A computing task processing method and edge computing network system

Abstract

The present application discloses a method for processing computing tasks and an edge computing network system. The method includes: determining a target computing task set, the target computing task set including multiple computing tasks initiated by a user terminal. The computing tasks in the target computing task set are classified to obtain a first type of computing task completed by local processing of the user terminal, a second type of computing task completed by the user terminal unloaded to a drone for processing, and a third type of computing task completed by the user terminal unloaded to a ground base station for processing using a drone as a relay. The first type of computing task is processed based on the user terminal; and the second type of computing task and the third type of computing task are unloaded to the drone based on the user terminal, so that the second type of computing task is processed by the drone, and the third type of computing task is unloaded to the ground base station for processing by the drone. The present application can effectively utilize drones in edge computing services to provide computing power support.

Description

Technical Field

[0001] This document relates to the field of edge computing technology, and in particular to a method for processing computing tasks and an edge computing network system. Background Art

[0002] With the rapid development of edge computing technology and drone technology, using drones as servers to provide edge computing services by leveraging their rapid deployment characteristics has become a hot topic that has attracted much attention.

[0003] At present, the battery capacity and processor power of drones have certain limitations. In the case of insufficient battery life and computing power, how to effectively use drones to provide edge computing services is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The embodiments of the present application provide a method for processing computing tasks and an edge computing network system, which can effectively utilize drones in edge computing services to provide computing power support.

[0005] To solve the above technical problems, the embodiments of the present application are implemented as follows:

[0006] In a first aspect, a method for processing a computing task is proposed, which is applied to a target edge computing network system, wherein the target edge computing network system includes a user terminal, a drone, and a ground base station, wherein the drone and the ground base station serve as edge computing servers for use by the user terminal; the method includes:

[0007] Determining a target computing task set, wherein the target computing task set includes a plurality of computing tasks initiated by the user terminal;

[0008] Classifying the computing tasks in the target computing task set to obtain a first type of computing tasks that are processed locally by the user terminal, a second type of computing tasks that are offloaded by the user terminal to the drone for processing, and a third type of computing tasks that are offloaded by the user terminal to the ground base station for processing using the drone as a relay;

[0009] The first type of computing task is processed based on a user terminal; and the second type of computing task and the third type of computing task are offloaded to the drone based on the user terminal, so that the second type of computing task is processed by the drone, and the third type of computing task is offloaded to the ground base station for processing by the drone.

[0010] In the second aspect, an edge computing network system is proposed, including: a user terminal, a drone and a ground base station, the drone and the ground base station serving as edge computing servers for the user terminal; an extraction module for extracting alarm logs from a business system;

[0011] The user terminal is used to: determine a target computing task set, the target computing task set includes multiple computing tasks initiated by the user terminal; classify the computing tasks in the target computing task set to obtain a first type of computing task completed by the user terminal locally, a second type of computing task completed by the user terminal unloaded to a drone for processing, and a third type of computing task completed by the user terminal unloaded to the ground base station via the drone as a relay; process the first type of computing task, and unload the second type of computing task and the third type of computing task to the drone;

[0012] The UAV is used to: process the received second type computing task, and offload the received third type computing task to the ground base station;

[0013] The ground base station is used to: process the received third type computing task.

[0014] This application configures a drone and a ground base station that can be used as an edge computing server for the user terminal in the edge computing network system. After the user terminal initiates multiple computing tasks, the multiple computing tasks are first classified to determine the first type of computing tasks that are processed locally by the user terminal, the second type of computing tasks that are unloaded by the user terminal to the drone for processing, and the third type of computing tasks that are unloaded by the user terminal to the ground base station for processing using the drone as a relay. In this way, for the third type of computing tasks that the drone is not suitable for processing, the user terminal can use the drone as a relay to establish contact with the ground base station, so as to take advantage of the high mobility of the drone and quickly unload the third type of computing tasks to the ground base station with a stable power supply from the power grid for processing, so as to maximize its role in edge computing under the limitations of the drone's endurance and insufficient computing power. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0016] Figure 1 This is a schematic diagram of the structure of the edge computing network system of an embodiment of the present application.

[0017] Figure 2 A flowchart of processing a computing task according to an embodiment of the present application is shown.

[0018] Figure 3 A schematic diagram of the structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0019] As mentioned earlier, the battery capacity and processor power of drones have certain limitations. In the case of insufficient battery life and computing power, how to effectively use drones to provide edge computing services is a technical problem that needs to be solved urgently.

