Resource exchange method and device in internet of vehicles edge computing

Abstract

The application relates to a resource exchange method and device in Internet of Vehicles edge computing. The method effectively solves the queue congestion problem caused by the limited computing capacity of the edge computing node by constructing a dynamic exchange mechanism of resource segments between jobs. The jobs are paired based on the queuing time difference and time interval constraint to ensure the feasibility of the exchange. The design of the utility function associates the QoE value with the soft deadline, so that the revenue and cost evaluation closely match the real-time requirements of the Internet of Vehicles jobs. Further, the double function structure of the target function balances the interests of both parties, and the resource exchange is converted into a mutually beneficial behavior through the compensation mechanism, avoiding system instability caused by unilateral resource preemption. Therefore, the resource utilization rate of the edge computing node is improved, and the overall response capability of the vehicle job is improved.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of Internet of Vehicles edge computing, and in particular to a resource exchange method and device in Internet of Vehicles edge computing. BACKGROUND

[0002] Internet of Vehicles applications have strict requirements for real-time response and high-bandwidth transmission. The traditional cloud computing architecture has significant round-trip communication delay due to the long distance between the data center and the vehicle, and it is difficult to support the low-latency requirements of key scenarios such as vehicle safety warning and automatic driving.

[0003] Vehicle edge computing technology deploys computing resources at edge nodes such as roadside units or base stations, effectively shortening the data transmission path by sinking data processing capabilities to the network edge. However, the physical resource capacity of edge computing nodes is limited, and when multiple vehicles submit computing tasks at the same time, the task queue is prone to backlog. Since each task occupies multiple discrete resource segments and has a soft deadline constraint, improper resource allocation can cause some tasks to wait too long and fail to complete the calculation within the soft deadline, resulting in a significant decline in user experience quality. Existing resource scheduling mechanisms usually use static allocation strategies or simple priority sorting, which cannot dynamically coordinate the resource usage of tasks with different queuing times, causing resource fragmentation to worsen and the overall efficiency of the system to be difficult to improve. SUMMARY

[0004] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0005] The main purpose of the embodiments of the present disclosure is to propose a resource exchange method in Internet of Vehicles edge computing, which can maximize the resource replacement problem of QoE of jobs under the consideration of the interests of the jobs themselves, to alleviate the queue congestion problem caused by the limited computing capacity of edge computing nodes.

[0006] In a first aspect, the present application provides a resource exchange method in Internet of Vehicles edge computing, the method comprising:

[0007] obtaining a plurality of jobs queuing and waiting for calculation in an edge computing node; the job is a task generated by a vehicle, and each job occupies a plurality of resource segments of the edge computing node;

[0008] determining a plurality of candidate job pairs from the plurality of jobs; wherein any one of the job pairs includes a first job and a second job, the queuing time of the first job is lower than the queuing time of the second job, and the queuing time interval between the first job and the second job is greater than a preset minimum time slice length;

[0009] determining a first utility function of a first job in each of the job pairs, and determining a second utility function of a second job; the first utility function comprises a difference between a first benefit item and a first cost item, the first benefit item is a QoE value obtained by the first job in a case of exchanging a latest available resource segment of the first job with an earliest available resource segment of the second job, the first cost item is a compensation paid by the first job to the second job in the case of exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job; the second utility function comprises a difference between a second benefit item and a second cost item, the second benefit item is a compensation paid by the first job in the case of exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job, the second cost item is a QoE value reduced by the second job in the case of exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job; wherein the QoE value is associated with a soft deadline of the first job or the second job;

[0010] determining an objective function and a corresponding decision item; the objective function comprises a first sub-function and a second sub-function, the first sub-function is to maximize a sum of the first utility functions corresponding to the plurality of job pairs, the second sub-function is to maximize a sum of the second utility functions corresponding to the plurality of job pairs; the decision item comprises all candidate plurality of job pairs, a compensation unit price between any one of the job pairs, and a resource exchange amount between any one of the job pairs; wherein the compensation is associated with a product of the compensation unit price and the resource exchange amount;

[0011] solving the objective function to obtain an optimal solution of the decision item;

[0012] completing exchange of the corresponding resource segments according to the optimal solution of the decision item.

[0013] The resource exchange method in the vehicle networking edge computing provided by the embodiment has at least the following beneficial effects:

[0014] The method effectively solves the queue congestion problem caused by limited computing capacity of the edge computing node by constructing a dynamic exchange mechanism of resource segments between jobs; wherein, the job pairs are screened based on the queuing time difference and the time interval constraint to ensure the feasibility of the exchange, and the design of the utility function associates the QoE value with the soft deadline, so that the benefit and cost evaluation closely fit the real-time requirement of the vehicle networking application; further, the double sub-function structure of the objective function balances the benefit demands of both jobs, and the resource exchange is converted into mutual beneficial behavior through the compensation mechanism, avoiding system instability caused by unilateral resource occupation; as a result, the completion time of the high-priority job is reduced, and the overall response capability of the vehicle job is improved.

[0015] In a second aspect, the application provides a resource exchange device in a vehicle Internet of Things edge computing, the device comprising:

[0016] a vehicle job obtaining module, configured to obtain a plurality of jobs in a queue waiting for computation in an edge computing node; the jobs are tasks generated by vehicles, and each job occupies a plurality of resource segments of the edge computing node;

[0017] a job pair determining module, configured to determine a plurality of candidate job pairs from the plurality of jobs; any one of the job pairs comprises a first job and a second job, the first job has a shorter queuing time than the second job, and a queuing time interval between the first job and the second job is greater than a preset minimum time slice length;

[0018] a utility function determining module, configured to determine a first utility function of the first job in each of the job pairs, and determine a second utility function of the second job; the first utility function comprises a difference between a first benefit item and a first cost item, the first benefit item is a QoE value obtained by the first job in a case of exchanging a latest available resource segment of the first job with an earliest available resource segment of the second job, and the first cost item is a compensation paid by the first job to the second job in the case of exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job; the second utility function comprises a difference between a second benefit item and a second cost item, the second benefit item is a compensation obtained by the second job paid by the first job in the case of exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job, and the second cost item is a reduced QoE value of the second job in the case of exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job; wherein the QoE value is associated with a soft deadline of the first job or the second job;

[0019] a target function determining module, configured to determine a target function and a corresponding decision item; the target function comprises a first sub-function and a second sub-function, the first sub-function is to maximize a sum of the first utility functions of the plurality of job pairs, and the second sub-function is to maximize a sum of the second utility functions of the plurality of job pairs; the decision item comprises the plurality of candidate job pairs, a compensation unit price between any one of the job pairs, and a resource exchange amount between any one of the job pairs; wherein the compensation is associated with a product of the compensation unit price and the resource exchange amount;

[0020] a target function solving module, configured to solve the target function to obtain an optimal solution of the decision item;

[0021] A resource segment interaction module is configured to exchange the corresponding resource segment according to the optimal solution of the decision item.

[0022] In a third aspect, the present application provides an electronic device, comprising at least one controller and a memory connected to the controller; the memory stores instructions executable by the at least one controller, and the instructions are executed by the at least one controller to enable the at least one controller to perform the resource exchange method in the vehicle networking edge computing as described above.

[0023] In a fourth aspect, the present application provides a computer-readable storage medium, which stores computer-executable instructions for enabling a computer to perform the resource exchange method in the vehicle networking edge computing as described above.

[0024] Additional aspects and advantages of the present application will be made apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments or related technical description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0026] Figure 1 is a flowchart of an embodiment of the resource exchange method in the vehicle networking edge computing provided by the present application;

[0027] Figure 2 is a structural schematic diagram of an embodiment of the resource exchange device in the vehicle networking edge computing provided by the present application;

[0028] Figure 3 is a structural schematic diagram of an embodiment of the electronic device provided by the present application. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solutions and advantages of the present application more clear, the following will further describe the present application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0030] In the description of the present application, if there is a description of first, second, etc., it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or the sequence of indicated technical features.

