Mobile edge computing task unloading method based on energy efficiency in Internet of Vehicles

Abstract

The invention discloses a mobile edge computing task unloading method based on energy efficiency in the Internet of Vehicles, and solves the problem that tasks cannot be unloaded to different link MEC servers due to the fact that an unloading method and energy consumption are not comprehensively considered in the prior art. The method comprises the following steps: accounting the energy efficiency of each vehicle-mounted task according to local energy consumption required by local operation of the vehicle-mounted tasks, and screening out high-energy-efficiency tasks; screening out a target roadside unit according to the distance between the vehicle and each roadside unit; for each high-energy-efficiency task, calculating total unloading time and total unloading energy consumption of the high-energy-efficiency task in each target roadside unit; carrying out normalization processing on the total time and the total energy consumption in combination with the load of the roadside unit, and carrying out filtering weighting calculation to obtain a judgment evaluation value; and for the high-energy-efficiency task, selecting the target roadside unit corresponding to the minimum judgment evaluation value to implement task unloading. According to the method, the stability of unloading link quality can be ensured to the greatest extent, and the success probability of task unloading is improved.

Description

Technical Field

[0001] The present invention relates to the field of 5G communication technology, and in particular to a mobile edge computing task offloading method based on energy efficiency in a vehicle network. Background Art

[0002] MEC (Mobile Edge Computation) is considered a key factor in the modernization of cellular base station models and the development of 5G technology. MEC introduces computing and storage resources to the edge of mobile networks, reduces computing latency and energy consumption of terminal devices, improves the user experience quality of mobile Internet applications, and reduces the high load of cloud computing centers.

[0003] In the Internet of Vehicles, the randomness of the direction and speed of vehicle movement makes dynamic resource scheduling very complex and challenging. Therefore, in the Internet of Vehicles, 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 the task offloading strategy of the Internet of Vehicles. For example, a complete offloading optimization algorithm and a method that considers joint task offloading and resource optimization in a multi-user system. Both of these methods are typical task offloading algorithms in the Internet of Vehicles. However, the first algorithm mentioned above considers the task offloading itself too much and does not solve the problem of energy consumption optimization. The second algorithm can only be applied to a single MEC server due to the complexity of the operation. Therefore, it is necessary to study how to comprehensively consider the offloading method and energy consumption, and even solve how to offload different tasks to MEC servers on different links, so that the offloading process becomes more stable. Summary of the invention

[0004] The present invention mainly solves the problem that the existing offloading method and energy consumption are not comprehensively considered and different tasks cannot be offloaded to MEC servers on different links. It provides a mobile edge computing task offloading method based on energy efficiency in the Internet of Vehicles.

[0005] The above technical problem of the present invention is mainly solved by the following technical solution: a mobile edge computing task unloading method based on energy efficiency in a vehicle network, comprising a roadside unit arranged in a vehicle driving area, a MEC edge computing server configured on the roadside unit, and the vehicle network including a vehicle-mounted terminal, comprising the following steps: Calculate the energy efficiency of each on-board task based on the local energy consumption required for local computing of the on-board task, and select high-efficiency tasks; Filter out target roadside units according to the distance between the vehicle and each roadside unit; For each high-energy-efficiency task, calculate the total time and total energy consumption for unloading in each target roadside unit; The total time and total energy consumption are normalized with the load of the roadside unit, and the judgment evaluation value is obtained through filtering and weighted calculation; For high energy efficiency tasks, the target roadside unit corresponding to the minimum judgment evaluation value is selected to implement task unloading.

[0006] The present invention measures the local computing time and local energy consumption through the complexity of the on-board tasks, calculates the energy efficiency of different on-board tasks according to the complexity of the on-board tasks and the local energy consumption, and selects the on-board tasks with high energy efficiency for unloading. The target roadside unit set for task unloading is selected according to the distance between the vehicle and the roadside unit, the transmission rate of the unloading link is calculated, and then the total time and total energy consumption of high energy efficiency task unloading are calculated, the data is normalized in combination with the roadside unit load, and the judgment evaluation value required for unloading different on-board tasks is obtained by filtering and weighting. The target roadside unit for unloading the target task is identified according to the minimization principle, which can ensure that the vehicle is always in a good unloading quality environment, and provide real-time guarantee for improving the success rate of task unloading. The present invention can reduce the heavy traffic burden brought by big data in the 5G Internet of Vehicles system, improve the transmission quality of content, and bring content closer to users, thereby significantly reducing the calculation delay and energy consumption of tasks, and improving the benefits of the Internet of Vehicles system and the usage perception of Internet of Vehicles users.

[0007] As a preferred solution, the local energy consumption required for local computing of the vehicle-mounted task is specifically: The computational complexity of the vehicle-mounted tasks is obtained, the time required for local operation of each vehicle-mounted task is calculated according to the computing power of the vehicle-mounted terminal, and the local energy consumption required for local operation of each vehicle-mounted task is calculated according to the computing power consumption of the vehicle-mounted terminal.

[0008] This solution calculates the local energy consumption required for local operation of on-board tasks based on the computational complexity of on-board tasks, the computing power of on-board terminals, and the computing power consumption.

[0009] The on-board task includes the attribute of computational complexity, that is, the CPU consumption of the on-board terminal required to calculate the on-board task. The computing capability and computing power consumption of the on-board terminal are the attributes of the on-board terminal. The computing capability of the on-board terminal is the CPU consumption of the on-board terminal per unit time, and the computing power consumption of the on-board terminal is the power consumption required per unit consumption.

[0010] The time required for local calculation of each vehicle task can be calculated based on the computing power of the vehicle terminal. After obtaining the time required for local calculation of the vehicle task, the local consumption required for local calculation of the vehicle task can be calculated based on the computing power consumption of the vehicle terminal.

[0011] The time required for local calculation of the vehicle-mounted task is the ratio of the calculation complexity of the vehicle-mounted task to the calculation capability of the vehicle-mounted terminal multiplied by 10 cubed, where 10 cubed is the conversion from the unit s to ms.

[0012] The local energy consumption required for local computing of the vehicle-mounted task is the product of the time required for local computing of the vehicle-mounted task and the computing power consumption of the vehicle-mounted terminal.

[0013] As a preferred solution, the energy efficiency of each vehicle-mounted task is calculated to select high energy efficiency tasks, specifically including: Obtaining the task size information of each on-board task, and obtaining the energy efficiency of the on-board task according to the ratio of the task size to the local energy consumption required for local operation of the on-board task; An energy efficiency threshold is set, and vehicle-mounted tasks that are not less than the energy efficiency threshold are screened as high-energy-efficiency tasks.

