Data security protection method and system for intelligent modular fire station

Abstract

The invention provides a data security protection method and system for an intelligent modular fire station, and relates to the technical field of data security, and the method comprises the steps: taking a plurality of fire-fighting sensing sensors arranged in a target region as sensing nodes, and obtaining the equipment state data of the target region through the fire-fighting sensing sensors; according to the equipment state data and the historical behavior data of the corresponding sensing nodes, the credibility of the sensing nodes is determined through a credibility mechanism, and the sensing nodes with the credibility lower than a preset credibility threshold value serve as malicious sensing nodes to be isolated; the edge computing layer corresponding to the sensing node determines the attribute sensitivity of the equipment state data of the non-malicious sensing node, distributes a privacy budget to the equipment state data in combination with the access frequency of the historical behavior data of the non-malicious sensing node, and adds privacy noise to the equipment state data based on the privacy budget; and constructing encrypted transmission data, and transmitting the encrypted transmission data to a cloud platform through a communication protocol.

Description

TECHNICAL FIELD

[0001] The present application relates to data security technology, and in particular to a data security protection method and system for an intelligent modular fire station. BACKGROUND

[0002] With the rapid development of smart city and Internet of Things technology, intelligent modular fire stations have become an important development direction in the modern fire field. Intelligent modular fire stations integrate various sensors, intelligent devices and data analysis technology to realize intelligent and efficient fire monitoring, early warning, rescue and other links. However, in the operation process of intelligent modular fire stations, a large amount of data faces many security risks in the process of collection, transmission and storage, such as data leakage, tampering and theft.

[0003] In the prior art, the data security protection of intelligent modular fire stations mainly adopts the following methods:

[0004] Access control: by setting access permissions and identity authentication mechanisms, limiting unauthorized users' access and operation of data. However, access control is vulnerable to malicious operations by internal personnel and bypassing by external attackers, and is difficult to adapt to the collaborative needs of multiple departments and roles in the fire scene.

[0005] Intrusion detection: by deploying an intrusion detection system, real-time monitoring of network traffic and system logs, identifying suspicious behavior and abnormal events. However, intrusion detection is vulnerable to false positives and false negatives, and monitoring rules are difficult to comprehensively cover all types of attack methods, making it difficult to achieve precise protection.

[0006] The above-mentioned prior art to some extent solves the data security problem of intelligent modular fire stations, but still has the limitations of insufficient security, large performance overhead, poor adaptability, etc., and is difficult to meet the increasingly high requirements of data security protection. SUMMARY

[0007] The embodiments of the present application provide a data security protection method and system for an intelligent modular fire station, which can solve the problems in the prior art.

[0008] The first aspect of the embodiments of the present application is,

[0009] A data security protection method for an intelligent modular fire station is provided, comprising:

[0010] A plurality of fire perception sensors disposed in a target area are used as perception nodes, and device state data of the target area is obtained through the fire perception sensors;

[0011] According to the device state data and historical behavior data of the corresponding sensing node, the reputation of the sensing node is determined through a reputation mechanism, and sensing nodes with a reputation lower than a preset reputation threshold are isolated as malicious sensing nodes;

[0012] The edge computing layer corresponding to the sensing node determines the attribute sensitivity of the device state data of the non-malicious sensing node, allocates a privacy budget to the device state data in combination with the access frequency of the historical behavior data of the non-malicious sensing node, adds privacy noise to the device state data based on the privacy budget, constructs encrypted transmission data, and transmits the encrypted transmission data to the cloud platform through a communication protocol.

[0013] In an optional implementation,

[0014] According to the device state data and historical behavior data of the corresponding sensing node, the reputation of the sensing node is determined through a reputation mechanism, and sensing nodes with a reputation lower than a preset reputation threshold are isolated as malicious sensing nodes, including:

[0015] According to the historical behavior data of the sensing node, the number of positive interactions and the number of negative interactions of the historical behavior data are determined, and the number of positive interactions and the number of negative interactions are iteratively updated, respectively, wherein the positive interaction is used to indicate that the Bernoulli evaluation of the historical behavior is positive, and the negative interaction is used to indicate that the Bernoulli evaluation of the historical behavior is negative;

[0016] According to the initial reputation probability of the sensing node corresponding to the device state data, in combination with the updated number of positive interactions and the number of negative interactions, a time decay factor is introduced to determine the reputation of the sensing node;

[0017] Sensing nodes with a reputation lower than a preset reputation threshold are isolated as malicious sensing nodes.

[0018] In an optional implementation,

[0019] According to the initial reputation probability of the sensing node corresponding to the device state data, in combination with the updated number of positive interactions and the number of negative interactions, a time decay factor is introduced to determine the reputation of the sensing node, including:

[0020]

[0021] wherein R i (t) represents the reputation of the sensing node i at time t, P i (0) represents the initial reputation probability of the sensing node i, λ represents a time decay factor, represents the number of positive interactions of the sensing node i at time k, represents the number of negative interactions of the sensing node i at time k.