[0031] In the description of the present application, it should be understood that the orientation description, such as the orientation or position relationship indicated by up, down and the like, is based on the orientation or position relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a particular orientation, be constructed and function in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0032] In a vehicle edge computing system, when the quantity of vehicle operation requests exceeds the processing capacity threshold of the edge computing node, the queued operations will cause the resource allocation mechanism to fail, and further cause the nonlinear growth of operation completion time. This problem is caused by the fact that the edge computing node cannot dynamically adjust the fragmented resource allocation strategy, so that high-priority operations cannot be executed in time due to resource congestion, and the overall response capability of the system is thus restricted.

[0033] For example, in the scenario of roadside unit deployed at the intersection of urban trunk roads, multiple networked vehicles simultaneously submit real-time traffic flow prediction tasks, and each task occupies multiple discrete resource fragments of the edge computing node. When traffic congestion occurs at the intersection, vehicles submit high-computing-complexity tasks in a centralized manner, which causes the resource queue of the edge computing node to quickly saturate. At this time, newly arrived operations are forced to prolong the waiting time because the resource fragments are continuously occupied by long-tail tasks, vehicles cannot obtain navigation update results within the soft deadline constraint, and the service continuity of the system is thus interrupted.

[0034] If this technical problem is not effectively solved, the resource queue of the edge computing node will continue to be in an overloaded state, and the operation completion time will uncontrollably prolong with the increase of system load. The continuous deterioration of resource allocation efficiency will cause the task backlog to intensify, eventually causing the risk of service quality decline and service interruption, and the system cannot meet the basic requirement of real-time response of vehicle networking applications.

[0035] As Figure 1 One embodiment of the present application provides a resource exchange method in vehicle networking edge computing, the method comprising:

[0036] Step S100, obtaining a plurality of operations queuing for calculation in an edge computing node; the operation is a task generated by a vehicle, and each operation occupies a plurality of resource fragments of the edge computing node;

[0037] Step S200, determining a plurality of operation pairs from the plurality of operations; wherein any operation pair includes a first operation and a second operation, the queuing time length of the first operation is lower than that of the second operation, and the queuing time interval between the first operation and the second operation is greater than a preset minimum time slice length;

[0038] In step S300, a first utility function of a first job in each job pair is determined, and a second utility function of a second job is determined; the first utility function includes a difference between a first benefit item and a first cost item, the first benefit item is a QoE value obtained by the first job in a case that a latest available resource segment of the first job is exchanged with an earliest available resource segment of the second job, and the first cost item is a compensation paid by the first job to the second job in the case that the latest available resource segment of the first job is exchanged with the earliest available resource segment of the second job; the second utility function includes a difference between a second benefit item and a second cost item, the second benefit item is a compensation paid by the first job in a case that the latest available resource segment of the first job is exchanged with the earliest available resource segment of the second job, and the second cost item is a reduced QoE value of the second job in the case that the latest available resource segment of the first job is exchanged with the earliest available resource segment of the second job; wherein the QoE value is associated with a soft deadline of the first job or the second job.

[0039] In step S400, a target function and a corresponding decision item are determined; the target function includes a first sub-function and a second sub-function, the first sub-function is to maximize a sum of the first utility functions corresponding to the plurality of job pairs, and the second sub-function is to maximize a sum of the second utility functions corresponding to the plurality of job pairs; the decision item includes all candidate plurality of job pairs, a compensation unit price between any one of the job pairs, and a resource exchange amount between any one of the job pairs; wherein the compensation is associated with a product of the compensation unit price and the resource exchange amount.

[0040] In step S500, the target function is solved to obtain an optimal solution of the decision item.

[0041] In step S600, the exchange of the corresponding resource segments is completed according to the optimal solution of the decision item.

[0042] In step S100, the edge computing node can be a road side unit (RSU). In this embodiment, it is assumed that there are a plurality of jobs in the edge computing node which are queuing and waiting for calculation, the job is a task generated by a vehicle, and the vehicle unloads the task in the edge computing node after generating the task, and the task enters a queuing queue and waits for the edge computing node to perform calculation. The edge computing node has a plurality of computing resource segments, and each job needs to occupy a certain resource segment.

[0043] This embodiment is to solve the problem of resource exchange between jobs in the queue of the edge computing node, and this step S200 selects all candidate plurality of job pairs from all jobs in the queuing queue, the candidate job pair refers to a job combination that can meet the queuing time difference and time interval constraints. Each job pair includes a first job and a second job, and the first job and the second job can determine whether to exchange the corresponding resource segments.

[0044] It should be noted that the candidate multiple job pairs are selected from all jobs, but the candidate multiple job pairs are not the multiple job pairs actually performing resource segment exchange, but need to pass subsequent target function calculation to obtain the optimal solution, and determine the multiple job pairs actually performing resource segment exchange based on the optimal solution.

[0045] In the embodiment, first, the queuing time length of the first job is lower than the queuing time length of the second job, that is, the second job is first in the queuing queue, and according to the existing queuing condition, in the case that the first job cannot meet the requirement of the soft deadline, the first job can request the second job to exchange the corresponding resource segment, so that the edge computing node can perform calculation for the first job as soon as possible to meet the soft deadline of the first job. The soft deadline refers to a flexible time limit for the completion time of the job.

[0046] Secondly, the embodiment sets the queuing time interval between the first job and the second job to be greater than the preset minimum time slice length, allowing the jobs to exchange the corresponding resource segments within the time interval of the minimum time slice length.

[0047] Then, the embodiment also provides an incentive strategy, that is, in the case that the first job requests to exchange the resource segment with the second job, the first job needs to give corresponding compensation for the exchanged resource segment, and the second job can decide whether to exchange the resource segment with the first job according to the compensation.

[0048] In step S300, for each job pair, a first utility function of the first job and a second utility function of the second job are calculated. Each utility function includes a benefit item and a cost item, and the net profit item is the difference between the benefit item and the cost item. Each utility function is described as follows:

[0049] The first utility function includes the difference between the first benefit item and the first cost item;

[0050] Because the second job enters the queuing queue first, it has a resource segment that can be used earlier than the first job, so the first benefit item is the QoE value obtained after exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job;

[0051] The first cost item is the compensation paid to the second job after exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job;

[0052] For the first job, it needs to request other jobs to exchange resources with it to improve the QoE value of the first job, and the cost is the compensation paid to the second job.

[0053] The second utility function includes the difference between a second benefit term and a second cost term;

[0054] The second benefit term is the compensation paid by the first job after exchanging the last available resource segment of the first job with the earliest available resource segment of the second job;

[0055] The second cost term is the reduced QoE value after exchanging the last available resource segment of the first job with the earliest available resource segment of the second job.

[0056] For the second job, it is the requested party, and it needs to choose whether to exchange resource segments between the first job according to the compensation given by the first job. Of course, its benefit is the compensation given by the first job, because the second job exchanges the resource segment that can be calculated as soon as possible to the first job, and the resource segment obtained in exchange is the resource segment that is calculated later in the queue, so the cost is the reduced QoE value.

[0057] In step S400, the objective function and the corresponding decision term are determined; the decision term includes:

[0058] All candidate multiple job pairs, compensation unit price, and resource exchange amount; different compensation unit prices and resource exchange amounts can determine whether the second job agrees to exchange resource segments between the first job in any job pair. Finally, the optimal solution of multiple job pairs needs to be calculated, so as to realize the resource exchange between jobs in the queue.

