Internet-of-vehicles safety calculation unloading method oriented to intelligent reflecting surface assistance

Abstract

The invention discloses an Internet of Vehicles safety calculation unloading method aided by an intelligent reflecting surface, and the method comprises the steps: obtaining vehicle basic information in the Internet of Vehicles, constructing a network system according to the vehicle basic information, and constructing an optimization problem under the constraint of the network system; obtaining an initial solution according to the optimization problem, defining the initial solution as an initial population, searching the initial population by adopting a swarm intelligence optimization algorithm to obtain a target population, and outputting the position of a global optimal individual in the target population; and performing safety calculation and efficiency optimization configuration according to the position of the global optimal individual. The method supports a multi-step unloading strategy, fully considers the selection and angle adjustment of the intelligent reflecting surface, and can achieve the minimization of the local total energy consumption of all vehicles and optimize the overall performance of an Internet of Vehicles system under the condition of meeting multiple constraints such as the transmission rate, the vehicle time delay and the security vulnerability cost.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method for offloading secure computing in an Internet of Vehicles assisted by an intelligent reflective surface. Background Art

[0002] In the past decade, with the rapid development of 5G networks, smart devices and Internet of Vehicles have been widely used. These smart devices require not only ultra-high computing performance, but also extremely low response time, especially in key application scenarios such as autonomous driving, in-vehicle entertainment, real-time traffic monitoring, and information exchange between vehicles. The widespread application of Internet of Vehicles technology has led to an increasing demand for data exchange and computing processing between vehicles and between vehicles and infrastructure. However, traditional in-vehicle computing devices have limitations in processing power and battery life, which directly affects the overall performance and user experience of the Internet of Vehicles system.

[0003] To meet this challenge, computation offloading technology has emerged. By offloading some computing tasks from on-board devices to remote cloud or edge computing platforms, the computing burden of on-board devices can be reduced, thereby improving the overall performance of the system. The advantages of computation offloading are particularly significant in application scenarios that require extremely high computing power, such as autonomous driving and real-time navigation. However, in autonomous driving and real-time traffic control, any communication delay or interruption may bring serious safety hazards and operational risks.

[0004] In order to solve the above wireless communication problems, intelligent reflecting surface (IRS) technology was proposed and applied in the field of wireless communication. IRS consists of a large number of adjustable reflection units, which can control and adjust the direction, strength and phase of the reflected signal, thereby optimizing the propagation path of the wireless signal. By intelligently controlling these reflection units, IRS can effectively enhance signal coverage, reduce interference, improve signal quality, and provide a more stable communication link in a dynamic environment. In the Internet of Vehicles, IRS technology is considered to be a solution with great potential. It can help improve the communication quality between vehicles and between vehicles and infrastructure, and solve the signal attenuation and interference problems that frequently occur in the Internet of Vehicles.

[0005] At the same time, in order to avoid serious interference problems and ensure the safe offloading of tasks, cryptographic algorithms are introduced. Researchers are exploring how to effectively combine IRS with edge servers in the Internet of Vehicles environment to minimize energy consumption while ensuring service quality.

[0006] In addition, although the existing sparrow algorithm and whale algorithm have fast convergence speed, they are insufficient in local search capabilities, which may limit their effectiveness in practical applications. Summary of the invention

[0007] In view of the shortcomings of the prior art, the present invention provides an energy consumption optimization method for vehicle network security computing assisted by intelligent reflective surfaces, which aims to solve the problem of increased system energy consumption caused by low vehicle uplink rate and high power consumption in complex road conditions.

[0008] To achieve the above object, the present invention provides the following technical solution: a method for offloading safe calculation of Internet of Vehicles assisted by intelligent reflective surface, comprising the following steps: Step S1: Obtain basic vehicle information of the Internet of Vehicles and build a network system, which includes a communication model, a computing model, and a security model; and construct an optimization problem for the network system based on constraints; Step S2: Obtain the initial solution of the optimization problem, and define the initial solution as the initial population. Use the swarm intelligence optimization algorithm SIOVS to search the initial population to obtain the target population, and obtain the global optimal solution in the target population; specifically: First, the population is initialized and the best and worst individuals in history are determined; then the weight factor is generated and applied to the solution search process; in the producer update phase, exponential decay exploration is performed to update the producer position; Secondly, perform Gaussian perturbation to further update the producer position; in the follower update phase, generate a random perturbation vector for each follower individual; if the follower's index exceeds half of the population, use Gaussian perturbation and exponential decay to update the position, otherwise, use the update relative to the leader and the random perturbation vector to adjust the follower position; then repeat the update of producers, followers and the historical best and worst individuals in sequence; finally obtain the global optimal solution; Step S3: Optimize the network system configuration of the Internet of Vehicles according to the global optimal solution.

[0009] Furthermore, the specific process of step S1 is: obtaining basic vehicle information of the Internet of Vehicles, further building a communication model, a computing model and a security model in the network system through the basic vehicle information, and constructing an optimization problem based on the communication model, the computing model and the security model; Furthermore, the specific process of obtaining basic vehicle information of the Internet of Vehicles in step S1 is as follows: Get a vehicle The information is expressed as: ; in, represents the index of any vehicle, Indicates the total number of vehicles; Get information about each vehicle's mission , expressed as: ; in, represents the index of any task for each user, Indicates the total number of tasks; Get the cloud CS information and roadside unit RSU information. The roadside unit RSU is represented as: ; in, and Respectively represent the indexes of different roadside units RSU, Indicates the total number of roadside units RSU in the network, Represents the collection of all connected vehicle infrastructure; Get the index collection of intelligent reflective surface IRS , expressed as: ; in, Represents the index of any intelligent reflective surface IRS; Indicates the total number of intelligent reflective surfaces IRS; Intelligent Reflective Surface IRS includes reflection unit coefficients, an indexed set of reflection unit coefficients , expressed as ; in, Represents the index of any reflection unit coefficient; represents the total number of reflection unit coefficients; The reflection unit coefficients are in matrix format and are expressed as: ; in, represents the reflection unit coefficient, Indicates The reflection coefficient of each reflective element on the intelligent reflective surface IRS is: represents a diagonal function; Get the physical location of all roadside units RSU and use it according to the physical location of all roadside units RSU The clustering algorithm will be divided into clusters; micro base stations are set based on the locations of the classification results; The communication range of each micro base station in the cluster includes several roadside units RSU, several vehicles and several intelligent reflective surfaces IRS. sub-channels are used by the roadside units RSU within the communication range, and the sub-channel index set is recorded as ;in, represents the index of any subchannel, Indicates the total number of subchannels; Get the total bandwidth of the network and subchannel bandwidth , the task is to pass the vehicle The Channel Transmit to intelligent reflective surface IRS The channel gain is denoted as , the task passes through the IRS via the sub-channel Transfer to RSU The channel gain is denoted as , the task is transmitted back to the vehicle through the roadside unit RSU The channel gain is denoted as and the noise is denoted as .

