A 5G MEC task offloading method based on throughput optimization in Internet of Vehicles

Abstract

The application discloses a 5G MEC task offloading method based on optimal throughput in Internet of Vehicles, solves the problem that the congestion and busy degree of roadside units are not considered in the prior art, the efficiency of the roadside units in processing tasks can be optimized, and high-throughput vehicle services cannot be maximally provided, and comprises the following steps: judging whether a current vehicle loading task can be offloaded according to a local processing delay and a MEC processing delay of the task; identifying an effective roadside unit set according to the load of each roadside unit; updating the priority of the task according to a waiting delay of the task, and calculating the total time required for processing the current task; combining a sampling weight and a transmission rate of the roadside unit to calculate the throughput approval value of the current vehicle loading task in all effective roadside units; and selecting the roadside unit with the maximum throughput approval value to offload. The method can ensure that the vehicle is always in a good offloading quality environment, maximally ensures the stability of the offloading link quality, and improves the success probability of task offloading.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of 5G communication, and particularly relates to a 5G MEC task offloading method based on optimal throughput in vehicle networking. BACKGROUND

[0002] The biggest feature of the 5G network architecture is decentralization, and MEC (Mobile Edge Computation) is the key to realizing the decentralization of the 5G network. The definition of MEC by ETSI (European Telecommunication Standards Institute) is to provide an IT service environment and cloud computing capability at the edge of the mobile network, that is, the wireless access and close to the user side, which is the natural evolution of the mobile base station and the integration of IT and CT technologies. MEC applications have many typical scenarios, and vehicle networking is one of them. In vehicle networking, the randomness of the moving direction and speed of vehicles makes dynamic resource scheduling very complex and challenging, so in vehicle networking, the formulation of offloading and caching decisions and the optimal allocation of computing and caching resources are crucial. At present, many studies have focused on task offloading strategies in vehicle networking.

[0003] CN117915402A is a 5G MEC task offloading method based on hierarchical delay in vehicle networking. According to the total delay of each link, the link with the lightest load is selected for offloading according to the hierarchical delay coefficient. In the 5G vehicle networking system, the heavy traffic burden caused by big data is reduced, the transmission quality of the content is improved, and the content is closer to the user, thereby significantly reducing the computing delay and energy consumption of the task. However, this method and most traditional offloading methods do not consider the congestion and busy degree of roadside units, so the efficiency of task processing by roadside units cannot be optimized, and high-throughput vehicle services cannot be maximized. SUMMARY

[0004] The present application aims to solve the problem that the congestion and busy degree of roadside units are not considered in the prior art, the efficiency of task processing by roadside units cannot be optimized, and high-throughput vehicle services cannot be maximized. A 5G MEC task offloading method based on optimal throughput in vehicle networking is provided, which can ensure that vehicles are always in a good offloading quality environment, maximize the stability of offloading link quality, improve the success probability of task offloading, and maximize the efficiency and benefit of vehicle networking task offloading while reducing the energy consumption of the 5G network.

[0005] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0006] A 5G MEC task offloading method based on throughput optimization in Internet of Vehicles, comprising the following steps:

[0007] S1: judging whether the current task loaded by the vehicle can be offloaded according to the local processing delay of the current task loaded by the vehicle and the MEC processing delay;

[0008] S2: identifying an effective roadside unit set according to the load state of the roadside unit;

[0009] S3: updating the priority of each task according to the waiting delay of the task and the corresponding upper limit value of the delay; the task here includes the current task loaded by the vehicle and the task in the task list in the roadside unit;

[0010] S4: calculating the throughput approval value of the current task loaded by the vehicle in all roadside units according to the sampling weight and the transmission rate of the roadside unit;

[0011] S5: selecting the link with the maximum throughput capacity to perform MEC task offloading.

[0012] The present application can reduce the heavy traffic burden caused by big data in the 5G Internet of Vehicles system, improve the transmission quality of the content, make the content closer to the user, thereby significantly reducing the computing delay and energy consumption of the task, and improving the income of the Internet of Vehicles system and the use perception of the Internet of Vehicles user.

[0013] As a preferred, the S3 comprises: calculating the mathematical expectation value of all the current waiting delays of the tasks in the task list currently running on each roadside unit in the effective roadside unit set and the corresponding upper limit value of the task waiting delay; updating the priority of the tasks in the task list currently running on the roadside unit and the priority of the current task loaded by the vehicle according to the relationship between the current waiting delay of the task and the upper limit value of the task waiting delay, and the roadside unit processes each task according to the updated priority; the roadside unit updates the waiting delay of the current task loaded by the vehicle and all other unfinished tasks after processing each task.

[0014] For the current task loaded by the vehicle, when updating the priority, the upper limit value of the task waiting delay of the roadside unit to which the current task loaded by the vehicle is offloaded is used for updating.