[0059] In this application, any job aims to maximize its own benefit, and the first job can dynamically adjust the completion time and resource exchange amount according to the compensation unit price of the resource segment to optimize its utility. In turn, the second job sets the compensation unit price according to the resource segment required by the first job to maximize its utility. The utility of the second job depends on the resource segment requested by the first job for exchange, which is affected by the compensation unit price decision provided by the second job.

[0060] The first job obtains resource exchange service by paying a certain compensation to the second job, and the first job hopes to obtain as much utility as possible, so when interacting with the second job, the first job determines the resource exchange amount according to the compensation unit price of the second job. This will maximize the utility by allowing the first job to pay the lowest cost while obtaining more benefits due to the improvement of QoE.

[0061] The second job needs to determine its compensation unit price according to the change of its QoE. Based on this, the second job should set a reasonable compensation unit price to attract the first job while ensuring its own utility. Obviously, the goal of the second job is also to maximize its own utility.

[0062] Therefore, the embodiment sets the following two objective functions:

[0063] The first sub-function is to maximize the sum of the first utility functions corresponding to the plurality of job pairs.

[0064] The second sub-function is to maximize the sum of the second utility functions corresponding to the plurality of job pairs.

[0065] In steps S500 and S600, the optimal solution of the decision term is obtained by solving the objective function, that is, the plurality of job pairs are obtained, and then the exchange of resources can be performed based on the plurality of job pairs.

[0066] The embodiment solves the queue congestion problem by constructing a dynamic exchange mechanism of resource segments between jobs. Specifically, by using the characteristic that the first job is shorter than the second job, combining the exchange operation of the earliest available resource segment and the latest available resource segment, quantifying the QoE value change and compensation relationship of the job based on the utility function, and maximizing the sum of the utilities of the two jobs by the objective function. Thus, under the condition that the computing capability of the edge computing node is limited, the efficient reallocation of computing resources is realized, and the completion time of the vehicle job is effectively reduced.

[0067] In the Internet of Vehicles edge computing scene, multiple vehicle jobs queuing for computing in the edge computing node are acquired, each job occupying multiple resource segments of the edge computing node, which provides a basic data source for resource exchange. Among them, all possible job pairs are determined from multiple jobs, any job pair includes a first job and a second job, the queuing time of the first job is shorter than that of the second job, and the queuing time interval between them is greater than the preset minimum time slice length, which ensures the feasibility of the exchange operation on the time slice scale and avoids the risk of resource fragmentation. Further, the first utility function and the second utility function in each job pair are determined, the first utility function includes the difference between the first income item and the first cost item, the first income item is the QoE value obtained after exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job, and the first cost item is the compensation paid to the second job; the second utility function includes the difference between the second income item and the second cost item, the second income item is the compensation obtained by the first job, and the second cost item is the reduced QoE value after exchange. The QoE value is directly related to the soft deadline of the job, which quantifies the dependence of user experience on time sensitivity. The objective function is determined, including the first sub-function and the second sub-function, the first sub-function is to maximize the sum of the first utility functions of the multiple job pairs, and the second sub-function is to maximize the sum of the second utility functions of the multiple job pairs; the decision item includes all possible job pairs, the unit price of compensation between any job pair, and the resource exchange amount, wherein the compensation is associated with the unit price of compensation and the resource exchange amount. Thus, the objective function is solved to obtain the optimal solution of the decision item, and the exchange of the corresponding resource segment is completed according to the optimal solution, which accurately matches the resource idle period and efficiently releases the computing resources.

[0068] The scheme constructs a dynamic exchange mechanism for resource segments between jobs, effectively solving the queue congestion problem caused by limited computing capacity of the edge computing node; among them, the job pairs are screened based on the queuing time difference and time interval constraint to ensure the feasibility of exchange, and the design of the utility function associates the QoE value with the soft deadline, making the benefit and cost evaluation closely match the real-time requirements of Internet of Vehicles applications; further, the double sub-function structure of the objective function balances the interests of both parties, and the exchange of resources is converted into mutual behavior through the compensation mechanism, avoiding system instability caused by unilateral resource preemption; thus, the resource utilization rate of the edge computing node is improved, and the overall response capability of the vehicle job is improved.

[0069] In practical applications, in the resource exchange optimization process, because the objective function contains both continuous variables (compensation unit price and resource exchange quantity) and discrete variables (combination of job pairings), traditional solving methods face the challenges of high computational complexity and slow convergence speed, making it difficult to quickly obtain a globally optimal solution in the real-time environment of vehicle networking edge computing, leading to an increase in queue congestion caused by resource exchange decision delays and affecting the timely completion of vehicle jobs.

[0070] To this end, the application further proposes solving the objective function in step S500 to obtain the optimal solution of the decision term, including:

[0071] Step S510, solving the first and second sub-functions based on the differential evolution pricing and resource exchange quantity solving method to obtain the optimal solution of the compensation unit price between any one job pairing and the resource exchange quantity between any one job pairing;

[0072] Step S520, solving the first and second sub-functions based on the Gale-Shapley many-to-many pairing method to obtain the optimal solution of all candidate many-to-many job pairings.

[0073] The differential evolution pricing and resource exchange quantity solving method refers to a global optimization algorithm based on population iteration, which can use mutation operations to generate candidate solutions, crossover operations to fuse solution vectors, and selection operations to retain high-quality solutions to achieve efficient search of continuous variables. The purpose is to avoid the traditional gradient descent method from falling into local optimization in non-convex utility functions, thereby quickly approaching the globally optimal solution of the compensation unit price and resource exchange quantity under limited computing resources. The Gale-Shapley many-to-many pairing method can be understood as a two-way stable matching mechanism, which can use the method of constructing a two-way preference list of jobs and performing an iterative proposal-acceptance process to achieve optimization of discrete pairings. The purpose is to avoid the combinatorial explosion problem caused by exhaustive search through the stable matching principle, ensuring that any job pairing combination satisfies the constraint of maximizing utility for both parties.

[0074] Specifically, the method decomposes the mixed variable optimization problem into two sub-problems of continuous variables and discrete variables by the divide-and-conquer strategy: the differential evolution algorithm optimizes the continuous variables in the first and second sub-functions, dynamically adjusts the search direction of the compensation unit price and the resource exchange amount by using the population iteration mechanism, and generates a feasible solution domain according to the nonlinear characteristics of the utility function; at the same time, the Gale-Shapley matching method constructs the one-way preference list of the first job to the second job and the second job to the first job based on the utility value in the solution domain, and determines the stable matching relationship through multiple rounds of matching iteration. Both of them form a closed-loop optimization mechanism - the continuous variable solution domain output by the differential evolution algorithm provides the utility value input for the Gale-Shapley matching, and the matching result is fed back to the continuous variable optimization process, which significantly reduces the overall solution complexity, thereby ensuring the economic efficiency of resource exchange while realizing the real-time convergence of the decision-making process.

[0075] In this embodiment, the differential evolution algorithm adopts a standard population initialization strategy, in which the population size is dynamically adjusted according to the real-time load of the edge computing node, the mutation operation generates a difference vector by randomly selecting parent individuals, and the crossover operation fuses the parent and child solution vectors with a fixed probability; in the Gale-Shapley matching stage, first, the utility value of each job is calculated based on the compensation unit price and the resource exchange amount output by the differential evolution, then the preference list of the first job in descending order of the utility value of the second job is constructed, and the preference list of the second job in descending order of the utility value of the first job is constructed, and through multiple rounds of two-way matching iteration until all jobs reach stable matching. This method can quickly obtain the global optimal solution of resource exchange decision in the real-time queue management of the edge computing node of the Internet of Vehicles, effectively shorten the solution time and reduce the consumption of computing resources, thereby timely relieving the queue congestion problem of the edge computing node and ensuring the efficient completion of vehicle jobs under the constraint of soft deadline.

