Edge-to-edge resource collaborative task unloading method, device, terminal, system and medium

Abstract

The invention relates to the technical field of power distribution networks, in particular to an edge-edge resource collaborative task unloading method, device, terminal and system and a medium, and the method comprises the steps: obtaining task processing capability information of at least one edge computing server corresponding to an intelligent terminal; obtaining a to-be-processed task, and dividing the to-be-processed task into a plurality of independent sub-tasks; based on the task level information of the to-be-processed task and the task processing capability information of the intelligent terminal and the edge computing server, determining the intelligent terminal or the edge computing server allocated to each sub-task; and unloading the subtasks allocated to the corresponding edge computing servers to the corresponding edge computing servers. According to the technical scheme, the to-be-processed task can be unloaded to the appropriate edge computing server for processing, the task processing efficiency is improved, the task processing resources of the intelligent terminal and the edge computing servers are reasonably used, the problems of collapse and jamming caused by excessive task processing of the intelligent terminal are avoided, and safe and stable operation of the power distribution network is ensured.

Description

Technical Field

[0001] The present disclosure relates to the technical field of distribution networks, and particularly to a task offloading method, device, terminal, system and medium for edge-edge resource collaboration. Background Art

[0002] In a distribution network system, the end side refers to the end sensing devices of the distribution network, including single-phase electricity meters, three-phase electricity meters, charging piles for initiating charging demands, photovoltaic grid-connected switches and other devices, which are used to realize the sensing and metering of various loads. The edge side includes intelligent terminals such as integrated terminals, substation terminals, feeder terminals, concentrators, etc., which are used to realize functions such as data acquisition from the same source, local analysis and calculation, and on-site fault handling. With the advancement of distribution network intelligence and the large-scale access of new types of sources and loads, when facing multi-task processing of massive data, the intelligent terminals have problems such as reduced task processing efficiency, overutilization of computing resources, and even crashes and freezes due to relatively limited computing resources, which affect the safe and stable operation of the distribution network. Summary of the Invention

[0003] To solve the problems in the related art, embodiments of the present disclosure provide a task offloading method, device, terminal, system and medium for edge-edge resource collaboration.

[0004] In a first aspect, embodiments of the present disclosure provide a task offloading method for edge-edge resource collaboration, which is applied to an intelligent terminal and includes:

[0005] Obtaining the task processing capability information of the intelligent terminal and at least one corresponding edge computing server;

[0006] Obtaining a task to be processed and dividing the task to be processed into multiple independent subtasks;

[0007] Based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, determining the allocation information for each subtask, where the allocation information includes the intelligent terminal or edge computing server assigned to the subtask;

[0008] Based on the allocation information of each subtask, offloading the subtasks assigned to the corresponding edge computing server to the corresponding edge computing server.

[0009] In a possible implementation manner, the task processing capability information includes the priority range of tasks that can be processed, the total computable resource information that can be allocated, and communication resource information;

[0010] The task level information includes a delay requirement level and a security level.

[0011] In a possible implementation manner, based on the task level information of the to-be-processed task and the task processing capability information of the intelligent terminal and the at least one edge computing server, determine the allocation information of each subtask, including:

[0012] Based on the task level information of the to-be-processed task and the task processing capability information of the intelligent terminal and the at least one edge computing server, construct a problem optimization model, where the objective function corresponding to the problem optimization model is a function for representing the minimization of the total processing overhead of all subtasks for processing the to-be-processed task, and solve the objective function to obtain the allocation information of each subtask when the total processing overhead is minimized;

[0013] Solve the problem optimization model to obtain the allocation information of each subtask.

[0014] In a possible implementation manner, the constructing a problem optimization model based on the task level information of the to-be-processed task and the task processing capability information of the intelligent terminal and the at least one edge computing server includes:

[0015] The objective function for constructing the problem optimization model is:

[0016]

[0017] where S is the total processing overhead, ω ∈ [0, 1], N 1 is the total number of subtasks corresponding to the to-be-processed task, N 2 is the total number of the intelligent terminal and the edge computing servers, j = 1 indicates that the subtask is assigned to the intelligent terminal, d i is the task data volume of the i-th subtask a i in the to-be-processed task, x i,j ∈ {0, 1} is the indicator factor to be solved, x i,j = 0 indicates that the subtask a i is not offloaded to the j-th edge computing server, x i,j = 1 indicates that the subtask a i is offloaded to the j-th edge computing server, b i,j is the bandwidth ratio to be solved when the intelligent terminal offloads the subtask a i to the j-th edge computing server, B is the known total transmission channel bandwidth of the intelligent terminal, p i,j is the transmission power to be solved when the intelligent terminal offloads the subtask a i to the j-th edge computing server, G is the known transmission channel condition status of the intelligent terminal, is the known transmission channel noise power of the intelligent terminal, I 1′Interference from other known intelligent terminals in the same frequency band The processing delay for the j-th edge computing server to complete the subtask a according to the predetermined task scheduling and processing rules i where ε is a predetermined constant parameter, and c i is the computational workload for processing the subtask a i and r i,j is the computing resource allocated by the j-th edge computing server for the subtask a to be solved i ;

[0018] The constraint conditions of the problem optimization model include:

[0019]

[0020] wherein, for all i, P is the maximum transmission power of the intelligent terminal, and R j is the total computing resource allocable by the j-th edge computing server

[0021] In a possible implementation, the method further includes:

[0022] Determining the priority of the i-th subtask a in the to-be-processed task on the j-th edge computing server according to the task level information of the to-be-processed task and the priority range of the tasks processable by the j-th edge computing server i on the j-th edge computing server;

[0023] According to the priority of the subtask a i on the j-th edge computing server, determining the processing delay i for the j-th edge computing server to complete the subtask a according to the predetermined task scheduling and processing rules

[0024] In a possible implementation, the determining the priority of the i-th subtask a in the to-be-processed task on the j-th edge computing server according to the task level information of the to-be-processed task and the priority range of the tasks processable by the j-th edge computing server includes: i :

[0025] Calculating the priority R i of the subtask a on the j-th edge computing server according to the following formula i,j :

[0026]

[0027] Among them, round(·) refers to the rounding operation, λ represents the preference degree of the task to be processed for low latency, and the priority range of the j-th edge computing server is [P j d , P j u , R t is the latency requirement level of the task to be processed, R s is the security level of the task to be processed, R t ranges from [R ta , R tb , and R s ranges from [R sa , R sb .