[0015] As a preferred, the S4 comprises: calculating the total time required for the roadside unit to process the current task loaded by the vehicle and the total size of the processed tasks, calculating the MEC offloading throughput of the current offloaded task of the vehicle in the roadside unit; calculating the transmission rate of the current task loaded by the vehicle offloaded to the roadside unit according to the channel gain and bandwidth of the roadside unit combined with the background noise power density; calculating the throughput approval value of the current offloaded task of the vehicle in the roadside unit according to the MEC offloading throughput, the transmission rate and the sampling weight.

[0016] As preferred, the updated priority is: selecting the minimum value between the current waiting time delay of the task and the upper limit value of the task waiting time delay, obtaining the ratio of the minimum value to the upper limit value of the task waiting time delay, calculating the absolute value of the difference between the ratio and 1, taking the inverse of the absolute value as the index of the exponential function with natural index, and the value of the exponential function plus the current priority of the task is the updated priority; for the vehicle current loaded task, the first time it is unloaded to each roadside unit, its current waiting time delay is 0.

[0017] As preferred, the S1 comprises: calculating the local processing time delay of the vehicle current loaded task according to the calculation complexity and the vehicle terminal calculation capability; calculating the MEC processing time delay of the vehicle current loaded task according to the calculation complexity and the MEC server calculation capability; if the local processing time delay of the vehicle current loaded task is less than or equal to the MEC processing time delay of the task, the vehicle current loaded task is kept for local processing, otherwise the vehicle current loaded task is processed for MEC unloading.

[0018] As preferred, according to the sum of the ratio of the current waiting time delay of each task to the total number of tasks in the list of tasks currently running on the roadside unit, the mathematical expectation value of the current waiting time delay is obtained; the upper limit value of the task waiting time delay is obtained according to the product of the set time delay coefficient and the task expectation value.

[0019] As preferred, the updated waiting time delay is obtained according to the current waiting time delay of the task and the MEC processing time delay of the task in the roadside processing unit which has just been processed due to the highest priority.

[0020] As preferred, the S2 comprises: setting a roadside unit load scheduling threshold, for all roadside units, selecting roadside units that meet the roadside unit load less than or equal to the roadside unit load scheduling threshold, and including them in the effective roadside unit set.

[0021] As preferred, the total time comprises the time for the roadside unit to process the vehicle current loaded task and the time for the roadside unit to process the tasks in the task list whose priority is higher than that of the vehicle current loaded task, and the total task size comprises the size of the vehicle current loaded task and the size of the tasks in the task list whose priority is higher than that of the vehicle current loaded task.

[0022] S3 is repeated until the vehicle current loaded task is processed by the roadside units in the effective roadside unit set, the total time required for the roadside unit to process the vehicle current loaded task and the total task size processed are obtained; the MEC unloading throughput of the vehicle current unloaded task in the roadside unit is obtained according to the ratio of the total task size to the total time.

[0023] As preferred, the S2 further comprises: calculating the MEC processing delay of all tasks in the list of tasks currently running on each road side unit in the effective road side unit set.

[0024] Therefore, the present application has the following beneficial effects: the local delay required for the vehicle operation task to be locally calculated is calculated to determine whether the current vehicle-mounted task can be offloaded; the effective road side unit set is identified according to the load of each road side unit; the MEC processing delay of all tasks currently processed in the effective road side unit set is calculated; the processing priority of each task is updated according to the waiting delay of the task and the upper limit value of the delay; the total time spent in completing the processing of the current vehicle-mounted task is counted; the throughput approval value of the current task in all effective road side units is calculated in combination with the sampling weight and the transmission rate of the road side unit; and the road side unit with the maximum throughput approval value is selected for offloading, which can reduce the heavy traffic burden caused by big data in the 5G vehicle networking system, improve the transmission quality of the content, make the content closer to the user, and thus significantly reduce the calculation delay and energy consumption of the task, and improve the benefits of the vehicle networking system and the use perception of the vehicle networking user. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 The overall step flowchart of the 5G MEC task offloading method based on the optimal throughput in the vehicle networking in the present application.

[0026] Figure 2 The road side unit distance and the offloadable real-time task comparison graph in the present application.

[0027] Figure 3 The task offloading throughput comparison graph of the method of the present application and other algorithms. DETAILED DESCRIPTION

[0028] The present application will be further described in detail below in combination with the drawings and specific embodiments.

[0029] Embodiment:

[0030] The present embodiment provides a 5G MEC task offloading method based on the optimal throughput in the vehicle networking, as shown in Figure 1 the operation flowchart is as follows: step one, judging whether the vehicle current loaded task can be offloaded according to the local processing delay and the MEC processing delay of the vehicle current loaded task; step two, identifying the effective road side unit set according to the load state of the road side unit; step three, updating the priority of each task according to the waiting delay of the task and the corresponding upper limit value of the delay; step four, calculating the throughput approval value of the vehicle current loaded task in all road side units according to the sampling weight and the transmission rate of the road side unit; and step five, selecting the link with the maximum throughput capacity for MEC task offloading.

[0031] In this embodiment, if no special description is made, only mentioning the task means the task currently loaded by the vehicle and the task in the task list in the roadside unit, that is, the task in the roadside unit.