[0076] In practical applications, in some embodiments of the present application, the first benefit item is used to quantify the influence of resource exchange on the QoE value of the job, however, in the implementation process, the calculation of the QoE value lacks an accurate mathematical model, which leads to the inability to accurately evaluate the net benefit change brought by the exchange, affecting the optimization efficiency and fairness of the resource exchange decision. Therefore, in some embodiments of the present application, the first benefit item includes: the difference between the QoE value of the first job after the resource segment exchange and the QoE value of the first job before the resource segment exchange; in this embodiment, the first benefit item refers to the net influence of the resource segment exchange operation on the QoE value of the first job, which can be quantified by the difference between the QoE values before and after the exchange, the purpose of which is to focus on the value increment of the exchange behavior itself, avoiding the defect of relying only on the absolute QoE value and ignoring the influence of dynamic adjustment.

[0077] The QoE value of the first job before the resource segment exchange includes: 1 minus a first natural exponential function, the first natural exponential function being a natural exponential of a negative ratio of a first function to a soft deadline of the first job, the first function being a difference between the soft deadline of the first job and a completion time of the first job;

[0078] The QoE value of the first job after the resource segment exchange includes: 1 minus a second function, the second function including a product between a second natural exponential function and a third natural exponential function, the second natural exponential function being a natural exponential of a negative ratio of a third function to a soft deadline of the first job, the third function being a difference between the soft deadline of the first job and a completion time of the resource segment of the first job after the exchange; and the third natural exponential function being a natural exponential of a negative ratio of a resource exchange amount to a maximum resource exchange amount.

[0079] The QoE value of the first job before the resource segment exchange can be understood as an exponential decay model based on time sensitivity, which can be implemented in the form of 1 minus a first natural exponential function, the first natural exponential function normalizing a time difference by a ratio of a first function to a soft deadline, the first function being a difference between the soft deadline of the first job and a completion time of the first job. In practical applications, the QoE value of the first job after the resource segment exchange is a composite function considering the completion time after the exchange and the resource exchange amount, which can be in the form of 1 minus a second function, the second function being a product between a second natural exponential function and a third natural exponential function, the second natural exponential function being a natural exponential of a negative ratio of a third function to a soft deadline of the first job, the third function being a difference between the soft deadline of the first job and a completion time of the resource segment of the first job after the exchange, and the third natural exponential function being a natural exponential of a negative ratio of a resource exchange amount to a maximum resource exchange amount. The purpose is to reflect the marginal effects of time delay and resource exchange amount on QoE at the same time, so as to balance resource utilization efficiency and user experience in the revenue calculation.

[0080] In particular, the embodiment directly quantifies the net revenue change brought by resource exchange by defining the first revenue item as the difference between the QoE values before and after the exchange. The QoE value before the exchange of the resource segment is calculated based on the difference between the soft deadline and the completion time, and a natural exponential function is used to capture the exponential decay characteristic of QoE with time delay; the QoE value after the exchange of the resource segment is calculated by introducing the completion time after the exchange and the amount of resource exchange, wherein the third function continues the consistency of the time-sensitive model, and the third natural exponential function normalizes the potential negative impact of the exchange operation on QoE by the ratio of the amount of exchange to the maximum amount of exchange. Overall, the model builds a dynamic and quantifiable QoE evaluation framework, enabling the objective function to be solved based on accurate revenue data to optimize job pairing and resource exchange decisions. The net revenue of resource exchange is accurately quantified, thereby optimizing the efficiency and fairness of resource exchange decisions and ensuring the adaptability and robustness of resource allocation in the edge computing environment of the Internet of Vehicles.

[0081] Similarly, in some embodiments of the application, the second revenue item includes the difference between the QoE value of the second job after the exchange of the resource segment and the QoE value of the second job before the exchange of the resource segment; the second revenue item is a quantitative indicator of the impact of resource exchange on the experience quality of the second job, which can be realized in the form of the difference between the QoE values before and after the exchange, with the purpose of converting the abstract experience quality into a calculable optimization parameter.

[0082] The QoE value of the second job before the exchange of the resource segment includes: 1 minus the first natural exponential function, the first natural exponential function being the natural exponential of the negative of the ratio of the first function to the soft deadline of the second job, the first function being the difference between the soft deadline of the second job and the completion time of the second job;

[0083] The QoE value of the second job after the exchange of the resource segment includes: 1 minus the second function, the second function including the product between the second natural exponential function and the third natural exponential function, the second natural exponential function being the natural exponential of the negative of the ratio of the third function to the soft deadline of the second job, the third function being the difference between the soft deadline of the second job and the completion time of the resource segment of the second job after the exchange; the third natural exponential function being the natural exponential of the negative of the ratio of the amount of resource exchange to the maximum amount of resource exchange.

[0084] In particular, the embodiment directly relates the actual impact of resource exchange on the second job experience quality by defining the second revenue item as the difference between the QoE values before and after the exchange. In the calculation of the QoE value before the exchange of the resource segment, the exponential function based on the difference between the soft deadline and the completion time is subtracted by 1. When the completion time approaches the soft deadline, the function value approaches 0, making the QoE close to 1, which reflects the criticality of low delay for vehicle networking applications. When the delay increases, the function value rises, causing the QoE to rapidly decrease, accurately reflecting the sensitive demand of vehicle operations for timeliness. In the calculation of the QoE value after the exchange of the resource segment, the second function quantifies the benefits brought by the improvement of the completion time after the exchange through the second natural exponential function, and reflects the additional cost caused by the increase of the exchange amount through the third natural exponential function. The product form of the two ensures that the joint influence of time delay and exchange amount on QoE is comprehensively evaluated. For example, when the exchange amount increases, the third natural exponential function decays faster to suppress the improvement of QoE, thereby avoiding excessive exchange of resources. Overall, these features collectively convert the abstract QoE into a differentiable mathematical model, enabling the solution of the objective function to accurately capture changes in revenue. The embodiment realizes accurate quantification of the second job revenue in resource exchange, avoids the problem of relying on fuzzy compensation definition, enables resource exchange decisions to be based on objective QoE value changes for global optimization, and effectively improves the efficiency and fairness of resource allocation in vehicle networking edge computing.

[0085] In some embodiments of the present application, the completion time of the resource segment of the first job after the exchange includes:

[0086] ;

[0087] wherein, is the completion time of the resource segment of the first job after the exchange, is the end time of the exchangeable resource segment of the first job , is the exchange amount between the first job and the second job , is the computing capacity corresponding to the edge computing node, is the start time of the exchangeable resource segment of the second job , is the judgment statement.

[0088] Specifically, the embodiment dynamically calculates the completion time after the exchange through a segmented function: when the exchange amount of the resource segment does not exceed half of the product of the time interval and the calculation capacity, the completion time is obtained by subtracting the time delay corresponding to the exchange amount from the original end time of the first job; when the exchange amount exceeds the threshold, the completion time is determined by adding the time delay corresponding to the exchange amount to the start time of the resource segment of the second job. The embodiment realizes accurate calculation of the completion time of the first job after the exchange, avoids estimation deviation of the QoE value, and makes the resource exchange decision more in line with actual physical constraints.

[0089] In some embodiments of the present application, the completion time of the resource segment of the second job after the exchange includes:

[0090] ;

[0091] wherein, is the completion time of the resource segment of the second job after the exchange, is the end time of the exchangeable resource segment of the first job , is the end time of the exchangeable resource segment of the second job , is a maximum function.

[0092] The embodiment can accurately calculate the completion time of the resource segment of the second job after the exchange, effectively avoid the deviation of the QoE value evaluation, and make the resource exchange decision more accurately reflect the actual benefit.

