A method for task offloading of non-uniform speed vehicle in complex road conditions based on MBS assistance

Abstract

The application claims a complex road condition non-uniform speed vehicle task offloading method based on MBS assistance, belonging to the field of communication technology. In view of the problem that the high time-varying nature of vehicle speed and trajectory in large traffic hubs, complex interchanges and other scenes may cause unstable vehicle communication link, thereby causing task offloading failure, a non-uniform speed vehicle task offloading method based on macro base station MBS assistance is proposed. In the method, MBS finds a candidate offloading vehicle set for each sub-task according to the offloading request of the task vehicle and the computing resource distribution of the vehicle network, selects a local or V2V or V2I2V offloading mode for each sub-task according to the vehicle link state change law, establishes a corresponding candidate offloading path set, and constructs a feasible offloading decision, and iteratively optimizes the offloading decision through the simulated annealing algorithm, thereby effectively improving the task completion rate and reducing the system unit task delay.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of communication network, and particularly relates to a complex road non-uniform speed vehicle task offloading method based on MBS assistance. BACKGROUND

[0002] In recent years, cellular vehicle to everything (C-V2X) technology has been widely applied in the field of intelligent networked vehicles due to its excellent performance in communication range and anti-interference. C-V2X supports vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communication through direct communication and cellular communication, enabling vehicles to offload tasks to other vehicles through direct communication or macro base station forwarding.

[0003] Existing research on vehicle edge network computing offloading is mainly based on scenarios with constant vehicle speed and direction, and the scheduling and resource allocation strategies for offloading tasks are difficult to meet the needs of actual application scenarios. For example, in large traffic hubs and complex interchanges, changes in vehicle speed and trajectory may cause rapid and disordered changes in network connectivity, leading to a decrease in network resource utilization and a deterioration in user service experience.

[0004] To solve the above problems, the application provides a complex road non-uniform speed vehicle task offloading method based on MBS assistance. In this method, the macro base station (MBS) finds a candidate offloading vehicle set for each sub-task based on the offloading request of the task vehicle and the distribution of available computing resources in the vehicle network. According to the change law of the link state between the task vehicle and the candidate offloading vehicle, the local (Local) or vehicle to vehicle (V2V) or vehicle to infrastructure to vehicle (V2I2V) offloading mode is selected for each sub-task, and the corresponding candidate offloading path set is established to construct a feasible offloading decision, and the simulated annealing algorithm is used for iterative optimization, thereby effectively improving the task completion rate and reducing the system unit task delay. SUMMARY

[0005] The application aims to solve the problems of the prior art. A complex road non-uniform speed vehicle task offloading method based on MBS assistance is proposed. The technical solution of the application is as follows:

[0006] A complex road non-uniform speed vehicle task offloading method based on MBS assistance includes the following steps:

[0007] 101, a simulated annealing algorithm is used to establish a vehicle network task offloading optimization model, and a task vehicle i* the offloading decision vector and the optimal objective function value are ψ = (ψ j | j∈J ), t * , where ψ j is the offloading path of the subtask j, the initial temperature T cur , the termination temperature T end , the temperature cooling coefficient ε and the maximum iteration number η max of the offloading decision vector, let the iteration count variable η = 0 and the temporary set J' = J;

[0008] 102, if the first element j is taken out from the set J', according to the data volume a j , the computational complexity c j of the subtask j, the candidate offloading path set P j = {p j,k} is established, where p j,k is a candidate offloading path of the subtask j with the code k, jump to step 102, otherwise, jump to step 103;

[0009] 103, a path p j | j∈J is randomly selected from each set P j,k initialize the offloading decision vector the objective function value t is calculated, the objective function value t cur of the current solution is initialized as t = t, the objective function value t * of the optimal solution is initialized as t = t, the optimal offloading decision ψ best is initialized as ψ, let the temporary set J' = J' - {j}, and jump to step 104, otherwise, jump to step 108;

[0010] 104, let the iteration count variable η = η + 1, if η ≤ η max , let the temporary set J'' = J - J', and jump to step 105, otherwise, jump to step 108; the count variable where ω is the proportionality coefficient, jump to step 105, otherwise, jump to step 106;

[0011] 105, let n = n - 1, if n ≥ 0, a subtask j is randomly taken out from the set J'' to join the temporary set J', and a path p j,k is randomly selected from the candidate offloading path set P j , let the offloading path ψ j of the subtask j = p j,k , jump to step 105, otherwise, jump to step 106;

