An anti-interference method for intention-driven system game theory

Abstract

This invention proposes an intent-driven system game-theoretic anti-interference method, solving the problem of network node interference detection and defense in wireless communication networks. The implementation includes: real-time sensing and storage of data from a network data monitoring model; constructing and utilizing an interference intent recognition model to determine the type of interference attack; constructing and utilizing an anti-interference game model based on incomplete information dynamic game theory to derive corresponding anti-interference strategies for three types of interference attacks; and deploying and implementing the anti-interference strategies. This invention introduces the concept of interference intent recognition into the anti-interference scenario of wireless communication networks and constructs an anti-interference game model based on incomplete information dynamic game theory, improving the anti-interference efficiency of network nodes under closed-loop management. This invention provides more and more accurate interference intent recognition types, more timely anti-interference decisions, and higher efficiency for use in wireless communication networks.

Description

Technical Field

[0001] This invention belongs to the field of network technology and mainly relates to anti-interference of wireless communication networks. Specifically, it is an intention-driven system game-theoretic anti-interference method for anti-interference of wireless communication networks. Background Technology

[0002] Currently, due to the shared and broadcast characteristics of wireless communication environments, external jammers emit radio signals to interfere with the normal communication of network nodes. This interference can occur by reducing spectrum utilization to disrupt communication or congesting channels, rendering them inaccessible or reducing their signal-to-noise ratio. Simultaneously, with the rapid development of software-defined radio, acquiring or generating jamming devices has become extremely easy, leading to interference in wireless communication and network insecurity. Security has become a crucial attribute for establishing communication networks, and anti-interference has become a significant issue for spectrum availability security. Furthermore, with the development of radio technology and artificial intelligence, highly intelligent jamming attacks pose a significant challenge to existing anti-interference mechanisms. As communication anti-interference technology rapidly advances, external jammers are also developing and deploying new jamming techniques, leading to the continuous emergence of new types of interference, which further increases the difficulty of anti-interference development. Since the success or failure of network node anti-interference is crucial for establishing communication links with other network nodes, both civilian and military communications urgently require radio communication systems with anti-interference capabilities. There is a pressing need to develop automatic, efficient, and flexible system-level anti-interference platforms to collaboratively plan communication service deployments, improve the security of network nodes in communication networks, accurately detect interference attacks, and effectively mitigate their impact on network nodes.

[0003] As attackers employ increasingly sophisticated interference methods, there is a growing need for more comprehensive, advanced, and rapidly evolving anti-interference technologies. This increases the difficulty of developing advanced anti-interference technologies. Existing interference attack identification technologies suffer from limitations such as prior knowledge of the attack type or insufficient identification parameters, making it difficult to accurately identify specific attack types and hindering flexible, real-time anti-interference processes. This makes the entire process of interference detection, anti-interference game-theoretic decision-making, and anti-interference strategy deployment difficult to manage, resulting in low anti-interference efficiency.

[0004] Existing research on the anti-interference capabilities of wireless communication networks includes:

[0005] In their paper titled "Detecting and mitigating jamming attacks in IoT networks using self-adaptation," published at the 2022 IEEE International Conference on Autonomic Computing and Self-Organizing Systems Companion (ACSOS-C), Maxim Reynvoet et al. utilize state-of-the-art technology based on the MAPE-K reference model to detect and mitigate the damage caused by jamming attacks in IoT systems. The Monitor component monitors characteristics and uncertainties associated with jamming attacks and uses this data to update the knowledge base. The Analyze component collects data monitored by the Monitor component, uses this data to identify whether the system state is "blocked" or "normal," and flags jamming detected in specific parts of the system. If a "blocked" state is determined, the Plan component collects mitigation plans, adaptation options, and historical monitoring data to determine the optimal mitigation plan and corresponding adaptation options. Finally, the Execute component implements the selected optimal mitigation plan on the IoT system. However, a drawback is that using the MAPE-K model can only detect one type of jamming attack, resulting in limited detection of jamming attack types and untimely jamming mitigation.

[0006] In their paper "An anti-jamming stochastic game for cognitive radio networks," published in the IEEE Journal on Selected Areas in Communications, Beibei Wang et al. proposed an anti-jamming defense mechanism for secondary users facing interference attacks, based on stochastic game theory. They observed spectrum availability, channel quality, and the interferer's strategy from the perspective of a blocked channel. This method, for an established spectrum pool system, extends the process of secondary users using the remaining spectrum bands of primary users through Markov decision processes. It also analyzes the strategies of secondary users and interferers under several typical states using minimax-Q learning. However, its drawbacks include the lack of a closed-loop anti-jamming management system, resulting in low anti-jamming efficiency for secondary users facing real-time interference attacks.

[0007] The problems with existing anti-interference schemes for wireless communication networks are summarized as follows:

[0008] (1) Insufficient parameters for identifying interference attack types: Existing interference attack identification technologies have problems such as insufficient known interference attack types or insufficient identification parameters, which leads to untimely interference mitigation, difficulty in accurately identifying specific interference attack types in real time, and inability to complete flexible real-time anti-interference processes.

[0009] (2) Low anti-interference efficiency: Existing anti-interference schemes require the establishment of an attack-defense game model based on known types of interference attacks to obtain the best anti-interference strategy, which makes the entire process of anti-interference management, from interference detection to anti-interference game decision-making to anti-interference strategy deployment, difficult and prone to low anti-interference efficiency. Summary of the Invention

[0010] The purpose of this invention is to address the problems and shortcomings of existing anti-interference schemes for wireless communication networks by proposing an intention-driven system game-based anti-interference method that offers more accurate and diverse interference intent identification, more timely anti-interference decision-making, and higher anti-interference efficiency.

[0011] This invention provides an intention-driven system game-theoretic anti-interference method. Each network node in the intention-driven wireless communication network includes functional blocks for forming closed-loop information transmission: network data monitoring, interference intention identification, anti-interference game decision-making, and anti-interference strategy execution, as well as a knowledge base. Each functional block interacts dynamically with the knowledge base. The intention-driven system game-theoretic anti-interference method comprises the following steps:

[0012] Step 1: Using a network data monitoring model, data is sensed and stored in the knowledge base in real time. Interferers in the wireless communication environment launch unknown types of interference attacks on each network node. Sensors in the wireless communication network use the SNMP network management protocol and a request / response model to monitor and collect abnormal traffic information inside each network node in real time, and store it in the knowledge base in the form of a management information database.

[0013] Step 2, Constructing an Interference Intent Recognition Model: Using network data stored in the knowledge base, an interference intent recognition model is constructed based on abnormal traffic data changes. This model sets a total interference time threshold to determine the dwell time of the wireless communication signal, recorded as the matching value for identifying forwarding-type interference intent; based on the set wireless communication channel threshold, it determines the interference bandwidth of the external interferer, recorded as the matching value for identifying broadband blocking interference intent; based on the set wireless communication signal center frequency threshold, it determines the interference power of the external interferer, recorded as the matching value for identifying band blocking interference intent.

[0014] Step 3: Using the interference intent identification model, determine the type of interference attack: Analyze the data changes of abnormal traffic in the wireless communication network to obtain the interference attack type matching results; compare the changes in the actual wireless communication signal dwell time with the forwarding interference intent matching value to determine the forwarding interference attack; compare the changes in the actual wireless communication channel bandwidth with the broadband blocking interference intent matching value to determine the broadband blocking interference attack; compare the actual interference power of the external interferer with the frequency band blocking interference intent matching value to determine the frequency band blocking interference attack; Each network node analyzes the above three types of interference attacks in real time through the interference intent matching value in the interference intent identification model to obtain the classification of the three types of interference attacks;

[0015] Step 4: Construct an anti-interference model based on incomplete information dynamic game theory: Based on the classification and judgment of interference attack types, and considering that neither the attacker nor the defender can grasp all the information about the other, an anti-interference model is constructed for each network node and the external interferencer, combined with incomplete information dynamic game theory. In the anti-interference model based on incomplete information dynamic game theory, each network node and the external interferencer are rational participants; each network node has three types of interference attacks and three corresponding anti-interference strategy classifications; each network node has an initial probability judgment of the interference attack type implemented by the external interferencer; each network node updates its inference of the interference attack type of the external interferencer through Bayes' theorem; each network node selects an anti-interference strategy based on the principle of maximizing its own gains.

[0016] Step 5: Using the anti-interference model based on incomplete information dynamic game, corresponding anti-interference strategies are obtained for the three types of interference attacks: Based on the constructed anti-interference model based on incomplete information dynamic game, the optimal payoff for each network node and the external interferer regarding interference is calculated, thereby obtaining an inference of the interference attack type. The posterior probability of each network node's inference of the interference attack type is calculated using Bayes' rule based on initial belief inference, so as to correct the inference of each network node's inference of the interference attack type, thus determining the interference attack type for each network node; For the interference attack type determined after the posterior probability correction, each network node solves the refined Bayesian equilibrium solution of the anti-interference model for the three interference attack types, thereby obtaining the corresponding anti-interference strategy.