[0093] In some embodiments of the present application, the completion time of any job includes:

[0094] ;

[0095] ;

[0096] ;

[0097] ;

[0098] wherein, is the completion time of the job , is the transmission energy, is the channel gain, is the background noise, is the spectral bandwidth, is the data transmission rate, is the amount of data transmitted by the job , is the transmission delay, is the job Generation time, For homework Total number of CPU cycles required The computing power corresponding to the edge computing nodes. Minimum time slice length, For homework computational delay, This refers to the idle time corresponding to the edge computing node. This is the function for finding the maximum value.

[0099] Specifically, this embodiment achieves dynamic calculation of job completion time through a hierarchical and progressive mechanism: First, the `max` function is used to compare the idle time of edge computing nodes with the job generation time plus transmission delay in real time to determine the starting point of the waiting window, avoiding underestimation of completion time due to ignoring resource fluctuations; second, based on the job's CPU cycle requirements, node computing power, and minimum time slice length, the computation delay is quantified by rounding up, adapting resource consumption assessment to discrete allocation characteristics; finally, the data volume is bound to the dynamic transmission rate, which strictly follows the wireless channel characteristic formula, integrating spectrum bandwidth, transmission energy, channel gain, and background noise parameters to ensure that the transmission process model conforms to the actual communication environment of the Internet of Vehicles. These steps form an organic whole, incorporating computation, transmission, and idle time into a unified framework, making completion time a quantifiable dynamic indicator and providing reliable input for QoE value calculation. This embodiment can more accurately quantify QoE value, enabling resource exchange decisions based on a dynamically coupled completion time model, effectively avoiding decision-making errors caused by deviations in completion time calculation.

[0100] One embodiment of this application provides a resource exchange method in edge computing for vehicle-to-everything (V2X) networks, the method comprising the following:

[0101] Step S910: Establishment of the resource exchange system;

[0102] Operations generated by vehicles within the coverage area of ​​the Roadside Unit (RSU) can be unloaded to the RSU.

[0103] When the computational load of a job exceeds its computational capacity, the job needs to exchange computational resources. This method aims to solve the resource swapping problem in the queuing queue, maximizing the job's QoE while considering the job's own interests.

[0104] During resource exchange, jobs in the resource exchange system are allowed to operate within a minimum time slice length. Exchange resource fragments within time intervals. ,by Taking time as an example, the queue contains JoEs and JoLs.

[0105] JoE refers to the job that enters the queue before and is still waiting for computation, denoted as .

[0106] JoL refers to the job that enters the queue between and and is still waiting for computation, denoted as .

[0107] The set of JoEs and JoLs is defined as .

[0108] If the existing queuing situation cannot meet the requirements of the soft deadline of JoL, JoL will initiate a resource exchange request. JoE will decide whether to provide resource exchange services according to its own interests.

[0109] In the RSU, there are service stations, each of which can compute jobs in parallel, and is denoted as . Each service station has a computing capacity . Considering the resource exchange, a job may have multiple resource segments, and when using time slices as the measurement unit, set as the minimum time slice length that can be used for unloading or exchange. After these configurations are in place, in order to achieve the expected effect of resource exchange, JoE will use its earliest available resource segment, and JoL will use its latest available resource segment for exchange.

[0110] In this embodiment, jobs are randomly generated by vehicles, each has attributes represented as , that is, is generated by , represents the total CPU cycle number that the job needs to compute, represents the amount of data that the job needs to transfer, represents the generation time of , represents the soft deadline of .

[0111] The RSU is responsible for matching the job pairing (resource segment exchange pair) of all jobs of IoLs and JoLs to maximize the interests of both parties and ensure the efficiency and accuracy of the pairing. Use to describe all possible job pairings, where , that is, equals the second job , Equivalent to the first assignment mentioned above In addition to determining the pairings, the resource exchange volume and compensation unit price must also be determined.

[0112] make for and The amount of resources exchanged between them, of which , This indicates the maximum number of exchanges. This indicates that the exchange quantity is zero. To indicate that it involves and The exchange quantity. Let for Towards The unit price of compensation collected, of which, and These represent the minimum and maximum limits for the compensation unit price, respectively.

[0113] Considering that the initial contiguous resources allocated by RSU may be replaced by multiple resource fragments, let and respectively This indicates the start and end times of the resource segment occupying the RSU. Let... and for The start and end times of exchangeable resource fragments are used. The service desk is used to identify the RSU to which the resource fragment belongs. .

[0114] at last, express The collection of all resource fragments occupied by the server. express At the service desk A collection of resource fragments on the platform. , express At the service desk The possession of the first The start and end times of each resource segment.

[0115] Assuming each vehicle is equipped with GPS, Orthogonal Partial Multiple Access (OFDMA) is used to ignore inter-channel interference in data transmission. The Shannon formula for the task data transmission rate is as follows:

[0116] ;

[0117] in, To transfer energy, For channel gain, For background noise, The spectrum bandwidth.

[0118] The transmission delay can be expressed as:

[0119] ;

[0120] The time at which the job arrives at the queue is:

[0121] ;

[0122] The computing capacity of the service station and the minimum time slice length , the computing delay can be calculated as:

[0123] ;

[0124] The next free time of the service station is the sum of the start time of the task currently being executed and the required execution duration. Once the service station becomes available, the RSU can provide computing resources, so the free time of the RSU is , and it is known that this free service station is , so the completion time of the job can be obtained as:

[0125] ;

[0126] Considering the impact of the short board effect on the completion time of the job, the impact on the completion time of the job needs to be exchanged. Specifically, let be the completion time of the exchanged resource segment, and let be the completion time of the exchanged resource segment. The solution is as follows:

[0127] ;

[0128] ;

[0129] If the RSU cannot meet the requirements of the job, the degree of user dissatisfaction will increase sharply with the increase of the completion time, resulting in a significant decline in QoE. However, if the RSU only meets the delay requirement, when the job completion time reaches a certain threshold, the QoE perceived by the user will tend to be blunt with the decrease of the completion time, and finally only converges to a stable value without significant improvement. The above characteristics are highly consistent with the definition of basic demand type of product in Kano model.

[0130] In order to simulate this feature, the QoE trend is fitted to the basic demand function in Kano model. When , the slope of the function is 1, and when The slope is greater than 1 and increases as decreases, which reflects the situation in real-world scenarios. If the basic requirement is not met, the user's QoE will decrease rapidly as the requirement satisfaction deteriorates, and the QoE can approach negative infinity. Conversely, when , the slope of the function is less than 1 and gradually decreases as increases, which means that if the basic requirement is met, the rate of QoE will gradually slow down as the requirement satisfaction increases, and the upper limit of QoE is 1.

[0131] Therefore, to better describe the relationship between job completion time and QoE, from the perspective of all resource segments occupied by a job, let the performance of the product be , represent the completion time of . Therefore, the QoE evaluation of all resource segments of a single job includes:

[0132] ;

[0133] By defining QoE in this way, the fact that the ratio equals 0 when the completion time of a job is exactly equal to the soft deadline can be reflected. As the completion time decreases, the ratio increases, and the enhancement of the vehicle's performance on the product becomes more and more sluggish, as fitted by the formula. As the completion time increases, the ratio decreases, and the deterioration of the vehicle's performance on the product becomes more sensitive, as fitted by the formula.

[0134] For JoE, the completion time of the resource segment after resource exchange is only affected by the pairing relationship. Once the pairing relationship is established, the change in QoE is also determined. Therefore, in order to achieve greater utility, JoE aims to increase the exchange amount to enhance the compensation income.

[0135] For JoL, the goal is to obtain more early available computing resources to improve QoE. However, simply obtaining earlier resources without considering the exchange amount is unreliable. For example, if a job only needs a very small amount of very early available resources, measuring the QoE gain in this way will overestimate the actual improvement. This is because the overall completion time is affected by the short board effect, and such a small amount of resources has little effect on the overall latest completion time.