[0032] The 5G MEC task offloading method based on optimal throughput in the vehicle networking provided by the embodiment starts from analyzing the local delay of the vehicle terminal, determines whether the current vehicle task can be offloaded, identifies the effective roadside unit set according to the load of each roadside unit, calculates the corresponding MEC processing delay for all tasks currently processed in the effective roadside unit set, calculates the upper limit value of the delay according to the waiting delay of the task, and then updates the processing priority of each task, and after completing the processing of the current vehicle task, the total time spent is counted, the throughput approval value of the current task in all roadside units is calculated in combination with the sampling weight and the transmission rate of the roadside unit, and then the roadside unit with the maximum throughput approval value is selected for offloading, so that the efficiency and benefit of the vehicle networking task offloading can be maximized while reducing the energy consumption of the 5G network.

[0033] The technical solutions and technical effects of the present application will be further described below through specific examples and specific application scenarios. The following examples are an explanation of the present application, and the present application is not limited to the following examples.

[0034] Specifically, a 5G MEC task offloading method based on optimal throughput in vehicle networking, wherein MEC includes a passing vehicle Vhc tar , the vehicle generates a vehicle task, the current vehicle task of the vehicle is Tk N , a roadside unit RSU installed in the driving area of the passing vehicle, the roadside unit includes Rsu1, Rsu2,..., Rsu m , wherein m represents the total number of roadside units. The roadside unit includes a corresponding load Rld1, Rld2,..., Rld m Each load corresponds to a roadside unit, and the task list TKL i currently running on any roadside unit includes Tk i,1 , Tk i,2 ,..., Tk i,ix , Tk i,ix represents the jth task in the task list currently running on the ith roadside unit. The size of task Tk i,j (j∈[1,i x ]) is Tkm i,j , the unit is kB, the calculation complexity (CPU consumption) is Tcp i,j (Cycles), the current waiting delay is Dlw i,j , the unit is ms, and the priority is Pri i,j∈[1,10], the larger the number, the higher the priority.

[0035] As shown in the following steps: Figure 1

[0036] First step: according to the local processing delay of the current vehicle task and the MEC processing delay, judge whether the current vehicle task can be unloaded.

[0037] (1-1) First, calculate the local processing delay of the current vehicle task, which is the ratio of the calculation complexity TCP (i.e. CPU consumption) to the computing power of the vehicle terminal, multiplied by the power function power(10,3), power(10,3) represents the value of the power function with base 10 and exponent 3.

[0038] (1-2) The MEC processing delay of the current vehicle task is the ratio of the calculation complexity TCP to the MEC computing power, multiplied by the power function power(10,3), power(10,3) represents the value of the power function with base 10 and exponent 3.

[0039] (1-3) According to the above calculation results, if the local processing delay of the current vehicle task is less than or equal to the MEC processing delay, the current vehicle task is kept in the local for processing, otherwise the current vehicle task is considered for MEC unloading processing.

[0040] Second step: according to the load state of roadside unit, identify the effective roadside unit set.

[0041] Set the roadside unit load scheduling threshold, select the roadside unit whose load is less than or equal to the roadside unit load scheduling threshold from all roadside units, and include it in the effective roadside unit set.

[0042] Calculate the MEC processing delay of all tasks in the task list in each roadside unit in the effective roadside unit set, which is calculated in the same way as the MEC processing delay in the first step.

[0043] Third step: update the priority of each task according to the waiting delay of the task and the corresponding delay upper limit value.

[0044] (3-1) Set the delay coefficient Pam dy , in this embodiment, the delay coefficient is greater than or equal to 1 and less than or equal to 2.

[0045] ​The mathematical expectation of the current waiting time delay of all tasks in the task list of each roadside unit in the effective roadside unit set is calculated. For a certain roadside unit, the mathematical expectation of the current waiting time delay is the sum of the current waiting time delay of each task in the task list currently running on the roadside unit and the total number of tasks in the task list currently running on the roadside unit.

[0046] The upper limit value of the task waiting time delay in the roadside unit is calculated. The upper limit value of the task waiting time delay is the product of the set time delay coefficient and the task expectation value.

[0047] (3-2) The priority of all tasks in the effective roadside unit set is updated. For example, for the jth task Tk in the task list in the ith roadside unit, the updated priority is the minimum value of the current waiting time delay and the upper limit value of the task waiting time delay, the ratio of the minimum value to the upper limit value of the task waiting time delay, the inverse of the absolute value of 1 minus the ratio, the inverse being used as the index of an exponential function with the natural constant e as the base, and the value of the exponential function being added to the current priority of the task. i,j

[0048] The priority of the vehicle's current loaded task is updated. For the vehicle's current loaded task, the waiting time delay is 0 ms when it is first pre-unloaded to each roadside unit. The updated priority is 1 minus the absolute value of the ratio of the minimum value of the current waiting time delay of the vehicle's current loaded task to the upper limit value of the task waiting time delay, multiplied by -1, and the value obtained is used as the index of an exponential function with the natural constant e as the base, and the value of the exponential function is added to the current priority of the vehicle's current loaded task.