[0017] Step 6: Deploy anti-interference strategies and terminate the anti-interference process of network nodes: The executor of each network node in the wireless communication network executes anti-interference strategies. For forwarding-type interference attacks, multiple signal copies are transmitted through multiple channels in time, frequency, or space, and the original communication signal is restored at the receiving end to achieve signal diversity anti-interference. For broadband blocking interference attacks, spread spectrum anti-interference is achieved by concealing communication signals and reducing narrowband interference effects. For frequency band blocking interference attacks, frequency hopping anti-interference is achieved by using the size of the frequency point and the speed of frequency change to avoid interference. Each network node deploys and executes the corresponding anti-interference strategies in real time for the three types of interference attacks from external interferers, completing the intention-driven system game anti-interference.

[0018] This invention solves the technical problem in interference attack identification technology where the type of interference attack is known in advance or there are too few identification parameters, making it difficult to accurately identify the specific type of interference attack; it also solves the technical problem that makes the entire process of interference attack detection, anti-interference strategy selection, and anti-interference strategy deployment difficult to manage, resulting in low anti-interference efficiency.

[0019] Compared with the prior art, the present invention has the following advantages:

[0020] More accurate interference intent identification: The network data monitoring model adopted in this invention, based on the SNMP network management protocol and the request / response pattern, monitors and collects a large amount of information on wireless communication network traffic and hardware device status, providing a large amount of abnormal network traffic information. On this basis, this invention uses the total interference time threshold as the identification of forwarding interference intent, the wireless communication channel threshold as the identification of broadband blocking interference intent, and the wireless communication signal center frequency threshold as the identification of frequency band blocking interference intent, thus constructing an interference intent identification model and improving the accuracy and reliability of the interference intent identification process.

[0021] More timely anti-interference decision-making: This invention uses Bayes' rule to calculate the posterior probability of each network node inferring the type of interference attack. By introducing a refined Bayesian equilibrium solution, an anti-interference game model based on incomplete information dynamic game is constructed. Each network node adaptively selects the deployment of the best anti-interference strategy for each of the three types of interference attacks in a short time, completing the anti-interference process of each network node, perceiving changes in the type of interference attack in real time, and bridging the gap between the correct adaptation between the type of interference attack and the anti-interference strategy.

[0022] Higher anti-interference efficiency: Based on the closed-loop information transmission functional block of network data monitoring - interference intent identification - anti-interference game decision-making - anti-interference strategy execution, this invention constructs an interference intent identification model, which accurately identifies three types of interference attacks implemented by external interferers. For the three types of interference attacks, an anti-interference game model based on incomplete information dynamic game is constructed to select the corresponding optimal anti-interference strategy for each network node. The anti-interference strategy is executed by the executor of each network node in the wireless communication network, which reduces the anti-interference configuration time inside each network node, enhances the interference suppression capability of each network node, and improves the anti-interference efficiency of each network node. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of the system game anti-interference method intended to drive the present invention;

[0025] Figure 2 This is a flowchart illustrating the specific implementation of the anti-interference method for game theory driven by the present invention.

[0026] Figure 3 This is a structural block diagram of the internal structure of the network node of the present invention;

[0027] Figure 4 This is a flowchart illustrating the implementation process of the interference intent recognition model of the present invention.

[0028] Figure 5 This is a flowchart illustrating the implementation process of selecting the optimal anti-interference strategy based on the anti-interference model of dynamic game theory with incomplete information, as described in this invention. Detailed Implementation

[0029] The present invention will now be described in detail with reference to the accompanying drawings and embodiments:

[0030] Example 1:

[0031] Existing anti-jamming technologies based on stochastic game theory are a type of anti-jamming defense mechanism for users facing interference attacks. They select the optimal anti-jamming strategy for the user by observing spectrum availability, channel quality, and the jammer's strategy based on the state of the blocked channel. However, these technologies require the user to already know the jammer's strategy. Even with prior knowledge of the jammer's strategy, anti-jamming efficiency remains low when facing real-time interference attacks, as users cannot make timely anti-jamming decisions.

[0032] To address the technical problems existing in the prior art, this invention proposes an anti-interference method for intention-driven system game theory, based on analysis and research. (See [link to relevant documentation]). Figure 3 , Figure 3 This is a structural block diagram of the network node internals of the present invention. Each network node in the wireless communication network driven by the present invention includes functional blocks for forming closed-loop information transmission: network data monitoring, interference intent identification, anti-interference game decision-making, and anti-interference strategy execution, as well as a knowledge base. Each functional block interacts dynamically with the knowledge base. See also... Figure 1 , Figure 1 This is a flowchart of the intention-driven system game anti-interference method of the present invention. The intention-driven system game anti-interference method of the present invention includes the following steps:

[0033] Step 1: Employing a network data monitoring model to sense and store data in a knowledge base in real time. Externally existing interferers (jammers) launch unknown types of interference attacks against each network node within the wireless communication network, disrupting its normal operation. This invention considers each network node sending a request to a sensor in the wireless communication network to collect network data. The sensor uses the SNMP network management protocol and a request / response model to monitor and collect abnormal traffic information within each network node in real time. This network data is then transmitted back to each network node in the form of a management information base and stored in the knowledge base to update the sensor environment and topology model located therein. When abnormal network data is generated, the sensor does not need to wait for a response from each network node; it will proactively send a Trap message representing an emergency to that network node, promptly reporting the abnormal network status. This invention, by employing a network data monitoring model based on the SNMP network management protocol and a request / response model, monitors and collects a large amount of information about wireless communication network traffic and hardware device status, providing network data support for the interference intent identification process. Based on the closed-loop information transmission functional block of network data monitoring - interference intent identification - anti-interference game decision-making - anti-interference strategy execution, this invention can sense and store data from the network data monitoring model in real time.

[0034] Step 2, Constructing an Interference Intent Recognition Model: This invention constructs an interference intent recognition model by calling network data stored in a knowledge base and based on changes in abnormal traffic data. The model matches interference intent information parameters in the knowledge base to transform interference intent into interference attack types. For three different interference attack types, the interference intent recognition model constructed in this invention determines the dwell time of the wireless communication signal by setting a threshold for the total interference time of the external interferer, and records this as the identification value for forwarding interference intent. Secondly, it determines the interference bandwidth of the external interferer by setting a threshold for the wireless communication channel of the wireless communication network, and records this as the identification value for broadband blocking interference intent. Finally, it determines the interference power of the external interferer by setting a threshold for the center frequency of the wireless communication signal, and records this as the identification value for bandgap blocking interference intent.

[0035] Step 3: Using the interference intent identification model to determine the type of interference attack: This invention calls upon and analyzes the data changes of abnormal traffic in the wireless communication network stored in the knowledge base, and obtains the interference attack type matching result based on the constructed interference intent identification model. First, this invention compares the changes in the actual dwell time of the wireless communication signal in the wireless communication network with the forwarding interference intent matching value set in the interference intent identification model to determine whether the interference attack type is a forwarding interference attack. Second, this invention compares the changes in the actual wireless communication channel bandwidth in the wireless communication network with the broadband blocking interference intent matching value set in the interference intent identification model to determine whether the interference attack type is a broadband blocking interference attack. Finally, this invention compares the actual interference power implemented by the external interferer with the frequency band blocking interference intent matching value set in the interference intent identification model to determine whether the interference attack type is a frequency band blocking interference attack. Each network node analyzes the above three types of interference attack in real time through the three types of interference intent matching values ​​in the interference intent identification model to obtain the classification of the three types of interference attack. Based on the availability of network resources, the conflict of interference attack types, and the correctness of the interference attack type judgment, the interference attack type is verified to prove whether the attack type is complete and correct.

[0036] Step 4: Constructing an anti-interference model based on incomplete information dynamic game theory: Based on the classification judgment of interference attack types obtained from the interference intent identification model, and considering the characteristic that neither the attacker nor the defender can grasp all the information of the other, as well as the sequential and dynamic interaction of actions between the attacker and the defender, this invention constructs an anti-interference game model between each network node and the external interferer, combined with the characteristics of incomplete information dynamic game theory. The anti-interference game model based on incomplete information dynamic game theory constructed in this invention mainly includes five parts: each network node and the external interferer are set as rational participants; the external interferer can launch three types of interference attacks; each network node has three corresponding anti-interference strategy classifications; each network node has an initial probability judgment of the interference attack type implemented by the external interferer; each network node updates its inference of the interference attack type of the external interferer through Bayesian rules; and each network node selects an anti-interference strategy based on the principle of maximizing its own payoff, according to the properties of the refined Bayesian equilibrium solution. The anti-interference game model based on incomplete information dynamic game theory constructed in this invention, by quantifying and representing the dynamic interaction relationship between network nodes and external interferers, can flexibly and in real-time select the best anti-interference strategy.

