Task offloading method and device for task vehicle in cellular vehicle networking

Abstract

The application provides a task vehicle task offloading method and device in cellular vehicle networking, belongs to the technical field of cellular vehicle networking, and aims to minimize the total task processing delay to establish a task offloading optimization model; the task offloading optimization model is solved to obtain the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle, and the optimal offloading ratio matrix is calculated; and the tasks offloaded to the roadside unit server and the auxiliary computing vehicle are distributed according to the optimal offloading ratio matrix. The application considers the distributed transmission and resource semi-persistent scheduling characteristics, models the task offloading optimization model, the task vehicle can offload the task to the auxiliary computing vehicle through the PC5 interface or offload the task to the VEC server through the Uu interface, and then the task processing delay is reduced; without dense base station coverage, the complexity and cost of task offloading and resource allocation are reduced, and the adaptability is high.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of cellular vehicle-to-everything (C-V2X) technology, and particularly relates to a task offloading method and device for a task vehicle in a C-V2X system. BACKGROUND

[0002] With the increase in the number of vehicles and the development of intelligent transportation services, the vehicle-to-everything (C-V2X) system puts forward higher requirements for system computing performance. The vehicular edge computing (VEC) system can configure a server of a road side unit (RSU) or other vehicles at the edge side to provide computing for vehicles with insufficient computing capability. In the VEC system, a vehicle can offload tasks to a VEC server or other vehicles for auxiliary computing to reduce task processing delay. At present, orthogonal resource allocation of a Uu interface requires a large number of base station coverages. In fact, due to the high mobility and uncertainty of C-V2X, it is difficult to ensure that vehicles are always within the coverage of a base station. Therefore, communication between vehicles is mainly realized through a PC5 interface. In the C-V2X system, due to the differentiated characteristics of the PC5 interface and the Uu interface, that is, the Uu interface is central orthogonal resource allocation, and the PC5 interface is distributed resource allocation, resource collision occurs, which affects task transmission and computing. In summary, the existing technology has the following disadvantages: the current communication mode of the C-V2X edge computing system mostly adopts an orthogonal access scheme, such as an orthogonal multiple access scheme, which requires a large number of base station coverages, and lacks an edge computing offloading strategy in the case of no base station coverage, and has the disadvantages of high complexity, high cost, and poor adaptability. SUMMARY

[0003] The present application aims to provide a task offloading method and device for a task vehicle in a C-V2X system, which considers the transmission characteristics of a PC5 interface and reduces task processing delay, so as to solve at least one of the technical problems in the background art.

[0004] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0005] In one aspect, the present application provides a task offloading method for a task vehicle in a C-V2X system. In the C-V2X system, the task vehicle is connected to a road side unit server through a Uu interface, and the road side unit server is connected to a vehicular edge computing system server. The task vehicle is connected to a plurality of auxiliary computing vehicles through a PC5 interface. The method comprises the following steps:

[0006] Defining an offloading ratio matrix of task offloading to the auxiliary computing vehicles and the road side unit server, and establishing a task offloading optimization model with the objective of minimizing the total task processing delay.

[0007] solving the task offloading optimization model, obtaining the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle;

[0008] According to the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle, an optimal offloading ratio matrix is calculated.

[0009] According to the optimal offloading ratio matrix, the tasks offloaded to the roadside unit server and the auxiliary computing vehicle are distributed.

[0010] Preferably, for the tasks offloaded to the auxiliary computing vehicle, the evaluation set is defined as the product of the successful transmission probability and the maximum computing frequency of the auxiliary computing vehicle, and then the evaluation set is used to determine the order and sequence of task offloading among the auxiliary vehicles.

[0011] Preferably, the auxiliary computing vehicles are sorted in descending order in the evaluation set, and the task vehicle preferentially offloads tasks to the auxiliary computing vehicles with high order. When the computing capacity of the current auxiliary computing vehicle exceeds the limit, the task vehicle offloads tasks to the auxiliary computing vehicle with the next order for further computation.

[0012] Preferably, the processing delay of each task includes two parts, one part is the sum of the offloading delay of the task vehicle offloading tasks to the roadside unit server and the computing delay of the roadside unit server for the task, and the other part is the sum of the offloading delay of the task vehicle offloading tasks to the auxiliary computing vehicle and the computing delay of the auxiliary computing vehicle for the task, and the larger sum is taken as the total processing delay of the task.

[0013] Preferably, the calculation of the offloading delay of the task vehicle offloading tasks to the roadside unit server includes: calculating the task transmission rate between the task vehicle and the roadside unit server; and calculating the offloading delay of the task from the task vehicle to the roadside unit server according to the calculated task transmission rate between the task vehicle and the roadside unit server and in combination with the task offloading ratio matrix.

[0014] Preferably, the calculation of the offloading delay of the task vehicle offloading tasks to the auxiliary computing vehicle includes:

[0015] According to the data transmission path loss from the task vehicle to the auxiliary computing vehicle, the sensing power threshold of the auxiliary computing vehicle, and the data transmission power from the task vehicle to the auxiliary computing vehicle, the data packet sensing probability of the auxiliary computing vehicle is calculated.

[0016] According to the sensing coefficient, the sensing distance of the multi-lane, and the number of subframes and subchannels in the selection window in the semi-persistent scheduling algorithm, the successful transmission probability of the task from the task vehicle to the auxiliary computing vehicle is calculated.

[0017] According to the calculated data packet perception probability and the task successful transmission probability, a normalized transmission rate between the task vehicle and the auxiliary computing vehicle is calculated;

[0018] According to the normalized transmission rate, an offloading delay of the task vehicle unloading the task to the auxiliary computing vehicle is calculated.

[0019] In a second aspect, the present application provides a task offloading device of a task vehicle in cellular vehicle networking, comprising:

[0020] A construction module is configured to define an offloading ratio matrix of the task offloading to the auxiliary computing vehicle and the roadside unit server, to establish a task offloading optimization model aiming at minimizing the total task processing delay, and to establish a task offloading optimization model aiming at minimizing the total task processing delay.

[0021] A solution module is configured to solve the task offloading optimization model to obtain an optimal roadside unit server computing frequency, an optimal auxiliary computing vehicle computing frequency, and a data packet transmission frequency between the task vehicle and the auxiliary computing vehicle.

