Industrial wireless network edge computing task unloading method, device, equipment, medium and product

Abstract

The invention discloses an industrial wireless network edge computing task unloading method and device, equipment, a medium and a product, and relates to the field of industrial internet and edge computing, and the method comprises the steps: constructing an industrial wireless network model; constructing a preference list of the to-be-unloaded tasks according to the preferences of the to-be-unloaded tasks on the matching objects; constructing a preference list of the communication path based on the preference of the communication path for each to-be-unloaded task; constructing a preference list of the edge server based on the preference of the edge server for each to-be-unloaded task; processing the preference lists of the to-be-unloaded tasks, the communication paths and the edge servers by adopting a distributed delay acceptance matching algorithm to obtain optimal matching objects of the to-be-unloaded tasks; and performing edge computing task unloading on each to-be-unloaded task based on the optimal matching object of each to-be-unloaded task. The method can adapt to the dynamic nature of an industrial wireless network, realizes task-path-server three-dimensional resource optimal matching, and ensures the real-time performance and reliability of the tasks.

Description

Technical Field

[0001] This application relates to the fields of industrial internet and edge computing, and in particular to a method, apparatus, device, medium and product for offloading edge computing tasks in an industrial wireless network. Background Technology

[0002] In smart factories, applications such as interactive remote operation of equipment, target recognition and tracking, and real-time fault diagnosis place extremely high demands on computing power and latency. However, terminal devices in industrial wireless networks typically have limited computing power, and wireless communication in the operating environment presents challenges related to latency and unpredictable reliability.

[0003] Edge computing is an effective way to solve the above problems by offloading computing tasks from the terminal to the network edge server. However, most existing task offloading methods have the following shortcomings: (1) Scenario simplification: Most studies are based on simple single-hop networks or "terminal-edge" binary models, which fail to fully consider the complex network topology of multi-hop and multi-edge servers common in industrial environments, and cannot guarantee the real-time performance and reliability of tasks.

[0004] (2) Single decision dimension: Existing methods usually only consider one of the computing resources or communication resources, or perform simple joint optimization, which cannot achieve optimal matching of three-dimensional resources of task, link and edge server.

[0005] (3) Poor dynamic adaptability: Centralized optimization algorithms have high computational overhead and slow response speed in scenarios where tasks arrive randomly and network states change over time; while distributed heuristic algorithms are prone to getting stuck in local optima, leading to edge hotspot problems, i.e., some servers or communication links are overloaded while other resources are idle, resulting in unbalanced resource utilization and task timeouts, which cannot adapt to the dynamic characteristics of industrial wireless networks.

[0006] In summary, there is an urgent need for an industrial wireless network edge computing task offloading method that can adapt to the dynamic nature of industrial wireless networks, achieve optimal matching of three-dimensional resources (tasks, links, and edge servers), and ensure the real-time performance and reliability of tasks. Summary of the Invention

[0007] The purpose of this application is to provide a method, apparatus, device, medium and product for offloading edge computing tasks in industrial wireless networks, which can adapt to the dynamic nature of industrial wireless networks, achieve optimal matching of three-dimensional resources of task-link-server, and ensure the real-time performance and reliability of tasks.

[0008] To achieve the above objectives, this application provides the following solution: In a first aspect, this application provides a method for offloading edge computing tasks in an industrial wireless network, comprising: constructing an industrial wireless network model based on the topology among terminal devices, routing nodes, cloud computing centers, and multiple edge servers within the target industrial wireless network.

[0009] For any task to be unloaded, a preference list for the task to be unloaded is constructed based on the task's preference for each matching object; the matching object is a combination scheme of communication path and edge server constructed based on the industrial wireless network model.

[0010] For any given communication path, a preference list for the communication path is constructed based on the preferences of the communication path for each task to be uninstalled; the preferences of the communication path for the tasks to be uninstalled are determined based on the task parameters of the tasks to be uninstalled.

[0011] For any given edge server, a preference list is constructed based on the edge server's preferences for each task to be uninstalled; the edge server's preferences for the tasks to be uninstalled are determined based on the task parameters of the tasks to be uninstalled.

[0012] A distributed delayed acceptance matching algorithm is adopted to perform three-dimensional matching based on the preference lists of each task to be uninstalled, the preference lists of each communication path, and the preference lists of each edge server, so as to obtain the optimal matching object for each task to be uninstalled.