[0022] In an optional implementation,

[0023] The method further comprises storing the reputation of the perception node in the form of a transaction in the blockchain:

[0024] Obtaining the node ID, reputation and timestamp of the perception node obtaining the device state data, constructing a reputation transaction;

[0025] The perception node is a consensus participant of the blockchain, and the distributed blockchain network is established by selecting a blockchain platform, configuring a blockchain node, establishing network connection between perception nodes, deploying a smart contract and initializing a genesis block.

[0026] The reputation transaction is encapsulated according to the format requirement of the distributed blockchain network, and the encapsulated reputation transaction is submitted to the distributed blockchain network, and the transaction is verified through a consensus mechanism, wherein the consensus mechanism includes at least one of transaction submission, transaction verification, transaction sorting, block packaging and block verification.

[0027] The distributed ledger of the reputation transaction is maintained by all perception nodes in the distributed blockchain network, so that each perception node can obtain the latest reputation state of other nodes by querying the reputation transaction of the distributed blockchain network, and the reputation transaction of the distributed blockchain network is verified.

[0028] In an optional implementation,

[0029] The edge computing layer corresponding to the perception node determines the attribute sensitivity of the device state data of the non-malicious perception node, allocates a privacy budget to the device state data in combination with the access frequency of the historical behavior data of the non-malicious perception node, adds privacy noise to the device state data based on the privacy budget, and constructs encrypted transmission data, comprising:

[0030] According to the preset attribute sensitivity evaluation rule, the attribute sensitivity of the device state data of the non-malicious perception node is determined, wherein the attribute sensitivity evaluation rule includes at least one of attribute uniqueness and attribute correlation;

[0031] The attribute sensitivity is normalized, the sensitivity weight is allocated to the attribute sensitivity, and the privacy budget is allocated to the device state data according to the attribute sensitivity, the sensitivity weight and the access frequency;

[0032] According to the privacy budget, a noise scale parameter is generated based on the differential privacy mechanism, and privacy noise is added to the device state data in combination with the utility function, and encrypted transmission data is constructed.

[0033] In an alternative embodiment,

[0034] allocating a privacy budget to the device state data according to the attribute sensitivity, the sensitivity weight, and the access frequency comprises:

[0035]

[0036] wherein B j represents the privacy budget allocated to the jth attribute, w j represents the sensitivity weight of the jth attribute, f l represents the access frequency of the lth historical behavior data, s j represents the attribute sensitivity of the jth attribute, N represents the total number of attributes, and L represents the total number of historical behavior data.

[0037] In an alternative embodiment,

[0038] generating a noise scale parameter based on a differential privacy mechanism according to the privacy budget, and adding privacy noise to the device state data in combination with a utility function comprises:

[0039]

[0040] wherein Pr represents the privacy noise, B j represents the privacy budget allocated to the jth attribute, u() represents the utility function, x represents the device state data, r represents the noise scale parameter, H represents the set of device state data, and Δu represents the global sensitivity of the utility function.

[0041] A second aspect of the embodiments of the application,

[0042] A data security protection system of an intelligent modular fire station is provided, comprising:

[0043] A first unit is configured to acquire device state data of a target area through a plurality of fire perception sensors arranged in the target area as perception nodes;

[0044] A second unit is configured to determine the credibility of the perception nodes through a credibility mechanism according to the device state data and historical behavior data of the corresponding perception nodes, and isolate the perception nodes with a credibility lower than a preset credibility threshold as malicious perception nodes.

[0045] A third unit is configured to determine, by an edge computing layer corresponding to the perception node, attribute sensitivity of device state data of a non-malicious perception node, assign a privacy budget to the device state data in combination with an access frequency of historical behavior data of the non-malicious perception node, add privacy noise to the device state data based on the privacy budget, construct encrypted transmission data, and transmit the encrypted transmission data to a cloud platform through a communication protocol.

[0046] A third aspect of the embodiment of the application,

[0047] An electronic device is provided, comprising:

[0048] A processor;

[0049] A memory for storing processor-executable instructions;

[0050] The processor is configured to invoke the instructions stored in the memory to execute the method described above.

[0051] A fourth aspect of the embodiment of the application,

[0052] A computer-readable storage medium is provided, which stores computer program instructions, and the computer program instructions are executed by a processor to implement the method described above.