[0028] The predetermined task scheduling and processing rule includes preferentially executing subtasks with higher priorities. For subtasks with the same priority, the time slice rotation scheduling algorithm is used for scheduling; then, according to the priority of subtask a i in the j-th edge computing server, it is determined that the j-th edge computing server processes and completes the subtask a i with the following processing latency including:

[0029] The processing latency is calculated according to the following formula

[0030]

[0031] where is the remaining execution time of subtasks in the j-th edge computing server that have higher priorities than subtask a i , N 3 is the number of subtasks in the j-th edge computing server that have the same priority as subtask a i , is the remaining execution time of N 3 subtasks that have the same priority as subtask a i , and θ j is the time slice of the time slice rotation scheduling algorithm in the j-th edge computing node.

[0032] In a second aspect, an edge-edge resource collaboration task offloading device provided in an embodiment of the present disclosure is applied to an intelligent terminal and includes:

[0033] An information acquisition module, configured to acquire the task processing capability information of the intelligent terminal and at least one corresponding edge computing server;

[0034] A task acquisition module, configured to acquire a task to be processed and divide the task to be processed into multiple independent subtasks;

[0035] An allocation determination module, configured to determine the allocation information of each subtask based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, where the allocation information includes the intelligent terminal or edge computing server assigned to the subtask;

[0036] A task offloading module, configured to offload the subtasks assigned to the corresponding edge computing server to the corresponding edge computing server based on the allocation information of each subtask.

[0037] In a possible implementation manner, the task processing capability information includes the priority range of tasks that can be processed, the total computable resource information that can be allocated, and communication resource information;

[0038] The task level information includes a latency requirement level and a security level.

[0039] In a possible implementation manner, the allocation determination module includes:

[0040] Based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, construct a problem optimization model, where the objective function corresponding to the problem optimization model is a function used to represent the minimization of the total processing overhead of all subtasks for processing the task to be processed, and solve the objective function to obtain the allocation information of each subtask when the total processing overhead is minimized;

[0041] Solve the problem optimization model to obtain the allocation information of each subtask.

[0042] In a possible implementation manner, the part of the allocation determination module that constructs a problem optimization model based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server is configured to:

[0043] The objective function for constructing the problem optimization model is:

[0044]

[0045] where S is the total processing overhead, ω ∈ [0, 1], N 1 is the total number of subtasks corresponding to the task to be processed, N 2 is the total number of the intelligent terminal and the edge computing server, j = 1 indicates that the subtask is assigned to the intelligent terminal, d i is the i-th subtask a in the task to be processedi The task data volume of, x i,j ∈ {0, 1} is the indication factor to be solved, x i,j = 0 indicates subtask a i is not unloaded to the j-th edge computing server, x i,j = 1 indicates subtask a i is unloaded to the j-th edge computing server, b i,j is the bandwidth ratio allocated when the intelligent terminal to be solved unloads subtask a i to the j-th edge computing server. B is the total transmission channel bandwidth of the known intelligent terminal, p i,j is the transmission power allocated when the intelligent terminal to be solved unloads subtask a i to the j-th edge computing server. G is the transmission channel condition status of the known intelligent terminal, is the transmission channel noise power of the known intelligent terminal, I 1′ is the interference from other intelligent terminals in the same frequency band, is the processing delay for the j-th edge computing server to process the subtask a according to the predetermined task scheduling and processing rules, ε is a predetermined constant parameter, c i is the computational workload for processing the subtask a i is for processing the subtask a i is the computational workload, r i,j is the computational resource allocated by the j-th edge computing server to subtask a to be solved i ;

[0046] The constraint conditions of the problem optimization model include:

[0047]

[0048] Among them, refers to for all i, P is the maximum transmission power of the intelligent terminal, R j is the total computational resource that can be allocated by the j-th edge computing server.

[0049] In a possible implementation manner, the device further includes:

[0050] A priority determination module, configured to determine the priority of the i-th subtask a in the to-be-processed task in the j-th edge computing server according to the task level information of the to-be-processed task and the priority range of the tasks that can be processed by the j-th edge computing server; i in the j-th edge computing server;

[0051] A delay determination module, configured to according to subtask a iDetermine the priority of the j-th edge computing server, and determine that the j-th edge computing server processes and completes the subtask a according to a predetermined task scheduling and processing rule i Processing delay

[0052] In a possible implementation manner, the priority determination module is configured to:

[0053] Calculate the subtask a according to the following formula i Priority R of the j-th edge computing server i,j :

[0054]

[0055] where round(·) refers to the rounding operation, λ represents the preference degree of the task to be processed for low latency, and the priority range of the j-th edge computing server is [P j d , P j u , R t is the latency requirement level of the task to be processed, R s is the security level of the task to be processed, R t The value range of is [R ta , R tb , R s The value range of is [R sa , R sb .

[0056] In a possible implementation manner, the predetermined task scheduling and processing rule includes preferentially executing subtasks with higher priorities. For subtasks with the same priority, a time slice rotation scheduling algorithm is used for scheduling; then the latency determination module is configured to:

[0057] Calculate the processing latency according to the following formula

[0058]

[0059] where is the remaining execution time of the subtasks with higher priorities than subtask a i in the j-th edge computing server, N 3 is the number of subtasks with the same priority as subtask a i in the j-th edge computing server, is N 3 The remaining execution time of the subtasks with the same priority as subtask a i , θ jIt is the time slice of the round-robin scheduling algorithm in the j-th edge computing node.

[0060] In a third aspect, embodiments of the present disclosure provide an intelligent terminal, including a memory and a processor. The memory is used to store one or more computer instructions, and the one or more computer instructions are executed by the processor to implement the method according to any one of the first aspects.

[0061] In a fourth aspect, embodiments of the present disclosure provide a task offloading system for edge-edge resource collaboration, including an intelligent terminal and at least one edge computing server, where:

[0062] The intelligent terminal is configured to use the method according to any one of the first aspects to offload the subtasks assigned to the corresponding edge computing server to the corresponding edge computing server;

[0063] The edge computing server is configured to obtain the subtasks offloaded by the intelligent terminal, execute the subtasks, and return the execution results to the intelligent terminal.