[0010] Furthermore, the process of constructing the communication model in step S1 is specifically as follows: Based on the condition that there is no intra-cluster interference in any task of the vehicle, the total bandwidth of the network system is divided into and They are used by the roadside unit RSU and the cloud respectively; is the frequency band division factor, and ; The number of subchannels used by each micro base station is expressed as ;in, represents the floor function, Indicates the number of subchannels; When a vehicle is associated with a roadside unit (RSU), the vehicle is connected to the roadside unit (RSU) by using the non-orthogonal multiple access (NOMA) technology. The part of the task to be unloaded is sent to multiple roadside units RSU at the same time; Calculate the uplink transmission rate, that is, calculate the uplink transmission rate in the subchannel On the vehicle Send the task to the roadside unit RSU Uplink NOMA transmission rate ; expressed as: ; ; ; ; ; ; ; in, Indicates the bandwidth used by the roadside unit, Indicates vehicle Associated Intelligent Reflecting Surface IRS The decision index coefficient of Indicates vehicle Associated Intelligent Reflecting Surface IRS The decision index coefficient of Indicates vehicle Its own transmission power; Indicates vehicle Its own transmission power; It means that in the process of uplink NOMA transmission, except for vehicles, and intelligent reflective surface IRS Other vehicles and other intelligent reflective surfaces other than IRS For vehicles and intelligent reflective surface IRS the interference caused; It represents any vehicle within the range of the micro base station. The tasks go through the subchannel Through uplink NOMA transmission to the intelligent reflector IRS The channel gain of Indicates other vehicles during the uplink NOMA transmission Send the task to the Intelligent Reflective Surface IRS The resulting channel gain interference; and denote the path loss factor and the Ricean fading factor, respectively. Indicates the vehicles within the range of the micro base station To the intelligent reflective surface IRS The distance , and They represent any vehicle Three-dimensional Cartesian coordinates, intelligent reflective surface IRS 3D Cartesian coordinates and roadside unit RSU The three-dimensional Cartesian coordinates of; It means vehicle To the IRS Line-of-sight link component; is the frequency; The elements in represent the intelligent reflective surface IRS The channel coefficient vector of each reflective element in ; represents the phase change caused by signal propagation, and They represent vehicles respectively To the IRS The direction angle of Indicates vehicle To the IRS Non-line-of-sight link component; represents the first noise;

[0011] When the vehicle is connected to the cloud, any vehicle Uplink NOMA transmission rate to the cloud It is expressed as: ; in, Indicates the bandwidth available for cloud use; Indicates vehicle IRS Link The decision index coefficient of represents any vehicle in the cluster The tasks go through the subchannel Transmitted to IRS via uplink NOMA The channel gain of It means the task from IRS Through the sub-channel Channel gain to the cloud; Indicates cloud to vehicle The channel gain of Represents the second noise.

[0012] Furthermore, the process of building a security model in step S1 is specifically as follows: When a vehicle is associated with a roadside unit (RSU), part of the vehicle's task can be offloaded to the roadside unit (RSU) after encryption; the roadside unit (RSU) decrypts the encrypted task and then encrypts it again before transmitting it to the cloud; When the vehicle is associated with the cloud CS, part of the vehicle's task can be offloaded to the cloud CS after encryption, and the cloud CS will perform the task after decryption; During the encryption or decryption process, different security levels are set. It is expressed as: middle; in, Representative Algorithm The security level index, Indicates the total number of encryption algorithms or decryption algorithms; algorithm Indicates encryption algorithm or decryption algorithm; Encryption and decryption algorithms The computing power is and ; Set the encryption energy consumption to be the same as the decryption energy consumption, algorithm The energy consumption of encryption or decryption is expressed as ; When the vehicle Mission Using encryption algorithm When part of its tasks can be safely offloaded, the failure probability is ; vehicle Mission Cost of security breaches It is expressed as: ; in, For the task financial loss in case of failure; A binary variable indicator indicating whether the vehicle v passes through the roadside units RSUs; For vehicles Mission Security decision indicators, set the selection algorithm To handle vehicles Mission ,but , otherwise 0; Based on this, the vehicle Total security breach cost It is expressed as: .

[0013] Furthermore, the processing flow of the calculation model constructed in step S1 is: Local Computing: When the vehicle When connected to a roadside unit (RSU) or cloud CS, the vehicle Mission The amount of local processing data is ,in, Indicates vehicle Execute the task The total amount of data; It's a task From vehicle The amount of data offloaded to the roadside unit (RSU) or cloud CS; Processing vehicles during local computing Tasks associated with the roadside unit RSU or cloud CS Local execution time used It is expressed as: ; In the formula, For vehicles computing power; To calculate time; For encryption time, Indicates vehicle Execute the task Costs; Unloading to roadside unit: When the vehicle When associated with a roadside unit RSU, the task Follow these steps; Will of After partial encryption, Unload to the roadside unit RSU; Roadside Unit RSU Decryption Then execute ;in, Indicates that during the decryption process, the vehicle Execute the task The amount of decrypted data in the process; right of After partial encryption, it is offloaded from the roadside unit RSU to the nearby cloud CS; Cloud CS executes after decryption ; Unload to the roadside unit to obtain the roadside unit RSU processing vehicle Mission Remote time , expressed as: ; In the formula, Indicates vehicle In subchannel The associated decision on the vehicle In subchannel Decision-making on association; is the limited backhaul rate between the roadside unit RSU and the cloud CS; It is the roadside unit RSU assigned to the vehicle Mission computing power; It is the cloud CS assigned to the vehicle Mission computing power; Indicates that From vehicle Time of uploading to the roadside unit RSU; Indicates calculation on the roadside unit RSU time; Indicates uploading from the roadside unit RSU Time to reach the nearby cloud CS; Indicates computing on the cloud CS time; Indicates decryption on the roadside unit RSU time; Indicates that it is used for encryption on the roadside unit RSU time; Indicates that it is used for decryption on a nearby cloud CS time; Offload to the cloud: When the vehicle is connected to the cloud CS, of After partial encryption, it is unloaded from the vehicle to the cloud CS; secondly, the cloud CS decrypts And execute; offload to the cloud for processing cloud CS related vehicles Mission The remote time is expressed as: ; in, Indicates that from the vehicle Upload Time to cloud CS; Indicates CS computing in the cloud time, Indicates decryption at the cloud CS time, 0 means cloud CS; vehicle The total time required for the task , expressed as: ; The local energy consumption of all vehicles is: ; In the formula, is the chip architecture energy coefficient, represents the calculated total energy consumption of the vehicle; represents the total energy consumption of encryption; Indicates the total energy consumption of uploading; Indicates vehicle consumption rate; Indicates the upload power.