[0012] 106, the objective function value t is calculated, if t < t* , let the objective function value of the current solution t cur = t, the objective function value of the optimal solution t * = t, the optimal offloading decision ψ best = ψ, jump to step 104, otherwise, jump to step 107;

[0013] 107, calculate the receiving probability of the offloading decision vector ψ and generate a random number If , let the objective function value of the current solution t cur = t, the objective function value of the optimal solution t * = t, the optimal offloading decision ψ best = ψ, jump to step 104, otherwise, jump to step 104;

[0014] 108, update the current temperature T cur = ε·T cur , if T cur > T end , let the iteration count variable η = 0, jump to step 104, otherwise, jump to step 109;

[0015] 109, output the optimal offloading decision ψ best , the step ends.

[0016] Further, the step 102 establishes a candidate offloading path set P j = {p j,k} for the subtask j, specifically including the following steps:

[0017] 1) According to the vehicle network state, initialize the whole vehicle set I = {i}, where i is any vehicle, and the candidate offloading vehicle set The candidate offloading path set of the subtask j The maximum offloading tolerant time delay t max of the task;

[0018] 2) If , take the first element i from the set I, according to the data volume a j , the computational complexity c j , calculate the minimum computing resource that the vehicle i needs to allocate for the subtask j within t max time Jump to step 3), otherwise, jump to step 4);

[0019] 3) If the remaining available computing resource f i r of the vehicle i is greater than or equal to Add vehicle i to the candidate unloading vehicle set I′ and proceed to step 2); otherwise, proceed to step 2.

[0020] 4) If Take the first vehicle i from set I′, and calculate the computing resources f that candidate unloading vehicle i needs to allocate for subtask j under fair allocation of onboard resources. Fr Let candidate unloading vehicle i be allocated computing resources f for subtask j. i,j =∞, jump to step 5); otherwise, jump to step 11.

[0021] 5) If i = i * Let the unloading mode be `model = Local`, and let candidate unloading vehicle `i` be the computing resource allocated to subtask `j`. If the previous step is incorrect, proceed to step 10; otherwise, proceed to step 6.

[0022] 6) If the mission vehicle i * Distance d between candidate unloading vehicle i and i*,i If the effective communication distance D of the vehicle is less than or equal to that of the vehicle, and the unloading mode model = V2V, calculate the task vehicle i. * Link duration τ with candidate unloading vehicle i i*,i Based on the upload latency of subtask j to candidate unloading vehicle i in V2V unloading mode Estimated backhaul delay of candidate unloading vehicle i along the original route Where r is the link transmission rate, σ j The ratio of the uploaded to the returned data volume of subtask j is calculated in... Minimum computing resources required for subtask j for candidate unloading vehicle i within the given time period If the previous step is incorrect, proceed to step 7; otherwise, proceed to step 8.

[0023] 7) If Let candidate unloading vehicle i be the computing resources allocated to subtask j. If the previous step is incorrect, proceed to step 8; otherwise, proceed to step 8.

[0024] 8) Let the unloading mode model = V2I2V, and calculate the upload latency of subtask j to candidate unloading vehicle i in the V2I2V unloading mode. Estimated backhaul delay of candidate unloading vehicle i along the original route Calculate at t max Minimum computing resources required for subtask j for candidate unloading vehicle i within the given time period if If the previous step is incorrect, proceed to step 9; otherwise, proceed to step 10.

[0025] 9) If Let candidate unloading vehicle i be the computing resources allocated to subtask j. If the previous step is not successful, proceed to step 10; otherwise, proceed to step 10.

[0026] 10) If f i,j ≠∞, f i,j The corresponding candidate unloading path p j,k Add to candidate unloading path set P j If yes, proceed to step 4); otherwise, proceed to step 4.

[0027] 11) Output the set of candidate unloading paths P j The steps are now complete.

[0028] Furthermore, in step 2), at t max Minimum computing resources that vehicle i needs to allocate for subtask j within a given time period The calculation method is shown in formula (1):

[0029]

[0030] In formula (1), a j c represents the amount of data in subtask j. j t represents the computational complexity of subtask j. max This indicates the maximum tolerable unloading delay for the task.