[0053] By deploying multiple fire-fighting perception sensors as perception nodes in the target area, and obtaining device state data through these sensors, comprehensive perception and collection of fire-fighting related data are realized, providing a data basis for subsequent data analysis and privacy protection. The reputation mechanism is used to evaluate the perception nodes, the reputation of the nodes is determined according to the device state data and historical behavior data, and malicious nodes with a reputation lower than a threshold are isolated, effectively identifying and preventing malicious behavior in the data collection process, improving the reliability and security of the data. The attribute sensitivity and historical access frequency are introduced in the edge computing layer, and the privacy budget is allocated to the device state data collected by the non-malicious nodes, combining the sensitivity of the data and the access characteristics, realizing the adaptability and refinement of privacy protection, and avoiding excessive or insufficient privacy protection. Based on the allocated privacy budget, privacy noise is added to the device state data, and encrypted transmission data is constructed, realizing privacy protection of the data while ensuring the effectiveness of the data, preventing leakage of sensitive information, and improving the security of the data in the transmission and storage process.

[0054] The overall scheme integrates reputation evaluation, privacy budget allocation, differential privacy protection, and encrypted transmission, forming a complete and multi-level data security protection system, and comprehensively improving the data security of the intelligent modular fire station, providing a solid security foundation for the analysis and application of fire data. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 A flowchart of a data security protection method of an intelligent modular fire station according to an embodiment of the present application is shown in

[0056] Figure 2 A structural diagram of a data security protection system of an intelligent modular fire station according to an embodiment of the present application is shown in DETAILED DESCRIPTION

[0057] In order to make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described below in connection with the drawings of the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0058] The technical solutions of the present application will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described in some embodiments.

[0059] Figure 1 A flowchart of a data security protection method of an intelligent modular fire station according to an embodiment of the present application is shown in Figure 1 The method comprises the following steps:

[0060] S101. A plurality of fire perception sensors arranged in a target area are taken as perception nodes, and device state data of the target area is acquired through the fire perception sensors;

[0061] S102. According to the device state data and historical behavior data of the corresponding perception nodes, the reputation degree of the perception nodes is determined through a reputation degree mechanism, and the perception nodes with a reputation degree lower than a preset reputation degree threshold are isolated as malicious perception nodes;

[0062] Exemplarily, the fire perception sensors refer to various sensors arranged in the target area for collecting fire-related data, such as smoke sensors, temperature sensors, and gas sensors; the device state data refers to data collected by the fire perception sensors and reflecting the running state of the devices in the target area, such as temperature, smoke concentration, and gas concentration; the historical behavior data refers to record data of the data collection, transmission, and processing behaviors of the perception nodes in the past period of time, and is used to evaluate the credibility and security of the nodes;

[0063] The reputation degree mechanism refers to a method for calculating the credibility of the nodes according to the current behavior and historical behavior of the perception nodes, and is commonly used to evaluate the security and data quality of the nodes. The reputation degree refers to the credibility of the perception nodes, and is calculated according to the behavior of the nodes.

[0064] In an optional implementation,

[0065] According to the device state data and the historical behavior data of the corresponding sensing node, the reputation of the sensing node is determined through a reputation mechanism, and the sensing node with a reputation lower than a preset reputation threshold is isolated as a malicious sensing node, comprising:

[0066] According to the historical behavior data of the sensing node, the number of positive interactions and the number of negative interactions of the historical behavior data are determined, and the number of positive interactions and the number of negative interactions are updated iteratively, wherein the positive interaction is used to indicate that the Bernoulli evaluation of the historical behavior is positive, and the negative interaction is used to indicate that the Bernoulli evaluation of the historical behavior is negative.

[0067] According to the initial reputation probability of the sensing node corresponding to the device state data, the updated number of positive interactions and the updated number of negative interactions, and a time decay factor, the reputation of the sensing node is determined.

[0068] The sensing node with a reputation lower than a preset reputation threshold is isolated as a malicious sensing node.

[0069] Exemplarily, the Bernoulli evaluation is a binary evaluation method, that is, the behavior or result of the evaluation object is divided into two mutually exclusive categories, which are usually represented by "positive" and "negative" or "success" and "failure". The present application uses "positive" and "negative" to evaluate the historical behavior of the sensing node.

[0070] The positive interaction refers to the behavior of the sensing node in the process of data collection, transmission, processing and the like, which is in line with the expectation and is beneficial to the normal operation of the system. These behaviors will be evaluated as "positive" and will play a role in improving the reputation of the node in the reputation calculation.

[0071] The negative interaction refers to the behavior of the sensing node in the process of data collection, transmission, processing and the like, which is not in line with the expectation and may endanger the normal operation of the system. These behaviors will be evaluated as "negative" and will play a role in reducing the reputation of the node in the reputation calculation.

[0072] The historical behavior data of the sensing node can include the following contents:

[0073] Data collection behavior: whether the time interval of data collection is stable, whether it is collected according to the preset frequency; whether the collected data is complete and accurate, whether there is loss or error; whether the working state of the sensor is normal, whether there is failure or abnormality.