[0013] Edge computing task uninstallation is performed on each task based on the optimal matching object for each task to be uninstalled.

[0014] Secondly, this application provides an industrial wireless network edge computing task offloading device, including: an industrial wireless network model construction module, used to construct an industrial wireless network model based on the topology between terminal devices, routing nodes, cloud computing centers and multiple edge servers within a target industrial wireless network.

[0015] The preference list construction module is used to construct a preference list for any task to be unloaded based on the task's preferences for each matching object; the matching object is a combination scheme of communication path and edge server constructed based on the industrial wireless network model; for any communication path, a preference list for the communication path is constructed based on the communication path's preferences for each task to be unloaded; the preferences of the communication path for the task to be unloaded are determined based on the task parameters of the task to be unloaded; for any edge server, a preference list for the edge server is constructed based on the edge server's preferences for each task to be unloaded; the preferences of the edge server for the task to be unloaded are determined based on the task parameters of the task to be unloaded.

[0016] The optimal matching object determination module is used to perform three-dimensional matching based on the preference lists of each task to be uninstalled, the preference lists of each communication path, and the preference lists of each edge server using a distributed delayed acceptance matching algorithm to obtain the optimal matching object for each task to be uninstalled.

[0017] The task unloading module is used to perform edge computing task unloading on each task based on the optimal matching object for each task to be unloaded.

[0018] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described industrial wireless network edge computing task offloading method.

[0019] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described industrial wireless network edge computing task offloading method.

[0020] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the above-described industrial wireless network edge computing task offloading method.

[0021] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method, apparatus, device, medium and product for offloading edge computing tasks in an industrial wireless network. This application constructs an industrial wireless network model based on the topology between terminal devices, routing nodes, cloud computing centers and multiple edge servers; it fully considers the complex network topology of multi-hop and multi-edge servers commonly found in industrial environments, ensuring the real-time performance and reliability of tasks.

[0022] Existing technologies fail to coordinate the three elements of "task", "communication path" and "edge server" as a whole for collaborative matching. This application adopts a distributed delayed acceptance matching algorithm to perform three-dimensional matching based on the preference lists of each task to be unloaded, the preference lists of each communication path and the preference lists of each edge server, to obtain the optimal matching object for each task to be unloaded. This achieves optimal matching of the three-dimensional resources of task, communication path and edge server. Furthermore, the distributed delayed acceptance matching algorithm can adapt to the dynamic nature of industrial wireless networks. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating an industrial wireless network edge computing task offloading method provided in an embodiment of this application.

[0025] Figure 2 A flowchart of a distributed delayed reception matching algorithm provided in an embodiment of this application.

[0026] Figure 3 This is a system architecture diagram of a target industrial wireless network provided in an embodiment of this application.

[0027] Figure 4 This is a schematic diagram of the functional modules of an industrial wireless network edge computing task offloading device provided in an embodiment of this application.

[0028] Figure 5 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0029] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0030] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] In one exemplary embodiment, such as Figure 1 As shown, a method for offloading edge computing tasks in an industrial wireless network is provided, including: Step 201: Constructing an industrial wireless network model based on the topology between terminal devices, routing nodes, cloud computing centers, and multiple edge servers within the target industrial wireless network. The industrial wireless network model includes terminal devices, routing nodes, multiple edge servers, and a cloud computing center. Figure 3 The connection relationships between terminal devices (industrial sensors, smart cameras, robotic arm controllers, and AGVs), routing nodes, multiple edge servers, and cloud computing centers are shown, and the data flow of task offloading is marked.

[0032] Step 202: For any task to be unloaded, construct a preference list for the task to be unloaded based on its preference for each matching object; the matching object is a combination scheme of communication path and edge server constructed based on the industrial wireless network model. Specifically, the matching object is represented as a "communication path P - edge server ES" tuple as (P, ES).

[0033] Step 203: For any communication path, construct a preference list for the communication path based on the preferences of the communication path for each task to be uninstalled; the preferences of the communication path for the tasks to be uninstalled are determined based on the task parameters of the tasks to be uninstalled.

[0034] Step 204: For any edge server, construct a preference list for the edge server based on the edge server's preferences for each task to be uninstalled; the edge server's preferences for the tasks to be uninstalled are determined based on the task parameters of the tasks to be uninstalled.