[0014] The on-board task includes the attribute of task size. Task size is the data size of the on-board task, which is expressed in KB. After calculating the local energy consumption required for local operation of the on-board task, the task size of the on-board task is divided by the local energy consumption to obtain the energy efficiency. By setting the energy efficiency threshold, all on-board tasks with energy efficiency greater than or equal to the energy efficiency threshold are selected as high-efficiency tasks.

[0015] As a preferred solution, the method of selecting the target roadside unit according to the distance between the vehicle and each roadside unit specifically includes: Get the distance between the vehicle and each roadside unit, set a distance threshold, and select the roadside unit whose distance is not greater than the distance threshold as the target roadside unit.

[0016] This solution selects roadside units that are not greater than a distance threshold as target roadside units based on the distance between the vehicle and each roadside unit.

[0017] As a preferred solution, the total time and total energy consumption of each high energy efficiency task for unloading in each target roadside unit are calculated separately, specifically including: Calculate the transmission rate of the MEC offloading link between the vehicle and each roadside unit; Compare the task size of the high-energy-efficiency task with the transmission rate corresponding to the target roadside unit to obtain the unloading transmission time of the high-energy-efficiency task in the target roadside unit; Obtain the computing power of the MEC edge computing server of the corresponding target roadside unit, compare the task computational complexity of the high-energy-efficiency task with the computing power of the MEC edge computing server, and obtain the time required for the high-energy-efficiency task to be computed on the MEC edge server; The total time for the high-efficiency task to be unloaded in the target roadside unit is the sum of the unloading transmission time of the high-efficiency task in the target roadside unit and the time required for the high-efficiency task to be calculated in the MEC edge server; The transmission power of the vehicle terminal is obtained, and the total energy consumption of the high-efficiency task unloading in the target roadside unit is obtained by multiplying the transmission power of the vehicle terminal with the transmission time of the high-efficiency task unloading in the target roadside unit.

[0018] This scheme performs the above calculations for each high-energy-efficiency task, and finally obtains the total time and total energy consumption of each high-energy-efficiency task for unloading in each target roadside unit.

[0019] First, calculate the transmission rate of the MEC unloading link between the vehicle and each roadside unit, and calculate the unloading transmission time of the high-energy-efficiency task in the target roadside unit according to the high-energy-efficiency task size and transmission rate. Specifically, divide the task size of the high-energy-efficiency task by the transmission rate to obtain the unloading transmission time of the high-energy-efficiency task in each target roadside unit. The specific calculation process of the unloading transmission time is to divide the task size by the transmission rate, then multiply by 8 / 1024, and multiply by 10 cubed, where 8 / 1024 and 10 cubed are unit conversions.

[0020] The MEC edge computing server includes computing power. The computing power of the MEC edge computing server is the CPU consumption of the MEC edge computing server per unit time. According to the computing power of the MEC edge computing server and the task size of the high-efficiency task, the time required for the high-efficiency task to be calculated on the MEC edge server can be obtained. Specifically, the task size of the high-efficiency task is divided by the computing power of the MEC edge computing server. At the same time, the obtained unloading transmission time and the time required for the MEC edge server calculation are added to obtain the total time for unloading. The vehicle terminal includes a transmission power attribute, and the transmission power is multiplied by the unloading transmission time to obtain the total energy consumption of unloading. The specific calculation process is 10 to the power of transmission power / 10 multiplied by the unloading transmission time, and then divided by 1000, where 10 to the power of transmission power / 10 and 1000 are the conversion between units.

[0021] As a preferred solution, the total time and total energy consumption are normalized with the load of the roadside unit, and the judgment evaluation value is obtained by filtering and weighted calculation, which specifically includes: The total time, total energy consumption, and load of the roadside unit of the high-efficiency task unloading are normalized to unity to obtain a normalized value of the total time, a normalized value of the total energy consumption, and a normalized value of the load of the target roadside unit; Set the unloading filter factor, perform weighted sum calculation on the normalized value of total time and the normalized value of total energy consumption, and obtain the evaluation value of the unloading process; The load filtering factor is set, and the unloading process evaluation value and the target roadside unit load normalization value are weighted and calculated to obtain the judgment evaluation value.

[0022] The normalization process tending to one means that the numerical processing tends to one with one as the center. The calculation process is to calculate the mean of the corresponding numerical sequence. The mean is the mathematical expectation value, and the ratio of each numerical value to the mean is the numerical normalization value. Through the normalization process tending to one, the total time, total energy consumption, and load are distributed in the range (2,0) centered on 1 to achieve comparability, where the normalized value equal to the mean is 1, the normalized value less than the mean is in the range (0,1), and the normalized value greater than the mean is in the range (2,1). The normalization process tending to one in this scheme is a transformation form of normalization, which is an existing conventional technical means. The unloading filter factor and the load filter factor are set as weights, and their ranges are both [0,1]. The weight is set with the unloading filter factor, and the weighted sum of the normalized value of the total time and the normalized value of the total energy consumption is calculated. Specifically, the difference between the unloading filter factor, 1 and the unloading filter factor is used as the weight, and they are multiplied with the normalized value of the total time and the normalized value of the total energy consumption respectively. The weighted sum value obtained by adding up their respective products is the unloading process evaluation value. The weight is set with the load filter factor, and the weighted sum of the unloading process evaluation value and the target roadside unit is calculated. Specifically, the difference between the load filter factor, 1 and the load filter factor is used as the weight, and they are multiplied with the normalized value of the target roadside unit load and the unloading process evaluation value respectively. The weighted sum value obtained by adding up their respective products is the judgment evaluation value. Through the above process, the judgment evaluation values ​​of all target roadside units corresponding to each high-efficiency task are obtained respectively.

[0023] As a preferred solution, the roadside unit corresponding to the minimum judgment evaluation value is selected for the high energy efficiency task to implement task unloading, specifically including: For high energy efficiency tasks, all judgment evaluation values ​​corresponding to the target roadside unit are arranged in ascending order, and the target roadside unit corresponding to the first judgment evaluation value is selected from the sequence to determine whether the target roadside unit is empty. If the target roadside unit is empty, the high-efficiency task is unloaded to the target roadside unit. If the target roadside unit is not empty, then select the target roadside unit corresponding to the next judgment evaluation value from the sequence, and repeat the judgment of whether the target roadside unit is empty; If the target roadside units corresponding to all judgment evaluation values ​​in the sequence are occupied, the high-efficiency task will be processed locally and the unloading operation will not be performed.

