Security decision-making method, device, equipment and storage medium of Internet of Things

Through the hierarchical decision model and decision tree algorithm, honeypot slices and service slices are generated, which solves the problem of network slice management neglecting security performance in industrial IoT environments, and reduces security risks and improves system stability under network performance guarantees .

CN115334096BInactive Publication Date: 2024-11-05BEIJING UNIV OF POSTS & TELECOMM
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210788091.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-07-04
Publication Date
2024-11-05
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The network slice management mechanism in the industrial Internet of Things environment ignores the security performance during network attacks, resulting in the service system being easily paralyzed and lacks effective security decision-making methods.

Method used

Using a hierarchical decision-making model, through a comprehensive decision-maker and slice generator, based on the decision-making conditions of the physical layer, device layer, simulation layer, resource layer and link layer, a decision tree algorithm is built to generate honeypot slices and service slices to improve the network Security.

Benefits of technology

On the basis of ensuring network performance, reduce network security risks, improve the security of IoT networks, prevent external attacks, and ensure the stable operation of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115334096B_ABST
    Figure CN115334096B_ABST
Patent Text Reader

Abstract

The present invention provides a security decision method, device, equipment and storage medium for the Internet of Things, the method comprising: based on at least one decision condition layer in a hierarchical decision model, determining the value of the decision condition corresponding to the at least one decision condition layer, wherein the hierarchical decision model comprises at least one decision condition layer and a slice layer, the slice layer comprises a comprehensive decision maker and a slice generator, the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things; using the comprehensive decision maker, making a comprehensive decision judgment on the value of the decision condition corresponding to the at least one decision condition layer, and obtaining a judgment result; generating a network slice based on the slice generator and the judgment result. The present invention reduces the security risk of the network and ensures the network security of the Internet of Things on the basis of ensuring network performance.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of network security technology, and in particular to a security decision-making method, device, equipment and storage medium for the Internet of Things. Background Art

[0002] In industrial applications, many devices, machines, and applications will rely on network connections, and industrial networks will be highly heterogeneous. This requires the network to provide services for various applications with different quality of service (QoS) requirements. Network slicing technology solves this problem well.

[0003] In the industrial Internet of Things environment, due to the richness and heterogeneity of devices, there will be huge attack risks. The current network slicing management mechanism often only considers the performance of resource allocation and the diversification of network functions to obtain better network services, but ignores the security performance of network slicing in the face of external attacks, which can easily lead to service system paralysis.

[0004] Therefore, how to provide a security decision-making method for industrial Internet of Things has become an urgent problem to be solved. Summary of the invention

[0005] The present invention provides a security decision method, device, equipment and storage medium for the Internet of Things, which are used to solve the defect in the prior art that the industrial Internet of Things is easily vulnerable to network attacks and causes network insecurity, and to ensure the network security of the Internet of Things.

[0006] The present invention provides a security decision-making method for the Internet of Things, comprising:

[0007] Based on at least one decision condition layer in a hierarchical decision model, determine the value of the decision condition corresponding to the at least one decision condition layer, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things;

[0008] Using the comprehensive decision maker, a comprehensive decision is made on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a decision result;

[0009] Based on the slice generator and the determination result, a network slice is generated.

[0010] According to a security decision-making method for an Internet of Things provided by the present invention, the at least one decision condition layer includes at least one of the following:

[0011] A physical layer, wherein the decision conditions corresponding to the physical layer include the physical trust of the Internet of Things;

[0012] A device layer, wherein the decision conditions corresponding to the device layer include the traffic information entropy of the Internet of Things and the average memory consumption of the device;

[0013] A simulation layer, wherein the decision condition corresponding to the simulation layer includes an information entropy difference of the Internet of Things;

[0014] A resource layer, wherein the decision conditions corresponding to the resource layer include the remaining physical resources of the Internet of Things;

[0015] Link layer, the decision condition corresponding to the link layer includes the number of attack paths of the Internet of Things.

[0016] According to a security decision-making method for the Internet of Things provided by the present invention, the comprehensive decision maker is used to perform comprehensive decision-making judgment on the values ​​of the decision conditions corresponding to the at least one decision condition layer to obtain the judgment result, including:

[0017] Using the comprehensive decision maker, constructing a decision tree algorithm based on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer;

[0018] Through the decision tree algorithm, a comprehensive decision is made on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a decision result.

[0019] According to a security decision-making method for the Internet of Things provided by the present invention, the comprehensive decision maker is used to construct a decision tree algorithm based on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer, including:

[0020] Using the comprehensive decision maker, based on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer, determine the decision order for the decision condition layers in the hierarchical decision model;

[0021] A decision tree algorithm is constructed based on the decision sequence of the decision condition layer in the hierarchical decision model.