[0136] To prevent valuation distortion, an exchange amount-based incentive design is considered. Therefore, for JoE and JoL, the QoE of a single task for a certain resource segment is:

[0137] ;

[0138] ;

[0139] The above two equations are set as the utility functions of the jobs respectively.

[0140] Meanwhile, the increase or decrease of the satisfaction of JoE and JoL due to the exchange can be expressed as follows:

[0141]

[0142]

[0143] The first half of the right side of the equation represents the QoE of the resource segment obtained after the exchange of resources, and the second half of the right side of the equation represents the QoE of the resource segment before the exchange of resources. In this way, the increase or decrease of QoE will be strictly related to the soft deadline, not just the completion time before and after the exchange of resources.

[0144] Step S920, constructing the utility function;

[0145] The benefit of the exchange of resources is the compensation obtained by exchanging the resource segment available earlier from The cost is the decrease of QoE value before and after the exchange. The utility function of JoE (i.e., the second utility function) is expressed as follows:

[0146]

[0147] The benefit of the exchange of resources is the compensation obtained by exchanging the resource segment available earlier from The cost is the decrease of QoE value before and after the exchange. The utility function of JoE (i.e., the first utility function) includes:

[0148]

[0149] wherein represents the cost weight factor, and represents the urgency of the job.

[0150] Step S930, determining the utility maximization problem based on the utility function;

[0151] Since JoL and JoE both aim to maximize their own benefits, JoL dynamically adjusts the completion time and the exchange amount according to the resource price to optimize its utility. JoE, in turn, sets the price according to the demand of JoL to maximize its utility. The utility of JoE depends on the resource segments proposed by JoL, which are affected by the compensation unit price decision provided by JoE. In order to maximize the utility of both parties, the optimal strategy of JoL must be consistent with the optimal response of JoE. Therefore, the above two utility functions are converted into the following two problems.​​​​

[0152] Utility maximization of JoL (i.e., the first sub-function):

[0153] JoL obtains resource exchange service by paying JoE compensation and hopes to obtain as much utility as possible. Therefore, when interacting with JoE, JoL must determine the amount of resource exchange according to the compensation unit price of JoE:

[0154] ;

[0155] wherein, is the optimal price given by . The constraint condition indicates that the end time of the exchangeable resource segment must be greater than the start time of the resource segment used for resource exchange, so that the resource segment exchange makes sense. The constraint condition provides that the total amount of exchanged resource segments must not exceed the maximum available exchange resource amount, wherein . The constraint condition indicates that the utility function of must be greater than 0.

[0156] Utility maximization for JoE: JoE needs to determine its compensation unit price according to the change of its QoE. Based on this, JoE should set a reasonable compensation unit price to attract JoL while ensuring its own utility. Obviously, the goal of JoE is also to maximize its own utility (i.e., the second sub-function):

[0157] ;

[0158] wherein, is the optimal resource exchange amount determined by . The constraint condition indicates that the utility function of must be greater than , and the constraint condition indicates that the pricing of the compensation unit price needs to be reasonable.

[0159] After defining the utility functions for JoEs and JoLs and constructing the utility maximization problem, the mutual exclusivity of interests must be considered. This requires integrating the problems and to find a Nash equilibrium solution. This solution will determine the job pairing set , the specific compensation unit price , and the specific resource exchange amount , ensuring that neither party has the motivation to change their behavior unilaterally while the other party's behavior remains unchanged.

[0160] ​Step S940, solution of the utility maximization problem;

[0161] First, the existence and uniqueness of this Nash equilibrium solution is proved in detail using backward induction.

[0162] Stackelberg equilibrium analysis;

[0163] In the case of the optimal compensation unit price of JoE and the optimal resource exchange amount of JoL , when satisfies the following conditions.

[0164] For the first stage JoE , if the pricing satisfies the following requirements.

[0165] ;

[0166] For each JoL in the second stage , if the resource exchange amount satisfies the following conditions.

[0167] ;

[0168] Let be denoted as the equilibrium point.

[0169] It will be proved that there is an equilibrium solution for this game, and the equilibrium point is obtained by using backward induction:

[0170] For about , the first-order derivative is as follows:

[0171] ;

[0172] where , and . Then, the second-order derivative of about is as follows:

[0173] ;

[0174] Obviously, the second-order derivative is less than zero, and has strict convexity with respect to . It can be proved that is unique. If the first-order derivative is zero, then the variational relationship between and can be obtained as follows:

[0175] ;

[0176] Based on the above relationship, one can calculate the maximum unit price and the minimum unit price when respectively. Meanwhile, by setting the first derivative to zero, one can find the optimal as follows.

[0177] ;

[0178] Thus one can get: and . Substituting the optimal solution into the utility function of formula JoE, one can get the following results.

[0179] ;

[0180] To simplify the following derivation, let . Then, one can find its first derivative as follows.

[0181] ;

[0182] Then one can find the second derivative as follows.

[0183] ;

[0184] where , which simplifies to , where , and takes any real value , so there is . In addition, , and , obviously there is . Therefore, one can conclude that: . That is, for , has strict convexity and there is a unique maximum. Therefore, the equilibrium point of the Stackelberg game exists and is unique. This means that both parties can find the optimal price and the corresponding resource exchange scheme, thus maximizing the utility of both parties. At the same time, by letting one can solve .

[0185] ​​Through the above analysis, there is a Stackelberg equilibrium between JoLs and JoEs, involving the exchange amount and unit price. Therefore, the embodiment proposes a QoE-aware differentiated pricing (QPDP) strategy, which includes a differential evolution (DE) pricing and a resource exchange amount solving (DPES) algorithm, and a Gale-Shapley-based many-to-many pairing (MPGS) algorithm. The DPES algorithm comprehensively considers the QoE loss of JoEs in resource exchange and the QoE gain of JoLs in resource exchange to determine the optimal resource pricing and job exchange amount. The MPGS algorithm involves finding the optimal pairing between JoEs and JoLs for computing resource exchange through multiple rounds of optimal pairing, thereby maximizing the utility.

[0186] Algorithm 1: Differential Evolution Pricing and Exchange Amount Solving Algorithm (DPES)

[0187] Requirements: Job information, differential evolution parameters: scaling factor F, crossover probability rc, population size Ns, number of rounds Nr

[0188] Solving: , ;

[0189] 1. Calculate and , and then initialize , ;

[0190] 2. Generate an initial population with a length of Ns, and set ;

[0191] 3. Perform the following for to :

[0192] 3-1. Perform the following for to :

[0193] 3-1-1. Obtain from ;

[0194] 3-1-2. ;

[0195] 3-1-3. Perform the following for to :

[0196] 3-1-3-1. If ;

[0197] 3-1-3-2. Calculate , with constraints in ;

[0198] 3-1-3-3. if ;

[0199] 3-1-4. end loop;

[0200] 3-2. if perform:

[0201] 3-2-1. ;

[0202] 3-2-2. ;

[0203] 3-3. end condition check;

[0204] 3-4. if if and ;

[0205] 3-5. end loop;

[0206] 4. for to perform:

[0207] 4-1. for to perform:

[0208] 4-1-1. ;

[0209] 4-1-2. if is obtained;

[0210] 4-2. end loop;

[0211] 5. return , ;

[0212] Algorithm 2, Gale-Shapley based many-to-many pairing (MPGS) algorithm

[0213] Requirements: job information, differential evolution parameters

[0214] Solution: optimal job pairing set ;

[0215] 1. initialize ;

[0216] 2. when or is executed:

[0217] 2-1. ← DPES algorithm;

[0218] 2-2. ;

[0219] 2-3. For is executed:

[0220] 2-3-1. If , then is added to ;

[0221] 2-4. End loop;

[0222] 2-5. For is executed:

[0223] 2-5-1. If , then ;

[0224] 2-5-2. Add to ;

[0225] 2-6. For and is executed:

[0226] 2-6-1. If and , then is added to ;

[0227] 2-7. Remove , from , ;

[0228] 2-8. Adjust job information;

[0229] 2-9. ;

[0230] 3. End loop;

[0231] 4. Return ;

[0232] As shown in Figure 2 , one embodiment of the present application provides a resource exchange device in Internet of Vehicles edge computing, the device comprising:

[0233] The vehicle job obtaining module 1001 is configured to obtain a plurality of jobs in a queue for computation in an edge computing node; the job is a task generated by a vehicle, and each job occupies a plurality of resource segments of the edge computing node.