[0020] In view of this, this application proposes a new edge computing solution, applying the target edge computing network system. Figure 1 This is a structural diagram of the target edge computing network system of this embodiment, including: user terminals, drones and ground base stations. Among them, the user terminal is responsible for initiating computing tasks; the drone and the ground base station serve as edge computing servers for the user terminal. In this application, for computing tasks that the drone can handle, the user terminal uses the drone as an edge computing server; for computing tasks that the drone is not suitable for processing, the user terminal uses the drone as a relay to connect with the ground base station, and unloads the computing tasks to the ground base station for processing through the drone.

[0021] In order to make the purpose, technical solutions and advantages of this document clearer, the technical solutions of this application will be clearly and completely described below in combination with the specific embodiments of this application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of this document, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this document.

[0022] Figure 2 It is a processing method of a computing task of the present application, which is applied to the above-mentioned target edge computing network system, including:

[0023] S102, determining a target computing task set, where the target computing task set includes a plurality of computing tasks initiated by a user terminal.

[0024] In this embodiment, the computing tasks in the target computing task set may come from the same user terminal or from different user terminals, which is not specifically limited herein.

[0025] S104, classify the computing tasks in the target computing task set to obtain the first type of computing tasks completed by local processing by the user terminal, the second type of computing tasks completed by the user terminal offloading to the drone for processing, and the third type of computing tasks completed by the user terminal offloading to the ground base station for processing using the drone as a relay.

[0026] As a feasible implementation method, this embodiment classifies the computing tasks in the target computing task set with the purpose of maximizing the number of computing tasks in the target computing task set to be completed within a preset time length T.

[0027] Here we take the policy gradient method as an example:

[0028] A strategy function for mapping states to actions can be constructed; the strategy parameter of the strategy function is the computing task classification result corresponding to the target computing task set; the total processing time corresponding to the state target computing task set under the computing task classification result; the actions include: the action of changing the computing task from being processed locally by the user terminal to being offloaded by the user terminal to the drone for processing, or the action of changing the computing task from being processed locally by the user terminal to being offloaded by the user terminal to the ground base station for processing using the drone as a relay;

[0029] Next, a reward function is defined. The reward function is used to represent the benefits of changing the computing task from being processed locally by the user terminal to being offloaded by the user terminal to the drone for processing, and the benefits of changing the computing task from being processed locally by the user terminal to being offloaded by the user terminal to the ground base station for processing using the drone as a relay.

[0030] Afterwards, based on the reward function, the policy parameters of the policy function are updated with the increase of the benefit within the preset time as the gradient to determine the first type of computing tasks in the target computing task set that are completed by the user terminal locally, the second type of computing tasks that are completed by the user terminal unloading to the drone for processing, and the third type of computing tasks that are completed by the user terminal unloading to the ground base station with the drone as a relay. If it is necessary to reduce the influence of variance in the gradient update process, a baseline (such as the estimated value of the state) can also be introduced. After determining the gradient, the policy parameters of the policy function are further updated based on the baseline.

[0031] It should be noted that:

[0032] The time taken for any target computing task to be unloaded from the user terminal to the drone includes: the time taken for the target computing task to be unloaded from the user terminal to the drone, and the time taken for the drone to process the target computing task.

[0033] The time taken for any target computing task to be offloaded from the user terminal to the drone via the relay and processed by the ground base station includes: the time taken for the target computing task to be offloaded from the user terminal to the drone, the time taken for the target computing task to be offloaded from the drone to the ground base station, and the time taken for the ground base station to process the target computing task, thereby obtaining the first type of computing task completed by the local processing of the user terminal, the second type of computing task offloaded from the user terminal to the drone for processing, and the third type of computing task completed by the user terminal to be offloaded from the user terminal to the ground base station for processing via the drone as the relay.

[0034] The time it takes for the target computing task to be offloaded from the user terminal to the drone is determined based on the noise power of the drone's receiving end and the transmission power of the terminal device; the time it takes for the target computing task to be offloaded from the drone to the ground base station is determined based on the noise power of the ground base station's receiving end and the transmission power of the drone. The time it takes for the drone to process the target computing task is determined based on the data volume and computational complexity of the target computing task, as well as the CPU clock power of the drone; the time it takes for the ground base station to process the target computing task is determined based on the data volume and computational complexity of the target computing task, as well as the CPU clock power of the ground base station.

[0035] S106, processing the first type of computing task based on the user terminal; and, offloading the second type of computing task and the third type of computing task to the drone based on the user terminal, so as to process the second type of computing task through the drone, and offloading the third type of computing task to the ground base station for processing through the drone.