[0022] A calculation module is configured to calculate an optimal offloading ratio matrix according to the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle.

[0023] A distribution module is configured to distribute the task offloading to the roadside unit server and the auxiliary computing vehicle according to the optimal offloading ratio matrix.

[0024] In a third aspect, the present application provides a non-transitory computer readable storage medium for storing computer instructions, which, when executed by a processor, implement the task offloading method of the task vehicle in cellular vehicle networking as described above.

[0025] In a fourth aspect, the present application provides a computer program product comprising a computer program, which, when running on one or more processors, is configured to implement the task offloading method of the task vehicle in cellular vehicle networking as described above.

[0026] In a fifth aspect, the present application provides an electronic device comprising a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory; when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes instructions for implementing the task offloading method of the task vehicle in cellular vehicle networking as described above.

[0027] The application has the advantages that: considering the distributed transmission and semi-persistent scheduling characteristics of the C-V2X system, a greedy offloading algorithm based on the Uu / PC5 interface is proposed, the success transmission probability of task offloading of the PC5 interface is modeled according to the distributed transmission and semi-persistent scheduling characteristics of the PC5 interface, the offloading strategy of the Uu / PC5 interface and the communication and computing resources are jointly designed, the task vehicle can offload the task to the auxiliary computing vehicle through the PC5 interface for task computing, or offload the task to the VEC server through the Uu interface for computing, thereby reducing the task processing delay; without dense base station coverage, the complexity and cost of task offloading and resource allocation are reduced, and the adaptability is strong.

[0028] The advantages of the additional aspects of the application will be more apparent from the following description section or will be understood through the practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0029] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0030] Figure 1 The cellular vehicle networking transmission model based on the vehicle edge computing system and the joint Uu interface and PC5 interface described in the embodiments of the application.

[0031] Figure 2 The task offloading method flow chart for the cellular vehicle networking edge computing and the joint Uu / PC5 interface described in the embodiments of the application.

[0032] Figure 3 The task processing delay performance graph under the influence of the number of tasks described in the embodiments of the application.

[0033] Figure 4 The task processing delay performance graph under the influence of the vehicle density described in the embodiments of the application.

[0034] Figure 5 The task processing delay performance graph under the influence of the VEC server computing frequency limit described in the embodiments of the application. DETAILED DESCRIPTION

[0035] The embodiments of the application will be described in detail below, and the examples of the embodiments are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below through the drawings are exemplary and are only used to explain the application, and cannot be explained as a limitation on the application.

[0036] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0037] It should also be understood that terms such as those defined in a general dictionary should be understood to have meanings consistent with meanings in the context of the prior art and should not be interpreted to be idealized or overly formal unless defined as such.

[0038] Those skilled in the art can understand that, unless otherwise stated, the singular forms "a", "an" and "the" used herein also include the plural forms. It should be further understood that the use of the phrase "comprising" in the specification of the present application means that the features, integers, steps, operations, elements and / or groups thereof are present, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements and / or groups thereof.

[0039] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "example", "specific example" or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, the skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples without contradiction.

[0040] In order to facilitate the understanding of the present application, the present application will be further explained and described in specific embodiments in connection with the accompanying drawings, and the specific embodiments do not constitute a limitation on the embodiments of the present application.

[0041] Those skilled in the art should understand that the drawings are only schematic of the embodiments, and the components in the drawings are not necessarily necessary for the implementation of the present application.

[0042] Embodiment 1

[0043] An embodiment 1 is provided in the present application, which is a task offloading device for task vehicles in cellular vehicle networking, comprising:

[0044] The construction module is configured to define an offloading ratio matrix of task offloading to auxiliary computing vehicles and roadside unit servers, to establish a task offloading optimization model with the goal of minimizing total task processing delay;

[0045] a solving module configured to solve the task offloading optimization model to obtain an optimal roadside unit server computing frequency, an optimal auxiliary computing vehicle computing frequency, and a data packet transmission frequency between the task vehicle and the auxiliary computing vehicle;

[0046] a calculating module configured to calculate an optimal offloading ratio matrix according to the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle;

[0047] a distributing module configured to distribute tasks offloaded to the roadside unit server and the auxiliary computing vehicle according to the optimal offloading ratio matrix.

[0048] In this embodiment 1, the above device is used to implement a task offloading method for a task vehicle in a cellular vehicle networking. In the cellular vehicle networking, the task vehicle is connected to a roadside unit server through a Uu interface, the roadside unit server is connected to a vehicle edge computing system server, and the task vehicle is connected to a plurality of auxiliary computing vehicles through a PC5 interface. The method includes the following steps:

[0049] An offloading ratio matrix of task offloading to the auxiliary computing vehicles and the roadside unit server is defined by the constructing module, and a task offloading optimization model is established to minimize the total task processing delay.

[0050] The solving module is used to solve the task offloading optimization model to obtain an optimal roadside unit server computing frequency, an optimal auxiliary computing vehicle computing frequency, and a data packet transmission frequency between the task vehicle and the auxiliary computing vehicle.

[0051] The calculating module is used to calculate an optimal offloading ratio matrix according to the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle.

[0052] Finally, the distributing module is used to distribute tasks offloaded to the roadside unit server and the auxiliary computing vehicle according to the optimal offloading ratio matrix.

[0053] For the tasks offloaded to the auxiliary computing vehicles, the evaluation set is defined as the product of the successful transmission probability and the maximum computing frequency of the auxiliary computing vehicle, and the evaluation set is used to determine the order and sequence of task offloading among the auxiliary vehicles. The auxiliary computing vehicles are sorted in descending order in the evaluation set, the task vehicle preferentially offloads tasks to the auxiliary computing vehicles with high order, and when the computing capacity of the current auxiliary computing vehicle exceeds the limit, the task vehicle offloads tasks to the next auxiliary computing vehicle with low order for further computation.

[0054] In the task offloading model, the processing delay of each task includes two parts, one part is the sum of the offloading delay of the task vehicle offloading the task to the roadside unit server and the computing delay of the roadside unit server to the task, and the other part is the sum of the offloading delay of the task vehicle offloading the task to the auxiliary computing vehicle and the computing delay of the auxiliary computing vehicle to the task, and the larger sum is the total processing delay of the task.