[0014] Furthermore, in step S1, an optimization problem is constructed for the network system based on constraints, specifically: ; in, They represent the parameters of the constraint-based optimization problem, which are specifically expressed as: ; also, They represent different constraints respectively; Indicates vehicle The execution time of a task cannot exceed its deadline ; Indicates vehicle The total security breach cost cannot exceed its maximum acceptable cost ; and It indicates that a vehicle can only be associated with one facility, and the facility includes the cloud CS side unit RSU; Indicates vehicle Mission Only one encryption algorithm can be selected; and It means that the vehicle can only select one sub-channel through IRS; Indicates the vehicle The lower and upper bounds of the transmit power ; Indicates uninstallation part and Are greater than or equal to , but less than or equal to the vehicle Mission Data size ,at the same time, Must be less than or equal to ; and Indicates vehicle Only one IRS can be selected for association; represents the angular range of the elements in the IRS reflective surface, Indicates the preset minimum parameter.

[0015] Furthermore, step S2 is specifically as follows: Step S21: Initialize the maximum number of iterations using the swarm intelligence optimization algorithm SIOVS , and the current number of iterations Set to 1; Step S22: Define population , individual use To represent the population set ; The constraint-based optimization problem parameters Encoded , Represents population Medium Vehicle An index set of status indicators for establishing a connection with a base station; Encoded , Represents population Medium Vehicle The index set with the selected cryptographic algorithm; Encoded , Represents population Medium Vehicle An index set of selected subchannels; Encoded , Represents population Medium Vehicle The transmit power index set; Encoded , Represents population Medium Vehicle Select the IRS index set; Encoded , Indicates vehicle The reflection coefficient index set of the reflective element of the associated IRS is coded as , Represents population Medium mission from vehicle The bit size index set of data offloaded to the roadside unit; Encoded , Indicates the data bit size index set of the task roadside unit unloaded to the cloud; Step S23: Evaluate the fitness of individuals based on constraints and is a nonlinear mixed integer form, and It is introduced into the fitness function as a penalty term to prevent individuals from falling into the infeasible area. In order to make the constraint and The energy consumed by all users is minimized, and the individual The fitness function is: ; in, Represents an individual The fitness function value of It is a vehicle The delay constraint penalty factor is It is a vehicle The total cost constraint penalty factor of the security vulnerability is Represents the total time of the calculation task, Represents the maximum allowable delay of task execution, For vehicles The total cost of security breaches, For vehicles The maximum allowable total cost of a security breach; Step S24: Population initialization; in order to satisfy the constraints , the initial population can be generated using the following rules; specifically, any single Can be initialized as: ; in, It means to randomly output an element from the collection. Indicates the generation of a and A random number between Indicates the preset parameters. Use the fitness function to calculate the fitness values ​​of all individuals in the population, and take the individual with the highest fitness value as the optimal individual, and find the worst individual at the same time; Step S25: Determine the current iteration index Is it less than or equal to the maximum number of iterations? ; If so, calculate the current weight factor , and update it with the following formula: ; If not, output the global optimal solution in the target population; in, is the initial weight value, used to start the initial global search; is the attenuation weight coefficient, and its value is , used to gradually reduce the weight value; is the current iteration number; In each iteration, in order to better control the update probability of individuals, normalization is also required; the normalized weight value The calculation formula is: in: is the minimum weight value set, which is the target weight at the end of the algorithm iteration; The value range is ; It is an important parameter in the algorithm to dynamically adjust the probability; if the calculation If it is less than 0, it is forced to be set to 0; Step S26: Generate for A random number in the range, is the number of producers, and its calculation formula is ,generate , is a A random number in the range, if , use exponential decay to update the producer's position; otherwise, use Gaussian perturbation to update the producer's position; Furthermore, exponential decay means that as the number of iterations increases, the movement amplitude of the producer gradually decreases, thereby performing a fine search in the solution space. The specific update formula is as follows: in, , , , , , , Replace Make updates; , , , , , , Replace the corresponding Make updates; Indicates the individual position with the best fitness; is a A random number within a range, used to introduce randomness; An index for producers; Represents the maximum number of iterations; if the producer does not choose exponential decay exploration, it adopts the random perturbation development method, that is, a small random perturbation is performed near the leader position to develop a local area of ​​the solution space; the specific update formula is as follows: ; in: , , , , , , replace Make updates; , , , , , , Replace the corresponding Make updates; Represents a random number with a Gaussian distribution; introduces small perturbations in local areas of the solution space to make producers develop around the leader; Step S27: For the followers in the upper half of the population after the producer, the update strategy is to move toward the individual with the worst fitness in the current population; the specific update formula is as follows: ; Among them, followers , , , , , , , replace Make updates; , , , , , , replace Make updates; Indicates the position of the individual with the worst fitness in the current population; For followers in the lower half of the population, the update strategy is to move toward the most abundant individual, thereby performing local search and development; the specific update formula is as follows: ; in, , , , , , , replace Make updates; , , , , , , Corresponding replacement Make updates; is a random perturbation vector, where the elements are randomly -1 or 1.

[0016] Furthermore, the specific process of step S3 is to restore the position of the global optimal individual into the format of constraint-based optimization problem parameters according to the index set of the optimal position corresponding to the global optimal solution obtained by the swarm intelligent optimization algorithm in step S2; and according to the obtained global optimal solution, perform vehicle task offloading, algorithm selection, vehicle computing resource allocation, infrastructure resource allocation and vehicle power control.

[0017] Compared with the existing technology, the present invention has the following beneficial effects:

[0018] (1) The algorithm of the present invention uses a swarm intelligence optimization method for search; by introducing an adaptive adjustment mechanism, the algorithm can dynamically attenuate the weight factor according to the number of iterations, achieving a smooth transition from global search to local optimization; when the producer position is updated, the algorithm uses exponential decay and Gaussian perturbation to refine the search of the solution space, thereby ensuring a balance between global convergence and local adjustment; when the follower position is updated, the algorithm uses different update strategies according to the position of the individual in the population. For followers in the upper half of the population, the update strategy is to move toward the individual with the worst current fitness to explore potential undeveloped areas in the solution space, thereby avoiding falling into the local optimal solution. Followers in the lower half perform local searches toward the individual with the best current fitness to further optimize the quality of the solution.

[0019] (2) The present invention constructs an optimization problem based on the basic information of vehicles in the Internet of Vehicles, and constructs the optimization problem under the constraints of the network system; obtains an initial solution based on the optimization problem, uses a swarm intelligence optimization algorithm to search the initial population to obtain a target population, and outputs the position of the global optimal individual in the target population; performs a safe computing efficiency optimization configuration based on the global optimal individual position. The present invention considers the intelligent reflective surface factor to optimize vehicle association, channel selection, IRS angle, IRS selection, safety decision and power allocation to minimize the local total energy consumption of the vehicle.