[0049] Each roadside unit processes the vehicle-mounted task and the tasks in the task list in the roadside unit according to the updated priority, with the higher the priority, the sooner the processing.

[0050] (3-3) After each roadside unit finishes processing a task, the waiting time delay of the vehicle's current loaded task and all other unfinished tasks is updated. The updated waiting time delay is the sum of the current waiting time delay of the task and the MEC processing time delay of the task that has just been processed in the roadside processing unit due to the highest priority.

[0051] For example, for the task in the roadside unit, the updated waiting time delay is the sum of the current waiting time delay of the task in the roadside unit and the MEC processing time delay of the task that has just been processed in the roadside processing unit due to the highest priority.

[0052] ​The updated waiting time of the current loaded task of the vehicle is the sum of the current waiting time of the current loaded task of the vehicle and the MEC processing time of the task with the highest priority that has just been processed in the roadside processing unit. If there are multiple tasks that have just been processed, the MEC processing time is the sum of the MEC processing time of each task.

[0053] The fourth step is to calculate the throughput approval value of the current loaded task of the vehicle in all roadside units according to the sampling weight and the transmission rate of the roadside unit.

[0054] (4-1) Determine whether the current roadside unit number is greater than the total number of roadside units. If not, repeat the third step to traverse all roadside units in the effective roadside unit set. If yes, it means that the current loaded task of the vehicle has been processed by all roadside units in the effective roadside unit set, and the subsequent steps are executed.

[0055] For each roadside unit in the effective roadside unit set, calculate the total time taken to process the current loaded task of the vehicle. The total time is the sum of the MEC processing time of the current loaded task of the vehicle and the MEC processing time of the roadside unit task in the task list with a higher processing priority than the current loaded task of the vehicle.

[0056] (4-2) Set the bandwidth, channel gain, and background noise power density of the roadside unit. For each roadside unit in the effective roadside unit set, calculate its signal-to-noise ratio (SNR), which is the difference between the channel gain and the background noise power density.

[0057] Calculate the transmission rate of the current loaded task of the vehicle pre-unloaded to each roadside unit. The transmission rate is the product of the value of the logarithm of the sum of the SNR and 1 with base 2, i.e., log2(1+SNi), and the bandwidth of the roadside unit.

[0058] The calculation method of the transmission rate of the current loaded task of the vehicle pre-unloaded to each roadside unit is the same.

[0059] (4-3) Calculate the MEC offload throughput of the current loaded task of the vehicle in the roadside unit according to the total time taken and the total task size.

[0060] For each roadside unit, the calculation method of the MEC offload throughput is the same, which is the ratio of the size of the current loaded task of the vehicle and the sum of the sizes of the roadside unit tasks in the task list with a higher processing priority than the current loaded task of the vehicle and the time taken by the roadside unit to process the current loaded task of the vehicle.

[0061] (4-4) Set the sampling value, in this embodiment, the sampling value is greater than or equal to 1 and less than or equal to 10, and calculate the sampling weight. The sampling weight is the reciprocal of the value of the power function with 2 as the base number and the sampling value as the exponent.

[0062] According to the sampling weight and the MEC throughput offload amount and the transmission rate, the vehicle current load task pre-offload in the effective roadside unit set is calculated. The calculation method of the vehicle current load task pre-offload in all roadside units is the same, which is the product of the sampling weight and the transmission rate of the vehicle current offload task offloaded to the roadside unit, plus the MEC offload throughput, minus the product of the sampling weight and the MEC offload throughput.

[0063] Step 5: Select the link with the maximum throughput capacity for MEC task offload.

[0064] Select the roadside unit with the maximum throughput approval value in the effective roadside unit set as the target offload link of the vehicle current offload task.

[0065] The 5G MEC task offload method based on throughput optimization in the vehicle networking provided in this embodiment can reduce the heavy traffic burden caused by big data in the 5G vehicle networking system, improve the transmission quality of content, make the content closer to the user, and thus significantly reduce the computing delay and energy consumption of the task, and improve the revenue of the vehicle networking system and the use perception of the vehicle networking user. The local computing delay can be calculated according to the complexity of the task; the effective roadside unit set can be identified according to the load state of the roadside unit; the priority of the task can be updated and dynamic processing can be implemented according to the waiting delay of the task; the total time required for processing the current task can be calculated; the throughput approval value of the current task can be calculated according to the sampling weight; the link with the maximum throughput capacity can be selected for task offload; and the vehicle can be ensured to be in a good offload quality environment, thereby providing real-time protection for improving the success rate of task offload.

[0066] The scheme of this embodiment will be described below with specific examples. It is assumed that the number of roadside units set in the vehicle form area is equal to 3, and the task status of the vehicle is as shown in Table 1:

[0067] Table 1 Vehicle running task distribution

[0068]

[0069] The roadside unit distribution is as shown in Table 2:

[0070] Table 2 Roadside unit distribution

[0071]

[0072] The basic data is as shown in Table 3:

[0073] Table 3 Basic Data

[0074]

[0075] This embodiment provides a 5G MEC task offloading method based on throughput optimization in vehicle-to-everything (V2X) communication, which includes the following steps: task pre-offloading decision, effective roadside unit identification, roadside unit task preprocessing, and vehicle-mounted task offloading calculation.

