Privacy conflict-sensitive vehicle-side cooperative task unloading method

Abstract

The invention requests to protect a privacy conflict-sensitive vehicle-side cooperative task unloading method, and belongs to the technical field of communication. In order to solve the problem of user privacy leakage possibly caused by privacy conflicts among tasks when the same user unloads a plurality of concurrent tasks to the same target server in Internet of Vehicles edge calculation, the invention provides a privacy conflict-sensitive vehicle-side collaborative task unloading method. According to the method, the risk of privacy conflict between tasks is quantified through a privacy conflict matrix, and an unloading decision is constructed for each concurrent task by using V2I, V2V and a vehicle-side collaborative unloading mode of local unloading according to the computing resource states of an edge server and collaborative vehicles and the link duration between the vehicles. Multi-task game modeling is carried out by adopting a precise potential game method, iterative optimization is carried out on unloading decisions of tasks through a probability greedy exploration strategy, and Nash equilibrium is finally realized, so that the privacy protection level is effectively improved, and the task completion rate is maximized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of communication network, and particularly relates to a privacy conflict sensitive vehicle edge collaborative task offloading method. BACKGROUND

[0002] With the rapid development of vehicle networking technology, a large number of delay-sensitive and computing-intensive vehicle-mounted applications emerge in an endless stream, the computing demand of vehicle-mounted devices is increasing day by day, and vehicle networking edge computing has become an effective means to solve the insufficient computing resources of vehicle terminals. In addition to utilizing local computing resources of vehicles to process tasks, the technology can also offload tasks to adjacent road side units (RSUs) through a vehicle-to-infrastructure (V2I) mode or offload tasks to collaborative vehicles through a vehicle-to-vehicle (V2V) mode, so as to effectively improve the resource utilization of a mobile edge network, optimize task processing efficiency, and improve user service experience.

[0003] The existing research on vehicle networking edge computing does not consider the privacy conflict problem that may exist between different tasks, that is, when multiple tasks with privacy conflicts are offloaded to the same target server, the high correlation between task data may expose sensitive information of users, thereby causing the risk of privacy leakage of users. In view of this problem, the application provides a privacy conflict sensitive vehicle edge collaborative task offloading method, which adopts an exact potential game method to model a multi-task game, maximizes the system utility through offloading game between concurrent tasks, and realizes Nash equilibrium, so as to effectively improve the task completion rate and the task privacy protection level. The exact potential game method is a special method in game theory, which maximizes the sum of utility functions of each participant through iterative update of strategy selection of each participant, so as to obtain the Nash equilibrium solution of the potential game. SUMMARY

[0004] The application aims to solve the problems of the prior art. A privacy conflict sensitive vehicle edge collaborative task offloading method is provided. The technical scheme of the application is as follows:

[0005] A privacy conflict sensitive vehicle edge collaborative task offloading method comprises the following steps:

[0006] 101. An exact potential game method is adopted to establish a vehicle networking task offloading optimization model, the system comprises an edge server k, a task vehicle j * and a plurality of collaborative vehicles J={j}, let I={i} be a concurrent task set of the vehicle j * proposed, wherein each task i is a game participant, a i is the offloading decision of the task i, and u iLet A = {a} be the utility function value for task i. i | i∈I} represents the global unloading decision, U = {u i | i∈I} represents the global utility function value, and the exploration threshold γ is initialized in the real number interval [0,1]. th ;

[0007] 102. Based on the data volume s of each task i in set I. i Computational complexity c i Maximum tolerable delay For each task i, establish a candidate unloading node set N. i ;

[0008] 103. For each task i participating in the game in the task set I, remove it from its candidate unloading node set N. i Randomly select a node to construct the unloading decision a i According to the global unloading decision A = {a i | i∈I} Calculate the global utility function value U = {u i | i∈I Let temporary set I′=I, temporary set

[0009] 104. If Take the first element i from set I′ and generate a random number γ in the real number interval [0,1]. Jump to step 105; otherwise, jump to step 108.

[0010] 105. If γ < γ th From the candidate unloading node set N of task i i Randomly select a candidate unloading node and calculate the unloading decision a′. i and utility function value u′ i If yes, proceed to step 106; otherwise, proceed to step 107.

[0011] 106. If u′ i >u i The unloading decision a′ for task i i Add to set A′ and jump to step 104; otherwise, jump to step 104.