[0024] The judgment evaluation values ​​of the target roadside units corresponding to the high-energy-efficiency tasks are arranged in ascending order, and the target roadside unit corresponding to the minimum judgment evaluation value is selected to implement task unloading. At the same time, it is also necessary to judge whether the target roadside unit is empty, that is, whether it is occupied. If it is not occupied, the high-energy-efficiency task is unloaded to the target roadside unit. If it is occupied, the target roadside unit corresponding to the next judgment evaluation value is selected, and the judgment of whether it is occupied is repeated, and so on, until the high-energy-efficiency task is unloaded to the target roadside unit. If all target roadside units are occupied, the high-energy-efficiency task is processed locally and the unloading operation is no longer performed. The above operation is performed for each high-energy-efficiency task, and all high-energy-efficiency tasks are unloaded.

[0025] As a preferred solution, the transmission rate of the MEC unloading link between the calculation vehicle and each roadside unit is specifically: According to the distance between the vehicle and the roadside unit, the quality status of the MEC offloading link is calculated, and the transmission rate of each link is calculated based on the signal-to-noise ratio of the MEC offloading link.

[0026] As a preferred solution, the calculation of the MEC offloading link quality status is specifically as follows: Set the path loss constant and path loss coefficient of the MEC offloading link and the vehicle terminal transmission power; According to the distance from the current vehicle to each roadside unit, the link path loss between the vehicle and each roadside unit is calculated. The link path loss is the product of the road loss constant plus the road loss coefficient and the logarithmic function of the distance from the current vehicle to the roadside unit / 1000. Calculate the vehicle's MEC offloading link channel gain, which is the difference between the vehicle terminal's transmit power and the link path loss.

[0027] The link path loss is calculated based on the distance between the vehicle and the roadside unit, and the vehicle MEC offloading link channel gain, i.e., the MEC offloading link quality status, is obtained by subtracting the link path loss from the vehicle terminal transmission power.

[0028] Specifically, set the path loss constant Ct of the MEC offloading link dd and path loss coefficient Cf dd , get the vehicle terminal transmission power Pwr snd (dBm); according to the distance between the vehicle and each roadside unit RDD={Rdd1,Rdd2,…, Rdd m}(m), calculate the link path loss between the vehicle and each roadside unit, PdL i =Ct dd +Cf dd *log(Rdd i / 1000)(dB); Where log() represents the common logarithm function.

[0029] Calculate the vehicle's MEC offloading link channel gain, i.e., the MEC offloading link quality status, Glink i =Pwr snd -PDLos i (dB).

[0030] As a preferred solution, the calculation of the transmission rate of each link in combination with the signal-to-noise ratio of the MEC offloading link specifically includes: Set the Boltzmann constant, Kelvin temperature, number of subcarriers occupied by a single resource block, subcarrier bandwidth, and carrier bandwidth of the roadside unit. The product of the Boltzmann constant, Kelvin temperature, the number of subcarriers occupied by a single resource block, the subcarrier bandwidth, and 1024 is converted into signal strength to obtain the Gaussian white noise power spectrum density; The signal-to-noise ratio of each MEC offloading link is obtained by subtracting the channel gain of each MEC offloading link from the power spectrum density of Gaussian white noise; The transmission rate of each MEC offload link is calculated based on the MEC offload link signal-to-noise ratio and carrier bandwidth.

[0031] According to the conversion formula between dBm and W, the Gaussian white noise power spectrum density is calculated by the Boltzmann constant, Kelvin temperature, the number of subcarriers occupied by a single resource block, and the subcarrier bandwidth. The specific expression is as follows: Set the Boltzmann constant K b , Kelvin temperature T k 、Number of subcarriers occupied by a single resource block Nm rb , subcarrier bandwidth Wd rb (kHz) , calculate the power spectral density of Gaussian white noise: Pp Gs =10*log(K b *T k *Nm rb *Wd rb *1024)+30(dBm / Hz) The signal-to-noise ratio of the MEC offloading link is: Snr i =Glink i -Pp Gs (dB), Set the carrier bandwidth Wd of the roadside unit cr (MHz), calculate the transmission rate of each MEC offloading link, Thr i = Wd cr *log2(1+ Snr i ) (Mbps), Among them, log2() represents the logarithmic function with base 2.

[0032] Therefore, the advantages of the present invention are: the local computing time and local energy consumption are calculated by the complexity of the on-board tasks, the energy efficiency of different on-board tasks is calculated according to the complexity of the on-board tasks and the local energy consumption, and the on-board tasks with high energy efficiency are screened out for unloading. The target roadside unit set for task unloading is screened according to the distance between the vehicle and the roadside unit, the transmission rate of the unloading link is calculated, and then the total time and total energy consumption for unloading high-energy-efficiency tasks are calculated, the data is normalized in combination with the roadside unit load, and filtering and weighting are performed to obtain the judgment evaluation values ​​required for unloading different on-board tasks, and the target roadside unit for unloading the target task is identified according to the minimization principle, which can ensure that the vehicle is always in a good unloading quality environment, and provide real-time guarantee for improving the success rate of task unloading. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of a process of the present invention.

[0034] Figure 2 It is a relationship diagram between the RSU distance and the offloadable real-time tasks under the same vehicle state in the present invention.

[0035] Figure 3 It is a comparison chart of the energy consumption of the vehicle terminals under the same vehicle state in the present invention.

[0036] Figure 4 It is a comparison chart of the MEC task offloading success rate under the same vehicle state in the present invention. DETAILED DESCRIPTION

[0037] The technical solution of the present invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0038] Embodiment 1:

[0039] This embodiment provides a method for offloading mobile edge computing tasks based on energy efficiency in an Internet of Vehicles, wherein the MEC includes a roadside unit arranged in a vehicle driving area, the roadside unit is equipped with a MEC edge computing server, and the Internet of Vehicles includes a vehicle-mounted terminal, such as Figure 1 As shown, the method comprises the following steps: According to the local energy consumption required for local computing of on-board tasks, the energy efficiency of each on-board task is calculated and high-efficiency tasks are selected.

[0040] Filter out target roadside units according to the distance between the vehicle and each roadside unit; For each high-energy-efficiency task, calculate the total time and total energy consumption for unloading in each target roadside unit; The total time and total energy consumption are normalized with the load of the roadside unit, and the judgment evaluation value is obtained through filtering and weighted calculation; For high energy efficiency tasks, the target roadside unit corresponding to the minimum judgment evaluation value is selected to implement task unloading.

[0041] As a preferred solution of this embodiment, the calculation of local energy consumption required for local operation of the vehicle-mounted task includes: The computational complexity of the vehicle-mounted tasks is obtained, the time required for local operation of each vehicle-mounted task is calculated according to the computing power of the vehicle-mounted terminal, and the local energy consumption required for local operation of each vehicle-mounted task is calculated according to the computing power consumption of the vehicle-mounted terminal.