[0076] Specifically:

[0077] Step 1: Task pre-unloading decision: Based on the local processing latency and MEC processing latency of the currently loaded task in the vehicle, determine whether the currently loaded task in the vehicle can be unloaded.

[0078] Step 1-1: Calculate the vehicle's Vhc tar Currently loading task Tk N Local processing latency T_Local N .

[0079] Local processing latency is the computational complexity of TCP and the computing power of the vehicle terminal (Ac). vhc The ratio is then multiplied by the power function power(10,3), and the result is 28.33ms.

[0080] Step 1-2: Calculate the MEC processing latency T_Edge for the vehicle's current loaded task. N .

[0081] MEC processing latency is the ratio of computational complexity (TCP) to MEC computational capability. MEC Then multiply by the power function power(10,3), which represents the value of a power function with base 10 and exponent 3. Substituting the specific value above, we get 4.25ms.

[0082] Steps 1-3: Based on the above calculation results, the local processing latency of the vehicle's current loading task is greater than the MEC processing latency. Therefore, we consider performing MEC unloading processing on the vehicle's current loading task.

[0083] Step 2: Identify valid roadside units: Identify the set of valid roadside units based on their load status.

[0084] Step 2-1: Set the roadside unit load scheduling threshold Rld th The percentage is 60%. Roadside units that meet the roadside unit load scheduling threshold (less than or equal to the roadside unit load) will be included in the effective roadside unit set SetR. According to the above data, the effective roadside unit set includes roadside unit Rsu1 and roadside unit Rsu3.

[0085] Step 2-2: Calculate the MEC processing latency for all tasks in the effective roadside unit task list.

[0086] All tasks in the effective roadside unit centralized task list include {task 1, task 2, task 4} in roadside unit Rsu1 and {task 5, task 6} in roadside unit Rsu3. Calculate their MEC processing latency and obtain the results {{4.25, 2.5, 1.5}, {1.88, 7.8}} ms.

[0087] Step 3: Roadside unit task preprocessing: Update the priority of the task according to the waiting delay and implement dynamic processing.

[0088] Step 3-1: Calculate the expected value of the current waiting time of all tasks in each roadside cell of the effective roadside cell set and the corresponding upper limit value of the waiting time DLW. i .

[0089] The expected value of the task is the roadside unit Rsu. i The list of currently running tasks (TKL) i The sum of the ratios of the current waiting time of each task to the total number of tasks in the currently running task list on the roadside unit is used to substitute the data in the table above, resulting in a mathematical expectation value of {3, 3.5} ms. For example, if there are three tasks currently running on roadside unit Rsu1, i.e., a total of 3 tasks, according to the data in the table above, the ratio of the current waiting time of task 1 to the total number of tasks is two-thirds, the ratio of the current waiting time of task 2 to the total number of tasks is 1, and the ratio of the current waiting time of task 4 to the total number of tasks is four-thirds. The sum of these three is 3, so the expected value of the task for roadside unit Rsu1 is 3ms. Similarly, there are two tasks currently running on the roadside unit Rsu3, so the total number of tasks is 2. The ratio of the current waiting time of task 5 to the total number of tasks is 0.5, and the ratio of the current waiting time of task 6 to the total number of tasks is 3. The sum of the two is 3.5, so the expected task time of roadside unit Rsu3 is 3.5ms.

[0090] The upper limit of task waiting delay is the product of the set delay coefficient and the expected value of the task. Substituting the data in the table above, we get the upper limit of task waiting delay as {4.5, 5.25} ms.

[0091] Step 3-2: Update the priority of all tasks in the valid roadside cell set.

[0092] For any roadside unit, the currently running task Tk in the task list i,j (j∈[1, i) x}, where i represents the i-th road-side unit, j represents the j-th task in the task list currently running on the road-side unit, and the updated priority Pri i,j is calculated as 1 minus the absolute value of the ratio of the minimum value of the current waiting delay and the upper limit value of the task waiting delay to the upper limit value of the task waiting delay, multiplied by -1, and the value obtained is taken as the index of an exponential function with the natural constant e as the base, and the value of the exponential function is added to the current priority of the task. Substituting the data in the above table, the updated priority Pri i,j is {{2.57, 7.72, 3.89}, {5.45, 5}}.