[0074] Data transmission behavior: whether the delay of data transmission is within an acceptable range, whether there is obvious delay or interruption, whether there are abnormal situations such as packet loss and retransmission during data transmission, and whether the security of data transmission is guaranteed and whether there is the risk of tampering and eavesdropping.

[0075] Data processing behavior: whether the collected data is pre-processed in time and correctly by the node, such as filtering, compression, encryption, etc.; whether the node performs data fusion, feature extraction, etc. according to the predetermined rules and algorithms; whether the node has the behavior of malicious tampering and falsification of data.

[0076] For example, the initial reputation probability: each sensing node is assigned an initial reputation probability when it joins the system, indicating the initial trust degree of the sensing node; the number of positive interactions and the number of negative interactions, which reflect the behavior performance of the sensing node in the past period of time. The number of positive interactions represents the number of normal and compliant behaviors of the sensing node, such as collecting data on time, reporting data truthfully, etc. The number of negative interactions represents the number of abnormal and suspicious behaviors of the sensing node, such as data falsification, malicious tampering, etc.

[0077] The time decay factor can reflect the importance of the recent behavior of the node, indicating the influence degree of the past behavior on the current reputation. The greater the time decay factor, the more emphasis on the recent performance of the sensing node; the smaller the time decay factor, the more emphasis on the historical performance of the node.

[0078] The formula for determining the reputation of the sensing node includes:

[0079]

[0080] wherein R i (t) represents the reputation of the sensing node i at time t, P i (0) represents the initial reputation probability of the sensing node i, λ represents the time decay factor, represents the number of positive interactions of the sensing node i at time k, represents the number of negative interactions of the sensing node i at time k.

[0081] As you can understand, the time decay factor formula allows for dynamic evaluation of the credibility of sensing nodes. When a sensing node's behavior changes, its credibility also changes accordingly. This dynamic nature enables real-time tracking of the trustworthiness of sensing nodes, enabling timely detection and response to unusual changes in their behavior. By accumulating historical behavior, it can mitigate short-term fluctuations in their behavior. Even if a sensing node exhibits unusual behavior at a certain point in time, its credibility will remain high as long as it performs well most of the time. This robustness makes the credibility evaluation results more stable and reliable.

[0082] In an optional embodiment,

[0083] The method further includes storing the reputation of the sensing node in the form of a transaction in the blockchain:

[0084] Obtain the node ID, reputation, and timestamp of the sensing node obtaining the device status data, and construct a reputation transaction;

[0085] Using sensor nodes as consensus participants in the blockchain, a distributed blockchain network is established by selecting a blockchain platform, configuring blockchain nodes, establishing network connections between sensor nodes, deploying smart contracts, and initializing the genesis block.

[0086] Encapsulating the reputation transaction in accordance with the format requirements of the distributed blockchain network, submitting the encapsulated reputation transaction to the distributed blockchain network, and verifying the transaction through a consensus mechanism, wherein the consensus mechanism includes at least one of transaction submission, transaction verification, transaction sorting, block packaging, and block verification;

[0087] All the perception nodes in the distributed blockchain network jointly maintain the distributed ledger of the reputation transactions, ensuring that each perception node obtains the latest reputation status of other nodes by querying the reputation transactions of the distributed blockchain network, and verifies the reputation transactions of the distributed blockchain network.

[0088] Exemplarily, the reputation transaction format may include fields such as node ID, reputation score, and timestamp; a consortium chain or a private chain may be deployed, with the perception node acting as a node of the blockchain, participating in the consensus of the reputation transaction; the reputation evaluation result of the perception node is submitted to the blockchain network in the form of a transaction, and after verification by a consensus mechanism (such as PBFT, Raft, etc.), it is packaged into a block; all perception nodes in the blockchain network jointly maintain a distributed ledger of reputation transactions, and each node can query and verify the reputation of other perception nodes; malicious nodes are judged and isolated based on the reputation information on the blockchain.

[0089] In traditional reputation mechanisms, the reputation values of nodes are usually stored centrally in a reputation management server. This centralized storage method is prone to single point of failure and security vulnerabilities, and reputation data is easy to tamper with or forge, affecting the reliability of reputation evaluation. To overcome this problem, this application introduces blockchain technology, which uses its decentralized, tamper-proof, and distributed consensus features to achieve secure storage and trusted sharing of perception layer node reputation.

[0090] To store the reputation information of nodes in the blockchain, the data format of the reputation transaction needs to be designed first. Reputation transaction refers to the process of propagating and recording the reputation evaluation results of nodes in the blockchain network. Each reputation transaction contains the following key fields:

[0091] Node ID: The unique identifier of the perception node, which can be represented by the public key address of the node; Reputation score: The comprehensive reputation evaluation result of the perception node, which can be a normalized real value, such as a range of [0, 1]; Timestamp: The time when the perception node obtains the device state data.