[0042] The local energy consumption required for local operation of the vehicle-borne task is calculated based on the computational complexity of the vehicle-borne task, the computing power of the vehicle-borne terminal, and the computing power consumption. The vehicle-borne task includes the attribute of computational complexity, that is, the CPU consumption of the vehicle-borne terminal required for the calculation of the vehicle-borne task, which is obtained by obtaining the vehicle-borne task. The vehicle-borne terminal includes the attributes of the computing power of the vehicle-borne terminal and the computing power consumption. According to the pre-setting of the vehicle-borne terminal, the computing power of the vehicle-borne terminal is the CPU consumption of the vehicle-borne terminal per unit time, and the computing power consumption of the vehicle-borne terminal is the power consumption required for the unit consumption. According to the computing power of the vehicle-borne terminal, the time required for local calculation of each vehicle-borne task can be calculated. After obtaining the time required for local operation of the vehicle-borne task, the local consumption required for local operation of the vehicle-borne task is calculated according to the computing power consumption of the vehicle-borne terminal. The specific calculation is that the time required for local calculation of the vehicle-borne task is the ratio of the computational complexity of the vehicle-borne task to the computing power of the vehicle-borne terminal, and the local energy consumption required for local operation of the vehicle-borne task is the product of the time required for local calculation of the vehicle-borne task and the computing power consumption of the vehicle-borne terminal.

[0043] As a preferred solution of this embodiment, calculating the energy efficiency of each vehicle-mounted task specifically includes: The task size information of each vehicle-borne task is obtained, and the energy efficiency of the vehicle-borne task is obtained according to the ratio of the task size to the local energy consumption required for local operation of the vehicle-borne task.

[0044] The vehicle-mounted task includes the attribute of task size, which is the data size of the vehicle-mounted task, expressed in KB, and can be obtained by obtaining the vehicle-mounted task. After calculating the local energy consumption required for the local operation of the vehicle-mounted task, the task size of the vehicle-mounted task is divided by the local energy consumption to obtain the energy efficiency.

[0045] Further, high energy-efficiency tasks are screened according to the energy efficiency of each on-board task. The specific operation is to set an energy efficiency threshold. All on-board tasks with energy efficiency greater than or equal to the energy efficiency threshold are selected as high energy-efficiency tasks and included in the high energy-efficiency task set.

[0046] As a preferred solution of this embodiment, the target roadside unit is selected according to the distance between the vehicle and each roadside unit, specifically including: Get the distance between the vehicle and each roadside unit, set a distance threshold, and select the roadside unit whose distance is not greater than the distance threshold as the target roadside unit.

[0047] The vehicle-mounted terminal obtains the distance information between itself and each roadside unit through communication. In this embodiment, a roadside unit with a shorter distance is selected as the target roadside unit for task unloading.

[0048] As a preferred solution of this embodiment, for each high energy efficiency task, the total time and total energy consumption for unloading in each target roadside unit are calculated respectively, specifically including: Calculate the transmission rate of the MEC offloading link between the vehicle and each roadside unit; Compare the task size of the high-energy-efficiency task with the transmission rate corresponding to the target roadside unit to obtain the unloading transmission time of the high-energy-efficiency task in the target roadside unit; Obtain the computing power of the MEC edge computing server of the corresponding target roadside unit, compare the task computational complexity of the high-energy-efficiency task with the computing power of the MEC edge computing server, and obtain the time required for the high-energy-efficiency task to be computed on the MEC edge server; The total time for the high-efficiency task to be unloaded in the target roadside unit is the sum of the unloading transmission time of the high-efficiency task in the target roadside unit and the time required for the high-efficiency task to be calculated in the MEC edge server; The transmission power of the vehicle terminal is obtained, and the total energy consumption of the high-efficiency task unloading in the target roadside unit is obtained by multiplying the transmission power of the vehicle terminal with the transmission time of the high-efficiency task unloading in the target roadside unit.

[0049] Perform the above steps for each high-energy-efficiency task, and finally obtain the total time and total energy consumption of each high-energy-efficiency task in each target roadside unit. First, calculate the transmission rate of the MEC unloading link between the vehicle and each roadside unit. Specifically, calculate the quality status of the MEC unloading link based on the distance between the vehicle and the roadside unit, and calculate the transmission rate of each link in combination with the signal-to-noise ratio of the MEC unloading link.

[0050] Calculate the MEC offloading link quality status, specifically: Set the path loss constant and path loss coefficient of the MEC offloading link and the vehicle terminal transmission power; According to the distance from the current vehicle to each roadside unit, the link path loss between the vehicle and each roadside unit is calculated. The link path loss is the product of the road loss constant plus the road loss coefficient and the logarithmic function of the distance from the current vehicle to the roadside unit / 1000. Calculate the vehicle's MEC offloading link channel gain, which is the difference between the vehicle terminal's transmit power and the link path loss.

[0051] This solution calculates the link path loss based on the distance between the vehicle and the roadside unit, and obtains the vehicle MEC offloading link channel gain, i.e., the MEC offloading link quality status, by subtracting the link path loss from the vehicle terminal transmission power.

[0052] The signal-to-noise ratio of the MEC offloading link is further calculated as follows: Set the Boltzmann constant, Kelvin temperature, number of subcarriers occupied by a single resource block, subcarrier bandwidth, and carrier bandwidth of the roadside unit. The product of the Boltzmann constant, Kelvin temperature, the number of subcarriers occupied by a single resource block, the subcarrier bandwidth, and 1024 is converted into signal strength to obtain the Gaussian white noise power spectrum density; The signal-to-noise ratio of each MEC offloading link is obtained by subtracting the channel gain of each MEC offloading link from the power spectrum density of Gaussian white noise.

[0053] Finally, the carrier bandwidth of the roadside unit is obtained, and the transmission rate of each MEC offloading link is calculated according to the MEC offloading link signal-to-noise ratio and carrier bandwidth.

[0054] After calculating the transmission rate, the unloading transmission time of the high-energy-efficiency task in the target roadside unit is calculated according to the high-energy-efficiency task size and the transmission rate. Specifically, the task size of the high-energy-efficiency task is divided by the transmission rate to obtain the unloading transmission time of the high-energy-efficiency task in each target roadside unit.

[0055] The MEC edge computing server includes computing power. The computing power of the MEC edge computing server is the CPU consumption of the MEC edge computing server per unit time. The vehicle terminal obtains the computing power information of the MEC edge computing server of the roadside unit through communication. According to the computing power of the MEC edge computing server and the task size of the high-energy-efficiency task, the time required for the high-energy-efficiency task to be calculated on the MEC edge server can be obtained. Specifically, the task size of the high-energy-efficiency task is divided by the computing power of the MEC edge computing server. At the same time, the obtained unloading transmission time and the time required for the MEC edge server calculation are added to obtain the total time for unloading. The vehicle terminal includes a transmission power attribute. The total energy consumption of unloading is obtained by multiplying the transmission power by the unloading transmission time.