[0022] According to a security decision-making method for the Internet of Things provided by the present invention, the decision tree algorithm is used to perform a comprehensive decision-making judgment on the values ​​of the decision conditions corresponding to the at least one decision condition layer to obtain a judgment result, including:

[0023] Adopting the decision tree algorithm, based on the decision sequence, sequentially making decisions on the values ​​of the decision conditions corresponding to the decision condition layers;

[0024] In the case where it is determined that the value of the decision condition corresponding to the decision condition layer is not within a preset range, determining that the determination result has a security risk;

[0025] When it is determined that the value of the decision condition corresponding to the decision condition layer does not exist and is not within a preset range, the determination result is determined to be that there is no security risk.

[0026] According to a security decision method for an Internet of Things provided by the present invention, generating a network slice based on the slice generator and the determination result includes:

[0027] If it is determined that the determination result has a security risk, a honeypot slice and a service slice are generated, wherein the honeypot slice is used to attract network attacks;

[0028] If it is determined that the judgment result does not pose a security risk, only service slices are generated.

[0029] According to a security decision-making method for an Internet of Things provided by the present invention, generating a honeypot slice includes:

[0030] The virtual network function (VNF) is used to simulate the various components of the devices in the Internet of Things, and all VNFs are connected to form a service function chain (SFC) of the simulated Internet of Things devices. Physical resources are allocated to the SFC and deployed on the honeypot slice.

[0031] According to a security decision-making method for an Internet of Things provided by the present invention, the method further includes:

[0032] Based on the operation log of the honeypot slice, the parameters of the hierarchical decision model are adjusted.

[0033] According to a security decision-making method for the Internet of Things provided by the present invention, adjusting the parameters of the hierarchical decision model based on the operation log of the honeypot slice includes:

[0034] The operation log of the honeypot slice is sent to the simulation layer, so that the simulation layer adjusts the activation function of the simulation layer based on the operation log of the honeypot slice.

[0035] The present invention provides a security decision-making device for the Internet of Things, comprising:

[0036] A first determination module is used to determine the value of the decision condition corresponding to the at least one decision condition layer based on at least one decision condition layer in the hierarchical decision model, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things;

[0037] A comprehensive decision judging module, used to use the comprehensive decision maker to perform comprehensive decision judging on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a judging result;

[0038] A network slice generation module is used to generate a network slice based on the slice generator and the determination result.

[0039] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a security decision-making method for the Internet of Things as described in any one of the above is implemented.

[0040] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the security decision-making methods for the Internet of Things described above.

[0041] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the security decision-making methods for the Internet of Things as described above.

[0042] The security decision-making method, device, equipment and storage medium of the Internet of Things provided by the present invention adopt a hierarchical decision-making model to perform comprehensive decision-making judgments on the values ​​of the decision conditions corresponding to the at least one decision condition layer, generate network slices based on the judgment results, and simulate the dynamic changes of the industrial Internet of Things environment through the values ​​of the decision conditions corresponding to the at least one decision condition layer. On the basis of ensuring network performance, the security risks of the network are reduced to ensure the network security of the Internet of Things. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 It is a flowchart of the security decision-making method of the Internet of Things provided by the present invention;

[0045] Figure 2 It is a structural diagram of a hierarchical decision model based on network slicing technology under the industrial Internet of Things provided by the present invention;

[0046] Figure 3 It is an example diagram of a decision tree for a slice layer in a hierarchical model provided by the present invention;

[0047] Figure 4It is a flow chart of slice generation management of the slice layer in the hierarchical model provided by the present invention;

[0048] Figure 5 It is a structural schematic diagram of the security decision-making device of the Internet of Things provided by the present invention;

[0049] Figure 6 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] A network slice is essentially a piece of physical infrastructure that contains resources from multiple network domains. Each slice is therefore an isolated end-to-end (E2E) network that can meet the requirements of the network services it hosts, such as quality of service (QoS), reliability, and security. Based on application requirements, a slice can be dynamically created or split, scaled up or down with more or less resources, and reconfigured by adding or removing network functions. This mechanism can maximize the efficiency of the practicality and reusability of resources in the physical infrastructure of the Industrial Internet of Things (IIoT).

[0052] However, network slices are vulnerable to external attacks and have low security performance, which can easily lead to service system paralysis. Some network attacks are likely to target a large number of IoT devices, affecting service quality or even destroying physical network resources. If resources are destroyed, network slice management will collapse, causing serious losses.

[0053] Therefore, the present invention provides a security decision-making method, device, equipment and storage medium for the Internet of Things.