[0055] Specifically, the offloading delay of the task vehicle offloading the task to the roadside unit server includes: calculating the task transmission rate between the task vehicle and the roadside unit server; according to the calculated task transmission rate between the task vehicle and the roadside unit server, and combining the task offloading ratio matrix, the offloading delay of the task from the task vehicle to the roadside unit server is calculated.

[0056] The offloading delay of the task vehicle offloading the task to the auxiliary computing vehicle includes:

[0057] According to the data transmission path loss of the task vehicle to the auxiliary computing vehicle, the sensing power threshold of the auxiliary computing vehicle and the data transmission power of the task vehicle to the auxiliary computing vehicle, the data packet sensing probability of the auxiliary computing vehicle is calculated;

[0058] According to the sensing coefficient, the sensing distance of the multi-lane and the number of subframes and subchannels in the selection window of the semi-persistent scheduling algorithm, the task successful transmission probability of the task vehicle to the auxiliary computing vehicle is calculated;

[0059] According to the calculated data packet sensing probability and the task successful transmission probability, the normalized transmission rate between the task vehicle and the auxiliary computing vehicle is calculated;

[0060] According to the normalized transmission rate, the offloading delay of the task vehicle offloading the task to the auxiliary computing vehicle is calculated.

[0061] Embodiment 2

[0062] In this embodiment 2, first, a task offloading device for a cellular vehicle-to-everything transmission model based on a vehicle edge computing system combined with Uu interface and PC5 interface is provided, the device is configured in a task vehicle, and the device is used to realize the allocation of task offloading of the task vehicle.

[0063] In this embodiment 2, in the cellular vehicle-to-everything (V2X) transmission model, a single RSU is connected to a VEC server. The RSU is equipped with a communication module, enabling communication and data transmission with the vehicle. The VEC server provides computing resources to assist the vehicle in task calculation. Vehicles can be divided into two types: the first type is the task vehicle, which generates a series of tasks requiring computation and thus needs more computing power; the second type is the auxiliary computing vehicle, typically a transportation service vehicle such as a bus or taxi, which can be deployed with a microcomputing server to assist the task vehicle in calculation. The auxiliary computing vehicle can be represented as N = {1, 2, ..., N}. Both types of vehicles are equipped with transmission and computing modules. The transmission module includes a Uu interface and a PC5 interface. The Uu interface is used for data communication between the vehicle and the RSU, and the PC5 interface is used for data communication between vehicles. The computing module processes the received tasks. The task vehicle offloads part of the task to the VEC server for computation via the Uu interface, and the remaining part of the task is offloaded to the auxiliary computing vehicle for computation via the PC5 interface. After computation, the VEC server and the auxiliary computing vehicle return the results to the task vehicle.

[0064] In this implementation 2, the successful transmission rate of the PC5 interface is characterized. Under the C-V2X architecture, a joint offloading strategy and resource allocation for the Uu interface and PC5 interface are designed to reduce task processing latency. The task vehicle generates M tasks, denoted as M∈{1,2,...,M}, and the data size of each task is denoted as L. m ,m∈M. Considering that each task is separable, partial offloading can be employed. For each task L m For m∈M, part of the computation is offloaded to RSU (VEC server) for computation, and the other part is offloaded to nearby auxiliary computing vehicles. For ease of description, RSU is defined as the (N+1)th auxiliary computing vehicle.

[0065] In this embodiment 2, the task offloading device includes: a construction module, used to define the offloading ratio matrix of tasks offloading to auxiliary computing vehicles and roadside unit servers, and to establish a task offloading optimization model with the goal of minimizing the total task processing latency; a solution module, used to solve the task offloading optimization model to obtain the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle; a calculation module, used to calculate the optimal offloading ratio matrix based on the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle; and an allocation module, used to allocate tasks offloaded to the roadside unit server and the auxiliary computing vehicle based on the optimal offloading ratio matrix.

[0066] In this embodiment 2, the offloading ratio matrix is defined as α, where the element of the matrix is α m,n ∈ [0, 1]. For offloading tasks to VEC servers for computation, α m,N+1 > 0 indicates that the task vehicle offloads a portion of the mth task to the VEC server; α m,N+1 L m to the VEC server; α m,N+1 = 0 indicates that the task vehicle does not offload any portion of the mth task to the VEC server. For offloading tasks to auxiliary computing vehicles for computation, α m,n > 0 indicates that the task vehicle offloads a portion of the mth task to the nth auxiliary computing vehicle; α m,n L m to the nth auxiliary computing vehicle; α m,n = 0 indicates that the task vehicle does not offload any portion of the mth task to the nth auxiliary computing vehicle.

[0067] When the above device is used for task offloading, a communication model and a computation model are established, and the task processing delay of the Uu interface and the PC5 interface is modeled. The communication mode is divided into two types of Uu interface and PC5 interface. The Uu interface is a communication interface between the vehicle and the RSU. The PC5 interface is a communication interface between vehicles.

[0068] For the Uu interface: the Uu interface accesses the channel through the Orthogonal Frequency Division Multiple Access (OFDMA) technology, and the transmission rate of the mth task between the task vehicle and the RSU is:

[0069]

[0070] where g m,N+1 represents the channel gain between the task vehicle and the RSU, B m,N+1 represents the bandwidth, σ 2 represents the noise power, represents the transmission power of the Uu interface of the task vehicle.

[0071] The offloading delay of the mth task from the task vehicle to the VEC server is:

[0072]

[0073] For the PC5 interface: the data packet transmission frequency of the PC5 interface of the task vehicle is represented as λ, which represents the number of data packets transmitted by the task vehicle per second. The transmission frequency is adaptive and the actual transmission frequency of the task vehicle is less than the ideal Shannon channel capacity, i.e. The task vehicle occupies resources in a semi-persistent scheduling manner, that is, it maintains a period of time after occupying resources, which describes the influence of C-V2X characteristics on V2V transmission rate. Since the wireless transmission mode of C-V2X is half duplex, the half duplex error refers to the fact that the vehicle cannot receive data packets when transmitting its own data packets in the same subframe. The HD error can be given as where N frame is the number of subframes per second. When the auxiliary computing vehicle receives a signal power greater than the sensing power threshold, the auxiliary computing vehicle can decode the data for processing, and the data packet sensing probability of the auxiliary computing vehicle can be represented as:

[0074]