[0056] As a preferred solution of this embodiment, the total time and the total energy consumption are normalized in combination with the load of the roadside unit, and the decision evaluation value is obtained by filtering and weighted calculation, which specifically includes: The total time, total energy consumption, and load of the roadside unit of the high-efficiency task unloading are normalized to unity to obtain a normalized value of the total time, a normalized value of the total energy consumption, and a normalized value of the load of the target roadside unit; Set the unloading filter factor, perform weighted sum calculation on the normalized value of total time and the normalized value of total energy consumption, and obtain the evaluation value of the unloading process; The load filtering factor is set, and the unloading process evaluation value and the target roadside unit load normalization value are weighted and calculated to obtain the judgment evaluation value.

[0057] The normalization process tending to one in this scheme is a form of calculation after the conversion of the normalization process, which is an existing conventional technical means. The normalization process tending to one means that the numerical value tends to one with one as the center after numerical processing. The calculation process is to calculate the mean of the corresponding numerical sequence. The mean is the mathematical expectation value, and the ratio of each numerical value to the mean is the numerical normalization value. Through the normalization process tending to one, the total time, total energy consumption, and load are distributed in the range (2,0) centered on 1 to achieve comparability, where the normalized value equal to the mean is 1, the normalized value less than the mean is in the range (0,1), and the normalized value greater than the mean is in the range (2,1). The unloading filter factor and the load filter factor are set as weights, and their ranges are both [0,1]. The weighted sum of the normalized value of the total time and the normalized value of the total energy consumption is calculated. The difference between the unloading filter factor, 1 and the unloading filter factor is used as the weight, and they are multiplied with the normalized value of the total time and the normalized value of the total energy consumption respectively. The weighted sum value obtained by adding up their respective products is the unloading process evaluation value. The weight is set with the load filter factor, and the weighted sum of the unloading process evaluation value and the target roadside unit is calculated. Specifically, the difference between the load filter factor, 1 and the load filter factor is used as the weight, and they are multiplied with the normalized value of the target roadside unit load and the unloading process evaluation value respectively. The weighted sum value obtained by adding up their respective products is the judgment evaluation value.

[0058] As a preferred solution of this embodiment, for the high energy efficiency task, the target roadside unit corresponding to the minimum judgment evaluation value is selected to implement task unloading, specifically including: For high energy efficiency tasks, all judgment evaluation values ​​corresponding to the target roadside unit are arranged in ascending order, and the target roadside unit corresponding to the first judgment evaluation value is selected from the sequence to determine whether the target roadside unit is empty. If the target roadside unit is empty, the high-energy-efficiency task is offloaded to the target roadside unit; If the target roadside unit is not empty, the target roadside unit corresponding to the next judgment evaluation value is selected from the sequence, and the judgment of whether the target roadside unit is empty is repeated until the high-efficiency task is unloaded to the target roadside unit; If the target roadside units corresponding to all judgment evaluation values ​​in the sequence are occupied, the high-efficiency task will be processed locally and the unloading operation will not be performed.

[0059] Each high-energy-efficiency task is operated once according to the above steps, and the target roadside unit corresponding to the minimum judgment evaluation value is selected to implement task unloading. After selecting the target roadside unit, it is necessary to judge whether the target roadside unit is empty or occupied. If it is not occupied, the high-energy-efficiency task is unloaded to the target roadside unit. If it is occupied, the target roadside unit corresponding to the next judgment evaluation value is selected, and the judgment of whether it is occupied is repeated, and so on, until the high-energy-efficiency task is unloaded to the target roadside unit. If all target roadside units are occupied after traversing, the high-energy-efficiency task will be processed locally and the unloading operation will no longer be performed.

[0060] Embodiment 2:

[0061] This embodiment discloses a method for offloading mobile edge computing tasks based on energy efficiency in a connected vehicle network. The method is described below using a specific formula, including a communication vehicle Vhc tar , MEC includes roadside units RSU={Rsu1,Rsu2,…, Rsu m}, the roadside unit is equipped with a MEC edge computing server, and the Internet of Vehicles includes a vehicle terminal, which is connected to the roadside unit. The roadside unit includes the corresponding load RLD={Rld1,Rld2,…, Rld m}, vehicle Vhc tar The distance from each roadside unit RDD={Rdd1,Rdd2,…, Rdd m}(m); Vehicle Vhc tar Accompanying the generation of vehicle tasks Tsl={Tsl1, Tsl2, …, Tsl n}, task size Tkm={Tkm1, Tkm2, …,Tkm n} (kB), computational complexity (CPU consumption) Tcp={ Tcp1, Tcp2, …, Tcp n} (Cycles).

[0062] like Figure 1 As shown, the method comprises the following steps: S1. Calculate the energy efficiency of each on-board task based on the local energy consumption required for local computing of the on-board task, and select high-efficiency tasks.

[0063] S11. Set the computing capacity AC of the vehicle terminal vhc (Cycles / s) and calculation power consumption Wt vhc (w), calculate each vehicle task Tsl j ,j∈[1,n] is the time required for local operation in the vehicle terminal T_Local j =Tcpj / AC vhc *power(10,3)(ms), where power() represents the power function. Calculate the local energy consumption E_Local required for local operations of each task j =T_Local j *Wt vhc (mJ).

[0064] S12. Obtain the task size information of each vehicle task and calculate Tsl for each vehicle task j Energy efficiency j =Tkm j / E_Local j / 1024*1000(MB / J).

[0065] S13. Set energy efficiency threshold Eff thr (MB / J), filter all the j ≥Eff thr The vehicle-borne tasks are high-efficiency tasks and are included in the high-efficiency task set Set heff .

[0066] S2. Filter out target roadside units according to the distance between the vehicle and each roadside unit.

[0067] Get the distance information between the vehicle and each roadside unit and set the distance threshold Dst the (m), filter to meet Rdd i ≤Dst the The roadside unit is the target roadside unit and is included in the short link set Set hlnk .

[0068] S3. For each high energy efficiency task, calculate the total time and total energy consumption for unloading in each target roadside unit.

[0069] S31. Calculate the transmission rate of the MEC offloading link between the vehicle and each roadside unit.

[0070] S311. Set the path loss constant Ct of the MEC offloading link dd and path loss coefficient Cf dd , vehicle terminal transmission power Pwr snd (dBm); Calculate the vehicle Vhc based on the distance from the current vehicle to each roadside unit tar Rsu i The link path loss PdL i =Ct dd +Cf dd *log(Rddi / 1000)(dB), Where log() represents the common logarithm function.

[0071] S312. Calculate the vehicle's MEC offloading link channel gain, i.e., the MEC offloading link quality status Glink i =Pwr snd -PDLos i (dB).