[0020] (3) Under the constraints of computing resources and delay, security cost, task offloading ratio, user transmission power ratio, minimum rate and maximum energy consumption, the method can well minimize the total local energy consumption of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flow chart of the method of the present invention.

[0022] Figure 2 This is a schematic diagram of the present invention showing the effect of the number of vehicles on the user's local total energy consumption.

[0023] Figure 3 The present invention discloses a schematic diagram of the impact of a vehicle's maximum computing capability on a user's local total energy consumption.

[0024] Figure 4 The present invention discloses the convergence diagram of the iterative process of the SIOVS fitness function value of the method of the present invention. DETAILED DESCRIPTION

[0025] like Figure 1 As shown, a method for offloading safe calculation of Internet of Vehicles assisted by intelligent reflective surface includes the following steps: Step S1: Obtain basic vehicle information of the Internet of Vehicles and build a network system, which includes a communication model, a computing model, and a security model; and construct an optimization problem for the network system based on constraints; Step S2: Obtain the initial solution of the optimization problem, and define the initial solution as the initial population. Use the swarm intelligence optimization algorithm SIOVS to search the initial population to obtain the target population, and obtain the global optimal solution in the target population; specifically: First, the population is initialized and the best and worst individuals in history are determined; then the weight factor is generated and applied to the solution search process; in the producer update phase, exponential decay exploration is performed to update the producer position; Secondly, perform Gaussian perturbation to further update the producer position; in the follower update phase, generate a random perturbation vector for each follower individual; if the follower's index exceeds half of the population, use Gaussian perturbation and exponential decay to update the position, otherwise, use the update relative to the leader and the random perturbation vector to adjust the follower position; then repeat the update of producers, followers and the historical best and worst individuals in sequence; finally obtain the global optimal solution; Step S3: Optimize the network system configuration of the Internet of Vehicles according to the global optimal solution.

[0026] Furthermore, the specific process of step S1 is: obtaining basic vehicle information of the Internet of Vehicles, further building a communication model, a computing model and a security model in the network system through the basic vehicle information, and constructing an optimization problem based on the communication model, the computing model and the security model; Furthermore, the specific process of obtaining basic vehicle information of the Internet of Vehicles in step S1 is as follows: Get a vehicle The information is expressed as: ; in, represents the index of any vehicle, Indicates the total number of vehicles; Get information about each vehicle's mission , expressed as: ; in, represents the index of any task for each user, Indicates the total number of tasks; Get the cloud CS information and roadside unit RSU information. The roadside unit RSU is represented as: ; in, and Respectively represent the indexes of different roadside units RSU, Indicates the total number of roadside units RSU in the network, Represents the collection of all connected vehicle infrastructure; Get the index collection of intelligent reflective surface IRS , expressed as: ; in, Represents the index of any intelligent reflective surface IRS; Indicates the total number of intelligent reflective surfaces IRS; Intelligent Reflective Surface IRS includes reflection unit coefficients, an indexed set of reflection unit coefficients , expressed as ; in, Represents the index of any reflection unit coefficient; represents the total number of reflection unit coefficients; The reflection unit coefficients are in matrix format and are expressed as: ; in, represents the reflection unit coefficient, Indicates The reflection coefficient of each reflective element on the intelligent reflective surface IRS is: represents a diagonal function; Get the physical location of all roadside units RSU and use it according to the physical location of all roadside units RSU The clustering algorithm will be divided into clusters; micro base stations are set based on the locations of the classification results; The communication range of each micro base station in the cluster includes several roadside units RSU, several vehicles and several intelligent reflective surfaces IRS. sub-channels are used by the roadside units RSU within the communication range, and the sub-channel index set is recorded as ;in, represents the index of any subchannel, Indicates the total number of subchannels; Get the total bandwidth of the network and subchannel bandwidth , the task is to pass the vehicle The Channel Transmit to intelligent reflective surface IRS The channel gain is denoted as , the task passes through the IRS via the sub-channel Transfer to RSU The channel gain is denoted as , the task is transmitted back to the vehicle through the roadside unit RSU The channel gain is denoted as and the noise is denoted as .

[0027] Furthermore, the process of constructing the communication model in step S1 is specifically as follows: Based on the condition that there is no intra-cluster interference in any task of the vehicle, the total bandwidth of the network system is divided into and They are used by the roadside unit RSU and the cloud respectively; is the frequency band division factor, and ; The number of subchannels used by each micro base station is expressed as ;in, represents the floor function, Indicates the number of subchannels; When a vehicle is associated with a roadside unit (RSU), the vehicle is connected to the roadside unit (RSU) by using the non-orthogonal multiple access (NOMA) technology. The part of the task to be unloaded is sent to multiple roadside units RSU at the same time; Calculate the uplink transmission rate, that is, calculate the uplink transmission rate in the subchannel On the vehicle Send the task to the roadside unit RSU Uplink NOMA transmission rate ; expressed as: ; ; ; ; ; ; ; in, Indicates the bandwidth used by the roadside unit, Indicates vehicle Associated Intelligent Reflecting Surface IRS The decision index coefficient of Indicates vehicle Associated Intelligent Reflecting Surface IRS The decision index coefficient of Indicates vehicle Its own transmission power; Indicates vehicle Its own transmission power; It means that in the process of uplink NOMA transmission, except for vehicles, and intelligent reflective surface IRS Other vehicles and other intelligent reflective surfaces other than IRS For vehicles and intelligent reflective surface IRS the interference caused; It represents any vehicle within the range of the micro base station. The tasks go through the subchannel Through uplink NOMA transmission to the intelligent reflector IRS The channel gain of Indicates other vehicles during the uplink NOMA transmission Send the task to the Intelligent Reflective Surface IRS The resulting channel gain interference; and denote the path loss factor and the Ricean fading factor, respectively. Indicates the vehicles within the range of the micro base station To the intelligent reflective surface IRS The distance , and They represent any vehicle Three-dimensional Cartesian coordinates, intelligent reflective surface IRS 3D Cartesian coordinates and roadside unit RSU The three-dimensional Cartesian coordinates of; It means vehicle To the IRS Line-of-sight link component; is the frequency; The elements in represent the intelligent reflective surface IRS The channel coefficient vector of each reflection element in ; represents the phase change caused by signal propagation, and They represent vehicles respectively. To the IRS The direction angle of Indicates vehicle To the IRS Non-line-of-sight link component; represents the first noise;

[0028] When the vehicle is connected to the cloud, any vehicle Uplink NOMA transmission rate to the cloud It is expressed as: ; in, Indicates the bandwidth available for cloud use; Indicates vehicle IRS Link The decision index coefficient of represents any vehicle in the cluster The tasks go through the subchannel Transmitted to IRS via uplink NOMA The channel gain of It means the task from IRS Through the sub-channel Channel gain to the cloud; Indicates cloud to vehicle The channel gain of Represents the second noise.