[0054] Combine the following Figure 1-Figure 6 The present invention describes a security decision-making method, device, equipment and storage medium for the Internet of Things.

[0055] Figure 1 : is a flow chart of the security decision-making method of the Internet of Things provided by the present invention, such as Figure 1 As shown, the method includes:

[0056] Step 100, based on at least one decision condition layer in a hierarchical decision model, determining the value of the decision condition corresponding to the at least one decision condition layer, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things;

[0057] Step 110, using the comprehensive decision maker to perform comprehensive decision determination on the values ​​of the decision conditions corresponding to the at least one decision condition layer, and obtain a determination result;

[0058] Step 120: Generate a network slice based on the slice generator and the determination result.

[0059] Specifically, based on at least one decision condition layer in the hierarchical decision model, the values ​​of the decision conditions corresponding to the at least one decision condition layer can be determined, that is, the network environment of the Internet of Things currently being secured can be simulated through the values ​​of the decision conditions that can characterize the network environment of the Internet of Things;

[0060] Specifically, the hierarchical decision model includes at least one decision condition layer for simulating a network environment. The hierarchical decision model also includes a slice layer, which includes a comprehensive decision maker and a slice generator. The comprehensive decision maker is used to make a comprehensive decision on the values ​​of the decision conditions corresponding to the at least one decision condition layer, and the slice generator is used to generate a network slice.

[0061] The present invention proposes a hierarchical model to simulate the Internet of Things environment, makes comprehensive decisions through hierarchical strategies, and implements a security management solution for network slices based on the results.

[0062] Optionally, the network slice may include a service slice;

[0063] Optionally, the network slice may include a honeypot slice;

[0064] Unlike traditional networks, the industrial Internet of Things is highly dynamic, involving issues such as cross-domain networks and heterogeneous devices. Therefore, the present invention proposes a hierarchical decision-making model. With the support of technologies such as network slicing, software-defined networks, and network technology virtualization, a decision-making solution for dynamically managing network slices is provided. The hierarchical decision model takes into account many factors and simulates the dynamic changes in the industrial Internet of Things environment to obtain an effective decision-making solution. And because honeypot slices can resist external network attacks to a certain extent, the present invention further reduces the security risks of the network on the basis of ensuring network performance, and realizes the stable operation of network slice management.

[0065] The security decision-making method for the Internet of Things provided by the present invention adopts a hierarchical decision model, performs comprehensive decision-making judgments on the values ​​of the decision conditions corresponding to the at least one decision condition layer, generates a network slice based on the judgment result, and simulates the dynamic changes of the industrial Internet of Things environment through the values ​​of the decision conditions corresponding to the at least one decision condition layer. On the basis of ensuring network performance, the security risks of the network are reduced to ensure the network security of the Internet of Things.

[0066] Optionally, the at least one decision condition layer includes at least one of the following:

[0067] A physical layer, wherein the decision conditions corresponding to the physical layer include the physical trust of the Internet of Things;

[0068] A device layer, wherein the decision conditions corresponding to the device layer include the traffic information entropy of the Internet of Things and the average memory consumption of the device;

[0069] A simulation layer, wherein the decision condition corresponding to the simulation layer includes an information entropy difference of the Internet of Things;

[0070] A resource layer, wherein the decision conditions corresponding to the resource layer include the remaining physical resources of the Internet of Things;

[0071] Link layer, the decision condition corresponding to the link layer includes the number of attack paths of the Internet of Things.

[0072] Optionally, the hierarchical decision model can be divided into a slice layer and the following N layers: any N items in the physical layer, device layer, simulation layer, resource layer, and link layer, where N is greater than 1 and N is less than or equal to 5. Each layer in the N layers will generate corresponding decision conditions, and finally generate a comprehensive strategy at the slice layer to achieve network slice resource management and generate network slices.

[0073] Optionally, the hierarchical decision model can be divided into a physical layer, a device layer, an analog layer, a resource layer, a link layer, and a slice layer.

[0074] Figure 2 It is a structural diagram of a hierarchical decision model based on network slicing technology under the industrial Internet of Things provided by the present invention. Figure 2 For example, Figure 2 The hierarchical decision model shown can be divided into the physical layer, device layer, simulation layer, resource layer, link layer and slice layer. The first five layers will generate their own decision conditions and generate the final strategy in the slice layer to achieve secure management of network slices.

[0075] Optionally, the physical layer of the hierarchical decision model can include three basic enabling technologies of the Internet of Things: Radio Frequency Identification (RFID), Wireless Sensor Networks (WSN), and Machine to Machine (M2M). This layer can generate decision conditions: physical trust.