[0075] where PL(d) is the path loss, is the sensing power threshold, P PC5 is the transmission power of the task vehicle at the PC5 interface. The successful sensing probability of the task vehicle to the auxiliary computing vehicle can be represented as:

[0076]

[0077] where K is the conversion coefficient considering multiple lanes, d sen is the sensing distance of the TV, N sub-f and N sub-c represent the number of subframes and subchannels in the selected window in the semi-persistent scheduling algorithm. Integrating the above model, the task successful transmission probability of the task vehicle is:

[0078] P STP =(1-δ HD )·P SEN ·PSR(d) (5)

[0079] The task successful transmission probability will further affect the transmission rate of the task, so the normalized transmission rate of the mth task between the task vehicle and the nth auxiliary computing vehicle is:

[0080]

[0081] The delay of the offloaded part of the mth task from the task vehicle to the nth auxiliary computing vehicle is

[0082]

[0083] For the computing model, the computing amount of the mth task of the task vehicle is represented by C m , the unit is cycle / bit, which represents the number of CPU cycles for computing 1 bit of task data. The computing frequencies of the VEC server and the nth auxiliary computing vehicle allocated to the mth task part are defined as and The unit is (CPU cycles / second). Therefore, the computation latency of the m-th task on the VEC server is:

[0084]

[0085] The computation delay of the m-th task in the n-th auxiliary computation vehicle is:

[0086]

[0087] Based on the above communication and computation models, an optimization problem is established that aims to minimize task processing latency while simultaneously satisfying the communication, computational resource constraints, and offloading factor constraints of both the Uu and PC5 interfaces, namely the task offloading optimization model.

[0088] Since the task vehicles are partially unloaded, the processing latency of each task consists of two parts: one part is the sum of the unloading latency and computation latency of the task vehicle to the VEC server, and the other part is the sum of the unloading latency and computation latency of the task vehicle to the auxiliary computation vehicle. The total processing latency of the m-th task is the larger of the two, expressed as:

[0089]

[0090] Jointly optimize the data packet transmission frequency λ of the PC5 interface and the VEC server frequency f. ser The CPU computing frequency f of the auxiliary calculation vehicle veh Given the unloading ratio matrix α, minimize the total processing latency of all tasks. Since the tasks on the task vehicle are processed sequentially, the total latency is the sum of the processing latencies of all tasks unloaded by the task vehicle. The task processing latency minimization problem is formulated as follows:

[0091]

[0092] subject to L m λ≤C v2v,max (11a)

[0093] λ∈(0,λ max (11b)

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100] where λ max = N frmae Constraints (11a) and (11b) represent the packet transmission frequency constraints. Constraints (11c) and (11d) are the computation frequency constraints for the VEC servers and the auxiliary computing vehicles. Constraint (11e) represents that the offloading factor ratio in the offloading matrix belongs to the range of 0 to 1, that is, partial offloading. Constraint (11f) represents that the sum of the ratios of the mth task assigned to the auxiliary computing vehicles or the VEC servers is equal to 1. Constraint (11g) guarantees that a part of the mth task is only offloaded to one auxiliary computing vehicle.

[0101] The optimization problem established above is decomposed into a resource allocation sub-problem and an offloading allocation sub-problem, and is solved, specifically including the following:

[0102] When the task processing delays transmitted to the VEC servers and the auxiliary computing vehicles are the same, that is, By reductio ad absurdum, the optimal solution of the problem can be obtained at this time, and therefore, the optimization objective can be converted to Therefore, the problem can be converted to

[0103]

[0104]

[0105] Due to the existence of variable coupling, the problem (12) is non-convex, and therefore, the problem is decomposed into a resource allocation sub-problem and an offloading allocation sub-problem, so as to realize the decoupling of the resource allocation variables.

[0106] The resource allocation sub-problem is:

[0107]

[0108] subject to (10a), (10b), (10c), (10d), (10h)

[0109] According to the convex optimization theory, the problem (13) is a convex function, and can be solved by using the KKT condition. The partial Lagrange function of the problem (13) can be expressed as

[0110]

[0111] where the dual variables are η, ζ and ψ. According to the KKT condition, the optimal computation frequencies of the VEC servers and the nth auxiliary computing vehicle solving the problem (13) are respectively

[0112]

[0113]

[0114] By equations (15) and (16), when the dual variables ζ and ψ n increase, i.e., VEC and the auxiliary computing vehicle's computing resource requirements are strict, the VEC server and vehicle computing frequency will decrease.

[0115] For the packet sending frequency λ, since it is not coupled with other optimization variables, it can be solved independently of the resource allocation subproblem.

[0116]

[0117] subject to(10a),(10b)

[0118] From the constraint of the packet sending frequency, λ∈(0,λ cons ], where For convenience of representation, define the objective of problem (17) as y(λ), whose second derivative satisfies In addition, since both constraints (10a) and (10b) are linear constraints, according to the second-order condition, problem (17) is a convex problem. By calculating , we can get Considering the constraint λ∈(0,λ cons ], when , the optimal packet sending frequency When , λ is not available, and since , the objective function of problem (17) is monotonically decreasing in the interval (0,λ cons ], and the optimal packet sending frequency is λ * =λ cons . In summary, the optimal packet sending frequency is

[0119]

[0120] For the dual variable update, according to the dual variables η, ζ and ψ updated by iteration and the optimization variables λ, f ser and f veh , the updated dual variables η (k) , ζ (k) and ψ (k) are obtained.

[0121] For the offloading allocation subproblem, by calculating The value of the elements of the offloading matrix can be obtained, that is, the optimal offloading ratio matrix of the task vehicle can be obtained when the transmission to the VEC server and the auxiliary computing vehicle task processing delay is the same, that is, the proportion of the task part offloaded to the auxiliary computing vehicle and the VEC server, as follows:

[0122]

[0123] wherein,

[0124] Finally, according to the proportion of the task part offloaded to the auxiliary computing vehicle and the VEC server, the offloading allocation of the task is carried out. Part of the task is offloaded to the auxiliary computing vehicle through the PC5 interface for calculation, and the other part is offloaded to the VEC server for calculation. For the task part offloaded to the auxiliary computing vehicle, the influence of the SL link of the PC5 interface is considered, and the evaluation set is defined as the product of the successful transmission probability and the maximum computing frequency of the auxiliary computing vehicle, that is, Further, the evaluation set is used to determine the order and sequence of offloading, wherein the auxiliary computing vehicles are sorted in descending order in the evaluation set. Based on the greedy criterion, the task vehicle preferentially offloads the task to the auxiliary computing vehicle with high order, that is, the auxiliary computing vehicle with high successful transmission probability of the transmission link and strong computing capability. When the computing capability of the current auxiliary computing vehicle exceeds the limit, the task vehicle offloads the task to the auxiliary computing vehicle with the next order for further calculation. Therefore, the task processing delay can be further reduced by using the greedy offloading algorithm based on the Uu / PC5 interface.