[0234] The job pair determining module 1002 is configured to determine a plurality of job pairs from the plurality of jobs; each job pair includes a first job and a second job, the first job has a shorter queuing time than the second job, and a queuing time interval between the first job and the second job is greater than a preset minimum time slice length.

[0235] The utility function determining module 1003 is configured to determine a first utility function of the first job in each job pair and a second utility function of the second job; the first utility function includes a difference between a first benefit item and a first cost item, the first benefit item is a QoE value obtained after exchanging a latest available resource segment of the first job with an earliest available resource segment of the second job, and the first cost item is a compensation paid to the second job after exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job; the second utility function includes a difference between a second benefit item and a second cost item, the second benefit item is a compensation paid by the first job after exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job, and the second cost item is a reduced QoE value after exchanging the latest available resource segment of the first job with the earliest available resource segment of the second job; the QoE value is associated with a soft deadline of the first job or the second job.

[0236] The objective function determining module 1004 is configured to determine an objective function and a corresponding decision item; the objective function includes a first sub-function and a second sub-function, the first sub-function is to maximize a sum of the first utility functions of the plurality of job pairs, and the second sub-function is to maximize a sum of the second utility functions of the plurality of job pairs; the decision item includes the plurality of job pairs, a compensation unit price between any one of the job pairs, and a resource exchange amount between any one of the job pairs; the compensation is associated with the compensation unit price and the resource exchange amount.

[0237] The objective function solving module 1005 is configured to solve the objective function to obtain an optimal solution of the decision item.

[0238] The resource segment interaction module 1006 is configured to complete exchange of corresponding resource segments according to the optimal solution of the decision item.

[0239] It should be noted that the vehicle networking edge computing resource exchange device provided in the embodiment and the vehicle networking edge computing resource exchange method provided in the embodiment are based on the same inventive concept, and therefore the content of the vehicle networking edge computing resource exchange method provided in the embodiment is also applicable to the content of the vehicle networking edge computing resource exchange device provided in the embodiment, which will not be described herein again.

[0240] As Figure 3 One embodiment of the present application further provides an electronic device, which comprises a memory and a processor. The memory stores a computer program. The processor executes the computer program to implement the vehicle networking edge computing resource exchange method.

[0241] at least one battery;

[0242] at least one memory;

[0243] at least one processor;

[0244] at least one program;

[0245] The program is stored in the memory, and the processor executes the at least one program to implement the vehicle networking edge computing resource exchange method provided in the present disclosure.

[0246] The electronic device can be any intelligent terminal, such as a mobile phone, a tablet computer, a personal digital assistant (PDA), a vehicle-mounted computer, etc.

[0247] The electronic device of the embodiment of the present application will be described in detail below.

[0248] The processor 1600 can be implemented in a general central processing unit (CPU), a microprocessor, an application specific integrated circuit (ASIC), or one or more integrated circuits, etc., and is used to execute a related program to implement the technical solutions provided in the embodiments of the present disclosure.

[0249] The memory 1700 can be implemented in the form of a Read Only Memory (ROM), a static storage device, a dynamic storage device, or a Random Access Memory (RAM), etc. The memory 1700 can store an operating system and other application programs. When the technical solutions provided by the embodiments of the present disclosure are implemented by software or firmware, the related program codes are stored in the memory 1700 and are invoked and executed by the processor 1600 to perform the resource exchange method in the vehicle networking edge computing.

[0250] The input / output interface 1800 is configured to realize information input and output.

[0251] The communication interface 1900 is configured to realize the communication interaction between the device and other devices. The communication can be realized in a wired manner (for example, a USB, a network cable, etc.) or in a wireless manner (for example, a mobile network, WIFI, Bluetooth, etc.).

[0252] The bus 2000 is configured to transmit information between various components (for example, the processor 1600, the memory 1700, the input / output interface 1800, and the communication interface 1900) of the device.

[0253] The processor 1600, the memory 1700, the input / output interface 1800, and the communication interface 1900 are connected to each other through the bus 2000 to realize the communication connection between the devices.

[0254] The present disclosure further provides a storage medium, which is a computer readable storage medium and stores computer executable instructions for causing a computer to perform the above-mentioned resource exchange method in the vehicle networking edge computing.

[0255] The memory is a non-transitory computer readable storage medium, which can be used to store non-transitory software programs and non-transitory computer executable programs. In addition, the memory can include a high-speed random access memory and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory remotely arranged relative to the processor, and these remote memories can be connected to the processor through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0256] The embodiments described in the present disclosure are to more clearly illustrate the technical solutions of the present disclosure, and do not constitute a limitation on the technical solutions provided by the present disclosure. Those skilled in the art can know that, as technology evolves and new application scenarios appear, the technical solutions provided by the present disclosure are also applicable to similar technical problems.

[0257] Those skilled in the art can understand that the technical solutions shown in the figures do not constitute a limitation on the present disclosure, and can include more or fewer steps than the figures shown, or combine certain steps, or different steps.

[0258] The device embodiments described above are merely illustrative, and units described as separate components can or can not be physically separated, that is, can be located in one place, or can be distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment.

[0259] Those skilled in the art can understand that all or some steps in the above disclosed method, functional modules / units in the system and device can be implemented as software, firmware, hardware and their appropriate combinations.

[0260] The terms "first", "second", "third", "fourth" and the like (if any) in the specification of the present application and the above-described drawings are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0261] It should be understood that, in the application, "at least one" refers to one or more, and "multiple" refers to two or more. "And / or" is used to describe the association relationship of the associated objects, which means that there can be three relationships, for example, "A and / or B" can represent three cases of only A, only B and A and B existing at the same time, wherein A and B can be singular or plural. The character " / " generally represents an "or" relationship between the associated objects before and after it. "At least one of the following" or similar expressions means any combination of these items, including any combination of single or multiple items. For example, at least one of a, b or c can represent a, b, c, "a and b", "a and c", "b and c", or "a and b and c", wherein a, b and c can be single or multiple.

[0262] In several embodiments provided in the application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed mutual units can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0263] The units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, that is, they can be located in one place, or they can be distributed on multiple network units. According to actual needs, some or all of the units can be selected to achieve the purpose of the embodiment scheme.

[0264] In addition, the functional units in each embodiment of the application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0265] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application, essentially or in other words, the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes multiple instructions for causing an electronic device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program storage media.

[0266] The above is a specific description of the preferred implementation of the embodiments of the present application, but the embodiments of the present application are not limited to the above implementation. Those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the embodiments of the present application, and these equivalent modifications or replacements are all included in the scope defined by the claims of the embodiments of the present application.