[0076] The value of physical trust can be determined by monitoring the frequency band range of RFID, that is, the bandwidth usage ratio; monitoring the hardware resource usage of WSN, that is, the memory usage ratio; and monitoring the communication status of M2M, that is, the namespace usage ratio.

[0077] Among them, the bandwidth usage ratio of RFID is negatively correlated with the physical trust Q. When the ratio is large, it is considered that there are more users in the network and the security risk is higher.

[0078] Among them, the memory usage ratio of WSN is negatively correlated with the physical trust Q. When the ratio is large, it is considered that the workload is large and the security risk is high.

[0079] Among them, the usage ratio of M2M namespace is negatively correlated with the physical trust Q. When the ratio is large, it is considered that there are more users in the network and the security risk is higher.

[0080] Physical trust can be defined as:

[0081] Q=e -(RFID×WSN×M2M) , where RFID in the formula refers to the bandwidth usage ratio of RFID, the memory usage ratio of WSN to WSN in the formula, and M2M in the formula refers to the namespace usage ratio of M2M;

[0082] This value Q is used as the decision condition ρ generated by the physical layer physical ,Right now:

[0083] ρ physical =Q;

[0084] Optionally, the value of the decision condition corresponding to the physical layer may be calculated based on the decision condition generated by the physical layer;

[0085] Optionally, the device layer of the hierarchical decision model may include all IoT devices within the coverage of the Industrial Internet of Things (IIoT), such as IIoT, and an edge decision maker. This layer may generate decision conditions: traffic information entropy and average device memory consumption.

[0086] The edge decision maker may refer to a decision maker including filtering and sampling functions, and may filter and sample data packets generated by the device layer.

[0087] Among them, the filtering function can include blacklist-based filtering, that is, if the source IP address of the incoming data packet matches an item in the blacklist, then the corresponding data packet can be blocked or processed according to predefined security rules. It also includes whitelist-based filtering, that is, whitelisted data packets can directly enter the target network. The whitelist can be modified by the security administrator. The sampling function can refer to sampling based on traffic. First, the data packets are classified into traffic, and then the sampling process is applied to the entire traffic according to the sampling rate indicated by the decision system.

[0088] Optionally, the traffic f generated by the edge decision maker can be n Calculate its information entropy to measure the amount of information generated by the device. For data packets X1, X2…X N , modeled as a discrete information source:

[0089] {{X1,X2…X N}};

[0090] For each data packet, it has a set of attributes (source IP address, destination IP address, port number, packet length), which is modeled as a set of random events:

[0091] Ω={ω 1, ω2,…,ω n}, where ω1 is a random event, …, ω n is one of the random events (source IP address, destination IP address, port number, packet length);

[0092] Then for each packet, attribute i occurs n times i Second, its information entropy can be expressed as:

[0093] Where k is the number of packets in the incoming network traffic;

[0094] For traffic, the overall information entropy can be expressed as:

[0095]

[0096] The device layer can monitor the average memory consumption of the device, which can be expressed as:

[0097] Among them, ram is the memory consumption of each device, and N is the number of devices;

[0098] The decision conditions that will be generated by the device layer are:

[0099] ρ equipment =[H,RAM], where RAM is the internal average consumption and H is the information entropy of the traffic;

[0100] Based on the decision condition of the device layer, the value of the decision condition corresponding to the device layer can be calculated;

[0101] Optionally, the simulation layer of the hierarchical decision model may include a QoS collector and a traffic predictor based on a deep neural network. This layer may generate decision conditions: information entropy difference;

[0102] The QoS collector may refer to extracting the QoS requirements from the traffic transmitted at the device layer and passing them to the resource layer.

[0103] The traffic predictor may refer to generating a traffic matrix TM according to the traffic transmitted at the device layer, which reflects all possible traffic flowing between nodes in the IIoT network within a specific time:

[0104] where h ij Represents every node in the Internet of Things;

[0105] The traffic matrix is ​​passed into the neural network for traffic prediction. The predicted result is the traffic matrix TM′, which represents all possible traffic flows between nodes in the IIoT network at a specific time in the future. The information entropy of the predicted traffic can be calculated according to the prediction results and the device layer information entropy calculation method:

[0106] Where TM′ represents the predicted traffic matrix, h[] represents the information entropy operator, and K represents the traffic matrix corresponding to time K in a time interval;

[0107] The information entropy of the device layer is then compared with that of the simulation layer, and the activation function Active is used to measure whether the difference is within the allowable error range. The activation function is defined as follows, where μ will be continuously resized according to the subsequent feedback from the slice layer:

[0108] Where x represents the incoming value to be activated, that is, the information entropy difference, and μ represents the threshold of the activation function;

[0109] This difference can be used as a decision condition for the simulation layer:

[0110] ρ simulate =Active(ΔH); where ΔH represents the information entropy difference;

[0111] Based on the decision condition of the simulation layer, the value of the decision condition corresponding to the simulation layer can be calculated;

[0112] Optionally, the simulation layer of the present invention can process the traffic matrix through a neural network algorithm to achieve traffic prediction.