[0125] Embodiment 3

[0126] In this embodiment 3, in order to solve the problem that the traditional offloading strategy does not consider the distributed transmission characteristics and semi-persistent scheduling of the PC5 interface in the C-V2X edge computing system, thereby lacking an effective offloading strategy for the C-V2X system, a greedy offloading algorithm based on the Uu / PC5 interface is proposed, and a joint offloading strategy suitable for the Uu / PC5 interface of the C-V2X edge computing system is designed. The communication and computing resources are optimized to ensure the task delay requirement of the C-V2X edge computing system, and an edge computing task allocation method for vehicle assistance of vehicle-road cooperation is realized,

[0127] In this embodiment 3, the edge computing task allocation method for vehicle assistance of vehicle-road cooperation specifically includes the following steps:

[0128] (I) Establish the architecture of the C-V2X edge computing system.

[0129] As Figure 1A vehicle-to-roadside cooperative edge computing system is shown, in which multiple vehicles cooperate with a single RSU, and the single RSU is connected with a VEC server. The RSU is connected with the VEC server, and the RSU is equipped with a communication module and can communicate and transmit data with vehicles, and the VEC server can provide computing resources to assist vehicle task computing. Vehicles can be divided into two types: the first type is a task vehicle, which generates a series of tasks that need to be calculated and requires more computing power; the second type is an auxiliary computing vehicle, which is generally a vehicle providing transportation services, such as a bus and a taxi, and can deploy a micro-computing server to assist the task vehicle in computing, and the auxiliary computing vehicle can be represented as N = {1, 2,..., N}. Both types of vehicles are equipped with a transmission module and a computing module, the transmission module includes a Uu interface and a PC5 interface, the Uu interface is used for data communication between the vehicle and the RSU, and the PC5 interface is used for data communication between the vehicle and the vehicle, and the computing module can be used to process the received tasks. The task vehicle offloads part of the task to the VEC server for calculation through the Uu interface, and offloads another part of the task to the auxiliary computing vehicle for calculation through the PC5 interface. After the calculation is completed, the VEC server and the auxiliary computing vehicle return the calculation result to the task vehicle. The present application characterizes the successful transmission rate of the PC5 interface, designs an offloading strategy and resource allocation of the joint Uu interface and PC5 interface under the C-V2X architecture, and reduces the task processing delay.

[0130] The task vehicle generates M tasks, represented as M ∈ {1, 2,..., M}, and the data size of the task is represented as L m ,m ∈ M. Considering that each task is separable, partial offloading can be used. For each task L m ,m ∈ M, part of it is offloaded to the RSU (VEC server) for calculation, and another part is offloaded to the nearby auxiliary computing vehicle. For the convenience of description, the RSU is defined as the (N+1)th auxiliary computing vehicle. Define the offloading ratio matrix as α, where the elements of the matrix are α m,n ∈ [0, 1]. For offloading tasks to the VEC server for calculation, α m,N+1 > 0 indicates that the task vehicle offloads a part of the mth task α m,N+1 L m to the VEC server; α m,N+1 = 0 indicates that the task vehicle will not offload any part of the mth task to the VEC server. For offloading tasks to the auxiliary computing vehicle for calculation, α m,n > 0 indicates that the task vehicle offloads a part of the mth task m α m,n L m to the nth auxiliary computing vehicle; α m,n = 0 indicates that the task vehicle will not assign any part of the mth task to the nth auxiliary computing vehicle.

[0131] The technical effect achieved by the step is to complete the function construction of each part of the entity module (task vehicle, auxiliary computing vehicle, RSU, VEC computing unit) in the C-V2X edge computing system, and to derive the task processing flow of the task vehicle.

[0132] (ii) A communication model and a calculation model are established to model the task processing delay of the Uu interface and the PC5 interface. The communication mode is divided into two types of Uu interface and PC5 interface. The Uu interface is the communication interface between the vehicle and the RSU. The PC5 interface is the communication interface between the vehicle and the vehicle.

[0133] For the Uu interface: the Uu interface accesses the channel through the Orthogonal Frequency Division Multiple Access (OFDMA) technology, and the transmission rate of the mth task between the task vehicle and the RSU is

[0134]

[0135] where g m,N+1 represents the channel gain between the task vehicle and the RSU, B m,N+1 represents the bandwidth, σ 2 represents the noise power, represents the transmission power of the Uu interface of the task vehicle.

[0136] The unloading delay of the mth task from the task unloading to the VEC server is

[0137]

[0138] For the PC5 interface: the data packet transmission frequency of the PC5 interface of the task vehicle is represented as λ, which represents the number of data packets transmitted by the task vehicle per second. The transmission frequency can be adaptively adjusted, and the actual transmission frequency of the task vehicle is less than the ideal Shannon channel capacity, that is, The resource occupation mode of the task vehicle is semi-persistent scheduling, that is, the resource is maintained for a period of time after being occupied, which describes the influence of C-V2X characteristics on the V2V transmission rate. Since the wireless transmission mode of C-V2X is half duplex, half duplex error means that the vehicle cannot receive data packets when transmitting its own data packets in the same subframe. HD can be given as where N frame is the number of subframes per second. When the signal power received by the auxiliary computing vehicle is greater than the sensing power threshold, the auxiliary computing vehicle can decode the data for processing, and the data packet sensing probability of the auxiliary computing vehicle can be represented as

[0139]

[0140] where PL(d) is the path loss, is the sensing power threshold, P PC5 is the transmit power of the task vehicle at the PC5 interface. The successful sensing probability of the task vehicle to the assistant computing vehicle can be expressed as

[0141]

[0142] where K is the conversion coefficient considering multi-lane, d sen is the sensing distance of the TV, N sub-f and N sub-c represent the number of subframes and subchannels in the selection window of the semi-persistent scheduling algorithm, respectively. Integrating the above model, the task successful transmission probability of the task vehicle is