[0012] 107. From the candidate unloading node set N of task i i Iterate through all candidate unloading nodes and calculate the corresponding unloading decision a′ for each candidate unloading node. i and utility function value u′ i And select the largest utility function value u′ from them. iand its corresponding unloading decision a′ i If a′ i ≠a i The corresponding unloading decision a′ i Add to set A′ and jump to step 104; otherwise, jump to step 104.

[0013] 108. If set Randomly select an unloading decision a′ for task i from set A′. i Let a i =a′ i Update the global uninstallation decision A = {a i | i∈I} and the global utility function value U = {u i | i∈I Let temporary set I′=I, temporary set Skip to step 104; otherwise, skip to step 109.

[0014] 109. The algorithm reaches Nash equilibrium and outputs the global unloading decision A, thus ending the process.

[0015] Furthermore, in step 102, a candidate unloading node set N is established for task i. i Specifically, it includes the following steps:

[0016] 1) Initialize the set of candidate unload nodes for task i Temporary assembly Based on the vehicle's effective communication range, the mission vehicle j * Cooperative vehicles within a one-hop range are added to set J′;

[0017] 2) If Take the first element j from set J′ and calculate the task vehicle j in V2V unloading mode. * Link duration with collaborating vehicle j Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay Calculation in The computing resources f that collaborative vehicle j needs to allocate for task i within the specified time period. i j If yes, proceed to step 3; otherwise, proceed to step 4.

[0018] 3) If vehicle j has remaining available computing resources f j Greater than or equal to f i j Add the cooperating vehicle j to the candidate unloading node set N. iIf yes, proceed to step 2); otherwise, proceed to step 2.

[0019] 4) Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay In the local unloading mode, the task vehicle j * Local computing resources to be allocated for task i If the mission vehicle j * Remaining available computing resources Greater than or equal to The mission vehicle j * Add to candidate unload node set N i If yes, proceed to step 5); otherwise, proceed to step 5.

[0020] 5) Calculate the task vehicle j in V2I unloading mode. * Link duration with edge server k in the current cell Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay Calculation in Within a given timeframe, the edge server k in the current cell needs to allocate computing resources f to task i. i k ;

[0021] 6) If the remaining available computing resources f of the edge server k in the current cell are... k Greater than or equal to f i k Add the edge server k of the current cell to the candidate offload node set N. i If yes, proceed to step 7); otherwise, proceed to step 7.

[0022] 7) Output the set of candidate unloading nodes N i The steps are now complete.

[0023] Furthermore, in step 2) during the V2V unloading mode, the task vehicle j * Link duration with collaborating vehicle j The calculation method is shown in formula (1), where the computing resources f to be allocated by the cooperating vehicle j for task i are... i j The calculation method is shown in formula (2):

[0024]

[0025] In formula (1), Indicates the mission vehicle j *The driving speed of the task vehicle j is given. * The direction of travel is positive, and v j v represents the speed of the cooperating vehicle j. j >0 indicates that vehicle j and vehicle j * Traveling in the same direction, v j <0 indicates that vehicle j and vehicle j * Reverse driving, R V2V This indicates the maximum communication distance between vehicles. Indicates the mission vehicle j * Position, x j This indicates the location of the collaborating vehicle j. In V2V offloading mode, it is determined based on the task vehicle j. * By analyzing the position and speed relationship between the cooperating vehicle j and the other vehicle j, the duration of the communication link between them can be calculated.

[0026]

[0027] In formula (2), s i c represents the amount of data for task i. i Let represent the computational complexity of task i. r represents the maximum tolerable delay for task i. j Indicates the mission vehicle j * The data transmission rate between the vehicle and the cooperating vehicle j is calculated as shown in formula (3):

[0028]

[0029] In formula (3), B V2V The bandwidth of the V2V link is represented by q, the transmit power of the mission vehicle is represented by g. j Indicates the mission vehicle j * V2V link channel gain between the cooperating vehicle j and the vehicle j, N0 represents noise power, L j Indicates the mission vehicle j * Path loss between the vehicle and the cooperating vehicle j.

[0030] Furthermore, in step 4) under local unloading mode, the task vehicle j * Local computing resources to be allocated for task i The calculation method is shown in formula (4):

[0031]

[0032] Furthermore, in step 5) under V2I unloading mode, the task vehicle j * Link duration with edge server k in the current cell The calculation method is shown in formula (5). The edge server k of the current cell needs to allocate computing resources f for task i. i k The calculation method is shown in formula (6):

[0033]

[0034] In formula (5), R V2I x represents the effective communication radius of the RSU. k This indicates the location of the edge server k in the current cell;

[0035]

[0036] In formula (6), r k Indicates the mission vehicle j * The data transmission rate between the current cell and the edge server k is calculated as shown in formula (7):

[0037]

[0038] In formula (7), B V2I G represents the bandwidth of the V2I link. k Indicates the mission vehicle j * The V2I link channel gain, L, between the current cell's edge server k and the edge server k is... k Indicates the mission vehicle j * The path loss between the current edge server k in the current cell.