[0037] Step 5: Using an anti-interference model based on incomplete information dynamic game theory, corresponding anti-interference strategies are obtained for the three types of interference attacks. This invention, based on the constructed anti-interference model based on incomplete information dynamic game theory, characterizes the offensive and defensive game process between network nodes and external interferers, obtains the payoff matrix for both sides, and considers the loss cost, the risk of network node anti-interference strategy selection, and the costs of interference and anti-interference as components of the objective functions of both sides. The expected payoff for each network node's anti-interference and the expected payoff for the external interferer's creation and generation of interference are calculated, thus obtaining an inference about the type of interference attack. The posterior probability of each network node's inference about the interference attack type is calculated using Bayes' theorem based on initial belief inference, to correct the inference of each network node's inference about the interference attack type, thus determining the interference attack type for each network node. For the interference attack types determined after posterior probability correction, each network node solves the refined Bayesian equilibrium solution of the anti-interference model for each of the three interference attack types to obtain the corresponding anti-interference strategy. This invention completes the anti-interference process of each network node by adaptively and quickly selecting the deployment of the best anti-interference strategy, and perceives changes in the type of interference attack in real time, bridging the gap between the type of interference attack and the anti-interference strategy for proper adaptation.

[0038] Step 6: Deploy anti-interference strategies and terminate the anti-interference process of network nodes: The actuator of each network node in the wireless communication network is directly connected to the internal infrastructure of the network node, deploying and executing the optimal anti-interference strategy selected for each network node. For three different types of interference attacks, this invention addresses forwarding interference attacks by transmitting multiple signal copies through multiple channels in time, frequency, or space, recovering the original communication signal at the receiving end of each network node, thus achieving signal diversity anti-interference for each network node. Secondly, for broadband congestion interference attacks, this invention addresses spreading spectrum anti-interference for each network node by concealing the communication signals of the wireless communication network and reducing narrowband interference effects. Finally, for frequency band congestion interference attacks, this invention addresses frequency hopping anti-interference for each network node by using the magnitude of the wireless communication center signal frequency and the rate of frequency change to evade interference. Each network node deploys and executes the corresponding anti-interference strategy in real time in the form of executable operations for the three types of interference attacks from external interferers, saves relevant information to the knowledge base, terminates the network node's anti-interference process, and completes the intention-driven system game-theoretic anti-interference of this invention.

[0039] This invention addresses the technical problems of low anti-interference efficiency and delayed anti-interference decisions in existing anti-interference technologies. It proposes an intent-driven system game-based anti-interference method. Based on a closed-loop information transmission functional block encompassing network data monitoring, interference intent identification, anti-interference game-based decision-making, and anti-interference strategy execution, this method employs a network data monitoring model to monitor and collect a large amount of information about wireless communication network traffic and hardware device status. Furthermore, based on the monitored network data, this invention sets interference intent matching values ​​for three types of interference attacks, thereby constructing an interference intent identification model that makes the interference intent identification process more accurate and reliable. To improve the anti-interference decision-making efficiency of each network node, this invention constructs an anti-interference game-based model for both attackers and defenders based on incomplete information dynamic game theory. This model adaptively and quickly selects the optimal anti-interference strategy to complete the anti-interference process for each network node, real-time sensing of changes in interference attack types, enhancing the interference suppression capability of each network node, and improving the anti-interference efficiency of each network node.

[0040] Example 2:

[0041] An anti-interference method for game theory in an intent-driven system is the same as in Embodiment 1. The abnormal traffic information monitored by network nodes in step 1 of this invention specifically includes:

[0042] This invention, based on each network node sending a request to sensors in the wireless communication network to collect network data, utilizes the SNMP network management protocol and a request / response pattern to monitor and collect abnormal traffic information within each network node in real time. It mainly includes two aspects:

[0043] 1.1 Network Traffic Information: Monitor and collect detailed information on network traffic in network link status, including: throughput, available bandwidth, packet loss rate, frequency, and GPS location.

[0044] 1.2 Hardware device status information: Monitor and collect the status information of network node hardware devices, including: CPU load, memory availability and load, and hard disk storage.

[0045] This invention stores abnormal traffic information of wireless communication networks monitored and collected by a network data monitoring model in real time into a knowledge base. The stored abnormal traffic information will provide network data support for the constructed interference intent identification model.

[0046] Example 3:

[0047] An anti-interference method for game theory in an intent-driven system is similar to that in Examples 1-2. Step 3, which utilizes an interference intent recognition model to determine the type of interference attack, specifically includes:

[0048] This invention constructs a interference intent identification model by calling network data stored in a knowledge base and based on changes in abnormal traffic data. See [link to relevant documentation]. Figure 4 , Figure 4 This is a flowchart illustrating the implementation process of the interference intent identification model of this invention. This invention accurately identifies forwarding interference intent, broadband blocking interference intent, and band blocking interference intent, determining the specific type of interference attack implemented by the external attacker. The interference attack types identified by the interference intent identification model of this invention include the following:

[0049] 3.1 Relay-based interference intent identification: This invention is based on the actual communication signal dwell time ηT n The system identifies internal interference intentions by analyzing changes in traffic data, determining whether external interference has initiated a forwarding-based interference attack. This is done by monitoring changes in the actual communication signal dwell time within the traffic data: if the total actual interference time exceeds the actual communication signal dwell time, forwarding-based interference is considered not to have occurred. If the total actual interference time is less than or equal to the actual communication signal dwell time, forwarding-based interference is considered to have actually occurred. The specific details of the forwarding-based interference intent identification method are as follows:

[0050] 3.1.1 If the actual total time of interference from external interference meets the following condition:

[0051] D n / c+T j >ηT n

[0052] Where c is the speed of light; D n T represents the distance between a network node and an external interferer. jThe expected interference time for the external interferer; η is the dwell time coefficient for successful interference, taking values ​​of [0,1]; T n This refers to the dwell time of the monitored communication signal. When the total actual interference time from an external interferer exceeds the actual dwell time of the communication signal, the external interferer has not engaged in repeater-type interference.

[0053] 3.1.2 If the actual total time of interference from external interference meets the following condition:

[0054] D n / c+T j ≤ηT n

[0055] When the total actual interference time of the external interferer is less than or equal to the actual dwell time of the communication signal, the type of interference attack implemented by the external interferer can be accurately determined to be a forwarding interference attack.

[0056] 3.2 Broadband Congestion Interference Intent Identification: This invention is based on the actual interference bandwidth B of the external interferer. j The changes in bandwidth are used to identify the interference intent within the system and determine whether an external interferer has launched a broadband blocking attack. This is determined by observing changes in the actual interference bandwidth of the external interferer in the traffic information: when the actual interference bandwidth of the external interferer is less than or equal to the actual channel bandwidth of the network node, it is determined that no broadband blocking attack has occurred. When the actual interference bandwidth of the external interferer is greater than the actual channel bandwidth of the network node, it is determined that broadband blocking attack has actually occurred. The specific details of the broadband blocking attack intent identification method are as follows:

[0057] 3.2.1 If the actual interference bandwidth of the external interferer meets the condition:

[0058]

[0059] Among them, B j B represents the actual interference bandwidth of the external interferator. n f represents the actual bandwidth of the network node. n f is the center frequency of the communication band; j This is the center frequency of the interference band. When the actual interference bandwidth of the external interferer is less than or equal to the bandwidth of the actual network node, the external interferer has not caused broadband blocking interference.

[0060] 3.2.2 If the actual interference bandwidth of the external interferor meets the condition:

[0061] B j >5B n ,f n ∈[f j -B j / 2,f j +B j / 2]

[0062] When the actual interference bandwidth of an external interferer is greater than the actual bandwidth of a network node, the type of interference attack implemented by the external interferer can be accurately determined to be a broadband blocking interference attack.

[0063] 3.3 Bandblocking Interference Intent Identification: This invention is based on the actual interference signal power p of the external interferer. j The change in (f) identifies the interference intent within the system at this time, determining whether an external interferer has launched a band-blocking interference attack. This is determined by the change in the actual interference signal power of the external interferer in the traffic information: when the actual interference signal power of the external interferer is less than or equal to the actual communication signal power of the network node, it is determined that no band-blocking interference has occurred. When the actual interference signal power of the external interferer is greater than the actual communication signal power of the network node, it is determined that band-blocking interference has actually occurred. The specific details of the band-blocking interference intent identification method are as follows:

[0064] 3.3.1 If the actual interference signal power of the external interfering party meets the following condition:

[0065] p j (f)>p n (f),f n ∈[f j -B j / 2,f j +B j / 2]

[0066] Among them, B j The actual interference bandwidth of the external interferator; p n (f) represents the actual communication signal power of the network node; f n f is the center frequency of the communication band; j The center frequency of the interference band is used; when the actual interference signal power of the external interferer is greater than the actual communication signal power of the network node, the type of interference attack implemented by the external interferer is accurately determined to be a band blocking interference attack.

