]

[0129] The strategy is a mixed-policy Nash equilibrium (p*, q*). In this case, the attacker chooses the attack target policy p*, the defender chooses the defense target policy q*, and the attack payoff is Q. A (p*, q*), where the defensive benefit is Q. D (p*,q*).

[0130] S42: Solve the evolutionary game model using the replication dynamic equation based on Q-learning to obtain the optimal jump strategy.

[0131] Construct a replication dynamic equation based on Q-learning and solve it by evolutionary equilibrium:

[0132]

[0133] Where, p i Let t be the number of game stages, representing the probability of choosing an attack strategy. attack strategy Expected returns Let p be the average attack payoff, x be the xth strategy, and p be the payoff. x Let q be the probability of choosing policy x. j Choose the probability for the defense strategy. For defensive strategies Expected returns To defend the expected return, let y be the y-th strategy, and q y Let y be the probability of choosing strategy y, n be the total number of strategies, and τ be the exploration factor.

[0134] When the strategies of both players in a game reach a mixed-strategy Nash equilibrium, the following exists:

[0135]

[0136] A hybrid Nash equilibrium can be obtained, which is an evolutionarily stable equilibrium solution. The corresponding strategy is the optimal jump strategy.

[0137] S5: Transmit data based on the optimal jump strategy.

[0138] Flow table rules are generated and distributed to network devices such as switches and routers through the control plane. Upon receiving the rules, the devices process data packets according to them, such as forwarding, dropping, or modifying packet header information. During transitions, network devices apply flow table rules in real time and dynamically adjust based on feedback from an evolutionary game model to address abnormal traffic and potential threats, ensuring network traffic flows according to predetermined strategies. This achieves device anonymity protection and enhances network security.

[0139] In summary, as Figure 5 As shown, this application first generates a virtual network topology, then performs a security assessment on the virtual network topology, then constructs a jump measurement based on the backtracking method, generates an optimal jump strategy using an evolutionary game strategy, and optimally distributes the flow table, and implements the jump strategy according to the optimal jump strategy.

[0140] Example 2:

[0141] Based on the same inventive concept, this invention also provides a data transmission system for anonymizing device identity based on route hopping, such as... Figure 6 As shown, it includes:

[0142] The generation module is used to dynamically generate a series of virtual network topologies based on real network nodes and fake nodes;

[0143] The evaluation module is used to evaluate the nodes in the virtual network topology, obtain the topology security level of the virtual network topology, and use the virtual network topology with a topology security level greater than a preset threshold as the jump pool.

[0144] The acquisition module is used to obtain a jump path that meets the preset constraints from the jump pool using a backtracking method;

[0145] The filtering module is used to filter the transition paths using an evolutionary game model to obtain the optimal transition strategy;

[0146] The transmission module is used to transmit data based on the optimal hopping strategy.

[0147] In a further preferred embodiment, the evaluation module evaluates the nodes in the virtual network topology to obtain the topology security level of the virtual network topology, and uses virtual network topologies with a topology security level greater than a preset threshold as a transition pool, including:

[0148] A general vulnerability scoring system is used to assess the vulnerabilities of nodes in a virtual network as the probability of node infection;

[0149] The topology security level of the virtual network topology is calculated based on the node infection probability using a topology security level assessment method.

[0150] Virtual network topologies with a topology security level greater than a preset threshold are used as the jump pool.

[0151] In a further preferred embodiment, the evaluation module uses virtual network topologies with a topology security level greater than a preset threshold as a jump pool, including:

[0152] The topology security level is compared with a preset threshold. If the topology security level is greater than the preset threshold, the corresponding virtual network topology is used as a jump pool, and the remaining virtual network topologies are discarded. If a preset proportion of virtual network topologies in a series have a security level less than or equal to the preset threshold, the virtual network topology is regenerated until the topology security level of the newly generated virtual network topology is greater than the preset threshold.

[0153] In a further preferred embodiment, the evaluation module calculates the topology security level of the virtual network topology based on the node infection probability using the topology security level evaluation method. The calculation formula is as follows:

[0154]

[0155] Where TSL(·) represents the topological security level of the virtual network topology, and G is the directed graph of the virtual network topology. i Let be a directed graph of subgraphs in the virtual network topology, e be the number of directed edges in the virtual network topology, N be the number of nodes in each subnet, and W(a,b) be the probability that node a will infect node b.

[0156] In a further preferred embodiment, the acquisition module uses a backtracking method to obtain a transition path that satisfies preset constraints from the transition pool, including:

[0157] The initial node is obtained based on the network topology information of the virtual network topology;

[0158] Randomly select adjacent nodes starting from the initial node;

[0159] Verify whether the path meets the preset constraints. If it does not meet the constraints, return to the previous node and select a new path. If it does meet the constraints, determine whether the node is the destination node or the source node. If it is the destination node or the source node, the generated path is a jump path and stored in the jump path space. If it is not the destination node or the source node, use the node as a new extension node to continue extending the path until the destination node or the source node is reached.

[0160] In a further preferred embodiment, the acquisition module acquires the initial node based on the network topology information of the virtual network topology, including:

[0161] Based on the network topology information of the virtual network topology, the source node and the destination node are obtained, and the node with fewer adjacent nodes between the source node and the destination node is selected as the initial node.

[0162] In a further preferred embodiment, the preset constraints in the acquisition module include: link capacity constraints, transmission delay constraints, reachability constraints, and non-overlapping constraints.