[0064] In a fifth aspect, embodiments of the present disclosure provide a computer-readable storage medium, on which computer instructions are stored, and when the computer instructions are executed by a processor, the method according to any one of the first aspects is implemented.

[0065] According to the technical solution provided by the embodiments of the present disclosure, the task processing capability information of the intelligent terminal and its corresponding at least one edge computing server can be obtained; after obtaining the task to be processed and dividing the task to be processed into multiple independent subtasks, based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the edge computing server, the allocation information of each subtask can be determined, and the allocation information includes the intelligent terminal or the edge computing server assigned to the subtask; based on the allocation information of each subtask, the subtasks assigned to the corresponding edge computing server are offloaded to the corresponding edge computing server; in this way, the intelligent terminal can reasonably offload the task to be processed to a suitable edge computing server for processing, improve the task processing efficiency, reasonably use the task processing resources of the intelligent terminal and each edge computing server, avoid the problems of crash and lag caused by excessive task processing of the intelligent terminal, and ensure the safe and stable operation of the distribution network.

[0066] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In combination with the drawings, through the following detailed description of non-limiting embodiments, other features, objects, and advantages of the present disclosure will become more obvious. In the drawings:

[0068] Figure 1 Flowchart showing the task offloading method for edge-edge resource collaboration provided by an embodiment of the present disclosure.

[0069] Figure 2 Block diagram showing the structure of the task offloading device for edge-edge resource collaboration provided by an embodiment of the present disclosure.

[0070] Figure 3 Block diagram showing the structure of an intelligent terminal according to an embodiment of the present disclosure.

[0071] Figure 4 Schematic diagram showing an edge-edge resource collaborative task offloading system provided by an embodiment of the present disclosure.

[0072] Figure 5 Schematic diagram showing the structure of a computer system suitable for implementing the method of an embodiment of the present disclosure. Detailed implementation manners

[0073] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. In addition, for clarity, parts irrelevant to the description of the exemplary embodiments are omitted in the drawings.

[0074] In the present disclosure, it should be understood that terms such as "including" or "having" are intended to indicate the presence of features, numbers, steps, actions, components, parts, or combinations thereof disclosed in this specification, and are not intended to exclude the possibility of the presence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0075] In addition, it should be noted that, without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0076] Figure 1 Flowchart showing the task offloading method for edge-edge resource collaboration provided by an embodiment of the present disclosure. As Figure 1 shown, the task offloading method for edge-edge resource collaboration includes the following steps S101-S104:

[0077] In step S101, obtain the task processing capability information of the intelligent terminal and its corresponding at least one edge computing server;

[0078] In step S102, obtain the task to be processed and divide the task to be processed into multiple independent subtasks;

[0079] In step S103, based on the task level information of the to-be-processed task and the task processing capability information of the intelligent terminal and the at least one edge computing server, determine the allocation information for each subtask, where the allocation information includes the intelligent terminal or edge computing server assigned to the subtask.

[0080] In step S104, based on the allocation information for each subtask, offload the subtasks assigned to the corresponding edge computing server to the edge computing server.

[0081] In a possible implementation manner, this task offloading method for edge resource collaboration is applicable to intelligent terminals capable of performing task offloading.

[0082] In a possible implementation manner, in a new type of distribution network system, not only intelligent terminals such as integrated terminals, substation terminals, feeder terminals, and concentrators are set on the edge side, but also at least one edge computing server is set. The intelligent terminal can, according to its own and the current task processing capabilities of these edge computing servers, process the task by itself or offload it to some edge computing servers for processing, improving the task processing efficiency, reducing the reasonable use of task processing resources, solving the problems of crashes and lags caused by excessive task processing of existing intelligent terminals, and ensuring the safe and stable operation of the distribution network.

[0083] In a possible implementation manner, each edge computing server automatically reports its own task processing capability information (such as reporting periodically), or the intelligent terminal can send a query request to each edge computing server when obtaining a to-be-processed task to query the task processing capability information of each edge computing server. When each edge computing server receives the query request, it reports the task processing capability information of each edge computing server.

[0084] In a possible implementation manner, the to-be-processed task can be uploaded by the end-side device to the intelligent terminal or generated by the intelligent terminal itself. After obtaining the to-be-processed task, the to-be-processed task can be divided into multiple independent subtasks, and there is no dependency relationship between the subtasks and they can be processed independently.

[0085] In a possible implementation manner, in a new type of distribution network system, there are significant differences in the processing requirements for different types of tasks. It is necessary to fully consider the quality of service requirements of various tasks and their impact on the stable operation of the distribution network. Therefore, the task level information of various tasks can be set accordingly to indicate the quality of service requirements of various tasks and their impact on the stable operation of the distribution network.

[0086] In a possible implementation, the intelligent terminal can determine the quality of service requirements of the task to be processed and its impact on the stable operation of the distribution network based on the task level information of the task to be processed, and determine the task processing capabilities of the intelligent terminal and the edge computing server based on the task processing capability information of the intelligent terminal and the edge computing server. In this way, the intelligent terminal can accordingly determine which sub-tasks to allocate to the intelligent terminal and which edge computing servers for task processing, and obtain the allocation information of each sub-task.

[0087] In a possible implementation, after determining the allocation information of each sub-task, the intelligent terminal can offload the sub-tasks allocated to the corresponding edge computing server to the corresponding edge computing server, and the corresponding edge computing server processes the sub-task. At the same time, the intelligent terminal also processes the sub-tasks allocated to itself. In this way, the task to be processed can be split into multiple sub-tasks and processed by multiple terminals.

[0088] This implementation can obtain the task processing capability information of the intelligent terminal and its corresponding at least one edge computing server. After obtaining the task to be processed and dividing the task to be processed into multiple independent sub-tasks, the allocation information of each sub-task can be determined based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the edge computing server. The allocation information includes the intelligent terminal or edge computing server allocated for the sub-task. Based on the allocation information of each sub-task, the sub-tasks allocated to the corresponding edge computing server are offloaded to the corresponding edge computing server. In this way, the intelligent terminal can reasonably offload the task to be processed to a suitable edge computing server for processing, improve the task processing efficiency, reasonably use the task processing resources of the intelligent terminal and each edge computing server, avoid the problems of crash and lag caused by excessive task processing of the intelligent terminal, and ensure the safe and stable operation of the distribution network.