[0093] The specific calculation process is as follows: for the road-side unit Rsu1, for task 1, according to the calculation result, the minimum value of the current waiting delay and the upper limit value of the task waiting delay is 2, then the index of the exponential function is about -0.56, and the value of the exponential function is 0.57, and the current priority of task 1 is 2, so the updated priority of task 1 is 2.57. Similarly, for task 2, according to the calculation result, the minimum value of the current waiting delay and the upper limit value of the task waiting delay is 3, then the index of the exponential function is about -0.33, and the value of the exponential function is about 0.72, and the current priority of task 2 is 7, so the updated priority is 7.72. For task 4, according to the calculation result, the minimum value of the current waiting delay and the upper limit value of the task waiting delay is 4, then the index of the exponential function is about -0.11, and the value of the exponential function is 0.89, and the current priority of task 4 is 3, so the updated priority of task 4 is 3.89.

[0094] For the road-side unit Rsu3, for task 5, according to the calculation result, the minimum value of the current waiting delay and the upper limit value of the task waiting delay is 1, then the index of the exponential function is about -0.81, and the value of the exponential function is 0.45, and the current priority of task 5 is 5, so the updated priority of task 5 is 5.45. For task 6, according to the calculation result, the minimum value of the current waiting delay and the upper limit value of the task waiting delay is 5.25, then the index of the exponential function is about 0, and the value of the exponential function is 1, and the current priority of task 6 is 4, so the updated priority of task 6 is 5.

[0095] For the tasks currently loaded by the vehicle, when they are unloaded to each road-side unit for the first time, their waiting delay Dlw i,N is 0 ms, and their priority Pri i,N is updated as 1 minus the absolute value of the ratio of the current waiting delay Dlw i,Nthe absolute value of the ratio of the minimum value in the task waiting time upper limit value and the task waiting time upper limit value, the absolute value multiplied by negative 1, as the index of the exponential function with the natural constant e as the base, the value of the exponential function plus the priority of the current vehicle current task, the updated priority Pri i,N is {6.37, 6.37}. Specifically, when the vehicle current task is pre-unloaded to the roadside unit Rsu1, the vehicle current task has been waiting for a time delay Dlw i,N the minimum value in the task waiting time upper limit value is 0, the index of the exponential function is -1, the value of the exponential function is about 0.37, the priority of the vehicle current task is 6, and when the vehicle current task is pre-unloaded to the roadside unit Rsu1, the updated priority is 6.37. Similarly, when the vehicle current task is pre-unloaded to the roadside unit Rsu3, the updated priority is 6.37.

[0096] Roadside unit Rsu i According to the updated priority, the processing order of the tasks in the roadside unit Rsu1 is {task 2, vehicle current task, task 4, task 1}, and the processing order of the tasks in the roadside unit Rsu3 is {vehicle current task, task 5, task 6}.

[0097] Step 3-3: Calculate the waiting time of the task according to the updated priority.

[0098] According to the foregoing calculation results, the roadside unit Rsu1 first processes task 2, and the waiting time of the other tasks {vehicle current task, task 4, task 1} increases by T_Edge 1,2 is 2.5 ms (the MEC processing time of task 2 calculated in step 2-2); the roadside unit Rsu3 first processes the vehicle current task, and the waiting time of the other tasks {task 5, task 6} increases by 4.25 ms (the MEC processing time of the vehicle current task obtained in step 1-2); the waiting time of the vehicle current task and all other unfinished tasks is updated, and the updated waiting time is the sum of the current waiting time and the time delay (i.e., the increased waiting time) of the task that has just been processed due to the highest priority in the roadside unit.

[0099] Then, the waiting time of {vehicle current loaded task, task 4, task 1} is increased to {2.5, 6.5, 4.5} ms respectively, and the waiting time of {task 5, task 6} is {5.25, 10.25} ms respectively.

[0100] Step 4: Vehicle task offloading calculation.

[0101] Step 4-1: Repeat step 3 until the vehicle current loaded task is processed by the road side unit in the effective road side unit set, and calculate the total time required to process the vehicle current loaded task.

[0102] The road side unit Rsu3 is processed due to the priority of the vehicle current loaded task, and the time T_MEC spent to process the vehicle current loaded task is calculated i,N , wherein the time spent is the sum of the MEC processing time delay of the vehicle current loaded task and the MEC processing time delay of the task with higher processing priority than the vehicle current loaded task in the task list of the road side unit Rsu3. Substituting the data in the above table, the MEC processing time delay of the vehicle current loaded task is 4.25 ms, the road side unit Rsu3 processes the vehicle current loaded task first, i.e. there is no task with higher processing priority than the vehicle current loaded task in the task list of the road side unit Rsu3, so the time T_MEC spent is 4.25 ms. i,N .

[0103] After the road side unit Rsu1 processes task 2, it continues to update the priority of the task {vehicle current loaded task, task 4, task 1}, and the updated priority is {6.9, 4.86, 3.29}. The next step is to process the vehicle current loaded task directly by the road side unit Rsu1, and calculate the time T_MEC spent to process the vehicle current loaded task i,N , wherein the time spent is the sum of the MEC processing time delay of the vehicle current loaded task and the MEC processing time delay of the task with higher processing priority than the vehicle current loaded task in the task list of the road side unit Rsu3. Substituting the calculation result, the MEC processing time delay of the vehicle current loaded task is 4.25 ms, and the road side unit Rsu1 processes the task with higher processing priority than the vehicle current loaded task in the task list, which is only task 2, and its MEC processing time delay is 2.5 ms, so the time spent is 6.75 ms.