[0113] Optionally, the neural network structure in the present invention is a neural network composed of multiple long short-term memory artificial neural network (LSTM) nodes.

[0114] Optionally, the resource layer of the hierarchical decision model may include a resource manager. This layer may generate decision conditions: remaining physical resources;

[0115] Optionally, the resource manager can allocate corresponding physical resources according to the quality of service requirements transmitted by the simulation layer. At the same time, the resource manager will monitor the remaining resources and make an evaluation.

[0116] The evaluation results can be used as decision conditions at the resource level:

[0117] ρ resource =δ resource ; where δ resource Represents the remaining resources;

[0118] Based on the decision condition of the resource layer, the value of the decision condition corresponding to the resource layer can be calculated;

[0119] The link layer of the hierarchical decision model may include a link builder and an attack graph decider. This layer may generate a decision condition: the number of attack paths.

[0120] Optionally, the link builder may refer to a link layer that builds a network topology map according to the allocation of the resource layer and the requirements of the physical devices, thereby forming an SFC to provide network services.

[0121] The attack graph decider may refer to generating a corresponding attack graph according to a network topology graph and analyzing possible attack paths. The attack graph may refer to the relationship between various vulnerabilities that can be exploited by an attacker and the permissions that the attacker can obtain. Different attack graphs may be generated according to the conditions of the nodes and edges in the topology graph. Generally, in an attack graph, a node represents a state (e.g., a host, a device, a permission, and a vulnerability), and a directed edge represents a state transition (e.g., node A to node B). The attack path may refer to a complete feasible path formed by combining directed edges in the attack graph. The number of possible attack paths will be obtained through the attack graph.

[0122] This structure serves as a decision condition for the link layer:

[0123] ρ link =G(H,L); where H represents the host node, L represents the path, and G represents the corresponding attack graph attack path operator;

[0124] Based on the decision condition of the link layer, the value of the decision condition corresponding to the link layer can be calculated.

[0125] The link layer in the present invention can generate an attack graph through a generated topology graph. The attack graph can be modeled as a directed cyclic graph. For example, for the generated attack graph, it can be modeled as a directed cyclic graph. After determining the starting point and the end point, all possible paths can be obtained. Two stacks can be prepared, respectively called the main stack and the auxiliary stack. Each element of the main stack is a single node for storing nodes on the current path. Each element of the auxiliary stack stores a list of adjacent nodes of the corresponding element of the main stack, and its length is consistent with that of the main stack.

[0126] You can build a stack first: put the starting point into the main stack, and put the adjacent list corresponding to the starting point into the auxiliary stack.

[0127] Then take out the first element of the node list in the auxiliary stack, push it into the main stack, and push the remaining node lists back to the auxiliary stack. For the newly pushed element into the main stack, query its adjacent node list and push it into the auxiliary stack. If some elements in the list have been pushed into the main stack, delete them from the node list and push them back.

[0128] This stacking method is used until all nodes are pushed into the main stack, at which point the top of the auxiliary stack is an empty list.

[0129] You can start trimming the stack, that is, popping the top of the main stack and the empty list at the top of the auxiliary stack.

[0130] Then, if the top of the auxiliary stack is an empty list, continue to cut the stack; if it is a non-empty list, continue to build the stack.

[0131] Repeat the above steps until the top of the main stack is the target vertex, and predict the attack path based on this principle.

[0132] Optionally, the adopting the comprehensive decision maker to perform comprehensive decision determination on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a determination result includes:

[0133] Using the comprehensive decision maker, constructing a decision tree algorithm based on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer;

[0134] Through the decision tree algorithm, a comprehensive decision is made on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a decision result.

[0135] Optionally, the comprehensive decision maker may refer to combining the decision conditions generated by the physical layer, device layer, simulation layer, resource layer and link layer to make a final decision. The comprehensive decision maker uses a decision tree algorithm to combine many decision conditions to obtain a final strategy. Generate a corresponding slice management solution and use a slice generator to generate service slices and / or honeypot slices.

[0136] Optionally, the adopting the comprehensive decision maker to construct a decision tree algorithm based on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer includes:

[0137] Using the comprehensive decision maker, based on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer, determine the decision order for the decision condition layers in the hierarchical decision model;

[0138] A decision tree algorithm is constructed based on the decision sequence of the decision condition layer in the hierarchical decision model.