[0039] Furthermore, in step 103, the unloading decision a for task i is constructed. i The method is as shown in formula (8):

[0040] a i =(n i ,f i ) (8) In formula (8), n i The candidate unloading node selected for task i is calculated as shown in formula (9):

[0041]

[0042] In formula (9), f i This represents the computing resources obtained by task i, calculated as shown in formula (10):

[0043]

[0044] Furthermore, in step 103, the global utility function value U = {u} is obtained. i |i∈I} is shown in equation (11):

[0045]

[0046] In equation (11), i∈I, i'∈I, and i≠i', a i and a i′ represent the offloading decision of task i or task i', respectively, a -i and a -i′ represent the set of offloading decisions of other tasks in set I except task i or task i', respectively, h i (a i , a -i ) and h i′ (a i′ , a -i′ ) represent the offloading benefit of task i or task i' when all tasks in set I participate in offloading, a -i′\i represents the set of offloading decisions of other tasks in set I except task i' when only task i does not participate in offloading, h i′ (a i′ , a -i′\i ) represents the offloading benefit of task i' when only task i does not participate in offloading, h i (a i , a -i ) and h i′ (a i′ , a -i′ ) are calculated as shown in equation (12), where task i or task i' represents any two different tasks in set I, h i′ (a i′ , a -i′\i ) is calculated as shown in equation (13):

[0047]

[0048] In equation (12) and equation (13), is a weight factor, θ i represents the completion status of task i when all tasks in set I participate in offloading, θ i = 1 if task i is completed within the maximum tolerated delay, otherwise θ i = 0, and θ i′\i represents the completion status of task i' when only task i does not participate in offloading, θ i′\i = 1 if task i' is completed within the maximum tolerated delay, otherwise θ i′\i = 0;

[0049] In equation (12), p idenotes the privacy protection level of task i when all tasks in set I participate in offloading, and the calculation method is shown in formula (14):

[0050]

[0051] In formula (14), denotes whether task i adopts V2I offloading mode, if task i is offloaded to edge server k in the current cell, let Otherwise, let denotes whether task i adopts V2V offloading mode, if task i is offloaded to cooperative vehicle j, let Otherwise, let |I| denotes the total number of concurrent tasks, m i,i′ denotes an element in the privacy conflict matrix M, and the calculation method is shown in formula (15):

[0052]

[0053] In formula (13), p i′\i denotes the privacy protection level of task i' when only task i in set I does not participate in offloading, and the calculation method is shown in formula (16):

[0054]

[0055] In formula (16), task i'' ∈ (I-i-i') denotes another task in set I except task i and task i'.

[0056] Further, the step 105 and the step 107 calculate the obtained offloading decision a' i and the utility function value u' i , and specifically include the following steps:

[0057] 1) Let temporary set A'' = A;

[0058] 2) According to the selected candidate offloading node of task i, construct the offloading decision a' i of task i in set A'' according to formula (8), let i a i of task i in set A'' according to formula (11), let i u i = u i ;

[0059] 3) Output the offloading decision a' and the utility function value u' i of task i, and the step ends.

[0060] An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the privacy conflict sensitive edge computing task offloading method when executing the program.

[0061] A non-transitory computer readable storage medium having stored thereon a computer program, the computer program implementing the privacy conflict sensitive edge computing task offloading method when executed by a processor.

[0062] The advantages and beneficial effects of the present application are as follows:

[0063] The present application discloses a privacy conflict sensitive edge computing task offloading method. Existing research on vehicle network edge computing offloading usually only focuses on task offloading performance and rarely considers privacy conflict between tasks. In view of the problem that when a user offloads multiple concurrent tasks to the same target server in vehicle network edge computing, privacy conflict between tasks may lead to user privacy leakage, the present application proposes a privacy conflict sensitive edge computing task offloading method. The method uses a privacy conflict matrix to represent the privacy conflict risk between concurrent tasks and quantifies the privacy protection level of the tasks. The edge computing task offloading mode of V2I, V2V and local offloading is used to select candidate offloading nodes for each concurrent task. A multi-task offloading game model is constructed using the exact potential game method, and the task offloading decision is iteratively optimized by a probability greedy exploration strategy to maximize the system utility and finally achieve Nash equilibrium, thereby effectively improving the task completion rate and the task privacy protection level. The exact potential game method is a special method in game theory, which maximizes the sum of the utility functions of each participant by iteratively updating the strategy selection of each participant, thereby obtaining the Nash equilibrium solution of the potential game. BRIEF DESCRIPTION OF DRAWINGS

[0064] Figure 1 The present application provides a privacy conflict sensitive edge computing task offloading method flow chart.