[0067] 3.3.2 If the actual interference signal power of the external interfering party meets the following condition:

[0068] p j (f)≤p n (f),f n ∈[-∞, f j -B j / 2)∪(f j +B j / 2,+∞]

[0069] When the actual interference signal power of the external interferer is less than or equal to the actual communication signal power of the network node, the external interferer has not implemented band blocking interference.

[0070] This invention constructs an interference intent identification model by analyzing changes in communication signal dwell time, actual interference bandwidth, and actual interference signal power of external interferers, based on network data stored in a knowledge base. For the characteristics of three types of interference attacks, this invention designs forwarding-based interference intent matching values, broadband blocking interference intent matching values, and frequency band blocking interference intent matching values, respectively. These values ​​are compared with actual network data from the wireless communication network to obtain the three types of interference attacks implemented by external interferers. This invention not only constructs the interference intent identification model but also provides mathematical expressions for different interference intent identification methods, enabling accurate identification of the three types of interference attacks in wireless communication networks. By constructing the interference intent identification model, this invention can promptly identify the type of interference attack from external interferers for each network node, improving the accuracy and reliability of the interference intent identification process.

[0071] Example 4:

[0072] An intention-driven system game anti-interference method is similar to that in Examples 1-3. The anti-interference model constructed in step 4 of this invention, based on incomplete information dynamic game theory, uses incomplete information dynamic game theory to characterize the offensive and defensive game process between the two sides, mathematically represented as a quintuple form. They are represented as follows:

[0073] 4.1 Each network node and external interferer are rational participants: In an anti-interference model based on dynamic game theory with incomplete information In this context, rational participants are represented as in, The number of network nodes represents the defensive side, which is expressed as a set. n represents the nth network node, and N represents the total number of network nodes; This represents the number of attackers, i.e., external interferers, and its representation is in set form. j represents the j-th external disruptor, and J represents the total number of external disruptors.

[0074] 4.2 Each network node has three types of interference attacks and three corresponding anti-interference strategy classifications: in the anti-interference model based on incomplete information dynamic game. In Chinese, a categorical set is represented as Q n Anti-interference strategy deployed for the nth network node Let T be the set of interference attack types for the j-th external interferer, where each external interferer has T interference attack types that can be implemented. This represents the t-th type of interference attack launched by the j-th external interferer against the n-th network node.

[0075] 4.3 Each network node has an initial probability judgment of the type of external interferer: the initial judgment set of the network node is as follows: Then, the initial judgment of the t-th type of interference attack implemented by the n-th network node against the j-th external interference agent. Represented as:

[0076]

[0077] 4.4 Each network node updates its inference of the interference attack type from external interferers using Bayes' theorem: After receiving the interference attack type identified by the interference intent recognition model, the nth network node updates its inference of the interference attack type from the jth external interferer using Bayes' theorem. The posterior probability set of the inference of the t-th interference attack type implemented by the nth network node against the j-th external interferer is represented as follows: Wherein, the posterior probability of the t-th type of interference attack implemented by the n-th network node against the j-th external interferer. for:

[0078]

[0079] 4.5 Each network node selects an anti-interference strategy based on its own profit maximization principle: a set of utility functions. for U n and U j Let $\frac{ ... The damage caused to the nth network node by the t-th type of interference attack implemented by the j-th external interferer is denoted as:

[0080]

[0081] Among them, C n The importance of the nth network node is represented by a value between 1 and 10. The lethality of the t-th type of interference attack from the j-th external interferer to the n-th network node is represented by a value between 0 and 1. The risk D represents the network node's anti-interference strategy selection. n This indicates that the nth network node will target the wth type of interference attack launched by the jth external interferer. Mistakenly identified as the t-th type of interference attack The risk and return at that time are expressed as:

[0082]

[0083] Where, λ n The degree of interference to the nth network node is represented by a value between 0 and 1; p(Q j,w |Q j,t () represents the w-th type of interference attack launched by the j-th external interferer by the n-th network node. Mistakenly identified as the t-th type of interference attack The probability of damage at time. The benefit of the t-th type of interference attack against the j-th external interferer is:

[0084]

[0085] in, A represents the interference cost incurred by the j-th external interference agent in launching the t-th type of interference attack against the n-th network node; n α1 represents the interference resistance cost of the nth network node; α2 represents the loss cost. The trade-off factor; α2 represents the interference resistance cost A of the nth network node. n The trade-off factor; α3 represents the interference cost of external interferers. The trade-off factors. Similarly, the payoff for the nth network node is:

[0086]

[0087] Where β1 represents the cost of loss. The trade-off factor; β2 represents the interference cost of external interferers. The trade-off factor; β3 represents the interference resistance cost A of the nth network node. n The trade-off factor; β4 represents the risk D of the anti-interference strategy selection of the nth network node. n The trade-off factors.

[0088] This invention, based on the classification judgment of interference attack types obtained from the interference intent recognition model, and considering the characteristic that neither the attacker nor the defender can possess all the information about the other, as well as the sequential and dynamic interaction of actions between the two sides, and combining the characteristics of incomplete information dynamic game theory, constructs an anti-interference game model between each network node and the external interferencer. In this anti-interference game model based on incomplete information dynamic game theory, the optimal anti-interference strategy is flexibly and in real-time selected during the anti-interference process by quantifying and representing the dynamic interaction relationship between the network node and the external interferencer. This invention not only constructs an anti-interference game model based on incomplete information dynamic game theory but also provides a mathematical expression for the anti-interference game model, enabling each network node to accurately select the corresponding anti-interference strategy for three types of interference attacks.

[0089] Example 5:

[0090] An intention-driven system game-theoretic anti-interference method is similar to embodiments 1-4. Specifically, step 5 of this invention describes obtaining corresponding anti-interference strategies for the three types of interference attacks. This invention utilizes a constructed anti-interference game model based on incomplete information dynamic game theory to obtain corresponding anti-interference strategies for the three types of interference attacks. See also... Figure 5 , Figure 5 This is a flowchart illustrating the implementation process of the anti-interference strategy selection based on the incomplete information dynamic game theory model of this invention. This invention targets three types of interference attacks and selects a corresponding anti-interference strategy for each network node using the anti-interference game theory model. The selection of the optimal anti-interference strategy for each network node based on the incomplete information dynamic game theory model of this invention includes the following steps:

[0091] 5.1 Network nodes use Bayesian rules to infer anti-interference strategies corresponding to three types of interference attacks. This invention infers anti-interference strategies for each network node against three types of interference attacks based on Bayesian rules derived from initial belief inference.

[0092]

[0093] In the formula For the nth network node, under Bayes' theorem, the tth type of interference attack is implemented against the jth external interferor. Posterior probability inference; This represents the gain of the nth network node. The anti-interference strategies for each network node against the three types of interference attacks inferred here by this invention are preliminary inferences.

[0094] 5.2 Solving for Refined Bayesian Equilibrium Solutions in Network Nodes Obtain the best anti-interference strategy This invention addresses three types of interference attacks from external interferers. The nth network node solves the refined Bayesian equilibrium solution. Obtain the optimal anti-interference strategy for the three types of interference attacks. Choose to make network nodes benefit Maximize the best anti-interference strategy The anti-interference strategy obtained in this invention is the optimal anti-interference strategy.

[0095] 5.3 Verify the optimal anti-interference strategy posterior probability Does a conflict exist? If the posterior probability of the nth network node implementing the tth type of interference against the jth external interferer is... With the best anti-interference strategy posterior probability If there is no conflict between them, then the refined Bayesian equilibrium solution obtained by this invention is... The best anti-interference strategy It is the best choice for each network node, meaning that for each of the three types of interference attacks, each network node can choose the best anti-interference strategy. To resist interference. Therefore, the present invention will This serves as the optimal anti-interference strategy for each network node.

[0096] This invention constructs an anti-interference game model based on incomplete information dynamic game theory, characterizing the offensive and defensive interaction process between each network node and an external interferer. Each network node calculates its posterior probability of inferring the type of interference attack using Bayes' rule based on initial belief inference, solves the refined Bayesian equilibrium of the anti-interference game model, and obtains the corresponding anti-interference strategy. By introducing the refined Bayesian equilibrium solution, this invention enables each network node to perceive changes in the type of interference attack in real time, adaptively and quickly selecting the optimal anti-interference strategy for each of the three types of interference attacks. This enhances the interference suppression capability and anti-interference efficiency of each network node, bridging the gap between the correct adaptation of interference attack types and anti-interference strategies.