[0036] The following is a detailed description of the process of classifying computing tasks by the method of this embodiment.

[0037] Consider a UAV, N ground user equipment (UE), and a ground base station (BS). The BS is equipped with a powerful edge computing server. Let N = {1, 2, ..., N} represent the set of ground users. The system runs in a time slot-by-time slot manner. The length of each time slot is Δ. This paper focuses on the long-term benefit optimization problem within T time slots, and let T = {1, 2, ..., T} represent the set of all time slots. Using a three-dimensional coordinate system, the horizontal coordinates of the BS and the nth UE can be expressed as l BS =(x b ,y b )and The drone flies at a fixed altitude H and along a predetermined trajectory. The length of each time slot is divided into small enough parts to ensure that the position of the drone in a time slot remains approximately unchanged. The horizontal coordinate of the drone in time slot t can be expressed as l UAV (t) = [x U (t),y U (t)].

[0038] Assuming that the UAV air-to-ground wireless channel is dominated by the line-of-sight channel, the channel power gains between the UAV and the ground BS and the nth ground UE in time slot t can be expressed as

[0039]

[0040]

[0041] where ρn ,n∈N represents the channel power gain when the reference distance is 1 meter.

[0042] At the beginning of each time slot, the user has a new pending computing task arriving, consisting of a triple Indicates that Represents a computing task The amount of input data, Represents a computing task The computational complexity of Represents a computing task Consider the scenario where tasks arrive randomly in practical applications, that is, computing tasks are distributed from a discrete task set The corresponding computational complexity and maximum tolerable processing delay can be expressed as and

[0043] At the beginning of each time slot, the user needs to make a decision on task processing, that is, to decide whether to execute the task locally, offload it to the UAV for execution, or use the UAV as a relay to further offload the task to the remote BS for auxiliary execution. Let α n,t ∈{0,-1,1},n∈N,t∈T represents the user uninstall decision variable, where α n,t =0 means the user task is executed locally; α n,t =-1 means the computing task is offloaded to the drone for execution; α n,t =1 means that the task is offloaded to the remote BS for auxiliary computing by using the UAV relay. Considering the actual delay constraint of the computing task, when the task processing delay exceeds its maximum tolerable delay, the task is discarded.

[0044] Each UE and UAV has a transmission queue and a calculation queue, which work in a first-in-first-out manner. n,t = 0, the pending tasks will be Stored in the calculation queue waiting for processing; if α n,t = ±1, the user stores the computing task in the transmission queue and waits for transmission to the drone; if α n,t = -1, the computing task offloaded to the drone will be further stored in the drone computing queue waiting for processing; if α n,t = -1, the computing task offloaded to the UAV will be further stored in the transmission queue of the UAV, waiting for further offloading to the BS.

[0045] Computational tasks The number of time slots consumed by local calculation can be expressed as

[0046]

[0047] in, Indicates rounding up operation. Indicates the CPU clock frequency of the user's local calculation. Considering the limited computing power of the user, the computing task Usually it cannot be completed within one time slot. Therefore, the subsequent tasks need to wait for the task to be processed. The corresponding waiting time can be expressed as:

[0048]

[0049] where [x] + =max(0,x), Represents a computing task The completed time slot can be specifically expressed as:

[0050]

[0051] The value of can be obtained by recursion. in Indicates the last task processed by the UE. Therefore, the processing of computing tasks The time slots that need to be consumed can be expressed as:

[0052]

[0053] At the same time, considering the maximum processing delay requirement of the computing task, the total number of time slots to complete the computing task is:

[0054]

[0055] In the user unloading stage to the drone, the frequency division multiple access protocol is used, and the overall available bandwidth is evenly distributed to each user to ensure fairness among users. The unloading rate of user n in time slot t can be expressed as:

[0056]

[0057] Among them, W N represents the total available bandwidth, σ 2 represents the noise power at the receiving end of the drone, represents the user transmission power. The number of time slots consumed by the offloading of the computing task can be expressed as:

[0058]

[0059] Where Z ++ represents the set of positive integers, represents the time slot when the computing task starts to be unloaded. The waiting delay of task offloading transmission can be expressed as:

[0060]

[0061] in, Represents a computing task The completed time slot can be expressed as:

[0062]

[0063] The value of can be obtained by recursion, in represents the first task that the user has unloaded. Then the total number of time slots required to complete the computing task unloading process can be expressed as:

[0064]