[0031] Furthermore, in step 4), the candidate unloading vehicle i needs to allocate computing resources f for subtask j under the fair allocation of onboard resources. Fr The calculation method is shown in formula (2):

[0032]

[0033] In formula (2), f i Represents the total computing resources of vehicle i, |I * | represents the total number of vehicles on the mission, |J max This represents the maximum number of subtasks that each unload request can contain, and |I| represents the total number of vehicles.

[0034] Furthermore, in step 6), the task vehicle i * Distance between candidate unloading vehicle i Link duration τ i*,i and in Minimum computing resources required for subtask j for candidate unloading vehicle i within the given time period The values ​​are obtained by formulas (3), (4), and (5), respectively:

[0035]

[0036] In formula (3), (x i ,y i ) respectively represent the position coordinates of the current task vehicle i * and the candidate offloading vehicle i, in formula (4), D represents the effective communication distance of the vehicle, μ max represents the maximum driving speed of the vehicle, γ i respectively represent the angles between the driving directions of the task vehicle i * and the candidate offloading vehicle i and the positive direction of the x-axis after the change of direction, and the calculation method is shown in formula (6):

[0037]

[0038] In formula (6), α i respectively represent the angles between the current driving directions of the task vehicle i * and the candidate offloading vehicle i and the positive direction of the x-axis, and β represents the angle change amount when the task vehicle i * and the candidate offloading vehicle i change from the current driving direction to the back driving direction, and the calculation method is shown in formula (7):

[0039]

[0040] In formula (7), θ represents the instantaneous maximum steering angle of the vehicle.

[0041] Further, in the V2I2V offloading mode in step 8), the calculation method of the minimum calculation resources max that the candidate offloading vehicle i needs to allocate to the subtask j within t is shown in formula (8):

[0042]

[0043] Further, the calculation method of the target function value t in step 103 is shown in formula (9):

[0044]

[0045] In formula (9), represents the offloading completion delay of the subtask j along the path p j,k to the candidate offloading vehicle i, and the calculation method is shown in formula (10), is a binary variable, which represents whether the subtask j is offloaded along the path p j,k to the last node vehicle i of p j,k , if yes, let otherwise, let The calculation method is shown in formula (11):

[0046]

[0047] In formula (9), denotes the subtask j along the path p j,k The upload time delay of the transmission to the candidate offloading vehicle i, denotes the subtask j along the path p j,k The return time delay of the original path return result.

[0048] Further, the calculation method of the receiving probability of the offloading decision vector ψ in the step 107 The calculation method is shown in formula (12):

[0049]

[0050] The advantages and beneficial effects of the present application are as follows:

[0051] The present application discloses a complex road non-uniform speed vehicle task offloading method based on MBS assistance. The existing research on vehicle network task offloading problem usually only considers that vehicles travel at a constant speed and direction, and does not consider the problem of real-time changes of vehicle speed and direction in large traffic hubs or complex interchange environments. In view of the above problems, the present application provides a complex road non-uniform speed vehicle task offloading method based on MBS assistance. In the method, the MBS finds a candidate offloading vehicle set for each subtask according to the offloading request of the task vehicle and the distribution of available computing resources in the vehicle network, selects a local or V2V or V2I2V offloading mode for each subtask according to the link connection state of the task vehicle and the candidate offloading vehicle, establishes a corresponding candidate offloading path set, constructs a feasible offloading decision, and iteratively optimizes the offloading decision through a simulated annealing algorithm, thereby effectively improving the task completion rate and reducing the system unit task time delay. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 It is a flow chart of a complex road non-uniform speed vehicle task offloading method based on MBS assistance provided by the present application. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present application will be described clearly and in detail below with reference to the drawings in the embodiments of the present application. The described embodiments are only a part of the embodiments of the present application.

[0054] The technical solutions of the present application to solve the above technical problems are:

[0055] The concepts and models involved in the content of the present application are as follows:

[0056] The present application assumes that each vehicle travels at an arbitrary speed and direction in a complex road network. Based on the effective communication distance constraint, tasks can be offloaded through V2V direct links or MBS assisted forwarding. Each vehicle is equipped with a server with certain computing capacity. Vehicle tasks are composed of multiple different types of sub-tasks, which can be offloaded to the local in parallel, or offloaded to other vehicles through direct links or macro base station assisted forwarding. Among them, the same task cannot be offloaded to two or more offloading vehicles.

