Edge IoT sensor network operation adjustment method and device

Abstract

The present invention is applicable to the field of wireless sensing and edge computing technology, and provides an edge IoT sensor network operation adjustment method and device, the method comprising: obtaining attack information of the edge IoT sensor network; according to the attack information, establishing a first utility function with the goal of maximizing the benefit of the edge IoT sensor network when attacked, and establishing a second utility function with the goal of maximizing the benefit of the attacker, and determining a Bayesian game model based on the first utility function and the second utility function; establishing the constraint conditions of the Bayesian game model and calculating the equilibrium solution of the Bayesian game model, and obtaining the optimal task offloading strategy when the edge IoT sensor network is attacked; and adjusting the operation of the edge IoT sensor network based on the optimal task offloading strategy. The present invention can improve the operational security of the existing edge IoT sensor network.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless sensing and edge computing, and in particular to a method and device for adjusting the operation of an edge IoT sensor network. Background Art

[0002] The existing monitoring methods of key industrial industries and environmental protection enterprises are far from meeting the needs of environmental management, making it difficult to detect and investigate enterprises for illegal discharge. Therefore, wireless sensor network monitoring and transmission methods built for the purpose of environmental protection enterprise emission monitoring came into being.

[0003] With the introduction and development of edge computing technology, wireless sensor networks and edge computing technologies have been further combined to produce edge IoT sensor networks. By relying on the cooperation of edge computing nodes (ECN) and sensor nodes, cloud computing capabilities are placed on the sensor network side, which can realize the collection and calculation of sensor data of environmental protection enterprise equipment on the edge side, ensuring that the status of environmental protection enterprises can be output in real time. However, the existing edge IoT sensor networks do not consider the potential safety hazards of nodes. If a node is attacked by external intrusion, it will lead to the loss of monitoring data, causing misjudgment of emission reduction policies and economic losses, and causing more illegal discharges. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides an edge IoT sensor network operation adjustment method and device to improve the operation security of the existing edge IoT sensor network.

[0005] A first aspect of an embodiment of the present invention provides an edge IoT sensor network operation adjustment method, including:

[0006] Obtain attack information of edge IoT sensor networks;

[0007] According to the attack information, a first utility function is established with the goal of maximizing the benefit of the edge IoT sensor network when it is attacked, and a second utility function is established with the goal of maximizing the benefit of the attacker, and a Bayesian game model is determined based on the first utility function and the second utility function;

[0008] Establish the constraints of the Bayesian game model and calculate the equilibrium solution of the Bayesian game model to obtain the optimal task offloading strategy when the edge IoT sensor network is attacked;

[0009] Adjust the operation of edge IoT sensor networks based on the optimal task offloading strategy.

[0010] A second aspect of an embodiment of the present invention provides an edge IoT sensor network operation adjustment device, including:

[0011] An acquisition module is used to obtain attack information of the edge IoT sensor network;

[0012] A determination module is used to establish a first utility function based on the attack information with the goal of maximizing the benefit of the edge IoT sensor network when it is attacked, and to establish a second utility function with the goal of maximizing the benefit of the attacker, and to determine a Bayesian game model based on the first utility function and the second utility function;

[0013] The solution module is used to establish the constraints of the Bayesian game model and calculate the equilibrium solution of the Bayesian game model to obtain the optimal task offloading strategy when the edge IoT sensor network is attacked;

[0014] The adjustment module is used to adjust the operation of the edge IoT sensor network based on the optimal task offloading strategy.

[0015] A third aspect of an embodiment of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned edge IoT sensor network operation adjustment method are implemented.

[0016] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the edge IoT sensor network operation adjustment method are implemented as described above.

[0017] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0018] The embodiment of the present invention can dynamically adjust the operation strategy of the edge IoT sensor network according to external attacks. That is, when the edge IoT sensor network is attacked by a network, a Bayesian game model is established to calculate and adjust the task offloading strategy of the network to ensure the collection, transmission, calculation and processing of most monitoring data, avoid data loss causing monitoring loss, and effectively improve the operation security of the edge IoT sensor network. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0020] Figure 1 It is a flow chart of an edge IoT sensor network operation adjustment method provided by an embodiment of the present invention;

[0021] Figure 2 It is a structural diagram of an edge IoT sensor network operation adjustment device provided by an embodiment of the present invention;

[0022] Figure 3 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0024] In order to illustrate the technical solution of the present invention, a specific embodiment is provided below for illustration.