[0072] S313. Setting the Boltzmann constant K b , Kelvin temperature T k 、Number of subcarriers occupied by a single resource block Nm rb , subcarrier bandwidth Wd rb (kHz), the carrier bandwidth Wd of the roadside unit cr (MHz); Calculate Gaussian white noise power spectral density Pp Gs =10*log(K b *T k *Nm rb *Wd rb *1024)+30(dBm / Hz); Calculate the signal-to-noise ratio of each MEC offloading link based on the quality status of each MEC offloading link and the power spectrum density of Gaussian white noise Snr i =Glink i -Pp Gs (dB), Calculate the transmission rate of each MEC offload link based on the MEC offload link signal-to-noise ratio and carrier bandwidth Thr i = Wd cr *log2(1+ Snr i ) (Mbps), Among them, log2() represents the logarithmic function with base 2.

[0073] S32. For high energy efficiency task set Set heff Each task Tsl j , respectively calculate its segment link set Set hlnk Each target roadside unit Rsu i Unload transfer time T_Ltran ji =Tkm j *8 / Thr i / 1024*power(10,3)(ms); Get the computing capacity Ac of the MEC edge computing server of the corresponding target roadside unit MEC (Cycles / s), the time required to calculate energy-efficient tasks on the MEC edge server T_Edge ji =Tcp j / Ac MEC * power(10,3)(ms); The total time for each energy-efficient task to be unloaded in the target roadside unit is T_MEC ji = T_Ltran ji + T_Edge ji (ms).

[0074] S33. Calculate the total energy consumption of each high-efficiency task unloading in the target roadside unit according to the vehicle terminal transmission power E_MEC ji =power(10,Pwr snd / 10)* T_Ltran ji / 1000(mJ).

[0075] S4. The total time and total energy consumption are normalized in combination with the load of the roadside unit, and a decision evaluation value is obtained through filtering and weighted calculation.

[0076] S41. For each high energy efficiency task Tsl j In the short link set Set hlnk Each target roadside unit Rsu i The total time of unloading, total energy consumption, and Rsu of each roadside unit i The load is normalized toward unity.

[0077] Compute energy-efficient task set Set heff The total number of tasks N s =|Set heff |; Calculate the total time T_MEC ji The mathematical expectation of T_MEC_A j =∑ Ns i=1 T_MEC ji / N s ; Calculate the mathematical expectation of total energy consumption E_MEC_A j =∑ Ns i=1 E_MEC ji / Ns ; Load Rld i The mathematical expectation of RLD_A=∑ Ns i=1 R i / N s .

[0078] Calculate the total time normalized value separately M_tmec ji = T_MEC ji / T_MEC_A j , Total energy consumption normalized value M_emec ji = E_MEC ji / E_MEC_A j , Target roadside unit load normalization value M_rld i =Rld i / RLD_A.

[0079] S42. Set the unloading filter factor k of ∈[0,1], calculate the evaluation value of all unloading processes P_of ji =k of * M_tmec ji +(1-k of )* M_emec ji ; Set the load filter factor k ld ∈[0,1], calculate the final judgment evaluation value P_ld ji =k ld * M_rld i +(1-k ld )* P_of ji .

[0080] S5. For high energy efficiency tasks, select the target roadside unit corresponding to the minimum judgment evaluation value to implement task unloading.

[0081] S51. Corresponding to each high energy efficiency task Tsl in the high energy efficiency task set j , all its P_ld ji Arrange in ascending order to get the sequence Set_rsu j , select the target roadside unit Rsu corresponding to the first judgment evaluation value from the sequence z1 , determine whether the target roadside unit is empty; S52. If the target roadside unit Rsu z1 If it is empty, the high energy efficiency task Tsl j Unload to target roadside unit Rsu z1 superior; If the target roadside unit Rsu z1 If it is not empty, the target roadside unit Rsu corresponding to the next judgment evaluation value is selected from the sequence z2 , repeat the judgment of whether the target roadside unit is empty, and so on, until the high energy efficiency task is unloaded to the target roadside unit; If the sequence Set_rsu j If all the target roadside units corresponding to the judgment evaluation values ​​in are occupied, the high energy efficiency task Tsl j Local processing is performed and no uninstallation is performed.

[0082] The method of the present invention is specifically described below by using an example. Taking m=3, n=3 as an example, the task status of the vehicle operation is shown in Table 1:

[0083] Table 1 Distribution of vehicle operation tasks The roadside unit situation is shown in Table 2:

[0084] Table 2 Distribution of roadside units The basic data are shown in Table 3:

[0085] Table 3 Basic data This embodiment provides a method for offloading mobile edge computing tasks based on energy efficiency in an Internet of Vehicles, including the following steps: S1. Calculate the energy efficiency of each on-board task based on the local energy consumption required for local computing of the on-board task, and select high-efficiency tasks.

[0086] S11. Calculate each vehicle task Tsl j ,j∈[1,n] is the time required for local operation in the vehicle terminal T_Local j =Tcp j / AC vhc *power(10,3)={25,1.33,28.33}(ms); Calculate the local energy consumption required for local computing of each task E_Local j =T_Local j *Wt vhc={2.5,0.13,2.83} (mJ).

[0087] S12. Obtain the task size information of each vehicle task and calculate Tsl for each vehicle task j Energy efficiency Eff j = Tkm j / E_Local j / 1024*1000={243.38,732.42,413.6} (MB / J).

[0088] S13. Set the energy efficiency threshold and filter all the j ≥Eff thr =250 (MB / J) vehicle-borne tasks are high-energy-efficiency tasks and are included in the high-energy-efficiency task set Set heff ={Tsl2, Tsl3}.

[0089] S2. Filter out target roadside units according to the distance between the vehicle and each roadside unit.

[0090] Set the distance threshold and filter those that meet Rdd i ≤Dst the =80 roadside units are the target roadside units and are included in the short link set Set hlnk ={Rsu1, Rsu3}.

[0091] S3. For each high energy efficiency task, calculate the total time and total energy consumption for unloading in each target roadside unit.

[0092] S31. Calculate the transmission rate of the MEC offloading link between the vehicle and each roadside unit.

[0093] S311. Set the path loss constant Ct of the MEC offloading link dd and path loss coefficient Cf dd , vehicle terminal transmission power Pwr snd (dBm); Calculate the vehicle Vhc based on the distance from the current vehicle to each roadside unit tar Rsu i , link path loss of i∈[1,m=3] PdL i =Ct dd +Cf dd *log(Rdd i / 1000) ={79.18,90.5,86.86} (dB).

[0094] S312. Calculate the vehicle's MEC offloading link channel gain, i.e., the MEC offloading link quality status Glink i =Pwr snd -PDLos i ={-55.18,-66.5,-62.68} (dB).