[0057] The technical solutions of the present application are as follows:

[0058] 1. The minimum computing resource allocated by vehicle i for sub-task j in t max time

[0059] The calculation method is shown in formula (1):

[0060]

[0061] In formula (1), a j represents the data volume of sub-task j, c j represents the computing complexity of sub-task j, t max represents the maximum offloading tolerance delay of the task;

[0062] 2. The computing resource f allocated by candidate offloading vehicle i for sub-task j under the fair allocation of vehicle resources Fr

[0063] The calculation method is shown in formula (2):

[0064]

[0065] In formula (2), f i represents the total computing resource of vehicle i, |I * | represents the total number of task vehicles, |J max represents the maximum number of sub-tasks contained in each offloading request, and |I| represents the total number of all vehicles.

[0066] 3. The distance d * between task vehicle i and candidate offloading vehicle i, the link duration i*,i , and the minimum computing resource allocated by candidate offloading vehicle i for sub-task j in t time

[0067] The calculation methods are obtained by formula (3), formula (4) and formula (5) respectively:

[0068]

[0069] In formula (3), (x i ,y i ) respectively represent the position coordinates of the current task vehicle i * and the candidate offloading vehicle i, in formula (4), D represents the effective communication distance of the vehicle, μ max represents the maximum driving speed of the vehicle, γ i respectively represent the angle between the driving direction of the task vehicle i * and the candidate offloading vehicle i and the positive direction of the x-axis after the direction change, the calculation method is shown in formula (6), in formula (5), represents the upload time delay of the subtask j transmitted to the candidate offloading vehicle i in the V2V offloading mode, represents the expected return time delay of the candidate offloading vehicle i returning the result along the original route in the V2V offloading mode;

[0070]

[0071] In formula (6), α i respectively represent the angle between the current driving direction of the task vehicle i * and the candidate offloading vehicle i and the positive direction of the x-axis, β represents the angle change amount when the task vehicle i * and the candidate offloading vehicle i change from the current driving direction to the back driving direction, the calculation method is shown in formula (7):

[0072]

[0073] In formula (7), θ represents the instantaneous maximum steering angle of the vehicle;

[0074] 4. In the V2I2V offloading mode, the candidate offloading vehicle i needs to allocate the minimum calculation resource for the subtask j within t max time

[0075] The calculation method is shown in formula (8):

[0076]

[0077] In formula (8), represents the upload time delay of the subtask j transmitted to the candidate offloading vehicle i in the V2I2V offloading mode, represents the expected return time delay of the candidate offloading vehicle i returning the result along the original route in the V2I2V offloading mode;

[0078] 5. The target function value t

[0079] The calculation method is shown in formula (9):

[0080]

[0081] In formula (9), f i,j represents the computing resource allocated by the candidate offloading vehicle i for the subtask j, p j,k represents f i,j the corresponding candidate offloading path, represents the subtask j along the path p j,k the offloading completion time delay when offloaded to the candidate offloading vehicle i, the calculation method is shown in formula (10), is a binary variable, which represents whether the subtask j along the path p j,k is the last node vehicle i of p j,k , if yes, let otherwise, let the calculation method is shown in formula (11):

[0082]

[0083] In formula (10), t represents the upload time delay of the subtask j along the path p j,k transmitted to the candidate offloading vehicle i, represents the backhaul time delay of the subtask j along the path p j,k backhauling the result in the original path, in formula (11), ψ j represents the offloading path of the subtask j;

[0084] 6. The receiving probability of the offloading decision vector ψ

[0085] The calculation method is shown in formula (12):

[0086]

[0087] In formula (12), t cur represents the objective function value of the current solution, T cur represents the current temperature;

[0088] 7. Sub-algorithm 1: Establishing the candidate offloading path set P j for the subtask j j,k

[0089] Step 1: According to the vehicle network state, initialize the whole vehicle set I = {i}, wherein i is any vehicle, and the candidate offloading vehicle set the candidate offloading path set of the subtask j the maximum offloading tolerance time delay t max of the task;

[0090] ​Step 2: If Take the first element i from set I, and determine the data size a of subtask j. j Computational complexity c j Calculate at t max Minimum computing resources that vehicle i needs to allocate for subtask j within a given time period Skip to step 3; otherwise, skip to step 4.

[0091] Step 3: If vehicle i has remaining available computing resources f i r Greater than or equal to Add vehicle i to the candidate unloading vehicle set I′ and proceed to step 2; otherwise, proceed to step 2.