[0029] Furthermore, the process of building a security model in step S1 is specifically as follows: When a vehicle is associated with a roadside unit (RSU), part of the vehicle's task can be offloaded to the roadside unit (RSU) after encryption; the roadside unit (RSU) decrypts the encrypted task and then encrypts it again before transmitting it to the cloud; When the vehicle is associated with the cloud CS, part of the vehicle's task can be offloaded to the cloud CS after encryption, and the cloud CS will perform the task after decryption; During the encryption or decryption process, different security levels are set. It is expressed as: middle; in, Representative Algorithm The security level index, Indicates the total number of encryption algorithms or decryption algorithms; algorithm Indicates encryption algorithm or decryption algorithm; Encryption and decryption algorithms The computing power is and ; Set the encryption energy consumption to be the same as the decryption energy consumption, algorithm The energy consumption of encryption or decryption is expressed as ; When the vehicle Mission Using encryption algorithm When part of its tasks can be safely offloaded, the failure probability is ; vehicle Mission Cost of security breaches It is expressed as: ; in, For the task financial loss in case of failure; A binary variable indicator indicating whether the vehicle v passes through the roadside units RSUs; For vehicles Mission Security decision indicators, set the selection algorithm To handle vehicles Mission ,but , otherwise 0; Based on this, the vehicle Total security breach cost It is expressed as: .

[0030] Furthermore, the processing flow of the calculation model constructed in step S1 is: Local Computing: When the vehicle When connected to a roadside unit (RSU) or cloud CS, the vehicle Mission The amount of local processing data is ,in, Indicates vehicle Execute the task The total amount of data; It's a task From vehicle The amount of data offloaded to the roadside unit (RSU) or cloud CS; Processing vehicles during local computing Tasks associated with roadside unit RSU or cloud CS Local execution time used It is expressed as: ; In the formula, For vehicles computing power; To calculate time; For encryption time, Indicates vehicle Execute the task Costs; Unloading to roadside unit: When the vehicle When associated with a roadside unit RSU, the task Follow these steps; Will of After partial encryption, Unload to the roadside unit RSU; Roadside Unit RSU Decryption Then execute ;in, Indicates that during the decryption process, the vehicle Execute the task The amount of decrypted data in the process; right of After partial encryption, it is offloaded from the roadside unit RSU to the nearby cloud CS; Cloud CS executes after decryption ; Unload to the roadside unit to obtain the roadside unit RSU processing vehicle Mission Remote time , expressed as: ; In the formula, Indicates vehicle In subchannel The associated decision on the vehicle In subchannel Decision-making on association; is the limited backhaul rate between the roadside unit RSU and the cloud CS; It is the roadside unit RSU assigned to the vehicle Mission computing power; It is the cloud CS assigned to the vehicle Mission computing power; Indicates that From vehicle Time of uploading to the roadside unit RSU; Indicates calculation on the roadside unit RSU time; Indicates uploading from the roadside unit RSU Time to reach the nearby cloud CS; Indicates computing on the cloud CS time; Indicates decryption on the roadside unit RSU time; Indicates that it is used for encryption on the roadside unit RSU time; Indicates that it is used for decryption on a nearby cloud CS time; Offload to the cloud: When the vehicle is connected to the cloud CS, of After partial encryption, it is unloaded from the vehicle to the cloud CS; secondly, the cloud CS decrypts And execute; offload to the cloud for processing cloud CS related vehicles Mission The remote time is expressed as: ; in, Indicates that from the vehicle Upload Time to cloud CS; Indicates CS computing in the cloud time, Indicates decryption at the cloud CS time, 0 means cloud CS; vehicle The total time required for the task , expressed as: ; The local energy consumption of all vehicles is: ; In the formula, is the chip architecture energy coefficient, represents the calculated total energy consumption of the vehicle; represents the total energy consumption of encryption; Indicates the total energy consumption of uploading; Indicates vehicle consumption rate; Indicates the upload power.

[0031] Furthermore, in step S1, an optimization problem is constructed for the network system based on constraints, specifically: ; in, They represent the parameters of the constraint-based optimization problem, which are specifically expressed as: ; also, They represent different constraints respectively; Indicates vehicle The execution time of a task cannot exceed its deadline ; Indicates vehicle The total security breach cost cannot exceed its maximum acceptable cost ; and It indicates that a vehicle can only be associated with one facility, and the facility includes the cloud CS side unit RSU; Indicates vehicle Mission Only one encryption algorithm can be selected; and It means that the vehicle can only select one sub-channel through IRS; Indicates the vehicle The lower and upper bounds of the transmit power ; Indicates uninstallation part and Are greater than or equal to , but less than or equal to the vehicle Mission Data size ,at the same time, Must be less than or equal to ; and Indicates vehicle Only one IRS can be selected for association; represents the angular range of the elements in the IRS reflective surface, Indicates the preset minimum parameter.

[0032] Furthermore, step S2 is specifically as follows: Step S21: Initialize the maximum number of iterations using the swarm intelligence optimization algorithm SIOVS , and the current number of iterations Set to 1; Step S22: Define the population , individual use To represent the population set ; The constraint-based optimization problem parameters Encoded , Represents population Medium Vehicle An index set of status indicators for establishing a connection with a base station; Encoded , Represents population Medium Vehicle The index set with the selected cryptographic algorithm; Encoded , Represents population Medium Vehicle a set of indices of the selected subchannels; Encoded , Represents population Medium Vehicle The transmit power index set of Encoded , Represents population Medium Vehicle Select the IRS index set; Encoded , Indicates vehicle The reflection coefficient index set of the reflective element of the associated IRS is coded as , Represents population Medium mission from vehicle The bit size index set of data offloaded to the roadside unit; Encoded , Indicates the data bit size index set of the task roadside unit unloaded to the cloud; Step S23: Evaluate the fitness of individuals based on constraints and is a nonlinear mixed integer form, and It is introduced into the fitness function as a penalty term to prevent individuals from falling into the infeasible area. In order to make the constraint and The energy consumed by all users is minimized, and the individual The fitness function is: ; in, Represents an individual The fitness function value of It is a vehicle The delay constraint penalty factor is It is a vehicle The total cost constraint penalty factor of the security vulnerability is Represents the total time of the calculation task, Represents the maximum allowable delay of task execution, For vehicles The total cost of security breaches, For vehicles The maximum allowable total cost of a security breach; Step S24: Population initialization; in order to satisfy the constraints , the initial population can be generated using the following rules; specifically, any single Can be initialized as: ; in, It means to randomly output an element from the collection. Indicates the generation of a and A random number between Indicates the preset parameters. Use the fitness function to calculate the fitness values ​​of all individuals in the population, and take the individual with the highest fitness value as the optimal individual, and find the worst individual at the same time; Step S25: Determine the current iteration index Is it less than or equal to the maximum number of iterations? ; If so, calculate the current weight factor , and update it with the following formula: ; If not, output the global optimal solution in the target population; in, is the initial weight value, used to start the initial global search; is the attenuation weight coefficient, and its value is , used to gradually reduce the weight value; is the current iteration number; In each iteration, in order to better control the update probability of individuals, normalization is also required; the normalized weight value The calculation formula is: in: is the minimum weight value set, which is the target weight at the end of the algorithm iteration; The value range is ; It is an important parameter in the algorithm to dynamically adjust the probability; if the calculation If it is less than 0, it is forced to be set to 0; Step S26: Generate for A random number in the range, is the number of producers, and its calculation formula is ,generate , is a A random number in the range, if , use exponential decay to update the producer's position; otherwise, use Gaussian perturbation to update the producer's position; Furthermore, exponential decay means that as the number of iterations increases, the movement amplitude of the producer gradually decreases, thereby performing a fine search in the solution space. The specific update formula is as follows: in, , , , , , , Replace Make updates; , , , , , , Replace the corresponding Make updates; Indicates the individual position with the best fitness; is a A random number within a range, used to introduce randomness; An index for producers; Represents the maximum number of iterations; if the producer does not choose exponential decay exploration, it adopts the random perturbation development method, that is, a small random perturbation is performed near the leader position to develop a local area of ​​the solution space; the specific update formula is as follows: ; in: , , , , , , replace Make updates; , , , , , , Replace the corresponding Make updates; Represents a random number with a Gaussian distribution; introduces small perturbations in local areas of the solution space to make producers develop around the leader; Step S27: For the followers in the upper half of the population after the producer, the update strategy is to move toward the individual with the worst fitness in the current population; the specific update formula is as follows: ; Among them, followers , , , , , , , replace Make updates; , , , , , , replace Make updates; Indicates the position of the individual with the worst fitness in the current population; For followers in the lower half of the population, the update strategy is to move toward the most abundant individuals to perform local search and development. The specific update formula is as follows: ; in, , , , , , , replace Make updates; , , , , , , Corresponding replacement Make updates; is a random perturbation vector, where the elements are randomly -1 or 1.