[0104] Step 4-2: Calculate the signal-to-noise ratio of the road side unit in the effective road side unit set, which is the difference between the road side unit channel gain and the background noise power density. Substituting the data in the above table, the signal-to-noise ratio of the road side unit in the effective road side unit set is {60, 70} dB / Hz.

[0105] Calculate the vehicle current loaded task pre-offloading to the road side unit Rsui transmission rate is the logarithm of the sum of the signal-to-noise ratio plus 1 with base 2, i.e. log2(1+SNi), and the value of the logarithm is multiplied by the bandwidth of the road side unit Rsu i . Substituting the above table data, the transmission rate is {593.07, 614.97} Mbps. Taking the road side unit Rsu1 as an example, the signal-to-noise ratio is 60, the bandwidth is 100, and the value of the logarithm of 61 with base 2 is about 5.930737, and the product of this value and the bandwidth is 593.07, i.e. the transmission rate of the road side unit Rsu1 is 593.07 Mbps. Similarly, the signal-to-noise ratio of the road side unit Rsu3 is 70, the bandwidth is 100, the value of the logarithm of 71 with base 2 is about 6.1497, and the product of this value and the bandwidth is 614.97, i.e. the transmission rate of the road side unit Rsu3 is 614.97 Mbps.

[0106] Step 4-3: Calculate the MEC offloading throughput THr i of the vehicle's current loaded task in the road side unit Rsu i,N .

[0107] The MEC offloading throughput is the ratio of the size of the vehicle's current loaded task to the sum of the sizes of the tasks in the task list of the road side unit Rsu i with a higher processing priority than the vehicle's current loaded task, and the time taken by the road side unit Rsu i to process the vehicle's current loaded task. Taking the road side unit Rsu1 as an example, the size of the vehicle's current loaded task is 1200 kb, and the road side unit Rsu1 has a task with a higher processing priority than the vehicle's current loaded task, which is task 2, and the size of task 2 is 1800 kb, the sum of the two is 3000 kb, and the time taken by the road side unit Rsu1 to process the vehicle's current loaded task is 6.75 ms, therefore, the MEC offloading throughput of the road side unit Rsu1 is 444.44 Mbps.

[0108] Therefore, substituting the above table data, the MEC offloading throughput is {444.44 (road side unit Rsu1), 282.35 (road side unit Rsu3)} Mbps.

[0109] Step 4-4: Calculate the throughput approval value.

[0110] Calculate the sampling weight Q sm , which is the inverse of the value of the power function with base 2 and the fourth of the sampling value as the exponent. Substituting the above table data, the sampling weight is 0.59.

[0111] Calculate the throughput approval value THO of the vehicle's current offloaded task in the road side unit Rsu i in the effective road side unit seti,N , the throughput approved value is the product of the sampling weight value and the transmission rate of the vehicle current unloading task unloading to the roadside unit, plus the MEC unloading throughput, minus the product of the sampling weight value and the MEC unloading throughput, and the throughput approved value is obtained by substituting the above table data, which is {532.82, 480.13} Mbps. Taking the roadside unit Rsu1 as an example, the sampling weight value is 0.59, the transmission rate is 593.07 Mbps, the MEC unloading throughput is 444.44 Mbps, and finally the obtained throughput approved value is 532.82 Mbps.

[0112] Step 4-5: Perform task offloading.

[0113] Take the roadside unit Rsu t=1 with the maximum throughput approved value, i.e. the throughput approved value is 532.82 Mbps, as the target unloading link of the current unloading task of the vehicle.

[0114] Simulation experiment:

[0115] The performance of the ODT-MOM task offloading method of the embodiment is compared and simulated on the MATLAB platform, and is compared with the CTD algorithm in the 5G MEC task offloading method based on hierarchical delay in CN117915402A in the background art. The basic data information is shown in Table 3 above, and the obtained results are shown in Figure 2 and Figure 3 .

[0116] As shown in Figure 2 , the relationship between the same vehicle at different RSU distances and the unloadable real-time tasks. Overall, the more vehicle-mounted real-time tasks, the more tasks can be unloaded according to the method provided in the embodiment; the farther the distance between the vehicle and the roadside unit, the worse the cellular link quality, the more serious the channel attenuation, and the lower the transmission rate, and the fewer tasks can be unloaded.

[0117] As can be seen by comparison, under the premise of the same vehicle-mounted task, since ODT aims to pursue higher throughput and is more suitable for eMBB scenarios, it will be slower than the CTD algorithm in response to delay, but the number of tasks that can be provided for offloading by the two algorithms at different distances is not much different.

[0118] As shown in Figure 3 , the task offloading throughput of the algorithm ODT provided in the embodiment is compared with the CTD algorithm. Whether it is long distance or short distance, ODT is obviously superior to the CTD algorithm; especially in close proximity, due to good wireless environment and high channel gain, the throughput of the two algorithms differs more greatly, and for vehicles with download, on-demand or video monitoring backhaul services, a more suitable ODT-MOM algorithm can be selected.