[0097] A more detailed example is given below to further illustrate the invention.

[0098] Example 6:

[0099] An anti-interference method for intention-driven system game theory is the same as in Examples 1-5, see [link / reference]. Figure 3 , Figure 3 This is a structural block diagram of the network node internals of the present invention. Each network node in the intent-driven wireless communication network of the present invention includes functional blocks for forming closed-loop information transmission: network data monitoring, interference intent identification, anti-interference game decision-making, and anti-interference strategy execution, as well as a knowledge base. Each functional block interacts dynamically with the knowledge base. See [link to relevant documentation]. Figure 2 , Figure 2 This is a flowchart illustrating the specific implementation of the intention-driven system game anti-interference method of the present invention. The intention-driven system game anti-interference method of the present invention includes the following steps:

[0100] Step 1: Employing a network data monitoring model, real-time sensing and storage of data in a knowledge base: External interference (jamming machines) can launch unknown types of interference attacks against each network node within the wireless communication network, disrupting their normal operation. This invention considers each network node sending a request to a sensor in the wireless communication network to collect network data. The sensor utilizes the SNMP network management protocol and a request / response model to monitor and collect abnormal traffic information within each network node in real time. This network data is then transmitted back to each network node in the form of a management information base and stored in the knowledge base to update the sensor environment and topology model located therein. When abnormal network data is generated, the sensor does not need to wait for a response from each network node; it will proactively send a Trap message representing an emergency to that network node, promptly reporting the abnormal network status.

[0101] Step 2, Constructing the Interference Intent Recognition Model: This invention constructs an interference intent recognition model by calling network data stored in a knowledge base and analyzing abnormal traffic data changes. The model matches interference intent information parameters in the knowledge base to transform interference intent into interference attack types. For three different interference attack types, the interference intent recognition model constructed in this invention determines the dwell time of the wireless communication signal by setting a threshold for the total interference time of the external interferer, recording this as the identification value for forwarding interference intent. Secondly, it determines the interference bandwidth of the external interferer by setting a threshold for the wireless communication channel of the wireless communication network, recording this as the identification value for broadband blocking interference intent. Finally, it determines the interference power of the external interferer by setting a threshold for the center frequency of the wireless communication signal, recording this as the identification value for bandgap blocking interference intent.

[0102] Step 3: Determine the type of interference attack using the interference intent identification model: This invention analyzes and calls upon data changes in abnormal traffic of wireless communication networks stored in the knowledge base. Based on the constructed interference intent identification model, it obtains the interference attack type matching result. First, this invention compares the changes in the actual dwell time of wireless communication signals in the wireless communication network with the forwarding interference intent matching value set in the interference intent identification model to determine whether the interference attack type is a forwarding interference attack. Second, this invention compares the changes in the actual bandwidth of the wireless communication channel in the wireless communication network with the broadband blocking interference intent matching value set in the interference intent identification model to determine whether the interference attack type is a broadband blocking interference attack. Finally, this invention compares the actual interference power implemented by the external interferer with the frequency band blocking interference intent matching value set in the interference intent identification model to determine whether the interference attack type is a frequency band blocking interference attack. Each network node analyzes the above three types of interference attack in real time using the three types of interference intent matching values ​​in the interference intent identification model to obtain a classification of the three types of interference attack. Based on the availability of network resources, the conflict of interference attack types, and the correctness of the interference attack type judgment, the interference attack type is verified to prove whether the attack type is complete and correct.

[0103] Step 4: Constructing an anti-interference model based on incomplete information dynamic game theory: Based on the classification judgment of interference attack types obtained from the interference intent identification model, and considering the characteristic that neither the attacker nor the defender can grasp all the information of the other, as well as the sequential and dynamic interaction of actions between the attacker and the defender, this invention constructs an anti-interference game model between each network node and the external interferer, combined with the characteristics of incomplete information dynamic game theory. The anti-interference game model based on incomplete information dynamic game theory constructed in this invention mainly includes five parts: each network node and the external interferer are set as rational participants; the external interferer can launch three types of interference attacks; each network node has three corresponding anti-interference strategy classifications; each network node has an initial probability judgment of the interference attack type implemented by the external interferer; each network node updates its inference of the interference attack type of the external interferer through Bayesian rules; and each network node selects an anti-interference strategy based on the property of the refined Bayesian equilibrium solution and the principle of maximizing its own profit.

[0104] Step 5: Using an anti-interference model based on incomplete information dynamic game theory, corresponding anti-interference strategies are obtained for the three types of interference attacks. This invention, based on the constructed anti-interference model based on incomplete information dynamic game theory, characterizes the offensive and defensive game process between network nodes and external interferers, obtains the payoff matrix for both sides, and considers the loss cost, the risk of network node anti-interference strategy selection, and the costs of interference and anti-interference as components of the objective functions of both sides. The expected payoff for each network node's anti-interference and the expected payoff for the external interferer's creation and generation of interference are calculated, thus obtaining an inference about the type of interference attack. The posterior probability of each network node's inference about the interference attack type is calculated using Bayes' theorem based on initial belief inference, to correct the inference of each network node's inference about the interference attack type, thus determining the interference attack type for each network node. For the interference attack types determined after posterior probability correction, each network node solves the refined Bayesian equilibrium solution of the anti-interference model for each of the three interference attack types to obtain the corresponding anti-interference strategy.

[0105] Step 6: Deploy anti-interference strategies and terminate the anti-interference process of network nodes: The actuator of each network node in the wireless communication network is directly connected to the internal infrastructure of the network node, deploying and executing the optimal anti-interference strategy selected for each network node. For three different types of interference attacks, this invention addresses forwarding interference attacks by transmitting multiple signal copies through multiple channels in time, frequency, or space, recovering the original communication signal at the receiving end of each network node, thus achieving signal diversity anti-interference for each network node. Secondly, for broadband congestion interference attacks, this invention addresses spreading spectrum anti-interference for each network node by concealing the communication signals of the wireless communication network and reducing narrowband interference effects. Finally, for frequency band congestion interference attacks, this invention addresses frequency hopping anti-interference for each network node by using the magnitude of the wireless communication center signal frequency and the rate of frequency change to evade interference. Each network node deploys and executes the corresponding anti-interference strategy in real time in the form of executable operations for the three types of interference attacks from external interferers, saves relevant information to the knowledge base, terminates the network node's anti-interference process, and completes the intention-driven system game-theoretic anti-interference of this invention.

[0106] Each network node uses the SNMP network management protocol and a request / response model to send requests to sensors in the wireless communication network to collect network data. It monitors and collects abnormal traffic information within each network node in real time, including network traffic information related to network link status such as throughput, available bandwidth, packet loss rate, frequency, and GPS location, as well as network node hardware status information such as CPU load, memory availability and load, and hard disk storage. The sensors transmit network data back to each network node in the form of a management information base, storing it in the knowledge base to update the sensor environment and topology model located therein. When abnormal network data is generated, the sensor does not need to wait for a response from the network status monitoring module; it will proactively send a Trap message representing an emergency to the network node, promptly reporting the network status anomaly.

[0107] By calling and analyzing data changes in abnormal traffic of wireless communication networks stored in a knowledge base, and matching the parameters of currently generated interference intent information with the interference intent recognition model existing in the knowledge base, this invention transforms interference intent into an interference attack type. First, this invention compares the changes in the actual dwell time of wireless communication signals in the wireless communication network with the forwarding interference intent matching value set in the interference intent recognition model to determine whether the interference attack type is a forwarding interference attack. Second, this invention compares the changes in the actual bandwidth of the wireless communication channel in the wireless communication network with the broadband blocking interference intent matching value set in the interference intent recognition model to determine whether the interference attack type is a broadband blocking interference attack. Finally, this invention compares the actual interference power implemented by the external interferer with the frequency band blocking interference intent matching value set in the interference intent recognition model to determine whether the interference attack type is a frequency band blocking interference attack. The interference intent recognition model constructed by this invention is described in [link to relevant documentation]. Figure 4 , Figure 4 This is a flowchart illustrating the implementation process of the interference intent recognition model of the present invention. The interference attack types identified by the interference intent recognition model of the present invention include the following:

[0108] 6.1 Relay-based interference intent identification: This invention is based on the actual communication signal dwell time ηT n The system identifies internal interference intentions by analyzing changes in traffic data, determining whether external interference has initiated a forwarding-based interference attack. This is done by monitoring changes in the actual communication signal dwell time within the traffic data: if the total actual interference time exceeds the actual communication signal dwell time, forwarding-based interference is considered not to have occurred. If the total actual interference time is less than or equal to the actual communication signal dwell time, forwarding-based interference is considered to have actually occurred. The specific details of the forwarding-based interference intent identification method are as follows:

[0109] 6.1.1 If the actual total interference time D of the external interference is...n / c+T j Conditions met:

[0110] D n / c+T j >ηT n

[0111] Where c is the speed of light; D n T represents the distance between a network node and an external interferer. j The expected interference time for the external interferer; η is the dwell time coefficient for successful interference, taking values ​​of [0,1]; T n The dwell time of the monitored communication signal. When the actual total interference time D from the external interferer... n / c+T j When the actual communication signal dwell time is greater than the actual dwell time, external interference does not cause forwarding interference.