[0033] Furthermore, the specific process of step S3 is to restore the position of the global optimal individual into the format of constraint-based optimization problem parameters according to the index set of the optimal position corresponding to the global optimal solution obtained by the swarm intelligent optimization algorithm in step S2; and according to the obtained global optimal solution, perform vehicle task offloading, algorithm selection, vehicle computing resource allocation, infrastructure resource allocation and vehicle power control.

[0034] The effects of the implementation cases of the present invention can be further illustrated by simulation.

[0035] The simulation conditions are set as follows: 31 infrastructures (30 roadside units and 1 cloud) are considered; 20 vehicles, each user has 5 subtasks; the system bandwidth is 50MHz; there are two intelligent reflective surfaces, and the number of reflective blocks in one IRS unit is 60; 6 cryptographic algorithms, which require [100 200 250 300 350 1050] CPU cycles to encrypt one bit of data, and [90 280 350 300 400 1700] CPU cycles to decrypt one bit of data, and the energy consumption for encrypting and decrypting one bit of data is [2.5296 5.0425 6.837 7.8528 8.7073 26.3643]*1e-7 joules respectively; the maximum allowable delay of the computing task is 5~10s; the user energy coefficient is 10 -24 ; The base station energy coefficient is 10 -26 ; Mission safety incurs costs due to protection failure, and the mission safety factor is {5,6}.

[0036] Figure 2 The present invention discloses the effect of the vehicle's maximum computing power on the vehicle's local total energy consumption. When the vehicle's maximum computing power increases from 0.5 GHz to 2 GHz, the vehicle's local total energy consumption of the three algorithms, WOA, SSA, and SIOVS, increases with the increase in the vehicle's maximum computing power, but the user's local total energy consumption of the SIOVS algorithm proposed in the present invention is always better than the other two algorithms.

[0037] Figure 3 The present invention discloses the effect of the number of vehicles on the local total energy consumption of the vehicles. In the network, when the number of network vehicles increases from 10 to 19, the total energy consumption value will increase. However, the local total energy consumption obtained by the SIOVS algorithm proposed in the present invention is always better than the energy consumption values ​​obtained by the other two algorithms.

[0038] like Figure 4 As shown, the SIOVS algorithm of the present invention will converge to a fixed value after the fitness function undergoes a certain number of iterations, so as to obtain a solution to the optimization problem.

[0039] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for offloading safe calculation of Internet of Vehicles assisted by intelligent reflective surface, characterized in that: The steps include: Step S1: Obtain basic vehicle information of the Internet of Vehicles and build a network system, which includes a communication model, a computing model, and a security model; and construct an optimization problem for the network system based on constraints; Step S2: Obtain the initial solution of the optimization problem, and define the initial solution as the initial population. Use the swarm intelligence optimization algorithm SIOVS to search the initial population to obtain the target population, and obtain the global optimal solution in the target population; specifically: First, the population is initialized and the best and worst individuals in history are determined; then the weight factor is generated and applied to the solution search process; in the producer update phase, exponential decay exploration is performed to update the producer position; Secondly, perform Gaussian perturbation to further update the producer position; in the follower update phase, generate a random perturbation vector for each follower individual; if the follower's index exceeds half of the population, use Gaussian perturbation and exponential decay to update the position, otherwise, use the update relative to the leader and the random perturbation vector to adjust the follower position; then repeat the update of producers, followers and the historical best and worst individuals in turn; finally obtain the global optimal solution; Step S3: Optimize the network system configuration of the Internet of Vehicles according to the global optimal solution.

2. According to claim 1, a method for offloading safe calculation of connected vehicles assisted by intelligent reflective surfaces, characterized in that: The specific process of step S1 is: obtaining basic vehicle information of the Internet of Vehicles, further building a communication model, a calculation model and a security model in the network system through the basic vehicle information, and constructing an optimization problem based on the communication model, the calculation model and the security model; Furthermore, the specific process of obtaining basic vehicle information of the Internet of Vehicles in step S1 is as follows: Get a vehicle The information is expressed as: ; in, represents the index of any vehicle, Indicates the total number of vehicles; Get information about each vehicle's mission , expressed as: ; in, represents the index of any task for each user, Indicates the total number of tasks; Get the cloud CS information and roadside unit RSU information. The roadside unit RSU is represented as: ; in, and Respectively represent the indexes of different roadside units RSU, Indicates the total number of roadside units RSU in the network, Represents the collection of all connected vehicle infrastructure; Get the index collection of intelligent reflective surface IRS , expressed as: ; in, Represents the index of any intelligent reflective surface IRS; Indicates the total number of intelligent reflective surfaces IRS; Intelligent Reflective Surface IRS includes reflection unit coefficients, an indexed set of reflection unit coefficients , expressed as ; in, Represents the index of any reflection unit coefficient; represents the total number of reflection unit coefficients; The reflection unit coefficients are in matrix format and are expressed as: ； in, represents the reflection unit coefficient, Indicates The reflection coefficient of each reflective element on the intelligent reflective surface IRS is: represents a diagonal function; Get the physical location of all roadside units RSU and use it according to the physical location of all roadside units RSU The clustering algorithm will be divided into clusters; micro base stations are set based on the locations of the classification results; The communication range of each micro base station in the cluster includes several roadside units RSU, several vehicles and several intelligent reflective surfaces IRS. sub-channels are used by the roadside units RSU within the communication range, and the sub-channel index set is recorded as ;in, represents the index of any subchannel, Indicates the total number of subchannels; Get the total bandwidth of the network and subchannel bandwidth , the task is to pass the vehicle The Channel Transmit to intelligent reflective surface IRS The channel gain is denoted as , the task passes through the IRS via the sub-channel Transfer to RSU The channel gain is denoted as , the task is transmitted back to the vehicle through the roadside unit RSU The channel gain is denoted as and the noise is denoted as .