[0065] Formula

[0066] The technical solutions in the embodiments of the present application will be described in detail below with reference to the accompanying drawings. The described embodiments are only a part of the embodiments of the present application.

[0067] The technical solutions of the present application to solve the above technical problems are:

[0068] The concepts and models involved in the present application are as follows.

[0069] 1. System model

[0070] The present application assumes that the system area is a two-way straight road, a plurality of RSUs are arranged on both sides of the road, and each RSU is equipped with a resource-limited edge server, which can provide task offloading services for vehicle users. The cooperative vehicle set on the road in the system is J={j}, which can be used as an offloading node to provide task offloading services for vehicle users. Vehicles can access nearby RSUs through V2I mode, or communicate with nearby vehicles through V2V mode.

[0071] The vehicle concurrently generates a plurality of task offloading requests, and each task cannot be split. The task set of the task vehicle j * is represented as I(i∈I). The task i of the task vehicle j * The offloading request of the task i can be represented as where s i represents the data volume of the task i, c i represents the computing complexity of the task i, and τi represents the maximum tolerable time delay of the task i. Considering that the maximum tolerable time delay of the task is small, it is assumed that the driving speed of the vehicle in the system is constant within the time period.

[0072] The task can be offloaded to the local, or offloaded to the edge server k of the current cell through the V2I offloading mode, or offloaded to the cooperative vehicle j based on the effective communication distance constraint of the vehicle through the V2V offloading mode, so the task has three offloading modes to choose from: local offloading, V2I, and V2V. The technical solutions of the present application are as follows.

[0073] 1. The link duration between the task vehicle j * and the cooperative vehicle j

[0074] The calculation method is shown in formula (1):

[0075]

[0076] In formula (1), represents the driving speed of the task vehicle j * , and the driving direction of the task vehicle j * is the positive direction, and v j represents the driving speed of the cooperative vehicle j, v j >0 indicates that the vehicle j and the vehicle j * travel in the same direction, v j <0 indicates that the vehicle j and the vehicle j * travel in opposite directions, R V2V represents the maximum communication distance between vehicles, represents the position of the task vehicle j * , and x jdenotes the position of the cooperative vehicle j;

[0077] 2. The computing resource f allocated by the cooperative vehicle j for the task i i j

[0078] The calculation method is shown in equation (2):

[0079]

[0080] In equation (2), s i denotes the data volume of the task i, c i denotes the computing complexity of the task i, denotes the maximum tolerable time delay of the task i, r j denotes the data transmission rate between the task vehicle j * and the cooperative vehicle j, the calculation method is shown in equation (3):

[0081]

[0082] In equation (3), B V2V denotes the V2V link bandwidth, q denotes the transmission power of the task vehicle, g j denotes the V2V link channel gain between the task vehicle j * and the cooperative vehicle j, N0 denotes the noise power, L j denotes the path loss between the task vehicle j * and the cooperative vehicle j;

[0083] 3. The local computing resource f allocated by the task vehicle j * for the task i

[0084] The calculation method is shown in equation (4):

[0085]

[0086] 4. In the V2I offloading mode, the link duration between the task vehicle j * and the edge server k of the current cell

[0087] The calculation method is shown in equation (5):

[0088]

[0089] In equation (5), R V2I denotes the effective communication radius of the RSU, x k denotes the position of the edge server k of the current cell;

[0090] 5. The computing resource f allocated by the edge server k of the current cell for the task i i k

[0091] The calculation method is shown in formula (6):

[0092]

[0093] In formula (6), r k represents the data transmission rate between the task vehicle j * and the edge server k of the current cell, and the calculation method is shown in formula (7):

[0094]

[0095] In formula (7), B V2I represents the bandwidth of the V2I link, g k represents the V2I link channel gain between the task vehicle j * and the edge server k of the current cell, and L k represents the path loss between the task vehicle j * and the edge server k of the current cell.