[0139] Specifically, when constructing a decision tree algorithm, the corresponding information entropy gain can be calculated based on the decision conditions generated by the physical layer, device layer, simulation layer, resource layer, and link layer. The decision conditions of each layer can be regarded as the division of network categories.

[0140] For example, the decision tree divides traffic into positive and negative. We can first use some sample data (positive and negative traffic, and their decision features) to calculate the information entropy of the tuple D, which is divided into negative and positive, corresponding to the probability pi, and bring it into the entropy formula of the training tuple D:

[0141] Where pi represents the probability corresponding to the partition interval, and D represents the sample tuple;

[0142] According to the decision conditions, each decision condition can be regarded as an attribute A, and then the probability of division according to attribute A can be calculated by dividing the number of elements belonging to this category by the total number of training tuple elements. Then the information entropy divided by A is:

[0143] Where D j Represents the number of intervals divided by different attributes under each sample;

[0144] That is, the expression of the expected information of A partitioning D;

[0145] Then the information gain is:

[0146] gain(A)=info(D)-info A (D); where D represents the sample tuple, info represents the information entropy operator, and gain represents the information gain;

[0147] By calculating the gain rate of each attribute and selecting the attribute with the largest gain for splitting, a decision tree can be generated until all decision attributes are constructed.

[0148] For example, the principle of generating a decision tree can be described as follows: first, the information entropy of a sample tuple D can be calculated, and then the information gain of each decision condition is calculated. The one with the largest information gain is selected to divide the decision tree. Then, for the generated branch nodes, the information gain corresponding to the remaining decision conditions is calculated again, and the above process is repeated until all decision conditions are used.

[0149] In this way, the comprehensive decision maker can generate a decision tree for comprehensive judgment.

[0150] Optionally, performing comprehensive decision determination on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer by the decision tree algorithm to obtain a determination result includes:

[0151] Adopting the decision tree algorithm, based on the decision sequence, sequentially making decisions on the values ​​of the decision conditions corresponding to the decision condition layers;

[0152] In the case where it is determined that the value of the decision condition corresponding to the decision condition layer is not within a preset range, the determination result is determined to have a security risk (which may be referred to as a "positive" result);

[0153] When it is determined that the value of the decision condition corresponding to the non-existent decision condition layer is not within a preset range, the determination result is determined to be without security risk (which may be referred to as a "negative" result).

[0154] Figure 3 is an example diagram of a decision tree for the slice layer in the hierarchical model provided by the present invention, such as Figure 3 As shown, based on the decision order, the values ​​of the decision conditions corresponding to the decision condition layers can be determined layer by layer in turn. When the values ​​of the decision conditions corresponding to the decision condition layers of the current layer are within a preset range, the values ​​of the decision conditions corresponding to the decision condition layers of the next layer are determined, until it is determined that the values ​​of the decision conditions corresponding to the decision condition layers are not within the preset range. In this case, it can be determined that the determination result has a security risk, and a result with a security risk can be directly output.

[0155] Optionally, the generating a network slice based on the slice generator and the determination result includes:

[0156] If it is determined that the determination result has a security risk, a honeypot slice and a service slice are generated, wherein the honeypot slice is used to attract network attacks;

[0157] If it is determined that the judgment result does not pose a security risk, only service slices are generated.

[0158] Specifically, for the comprehensive decision-making judgment of the comprehensive decision maker, if there is a security risk, a honeypot slice and a service slice are generated; if there is no security risk, only a service slice is generated.

[0159] Figure 4 is a flow chart of slice generation management of the slice layer in the hierarchical model provided by the present invention, such as Figure 4 As shown in the figure, a decision tree can be constructed based on the value of the decision condition of the decision condition layer and the final strategy can be generated accordingly. If the judgment result is negative, a service slice is directly generated, the network service is completed and the result is fed back; if the judgment result is positive, a honeypot slice is first generated to induce external attacks and obtain the operation result. If the result is a true positive, the service is stopped; if it is a false positive, a service slice is generated and the network service is completed. The result is finally fed back.

[0160] Specifically, network slices can be divided into service slices and honeypot slices. Service slices can refer to slices that provide normal network service functions. Honeypot slices can refer to slices disguised as IoT services to deliberately induce external attacks.

[0161] Optionally, generating a honeypot slice includes:

[0162] VNF is used to simulate various components of the devices in the Internet of Things, all VNFs are connected to form an SFC that simulates the Internet of Things devices, physical resources are allocated to the SFC, and it is deployed on the honeypot slice.