[0025] Manual monitoring methods are difficult to reflect the environmental status quo such as environmental capacity, ecological status, ecological deterioration rate and structure of pollution sources, resulting in a disconnect between the ability of environmental monitoring and supervision and the demand for environmental governance, which is far from meeting the monitoring needs of key industrial industries and environmental protection enterprises. In addition, the lack of real-time monitoring means also makes it difficult to discover and investigate enterprises' illegal discharges, and it is difficult to perceive changes in the regional environment in a timely manner, and it is impossible to timely predict or discover major environmental accident hazards. Therefore, wireless sensor network monitoring and transmission methods built for the purpose of emission monitoring of environmental protection enterprises have emerged. Wireless sensor networks monitor the environment by collecting data. In today's Internet of Things era, they are widely used in agricultural and forestry environmental monitoring, military intelligence detection, urban structural safety monitoring, and health care. Among them, D2D (Device to Device) wireless sensor networks integrate the advantages and characteristics of wireless power transmission technology and AD-hoc networks, which can extend the service life of wireless sensor networks and maintain connectivity, ensure the realization of full-time environmental protection enterprise equipment operation status monitoring and data transmission, and prevent environmental protection enterprises from taking advantage of leaks to discharge illegally. With the proposal and development of edge computing technology, D2D wireless sensor networks are further combined with edge computing technology to produce edge IoT sensor networks. By relying on the cooperation between edge computing nodes and sensor nodes, and placing cloud computing capabilities on the sensor network side, it is possible to collect and calculate sensor data from environmental protection enterprise equipment on the edge side, ensure that the status of environmental protection enterprises can be output in real time, and cooperate with technologies such as knowledge graphs to achieve real-time portraits of the environmental protection status of enterprises, so as to avoid the monitoring data being unable to be transmitted and processed for too long, resulting in information losing timeliness and hindering the formulation of effective carbon reduction policies. For example, the existing technical document "Research on Computational Offloading Methods for Marine Observation and Monitoring Sensor Networks" confirms the feasibility of integrating wireless sensor networks and edge computing in marine environmental monitoring scenarios.

[0026] However, edge IoT sensor networks are facing new difficulties. Existing studies have not considered the potential security risks of nodes. If a node is attacked by external intrusion, it will lead to the loss of monitoring data, causing misjudgment of emission reduction policies and economic losses, and even causing more illegal emissions.

[0027] In view of this, an embodiment of the present invention provides an edge IoT sensor network operation adjustment method, see Figure 1 As shown, the method includes:

[0028] Step S101, acquiring attack information of the edge IoT sensor network.

[0029] In an embodiment of the present invention, when the edge IoT sensor network is not attacked, it runs with the task offloading strategy that maximizes the network benefit. Once the network is attacked by an attacker, the attack information, including the attack type and attack probability, can be analyzed based on historical data. In this embodiment, the task offloading strategy includes the resource allocation of edge computing nodes, which edge computing nodes the sensor node specifically offloads tasks to, and the task offloading time. The attack information considers two types of attacks: distributed denial of service attacks (DDOS) and privacy eavesdropping attacks. It should be noted that when a distributed denial of service attack is launched, the edge computing node will be directly paralyzed and unable to communicate. The emission equipment sensors of environmental protection enterprises can only select other edge computing nodes to complete communication and task offloading, which will cause the energy consumption of sensor equipment to increase; when a privacy eavesdropping attack is launched, the edge computing node maintains communication capabilities, but the communication and computing resources will be hijacked, causing the transmitting antenna power and CPU chip to operate across the board, resulting in an increase in the energy consumption of the edge computing node.

[0030] Step S102, based on the attack information, a first utility function is established with the goal of maximizing the benefit of the edge IoT sensor network when it is attacked, and a second utility function is established with the goal of maximizing the benefit of the attacker, and a Bayesian game model is determined based on the first utility function and the second utility function.

[0031] In the embodiment of the present invention, a Bayesian game model is established to maximize the benefits of both parties based on the attack information. The edge IoT sensor network central manager in the environmental monitoring sensor communication scenario is the leader of the master-slave game, and the attacker in the dark will determine his attack strategy based on the response benefits that the defender may obtain when he is not attacked, and is the follower of the game.

[0032] Step S103, establishing the constraint conditions of the Bayesian game model and calculating the equilibrium solution of the Bayesian game model to obtain the optimal task offloading strategy when the edge IoT sensor network is attacked.

[0033] Step S104: adjusting the operation of the edge IoT sensor network based on the optimal task offloading strategy.

[0034] In an embodiment of the present invention, by establishing constraint conditions, the edge IoT sensor network continuously adjusts the task offloading strategy according to the attack strategy within the constraint to maximize its own benefit, and the attacker continuously adjusts the attack strategy according to the task offloading strategy of the edge IoT sensor network to maximize its own benefit, and finally obtains a balanced solution for both parties, thereby obtaining the optimal task offloading strategy for operation adjustment.