[0096] 6. The offloading decision a i

[0097] The construction method is shown in formula (8):

[0098] a i = (n i , f i ) (24)

[0099] In formula (8), n i represents the candidate offloading node selected by the task i, and the calculation method is shown in formula (9):

[0100]

[0101] In formula (8), f i represents the computing resource obtained by the task i, and the calculation method is shown in formula (10):

[0102]

[0103] 7. The global utility function value U = {u i | i∈I}

[0104] The calculation method is shown in formula (11):

[0105]

[0106] In formula (11), i∈I, i'∈I, and i≠i', a i and a i′ respectively represent the offloading decision of task i or task i', a -i and a -i′ respectively represent the set of offloading decisions of other tasks in set I except task i or task i', h i (a i ,a -i ) and h i′ (a i′ ,a -i′ ) respectively represent the offloading benefit of task i or task i' when all tasks in set I participate in offloading, a -i′\i represents the set of offloading decisions of other tasks in set I except task i' when only task i does not participate in offloading, h i′ (a i′ ,a -i′\i ) represents the offloading benefit of task i' when only task i does not participate in offloading, h i (a i ,a -i ) and h i′ (a i′ ,a -i′ ) are calculated as shown in formula (12), wherein task i or task i' represents any two different tasks in set I, h i′ (a i′ ,a -i′\i ) is calculated as shown in formula (13):

[0107]

[0108]

[0109] In formula (12) and formula (13), is a weight factor, θ i represents the completion state of task i when all tasks in set I participate in offloading, θ i =1 if task i is completed within the maximum tolerable delay, otherwise θ i =0, and θ i′\i represents the completion state of task i' when only task i does not participate in offloading, θ i′\i =1 if task i' is completed within the maximum tolerable delay, otherwise θ i′\i =0.

[0110] In formula (12), p i represents the privacy protection level of task i when all tasks in set I participate in offloading, and the calculation method is shown in formula (14):

[0111]

[0112] In formula (14), denotes whether task i adopts V2I offloading mode, if task i is offloaded to edge server k in the current cell, let Otherwise, let denotes whether task i adopts V2V offloading mode, if task i is offloaded to cooperative vehicle j, let Otherwise, let |I| denotes the total number of concurrent tasks, m i,i′ denotes an element in the privacy conflict matrix M, the calculation method is shown in formula (15):

[0113]

[0114] In formula (13), p i′\i denotes the privacy protection level of task i' in set I when only task i does not participate in offloading, the calculation method is shown in formula (16):

[0115]

[0116] In formula (16), task i'' ∈ (I-i-i') denotes another task in set I except task i and task i';

[0117] 8. Sub-algorithm 1: Establishing candidate offloading node set N for task i i

[0118] Step 1: Initialize the candidate offloading node set of task i Let temporary set According to the effective communication distance of the vehicle, cooperative vehicles within one-hop range of task vehicle j * are added to set J';

[0119] Step 2: If Take the first element j from set J', calculate the link duration between task vehicle j * and cooperative vehicle j in V2V offloading mode According to the data volume s i of task i, the calculation complexity c i , and the maximum tolerable delay , calculate the computing resources f i allocated by cooperative vehicle j for task i within time j , jump to step 3, otherwise, jump to step 4;

[0120] Step 3: If vehicle j has remaining available computing resources f j Greater than or equal to f i j Add the cooperating vehicle j to the candidate unloading node set N. i If yes, proceed to step 2; otherwise, proceed to step 2.

[0121] Step 4: Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay In the local unloading mode, the task vehicle j * Local computing resources to be allocated for task i If the mission vehicle j * Remaining available computing resources Greater than or equal to The mission vehicle j * Add to candidate unload node set N i If yes, proceed to step 5; otherwise, proceed to step 5.

[0122] Step 5: Calculate the mission vehicle j in V2I unloading mode. * Link duration with edge server k in the current cell Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay Calculation in Within a given timeframe, the edge server k in the current cell needs to allocate computing resources f to task i. i k ;

[0123] Step 6: If the remaining available computing resources f of the edge server k in the current cell are... k Greater than or equal to f i k Add the edge server k of the current cell to the candidate offload node set N. i If yes, proceed to step 7; otherwise, proceed to step 7.

[0124] Step 7: Output the set of candidate unloading nodes N i The steps are now complete.

[0125] 9. Sub-algorithm 2: Calculate and obtain the unloading decision a′ i and utility function value u′ i

[0126] Step 1: Let the temporary set A″ = A;

[0127] Step 2: According to the selected candidate offloading node of task i, the offloading decision a' is constructed by using formula (8) i , let a i =a' i , according to A'', the utility function value u i of task i is calculated by using formula (11) i , let u' i =u i ;

[0128] Step 3: Output the offloading decision a' and the utility function value u' of task i i , and the step is ended.