[0163] Specifically, in order to achieve the management of network slices, the requirements of network services should be determined first. Then, with the help of software-defined networking (SDN) and network function virtualization (NFV) technologies, corresponding slices are created according to specific network service requirements. That is, virtual network functions (VNFs) are connected in series according to network requirements to form service function chains (SFCs) to provide network services. Then, according to the changing network service requirements, physical resources are allocated and adjusted, and network functions are added or deleted to reposition network functions.

[0164] Specifically, honeypot slices can use NFV technology to simulate the various components of IoT devices using VNFs. A series of VNFs are connected according to the IoT network environment to form an SFC that simulates IoT devices. Physical resources are then allocated to them and deployed on the honeypot slice. The slice acts as a "fake" IoT device to confuse network attacks. And the running results are generated to determine whether the attack is a true positive or a false positive.

[0165] Optionally, the method further comprises:

[0166] Based on the operation log of the honeypot slice, the parameters of the hierarchical decision model are adjusted.

[0167] Specifically, network traffic can generate running logs on the honeypot slices, and the running logs generated on the honeypot slices can be used as feedback results to dynamically modify the hierarchical model to make the decision results more accurate.

[0168] Optionally, adjusting the parameters of the hierarchical decision model based on the operation log of the honeypot slice includes:

[0169] The operation log of the honeypot slice is sent to the simulation layer, so that the simulation layer adjusts the activation function of the simulation layer based on the operation log of the honeypot slice.

[0170] Optionally, the slicing layer may feed back the operation log of the honeypot slice as a result to the activation function of the simulation layer, so that the simulation layer adjusts the parameters of the activation function of the simulation layer based on the operation log of the honeypot slice to make the decision result more accurate.

[0171] Optionally, a decision tree can be constructed based on the value of the decision condition of the decision condition layer and the final strategy can be generated accordingly. If the judgment result is negative, the service slice is directly generated, the network service is completed and the result is fed back; if the judgment result is positive, a honeypot slice is first generated to induce external attacks and obtain the running result. If the result is a true positive, the service is stopped; if it is a false positive, a service slice is generated and the network service is completed. The final feedback result; wherein, the feedback result can be used to modify the activation function of the simulation layer. The principle of the feedback algorithm can be: if the result generated by the honeypot slice is a true positive, then the μ value of the simulation layer activation function is halved, if the result is a false positive, then μ will be an exponential function of time t and keep growing: μ = e t .

[0172] The security decision-making method for the Internet of Things provided by the present invention adopts a hierarchical decision model, performs comprehensive decision-making judgments on the values ​​of the decision conditions corresponding to the at least one decision condition layer, generates a network slice based on the judgment result, and simulates the dynamic changes of the industrial Internet of Things environment through the values ​​of the decision conditions corresponding to the at least one decision condition layer. On the basis of ensuring network performance, the security risks of the network are reduced to ensure the network security of the Internet of Things.

[0173] The security decision-making device of the Internet of Things provided by the present invention is described below. The security decision-making device of the Internet of Things described below and the security decision-making method of the Internet of Things described above can be referenced to each other.

[0174] Figure 5 : is a schematic diagram of the structure of the security decision-making device of the Internet of Things provided by the present invention, such as Figure 5As shown, the device comprises:

[0175] A first determination module 510 is used to determine the value of the decision condition corresponding to the at least one decision condition layer based on at least one decision condition layer in the hierarchical decision model, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things;

[0176] A comprehensive decision judging module 520 is used to use the comprehensive decision maker to perform comprehensive decision judging on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a judging result;

[0177] The network slice generation module 530 is used to generate a network slice based on the slice generator and the determination result.

[0178] The security decision-making device of the Internet of Things provided by the present invention can implement each process implemented by the above-mentioned method embodiments and achieve the same technical effect. To avoid repetition, it will not be repeated here. The security decision-making device of the Internet of Things provided by the present invention adopts a hierarchical decision model to make a comprehensive decision judgment on the values ​​of the decision conditions corresponding to the at least one decision condition layer, and generates a network slice based on the judgment result. The values ​​of the decision conditions corresponding to the at least one decision condition layer simulate the dynamic changes of the industrial Internet of Things environment, and on the basis of ensuring network performance, reduce the security risks of the network and ensure the network security of the Internet of Things.

[0179] Figure 6 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 6 As shown, the electronic device may include: a processor 610, a communications interface 620, a memory 630 and a communication bus 640, wherein the processor 610, the communications interface 620 and the memory 630 communicate with each other through the communication bus 640. The processor 610 may call the logic instructions in the memory 630 to execute the security decision method of the Internet of Things, and the method includes:

[0180] Based on at least one decision condition layer in a hierarchical decision model, determine the value of the decision condition corresponding to the at least one decision condition layer, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things;

[0181] Using the comprehensive decision maker, a comprehensive decision is made on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a decision result;

[0182] Based on the slice generator and the determination result, a network slice is generated.