[0035] It can be seen that the embodiments of the present invention can dynamically adjust the operation strategy of the edge IoT sensor network according to external attacks. That is, when the edge IoT sensor network is attacked by a network, the task offloading strategy of the network is calculated and adjusted through the established Bayesian game model to ensure the collection, transmission, calculation and processing of most monitoring data, avoid data loss and monitoring loss, and effectively improve the operation security of the edge IoT sensor network.

[0036] Optionally, the attack information includes the type and probability of attacks launched by the attacker to the edge IoT sensor network, and the attack types include distributed denial of service attacks and privacy eavesdropping attacks.

[0037] The first utility function is:

[0038]

[0039] Where U(x,f) is the revenue of the edge IoT sensor network when it is attacked; p1+p2=1, p1 is the probability of distributed denial of service attack, p2 is the probability of privacy eavesdropping attack; K is the number of edge computing nodes in the edge IoT sensor network; U k,1 (x,f) is the benefit of the edge IoT sensor network when it is attacked by a distributed denial of service; U k,2 (x,f) is the gain of the edge IoT sensor network when it is attacked by privacy eavesdropping.

[0040] Optional, U k,1 (x,f) and U k,2 The calculation formula for (x,f) is:

[0041]

[0042]

[0043] In the formula, a k ={0,1} represents the attack type, where 0 represents a distributed denial of service attack and 1 represents a privacy eavesdropping attack; b k ={0,1} represents the attack decision, where 0 means not to launch an attack and 1 means to launch an attack; I is the number of sensor nodes in the edge IoT sensor network; rik r′ is the wireless communication rate from sensor node i to edge computing node k before attack; ik is the wireless transmission rate of sensor node i offloading tasks to edge computing node k after being attacked and making task decisions; ik is the resource usage of sensor node i offloading tasks to edge computing node k; is the communication delay between sensor node i and edge computing node k during task offloading; is the computing delay required for edge computing node k to process tasks.

[0044] Optionally, the second utility function is:

[0045]

[0046] Where U a (a, b) is the attacker’s profit; L is the number of edge computing nodes attacked by the attacker and L <K; is the probability of an attacker attacking edge computing node k; C k is the electricity cost of edge computing node k; μ k is the attack cost required for an attacker to attack edge computing node k; γ and θ are market risk control coefficients.

[0047] Optionally, constraints include:

[0048] The attacker's maximum number of attack nodes and maximum number of attacks

[0049]

[0050] Offloading strategy constraints for sensor nodes

[0051]

[0052] Computing power constraints of edge computing nodes

[0053]

[0054] Energy Consumption Constraints of Sensor Nodes

[0055]

[0056] Where A is the maximum attack limit; i is the total number of tasks that need to be offloaded from sensor node i; F k is the maximum allowed number of tasks for edge computing node k; f ik is the computing capacity occupied by edge computing node k by sensor node i, f max is the calculation capacity limit; is the communication energy consumption between sensor node i and edge computing node k during task offloading, E imax is the communication energy consumption limit.

[0057] In an embodiment of the present invention, in order to construct the first utility function and its constraints, it is necessary to construct a lightweight monitoring and computing model of the edge IoT sensor network in advance, specifically including a wireless sensor network communication model, an edge computing node data calculation model, a sensor node monitoring data processing and a node energy consumption model.

[0058] The wireless sensor network communication model is:

[0059] Assume that the number set of edge computing nodes in the network is recorded as K@{1,2,...,k}, the edge computing nodes are mainly deployed around the emission equipment of environmental protection enterprises, the communication ranges between edge computing nodes overlap, and all emission equipment information within the communication range is monitored and transmitted. The set of sensor nodes attached to the emission equipment of environmental protection enterprises in the communication scenario is recorded as I@{1,2,...,i}, then the edge IoT sensor network communication model is constructed (that is, the wireless communication rate r in the process of sensor node i offloading tasks to edge computing node k) ik )

[0060]

[0061]

[0062] In the formula, g ik is the wireless channel gain, which is related to the communication distance between the sensor node i and the edge computing node k and decays as the distance between them increases; is the antenna transmission power allocated to the wireless transmission communication of the edge computing node k; α is the path loss index; d0 is the optimal communication distance of the edge computing node; d ik is the communication distance between sensor node i and edge computing node k; A is the product of Rayleigh fading coefficient and path loss constant; σ 2 is the noise power; ζ is the average inter-cell interference power; w is the allocated spectrum.