[0119] The above-described embodiments are only the preferred ones of the present application, and do not limit the present application in any form, and other variants and modifications can be made without departing from the technical solutions recited in the claims.

Claims

1.A method for 5G MEC task offloading based on throughput optimization in Internet of Vehicles, characterized in that, The method comprises the following steps: S1: judging whether the current vehicle task can be offloaded according to the local processing delay and the MEC processing delay of the current vehicle task; S2: identifying an effective roadside unit set according to the load state of the roadside unit; S3: updating the priority of each task according to the waiting delay and the corresponding upper limit value of the delay of the task; S4: calculating the throughput approval value of the current vehicle task in all roadside units according to the sampling weight and the transmission rate of the roadside unit; S5: selecting the link with the maximum throughput capacity for MEC task offloading. 2.The 5G MEC task offloading method based on optimal throughput in a vehicle Internet of Things according to claim 1, wherein The S3 comprises the following steps: calculating the mathematical expectation value of the current waiting delay of all tasks in the task list currently running on each roadside unit in the effective roadside unit set and the corresponding upper limit value of the task waiting delay; updating the priority of the task according to the relationship between the current waiting delay of the task and the corresponding upper limit value of the task waiting delay in the roadside unit; the roadside unit processes each task according to the updated priority; and updating the waiting delay of the current vehicle task and all other unfinished tasks after the roadside unit finishes processing each task. 3.The 5G MEC task offloading method based on throughput optimization in a vehicle Internet of Things according to claim 1, characterized in that, The S4 comprises the following steps: calculating the MEC offloading throughput of the current vehicle task in the roadside unit according to the total time required by the roadside unit for processing the current vehicle task and the total size of the processed tasks; calculating the transmission rate of the current vehicle task offloaded to the roadside unit according to the channel gain and bandwidth of the roadside unit and the background noise power density; and calculating the throughput approval value of the current vehicle task in the roadside unit according to the MEC offloading throughput, the transmission rate and the sampling weight. 4.The 5G MEC task offloading method based on throughput optimization in a vehicle Internet of Things according to claim 2, characterized in that, The updated priority is obtained by selecting the minimum value of the current waiting delay of the task and the upper limit value of the task waiting delay, obtaining the reciprocal of the difference between the minimum value and the upper limit value of the task waiting delay, calculating the absolute value of the difference between the reciprocal and 1, and taking the reciprocal as the index of an exponential function with a natural index as the base, and the value of the exponential function plus the current priority of the task is the updated priority; for the current vehicle task, the current waiting delay is 0 when the task is offloaded to each roadside unit for the first time. 5.The 5G MEC task offloading method based on optimal throughput in a vehicle Internet according to claim 1 or 2 or 3 or 4, characterized in that, The S1 comprises the following steps: calculating the local processing delay of the current vehicle task according to the calculation complexity and the computing capacity of the vehicle terminal; calculating the MEC processing delay of the current vehicle task according to the calculation complexity and the computing capacity of the MEC server; if the local processing delay of the current vehicle task is less than or equal to the MEC processing delay of the task, the current vehicle task is kept for local processing, otherwise the current vehicle task is processed by MEC offloading. 6.The 5G MEC task offloading method based on optimal throughput in a vehicle Internet according to claim 2 or 4, characterized in that, The mathematical expectation value of the current waiting delay is obtained according to the sum of the ratio of the current waiting delay of each task to the total number of tasks in the task list currently running on the roadside unit; and the upper limit value of the task waiting delay is obtained according to the product of the set delay coefficient and the task expectation value. 7.The 5G MEC task offloading method based on optimal throughput in a vehicle Internet of Things according to claim 2 or 4, characterized in that, The updated waiting delay is obtained according to the current waiting delay of the task and the MEC processing delay of the task that has just been processed in the roadside processing unit due to the highest priority. 8.The 5G MEC task offloading method based on optimal throughput in a vehicle Internet of Things according to claim 1 or 2 or 3 or 4, characterized in that, The S2 comprises: setting a roadside unit load scheduling threshold, selecting, for all roadside units, a roadside unit satisfying roadside unit load less than or equal to the roadside unit load scheduling threshold, and including the roadside unit into an effective roadside unit set. 9.The 5G MEC task offloading method based on throughput optimization in a vehicle Internet of Things according to claim 3, characterized in that, The total time comprises a time for the roadside unit to process a current loading task of the vehicle and a time for processing a roadside unit task in a task list of the roadside unit and having a priority higher than the current loading task of the vehicle, and the total task size comprises a size of the current loading task of the vehicle and a size of the roadside unit task in the task list of the roadside unit and having a priority higher than the current loading task of the vehicle. 10.The 5G MEC task offloading method based on throughput optimization in a vehicle Internet of Things according to claim 9, wherein, The S2 further comprises: calculating a MEC processing delay of all tasks in a task list currently running on each roadside unit in the effective roadside unit set.

Images