[0095] S313. Setting the Boltzmann constant K b , Kelvin temperature T k 、Number of subcarriers occupied by a single resource block Nm rb , subcarrier bandwidth Wd rb (kHz), the carrier bandwidth Wd of the roadside unit cr (MHz); Calculate Gaussian white noise power spectral density Pp Gs =10*log(K b *T k *Nm rb *Wd rb *1024)+30=-118.31 (dBm / Hz); Calculate the signal-to-noise ratio of each MEC offloading link based on the quality status of each MEC offloading link and the power spectrum density of Gaussian white noise Snr i =Glink i -Pp Gs ={63.13,51.81,55.46} (dB).

[0096] Calculate the transmission rate of each MEC offload link based on the MEC offload link signal-to-noise ratio and carrier bandwidth Thr i = Wd cr *log2(1+ Snr i ) ={600.29,572.28,581.9} (Mbps).

[0097] S32. For high energy efficiency task set Set heff = Each task Tsl in {Tsl2, Tsl3} j , respectively calculate its segment link set Set hlnk = Each target roadside unit Rsu in {Rsu1, Rsu3} i Unload transfer time T_Ltran ji =Tkm j *8 / Thr i / 1024*power(10,3) ={{1.3,1.34},{15.62,16.11}} (ms); Get the computing capacity Ac of the MEC edge computing server of the corresponding target roadside unit MEC (Cycles / s), the time required to calculate energy-efficient tasks on the MEC edge server T_Edge ji =Tcp j / Ac MEC * power(10,3) ={{0.2,0.2},{4.25,4.25}} (ms); The total time for each energy-efficient task to be unloaded in the target roadside unit is T_MEC ji = T_Ltran ji + T_Edge ji ={{1.5,1.54},{19.87,20.36}} (ms).

[0098] S33. Calculate the total energy consumption of each high-efficiency task unloading in the target roadside unit according to the vehicle terminal transmission power E_MEC ji =power(10,Pwr snd / 10)* T_Ltran ji / 1000 ={{0.33,0.34},{3.92,4.05}} (mJ).

[0099] S4. The total time and total energy consumption are normalized in combination with the load of the roadside unit, and a decision evaluation value is obtained through filtering and weighted calculation.

[0100] S41. For each high energy efficiency task Tsl j In the short link set Set hlnk Each target roadside unit Rsu i The total time of unloading, total energy consumption, and Rsu of each roadside unit i The load is normalized toward unity.

[0101] Compute energy-efficient task set Set heff The total number of tasks N s =|Set heff |=2; Calculate the total time T_MEC ji The mathematical expectation of T_MEC_A j =∑Ns i=1 T_MEC ji / N s ={1.52,20.11}(ms); Calculate the mathematical expectation of total energy consumption E_MEC_A j =∑ Ns i=1 E_MEC ji / N s ={0.33,3.98}(mJ); Load Rld i The mathematical expectation of RLD_A=∑ Ns i=1 R i / N s =0.55.

[0102] Calculate the total time normalized value separately M_tmec ji = T_MEC ji / T_MEC_A j ={{0.9865,1.0135},{0.9877,1.0123}}, Total energy consumption normalized value M_emec ji = E_MEC ji / E_MEC_A j ={{0.9844,1.0156},{0.9844,1.0156}}, Target roadside unit load normalization value M_rld i =Rld i / RLD_A ={{1.0909,0.9091},{1.0909,0.9091}}.

[0103] S42. Set the unloading filter factor k of , calculate the evaluation value of all unloading processes P_of ji =k of * M_tmec ji +(1-k of )* M_emec ji ={{0.9857,1.0143},{0.9864,1.0136}}; Set the load filter factor k ld , calculate the final judgment evaluation value P_ld ji =k ld * M_rld i +(1-k ld )* P_of ji ={{1.0383,0.9617},{1.0387,0.9613}}.

[0104] S5. For high energy efficiency tasks, select the target roadside unit corresponding to the minimum judgment evaluation value to implement task unloading.

[0105] S51. Corresponding to each high energy efficiency task Tsl in the high energy efficiency task set j , all its P_ld ji Arrange in ascending order to get the sequence Set_rsu j ={{0.9617,1.0383},{0.9613,1.0387,}}; Select the target roadside unit Rsu corresponding to the first judgment evaluation value from the sequence z1=3 , determine whether the target roadside unit is empty; If the target roadside unit Rsu z1=3 If it is empty, the high energy efficiency task Tsl j Unload to target roadside unit Rsu z1=3 superior.

[0106] If the target roadside unit Rsu z1=3 If it is not empty, that is, it is occupied, then select the target roadside unit Rsu corresponding to the next judgment evaluation value from the sequence z2=1 , target roadside unit Rsu z2=1 If it is empty, the high energy efficiency task Tsl j Unload to target roadside unit Rsu z1=1 superior.

[0107] The effect of the present invention is explained below through simulation experiments. The performance of the method (EFJ-MOM) of the present invention was simulated on the MATLAB platform and compared with a typical complete unloading optimization algorithm (hereinafter referred to as the first algorithm) and a joint task unloading algorithm in a multi-user system (hereinafter referred to as the second algorithm). The basic data information is shown in Table 3 above, and the results are shown in Table 3. Figures 2 to 4 shown.

[0108] like Figure 2The figure shows the relationship between the same vehicle and the real-time tasks that can be unloaded at different RSU distances. In general, for this algorithm, the more real-time tasks there are on the vehicle, the more tasks that can be unloaded; the farther the vehicle is from the RSU, the worse the cellular link quality, the more severe the channel attenuation, the lower the transmission rate, and the fewer tasks that can be unloaded; like Figure 3 The figure shows the comparison of unloading energy consumption of the vehicle terminal of the same vehicle under different link distance states. Similarly, the more real-time tasks there are on the vehicle, the greater the energy consumption of the vehicle terminal; the farther the unloading distance, the greater the energy consumption; but compared with local computing, the energy consumption in the unloading state is generally lower, reducing by about 45%; like Figure 4 The task offloading success rate of the algorithm EFJ-MOM shown in the figure is compared with that of the first and second algorithms. The first algorithm adopts a complete offloading method, that is, it does not consider the resource status of the local and MEC servers, and all tasks are offloaded to the edge server, which will directly lead to task offloading conflicts, and may cause task offloading failures. The second algorithm considers the joint offloading of multiple users and multiple tasks, which can better improve the success rate of task offloading, but the operation complexity is relatively high. EFI-MOM prefers to offload high-efficiency tasks to links with better quality, and is more inclined to offload to roadside units with light loads. Therefore, the offloading success rate is higher.