[0063] The edge computing node data calculation model is:

[0064] The limited computing resources owned by edge computing node k are denoted as f k , represented by the CPU operating frequency of the edge computing node, in GHz. Sensor node i has λ i The tasks need to be offloaded to the edge computing node to complete data analysis and other calculations. The offloading decision of sensor node i is x i ={x i1 ,x i2 ,...,x iK, where the offloading decision needs to satisfy the offloading data conservation and the computing resource limitation of the edge computing node

[0065]

[0066]

[0067] The monitoring data processing of the sensor node and the node energy consumption model are as follows:

[0068] The data attributes of the demand response tasks of the sensor nodes are represented by the data size s (unit: mb) and the computing resources h (unit: CPU cycles) required for data processing. Without loss of generality, assume that all tasks of each sensor node have the same data attributes. Then, during the task offloading process, the communication delay and energy consumption between the sensor node i and the edge computing node k are expressed as

[0069]

[0070]

[0071] In the formula, γ is the transmission energy consumption coefficient.

[0072] After the monitoring data of the sensor node arrives at the edge computing node, the edge computing node needs to perform processing such as data cleaning. The computing delay and energy consumption required for the edge computing node k to process the task are expressed as:

[0073]

[0074]

[0075] In the formula, κ represents the computing energy consumption coefficient, and κ > 0.

[0076] In the embodiment of the present invention, in order to construct the second utility function and its constraints, it is necessary to construct an attacker decision model and an attack probability model in advance.

[0077] The attacker decision model is as follows:

[0078] Assume that the attacker has limited attack capabilities and can only select L edge computing nodes as attack targets, where L < k. Let a k = {0, 1} represent the attack type launched by the attacker against the edge computing node k, where 0 represents a distributed denial-of-service attack and 1 represents a privacy eavesdropping attack. Denote the probability that the attacker selects the edge computing node k as an attack target as This probability is related to the protection intensity of defense means such as firewalls pre-placed on the edge computing node, and denote the attack decision set of the attacker as When the attack edge computing node k is selected, the value is 1. If the attacker does not plan to launch an attack, the value is 0. The number of attackable nodes and attack types that can be launched must meet

[0079]

[0080]

[0081] Among them, A represents the total number of times the two attacks can be launched.

[0082] The attack probability model is:

[0083] Different from the openness of the information policy of the edge IoT sensor network, the attacker is hidden in the dark, and the network defender cannot know the attacker's information. The attacker can only obtain prior knowledge of the attacker through historical attack reports and system operation reports. Therefore, the cognitive probability of the attack type that the attacker may launch is expressed as {p1, p2}, which represents the distributed denial of service attack and privacy eavesdropping attack type respectively, which must satisfy

[0084] p1+p2=1

[0085] Based on the above model and attack information, the first utility function and its constraints, the second utility function and its constraints can be determined, and then the Bayesian game model can be obtained, which is transformed into finding the optimal task offloading strategy for the defender and the optimal attack strategy for the attacker in this network scenario, and calculating the equilibrium solution.

[0086] Optionally, compute the equilibrium solution of the Bayesian game model, including:

[0087] Obtain the maximum benefit of the edge IoT sensor network when there is no attack, obtain the initial upper bound value, and calculate the maximum benefit of the attacker under the task offloading strategy corresponding to the initial upper bound value to obtain the initial lower bound value;

[0088] Based on the branch and bound method, the upper and lower bound values ​​are iteratively updated until the difference between the upper and lower bound values ​​is less than the preset threshold. The task offloading strategy corresponding to the final upper bound value and the attack strategy corresponding to the final lower bound value are the equilibrium solutions of the Bayesian game model; among them, the task offloading strategy corresponding to the final upper bound value is the optimal task offloading strategy when the edge IoT sensor network is attacked.

[0089] Optionally, based on the attack strategy corresponding to the initial lower bound value and the branch and bound method, iteratively updating the upper bound value and the lower bound value until the difference between the upper bound value and the lower bound value is less than a preset threshold value, including:

[0090] Step 1: branch the attack strategy corresponding to the initial lower bound value;

[0091] Step 2: For any branch, solve the maximum benefit of the edge IoT sensor network under the branch attack strategy, and update the upper bound value according to the maximum benefit, and determine the maximum benefit of the attacker according to the task offloading strategy corresponding to the maximum benefit, and update the lower bound value according to the maximum benefit of the attacker;

[0092] Step 3: If the latest lower bound value is smaller than the upper bound value and larger than the original lower bound value, the attack strategy corresponding to the latest lower bound value is used as the attack strategy of the branch;

[0093] Step 4: Repeat steps 2 and 3 until the difference between the lower limit value and the upper limit value is less than a preset threshold.

[0094] In the embodiment of the present invention, the problem is solved based on the generalized Benders decomposition method, and the solution process includes designing heuristic search rules. Based on the fixed attack decision, the task offloading and collection information transmission sub-problems are solved. Based on the fixed task offloading strategy, the attack strategy is solved until the leader utility function converges.

[0095] Designing heuristic search rules includes: for an attack decision with attack type a and satisfying the following formula

[0096]

[0097] If it does not satisfy (a·b) T · χ a <μ, then it is definitely not a Stackelberg equilibrium solution.