3. The method for offloading secure computing in connected vehicles assisted by intelligent reflective surfaces according to claim 2, characterized in that: The process of constructing the communication model in step S1 is specifically as follows: Based on the condition that there is no intra-cluster interference in any task of the vehicle, the total bandwidth of the network system is divided into and They are used by the roadside unit RSU and the cloud respectively; is the frequency band division factor, and ; The number of subchannels used by each micro base station is expressed as ;in, represents the floor function, Indicates the number of subchannels; When a vehicle is associated with a roadside unit (RSU), the vehicle is connected to the roadside unit (RSU) by using the non-orthogonal multiple access (NOMA) technology. The part of the task to be unloaded is sent to multiple roadside units RSU at the same time; Calculate the uplink transmission rate, that is, calculate the uplink transmission rate in the subchannel On the vehicle Send the task to the roadside unit RSU Uplink NOMA transmission rate ; expressed as: ； ； ； ； ； ； ； in, Indicates the bandwidth used by the roadside unit, Indicates vehicle Associated Intelligent Reflecting Surface IRS The decision index coefficient of Indicates vehicle Associated Intelligent Reflecting Surface IRS The decision index coefficient of Indicates vehicle Its own transmission power; Indicates vehicle Its own transmission power; It means that in the process of uplink NOMA transmission, except for vehicles, and intelligent reflective surface IRS Other vehicles and other intelligent reflective surfaces other than IRS For vehicles and intelligent reflective surface IRS the interference caused; It represents any vehicle within the range of the micro base station. The tasks go through the subchannel Through uplink NOMA transmission to the intelligent reflector IRS The channel gain of Indicates other vehicles during the uplink NOMA transmission Send the task to the Intelligent Reflective Surface IRS The resulting channel gain interference; and denote the path loss factor and the Ricean fading factor, respectively. Indicates the vehicles within the range of the micro base station To the intelligent reflective surface IRS The distance , and They represent any vehicle Three-dimensional Cartesian coordinates, intelligent reflective surface IRS 3D Cartesian coordinates and roadside unit RSU The three-dimensional Cartesian coordinates of; It means vehicle To the IRS Line-of-sight link component; is the frequency; The elements in represent the intelligent reflective surface IRS The channel coefficient vector of each reflective element in ; represents the phase change caused by signal propagation, and They represent vehicles respectively To the IRS The direction angle of Indicates vehicle To the IRS Non-line-of-sight link component; represents the first noise; When the vehicle is connected to the cloud, any vehicle Uplink NOMA transmission rate to the cloud It is expressed as: ; in, Indicates the bandwidth available for cloud use; Indicates vehicle IRS Link The decision index coefficient of represents any vehicle in the cluster The tasks go through the subchannel Transmitted to IRS via uplink NOMA The channel gain of It means the task from IRS Through the sub-channel Channel gain to the cloud; Indicates cloud to vehicle The channel gain of Represents the second noise.

4. The method for offloading secure computing in connected vehicles assisted by intelligent reflective surfaces according to claim 3, characterized in that: The process of building a security model in step S1 is specifically as follows: When a vehicle is associated with a roadside unit (RSU), part of the vehicle's task can be offloaded to the roadside unit (RSU) after encryption; the roadside unit (RSU) decrypts the encrypted task and then encrypts it again before transmitting it to the cloud; When the vehicle is associated with the cloud CS, part of the vehicle's task can be offloaded to the cloud CS after encryption, and the cloud CS will perform the task after decryption; During the encryption or decryption process, different security levels are set. It is expressed as: middle; in, Representative Algorithm The security level index, Indicates the total number of encryption algorithms or decryption algorithms; algorithm Indicates encryption algorithm or decryption algorithm; Encryption and decryption algorithms The computing power is and ; Set the encryption energy consumption to be the same as the decryption energy consumption, algorithm The energy consumption of encryption or decryption is expressed as ; When the vehicle Mission Using encryption algorithm When part of its tasks are safely offloaded, the failure probability is ; vehicle Mission Cost of security breaches It is expressed as: ； in, For the task financial loss in case of failure; Indicates whether the vehicle v passes the roadside unit RSU Binary variable indicator of ; For vehicles Mission Security decision indicators, set the selection algorithm To handle vehicles Mission ,but , otherwise 0; Based on this, the vehicle Total security breach cost It is expressed as: 。 5. The method for offloading secure computing in connected vehicles assisted by intelligent reflective surfaces according to claim 4, characterized in that: The processing flow of the calculation model constructed in step S1 is: Local Computing: When the vehicle When connected to a roadside unit (RSU) or cloud CS, the vehicle Mission The amount of local processing data is ,in, Indicates vehicle Execute the task The total amount of data; It's a task From vehicle The amount of data offloaded to the roadside unit (RSU) or cloud CS; Processing vehicles during local computing Tasks associated with roadside unit RSU or cloud CS Local execution time used It is expressed as: ； In the formula, For vehicles computing power; To calculate time; For encryption time, Indicates vehicle Execute the task Costs; Unloading to roadside unit: When the vehicle When associated with a roadside unit RSU, the task Follow these steps; Will of After partial encryption, Unload to the roadside unit RSU; Roadside Unit RSU Decryption Then execute ;in, Indicates that during the decryption process, the vehicle Execute the task The amount of decrypted data in the process; right of After partial encryption, it is offloaded from the roadside unit RSU to the nearby cloud CS; Cloud CS executes after decryption ; Unload to the roadside unit to obtain the roadside unit RSU processing vehicle Mission Remote time , expressed as: ； In the formula, Indicates vehicle In subchannel The associated decision on the vehicle In subchannel Decision-making on association; is the limited backhaul rate between the roadside unit RSU and the cloud CS; It is the roadside unit RSU assigned to the vehicle Mission computing power; It is the cloud CS assigned to the vehicle Mission computing power; Indicates that From vehicle Time of uploading to the roadside unit RSU; Indicates calculation on the roadside unit RSU time; Indicates uploading from the roadside unit RSU Time to reach the nearby cloud CS; Indicates computing on the cloud CS time; Indicates decryption on the roadside unit RSU time; Indicates that it is used for encryption on the roadside unit RSU time; Indicates that it is used for decryption on a nearby cloud CS time; Offload to the cloud: When the vehicle is connected to the cloud CS, of After partial encryption, it is unloaded from the vehicle to the cloud CS; secondly, the cloud CS decrypts And execute; offload to the cloud for processing cloud CS related vehicles Mission The remote time is expressed as: ； in, Indicates that from the vehicle Upload Time to cloud CS; Indicates CS computing in the cloud time, Indicates decryption at the cloud CS time, 0 means cloud CS; vehicle The total time required for the task , expressed as: ； The local energy consumption of all vehicles is: ； In the formula, is the chip architecture energy coefficient, represents the calculated total energy consumption of the vehicle; represents the total energy consumption of encryption; Indicates the total energy consumption of uploading; Indicates vehicle consumption rate; Indicates the upload power.