[0092] To store and share reputation transactions, a blockchain network needs to be deployed among perception nodes. Alliance chain or private chain can be used, and perception nodes act as consensus participants in the blockchain to jointly maintain the distributed ledger of reputation transactions.

[0093] The deployment steps of the blockchain network are as follows:

[0094] Select a blockchain platform: According to the size, performance requirements, and development environment of perception nodes, select a suitable blockchain platform, such as Hyperledger Fabric, Ethereum, etc.; Configure the blockchain node: Install and configure the blockchain node software on the perception node, including the identity information, communication protocol, consensus algorithm, and other parameters of the node. Establish network connection between perception nodes: Through P2P protocol or other network communication methods, establish connection between perception nodes to form the blockchain network topology. Deploy smart contract: According to the data format and processing logic of reputation transactions, write and deploy smart contracts to realize the verification, storage, and query functions of reputation transactions. Initialize the genesis block: Generate a genesis block containing the initial reputation state as the starting point of the blockchain, ensuring that the initial state of all perception nodes is consistent.

[0095] Through the above steps, a distributed blockchain network is established among perception nodes, laying the foundation for the secure storage and consensus of reputation transactions.

[0096] The reputation evaluation results of the perception nodes need to be submitted in the form of transactions to the blockchain network, and after the verification of the consensus mechanism, they can be packaged into blocks and recognized and accepted by all perception nodes. The consensus mechanism is the core of the blockchain, ensuring the consistency and tamper resistance of reputation transactions.

[0097] The consensus process of reputation transactions is as follows:

[0098] Transaction submission: The perception node encapsulates the locally generated reputation evaluation results into a transaction and sends it to other perception nodes in the blockchain network through broadcast or peer-to-peer.

[0099] Transaction verification: The perception node that receives the transaction verifies the legality of the transaction, including the transaction format, signature, timestamp, etc., and checks whether the reputation score in the transaction is reasonable. The perception node sorts the received transactions to form a consistent transaction sequence across the network.

[0100] Block packaging: According to the transaction sequence, a perception node in the blockchain network (such as the main node or the rotating block node) packages the transactions into a block and broadcasts the block to other perception nodes.

[0101] Block verification: The perception node that receives the block verifies the block, including the legality of the transaction and the validity of the block header. If the verification is passed, the perception node adds the block to the local blockchain and updates the reputation of the perception node.

[0102] Based on the reputation stored in the blockchain, the perception node can easily query and verify the reputation of other perception nodes. The specific steps are as follows:

[0103] Reputation query: The perception node queries the latest reputation transaction of the target perception node stored on the blockchain through the API or smart contract interface provided by the blockchain according to the ID of the target perception node.

[0104] Reputation verification: The perception node verifies the queried reputation transaction, including the integrity of the transaction, the validity of the signature, and the reasonableness of the timestamp. Through verification, the perception node can confirm the authenticity and credibility of the reputation.

[0105] Reputation tracking: Since the blockchain stores the historical transaction records of the reputation, the perception node can also track the change history of the reputation of the target perception node and analyze the stability and trend of the reputation.

[0106] It can be understood that by storing the reputation of the perception node in the form of transactions in the blockchain, the trusted storage of the reputation can be realized. The tamper-proof nature and distributed consensus mechanism of the blockchain ensure the integrity and consistency of the reputation data, preventing malicious tampering or forgery of the reputation by a single perception node. Each perception node can obtain the real and trusted reputation status of other perception nodes, providing a reliable basis for subsequent decision-making and action. By including the timestamp of the perception node obtaining the device state data in the reputation transaction, real-time updating of the reputation can be achieved. Each time the perception node collects new device state data, a new reputation transaction is generated and submitted to the blockchain network. This real-time reputation updating mechanism enables the reputation status of the perception node to reflect its latest behavior and performance in a timely manner, improving the timeliness and accuracy of reputation evaluation. By including the perception node as a consensus participant of the blockchain and building a distributed blockchain network, decentralized reputation management is achieved. Each perception node participates in the generation, verification, and storage of reputation transactions, no longer relying on centralized reputation management institutions. This decentralized architecture improves the fault tolerance and availability of the system, eliminating the single point of failure risk and performance bottleneck of centralized management, making reputation management more secure and efficient.

[0107] S103. The edge computing layer corresponding to the perception node determines the attribute sensitivity of the device state data of the non-malicious perception node, allocates a privacy budget to the device state data in combination with the access frequency of the historical behavior data of the non-malicious perception node, adds privacy noise to the device state data based on the privacy budget, constructs encrypted transmission data, and transmits the encrypted transmission data to the cloud platform through a communication protocol.

[0108] Illustratively, the edge computing layer has computing and processing capabilities deployed near the network edge of the Internet of Things device or data source. Compared to transmitting all data to the cloud for processing, the edge computing layer can preprocess, filter, and analyze data locally, reducing network transmission delay and bandwidth pressure.