[0098] Among them, μ represents the attacker's attack cost, and χ represents the mathematical expectation of the attack benefit of a certain attack type, which is determined by the selected attack type and is recorded as

[0099] (corresponding to a=0).

[0100] or (corresponding to a=1).

[0101] In order to improve the efficiency of algorithm execution and reduce the time of searching for the optimal solution, the proposed heuristic rule stipulates that the attacker uses the attack utility value in the absence of an attack as the basis to find the attack decision with the best benefit. If the preset defense measures of the edge computing node are good, or the attack value is low, then the attacker will not choose to attack the edge computing node and the corresponding device group. The heuristic rule allows the algorithm to first eliminate the attack decisions that will not necessarily meet the Stackelberg equilibrium solution based on the attack probability and the maximum obtainable benefit before execution, so that the algorithm reduces the number of branches required in the branch solution process. After reducing invalid attack strategies based on the heuristic rule, the problem is solved.

[0102] Furthermore, the problem solving process includes:

[0103] Solve the optimal task offloading and resource allocation strategy in the absence of attacks. First, solve the optimal task offloading decision when there is no attacker launching a network attack, and record it as the initial upper bound of the original problem. Then, initialize all possible attack decisions of an attacker of a certain attack type corresponding to each MIP problem according to the given upper bound decision variable, and solve to obtain other attack decisions. The maximum attack utility among all decisions is recorded as the lower bound value of the original problem. Update the attack decision based on the optimal task offloading and resource allocation strategy without attacks, and search for the optimal task offloading strategy under fixed attack decisions based on the branch and bound method.

[0104] Based on the current fixed optimal attack decision, the idea of ​​introducing the branch and bound method is to solve the optimal leader defense decision corresponding to the branch, and generate the possible attack decision for the second attack of this attack type, and continue to solve the lower bound of the problem and compare. If the newly obtained lower bound value is greater than the upper bound, the branch is removed according to the heuristic rule. On the contrary, if the new lower bound value is greater than the original lower bound value but less than the upper bound value, the above operation is repeated based on the attack branch until the upper and lower bounds are close to a very small real value. The problem solution is considered to converge, and the strong Stackelberg equilibrium solution of the Bayesian game is obtained.

[0105] Furthermore, the computing resource allocation subproblem of the fixed attack strategy and the fixed task offloading strategy and the task strategy main problem of the fixed attack strategy and the fixed computing resource allocation obtained based on the generalized Benders decomposition are as follows:

[0106]

[0107] constraint:

[0108]

[0109]

[0110]

[0111]

[0112]

[0113] x∈{0,1}

[0114] Among them, z is the introduced auxiliary variable, Γ and Γ′ are the optimal solution cut set and the suboptimal solution cut set respectively, and the Lagrangian auxiliary equation is generated based on the Lagrangian method KKT condition, where ω1 and ω2 are the Lagrangian multipliers of the optimal solution and the suboptimal solution respectively. In each branch, the computing resource allocation subproblem based on the generalized Benders decomposition is obviously a convex function, and the subproblem can be solved by traditional methods such as the interior point method; and the task offloading main problem is relaxed into a MIP problem in the range of [0,1], and the corresponding non-relaxed optimal solution can be obtained by directly solving and relying on various reduction methods, thereby reducing the complexity of solving the MIP problem.

[0115] First, the task offloading decision is initialized according to the attack decision of the current branch, denoted as x (0) , solve the subproblem based on the given attack and unloading decisions and obtain the optimal resource allocation solution, and record the current utility value as the initial upper bound Then enter the iteration and solve the relaxed main problem to obtain the task offloading decision x for the current qth iteration (q) And bring it into the subproblem. If the solution of the subproblem satisfies the constraint, update the upper bound value And the new resource allocation decision f (q) Substitute and solve the constrained main problem; otherwise, introduce a non-negative auxiliary value ε to relax the constraint to G≤ε, and solve to obtain the suboptimal solution. Substitute the resource allocation decision obtained based on the above process into the main problem with constraints, and obtain the lower bound value of the current qth iteration By continuously iterating until convergence, the problem of the current branch is solved and pruning and a new round of branch solving are started, thereby realizing the search for a strong Stackelberg equilibrium.

[0116] The advantages of the full-time environmental protection data edge monitoring and transmission method taking data security into account in the embodiment of the present invention are: (1) considering the data collection and transmission security risks caused by the lack of security capabilities during wireless sensor network data transmission, wireless sensor network collaborative computing and mitigation defense measures are designed according to the characteristics of wireless DDOS attacks and privacy attacks, so as to improve the network communication security when building an edge sensor Internet of Things network for environmental protection enterprise monitoring; (2) considering the impact of network attacks on the communication performance of sensor nodes, it ensures that sensor nodes can complete data monitoring and transmission within limited energy consumption, thereby improving network reliability.