6. The method for offloading secure computing in connected vehicles assisted by intelligent reflective surfaces according to claim 5, characterized in that: In step S1, an optimization problem is constructed for the network system based on constraints, specifically: ； in, They represent the parameters of the constraint-based optimization problem, which are specifically expressed as: ； also, They represent different constraints respectively; Indicates vehicle The execution time of a task cannot exceed its deadline ; Indicates vehicle The total security breach cost cannot exceed its maximum acceptable cost ; and It indicates that a vehicle can only be associated with one facility, and the facility includes the cloud CS side unit RSU; Indicates vehicle Mission Only one encryption algorithm can be selected; and It means that the vehicle can only select one sub-channel through IRS; Indicates the vehicle The lower and upper bounds of the transmit power ; Indicates uninstall part and Are greater than or equal to , but less than or equal to the vehicle Mission Data size ,at the same time, Must be less than or equal to ; and Indicates vehicle Only one IRS can be selected for association; represents the angular range of the elements in the IRS reflective surface, Indicates the preset minimum parameter.

7. The method for offloading safe calculation of Internet of Vehicles assisted by intelligent reflective surface according to claim 6, characterized in that: Step S2 is specifically as follows: Step S21: Initialize the maximum number of iterations using the swarm intelligence optimization algorithm SIOVS , and the current number of iterations Set to 1; Step S22: Define the population , individual use To represent the population set ; The constraint-based optimization problem parameters Encoded , Represents population Medium Vehicle An index set of status indicators for establishing a connection with a base station; Encoded , Represents population Medium Vehicle The index set with the selected cryptographic algorithm; Encoded , Represents population Medium Vehicle a set of indices of the selected subchannels; Encoded , Represents population Medium Vehicle The transmit power index set; Encoded , Represents population Medium Vehicle Select the IRS index set; Encoded , Indicates vehicle The reflection coefficient index set of the reflective element of the associated IRS is coded as , Represents population Medium mission from vehicle The bit size index set of data offloaded to the roadside unit; Encoded , Indicates the data bit size index set of the task roadside unit unloaded to the cloud; Step S23: Evaluate the fitness of individuals based on constraints and is a nonlinear mixed integer form, and It is introduced into the fitness function as a penalty term to prevent individuals from falling into the infeasible area. In order to make the constraint and The energy consumed by all users is minimized, and the individual The fitness function is: ； in, Represents an individual The fitness function value of It is a vehicle The delay constraint penalty factor is It is a vehicle The total cost constraint penalty factor of the security vulnerability is Represents the total time of the calculation task, Represents the maximum allowable delay of task execution, For vehicles The total cost of security breaches, For vehicles The maximum allowable total cost of a security breach; Step S24: Population initialization; in order to satisfy the constraints , the initial population can be generated using the following rules; specifically, any single Can be initialized as: ； in, It means to randomly output an element from the collection. Indicates the generation of a and A random number between Indicates the preset parameters. Use the fitness function to calculate the fitness values ​​of all individuals in the population, and take the individual with the highest fitness value as the optimal individual, and find the worst individual at the same time; Step S25: Determine the current iteration index Is it less than or equal to the maximum number of iterations? ; If so, calculate the current weight factor , and update it with the following formula: ; If not, output the global optimal solution in the target population; in, is the initial weight value, used to start the initial global search; is the attenuation weight coefficient, and its value is , used to gradually reduce the weight value; is the current iteration number; In each iteration, in order to better control the update probability of individuals, normalization is also required; the normalized weight value The calculation formula is: in: is the minimum weight value set, which is the target weight at the end of the algorithm iteration; The value range is ; It is an important parameter in the algorithm to dynamically adjust the probability; if the calculation If it is less than 0, it is forced to be set to 0; Step S26: Generate for A random number in the range, is the number of producers, and its calculation formula is ,generate , is a A random number in the range, if , use exponential decay to update the producer's position; otherwise, use Gaussian perturbation to update the producer's position; Furthermore, exponential decay means that as the number of iterations increases, the movement amplitude of the producer gradually decreases, thereby performing a fine search in the solution space. The specific update formula is as follows: in, , , , , , , Replace Make updates; , , , , , , Replace the corresponding Make updates; Indicates the individual position with the best fitness; is a A random number within a range, used to introduce randomness; An index for producers; Represents the maximum number of iterations; if the producer does not choose exponential decay exploration, it adopts the random perturbation development method, that is, a small random perturbation is performed near the leader position to develop a local area of ​​the solution space; the specific update formula is as follows: ； in: , , , , , , replace Make updates; , , , , , , Replace the corresponding Make updates; Represents a random number with a Gaussian distribution; introduces small perturbations in local areas of the solution space to make producers develop around the leader; Step S27: For the followers in the upper half of the population after the producer, the update strategy is to move toward the individual with the worst fitness in the current population; the specific update formula is as follows: ； Among them, followers , , , , , , , replace Make updates; , , , , , , replace Make updates; Indicates the position of the individual with the worst fitness in the current population; For followers in the lower half of the population, the update strategy is to move toward the most abundant individuals to perform local search and development. The specific update formula is as follows: ； in, , , , , , , replace Make updates; , , , , , , Corresponding replacement Make updates; is a random perturbation vector, where the elements are randomly -1 or 1.

8. The method for offloading secure computing in connected vehicles assisted by intelligent reflective surfaces according to claim 7, characterized in that: The specific process of step S3 is to restore the position of the global optimal individual into the format of constraint-based optimization problem parameters according to the index set of the optimal position corresponding to the global optimal solution obtained by the swarm intelligence optimization algorithm in step S2; And based on the obtained global optimal solution, vehicle task offloading, algorithm selection, vehicle computing resource allocation, infrastructure resource allocation and vehicle power control are performed.

Images