[0089] In a possible implementation, the task processing capability information includes the priority range of tasks that can be processed, the total computable resource information that can be allocated, and the communication resource information; the task level information includes the delay requirement level and the security level.

[0090] In this implementation, the priority range of tasks that can be processed indicates the priority range of tasks that the edge computing server can process, the total computable resource information that can be allocated indicates the resource information that the edge computing server can allocate to process new tasks, and the communication resource information indicates the communication resource information that the edge computing server can use to receive the sub-tasks offloaded by the intelligent terminal.

[0091] In this embodiment, the characteristics of the new distribution network system are fully explored, and the tasks in the distribution network system are classified, mainly including operation control tasks, status monitoring tasks, fault handling tasks, asset management tasks, power marketing tasks, etc. At the same time, according to expert experience, each type of task is further divided into multiple sub-tasks, and the delay requirements of each sub-task and the impact on the safe operation of the distribution network are graded. The specific grading results are shown in Table 1 below:

[0092]

[0093]

[0094] Table 1

[0095] In Table 1, the higher the delay requirement level, the lower the delay requirement. In addition, most of the existing task offloading methods overemphasize indicators such as data transmission delay and ignore the importance of safety and stability for the distribution network; while in this embodiment, expert experience is used to grade the safety level of each sub-task, and the higher the safety level, the smaller the impact of the task on the safe operation of the distribution network. It should be noted here that the task types and task level information in Table 1 are only examples. In other embodiments, there may be other task types and their corresponding task level information, which are not limited here.

[0096] In a possible implementation manner, determining the allocation information of each sub-task based on the task level information of the to-be-processed task and the task processing capabilities of the intelligent terminal and the at least one edge computing server includes:

[0097] Based on the task level information of the to-be-processed task and the task processing capabilities of the intelligent terminal and the at least one edge computing server, a problem optimization model is constructed. The objective function corresponding to the problem optimization model is a function used to represent the minimization of the total processing cost of all sub-tasks for processing the to-be-processed task. Solving the objective function obtains the allocation information of each sub-task when the total processing cost is minimized;

[0098] Solving the problem optimization model to obtain the allocation information of each sub-task.

[0099] In this embodiment, the problem optimization model involves finding the optimal solution of the objective function under given constraints. When constructing the problem optimization model in this embodiment, the corresponding objective function is the function to be minimized, that is, the function representing the minimization of the total processing overhead of all subtasks for processing the to-be-processed task. The given constraints are that the task processing capacity information of the intelligent terminal or edge computing server assigned to each subtask can ensure that the assigned subtasks can be processed (for example, ensuring that tasks at this task level information can be processed, and ensuring that the total computable resource information and communication resource information that can be allocated can process this subtask, etc.). Solving this objective function can obtain the allocation information of each subtask when the total processing overhead of all these subtasks is minimized.

[0100] In this embodiment, the solving methods of this problem optimization model include various methods such as genetic algorithms, reinforcement learning, and deep learning, which are not limited here.

[0101] In a possible implementation manner, constructing a problem optimization model based on the task level information of the to-be-processed task and the task processing capacity information of the intelligent terminal and the at least one edge computing server includes:

[0102] The objective function for constructing the problem optimization model is:

[0103]

[0104] where S is the total processing overhead, ω ∈ [0, 1], N 1 is the total number of subtasks corresponding to the to-be-processed task, N 2 is the total number of the intelligent terminal and edge computing servers, j = 1 indicates that the subtask is assigned to the intelligent terminal, d i is the task data volume of the i-th subtask a i in the to-be-processed task, x i,j ∈ {0, 1} is the indication factor to be solved, x i,j = 0 indicates that the subtask a i is not offloaded to the j-th edge computing server, x i,j = 1 indicates that the subtask a i is offloaded to the j-th edge computing server, b i,j is the bandwidth ratio to be solved when the intelligent terminal offloads the subtask a i to the j-th edge computing server, B is the known total transmission channel bandwidth of the intelligent terminal, which belongs to the communication resource information of the intelligent terminal, p i,j is the one to be solved when the intelligent terminal offloads the subtask a iThe transmission power allocated when offloading to the j-th edge computing server, where G is the known transmission channel condition status of the intelligent terminal, such as channel power gain, etc., which belongs to the communication resource information of the intelligent terminal; The transmission channel noise power of the known intelligent terminal, which belongs to the communication resource information of the intelligent terminal, I 1′ The interference from other intelligent terminals in the same frequency band, which belongs to the communication resource information of the intelligent terminal, The processing delay for the j-th edge computing server to process the subtask a according to the predetermined task scheduling and processing rules, where ε is a predetermined constant parameter, c i The processing delay for the j-th edge computing server to process the subtask a according to the predetermined task scheduling and processing rules, where ε is a predetermined constant parameter, c i The processing workload for processing the subtask a i The processing workload for processing the subtask a i,j The computing resource to be solved for the j-th edge computing server to allocate to the subtask a i The computing resource to be solved for the j-th edge computing server to allocate to the subtask a

[0105] The constraint conditions of the problem optimization model include:

[0106]

[0107] Among them, P is the maximum transmission power of the intelligent terminal, which belongs to the communication resource information of the intelligent terminal, R j The total computing resources that can be allocated by the j-th edge computing server, which belongs to the information of the total computing resources that can be allocated reported by the edge computing server.

[0108] In this embodiment, the task A to be processed is split into N 1 subtasks There is 1 intelligent terminal and N 2 -1 edge computing servers on the edge side, which can be expressed as Among them, e 1 represents the intelligent terminal, is N 2 -1 edge computing servers.

[0109] In this embodiment, to improve the resource utilization rate and task offloading efficiency of the edge devices, the total processing overhead S considered in this embodiment includes the total delay T and total energy consumption EG for processing the task A to be processed. Among them, the total delay includes the transmission delay for the smart terminal to offload the subtask to the corresponding edge computing server and the computing delay for the corresponding edge computing server to process the subtask; the total energy consumption includes the total energy consumption of each subtask on the corresponding edge computing server. It should be noted here that since the end-side device performs resource collaboration and task offloading only after uploading the task to be processed to the smart terminal, the transmission time generated during this period is an established fact and does not need to be considered; since the amount of data of the calculation result is usually very small, the transmission delay and energy consumption caused by the corresponding edge computing server returning the calculation result to the smart terminal will be ignored.