[0109] Attribute sensitivity refers to the sensitivity of each attribute in the device state data. Different attributes may contain different levels of privacy information. Attribute sensitivity reflects the degree of influence on privacy control when these attributes are leaked. In the field of intelligent modular fire stations, attribute sensitivity can refer to the control privacy value of different fire equipment. Different fire equipment plays different roles when a fire occurs, and to avoid illegal intrusion, their control values must be kept secret.

[0110] Privacy budget refers to the privacy resources allocated to each query or data access in the differential privacy protection mechanism. Privacy budget controls the intensity of adding privacy noise, the larger the budget, the smaller the noise, and the lower the privacy protection level of the data. Privacy noise refers to the random noise added to the original data in the differential privacy protection. By introducing random noise, the characteristics of individual data can be masked to prevent attackers from inferring sensitive information from the data. In the embodiments of the present application, sensitive information can include API interface information of fire-fighting equipment, control instructions of fire-fighting equipment, etc.

[0111] In an alternative embodiment,

[0112] The edge computing layer corresponding to the perception node determines the attribute sensitivity of the device state data of the non-malicious perception node, allocates a privacy budget to the device state data in combination with the access frequency of the historical behavior data of the non-malicious perception node, adds privacy noise to the device state data based on the privacy budget, and constructs encrypted transmission data, including:

[0113] According to a preset attribute sensitivity evaluation rule, the attribute sensitivity of the device state data of the non-malicious perception node is determined, wherein the attribute sensitivity evaluation rule includes at least one of attribute uniqueness and attribute correlation;

[0114] The attribute sensitivity is normalized, a sensitivity weight is allocated to the attribute sensitivity, and a privacy budget is allocated to the device state data according to the attribute sensitivity, the sensitivity weight, and the access frequency;

[0115] According to the privacy budget, a noise scale parameter is generated based on a differential privacy mechanism, and privacy noise is added to the device state data in combination with a utility function to construct encrypted transmission data.

[0116] Optionally, the privacy risk of each attribute of the device state data is evaluated to determine the sensitivity level of each attribute. Expert evaluation, user survey, data analysis, etc. can be used to consider factors such as attribute semantics, application scenarios, and privacy impact to obtain a sensitivity score. The attribute sensitivity is divided into multiple levels, such as high, medium, and low, which correspond to different privacy protection strengths and privacy budget allocation strategies. The attribute sensitivity evaluation rule of the present application includes at least one of attribute uniqueness, attribute correlation, and attribute application range.

[0117] The attribute uniqueness evaluation is used to evaluate the uniqueness of each attribute in the device state data. The higher the uniqueness, the higher the attribute sensitivity. The frequency distribution of the attribute value can be counted: for each attribute, the frequency of different values is counted. The more concentrated the frequency distribution, the lower the uniqueness of the attribute.

[0118] Or calculate the information entropy of the attribute value: according to the frequency distribution of the attribute value, calculate the information entropy index such as Shannon entropy or Gini coefficient. The higher the information entropy, the higher the uniqueness of the attribute.

[0119] Set a uniqueness threshold: set a uniqueness threshold according to experience or data analysis. When the information entropy of the attribute exceeds the threshold, it is considered that the uniqueness of the attribute is high, and the sensitivity is high.

[0120] Among them, the attribute correlation evaluation is used to evaluate the correlation between attributes in the device state data. The stronger the correlation, the higher the attribute sensitivity.

[0121] An attribute correlation graph can be constructed: according to the semantic relationship between attributes or data analysis, an attribute correlation graph is constructed. The nodes in the graph represent attributes, and the edges represent the correlation between attributes. Or calculate the centrality of the attribute: in the attribute correlation graph, calculate the centrality index of each attribute node, such as degree centrality, betweenness centrality, etc. The higher the centrality, the stronger the correlation between the attribute and other attributes.

[0122] In an optional implementation,

[0123] According to the attribute sensitivity, the sensitivity weight and the access frequency, the device state data is allocated a privacy budget, which includes:

[0124]

[0125] Where B j represents the privacy budget allocated to the jth attribute, w j represents the sensitivity weight of the jth attribute, f l represents the access frequency of the lth historical behavior data, s j represents the attribute sensitivity of the jth attribute, N represents the total number of attributes, and L represents the total number of historical behavior data.

[0126] The access frequency reflects the influence degree of the historical behavior data on the current attribute. The higher the frequency, the more privacy budget is allocated, which means that the attribute is frequently accessed in the historical behavior and needs more privacy protection.

[0127] The attribute sensitivity in the formula quantifies the privacy sensitivity of the attribute as a numerical value. By quantifying the attribute sensitivity, the demand of different attributes for privacy protection can be more accurately evaluated. The quantified attribute sensitivity can be combined with the sensitivity weight to determine the privacy budget allocation of the attribute. The quantified attribute sensitivity makes the privacy protection decision more objective and measurable, reducing the influence of subjective judgment.