[0143] P STP = (1 - δ HD ) · P SEN · PSR(d) (5)

[0144] The task successful transmission probability will further affect the transmission rate of the task, so the normalized transmission rate of the mth task between the task vehicle and the nth assistant computing vehicle is

[0145]

[0146] The offloading part of the mth task is offloaded from the task vehicle to the nth assistant computing vehicle with a delay of

[0147]

[0148] For the computing model, the computing amount of the mth task of the task vehicle is represented by C m , with the unit of cycle / bit, which represents the number of CPU cycles for computing 1 bit of task data. The computing frequencies of the VEC server and the nth assistant computing vehicle allocated to the mth task part are defined as and with the unit of (CPU cycle / sec). Then the computing delay of the mth task at the VEC server is

[0149]

[0150] The computing delay of the mth task at the nth assistant computing vehicle is

[0151]

[0152] The technical effect achieved by the second step is to construct the communication model and the calculation model of the C-V2X edge computing system, specifically to establish the transmission rate of the Uu interface and the success transmission probability and transmission rate of the PC5 interface, and then to deduce the task processing delay of the task through the Uu and PC5 interfaces for unloading.

[0153] (Three) Establish an optimization problem with the minimum task processing delay as the target while meeting the communication, calculation resource constraints and unloading factor constraints of the Uu interface and the PC5 interface.

[0154] Since the task vehicle adopts partial unloading, the processing delay of each task includes two parts, one part is the sum of the unloading delay and the calculation delay of the task vehicle unloading to the VEC server, and the other part is the sum of the unloading delay and the calculation delay of the task vehicle unloading to the auxiliary computing vehicle. The total processing delay of the mth task is the larger delay between the two, which is expressed as

[0155]

[0156] Jointly optimize the data packet transmission frequency λ of the PC5 interface, the VEC server frequency f ser , the CPU calculation frequency f veh of the auxiliary computing vehicle and the unloading ratio matrix α to minimize the total processing delay of all tasks. Since the tasks of the task vehicle are processed in series, the total delay is the sum of the processing delays of all tasks unloaded by the task vehicle. The task processing delay minimization problem is described as follows:

[0157]

[0158] subject to L m λ≤C v2v,max (11a)

[0159] λ∈(0,λ max ] (11b)

[0160]

[0161]

[0162]

[0163]

[0164]

[0165]

[0166] where λ max =N frmaeConstraints (11a) and (11b) represent the packet transmission frequency constraints. Constraints (11c) and (11d) are the VEC server and auxiliary computing vehicle's computing frequency constraints. Constraint (11e) represents that the offloading factor ratio in the offloading matrix belongs to the range of 0 to 1, that is, partial offloading. Constraint (11f) represents that the sum of the ratios of the mth task assigned to the auxiliary computing vehicle or the VEC server is equal to 1. Constraint (11g) guarantees that part of the mth task is only offloaded to one auxiliary computing vehicle.

[0167] The technical effect achieved by the third step is that the overall task processing delay and the resource constraints in the system are modeled, and the task delay minimization problem is derived.

[0168] (4) Decompose the optimization problem established in step 3 into a resource allocation sub-problem and an offloading allocation sub-problem, and solve them.

[0169] When the task processing delay of the transmission to the VEC server and the auxiliary computing vehicle is the same, that is, By reductio ad absurdum, the optimal solution of the problem can be obtained at this time, and therefore, the optimization objective can be converted to Therefore, the problem can be converted to

[0170]

[0171]

[0172] Due to the existence of variable coupling, the problem (12) is non-convex, and therefore the problem is decomposed into a resource allocation sub-problem and an offloading allocation sub-problem, so as to realize the decoupling of the resource allocation variable.

[0173] The resource allocation sub-problem is:

[0174]

[0175] subject to (10a), (10b), (10c), (10d), (10h)

[0176] According to the convex optimization theory, the problem (13) is a convex function, and can be solved using the KKT condition. The partial Lagrange function of the problem (13) can be expressed as

[0177]

[0178] where the dual variables are η, ζ and ψ. According to the KKT condition, the optimal computing frequencies of the VEC server and the nth auxiliary computing vehicle for solving the problem (13) are respectively

[0179]

[0180]

[0181] By equations (15) and (16), when the dual variables ζ and ψ n increase, i.e., VEC and the auxiliary computing vehicle's computing resource requirements are strict, the VEC server and vehicle computing frequency will decrease.

[0182] For the packet sending frequency λ, since it is not coupled with other optimization variables, it can be solved independently of the resource allocation subproblem.

[0183]

[0184] subject to(10a),(10b)

[0185] From the constraint of the packet sending frequency, λ∈(0,λ cons ], where For convenience of representation, define the objective of problem (17) as y(λ), whose second derivative satisfies In addition, since both constraints (10a) and (10b) are linear constraints, according to the second-order condition, problem (17) is a convex problem. By calculating , we can get Considering the constraint λ∈(0,λ cons ], when , the optimal packet sending frequency When , is not available, and since , the objective function of problem (17) is monotonically decreasing in the interval (0,λ cons ], and the optimal packet sending frequency is λ * =λ cons . In summary, the optimal packet sending frequency is

[0186]

[0187] For the dual variable update, according to the dual variables η, ζ and ψ updated by iteration and the optimization variables λ, f ser and f veh , the updated dual variables η (k) , ζ (k) and ψ (k) are obtained.

[0188] For the offloading allocation subproblem, by calculating The value of the elements of the offloading matrix can be obtained, that is, the optimal offloading ratio matrix of the task vehicle can be obtained when the transmission to the VEC server and the task processing delay of the auxiliary computing vehicle are the same, that is, the ratio of the task part offloaded to the auxiliary computing vehicle to the VEC server, as follows

[0189]

[0190] wherein

[0191] The technical effect that can be achieved by this step (four) is to decompose the problem in step three into two sub-problems, in the resource allocation sub-problem, the packet transmission frequency of the PC5 interface and the calculation frequency of the auxiliary computing vehicle and the VEC server are derived; in the offloading allocation sub-problem, the ratio of the offloading matrix is derived.

[0192] (five) Establish a greedy offloading algorithm based on Uu / PC5 interface.