[0065] After the offloading process is completed, the computing task will be stored in the computing queue of the drone and wait for processing. In order to ensure fairness among users, the computing resources of the drone are evenly distributed to each user. The processing time of the computing task can be expressed as:

[0066]

[0067] Among them, f max Indicates the maximum available CPU clock frequency of the drone. The waiting time of the computing task can be expressed as:

[0068]

[0069] in, Represents a computing task The completed time slot index can be expressed as:

[0070]

[0071] Available Indicates the time when the first task on the drone starts processing. Therefore, the calculation task The overall processing time can be expressed as:

[0072]

[0073] The number of time slots consumed to complete the computing task can be expressed as:

[0074]

[0075] When the UAV further offloads computing tasks to the BS, the time division multiple access protocol is used to evenly divide the time slot into N sub-time slots of equal length, that is, δ = Δ / N. The transmission rate of the UAV can be expressed as:

[0076]

[0077] in, represents the noise power at the BS receiving end, P UAV represents the UAV transmission power. The number of time slots required to perform computing task offloading can be expressed as:

[0078]

[0079] in, Represents a computing task The time slot index when starting to offload. The number of waiting time slots for computing task offloading can be expressed as:

[0080]

[0081] in, Indicates the time slot index when the computing task is offloaded, which can be specifically expressed as:

[0082]

[0083] Available, in, represents the first task offloaded to the UAV. Therefore, BS performs the computation task The total number of time slots consumed is:

[0084]

[0085] Taking into account the actual computing task latency requirements, the overall task processing time can be expressed as:

[0086]

[0087] In order to maximize the amount of computing tasks that users can complete in the long term, the optimization problem is established as follows:

[0088]

[0089] stα n,t ∈{-1,0,1},n∈N,t∈T

[0090] Among them, γ∈[0,1] is the attenuation factor, which indicates the importance of future benefits compared to current benefits; is the indicator function. Specifically, Indicates that the task is completed within the latency requirement. It means that the task processing time exceeds the delay requirement. In order to solve the above non-convex optimization problem, an optimization algorithm based on deep reinforcement learning is developed.

[0091] Decompose the original optimization problem into N independent sub-optimization problems:

[0092]

[0093] stα n,t ∈{-1,0,1},t∈T

[0094] The parameterized policy gradient method and baseline method are used to obtain the optimal task allocation. The system state, action, reward function and strategy are defined as follows:

[0095] (1) System state S n (t): The user state in time slot n is expressed as:

[0096] S n (t)=(s1(t),s2(t),s3(t),s4(t),s5(t),s6(t),s7(t),s8(t),s9(t))

[0097] in:

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105]

[0106]

[0107] (2) Action A n (t): A n (t) represents the user uninstall decision variable, that is,

[0108] A n (t) = α n,t ,n∈N,t∈T

[0109] (3) Reward function R n (t): The reward function represents the immediate reward after the user makes the uninstall decision. It is related to the objective function in the optimization problem and can be defined as:

[0110]

[0111] (4) Strategy The policy represents the user state S n (t) Towards action A n (t), that is, where θ n Represents the parameters of the policy.

[0112] Based on the above analysis, the optimization problem is transformed into:

[0113]

[0114] stα n,t ∈{-1,0,1},t∈T

[0115] The parameterized policy gradient method uses a parameterized policy to update the policy parameters through a gradient iteration method. A typical performance metric U(θ n ) can be expressed as It is the value function of the policy. The policy value gradient can be expressed as:

[0116]

[0117] In order to reduce the variance in the learning process, an action-independent baseline b[S n (t)], and obtain the policy gradient method with baseline:

[0118]

[0119] Select the estimated value of the state value v[S n (t); w] as the baseline, where is the weight vector of the state value function w. Therefore, θ n Can be updated to:

[0120]

[0121] Among them, α represents the learning rate, θ n,t represents the time slot t in θ n The estimated value of .

[0122] Based on the above, it can be seen that in the edge computing network system of this embodiment:

[0123] The user terminal is used to: determine a target computing task set, which includes multiple computing tasks initiated by the user terminal; classify the computing tasks in the target computing task set to obtain a first type of computing task completed by the user terminal locally, a second type of computing task completed by the user terminal unloaded to a drone for processing, and a third type of computing task completed by the user terminal unloaded to the ground base station via the drone as a relay; process the first type of computing task, and use a frequency division multiple access protocol to unload the second type of computing task and the third type of computing task to the drone.