[0128] By comprehensively considering the attribute sensitivity, sensitivity weight and access frequency, the formula can optimize the allocation and utilization of privacy budget. The privacy budget is allocated to the attribute that needs protection the most, avoiding over-protection of insensitive attributes and improving the efficiency of privacy protection. At the same time, more privacy budget is allocated to the attribute with high sensitivity and frequent access, ensuring the effectiveness of privacy protection. By optimizing the utilization of privacy budget, the maximum privacy protection effect can be achieved under limited privacy protection resources.

[0129] In an optional implementation,

[0130] According to the privacy budget, a noise scale parameter is generated based on a differential privacy mechanism, and privacy noise is added to the device state data in combination with a utility function, comprising:

[0131]

[0132] wherein Pr represents the privacy noise, B j represents the privacy budget allocated to the jth attribute, u() represents the utility function, x represents the device state data, r represents the noise scale parameter, H represents the set of device state data, and Δu represents the global sensitivity of the utility function.

[0133] Exemplarily, the utility function is used to measure the utility value of the device state data under the noise scale parameter, and the utility function reflects the usefulness and quality of the data. The higher the utility value, the smaller the influence of noise on the data, and the better the usability of the data. By introducing the utility function, the actual utility of the data can be considered when generating the privacy noise, so that the data after adding noise still has certain usability.

[0134] The noise scale parameter is used to control the intensity and range of the privacy noise. The noise scale parameter can be adaptively adjusted according to the values of the privacy budget and the utility function. When the privacy budget is large, the noise scale parameter can be correspondingly reduced, the generated privacy noise has low intensity, and more data utility is retained. When the privacy budget is small, the noise scale parameter can be correspondingly increased, the generated privacy noise has high intensity, and stronger privacy protection is provided. Through adaptive adjustment of the noise scale parameter, the generation of privacy noise can be flexibly controlled under different privacy protection requirements, and dynamic balance of privacy protection can be achieved.

[0135] The global sensitivity reflects the sensitivity of the utility function to data changes. The higher the sensitivity, the more sensitive the utility function is to data perturbation. When generating privacy noise, the global sensitivity is used to calibrate the privacy budget and the utility function to meet the requirements of differential privacy. By considering the global sensitivity, it can be ensured that the generated privacy noise can still provide sufficient privacy protection in the worst case.

[0136] Figure 2FIG. 1 is a structural diagram of a data security protection system for an intelligent modular fire station according to an embodiment of the present invention. Figure 2 As shown, the system includes:

[0137] The first unit is configured to use a plurality of fire sensing sensors deployed in a target area as sensing nodes and obtain equipment status data of the target area through the fire sensing sensors;

[0138] The second unit is configured to determine the credibility of the sensing node through a credibility mechanism based on the device status data and the historical behavior data of the corresponding sensing node, and isolate the sensing node whose credibility is lower than a preset credibility threshold as a malicious sensing node;

[0139] The third unit is used for the edge computing layer corresponding to the perception node to determine the attribute sensitivity of the device status data of the non-malicious perception node, allocate a privacy budget to the device status data based on the access frequency of the historical behavior data of the non-malicious perception node, add privacy noise to the device status data based on the privacy budget, construct encrypted transmission data, and transmit the encrypted transmission data to the cloud platform through the communication protocol.

[0140] According to a third aspect of the embodiments of the present invention,

[0141] An electronic device is provided, comprising:

[0142] processor;

[0143] a memory for storing processor-executable instructions;

[0144] The processor is configured to call the instructions stored in the memory to execute the aforementioned method.

[0145] According to a fourth aspect of the embodiments of the present invention,

[0146] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0147] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0148] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions recorded in the above embodiments can be modified, or some or all of the technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A data security protection method for an intelligent modular fire station, characterized in that: include: Multiple fire sensing sensors deployed in the target area are used as sensing nodes, and the equipment status data of the target area is obtained through the fire sensing sensors; Determine the credibility of the sensing node through a credibility mechanism based on the device status data and the historical behavior data of the corresponding sensing node, and isolate the sensing node whose credibility is lower than a preset credibility threshold as a malicious sensing node; The edge computing layer corresponding to the perception node determines the attribute sensitivity of the device status data of the non-malicious perception node, allocates a privacy budget to the device status data based on the access frequency of the historical behavior data of the non-malicious perception node, adds privacy noise to the device status data based on the privacy budget, constructs encrypted transmission data, and transmits the encrypted transmission data to the cloud platform through the communication protocol.