[0110] In this embodiment, for the i-th subtask a of the task A to be processed i , v i,j is the transmission rate at which the smart terminal offloads the subtask a i to the j-th edge computing server, is the transmission delay at which the smart terminal offloads the subtask a i to the j-th edge computing server, is the processing delay for the j-th edge computing server to process and complete the subtask a i in accordance with the predetermined task scheduling and processing rules. Thus, the total delay for completing the task A

[0111] In this embodiment, EG i,j is the total energy consumption of the subtask a i on the j-th edge computing server. Then, the total energy consumption for processing the task A to be processed

[0112] In this embodiment, the total processing overhead S can be obtained by taking the weighted sum of the total delay T and the total energy consumption EG, i.e., S = ωT + (1 - ω)EG.

[0113] In this embodiment, the objective of the present disclosure is to formulate appropriate offloading strategies for all subtasks and determine the computing and bandwidth resources allocated to each subtask so that the overhead generated by all subtasks to complete the computing task is as small as possible. Therefore, the objective function of the problem optimization model is designed as follows:

[0114]

[0115] That is, solve for x i,j , b i,j , p i,j , r i,j , such that the total processing overhead S is minimized.

[0116] In the constraint conditions of the above problem optimization model, x i,j ∈{0, 1}; is the constraint on the indication factor x i,j , which restricts that a subtask can only be offloaded to one edge computing server, 0 ≤ b i,j ≤ 1; is the constraint on the bandwidth ratio b i,j ; 0 ≤ p i,j ≤ P; is the constraint on the transmission power p i,j , which restricts that the power of the transmission subtask shall not exceed the maximum power P; r i,j ≥ 0; is the constraint on the computing resource r i,j , which restricts that the computing resource for each edge computing server to execute the subtask shall not exceed the maximum computing resource R j .

[0117] In a possible implementation manner, the method further includes:

[0118] Determine the priority of the i-th subtask a in the to-be-processed task on the j-th edge computing server according to the task level information of the to-be-processed task and the priority range of the tasks that the j-th edge computing server can process; i

[0119] According to the priority of the subtask a on the j-th edge computing server, the j-th edge computing server processes the processing delay of the subtask a i in accordance with a predetermined task scheduling and processing rule; i

[0120] In this implementation manner, the task level information of the to-be-processed task can be mapped to the priority range of the tasks that the j-th edge computing server can process, so as to obtain the priority of the i-th subtask a in the to-be-processed task on the j-th edge computing server. In this way, the priorities of each subtask in the j-th edge computing server can be determined. i

[0121] In this implementation manner, a corresponding task scheduling and processing rule is preset in the edge computing server. For example, tasks with higher priorities are processed first, and for tasks with the same priority, the tasks received earlier are processed first according to the time when the tasks are received; or, it can also be that tasks with higher priorities are processed first, and for tasks with the same priority, a time slice rotation scheduling algorithm is used for scheduling, etc. For different task scheduling and processing rules, the processing delay i of the subtask a will also be correspondingly different, and it can be determined according to the subtask a i ​​​The priority of the j-th edge computing server, and the j-th edge computing server processes and completes the subtask a according to a predetermined task scheduling and processing rule i processing delay

[0122] In a possible implementation manner, determining the i-th subtask a in the to-be-processed task according to the task level information of the to-be-processed task and the priority range of the tasks that the j-th edge computing server can process i The priority of the j-th edge computing server includes:

[0123] Calculate the subtask a according to the following formula i The priority R of the j-th edge computing server i,j :

[0124]

[0125] where, round(·) refers to the rounding operation, λ represents the preference degree of the to-be-processed task for low latency, and the priority range of the j-th edge computing server is [P j d , P j u , R t is the latency requirement level of the to-be-processed task, R s is the security level of the to-be-processed task, R t The value range of is [R ta , R tb , R s The value range of is [R sa , R sb .

[0126] In this implementation manner, according to the example in Table 1, the latency requirement level of the to-be-processed task A is R t ∈[R ta , R tb = [1, 4], the security level is R s ∈[R sa , R sb = [1, 5], at this time, where, the smaller R i,j , the higher the priority, round(·) refers to the rounding operation, ensuring that the priority is a positive integer; λ represents the preference degree of the to-be-processed task for low latency, the closer λ is to 1, the higher the requirement of the task for latency, the closer λ is to 0, the higher the requirement of the task for security, and λ can be determined according to the type of the to-be-processed task A. It can be seen that the subtask a i Although it inherits the priority λR of the to-be-processed task At +(1 - λ)R s , but if subtask a i is offloaded to the j-th edge computing node, the priority range [P j d , P j u also affects its priority determination.

[0127] The described predetermined task scheduling processing rule includes preferentially executing subtasks with higher priorities. For subtasks with the same priority, a round-robin scheduling algorithm is used for scheduling; then, according to the priority of subtask a i on the j-th edge computing server, the j-th edge computing server processes and completes subtask a i according to the predetermined task scheduling processing rule, and the processing delay includes:

[0128] The processing delay is calculated according to the following formula

[0129]

[0130]

[0131] where is the remaining execution time of subtasks with higher priorities than subtask a in the j-th edge computing server, N i is the number of subtasks with the same priority as subtask a in the j-th edge computing server, 3 is the remaining execution time of N i subtasks with the same priority as subtask a, and θ is the time slice of the round-robin scheduling algorithm in the j-th edge computing node. 3 subtasks with the same priority as subtask a, and θ i is the time slice of the round-robin scheduling algorithm in the j-th edge computing node. j is the time slice of the round-robin scheduling algorithm in the j-th edge computing node.

[0132] In this embodiment, the working principle of the round-robin scheduling algorithm is that each subtask with the same priority is assigned a time slice θ of a fixed length j , and the scheduler sequentially selects subtasks from the ready queue. If the remaining execution time of the subtask does not exceed the time slice θ j , it is directly executed and completed. If the remaining time of the subtask exceeds one time slice, the subtask will be paused after the time slice ends, the computing resource will be allocated to the next subtask, the subtask will be re-queued and wait for the next scheduling, and the subtask will be removed from the queue after completion.