[0109] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A mobile edge computing task offloading method based on energy efficiency in an Internet of Vehicles, comprising a roadside unit, the roadside unit is configured with a MEC edge computing server, and the Internet of Vehicles includes an on-board terminal, characterized in that: The following steps are involved: Calculate the energy efficiency of each on-board task based on the local energy consumption required for local computing of the on-board task, and select high-efficiency tasks; Filter out target roadside units according to the distance between the vehicle and each roadside unit; For each high-energy-efficiency task, calculate the total time and total energy consumption for unloading in each target roadside unit; The total time and total energy consumption are normalized with the load of the roadside unit, and the judgment evaluation value is obtained through filtering and weighted calculation; For high energy efficiency tasks, the target roadside unit corresponding to the minimum judgment evaluation value is selected to implement task unloading.

2. According to claim 1, a method for offloading mobile edge computing tasks based on energy efficiency in a connected vehicle network is characterized by: The local energy consumption required for local computing of the vehicle-mounted tasks is specifically: The computational complexity of the vehicle-mounted tasks is obtained, the time required for local operation of each vehicle-mounted task is calculated according to the computing power of the vehicle-mounted terminal, and the local energy consumption required for local operation of each vehicle-mounted task is calculated according to the computing power consumption of the vehicle-mounted terminal.

3. According to claim 2, a method for offloading mobile edge computing tasks based on energy efficiency in a connected vehicle network is characterized in that The energy efficiency of each vehicle-mounted task is calculated to select high-efficiency tasks, specifically including: Obtaining the task size information of each on-board task, and obtaining the energy efficiency of the on-board task according to the ratio of the task size to the local energy consumption required for local operation of the on-board task; An energy efficiency threshold is set, and vehicle-mounted tasks that are not less than the energy efficiency threshold are screened as high-energy-efficiency tasks.

4. According to claim 1, a method for offloading mobile edge computing tasks based on energy efficiency in a connected vehicle network is characterized by: The method of selecting the target roadside unit according to the distance between the vehicle and each roadside unit specifically includes: Get the distance between the vehicle and each roadside unit, set a distance threshold, and select the roadside unit whose distance is not greater than the distance threshold as the target roadside unit.

5. A method for offloading mobile edge computing tasks based on energy efficiency in a vehicle network according to claim 1, 2, 3 or 4, characterized in that: The total time and total energy consumption of each high energy efficiency task for unloading in each target roadside unit are calculated separately, specifically including: Calculate the transmission rate of the MEC offloading link between the vehicle and each roadside unit; Compare the task size of the high-energy-efficiency task with the transmission rate corresponding to the target roadside unit to obtain the unloading transmission time of the high-energy-efficiency task in the target roadside unit; Obtain the computing power of the MEC edge computing server of the corresponding target roadside unit, compare the task computational complexity of the high-energy-efficiency task with the computing power of the MEC edge computing server, and obtain the time required for the high-energy-efficiency task to be computed on the MEC edge server; The total time for the high-efficiency task to be unloaded in the target roadside unit is the sum of the unloading transmission time of the high-efficiency task in the target roadside unit and the time required for the high-efficiency task to be calculated in the MEC edge server; The transmission power of the vehicle terminal is obtained, and the total energy consumption of the high-efficiency task unloading in the target roadside unit is obtained by multiplying the transmission power of the vehicle terminal with the transmission time of the high-efficiency task unloading in the target roadside unit.

6. A method for offloading mobile edge computing tasks based on energy efficiency in a vehicle network according to claim 5, characterized in that The total time and total energy consumption are combined with the load of the roadside unit for normalization, and the judgment evaluation value is obtained by filtering and weighted calculation, specifically including: The total time, total energy consumption, and load of the roadside unit of the high-efficiency task unloading are normalized to unity to obtain a normalized value of the total time, a normalized value of the total energy consumption, and a normalized value of the load of the target roadside unit; Set the unloading filter factor, perform weighted sum calculation on the normalized value of total time and the normalized value of total energy consumption, and obtain the evaluation value of the unloading process; The load filtering factor is set, and the unloading process evaluation value and the target roadside unit load normalization value are weighted and calculated to obtain the judgment evaluation value.

7. A method for offloading mobile edge computing tasks based on energy efficiency in a vehicle network according to claim 1, 2, 3 or 4, characterized in that: The selecting of the roadside unit corresponding to the minimum judgment evaluation value for the high energy efficiency task to implement task unloading specifically includes: For high energy efficiency tasks, all judgment evaluation values ​​corresponding to the target roadside unit are arranged in ascending order, and the target roadside unit corresponding to the first judgment evaluation value is selected from the sequence to determine whether the target roadside unit is empty. If the target roadside unit is empty, the high-efficiency task is unloaded to the target roadside unit. If the target roadside unit is not empty, then select the target roadside unit corresponding to the next judgment evaluation value from the sequence, and repeat the judgment of whether the target roadside unit is empty; If the target roadside units corresponding to all judgment evaluation values ​​in the sequence are occupied, the high-efficiency task will be processed locally and the unloading operation will not be performed.

8. The method for offloading mobile edge computing tasks based on energy efficiency in a vehicle network according to claim 5 is characterized in that The transmission rate of the MEC unloading link between the computing vehicle and each roadside unit is specifically: According to the distance between the vehicle and the roadside unit, the quality status of the MEC offloading link is calculated, and the transmission rate of each link is calculated based on the signal-to-noise ratio of the MEC offloading link.

9. A method for offloading mobile edge computing tasks based on energy efficiency in a vehicle network according to claim 8, characterized in that The calculation of the MEC offloading link quality status is specifically: Set the path loss constant and path loss coefficient of the MEC offloading link and the vehicle terminal transmission power; According to the distance from the current vehicle to each roadside unit, the link path loss between the vehicle and each roadside unit is calculated. The link path loss is the product of the road loss constant plus the road loss coefficient and the logarithmic function of the distance from the current vehicle to the roadside unit / 1000. Calculate the vehicle's MEC offloading link channel gain, which is the difference between the vehicle terminal's transmit power and the link path loss.

10. A method for offloading mobile edge computing tasks based on energy efficiency in a connected vehicle network according to claim 9, characterized in that The calculation of the transmission rate of each link in combination with the signal-to-noise ratio of the MEC offloading link specifically includes: Set the Boltzmann constant, Kelvin temperature, number of subcarriers occupied by a single resource block, subcarrier bandwidth, and carrier bandwidth of the roadside unit. The product of the Boltzmann constant, Kelvin temperature, the number of subcarriers occupied by a single resource block, the subcarrier bandwidth, and 1024 is converted into signal strength to obtain the Gaussian white noise power spectrum density; The signal-to-noise ratio of each MEC offloading link is obtained by subtracting the channel gain of each MEC offloading link from the power spectrum density of Gaussian white noise; The transmission rate of each MEC offload link is calculated based on the MEC offload link signal-to-noise ratio and carrier bandwidth.

Images