[0193] For the offloading algorithm, each task of the task vehicle adopts partial offloading, part of the task is offloaded to the auxiliary computing vehicle for calculation through the PC5 interface, and the other part is offloaded to the VEC server for calculation. For the task part offloaded to the auxiliary computing vehicle, the influence of the SL link of the PC5 interface is considered, and the evaluation set is defined as the product of the successful transmission probability and the maximum calculation frequency of the auxiliary computing vehicle, that is, Further, the evaluation set is used to determine the order and sequence of offloading, wherein the auxiliary computing vehicles are sorted in descending order in the evaluation set. Based on the greedy criterion, the task vehicle preferentially offloads the task to the auxiliary computing vehicle with high order, that is, the auxiliary computing vehicle with high successful transmission probability of the transmission link and strong computing ability, and when the computing ability of the current auxiliary computing vehicle exceeds the limit, the task vehicle offloads the task to the auxiliary computing vehicle with the next order for further calculation, so that the task processing delay can be further reduced through the greedy offloading algorithm based on Uu / PC5 interface.

[0194] The technical effect that can be achieved by this step (five) is to derive the greedy offloading algorithm based on Uu / PC5 interface based on the results of resource allocation and offloading allocation derived in step four.

[0195] (six) The task vehicle offloads the task to the VEC server or the auxiliary computing vehicle for calculation through the Uu interface and the PC5 interface according to the greedy offloading algorithm based on Uu / PC5 interface, and sends the result back to the task vehicle after the calculation is completed. The technical effect that can be achieved by this step is to execute the greedy offloading algorithm based on Uu / PC5 interface, offload the task to the VEC server or the auxiliary computing vehicle for calculation through the Uu interface and the PC5 interface, complete the task calculation, and then return the result to the task vehicle.

[0196] Without loss of generality, to verify that the method can reduce the task processing delay of task processing, taking the parameter configuration in Table 1 as an example, the effect of the task offloading method described in the embodiment is illustrated.

[0197] Figure 1 The figure is a C-V2X edge computing system architecture established in the embodiment. A single RSU is deployed and connected with a VEC server. A task vehicle has a large amount of computing tasks to be processed. The task vehicle offloads part of the tasks to an auxiliary computing vehicle for computation according to a greedy offloading algorithm based on a Uu / PC5 interface, and offloads another part of the tasks to the VEC server for computation. After the computation is completed, the results are sent back to the task vehicle.

[0198] Figure 2 The figure is a method flowchart of the joint resource allocation and offloading algorithm selection method in the embodiment. The method first solves a resource allocation subproblem, calculates the optimal task vehicle package sending frequency, the CPU computation frequency of the auxiliary computing vehicle and the VEC server, and the proportion of the offloading ratio matrix, then solves a and offloading allocation subproblem, and obtains the factor of the offloading matrix, that is, the offloading strategy, by using a greedy search algorithm, and further minimizes the task processing delay.

[0199] Table 1 simulation parameter table

[0200]

[0201]

[0202] Under the simulation parameter settings as shown in Table 1, three comparison schemes of a VEC server computation scheme, an auxiliary computing vehicle computation scheme, and an orthogonal allocation offloading computation scheme are set to compare the performance of the method proposed in the present embodiment. Simulation is performed to obtain the task processing delay performance graph of the present embodiment under the influence of the number of tasks, the task processing delay performance graph under the influence of vehicle density, and the task processing delay performance graph under the influence of the VEC server computation frequency limit, as shown in Figure 3 , Figure 4 and Figure 5 .

[0203] Figure 3 The figure shows the task processing delay performance graph under the influence of the number of tasks. As can be observed from the figure, the present embodiment can achieve lower task processing delay compared with the VEC server computation scheme, the auxiliary computing vehicle computation scheme, and the orthogonal allocation offloading computation scheme. When the number of tasks increases, the task processing delay increases with the increase of the number of tasks due to the limitation of system resources.

[0204] Figure 4The figure shows the latency performance of task processing under the influence of vehicle density. It can be observed that the present scheme and the orthogonal allocation offloading computing scheme have lower task processing latency than the VEC server computing scheme and the auxiliary computing vehicle computing scheme, which embodies the advantage of partial offloading. In addition, as the vehicle density increases, the task processing latency of the present scheme, the orthogonal allocation offloading computing scheme and the auxiliary computing vehicle computing scheme also increases accordingly. When the vehicle density is greater than 0.11, the task processing latency of the present scheme is greater than that of the orthogonal allocation offloading computing scheme, because the successful transmission probability decreases with the increase of vehicle density, so the orthogonal allocation offloading computing scheme is more suitable for high vehicle density.

[0205] Figure 5 The figure shows the latency performance of task processing under the influence of VEC server computing frequency limit. The present scheme can achieve lower task processing latency than the VEC server computing scheme, the auxiliary computing vehicle computing scheme, and the orthogonal allocation offloading computing scheme. In addition, the system latency of the auxiliary computing vehicle computing scheme does not change with the change of VEC server computing frequency. As the VEC server computing frequency increases, the task processing latency of the present scheme, the orthogonal allocation offloading computing scheme and the VEC server computing scheme decreases accordingly, and as the computing frequency increases, the task processing latency of the three schemes tends to be stable due to the limitation of wireless resources.

[0206] Embodiment 4

[0207] Embodiment 4 of the present application provides a non-transitory computer readable storage medium for storing computer instructions, which, when executed by a processor, implements a task offloading method for a task vehicle in cellular vehicle networking.

[0208] Embodiment 5

[0209] Embodiment 5 of the present application provides a computer program (product) comprising a computer program for implementing a task offloading method for a task vehicle in cellular vehicle networking when running on one or more processors.

[0210] Embodiment 6

[0211] Embodiment 6 of the present application provides an electronic device comprising a processor, a memory and a computer program; wherein the processor is connected with the memory, and the computer program is stored in the memory; when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes instructions for implementing a task offloading method for a task vehicle in cellular vehicle networking.