[0117] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0118] The embodiment of the present invention provides an edge IoT sensor network operation adjustment device, see Figure 2 As shown, the device 20 comprises:

[0119] The acquisition module 21 is used to obtain attack information of the edge IoT sensor network.

[0120] The determination module 22 is used to establish a first utility function based on the attack information, with the goal of maximizing the benefit of the edge IoT sensor network when it is attacked, and to establish a second utility function with the goal of maximizing the benefit of the attacker, and to determine the Bayesian game model based on the first utility function and the second utility function.

[0121] The solution module 23 is used to establish the constraint conditions of the Bayesian game model and calculate the equilibrium solution of the Bayesian game model to obtain the optimal task offloading strategy when the edge IoT sensor network is attacked.

[0122] The adjustment module 24 is used to adjust the operation of the edge IoT sensor network based on the optimal task offloading strategy.

[0123] Optionally, the attack information includes the type and probability of attacks launched by the attacker to the edge IoT sensor network, and the attack types include distributed denial of service attacks and privacy eavesdropping attacks.

[0124] The first utility function is:

[0125]

[0126] Where U(x,f) is the revenue of the edge IoT sensor network when it is attacked; p1+p2=1, p1 is the probability of distributed denial of service attack, p2 is the probability of privacy eavesdropping attack; K is the number of edge computing nodes in the edge IoT sensor network; U k,1 (x,f) is the benefit of the edge IoT sensor network when it is attacked by a distributed denial of service; U k,2 (x,f) is the gain of the edge IoT sensor network when it is attacked by privacy eavesdropping.

[0127] Optional, U k,1 (x,f) and U k,2 The calculation formula for (x,f) is:

[0128]

[0129]

[0130] In the formula, a k ={0,1} represents the attack type, where 0 represents a distributed denial of service attack and 1 represents a privacy eavesdropping attack; b k ={0,1} represents the attack decision, where 0 means not to launch an attack and 1 means to launch an attack; I is the number of sensor nodes in the edge IoT sensor network; r ikr′ is the wireless communication rate from sensor node i to edge computing node k before attack; ik is the wireless transmission rate of sensor node i offloading tasks to edge computing node k after being attacked and making task decisions; ik is the resource usage of sensor node i offloading tasks to edge computing node k; is the communication delay between sensor node i and edge computing node k during task offloading; is the computing delay required for edge computing node k to process tasks.

[0131] Optionally, the second utility function is:

[0132]

[0133] Where U a (a, b) is the attacker’s profit; L is the number of edge computing nodes attacked by the attacker and L <K; is the probability of an attacker attacking edge computing node k; C k is the electricity cost of edge computing node k; μ k is the attack cost required for an attacker to attack edge computing node k; γ and θ are market risk control coefficients.

[0134] Optionally, constraints include:

[0135] The attacker's maximum number of attack nodes and maximum number of attacks

[0136]

[0137] Offloading strategy constraints for sensor nodes

[0138]

[0139] Computing power constraints of edge computing nodes

[0140]

[0141] Energy Consumption Constraints of Sensor Nodes

[0142]

[0143] Where A is the maximum attack limit; i is the total number of tasks that need to be offloaded from sensor node i; F k is the maximum allowed number of tasks for edge computing node k; f ik is the computing capacity occupied by edge computing node k by sensor node i, f max is the calculation capacity limit; is the communication energy consumption between sensor node i and edge computing node k during task offloading, E imax is the communication energy consumption limit.

[0144] Optionally, the solution module 23 is specifically used for:

[0145] Obtain the maximum benefit of the edge IoT sensor network when there is no attack, obtain the initial upper bound value, and calculate the maximum benefit of the attacker under the task offloading strategy corresponding to the initial upper bound value to obtain the initial lower bound value;

[0146] Based on the branch and bound method, the upper and lower bound values ​​are iteratively updated until the difference between the upper and lower bound values ​​is less than the preset threshold. The task offloading strategy corresponding to the final upper bound value and the attack strategy corresponding to the final lower bound value are the equilibrium solutions of the Bayesian game model; among them, the task offloading strategy corresponding to the final upper bound value is the optimal task offloading strategy when the edge IoT sensor network is attacked.

[0147] Optionally, the solution module 23 is specifically used for:

[0148] Step 1: branch the attack strategy corresponding to the initial lower bound value;

[0149] Step 2: For any branch, solve the maximum benefit of the edge IoT sensor network under the branch attack strategy, and update the upper bound value according to the maximum benefit, and determine the maximum benefit of the attacker according to the task offloading strategy corresponding to the maximum benefit, and update the lower bound value according to the maximum benefit of the attacker;

[0150] Step 3: If the latest lower bound value is smaller than the upper bound value and larger than the original lower bound value, the attack strategy corresponding to the latest lower bound value is used as the attack strategy of the branch;

[0151] Step 4: Repeat steps 2 and 3 until the difference between the lower limit value and the upper limit value is less than a preset threshold.