[0092] Step 4: If Take the first vehicle i from set I′, and calculate the computing resources f that candidate unloading vehicle i needs to allocate for subtask j under fair allocation of onboard resources. Fr Let candidate unloading vehicle i be allocated computing resources f for subtask j. i,j =∞, jump to step 5; otherwise, jump to step 11;

[0093] Step 5: If i = i * Let the unloading mode be `model = Local`, and let candidate unloading vehicle `i` be the computing resource allocated to subtask `j`. Skip to step 10; otherwise, skip to step 6.

[0094] Step 6: If the mission vehicle i * Distance d between candidate unloading vehicle i and i*,i If the effective communication distance D of the vehicle is less than or equal to that of the vehicle, and the unloading mode model = V2V, calculate the task vehicle i. * Link duration τ with candidate unloading vehicle i i*,i Based on the upload latency of subtask j to candidate unloading vehicle i in V2V unloading mode Estimated backhaul delay of candidate unloading vehicle i along the original route Where r is the link transmission rate, σ j Let be the ratio of the uploaded to the returned data volume of subtask j, calculated in min{t max ,τ i*,i The minimum amount of computing resources that candidate unloading vehicle i needs to allocate for subtask j within the specified time. Skip to step 7; otherwise, skip to step 8.

[0095] Step 7: If Let candidate unloading vehicle i be the computing resources allocated to subtask j. Go to step 8, otherwise, go to step 8;

[0096] Step 8: Let the offloading mode model = V2I2V, according to the upload time delay of subtask j transmitted to the candidate offloading vehicle i in the V2I2V offloading mode The predicted return time delay of the candidate offloading vehicle i along the original route to return the result Calculate the minimum computing resources that the candidate offloading vehicle i needs to allocate to the subtask j within t max Time If Go to step 9, otherwise, go to step 10;

[0097] Step 9: If Let the computing resources allocated by the candidate offloading vehicle i to the subtask j Go to step 10, otherwise, go to step 10;

[0098] Step 10: If f i,j ≠∞, f i,j The corresponding candidate offloading path p j,k is added to the candidate offloading path set P j , go to step 4, otherwise, go to step 4;

[0099] Step 11: Output the candidate offloading path set P j , and the step ends.

[0100] A complex road condition non-uniform speed vehicle task offloading method based on MBS assistance, specifically comprising the following steps:

[0101] Step 1: Use the simulated annealing algorithm to establish a vehicle network task offloading optimization model, set the task vehicle i * 's offloading request J = {j}, where j is a parallel subtask, the offloading decision vector and the optimal objective function value are ψ = (ψ j | j∈J ), t * , where ψ j is the offloading path of subtask j, the current temperature T cur , the termination temperature T end , the temperature cooling coefficient ε and the maximum iteration number η of the offloading decision vector are initialized max , let the iteration variable η = 0, and the temporary set J' = J;

[0102] Step 2: If Take the first element j from the set J', according to the data volume a j , the computing complexity c j , call sub-algorithm 1 to establish the candidate offloading path set P j={p j,k}, where p j,k If a candidate unloading path with code k is found for subtask j, proceed to step 2; otherwise, proceed to step 3.

[0103] Step 3: From each set P j | j∈J Choose any path p j,k Initialize the unloading decision vector Calculate the objective function value t, and initialize the objective function value t of the current solution. cur =t, the objective function value of the optimal solution t * =t, optimal unloading decision ψ best =ψ, let the temporary set be ψ.

[0104] Step 4: Let the iteration count variable η = η + 1. If η ≤ η max Let the temporary set J″ = JJ′, Count variable Where ω is the proportionality coefficient, skip to step 5; otherwise, skip to step 8.

[0105] Step 5: Let n = n-1. If n ≥ 0, randomly select subtask j from set J″ and add it to temporary set J′, and select it from candidate unloading path set P. j Randomly select path p j,k Let subtask j's unloading path ψ j =p j,k If yes, proceed to step 5; otherwise, proceed to step 6.

[0106] Step 6: Calculate the objective function value t. If t < t * Let the objective function value of the current solution be t. cur =t, the objective function value of the optimal solution t * =t, optimal unloading decision ψ best =ψ, then proceed to step 4; otherwise, proceed to step 7.

[0107] Step 7: Calculate the acceptance probability of the unloading decision vector ψ And generate random numbers if Let the objective function value of the current solution be t. cur =t, the objective function value of the optimal solution t * =t, optimal unloading decision ψ best =ψ, then proceed to step 4; otherwise, proceed to step 4.