[0133] The present disclosure also provides a task offloading device for edge-edge resource collaboration,Figure 2 The structure block diagram of the task offloading device for edge-edge resource collaboration provided by an embodiment of the present disclosure is shown. This device can be implemented as part or all of an electronic device through software, hardware, or a combination of both. As Figure 2 shown, the task offloading device for edge-edge resource collaboration includes:

[0134] An information acquisition module 201, configured to acquire the task processing capability information of the smart terminal and at least one corresponding edge computing server;

[0135] A task acquisition module 202, configured to acquire the task to be processed and divide the task to be processed into multiple independent subtasks;

[0136] An allocation determination module 203, configured to determine the allocation information of each subtask based on the task level information of the task to be processed and the task processing capability information of the smart terminal and the at least one edge computing server, where the allocation information includes the smart terminal or edge computing server assigned to the subtask;

[0137] A task offloading module 204, configured to offload the subtasks assigned to the corresponding edge computing server to the corresponding edge computing server based on the allocation information of each subtask.

[0138] In a possible implementation manner, the task processing capability information includes the priority range of tasks that can be processed, the total computable resource information that can be allocated, and communication resource information;

[0139] The task level information includes a latency requirement level and a security level.

[0140] In a possible implementation manner, the allocation determination module includes:

[0141] Based on the task level information of the task to be processed and the task processing capability information of the smart terminal and the at least one edge computing server, construct a problem optimization model. The objective function corresponding to the problem optimization model is a function used to represent the minimization of the total processing overhead of all subtasks for processing the task to be processed. Solve the objective function to obtain the allocation information of each subtask when the total processing overhead is minimized;

[0142] Solve the problem optimization model to obtain the allocation information of each subtask.

[0143] In a possible implementation manner, the part in the allocation determination module that constructs a problem optimization model based on the task level information of the task to be processed and the task processing capability information of the smart terminal and the at least one edge computing server is configured to:

[0144] The objective function for constructing the problem optimization model is as follows:

[0145]

[0146] where S is the total processing cost, ω ∈ [0, 1], N 1 is the total number of subtasks corresponding to the task to be processed, N 2 is the total number of the intelligent terminals and edge computing servers, j = 1 indicates that the subtask is assigned to the intelligent terminal, d i is the task data volume of the i-th subtask a i in the task to be processed, x i,j ∈ {0, 1} is the indication factor to be solved, x i,j = 0 indicates that the subtask a i is not offloaded to the j-th edge computing server, x i,j = 1 indicates that the subtask a i is offloaded to the j-th edge computing server, b i,j is the bandwidth ratio allocated by the intelligent terminal when offloading the subtask a i to the j-th edge computing server, B is the known total transmission channel bandwidth of the intelligent terminal, p i,j is the transmission power to be solved when the intelligent terminal offloads the subtask a i to the j-th edge computing server, G is the known transmission channel condition status of the intelligent terminal, is the known transmission channel noise power of the intelligent terminal, I 1′ is the known interference from other intelligent terminals in the same frequency band, is the processing delay for the j-th edge computing server to process the subtask a i according to the predetermined task scheduling and processing rules, ε is a predetermined constant parameter, c i is the computing workload for processing the subtask a i r i,j is the computing resource allocated by the j-th edge computing server to the subtask a i ;

[0147] The constraint conditions of the problem optimization model include:

[0148]

[0149] where, refers to for all i, P is the maximum transmission power of the intelligent terminal, R j is the total computing resource that can be allocated by the j-th edge computing server.

[0150] In a possible implementation, the device further includes:

[0151] A priority determination module, configured to determine the priority of the i-th subtask a in the to-be-processed task according to the task level information of the to-be-processed task and the priority range of the tasks that can be processed by the j-th edge computing server; i Among the priorities of the j-th edge computing server;

[0152] A latency determination module, configured to determine the processing latency of the j-th edge computing server to complete the subtask a according to the priority of the subtask a i Among the priorities of the j-th edge computing server, according to a predetermined task scheduling and processing rule; i Of the processing;

[0153] In a possible implementation, the priority determination module is configured to:

[0154] Calculate the priority R of the subtask a i Among the j-th edge computing servers according to the following formula: i,j :

[0155]

[0156] Where round(·) refers to the rounding operation, λ represents the preference degree of the to-be-processed task for low latency, the priority range of the j-th edge computing server is [P j d , P j u , R t Is the latency requirement level of the to-be-processed task, R s Is the security level of the to-be-processed task, R t The value range of is [R ta , R tb , R s The value range of is [R sa , R sb .

[0157] In a possible implementation, the predetermined task scheduling and processing rule includes preferentially executing subtasks with higher priorities. For subtasks with the same priority, a time slice rotation scheduling algorithm is used for scheduling; then the latency determination module is configured to:

[0158] Calculate the processing latency according to the following formula

[0159]

[0160] Where, is the remaining execution time of the subtasks with higher priority than subtask a in the j-th edge computing server, N i is the remaining execution time of the subtasks with higher priority than subtask a in the j-th edge computing server, N 3 is the number of subtasks with the same priority as subtask a in the j-th edge computing server, i is the number of subtasks with the same priority as subtask a in the j-th edge computing server, is N 3 is the remaining execution time of N subtasks with the same priority as subtask a, θ i is the remaining execution time of N subtasks with the same priority as subtask a, θ j is the time slice of the round-robin scheduling algorithm in the j-th edge computing node.

[0161] The technical terms and technical features mentioned in the embodiments of this device are the same as or similar to those mentioned in the above method embodiments. For the explanations and descriptions of the technical terms and technical features involved in this device, reference can be made to the explanations and descriptions of the above method embodiments, which will not be elaborated here.

[0162] This disclosure also discloses an intelligent terminal, Figure 3 showing a structural block diagram of the intelligent terminal according to an embodiment of this disclosure.

[0163] As Figure 3 shown, the intelligent terminal 300 includes a memory 301 and a processor 302. Among them, the memory 301 is used to store one or more computer instructions, and the one or more computer instructions are executed by the processor 302 to implement the method according to the embodiments of this disclosure.