[0183] In addition, the logic instructions in the above-mentioned memory 630 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0184] On the other hand, the present invention further provides a computer program product, the computer program product includes a computer program, the computer program can be stored on a non-transitory computer-readable storage medium, when the computer program is executed by a processor, the computer can execute the security decision-making method of the Internet of Things provided by the above methods, the method includes:

[0185] Based on at least one decision condition layer in a hierarchical decision model, determine the value of the decision condition corresponding to the at least one decision condition layer, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things;

[0186] Using the comprehensive decision maker, a comprehensive decision is made on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a decision result;

[0187] Based on the slice generator and the determination result, a network slice is generated.

[0188] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for making a security decision for the Internet of Things provided by the above methods is implemented, and the method comprises:

[0189] Based on at least one decision condition layer in a hierarchical decision model, determine the value of the decision condition corresponding to the at least one decision condition layer, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things;

[0190] Using the comprehensive decision maker, a comprehensive decision is made on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a decision result;

[0191] Based on the slice generator and the determination result, a network slice is generated.

[0192] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0193] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0194] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A security decision-making method for the Internet of Things, characterized in that: include: Based on at least one decision condition layer in a hierarchical decision model, determine the value of the decision condition corresponding to the at least one decision condition layer, wherein the hierarchical decision model includes at least one decision condition layer and a slice layer, the slice layer includes a comprehensive decision maker and a slice generator, and the value of the decision condition corresponding to the at least one decision condition layer is used to characterize the network environment of the Internet of Things; Using the comprehensive decision maker, a comprehensive decision is made on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a decision result; Generate a network slice based on the slice generator and the determination result; The at least one decision condition layer includes at least one of the following: A physical layer, wherein the decision conditions corresponding to the physical layer include the physical trust of the Internet of Things; A device layer, wherein the decision conditions corresponding to the device layer include the traffic information entropy of the Internet of Things and the average memory consumption of the device; A simulation layer, wherein the decision condition corresponding to the simulation layer includes an information entropy difference of the Internet of Things; A resource layer, wherein the decision conditions corresponding to the resource layer include the remaining physical resources of the Internet of Things; A link layer, wherein the decision condition corresponding to the link layer includes the number of attack paths of the Internet of Things; The comprehensive decision maker is used to perform comprehensive decision determination on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer to obtain a determination result, including: Using the comprehensive decision maker, based on the values ​​of the decision conditions respectively corresponding to the at least one decision condition layer, determine the decision order for the decision condition layers in the hierarchical decision model; Constructing a decision tree algorithm based on the decision sequence of the decision condition layer in the hierarchical decision model; Adopting the decision tree algorithm, based on the decision sequence, sequentially making decisions on the values ​​of the decision conditions corresponding to the decision condition layers; In the case where it is determined that the value of the decision condition corresponding to the decision condition layer is not within a preset range, determining that the determination result has a security risk; In the case where it is determined that the value of the decision condition corresponding to the non-existent decision condition layer is not within a preset range, determining that the determination result is that there is no security risk; The generating a network slice based on the slice generator and the determination result includes: If it is determined that the determination result has a security risk, a honeypot slice and a service slice are generated, wherein the honeypot slice is used to attract network attacks; If it is determined that the judgment result does not pose a security risk, only service slices are generated.

2. The security decision-making method of the Internet of Things according to claim 1 is characterized in that: The generating of honeypot slices includes: The virtual network function VNF is used to simulate the various components of the device in the Internet of Things, and all VNFs are connected to form a service function chain SFC that simulates the Internet of Things device. Physical resources are allocated to the SFC and deployed on the honeypot slice.

3. The security decision-making method for the Internet of Things according to claim 1 or 2, characterized in that: The method further comprises: The operation log of the honeypot slice is sent to the simulation layer, so that the simulation layer adjusts the activation function of the simulation layer based on the operation log of the honeypot slice.

4. A security decision-making device for the Internet of Things, characterized in that: The security decision-making device of the Internet of Things is used to implement the security decision-making method of the Internet of Things as claimed in claim 1.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the security decision-making method for the Internet of Things as described in any one of claims 1 to 3 is implemented.

6. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the security decision-making method for the Internet of Things as described in any one of claims 1 to 3 is implemented.

7. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the security decision-making method for the Internet of Things as described in any one of claims 1 to 3 is implemented.