[0108] Step 8: Update the current temperature T cur =ε·T cur If Tcur T end set the iteration count variable n = 0, jump to step 4, otherwise, jump to step 9;

[0109] Step 9: output the optimal unloading decision best Step end.

[0110] The system, apparatus, module or unit illustrated in the above embodiments can be specifically implemented by a computer chip or entity, or by a product with certain functions.

[0111] It should also be noted that the terms "comprising", "including", or any other variant thereof are intended to cover non-exclusive inclusion, such that processes, methods, articles or devices that comprise a series of elements are not limited to those elements, but can include other elements not expressly listed, or inherent to such processes, methods, articles or devices. 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 device that includes the element.

[0112] The above embodiments are understood as merely illustrating the present application and not limiting the protection scope of the present application. After reading the content of the present application, the 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 MBS-assisted complex road non-uniform speed vehicle task offloading, characterized in that, The method comprises the following steps: 101、Using the simulated annealing algorithm to establish a vehicle network task offloading optimization model, set the task vehicle i * 's offloading request J={j}, where j is a parallel subtask, the offloading decision vector and the optimal objective function value are ψ=(ψ j | j∈J ), t * , where ψ j is the offloading path of subtask j, the initial temperature T cur , the termination temperature T end , the temperature cooling coefficient ε and the maximum iteration number η of the offloading decision vector max , let the iteration count variable η=0, and the temporary set J'=J; 102、If Take the first element j from the set J', according to the data volume a j , the computational complexity c j , of the subtask j, establish a candidate offload path set P j ={p j,k}, where p j,k is a candidate offload path of the subtask j with the label k, jump to step 102, otherwise, jump to step 103; 103. Selecting one path p from each set P j | j∈J arbitrarily j,k Initializing the offloading decision vector Computing the objective function value t, initializing the objective function value t of the current solution cur = t, the objective function value t of the optimal solution * = t, the optimal offloading decision ψ best = ψ, setting the temporary set 104. Let the iteration counter variable η = η + 1, if η < η max , let the temporary set J" = J - J', counter variable where ω is a proportionality coefficient, jump to step 105, otherwise, jump to step 108; 105. Let n = n-1. If n ≥ 0, randomly select subtask j from set J″ and add it to temporary set J′, and select it from candidate unloading path set P. j Randomly select path p j,k Let subtask j's unloading path ψ j =p j,k If yes, proceed to step 105; otherwise, proceed to step 106.

106. Calculate the objective function value t, if t < t * , let the objective function value t cur of the current solution t * = t, the optimal unloading decision ψ best = ψ, jump to step 104, otherwise, jump to step 107; 107. Calculate the reception probability of the offloading decision vector ψ and generate a random number If Let the objective function value of the current solution t cur = t, the objective function value of the optimal solution t * = t, the optimal offloading decision ψ best = ψ, jump to step 104, otherwise, jump to step 104; 108. Update the current temperature T cur = ε · T cur If T cur > T end , let the iteration count variable η = 0, jump to step 104, otherwise, jump to step 109; 109. outputting the optimal offloading decision ψ best , the step ends.