[0152] Figure 3 is a schematic diagram of an electronic device 30 provided in an embodiment of the present invention.

[0153] like Figure 3 As shown, the electronic device 30 of this embodiment includes: a processor 31, a memory 32, and a computer program 33 stored in the memory 32 and executable on the processor 31, such as an edge IoT sensor network operation adjustment program. When the processor 31 executes the computer program 33, the steps in the above-mentioned edge IoT sensor network operation adjustment method embodiments are implemented, such as Figure 1 Alternatively, when the processor 31 executes the computer program 33, the functions of each module in the above-mentioned device embodiments are implemented, for example Figure 2 The functions of modules 21 to 24 are shown.

[0154] Exemplarily, the computer program 33 may be divided into one or more modules / units, one or more modules / units are stored in the memory 32, and are executed by the processor 31 to complete the present invention. One or more modules / units may be a series of computer program instruction segments that can complete specific functions, and the instruction segments are used to describe the execution process of the computer program 33 in the electronic device 30. For example, the computer program 33 may be divided into an acquisition module 21, a determination module 22, a solution module 23, and an adjustment module 24 (modules in the virtual device), and the specific functions of each module are as follows:

[0155] The acquisition module 21 is used to obtain attack information of the edge IoT sensor network.

[0156] The determination module 22 is used to establish a first utility function based on the attack information, with the goal of maximizing the benefit of the edge IoT sensor network when it is attacked, and to establish a second utility function with the goal of maximizing the benefit of the attacker, and to determine the Bayesian game model based on the first utility function and the second utility function.

[0157] The solution module 23 is used to establish the constraint conditions of the Bayesian game model and calculate the equilibrium solution of the Bayesian game model to obtain the optimal task offloading strategy when the edge IoT sensor network is attacked.

[0158] The adjustment module 24 is used to adjust the operation of the edge IoT sensor network based on the optimal task offloading strategy.

[0159] The electronic device 30 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device 30 may include, but is not limited to, a processor 31 and a memory 32. Those skilled in the art will appreciate that Figure 3 It is only an example of the electronic device 30 and does not constitute a limitation of the electronic device 30. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device 30 may also include input and output devices, network access devices, buses, etc.

[0160] The processor 31 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0161] The memory 32 may be an internal storage unit of the electronic device 30, such as a hard disk or memory of the electronic device 30. The memory 32 may also be an external storage device of the electronic device 30, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 30. Further, the memory 32 may also include both an internal storage unit of the electronic device 30 and an external storage device. The memory 32 is used to store computer programs and other programs and data required by the electronic device 30. The memory 32 may also be used to temporarily store data that has been output or is to be output.

[0162] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0163] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0164] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0165] In the embodiments provided by the present invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0166] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0168] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electric carrier signals and telecommunication signals.

[0169] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such 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, and should be included in the protection scope of the present invention.

Claims

1. A method for adjusting the operation of an edge IoT sensor network, characterized in that: include: Obtain attack information of edge IoT sensor networks; According to the attack information, a first utility function is established with the goal of maximizing the benefit of the edge IoT sensor network when it is attacked, and a second utility function is established with the goal of maximizing the benefit of the attacker, and a Bayesian game model is determined based on the first utility function and the second utility function; Establishing the constraint conditions of the Bayesian game model and calculating the equilibrium solution of the Bayesian game model to obtain the optimal task offloading strategy when the edge IoT sensor network is attacked; Adjusting the operation of the edge IoT sensor network based on the optimal task offloading strategy; Calculating the equilibrium solution of the Bayesian game model includes: Obtain the maximum benefit of the edge IoT sensor network when there is no attack, obtain an initial upper bound value, and calculate the maximum benefit of the attacker under the task offloading strategy corresponding to the initial upper bound value, and obtain an initial lower bound value; Based on the branch and bound method, the upper bound value and the lower bound value are iteratively updated until the difference between the upper bound value and the lower bound value is less than the preset threshold value. The task offloading strategy corresponding to the final upper bound value and the attack strategy corresponding to the final lower bound value are the equilibrium solutions of the Bayesian game model; wherein, the task offloading strategy corresponding to the final upper bound value is the optimal task offloading strategy when the edge IoT sensor network is attacked.