[0124] The drone is used to: process the received second type of computing tasks, and unload the received third type of computing tasks to the ground base station using a time division multiple access protocol.

[0125] The ground base station is used to: process the received third type computing task.

[0126] In summary, the solution of the embodiment of the present application uses drone relay technology as an effective means to improve the quality of ground communication network channels, and can build a high-quality line-of-sight communication channel between ground users and remote high-performance ground edge servers. Based on drone relay, user computing tasks can be further offloaded to ground edge servers, and ground edge servers with stable power supply can be used to provide additional computing power support for drone edge computing networks, which can improve the stability of the network during the high-incidence stage of computing tasks and extend the drone's flight time.

[0127] Figure 3 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 3 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0128] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0129] The memory is used to store programs. Specifically, the program may include program code, and the program code includes computer operation instructions. The memory may include internal memory and non-volatile memory, and provides instructions and data to the processor. Among them, the processor reads the corresponding computer program from the non-volatile memory into the internal memory and then runs it, forming a user terminal of the edge computing network system in this embodiment at the logical level. The processor executes the program stored in the memory, and is specifically used to perform the following operations:

[0130] A target computing task set is determined, where the target computing task set includes a plurality of computing tasks initiated by the user terminal.

[0131] The computing tasks in the target computing task set are classified to obtain a first type of computing task completed by local processing by the user terminal, a second type of computing task completed by offloading from the user terminal to the drone for processing, and a third type of computing task completed by the user terminal by offloading from the user terminal to the ground base station for processing using the drone as a relay.

[0132] Processing the first type of computing tasks; and offloading the second type of computing tasks and the third type of computing tasks to the drone, so as to process the second type of computing tasks through the drone, and offloading the third type of computing tasks to the ground base station for processing through the drone.

[0133] The above application Figure 2The method disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in one or more embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with one or more embodiments of the present application can be directly embodied as a hardware decoding processor for execution, or a combination of hardware and software modules in the decoding processor for execution. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0134] Of course, in addition to software implementation methods, the electronic device of the present application does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc. That is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0135] The present application also provides a computer-readable storage medium that stores one or more programs. The one or more programs include instructions that, when executed by a portable electronic device that includes multiple application programs, enable the portable electronic device to execute Figure 1 The method of the embodiment shown is specifically used to perform the following operations:

[0136] After a first user starts the RPA task of the target browser, based on the user history queue of the target browser in the RPA system, it is determined whether the first user is the same user as the second user who last started the RPA task of the target browser.

[0137] In the case where the first user and the second user are not the same user, the browser cache of the current target browser of the RPA system is used as the history cache of the second user to be migrated to an external history cache storage area set for the target browser. And,

[0138] The RPA task initiated by the first user is executed.

[0139] In short, the above description is only a preferred embodiment of the present application and is not intended to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of the present application shall be included in the protection scope of one or more embodiments of the present application.

[0140] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0141] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0142] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0143] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

Claims

1. A method for processing computing tasks, applied to a target edge computing network system, characterized in that: The target edge computing network system includes a user terminal, a drone and a ground base station, and the drone and the ground base station serve as edge computing servers used by the user terminal; the method includes: Determining a target computing task set, wherein the target computing task set includes a plurality of computing tasks initiated by the user terminal; Classifying the computing tasks in the target computing task set to obtain a first type of computing tasks that are processed locally by the user terminal, a second type of computing tasks that are offloaded by the user terminal to the drone for processing, and a third type of computing tasks that are offloaded by the user terminal to the ground base station for processing using the drone as a relay; Processing the first type of computing tasks based on the user terminal; and offloading the second type of computing tasks and the third type of computing tasks to the drone based on the user terminal, so that the second type of computing tasks are processed by the drone, and the third type of computing tasks are offloaded to the ground base station for processing by the drone; The computing tasks in the target computing task set are classified to obtain a first type of computing task that is processed locally by a user terminal, a second type of computing task that is offloaded by the user terminal to a drone for processing, and a third type of computing task that is offloaded by the user terminal to the ground base station for processing using the drone as a relay, including: With the purpose of maximizing the number of computing tasks in the target computing task set within a preset time, the computing tasks in the target computing task set are classified to obtain a first type of computing task completed by local processing by the user terminal, a second type of computing task completed by offloading from the user terminal to the drone for processing, and a third type of computing task completed by the user terminal offloading to the ground base station for processing using the drone as a relay.