[0164] This disclosure also discloses a task offloading system for edge-edge resource collaboration, including an intelligent terminal and at least one edge computing server, where:

[0165] The intelligent terminal is configured to use the above-mentioned task offloading method for edge-edge resource collaboration to offload the subtasks assigned to the corresponding edge computing server to the corresponding edge computing server;

[0166] The edge computing server is configured to obtain the subtasks offloaded by the intelligent terminal, execute the subtasks, and return the execution results to the intelligent terminal.

[0167] Exemplarily, Figure 4 shows a schematic diagram of a task offloading system for edge-edge resource collaboration provided by an embodiment of this disclosure. As Figure 4 shown, the system includes an end-side device 401, an intelligent terminal 402, and two edge computing servers 403A and 403B.

[0168] As Figure 4As shown, the edge computing servers 403A and 403B can report their task processing capacity information. The edge device 401 can upload the tasks to be processed to the intelligent terminal 402. The intelligent terminal 402 can execute the above-mentioned edge-edge resource collaboration task offloading method, execute the task offloading policy, determine the intelligent terminal or edge computing server assigned to each subtask of the task to be processed, and then offload the subtask assigned to the edge computing server 403A to the edge computing server 403A, and offload the subtask assigned to the edge computing server 403B to the edge computing server 403B. The edge computing servers 403A and 403B process the tasks and return the processing results to the intelligent terminal 402. The intelligent terminal 402 integrates the processing results of each subtask of the task to be processed to obtain the overall processing result, and returns the overall processing result to the edge device 401.

[0169] Figure 5 The structural diagram of a computer system suitable for implementing the method of the embodiments of the present disclosure is shown.

[0170] As Figure 5 As shown, the computer system 500 includes a processing unit 501, which can execute various processes in the above embodiments according to the program stored in the read-only memory (ROM) 502 or the program loaded from the storage section 508 into the random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the computer system 500 are also stored. The processing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. The input / output (I / O) interface 505 is also connected to the bus 504.

[0171] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed, so that the computer program read from it can be installed into the storage section 508 as needed. Among them, the processing unit 501 can be implemented as a processing unit such as a CPU, a GPU, a TPU, an FPGA, an NPU, etc.

[0172] In particular, according to an embodiment of the present disclosure, the method described above can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product that includes computer instructions which, when executed by a processor, implement the method steps described above. In such an embodiment, the computer program product can be downloaded and installed from a network via the communication section 509, and / or installed from the removable medium 511.

[0173] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that, in some alternative implementations, the functions noted in the blocks may occur in a different order than that noted in the drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0174] The units or modules described in the embodiments of the present disclosure can be implemented in software or in programmable hardware. The described units or modules can also be provided in a processor, and the names of these units or modules do not, in some cases, constitute a limitation on the units or modules themselves.

[0175] On the other hand, the present disclosure also provides a computer-readable storage medium, which can be the computer-readable storage medium included in the electronic device or computer system in the above embodiments; or can be a computer-readable storage medium that exists separately and is not assembled into the device. The computer-readable storage medium stores one or more programs, and the one or more programs are used by one or more processors to execute the method described in the present disclosure.

[0176] The above description is only a preferred embodiment of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, technical solutions formed by mutually replacing the above features with technical features having similar functions (but not limited to) disclosed in the present disclosure.

Claims

1. A task offloading method with edge resource collaboration, characterized in that: Applied to smart terminals, including: Obtaining task processing capability information of the smart terminal and at least one edge computing server corresponding to the smart terminal; Obtaining a task to be processed, and dividing the task to be processed into multiple independent subtasks; Determine allocation information of each subtask based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, wherein the allocation information includes the intelligent terminal or edge computing server allocated to the subtask; Based on the allocation information of each subtask, the subtask allocated to the corresponding edge computing server is unloaded to the corresponding edge computing server.

2. The method according to claim 1, characterized in that The task processing capability information includes the priority range of processable tasks, the total computing resource information that can be allocated, and the communication resource information; The task level information includes a delay requirement level and a safety level.

3. The method according to claim 1 or 2, characterized in that: The determining, based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, allocation information of each subtask comprises: Based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, a problem optimization model is constructed, wherein the objective function corresponding to the problem optimization model is a function for minimizing the total processing overhead of all subtasks of the task to be processed, and the allocation information of each subtask is obtained by solving the objective function to minimize the total processing overhead; Solve the problem optimization model to obtain the allocation information of each subtask.

4. The method according to claim 3, characterized in that The constructing a problem optimization model based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server includes: The objective function of constructing the optimization model for the problem is: Where S is the total processing overhead, ω∈[0,1], N1 is the total number of subtasks corresponding to the task to be processed, N2 is the total number of smart terminals and edge computing servers, j=1 means that the subtask is assigned to the smart terminal, d i is the i-th subtask a in the task to be processed i The amount of task data, x i,j ∈{0,1} is the indicator factor to be solved, x i,j =0 indicates subtask a i Not offloaded to the jth edge computing server, x i,j =1 indicates subtask a i Offloaded to the jth edge computing server, b i,j For the intelligent terminal that needs to be solved, assign subtask a i The bandwidth ratio allocated when offloading to the jth edge computing server, B is the known total transmission channel bandwidth of the smart terminal, p i,j For the intelligent terminal that needs to be solved, subtask a i The transmission power allocated when unloading to the jth edge computing server, G is the known transmission channel condition state of the smart terminal, is the known transmission channel noise power of the intelligent terminal, I 1′ is the known interference from other intelligent terminals in the same frequency band, The jth edge computing server completes the subtask a according to the predetermined task scheduling processing rules. i processing delay, ε is a predetermined constant parameter, c i Is to process the subtask a i The computational workload, r i,j Give subtask a to the jth edge computing server that needs to be solved i Allocated computing resources; The constraints of the problem optimization model include: in, It means that for all i, P is the maximum transmission power of the intelligent terminal, R j is the total computing resources that can be allocated to the jth edge computing server.