2. The method according to claim 1, characterized in that Determining the credibility of the sensing node through a credibility mechanism based on the device status data and the historical behavior data of the corresponding sensing node, and isolating the sensing node whose credibility is lower than a preset credibility threshold as a malicious sensing node includes: Determining, based on the historical behavior data of the sensing node, the number of positive interactions and the number of negative interactions of the historical behavior data, and iteratively updating the number of positive interactions and the number of negative interactions, respectively, wherein a positive interaction is used to indicate a Bernoulli evaluation of the historical behavior as positive, and a negative interaction is used to indicate a Bernoulli evaluation of the historical behavior as negative; According to the initial credibility probability of the sensing node corresponding to the device status data, combined with the updated number of positive interactions and negative interactions, a time decay factor is introduced to determine the credibility of the sensing node; Sensing nodes whose credibility is lower than the preset credibility threshold are isolated as malicious sensing nodes.

3. The method according to claim 2, characterized in that Based on the initial credibility probability of the sensing node corresponding to the device status data, combined with the updated number of positive interactions and negative interactions, a time decay factor is introduced to determine the credibility of the sensing node, including: Among them, R i (t) represents the credibility of sensing node i at time t, P i (0) represents the initial credibility probability of the sensing node i, λ represents the time decay factor, represents the number of positive interactions of sensing node i at time k, represents the number of negative interactions of sensor node i at time k.

4. The method according to claim 2, characterized in that The method further includes storing the reputation of the sensing node in the form of a transaction in the blockchain: Obtain the node ID, reputation, and timestamp of the sensing node obtaining the device status data, and construct a reputation transaction; Using sensor nodes as consensus participants in the blockchain, a distributed blockchain network is established by selecting a blockchain platform, configuring blockchain nodes, establishing network connections between sensor nodes, deploying smart contracts, and initializing the genesis block. Encapsulating the reputation transaction in accordance with the format requirements of the distributed blockchain network, submitting the encapsulated reputation transaction to the distributed blockchain network, and verifying the transaction through a consensus mechanism, wherein the consensus mechanism includes at least one of transaction submission, transaction verification, transaction sorting, block packaging, and block verification; All the perception nodes in the distributed blockchain network jointly maintain the distributed ledger of the reputation transactions, ensuring that each perception node obtains the latest reputation status of other nodes by querying the reputation transactions of the distributed blockchain network, and verifies the reputation transactions of the distributed blockchain network.

5. The method according to claim 1, wherein The edge computing layer corresponding to the sensing node determines the attribute sensitivity of the device status data of the non-malicious sensing node, allocates a privacy budget to the device status data based on the access frequency of the historical behavior data of the non-malicious sensing node, adds privacy noise to the device status data based on the privacy budget, and constructs encrypted transmission data, including: Determining the attribute sensitivity of the device status data of the non-malicious sensing node according to a preset attribute sensitivity evaluation rule, wherein the attribute sensitivity evaluation rule includes at least one of attribute uniqueness and attribute relevance; Normalizing the attribute sensitivity, assigning a sensitivity weight to the attribute sensitivity, and allocating a privacy budget to the device state data according to the attribute sensitivity, the sensitivity weight, and the access frequency; A noise scale parameter is generated based on the privacy budget and a differential privacy mechanism, and privacy noise is added to the device state data in combination with a utility function to construct encrypted transmission data.

6. The method according to claim 5, characterized in that Allocating a privacy budget to the device state data according to the attribute sensitivity, the sensitivity weight, and the access frequency includes: Among them, B j represents the privacy budget allocated to the j-th attribute, w j represents the sensitivity weight of the jth attribute, f l Indicates the access frequency of the lth historical behavior data, s j represents the attribute sensitivity of the j-th attribute, N represents the total number of attributes, and L represents the total amount of historical behavior data.

7. The method according to claim 5, characterized in that Generating a noise scale parameter based on the differential privacy mechanism according to the privacy budget, and adding privacy noise to the device state data in combination with the utility function includes: Among them, Pr represents privacy noise, B j represents the privacy budget assigned to the j-th attribute, u() represents the utility function, x represents the device state data, r represents the noise scale parameter, H represents the set of device state data, and Δu represents the global sensitivity of the utility function.

8. A data security protection system for an intelligent modular fire station, used to implement the method according to any one of claims 1 to 7, characterized in that: include: The first unit is configured to use a plurality of fire sensing sensors deployed in a target area as sensing nodes and obtain equipment status data of the target area through the fire sensing sensors; The second unit is configured to determine the credibility of the sensing node through a credibility mechanism based on the device status data and the historical behavior data of the corresponding sensing node, and isolate the sensing node whose credibility is lower than a preset credibility threshold as a malicious sensing node; The third unit is used for the edge computing layer corresponding to the perception node to determine the attribute sensitivity of the device status data of the non-malicious perception node, allocate a privacy budget to the device status data based on the access frequency of the historical behavior data of the non-malicious perception node, add privacy noise to the device status data based on the privacy budget, construct encrypted transmission data, and transmit the encrypted transmission data to the cloud platform through the communication protocol.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images