[0129] A privacy conflict sensitive task offloading method at the edge of the vehicle, specifically comprising the following steps:

[0130] Step 1: An accurate potential game method is used to establish a task offloading optimization model of Internet of Vehicles, the system contains an edge server k, a task vehicle j * and a plurality of cooperative vehicles J={j}, let I={i} be the concurrent task set proposed by vehicle j * , wherein each task i is a game participant, a i is the offloading decision of task i, u i is the utility function value of task i, A={a i | i∈I} is the global offloading decision, U={u i | i∈I} is the global utility function value, and the exploration threshold value γ is initialized in the real number interval [0, 1] th ;

[0131] Step 2: According to the data amount s i , the calculation complexity c i , and the maximum tolerable time delay of each task i, a candidate offloading node set N i is established for each task i

[0132] Step 3: For each task i participating in the game in the task set I, an offloading decision a i is constructed by randomly selecting a node from the candidate offloading node set N i of the task i, the global utility function value U={u i | i∈I} is calculated according to the global offloading decision A={a i | i∈I}, let the temporary set I'=I, and the temporary set

[0133] Step 4: If Take the first element i from the set I', and generate a random number y in the real number interval [0, 1], jump to step 5, otherwise, jump to step 8;

[0134] Step 5: If y < y th , randomly select a candidate offloading node from the candidate offloading node set N i of the task i, calculate the obtained offloading decision a' i and the utility function value u' i , jump to step 6, otherwise, jump to step 7;

[0135] Step 6: If u' i > u i , add the offloading decision a' i of the task i to the set A', jump to step 104, otherwise, jump to step 4;

[0136] Step 7: Traverse all candidate offloading nodes from the candidate offloading node set N i of the task i, respectively calculate the obtained offloading decision a' i and the utility function value u' i of each candidate offloading node, and select the maximum utility function value u' i and the corresponding offloading decision a' i , if a' i ≠ a i , add the corresponding offloading decision a' i to the set A', jump to step 4, otherwise, jump to step 4;

[0137] Step 8: If the set is not empty, randomly select an offloading decision a' i of the task i from the set A', let a i = a' i , update the global offloading decision A = {a i | i∈I} and the global utility function value U = {u i | i∈I}, let the temporary set I' = I, the temporary set jump to step 4, otherwise, jump to step 9;

[0138] Step 9: The algorithm reaches Nash equilibrium, output the global offloading decision A, and the step ends.

[0139] The system, device, module or unit illustrated in the above embodiment can be specifically implemented by a computer chip or entity, or by a product with certain functions.

[0140] Computer-readable media includes permanent and non-permanent, movable and non-movable media that can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0141] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusions, so that a process, method, article or apparatus that includes a list of elements not only includes those elements, but also includes other elements not explicitly listed, or inherent to such a process, method, article or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0142] The above embodiments should be understood as merely illustrative of the present application and not restrictive of the scope of protection of the present application. After reading the description of the present application, those skilled in the art can make various modifications or changes to the present application, and these equivalent changes and modifications also fall within the scope defined by the claims of the present application.

Claims

1. A method for unloading privacy-sensitive vehicle-side collaborative tasks, characterized in that, Includes the following steps:

101. An optimization model for task offloading in the Internet of Vehicles (IoV) is established using the exact potential game method. The system includes an edge server k and a task vehicle j. * And several cooperating vehicles J = {j}, let I = {i} be the vehicles j * The proposed set of concurrent tasks, where each task i is a game participant, a i For the unloading decision of task i, u i Let A = {a} be the utility function value for task i. i | i∈I } represents the global unloading decision, U = {u i | i∈I } represents the global utility function value, and the exploration threshold γ is initialized in the real number interval [0,1]. th ; 102. Based on the data volume s of each task i in set I. i Computational complexity c i Maximum tolerable delay For each task i, establish a candidate unloading node set N. i ; 103. For each task i participating in the game in the task set I, remove it from its candidate unloading node set N. i Randomly select a node to construct the unloading decision a i According to the global unloading decision A = {a i | i∈I } Calculate the global utility function value U = {u i | i∈I Let temporary set I′=I, temporary set 104. If Take the first element i from set I′ and generate a random number γ in the real number interval [0,1]. Jump to step 105; otherwise, jump to step 108.