[0163] In a further preferred embodiment, the screening module uses an evolutionary game model to screen the transition paths to obtain the optimal transition strategy, including:

[0164] An evolutionary game model is constructed using the aforementioned jump path as a defense strategy;

[0165] The evolutionary game model is solved using a replication dynamic equation based on Q-learning to obtain the optimal transition strategy.

[0166] In a further preferred embodiment, the generation module further includes the following before generating a series of virtual network topologies based on real network nodes and dummy nodes:

[0167] The system determines the number of times a fake node is scanned within a preset time. If the number of scans exceeds the preset number or more than half of the fake nodes are detected, the source host of the scanned traffic is identified as a malicious host. The malicious host is then isolated and excluded from virtual network topology construction.

[0168] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A data transmission method for anonymizing device identity based on route hopping, characterized in that, include: A series of virtual network topologies are dynamically generated based on real network nodes and fake nodes; The nodes in the virtual network topology are evaluated to obtain the topology security level of the virtual network topology, and the virtual network topology with a topology security level greater than a preset threshold is used as the jump pool. The backtracking method is used to obtain the jump path that meets the preset constraints in the jump pool; An evolutionary game theory model is used to screen the transition paths to obtain the optimal transition strategy; Data is transmitted based on the optimal jump strategy described above.

2. The data transmission method for anonymizing device identity based on route hopping according to claim 1, characterized in that, The step of evaluating the nodes in the virtual network topology to obtain the topology security level of the virtual network topology, and using virtual network topologies with a topology security level greater than a preset threshold as a transition pool, includes: A general vulnerability scoring system is used to assess the vulnerabilities of nodes in a virtual network as the probability of node infection; The topology security level of the virtual network topology is calculated based on the node infection probability using a topology security level assessment method. Virtual network topologies with a topology security level greater than a preset threshold are used as the jump pool.

3. The data transmission method for anonymizing device identity based on route hopping according to claim 2, characterized in that, The step of using virtual network topologies with a topology security level greater than a preset threshold as a jump pool includes: The topology security level is compared with a preset threshold. If the topology security level is greater than the preset threshold, the corresponding virtual network topology is used as a jump pool, and the remaining virtual network topologies are discarded. If a preset proportion of virtual network topologies in a series have a security level less than or equal to the preset threshold, the virtual network topology is regenerated until the topology security level of the newly generated virtual network topology is greater than the preset threshold.

4. The data transmission method for anonymizing device identity based on route hopping according to claim 2, characterized in that, The formula for calculating the topology security level of the virtual network topology based on the node infection probability using the topology security level assessment method is as follows: Where TSL(·) represents the topological security level of the virtual network topology, and G is the directed graph of the virtual network topology. i Let be a directed graph of subgraphs in the virtual network topology, e be the number of directed edges in the virtual network topology, N be the number of nodes in each subnet, and W(a,b) be the probability that node a will infect node b.

5. The data transmission method for anonymizing device identity based on route hopping according to claim 1, characterized in that, The step of obtaining a transition path that satisfies preset constraints in the transition pool using a backtracking method includes: The initial node is obtained based on the network topology information of the virtual network topology; Randomly select adjacent nodes starting from the initial node; Verify whether the path meets the preset constraints. If it does not meet the constraints, return to the previous node and select a new path. If it does meet the constraints, determine whether the node is the destination node or the source node. If it is the destination node or the source node, the generated path is a jump path and stored in the jump path space. If it is not the destination node or the source node, use the node as a new extension node to continue extending the path until the destination node or the source node is reached.

6. The data transmission method for anonymizing device identity based on route hopping according to claim 5, characterized in that, The process of obtaining the initial node based on the network topology information of the virtual network topology includes: Based on the network topology information of the virtual network topology, the source node and the destination node are obtained, and the node with fewer adjacent nodes between the source node and the destination node is selected as the initial node.

7. The data transmission method for anonymizing device identity based on route hopping according to claim 5, characterized in that, The preset constraints include: link capacity constraints, transmission delay constraints, reachability constraints, and non-overlapping constraints.

8. The data transmission method for anonymizing device identity based on route hopping according to claim 1, characterized in that, The step of using an evolutionary game model to select the transition path and obtain the optimal transition strategy includes: An evolutionary game model is constructed using the aforementioned jump path as a defense strategy; The evolutionary game model is solved using a replication dynamic equation based on Q-learning to obtain the optimal transition strategy.

9. The data transmission method for anonymizing device identity based on route hopping according to claim 1, characterized in that, Before generating a series of virtual network topologies based on real and fake network nodes, the following also includes: The system determines the number of times a fake node is scanned within a preset time. If the number of scans exceeds the preset number or more than half of the fake nodes are detected, the source host of the scanned traffic is identified as a malicious host. The malicious host is then isolated and excluded from virtual network topology construction.

10. A data transmission system for anonymizing device identity based on route hopping, characterized in that, include: The generation module is used to dynamically generate a series of virtual network topologies based on real network nodes and fake nodes; The evaluation module is used to evaluate the nodes in the virtual network topology, obtain the topology security level of the virtual network topology, and use the virtual network topology with a topology security level greater than a preset threshold as the jump pool. The acquisition module is used to obtain a jump path that meets the preset constraints from the jump pool using a backtracking method; The filtering module is used to filter the transition paths using an evolutionary game model to obtain the optimal transition strategy; The transmission module is used to transmit data based on the optimal hopping strategy.

Images