[0112] 6.1.2 If the actual total interference time D from external interference is... n / c+T j Conditions met:

[0113] D n / c+T j ≤ηT n

[0114] When the actual total time of external interference is D n / c+T j When the actual communication signal dwell time is less than or equal to the actual dwell time, the type of interference attack carried out by the external interferer can be accurately determined to be a forwarding interference attack.

[0115] 6.2 Broadband Congestion Interference Intent Identification: This invention is based on the actual interference bandwidth B of the external interferer. j The changes in bandwidth are used to identify the interference intent within the system and determine whether an external interferer has launched a broadband blocking attack. This is determined by observing changes in the actual interference bandwidth of the external interferer in the traffic information: when the actual interference bandwidth of the external interferer is less than or equal to the actual channel bandwidth of the network node, it is determined that no broadband blocking attack has occurred. When the actual interference bandwidth of the external interferer is greater than the actual channel bandwidth of the network node, it is determined that broadband blocking attack has actually occurred. The specific details of the broadband blocking attack intent identification method are as follows:

[0116] 6.2.1 If the actual interference bandwidth of the external interferer meets the condition:

[0117]

[0118] Among them, B j B represents the actual interference bandwidth of the external interferator. n f is the actual bandwidth of the network node;n f is the center frequency of the communication band; j This is the center frequency of the interference band. When the actual interference bandwidth of the external interferer is less than or equal to the bandwidth of the actual network node, the external interferer has not caused broadband blocking interference.

[0119] 6.2.2 If the actual interference bandwidth of the external interferer meets the following conditions:

[0120] B j >5B n ,f n ∈[f j -B j / 2,f j +B j / 2]

[0121] When the actual interference bandwidth of an external interferer is greater than the actual bandwidth of a network node, the type of interference attack implemented by the external interferer can be accurately determined to be a broadband blocking interference attack.

[0122] 6.3 Bandwidth blocking interference intent identification: This invention is based on the actual interference signal power p of the external jammer. j The change in (f) identifies the interference intent within the system at this time, determining whether an external interferer has launched a band-blocking interference attack. This is determined by the change in the actual interference signal power of the external interferer in the traffic information: when the actual interference signal power of the external interferer is less than or equal to the actual communication signal power of the network node, it is determined that no band-blocking interference has occurred. When the actual interference signal power of the external interferer is greater than the actual communication signal power of the network node, it is determined that band-blocking interference has actually occurred. The specific details of the band-blocking interference intent identification method are as follows:

[0123] 6.3.1 If the actual interference signal power of the external interfering party meets the following condition:

[0124] p j (f)>p n (f),f n ∈[f j -B j / 2,f j +B j / 2]

[0125] Among them, B j p represents the actual interference bandwidth of the external interferator. n (f) represents the actual communication signal power of the network node; f n f is the center frequency of the communication band; jThe center frequency of the interference band is used; when the actual interference signal power of the external interferer is greater than the actual communication signal power of the network node, the type of interference attack implemented by the external interferer is accurately determined to be a band blocking interference attack.

[0126] 6.3.2 If the actual interference signal power of the external interfering party meets the following condition:

[0127] p j (f)≤p n (f),f n ∈[-∞, f j -B j / 2)∪(f j +B j / 2,+∞]

[0128] When the actual interference signal power of the external interferer is less than or equal to the actual communication signal power of the network node, the external interferer has not implemented band blocking interference.

[0129] Based on the identified interference attack types in the interference intent identification model, the sequential execution of actions by network nodes and external interferers, and the incompleteness of information between the two parties, this invention establishes an anti-interference game model for both attackers and defenders based on incomplete information dynamic game theory, mathematically represented as a quintuple form. They are represented as follows:

[0130] 6.4 Each network node and external interferer are rational participants: In an anti-interference model based on dynamic game theory with incomplete information In this context, rational participants are represented as in, The number of network nodes represents the defensive side, which is expressed as a set. n represents the nth network node, and N represents the total number of network nodes; This represents the number of attackers, i.e., external interferers, and its representation is in set form. j represents the j-th external disruptor, and J represents the total number of external disruptors.

[0131] 6.5 Each network node has three types of interference attacks and three corresponding anti-interference strategy classifications: in the anti-interference model based on incomplete information dynamic game. In Chinese, a categorical set is represented as Q n Anti-interference strategy deployed for the nth network node Let T be the set of interference attack types for the j-th external interferer, where each external interferer has T interference attack types that can be implemented. This represents the t-th type of interference attack launched by the j-th external interferer against the n-th network node.

[0132] 6.6 Each network node has an initial probability judgment of the type of external interferer: the initial judgment set of the network node is as follows: Then, the initial judgment of the t-th type of interference attack implemented by the n-th network node against the j-th external interference agent. Represented as:

[0133]

[0134] 6.7 Each network node updates its inference of the interference attack type from external interferers using Bayes' theorem: After receiving the interference attack type identified by the interference intent recognition model, the nth network node updates its inference of the interference attack type from the jth external interferer using Bayes' theorem. The posterior probability set of the inference of the t-th interference attack type implemented by the nth network node against the j-th external interferer is represented as follows: Wherein, the posterior probability of the t-th type of interference attack implemented by the n-th network node against the j-th external interferer. for:

[0135]

[0136] 6.8 Each network node selects an anti-interference strategy based on its own profit maximization principle: a set of utility functions. for U n and U j Let $\frac{ ... The damage caused to the nth network node by the t-th type of interference attack implemented by the j-th external interferer is denoted as:

[0137]

[0138] Among them, C n The importance of the nth network node is represented by a value between 1 and 10. The lethality of the t-th type of interference attack from the j-th external interferer to the n-th network node is represented by a value between 0 and 1. The risk D represents the network node's anti-interference strategy selection. n This indicates that the nth network node will target the wth type of interference attack launched by the jth external interferer. Mistakenly identified as the t-th type of interference attack The risk and return at that time are expressed as:

[0139]

[0140] Where, λ nThe degree of interference to the nth network node is represented by a value between 0 and 1; p(Q j,w |Q j,t () represents the w-th type of interference attack launched by the j-th external interferer by the n-th network node. Mistakenly identified as the t-th type of interference attack The probability of damage at time. The benefit of the t-th type of interference attack against the j-th external interferer is:

[0141]

[0142] in, A represents the interference cost incurred by the j-th external interference agent in launching the t-th type of interference attack against the n-th network node; n α1 represents the interference resistance cost of the nth network node; α2 represents the loss cost. The trade-off factor; α2 represents the interference resistance cost A of the nth network node. n The trade-off factor; α3 represents the interference cost of external interferers. The trade-off factors. Similarly, the payoff for the nth network node is:

[0143]

[0144] Where β1 represents the cost of loss. The trade-off factor; β2 represents the interference cost of external interferers. The trade-off factor; β3 represents the interference resistance cost A of the nth network node. n The trade-off factor; β4 represents the risk D of the anti-interference strategy selection of the nth network node. n The trade-off factors.

[0145] This invention establishes an anti-interference game model for both attackers and defenders based on a constructed dynamic game of incomplete information. It characterizes the attack-defense game process between network nodes and external interferers, obtains the payoff matrix for both sides, and considers the loss cost, the risk of network node anti-interference strategy selection, and the costs of interference and anti-interference as components of the objective function for both sides. It calculates the expected payoff for each network node's anti-interference efforts and the expected payoff for the external interferer's creation and generation of interference, thereby inferring the type of interference attack. By introducing the refined Bayesian Nash equilibrium method, network nodes calculate the posterior probability of their inference of the interference attack type based on the type of interference attack implemented by the external interferer, using Bayes' rule based on initial belief inference. This corrects each network node's inference of the interference attack type, thus determining the interference attack type for each network node. For the interference attack type determined after posterior probability correction, each network node solves the refined Bayesian equilibrium solution of the anti-interference model for each of the three interference attack types, obtaining the corresponding anti-interference strategy.