105. If γ < γ th From the candidate unloading node set N of task i i Randomly select a candidate unloading node and calculate the unloading decision a′. i and utility function value u′ i If yes, proceed to step 106; otherwise, proceed to step 107.

106. If u′ i >u i The unloading decision a′ for task i i Add to set A′ and jump to step 104; otherwise, jump to step 104.

107. From the candidate unloading node set N of task i i Iterate through all candidate unloading nodes and calculate the corresponding unloading decision a′ for each candidate unloading node. i and utility function value u′ i And select the largest utility function value u′ from them. i and its corresponding unloading decision a′ i If a′ i ≠a i The corresponding unloading decision a′ i Add to set A′ and jump to step 104; otherwise, jump to step 104.

108. If set Randomly select an unloading decision a′ for task i from set A′. i Let a i =a′ i Update the global uninstallation decision A = {a i | i∈I } and global utility function value U={u i | i∈I Let temporary set I′=I, temporary set Skip to step 104; otherwise, skip to step 109.

109. The algorithm reaches Nash equilibrium and outputs the global unloading decision A, thus ending the process.

2. The method for offloading privacy-sensitive vehicle-side collaborative tasks according to claim 1, characterized in that, In step 102, a candidate unloading node set N is established for task i. i Specifically, it includes the following steps: 1) Initialize the set of candidate unload nodes for task i Temporary assembly Based on the vehicle's effective communication range, the mission vehicle j * Cooperative vehicles within a one-hop range are added to set J′; 2) If Take the first element j from set J′ and calculate the task vehicle j in V2V unloading mode. * Link duration with collaborating vehicle j Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay Calculation in The computing resources f that collaborative vehicle j needs to allocate for task i within the specified time period. i j If yes, proceed to step 3; otherwise, proceed to step 4. 3) If vehicle j has remaining available computing resources f j Greater than or equal to f i j Add the cooperating vehicle j to the candidate unloading node set N. i If yes, proceed to step 2); otherwise, proceed to step 2. 4) Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay In the local unloading mode, the task vehicle j * Local computing resources to be allocated for task i If the mission vehicle j * Remaining available computing resources Greater than or equal to The mission vehicle j * Add to candidate unload node set N i If yes, proceed to step 5; otherwise, proceed to step 5. 5) Calculate the task vehicle j in V2I unloading mode. * Link duration with edge server k in the current cell Based on the data volume s of task i i Computational complexity c i Maximum tolerable delay Calculation in Within a given timeframe, the edge server k in the current cell needs to allocate computing resources f to task i. i k ; 6) If the remaining available computing resources f of the edge server k in the current cell are... k Greater than or equal to f i k Add the edge server k of the current cell to the candidate offload node set N. i If yes, proceed to step 7); otherwise, proceed to step 7. 7) Output the set of candidate unloading nodes N i The steps are now complete.

3. The method for unloading privacy-sensitive vehicle-side collaborative tasks according to claim 2, characterized in that, In step 2), under the V2V unloading mode, the task vehicle j * Link duration with collaborating vehicle j The calculation method is shown in formula (1), where the computing resources f to be allocated by the cooperating vehicle j for task i are... i j The calculation method is shown in formula (2): In formula (1), Indicates the mission vehicle j * The driving speed of the task vehicle j is given. * The direction of travel is positive, and v j V represents the speed of the cooperating vehicle j. j >0 indicates that vehicle j and vehicle j * Traveling in the same direction, v j <0 indicates that vehicle j and vehicle j * Reverse driving, R V2V This indicates the maximum communication distance between vehicles. Indicates the mission vehicle j * Position, x j This indicates the location of the collaborating vehicle j. In V2V offloading mode, it is determined based on the task vehicle j. * By analyzing the position and speed relationship between the cooperating vehicle j and the other vehicle j, the duration of the communication link between them can be calculated. In formula (2), s i c represents the amount of data for task i. i Let represent the computational complexity of task i. r represents the maximum tolerable delay for task i. j Indicates the mission vehicle j * The data transmission rate between the vehicle and the cooperating vehicle j is calculated as shown in formula (3): In formula (3), B V2V The bandwidth of the V2V link is represented by q, the transmit power of the mission vehicle is represented by g. j Indicates the mission vehicle j * V2V link channel gain between the cooperating vehicle j and the vehicle j, N0 represents noise power, L j Indicates the mission vehicle j * Path loss between the vehicle and the cooperating vehicle j.