[0212] In summary, the task offloading method of the joint Uu / PC5 interface for cellular vehicle-to-everything edge computing is described in the embodiment of the application. The differentiated features of task vehicles performing task offloading through Uu and PC5 interfaces in the cellular vehicle-to-everything (C-V2X) edge computing system are studied. The success transmission probability of task offloading through PC5 interface is modeled according to the distributed transmission and resource semi-persistent scheduling characteristics of the C-V2X system. A task offloading method combining Uu and PC5 interfaces is designed. The application optimizes the offloading ratio matrix and packet transmission frequency of the PC5 interface, optimizes the allocation of computing resources of the auxiliary computing vehicle and the VEC server, and designs the task offloading method with the goal of minimizing the task processing delay. The task processing delay minimization problem is decomposed into resource allocation and offloading strategy sub-problems, and a greedy offloading algorithm based on Uu / PC5 interface is proposed. For the resource allocation sub-problem, the optimal packet transmission frequency of the PC5 interface and the CPU computing frequency of the vehicle and the VEC server are designed. For the offloading strategy sub-problem, the offloading ratio matrix and the task offloading scheme are obtained through the proposed PC5-GO algorithm. Based on the task offloading method, the task vehicle can offload tasks to the auxiliary computing vehicle through the PC5 interface for task computation, or offload tasks to the VEC server through the Uu interface for computation. The application effectively completes the task offloading method and reduces the task processing delay of the cellular vehicle-to-everything edge computing system.

[0213] Although the specific embodiments of the application are described above with reference to the accompanying drawings, the description is not a limitation on the scope of protection of the application. Those skilled in the art should understand that various modifications or variations can be made to the disclosed technical solutions without creative labor, and all such modifications or variations should be covered within the scope of protection of the application.

Claims

1. A method for task offloading of a task vehicle in a cellular vehicle-to-everything (V2X) network, characterized in that, The method comprises the following steps: Defining an offloading ratio matrix of task offloading to an auxiliary computing vehicle and a roadside unit server, establishing a task offloading optimization model aiming at minimizing total task processing latency; Solving the task offloading optimization model to obtain an optimal roadside unit server computing frequency, an optimal auxiliary computing vehicle computing frequency and a data packet transmission frequency between a task vehicle and the auxiliary computing vehicle; Calculating an optimal offloading ratio matrix according to the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency and the data packet transmission frequency between the task vehicle and the auxiliary computing vehicle; Allocating tasks offloaded to the roadside unit server and the auxiliary computing vehicle according to the optimal offloading ratio matrix. 2.The task offloading method for task vehicles in cellular V2X according to claim 1, wherein, For the tasks offloaded to the auxiliary computing vehicle, defining an evaluation set as the product of a successful transmission probability and a maximum computing frequency of the auxiliary computing vehicle, and then using the evaluation set to determine the order and sequence of task offloading among the auxiliary vehicles. 3.The task offloading method for task vehicles in cellular V2X of claim 2, wherein, The auxiliary computing vehicles are sorted in descending order in the evaluation set, and the task vehicle preferentially offloads tasks to the auxiliary computing vehicles with high order, and when the computing capacity of the current auxiliary computing vehicle exceeds the limit, the task vehicle offloads tasks to the auxiliary computing vehicle with the next order for continuous computing. 4.The method of claim 1, wherein, The processing latency of each task comprises two parts, one part is the sum of the offloading latency of the task vehicle offloading tasks to the roadside unit server and the computing latency of the roadside unit server to the tasks, and the other part is the sum of the offloading latency of the task vehicle offloading tasks to the auxiliary computing vehicle and the computing latency of the auxiliary computing vehicle to the tasks, and the larger sum is the total task processing latency. 5.The method of claim 4, wherein, The offloading latency of the task vehicle offloading tasks to the roadside unit server comprises: calculating a task transmission rate between the task vehicle and the roadside unit server; and calculating the offloading latency of the task offloaded from the task vehicle to the roadside unit server according to the calculated task transmission rate between the task vehicle and the roadside unit and in combination with the task offloading ratio matrix. 6.The method of claim 5, wherein, The offloading latency of the task vehicle offloading tasks to the auxiliary computing vehicle comprises: Calculating a data packet sensing probability of the auxiliary computing vehicle according to a data transmission path loss from the task vehicle to the auxiliary computing vehicle, a sensing power threshold of the auxiliary computing vehicle and a data transmission power from the task vehicle to the auxiliary computing vehicle; Calculating a task successful transmission probability from the task vehicle to the auxiliary computing vehicle according to a sensing coefficient of a multi-lane, a sensing distance and the number of subframes and subchannels in a selection window in a semi-persistent scheduling algorithm; Calculating a normalized transmission rate between the task vehicle and the auxiliary computing vehicle according to the calculated data packet sensing probability and the task successful transmission probability; Calculating the offloading latency of the task vehicle offloading tasks to the auxiliary computing vehicle according to the normalized transmission rate. 7.A task offloading device for a task vehicle in a cellular vehicle networking, characterized in that, The method comprises the following steps: A construction module is configured to define an offloading ratio matrix of task offloading to an auxiliary computing vehicle and a roadside unit server, and establish a task offloading optimization model aiming at minimizing total task processing latency; A solving module is configured to solve the task offloading optimization model to obtain an optimal roadside unit server computing frequency, an optimal auxiliary computing vehicle computing frequency and a data packet transmission frequency between a task vehicle and the auxiliary computing vehicle; and The computing module is configured to calculate an optimal roadside unit server computing frequency, an optimal auxiliary computing vehicle computing frequency, a data packet transmission frequency between the task vehicle and the auxiliary computing vehicle, and an optimal offloading ratio matrix according to the optimal roadside unit server computing frequency, the optimal auxiliary computing vehicle computing frequency, and the data packet transmission frequency. The distribution module is configured to distribute tasks to the roadside unit server and the auxiliary computing vehicle according to the optimal offloading ratio matrix.

8. A non-transitory computer-readable storage medium, comprising: The non-transitory computer readable storage medium is configured to store computer instructions, and the computer instructions are executed by a processor to implement the task offloading method for a task vehicle in cellular vehicle networking according to any one of claims 1-6.

9. A computer program product, characterised in that, The computer program is configured to implement the task offloading method for a task vehicle in cellular vehicle networking according to any one of claims 1-6 when the computer program is run on one or more processors.

10. An electronic device, comprising: The computer program product comprises: A processor, a memory, and a computer program; wherein the processor is connected with the memory, the computer program is stored in the memory, and the processor executes the computer program stored in the memory to enable the electronic device to execute instructions for implementing the task offloading method for a task vehicle in cellular vehicle networking according to any one of claims 1-6 when the electronic device is running.

Images