2. The edge IoT sensor network operation adjustment method according to claim 1, characterized in that: The attack information includes the type and probability of attacks launched by the attacker to the edge IoT sensor network, and the attack type includes distributed denial of service attacks and privacy eavesdropping attacks; The first utility function is: Where U(x, f) is the benefit of the edge IoT sensor network when it is attacked; p1+p2=1, p1 is the probability of distributed denial of service attack, p2 is the probability of privacy eavesdropping attack; K is the number of edge computing nodes in the edge IoT sensor network; U k，1 (x, f) is the benefit of the edge IoT sensor network when it is attacked by a distributed denial of service; U k，2 (x, f) is the gain of the edge IoT sensor network when it is attacked by privacy eavesdropping.

3. The edge IoT sensor network operation adjustment method as claimed in claim 2, characterized in that: The U k，1 (x, f) and the U k，2 The calculation formula for (x, f) is: In the formula, α k ={0, 1} represents the attack type, where 0 represents a distributed denial of service attack and 1 represents a privacy eavesdropping attack; b k ={0, 1} represents the attack decision, where 0 means not to launch an attack and 1 means to launch an attack; I is the number of sensor nodes in the edge IoT sensor network; r ik r′ is the wireless communication rate from sensor node i to edge computing node k before attack; ik is the wireless transmission rate of sensor node i offloading tasks to edge computing node k after being attacked and making task decisions; ik is the resource usage of sensor node i offloading tasks to edge computing node k; is the communication delay between sensor node i and edge computing node k during task offloading; The computing delay required for edge computing node k to process tasks; is the communication energy consumption between sensor node i and edge computing node k during task offloading; f ik The computing power of edge computing node k is occupied by sensor node i.

4. The edge IoT sensor network operation adjustment method as claimed in claim 3, characterized in that: The second utility function is: Where U a (a, b) is the attacker's profit; L is the number of edge computing nodes attacked by the attacker and L<K; is the probability of an attacker attacking edge computing node k; C k is the electricity cost of edge computing node k; μ k is the attack cost required for an attacker to attack edge computing node k; γ and θ are market risk control coefficients.

5. The edge IoT sensor network operation adjustment method as claimed in claim 4, characterized in that: The constraints include: The attacker's maximum number of attack nodes and maximum number of attacks Task offloading strategy constraints for sensor nodes Computing power constraints of edge computing nodes Energy Consumption Constraints of Sensor Nodes Where A is the maximum attack limit; i is the total number of tasks that need to be offloaded from sensor node i; F k is the maximum allowed number of tasks for edge computing node k; f ik is the computing capacity occupied by edge computing node k by sensor node i, f max is the calculation capacity limit; is the communication energy consumption between sensor node i and edge computing node k during task offloading, E imax is the communication energy consumption limit.

6. The edge IoT sensor network operation adjustment method according to claim 1, characterized in that: Iteratively update the upper and lower bound values ​​based on the branch and bound method until the difference between the upper and lower bound values ​​is less than a preset threshold, including: Step 1, branching the attack strategy corresponding to the initial lower bound value; Step 2: For any branch, solve the maximum benefit of the edge IoT sensor network under the branch attack strategy, and update the upper bound value according to the maximum benefit, and determine the maximum benefit of the attacker according to the task offloading strategy corresponding to the maximum benefit, and update the lower bound value according to the maximum benefit of the attacker; Step 3: If the latest lower bound value is smaller than the upper bound value and larger than the original lower bound value, the attack strategy corresponding to the latest lower bound value is used as the attack strategy of the branch; Step 4: Repeat steps 2 and 3 until the difference between the lower limit value and the upper limit value is less than a preset threshold.

7. An edge IoT sensor network operation adjustment device, characterized in that: include: An acquisition module is used to obtain attack information of the edge IoT sensor network; A determination module, configured to establish a first utility function based on the attack information with the goal of maximizing the benefit of the edge IoT sensor network when it is attacked, and to establish a second utility function with the goal of maximizing the benefit of the attacker, and to determine a Bayesian game model based on the first utility function and the second utility function; A solution module, used to establish the constraint conditions of the Bayesian game model and calculate the equilibrium solution of the Bayesian game model to obtain the optimal task offloading strategy when the edge IoT sensor network is attacked; An adjustment module, configured to adjust the operation of the edge IoT sensor network based on the optimal task offloading strategy; Calculating the equilibrium solution of the Bayesian game model includes: Obtain the maximum benefit of the edge IoT sensor network when there is no attack, obtain an initial upper bound value, and calculate the maximum benefit of the attacker under the task offloading strategy corresponding to the initial upper bound value, and obtain an initial lower bound value; Based on the branch and bound method, the upper bound value and the lower bound value are iteratively updated until the difference between the upper bound value and the lower bound value is less than the preset threshold value. The task offloading strategy corresponding to the final upper bound value and the attack strategy corresponding to the final lower bound value are the equilibrium solutions of the Bayesian game model; wherein, the task offloading strategy corresponding to the final upper bound value is the optimal task offloading strategy when the edge IoT sensor network is attacked.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images