Claims

1. A resource exchange method in edge computing for vehicle-to-everything (V2X) networks, characterized in that, The method includes: Retrieve multiple jobs that are queued for computation in the edge computing node; the jobs are tasks generated by the vehicle, and each job occupies multiple resource fragments of the edge computing node; All candidate job pairs are determined from the plurality of jobs; wherein any job pair includes a first job and a second job, the queuing time of the first job is shorter than the queuing time of the second job, and the queuing time interval between the first job and the second job is greater than a preset minimum time slice length. A first utility function is determined for a first job in each job pair, and a second utility function is determined for a second job. The first utility function includes the difference between a first benefit and a first cost, where the first benefit is the QoE value gained by the first job when the latest available resource fragment of the first job is exchanged for the earliest available resource fragment of the second job, and the first cost is the compensation paid by the first job to the second job when the latest available resource fragment of the first job is exchanged for the earliest available resource fragment of the second job. The second utility function includes the difference between a second benefit and a second cost, where the second benefit is the compensation paid by the first job to the second job when the latest available resource fragment of the first job is exchanged for the earliest available resource fragment of the second job, and the second cost is the reduced QoE value of the second job when the latest available resource fragment of the first job is exchanged for the earliest available resource fragment of the second job. The QoE value is associated with the soft deadline of either the first or second job. The objective function and corresponding decision terms are determined. The objective function includes a first sub-function and a second sub-function. The first sub-function maximizes the sum of the first utility functions corresponding to the multiple job pairs, and the second sub-function maximizes the sum of the second utility functions corresponding to the multiple job pairs. The decision terms include all candidate job pairs, the compensation unit price between any of the job pairs, and the resource exchange volume between any of the job pairs. The compensation is associated with the product of the compensation unit price and the resource exchange volume. Solve the objective function to obtain the optimal solution for the decision item; The corresponding resource fragments are exchanged based on the optimal solution of the decision item.

2. The resource exchange method in edge computing for vehicle-to-everything (V2X) networks according to claim 1, characterized in that, Solving the objective function to obtain the optimal solution for the decision term includes: Based on the differential evolution pricing and resource exchange volume solution method, the first sub-function and the second sub-function are solved to obtain the optimal solution of the compensation unit price between any one of the job pairs and the optimal solution of the resource exchange volume between any one of the job pairs. The first and second sub-functions are solved using the Gale-Shapley many-to-many pairing method to obtain the optimal solution for all candidate multiple job pairings.

3. The resource exchange method in edge computing for vehicle-to-everything (V2X) networks according to claim 1, characterized in that, The first benefit item includes: the difference between the QoE value of the first job after the resource fragment swap and the QoE value of the first job before the resource fragment swap; The QoE value of the first job before the resource fragment exchange includes: the difference between 1 and the first natural exponential function, where the first natural exponential function is the natural exponent of the negative of the ratio of the first function to the soft deadline of the first job, and the first function is the difference between the soft deadline of the first job and the completion time of the first job. The QoE value of the first job after the resource fragment exchange includes: 1 and the difference of a second function, the second function being the product of a second natural exponential function and a third natural exponential function, the second natural exponential function being the natural exponent of the ratio of the third function to the soft deadline of the first job, the third function being the difference between the soft deadline of the first job and the completion time of the resource fragment of the first job after the exchange; the third natural exponential function being the natural exponent of the ratio of the resource exchange volume to the maximum resource exchange volume.

4. The resource exchange method in edge computing for vehicle-to-everything (V2X) networks according to claim 1, characterized in that, The second benefit item includes: the difference between the QoE value of the second job after the resource fragment swap and the QoE value of the second job before the resource fragment swap; The QoE value of the second job before the resource fragment exchange includes: the difference between 1 and the first natural exponential function, where the first natural exponential function is the natural exponent of the negative of the ratio of the first function to the soft deadline of the second job, and the first function is the difference between the soft deadline of the second job and the completion time of the second job. The QoE value of the second job after the resource fragment exchange includes: 1 and the difference of a second function, the second function being the product of a second natural exponential function and a third natural exponential function, the second natural exponential function being the natural exponent of the ratio of the third function to the soft deadline of the second job, the third function being the difference between the soft deadline of the second job and the completion time of the resource fragment of the second job after the exchange; the third natural exponential function being the natural exponent of the ratio of the resource exchange volume to the maximum resource exchange volume.

5. The resource exchange method in edge computing for vehicle-to-everything (V2X) networks according to claim 3, characterized in that, The completion time of the resource fragment of the first job after the exchange includes: ； in, The first assignment after the exchange The completion time of the resource fragment. As the first assignment The end time of exchangeable resource fragments. As the first assignment Second assignment The amount of resources exchanged between them The computing power corresponding to the edge computing nodes. For the second assignment The start time of exchangeable resource fragments. This is a conditional statement.

6. The resource exchange method in edge computing for vehicle-to-everything (V2X) networks according to claim 4, characterized in that, The completion time of the resource fragment of the second job after the exchange includes: ； in, For the second assignment after the exchange The completion time of the resource fragment. As the first assignment The end time of exchangeable resource fragments. For the second assignment The end time of exchangeable resource fragments. This is the function for finding the maximum value.

7. A resource exchange method in edge computing for vehicle-to-everything (V2X) networks according to claim 3 or 4, characterized in that, The completion time for any assignment includes: ； ； ； ； in, For homework Completion time, To transfer energy, For channel gain, For background noise, For spectrum bandwidth, For data transmission rate, For homework The amount of data transmitted For transmission delay, For homework Generation time, For homework Total number of CPU cycles required The computing power corresponding to the edge computing nodes. Minimum time slice length, For homework computational delay, This refers to the idle time corresponding to the edge computing node. To find the maximum value function, To Perform the rounding up operation.

8. A resource exchange device in edge computing for vehicle-to-everything (V2X) networks, characterized in that, The device includes: The vehicle job acquisition module is used to acquire multiple jobs that are queuing for computation in the edge computing node; the job is a task generated by the vehicle, and each job occupies multiple resource fragments of the edge computing node; The job pairing determination module is used to determine all candidate job pairs from the plurality of jobs; wherein any job pair includes a first job and a second job, the queuing time of the first job is shorter than the queuing time of the second job, and the queuing time interval between the first job and the second job is greater than a preset minimum time slice length. A utility function determination module is used to determine a first utility function for a first job in each job pair, and a second utility function for a second job. The first utility function includes the difference between a first benefit and a first cost, where the first benefit is the QoE value obtained by the first job when the latest available resource fragment of the first job is exchanged with the earliest available resource fragment of the second job, and the first cost is the compensation paid by the first job to the second job when the latest available resource fragment of the first job is exchanged with the earliest available resource fragment of the second job. The second utility function includes the difference between a second benefit and a second cost, where the second benefit is the compensation paid by the first job to the second job when the latest available resource fragment of the first job is exchanged with the earliest available resource fragment of the second job, and the second cost is the reduced QoE value of the second job when the latest available resource fragment of the first job is exchanged with the earliest available resource fragment of the second job. The QoE value is associated with the soft deadline of either the first or second job. The objective function determination module is used to determine the objective function and the corresponding decision items. The objective function includes a first sub-function and a second sub-function. The first sub-function maximizes the sum of the first utility functions corresponding to the multiple job pairs, and the second sub-function maximizes the sum of the second utility functions corresponding to the multiple job pairs. The decision items include all candidate multiple job pairs, the compensation unit price between any of the job pairs, and the resource exchange volume between any of the job pairs. The compensation is associated with the product of the compensation unit price and the resource exchange volume. The objective function solving module is used to solve the objective function to obtain the optimal solution of the decision item; The resource fragment interaction module is used to complete the exchange of corresponding resource fragments based on the optimal solution of the decision item.

9. An electronic device, characterized in that, It includes at least one controller and a memory for communicatively connecting with the controller; the memory stores instructions executable by the at least one controller, which, when executed by the at least one controller, causes the at least one controller to perform a resource exchange method in edge computing for a vehicle network as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions for causing a computer to perform a resource exchange method in edge computing for a vehicle network as described in any one of claims 1 to 7.

Images