[0146] This invention utilizes a constructed anti-interference game model based on incomplete information dynamic game theory to derive corresponding anti-interference strategies for three types of interference attacks. See also Figure 5 , Figure 5 This is a flowchart illustrating the implementation process of the anti-interference strategy selection based on the incomplete information dynamic game theory model of this invention. This invention targets three types of interference attacks and selects a corresponding anti-interference strategy for each network node using the anti-interference game theory model. The selection of the optimal anti-interference strategy for each network node based on the incomplete information dynamic game theory model of this invention includes the following steps:

[0147] 6.9 Network nodes use Bayes' theorem to infer anti-interference strategies corresponding to three types of interference attacks. This invention infers anti-interference strategies for each network node against three types of interference attacks based on Bayesian rules derived from initial belief inference.

[0148]

[0149] In the formula For the nth network node, under Bayes' theorem, the tth type of interference attack is implemented against the jth external interferor. Posterior probability inference; This represents the gain of the nth network node. The anti-interference strategies for each network node against the three types of interference attacks inferred here by this invention are preliminary inferences.

[0150] 6.10 Solving for Refined Bayesian Equilibrium Solutions in Network Nodes Obtain the best anti-interference strategy This invention addresses three types of interference attacks from external interferers. The nth network node solves the refined Bayesian equilibrium solution. Obtain the optimal anti-interference strategy for the three types of interference attacks. Choose to make network nodes benefit Maximize the best anti-interference strategy The anti-interference strategy obtained in this invention is the optimal anti-interference strategy.

[0151] 6.11 Verify the optimal anti-interference strategy posterior probability Does a conflict exist? If the posterior probability of the nth network node implementing the tth type of interference against the jth external interferer is... With the best anti-interference strategy posterior probability If there is no conflict between them, then the refined Bayesian equilibrium solution obtained by this invention is... The best anti-interference strategy It is the best choice for each network node, meaning that for each of the three types of interference attacks, each network node can choose the best anti-interference strategy. To resist interference. Therefore, the present invention will This serves as the optimal anti-interference strategy for each network node.

[0152] In a wireless communication network, the actuator of each network node is directly connected to the internal infrastructure of the network node, deploying and executing the optimal anti-interference strategy selected by this invention for each network node. For three different types of interference attacks, this invention addresses forwarding interference attacks by transmitting multiple signal copies across multiple channels in time, frequency, or space, recovering the original communication signal at the receiving end of each network node, thus achieving signal diversity anti-interference for each network node. Secondly, for broadband congestion interference attacks, this invention addresses spreading spectrum anti-interference for each network node by concealing the communication signals of the wireless communication network and reducing narrowband interference effects. Finally, for frequency band congestion interference attacks, this invention addresses frequency hopping anti-interference for each network node by using the magnitude of the wireless communication center signal frequency and the rate of frequency change to evade interference. In this invention, each network node deploys and executes the corresponding anti-interference strategy in real time in the form of executable operations for the three types of interference attacks from external interferers, saves relevant information to a knowledge base, terminates the network node's anti-interference process, and completes the intention-driven system game-theoretic anti-interference process of this invention.

[0153] In summary, the intention-driven system game-based anti-interference method proposed in this invention solves the problem of network node interference detection and defense in wireless communication networks. Implementation includes: based on a closed-loop information transmission functional block of network data monitoring-interference intention identification-anti-interference game decision-making-anti-interference strategy execution, real-time sensing and storage of data from the network data monitoring model; constructing and utilizing an interference intention identification model to determine the type of interference attack; constructing and utilizing an anti-interference game model based on incomplete information dynamic game theory to obtain corresponding anti-interference strategies for the three types of interference attacks; and deploying and implementing the anti-interference strategies. This invention introduces the concept of interference intention identification into the anti-interference scenario of wireless communication networks, constructs an interference intention identification model, accurately identifies three types of interference attacks, and constructs an anti-interference game model based on incomplete information dynamic game theory. For each network node, it selects a corresponding anti-interference strategy in real time for each of the three types of interference attacks, improving the anti-interference efficiency of network nodes. This invention can perceive changes in interference attack types in real time, bridging the gap between the correct adaptation of interference attack types and anti-interference strategies, and enhancing the interference suppression capability of network nodes.

[0154] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An interference-resistant method for intent-driven system game, each network node in the intent-driven wireless communication network comprises functional blocks forming a closed-loop information transmission: network data monitoring, interference intent identification, interference-resistant game decision and interference-resistant strategy execution, and a knowledge base, wherein there is dynamic data information interaction between each functional block and the knowledge base, characterized in that, The intention-driven system game anti-interference method comprises the following steps: Step 1, using a network data monitoring model, real-time sensing and storing data to a knowledge base: the disturbers existing in real time in a wireless communication environment launch unknown type of interference attacks on each network node, sensors in the wireless communication network utilize an SNMP network management protocol and collect abnormal flow information in each network node in real time based on a request / response mode, and store the information to the knowledge base in the form of a management information base; Step 2, constructing an interference intention recognition model: calling the network data stored in the knowledge base, constructing an interference intention recognition model according to the abnormal flow data changes, setting a total interference time threshold value, judging the wireless communication signal residence time, and recording the recognition retransmission interference intention matching value; judging the interference bandwidth of the external disturber according to the set wireless communication channel threshold value, recording the recognition wideband blocking interference intention matching value; judging the interference power of the external disturber according to the set wireless communication signal center frequency threshold value, recording the recognition frequency band blocking interference intention matching value; Step 3, using the interference intention recognition model to judge the interference attack type: analyzing the data changes of the abnormal flow of the wireless communication network to obtain the interference attack type matching result; comparing the actual wireless communication signal residence time changes with the retransmission interference intention matching value to judge the retransmission interference attack; comparing the actual wireless communication channel bandwidth changes with the wideband blocking interference intention matching value to judge the wideband blocking interference attack; comparing the actual interference power of the external disturber with the frequency band blocking interference intention matching value to judge the frequency band blocking interference attack; each network node analyzes the above three interference attack types in real time through the interference intention matching value in the interference intention recognition model to obtain the classification of the three interference attack types; Step 4, constructing an anti-interference game model based on an incomplete information dynamic game: based on the classification judgment of the interference attack type, considering the characteristics that both the attacker and the defender cannot master all the information of the other party, combining the incomplete information dynamic game to construct an anti-interference game model of each network node and the external disturber; in the anti-interference game model based on the incomplete information dynamic game, each network node and the external disturber are rational participants; each network node has three interference attack types and three corresponding anti-interference strategy classifications; each network node has an initial probability judgment on the interference attack type implemented by the external disturber; each network node updates the inference on the interference attack type of the external disturber through the Bayes rule; each network node selects an anti-interference strategy according to the principle of maximizing its own benefit. Step 5, using the anti-interference game model based on incomplete information dynamic game, the corresponding anti-interference strategy is obtained for three types of interference attacks: according to the constructed anti-interference game model based on incomplete information dynamic game, the best income of each network node and external interferer about interference is calculated, and then the inference about the type of interference attack is obtained, the posterior probability of each network node about the inference of the type of interference attack is calculated by using the Bayesian rule based on the initial belief inference, to correct the inference of each network node about the type of interference attack, and the determination of the type of interference attack of each network node is formed; for the type of interference attack determined after the posterior probability correction, the refined Bayesian equilibrium solution of the anti-interference game model is solved for each network node for three types of interference attacks, and the corresponding anti-interference strategy is obtained; Step 6, deploy anti-interference strategy, end network node anti-interference process: the executor of each network node in wireless communication network executes anti-interference strategy, for forwarding interference attack, multiple signal copies are transmitted in time, frequency or space multiple channels, the original communication signal is recovered at the receiving end, signal diversity anti-interference is realized; for wideband jamming attack, by hiding communication signal and reducing the effect of narrowband interference, spread spectrum anti-interference is realized; for frequency band jamming attack, by frequency point size and the fast and slow of constantly changing frequency to avoid interference, frequency hopping anti-interference is realized; each network node deploys and executes corresponding anti-interference strategy in real time for three types of interference attacks of external interferer, and completes the intention driven system game anti-interference.

2. The intent-driven system game anti-interference method of claim 1, wherein, The abnormal traffic information monitored by the network node in step 1 specifically includes: The abnormal traffic information of the network data collected from the wireless communication network mainly includes two aspects: 1.1 Network traffic information: monitor and collect detailed information of network traffic in network link state, including: throughput, available bandwidth, packet loss rate, frequency, GPS position; 1.2 Hardware device state information: monitor and collect state information of network node hardware device, including: CPU load condition, memory availability and load condition, hard disk storage capacity; All network data abnormal traffic information is collected and saved to the knowledge base.