5. The method according to claim 4, characterized in that The method further comprises: According to the task level information of the task to be processed and the priority range of the tasks that can be processed by the jth edge computing server, determine the i-th subtask a in the task to be processed i The priority of the j-th edge computing server; According to subtask a i At the priority of the jth edge computing server, determine that the jth edge computing server completes the subtask a according to the predetermined task scheduling processing rules i Processing delay 6. The method according to claim 5, characterized in that According to the task level information of the task to be processed and the priority range of the tasks that can be processed by the jth edge computing server, determine the i-th subtask a in the task to be processed i The priority of the j-th edge computing server includes: The subtask a is calculated according to the following formula i The priority R of the jth edge computing server i,j : Among them, round(·) refers to the rounding operation, λ represents the preference of the task to be processed for low latency, and the priority range of the jth edge computing server is R t is the latency requirement level of the task to be processed, R s is the security level of the task to be processed, R t The value range is [R ta ,R tb ], R s The value range is [R sa ,R sb ].

7. The method according to claim 5, characterized in that The predetermined task scheduling processing rules include giving priority to executing subtasks with higher priorities, and using a time slice round-robin scheduling algorithm to schedule subtasks with the same priority; i At the priority of the jth edge computing server, determine that the jth edge computing server completes the subtask a according to the predetermined task scheduling processing rules i Processing delay include: The processing delay is calculated according to the following formula: in, is the number of subtasks a in the jth edge computing server i The remaining execution time of subtasks with higher priority, N3 and subtask a i The remaining execution time of subtasks with the same priority, N3 is the remaining execution time of subtask a in the jth edge computing server i The number of subtasks with the same priority, θ j is the time slice of the round-robin scheduling algorithm in the jth edge computing node.

8. A task offloading device with edge-to-edge resource collaboration, characterized in that: Applied to smart terminals, including: An information acquisition module, configured to acquire task processing capability information of the smart terminal and at least one edge computing server corresponding to the smart terminal; A task acquisition module is configured to acquire tasks to be processed and divide the tasks to be processed into multiple independent subtasks; an allocation determination module, configured to determine allocation information of each subtask based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, the allocation information including the intelligent terminal or edge computing server allocated to the subtask; The task offloading module is configured to offload the subtasks assigned to the corresponding edge computing server to the corresponding edge computing server based on the assignment information of each subtask.

9. The device according to claim 8, characterized in that The task processing capability information includes the priority range of processable tasks, the total computing resource information that can be allocated, and the communication resource information; The task level information includes a delay requirement level and a safety level.

10. The device according to claim 8 or 9, characterized in that The allocation determination module comprises: Based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server, a problem optimization model is constructed, wherein the objective function corresponding to the problem optimization model is a function for minimizing the total processing overhead of all subtasks of the task to be processed, and the allocation information of each subtask is obtained by solving the objective function to minimize the total processing overhead; Solve the problem optimization model to obtain the allocation information of each subtask.

11. The device according to claim 10, characterized in that The part of constructing the problem optimization model based on the task level information of the task to be processed and the task processing capability information of the intelligent terminal and the at least one edge computing server is configured as follows: The objective function of constructing the optimization model for the problem is: Where S is the total processing overhead, ω∈[0,1], N1 is the total number of subtasks corresponding to the task to be processed, N2 is the total number of smart terminals and edge computing servers, j=1 means that the subtask is assigned to the smart terminal, d i is the i-th subtask a in the task to be processed i The amount of task data, x i,j ∈{0,1} is the indicator factor to be solved, x i,j =0 indicates subtask a i Not offloaded to the jth edge computing server, x i,j =1 indicates subtask a i Offloaded to the jth edge computing server, b i,j For the intelligent terminal that needs to be solved, assign subtask a i The bandwidth ratio allocated when offloading to the jth edge computing server, B is the known total transmission channel bandwidth of the smart terminal, p i,j For the intelligent terminal that needs to be solved, subtask a i The transmission power allocated when unloading to the jth edge computing server, G is the known transmission channel condition state of the smart terminal, is the known transmission channel noise power of the intelligent terminal, I 1′ is the known interference from other intelligent terminals in the same frequency band, The jth edge computing server completes the subtask a according to the predetermined task scheduling processing rules. i processing delay, ε is a predetermined constant parameter, c i Is to process the subtask a i The computational workload, r i,j Give subtask a to the jth edge computing server that needs to be solved i Allocated computing resources; The constraints of the problem optimization model include: in, It means that for all i, P is the maximum transmission power of the intelligent terminal, R j is the total computing resources that can be allocated to the jth edge computing server.

12. The device according to claim 11, characterized in that The device also includes: The priority determination module is configured to determine the i-th subtask a in the task to be processed according to the task level information of the task to be processed and the priority range of the task that can be processed by the j-th edge computing server. i The priority of the j-th edge computing server; The delay determination module is configured to determine the time delay according to the subtask a i At the priority of the jth edge computing server, determine that the jth edge computing server completes the subtask a according to the predetermined task scheduling processing rules i Processing delay 13. The device according to claim 12, characterized in that The priority determination module is configured to: The subtask a is calculated according to the following formula i The priority R of the jth edge computing server i,j : Among them, round(·) refers to the rounding operation, λ represents the preference of the task to be processed for low latency, and the priority range of the jth edge computing server is R t is the latency requirement level of the task to be processed, R s is the security level of the task to be processed, R t The value range is [R ta ,R tb ], R s The value range is [R sa ,R sb ].

14. The device according to claim 12, characterized in that The predetermined task scheduling processing rule includes giving priority to executing subtasks with higher priorities, and scheduling subtasks with the same priority using a time slice round-robin scheduling algorithm; the delay determination module is configured as follows: The processing delay is calculated according to the following formula: in, is the number of subtasks a in the jth edge computing server i The remaining execution time of subtasks with higher priority, N3 and subtask a i The remaining execution time of subtasks with the same priority, N3 is the remaining execution time of subtask a in the jth edge computing server i The number of subtasks with the same priority, θ j is the time slice of the round-robin scheduling algorithm in the jth edge computing node.

15. An intelligent terminal, characterized in that: The method comprises a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method according to any one of claims 1 to 7.

16. A task offloading system with edge resource collaboration, characterized in that: It includes a smart terminal and at least one edge computing server, wherein: The smart terminal is configured to adopt the method described in any one of claims 1 to 7 to offload the subtask assigned to the corresponding edge computing server to the corresponding edge computing server; The edge computing server is configured to obtain the subtask uninstalled by the smart terminal, execute the subtask, and return the execution result to the smart terminal.

17. A readable storage medium, characterized in that: Computer instructions are stored thereon, and when the computer instructions are executed by a processor, the method described in any one of claims 1 to 7 is implemented.

Images