2. The method of claim 1, wherein, The step 102 of establishing a candidate offloading path set P for a subtask j j = {p j,k}, specifically comprising the following steps: 1) Initialize the set of all vehicles I = {i}, where i is any vehicle, according to the vehicle network status, and the set of candidate offloading vehicles The set of candidate offloading paths for subtask j The maximum offloading tolerant time delay t for task max ; 2) If Take the first element i from the set I, and compute the minimum computing resources that vehicle i needs to allocate to sub-task j in time t j , the data volume a j , the computational complexity c max , the computation time t Go to step 3), otherwise, go to step 4); 3) If the remaining available computing resources of vehicle i are greater than or equal to add vehicle i to the set of candidate offloading vehicles I', go to step 2), otherwise, go to step 2); 4) If Take the first vehicle i from the set I', calculate the computing resource f that vehicle i needs to allocate to the subtask j under the fair allocation of on-board resources Fr Let the computing resource f that vehicle i allocates to the subtask j be f i,j = ∞, jump to step 5), otherwise, jump to step 11); 5) if i = i * , let the offloading mode model = Local, the candidate offloading vehicle i is assigned to the computing resource of the subtask j jump to step 10), otherwise, jump to step 6); 6) If task vehicle i * Distance between task vehicle i and candidate offloading vehicle i is less than or equal to the effective communication distance D of vehicles, let offloading mode model = V2V, calculate the task vehicle i * Link duration between task vehicle i and candidate offloading vehicle i According to the upload time delay of subtask j transmitted to candidate offloading vehicle i in V2V offloading mode The predicted return time delay of candidate offloading vehicle i along the original route to return the result Wherein, r is the link transmission rate, σ j is the ratio of the upload and return data volume of subtask j, calculate the computing resources that candidate offloading vehicle i needs to allocate for subtask j in Time Jump to step 7), otherwise, jump to step 8); 7) if Let candidate offloading vehicle i be assigned computing resources for subtask j Go to step 8), otherwise, go to step 8); 8) Let the offloading mode model = V2I2V, according to the upload time delay of subtask j transmitted to the candidate offloading vehicle i in the V2I2V offloading mode The predicted return time delay of the candidate offloading vehicle i along the original route to return the result Calculate the computing resources that the candidate offloading vehicle i needs to allocate to the subtask j at least in the time t max If Jump to step 9), otherwise, jump to step 10);​ 9) if Let candidate offloading vehicle i be assigned computing resources for subtask j Go to step 10), otherwise, go to step 10); 10) If f i,j ≠∞, f i,j The corresponding candidate unloading path p j,k Add to candidate unloading path set P j If yes, proceed to step 4); otherwise, proceed to step 4. 11) output the set of candidate offload paths P j The step ends.

3. The method of claim 2, wherein, The minimum computing resources that the vehicle i needs to allocate for the subtask j in t max time The computing method is shown as formula (1): In formula (1), a j represents the data volume of the subtask j, c j represents the computing complexity of the subtask j, t max represents the maximum offloading tolerance time delay of the task.

4. The method of claim 2, wherein, The computing method of the candidate offloading vehicle i under the fair allocation of the on-vehicle resources in step 4) for the computing resources f allocated by the candidate offloading vehicle i for the subtask j Fr The computing method is shown as formula (2): In formula (2), f i represents the total computing resources of vehicle i, |I * | represents the total number of vehicles, |J| max represents the maximum number of sub-tasks contained in each unloading request, |I| represents the total number of vehicles.

5. The method of claim 2, wherein, the task vehicle i in the step 6) * the distance between the candidate offloading vehicle i and the task vehicle link duration and in the minimum computing resources that the candidate offloading vehicle i needs to allocate to the sub-task j within the time the values of the above are calculated by formula (3), formula (4) and formula (5) respectively: In formula (3), (x i ,y i ) respectively represent the position coordinates of the current task vehicle i * and the candidate unloading vehicle i, in formula (4), D represents the effective communication distance of the vehicle, μ max represents the maximum driving speed of the vehicle, γ i respectively represent the angles between the driving directions of the task vehicle i * and the candidate unloading vehicle i and the positive direction of the x-axis after the direction change, and the calculation method is shown in formula (6): In formula (6), α i respectively represent the current driving direction of the task vehicle i * and the candidate unloading vehicle i, and β represents the angle between the current driving direction of the task vehicle i * and the candidate unloading vehicle i when the candidate unloading vehicle i changes from the current driving direction to the backward driving direction, and the calculation method is shown in formula (7): In formula (7), θ represents the instantaneous maximum steering angle of the vehicle.

6. The method of claim 2, wherein, in step 8) in the V2I2V offloading mode, t max the minimum computing resources that the candidate offloading vehicle i needs to allocate to the subtask j within the time The computing method is shown as formula (8):

7. The method of claim 1, wherein, The calculation method of the target function value t in the step 103 is shown in formula (9): In formula (9), denotes whether subtask j is along path p j,k The unloading completion delay when unloading to candidate unloading vehicle i, the calculation method is shown in formula (10), is a binary variable, which denotes whether subtask j is along path p j,k The last node vehicle i of p j,k , if yes, let Otherwise, let The calculation method is shown in formula (11): In formula (9), denotes the subtask j along the path p j,k the upload latency to the candidate offloading vehicle i, denotes the subtask j along the path p j,k the return latency of the original route return result.

8. The method of claim 1, wherein, The receiving probability of the unloading decision vector ψ in the step 107 The calculation method of the probability is shown in formula (12):

Images