3. The intent-driven system game anti-interference method of claim 1, wherein, The interference intention recognition model used in step 3 to determine the type of interference attack specifically includes: 3.1 Repeater jamming intention recognition: According to the change of actual communication signal residence time ηT n , the jamming intention inside the system at this time is recognized, and it is determined whether the repeater jamming attack is implemented by the external jammer; the judgment is made through the change of actual communication signal residence time in the traffic information: when the actual total jamming time of the external jammer is greater than the actual communication signal residence time, it is judged that the repeater jamming does not occur; when the actual total jamming time of the external jammer is less than or equal to the actual communication signal residence time, it is judged that the repeater jamming actually occurs; the specific content is as follows, 3.1.1 If the actual total interference time of external interferer meets the condition: D n / c+T j >ηT n Wherein, c is the speed of light; D n is the distance between the network node and the external interferer; T j is the expected interference time of the external interferer; η is the residence time coefficient of successful interference, and the value is [0, 1]; T n is the residence time of the monitored communication signal; when the actual total interference time of the external interferer is greater than the actual residence time of the communication signal, the external interferer does not occur the forwarding type interference; 3.1.2 If the actual total interference time of external interferer meets the condition: D n / c+T j ≤ηT n When the actual total interference time of external interferer is less than or equal to the actual communication signal residence time, the external interferer has successfully implemented forwarding interference attack; 3.2 Broadband Congestion Intent Identification: Based on the actual interference bandwidth B of the external interferer. j The changes in traffic data are used to identify the interference intent within the system and determine whether an external interferer has launched a broadband blocking attack. The judgment is made by observing the changes in the actual interference bandwidth of the external interferer in the traffic information: when the actual interference bandwidth of the external interferer is less than or equal to the actual channel bandwidth of the network node, it is determined that no broadband blocking interference has occurred; when the actual interference bandwidth of the external interferer is greater than the actual channel bandwidth of the network node, it is determined that broadband blocking interference has actually occurred. 3.2.1 If the actual interference bandwidth of external interferer meets the condition: B j is the actual interference bandwidth of the external interferer; B n is the actual bandwidth of the network node; f n is the center frequency of the communication frequency band; f j is the center frequency of the interference frequency band; when the actual interference bandwidth of the external interferer is less than or equal to the actual bandwidth of the network node, the external interferer does not occur broadband blocking interference; 3.2.2 If the actual interference bandwidth of external interferer meets the condition: B j > 5B n f n ∈ [f j -B j / 2, f j +B j / 2] When the actual interference bandwidth of external interferer is greater than the actual bandwidth of network node, the external interferer has successfully implemented wideband jamming attack; 3.3 Band jamming interference intent recognition: according to the actual interference signal power p j (f) changes, identify the interference intent of the system at this time, determine whether the external interferer has implemented a band jamming interference attack; by the change of the actual interference signal power of the external interferer in the traffic information: when the actual interference signal power of the external interferer is less than or equal to the actual communication signal power of the network node, it is judged that no band jamming interference has occurred; when the actual interference signal power of the external interferer is greater than the actual communication signal power of the network node, it is judged that the band jamming interference has actually occurred; 3.3.1 If the actual interference signal power of external interferer meets the condition: p j (f) > p n (f), f n ∈ [f j -B j / 2, f j +B j / 2] wherein B j is the actual interference bandwidth of the external interferer; p n (f) is the actual communication signal power of the network node; f n is the center frequency of the communication frequency band; f j is the center frequency of the interference frequency band; when the actual interference signal power of the external interferer is greater than the actual communication signal power of the network node, the external interferer successfully implements the frequency band blocking interference; 3.3.2 If the actual interference signal power of external interferer meets the condition: p j (f)≤p n (f),f n ∈[-∞,f j -B j / 2)∪(f j +B j / 2,+∞] The external interferer does not implement the frequency band jamming interference when the actual interference signal power of the external interferer is less than or equal to the actual communication signal power of the network node.

4. The intent-driven system game anti-interference method of claim 1, wherein, Step 4, the anti-interference game model based on the incomplete information dynamic game is constructed, which is to describe the attack and defense game process of both sides by using the incomplete information dynamic game, and is expressed in the form of five tuples respectively expressed as: 4.1 Each network node and external disturber is a rational participant: Anti-interference game model based on dynamic game with incomplete information , where the rational participant is represented as , where is the number of defense sides, i.e., network nodes, which is represented in set form as n represents the nth network node, and N represents the total number of network nodes; is the number of attack sides, i.e., external disturbers, which is represented in set form as j represents the jth external disturber, and J represents the total number of external disturbers; 4.2 Each network node has three types of interference attack and three corresponding anti-interference strategy categories: in the anti-interference game model based on dynamic game with incomplete information , the classification set is represented as = {Q n , Q j}, Q n is the anti-interference strategy deployed by the nth network node, is the interference attack type set of the jth external interferer, each external interferer has T types of interference attack types that can be implemented, represents the tth interference attack type implemented by the jth external interferer on the nth network node; 4.3 Each network node has an initial probability judgment of the type of external interferer: the initial judgment set of network nodes is The initial judgment of the nth network node on the jth external interferer implementing the tth type of interference attack is denoted as ​ 4.4 Each network node updates the inference of the interference attack type of the external interferer by Bayes rule: after receiving the interference attack type identified by the interference intent identification model, the nth network node updates the inference of the interference attack type of the jth external interferer by Bayes rule, and the posterior probability set of the nth network node for the inference of the tth interference attack type implemented by the jth external interferer is represented as wherein the posterior probability of the nth network node for the inference of the tth interference attack type implemented by the jth external interferer is ​ 4.5 Each network node selects an anti-interference strategy according to the principle of maximizing its own benefits: a set of utility functions For U n and U j are the utility functions of the nth network node and the jth external interferer, respectively, and the loss cost, the risk of the network node's anti-interference strategy selection, and the interference and anti-interference costs are considered as the components of the objective functions of both parties in the game; the loss cost represents the damage caused by the jth external interferer to the nth network node by implementing the tth interference attack type, which is represented as: wherein C n is the importance of the nth network node, represented by a value between 1 and 10; represents the lethality of the tth interference attack type of the jth external interferer to the nth network node, represented by a value between 0 and 1; the network node anti-interference strategy selection risk D n represents the risk benefit of the nth network node when the wth interference attack type implemented by the jth external interferer is mistaken as the tth interference attack type .​ Where, λ n The degree of interference to the nth network node is represented by a value between 0 and 1; p(Q j,w |Q j,t () represents the w-th type of interference attack carried out by the j-th external interferer by the n-th network node. Mistakenly identified as the t-th type of interference attack The probability of damage at time; the benefit generated by the t-th type of interference attack against the j-th external interference is: wherein, denotes the interference cost produced by the jth external interferer implementing the tth interference attack type against the nth network node; A n denotes the anti-interference cost of the nth network node; a1 denotes the trade-off factor of the loss cost ; a2 denotes the trade-off factor of the anti-interference cost A n of the nth network node; a3 denotes the trade-off factor of the interference cost of the external interferer; and the profit of the nth network node is obtained in the same way. Where β1 represents the cost of loss. The trade-off factor; β2 represents the interference cost of external interferers. The trade-off factor; β3 represents the interference resistance cost A of the nth network node. n The trade-off factor; β4 represents the risk D of the anti-interference strategy selection of the nth network node. n The trade-off factors.

5. The intent-driven system game anti-interference method of claim 1, wherein the corresponding anti-interference strategies for the three types of interference attacks are obtained in step 5. comprising the steps of: 5.1 The network nodes use Bayesian rule to infer the anti-jamming strategy corresponding to the three types of jamming attacks The Bayesian rule based on the initial belief inference infers the anti-jamming strategy corresponding to the three types of jamming attacks for each network node In the formula is the type of the jth external interferer implemented by the nth network node under the Bayesian rule for the tth interference attack The posterior probability inference of the jth external interferer; represents the benefit of the nth network node; here the inferred anti-interference strategy is the preliminary inference result; 5.2 Network nodes solve the refined Bayesian equilibrium solution Obtain the best anti-interference strategy For the three types of interference attacks of external interferers, the nth network node solves the refined Bayesian equilibrium solution Obtain the best anti-interference strategy corresponding to the three types of interference attacks Select the best anti-interference strategy that maximizes the network node's revenue The anti-interference strategy obtained at this time is the best anti-interference strategy The anti-interference strategy obtained at this time is the best anti-interference strategy 5.3 Verify the optimal anti-interference strategy posterior probability Does a conflict exist? If the posterior probability of the nth network node implementing the tth type of interference against the jth external interferer is... With the best anti-interference strategy posterior probability If there is no conflict between them, then the refined Bayesian equilibrium solution is obtained. The best anti-interference strategy It is the best choice for each network node, meaning that for each of the three types of interference attacks, each network node can choose the best anti-interference strategy. To resist interference, This serves as the optimal anti-interference strategy for each network node.

Images