4. The method for offloading privacy-sensitive vehicle-side collaborative tasks according to claim 2, characterized in that, In step 4), under the local unloading mode, the task vehicle j * Local computing resources to be allocated for task i The calculation method is shown in formula (4):

5. A method for offloading privacy-sensitive vehicle-side collaborative tasks according to claim 2, characterized in that, In step 5), under V2I unloading mode, the task vehicle j * Link duration with edge server k in the current cell The calculation method is shown in formula (5). The edge server k of the current cell needs to allocate computing resources f for task i. i k The calculation method is shown in formula (6): In formula (5), R V2I x represents the effective communication radius of the RSU. k This indicates the location of the edge server k in the current cell; In formula (6), r k Indicates the mission vehicle j * The data transmission rate between the current cell and the edge server k is calculated as shown in formula (7): In formula (7), B V2I G represents the bandwidth of the V2I link. k Indicates the mission vehicle j * The V2I link channel gain, L, between the current cell's edge server k and the edge server k is... k Indicates the mission vehicle j * The path loss between the current edge server k in the current cell.

6. The method for offloading privacy-sensitive vehicle-side collaborative tasks according to claim 1, characterized in that, In step 103, the unloading decision a for task i is constructed. i The method is as shown in formula (8): a i =(n i ,f i ) (8) In formula (8), n i The candidate unloading node selected for task i is calculated as shown in formula (9): In formula (9), f i This represents the computing resources obtained by task i, calculated as shown in formula (10):

7. A method for unloading privacy-sensitive vehicle-side collaborative tasks according to claim 1, characterized in that, In step 103, the global utility function value U = {u} is obtained. i | i∈I The method is as shown in formula (11): In formula (11), i∈I, i′∈I, and i≠i′, a i and a i′ Let a represent the unloading decision for task i or task i′, respectively. -i and a -i′ Let h represent the unloading decision sets for all tasks in set I except task i or task i′. i (a i ,a -i ) and h i′ (a i′ ,a -i′ ) represent the unloading benefit of task i or task i′ when all tasks in set I participate in unloading, respectively. -i′\i h represents the set of unloading decisions for all tasks other than task i′ in set I when only task i is not involved in unloading. i′ (a i′ ,a -i′\i ) represents the unloading benefit of task i′ when only task i in set I does not participate in the unloading, h i (a i ,a -i ) and h i′ (a i′ ,a -i′ The calculation method for h is shown in formula (12), where task i or task i′ represents any two different tasks in set I, and h i′ (a i′ ,a -i′\i The calculation method for ) is shown in formula (13): In formulas (12) and (13), θ is the weighting factor. i θ represents the completion status of task i when all tasks in set I participate in unloading. If task i completes within the maximum tolerable delay, then... i =1, otherwise, θ i =0, θ i′\i θ represents the completion status of task i′ when only task i in set I is not involved in unloading. If task i′ completes within the maximum tolerable delay, θ i′\i =1, otherwise, θ i′\i =0; In formula (12), p i This represents the privacy protection level of task i when all tasks in set I participate in the unloading process. The calculation method is shown in formula (14): In formula (14), Indicates whether task i adopts V2I offload mode. If task i is offloaded to the edge server k of the current cell, let Otherwise, let Indicates whether task i adopts V2V offload mode. If task i is offloaded to collaborating vehicle j, then... Otherwise, let |I| represents the total number of concurrent tasks, m i,i′ The elements in the privacy conflict matrix M are represented by the formula (15): In formula (13), p i′\i This represents the privacy protection level of task i′ when only task i in set I is not involved in the unloading process. The calculation method is shown in formula (16): In formula (16), task i″∈(Iii′) represents another task in set I besides task i and task i′.

8. A method for offloading privacy-sensitive vehicle-side collaborative tasks according to claim 1, characterized in that, The unloading decision a′ is calculated in steps 105 and 107. i and utility function value u′ i Specifically, it includes the following steps: 1) Let temporary set A″ = A; 2) Based on the candidate unloading nodes selected for task i, construct the unloading decision a′ using formula (8). i Let A″ be ​​the unloading decision for task i. i =a′ i According to A″, the utility function value u of task i is calculated using formula (11). i , let u′ i =u i ; 3) Output the unloading decision a′ and utility function value u′ for task i. i The steps are now complete.

9. An electronic device, characterized in that, The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the privacy-sensitive vehicle-side collaborative task offloading method as described in any one of claims 1 to 8.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the privacy conflict-sensitive vehicle-side collaborative task offloading method as described in any one of claims 1 to 8.

Images