2. The method according to claim 1, characterized in that For the purpose of maximizing the number of computing tasks in the target computing task set within a preset time, the computing tasks in the target computing task set are classified to obtain a first type of computing task completed by local processing by the user terminal, a second type of computing task completed by unloading from the user terminal to the drone for processing, and a third type of computing task completed by unloading from the user terminal to the ground base station for processing using the drone as a relay, including: Construct a strategy function for mapping states to actions; the strategy parameter of the strategy function is the computing task classification result corresponding to the target computing task set; the state is the total processing time corresponding to the target computing task set under the computing task classification result; the action includes: the action of changing the computing task from being processed locally by the user terminal to being offloaded by the user terminal to the drone for processing, or the action of changing the computing task from being processed locally by the user terminal to being offloaded by the user terminal to the ground base station for processing using the drone as a relay; Constructing a reward function, the reward function is used to represent the benefits of changing the computing task from being completed by the user terminal locally to being offloaded by the user terminal to the drone for processing, and the benefits of changing the computing task from being completed by the user terminal locally to being offloaded by the user terminal to the ground base station for processing using the drone as a relay; Based on the reward function, the strategy parameters of the strategy function are updated with the increase of benefits within a preset time as the gradient to determine the first type of computing tasks in the target computing task set that are completed by local processing by the user terminal, the second type of computing tasks that are completed by offloading from the user terminal to the drone for processing, and the third type of computing tasks that are completed by the user terminal offloading to the ground base station for processing using the drone as a relay.

3. The method according to claim 2, characterized in that Based on the reward function, the policy parameters of the policy function are updated with the increase of the revenue within the preset time as the gradient, including: Based on the reward function, the policy parameters of the policy function are updated on the basis of a preset baseline with the increase of the benefit within a preset time period as the gradient; wherein the baseline is an estimated value of the state.

4. The method according to claim 2, characterized in that The time taken for any target computing task to be unloaded from the user terminal to the drone includes: the time taken for the target computing task to be unloaded from the user terminal to the drone, and the time taken for the drone to process the target computing task; The time taken for any target computing task to be offloaded from the user terminal to the ground base station via the drone as a relay includes: the time taken for the target computing task to be offloaded from the user terminal to the drone, the time taken for the target computing task to be offloaded from the drone to the ground base station, and the time taken for the ground base station to process the target computing task, thereby obtaining a first type of computing task completed by local processing by the user terminal, a second type of computing task completed by offloading from the user terminal to the drone, and a third type of computing task completed by offloading from the user terminal to the ground base station via the drone as a relay.

5. The method according to claim 3, The time taken for the target computing task to be offloaded from the user terminal to the drone is determined based on the noise power at the receiving end of the drone and the transmission power of the user terminal; The time taken for the target computing task to be offloaded from the UAV to the ground base station is determined based on the receiving end noise power of the ground base station and the transmission power of the UAV.

6. The method according to claim 3, characterized in that The time consumed by the drone to process the target computing task is determined based on the data volume and computing complexity of the target computing task, and the CPU clock power of the drone; The time consumed by the ground base station to process the target computing task is determined based on the data volume and computational complexity of the target computing task, and the CPU clock power of the ground base station.

7. An edge computing network system, characterized in that: include: A user terminal, a drone and a ground base station, wherein the drone and the ground base station serve as edge computing servers used by the user terminal; The user terminal is used to: determine a target computing task set, wherein the target computing task set includes a plurality of computing tasks initiated by the user terminal; Classifying the computing tasks in the target computing task set to obtain a first type of computing task that is processed locally by the user terminal, a second type of computing task that is offloaded by the user terminal to a drone for processing, and a third type of computing task that is offloaded by the user terminal to the ground base station for processing using the drone as a relay; processing the first type of computing task, and offloading the second type of computing task and the third type of computing task to the drone; The UAV is used to: process the received second type computing task, and offload the received third type computing task to the ground base station; The ground base station is used to: process the received third type of computing tasks; The user terminal is also specifically used to: classify the computing tasks in the target computing task set with the purpose of maximizing the number of computing tasks in the target computing task set completed within a preset time period, and obtain a first type of computing task completed by local processing by the user terminal, a second type of computing task completed by offloading from the user terminal to a drone for processing, and a third type of computing task completed by the user terminal offloading to the ground base station for processing using the drone as a relay.

8. The system according to claim 7, characterized in that The user terminal uses a frequency division multiple access protocol to offload the second type of computing tasks and the third type of computing tasks to the drone.

9. The system according to claim 7, characterized in that The UAV uses a time division multiple access protocol to offload the received third type of computing tasks to the ground base station.

Images