[0035] Step 205: Using a distributed delayed acceptance matching algorithm, a three-dimensional matching is performed based on the preference lists of each task to be uninstalled, the preference lists of each communication path, and the preference lists of each edge server to obtain the optimal matching object for each task to be uninstalled.

[0036] Step 206: Perform edge computing task uninstallation on each task based on the optimal matching object of each task to be uninstalled.

[0037] In another exemplary embodiment of this application, the calculation process for the preference of the task to be unloaded for each matching object is as follows: multiple communication paths are constructed based on the industrial wireless network model.

[0038] For any task to be unloaded, multiple matching objects corresponding to the task to be unloaded are constructed based on the industrial wireless network model and each communication path.

[0039] Based on the transmission reliability and latency of each communication path, the unloading cost of the task to be unloaded, and the waiting latency and computation latency of each edge server, the utility of the task to be unloaded in selecting each matching object is calculated, and the preference of the task to be unloaded for each matching object is obtained (utility is preference).

[0040] In practical applications, according to the formula U_task=α (1 / T_total)+β Reliability(P)-γ Cost calculates the utility U_task for each matching object (P, ES), where T_total = T_trans(P) + T_queue(ES) + T_comp(ES) is the estimated total latency, Reliability(P) is the transmission reliability of communication path P, Cost is the unloading cost of the task to be unloaded, and α, β, and γ are weighting coefficients. T_total represents the estimated total completion latency of the task, which is the total time from the start of task transmission to completion of execution on the edge server. T_trans(P) represents the transmission latency of task data transmitted via communication path P, i.e., the transmission latency of communication path P, which depends on path bandwidth, data volume, and path hop count. T_queue(ES) represents the waiting latency of the task in the computation queue of edge server ES, i.e., the waiting latency of edge server ES. Its calculation depends on the current server load and task scheduling strategy. T_comp(ES) represents the computation latency required for the task to be actually executed on edge server ES, i.e., the computation latency of edge server ES, which depends on the number of CPU cycles required by the task and the server's computing power.

[0041] Generate a preference list based on U_task sorted in descending order.

[0042] In another exemplary embodiment of this application, the calculation process for the preference of the communication path for each task to be unloaded is as follows: obtain the task parameters of each task to be unloaded and the available bandwidth and load status of each communication path.

[0043] For any given communication path, the preference of the communication path for each task to be unloaded is obtained based on the available bandwidth and load status of the communication path and the task parameters of each task to be unloaded.

[0044] In another exemplary embodiment of this application, the calculation process of the edge server's preference for each task to be unloaded is as follows: obtain the task parameters of each task to be unloaded, as well as the computing load and remaining energy of each edge server.

[0045] For any edge server, the preference of the communication path for each task to be unloaded is obtained based on the edge server's computing load and remaining energy, as well as the task parameters of each task to be unloaded.

[0046] In practical applications, after constructing the industrial wireless network model, the following steps are also included: Creating a model Task for each dynamically arriving task (task to be unloaded), which can be represented by a four-tuple: Task = {Data_Size, CPU_Cycles, D_max, Priority}, where Data_Size is the task data size, CPU_Cycles is the number of CPU cycles required by the task, D_max is the maximum tolerable latency of the task, and Priority is the task priority. Data_Size, CPU_Cycles, D_max, and Priority are all task parameters.

[0047] In practical applications, each communication path P evaluates task preferences based on its current available bandwidth, load status, and the rewards that the task can provide (such as Priority and Data_Size), and generates a preference list, with tasks with higher preference values ​​ranked first.

[0048] For a communication path P, its preference U_path(P,Task_i) for Task_i can be defined as: U_path(P,Task_i)=δ1 (Priority_i / Data_Size_i)-δ2 Load_P+δ3 Available_Bandwidth_P. Here, Priority_i represents the task priority of Task_i, Data_Size_i represents the data volume of Task_i, Priority_i / Data_Size_i represents the benefit of a unit of data transmission (tasks with higher priority and smaller data volumes are more popular), Load_P is the current load of communication path P (e.g., the percentage of bandwidth already used), δ1, δ2, and δ3 are weighting coefficients, and Available_Bandwidth_P is the current available bandwidth of communication path P. The larger the available bandwidth, the higher the path's preference value for that task. Communication path P sorts tasks in descending order according to U_path(P,Task_i) to generate a preference list.

[0049] In practical applications, each edge server (ES) evaluates task preferences based on its current computing load, remaining energy, and the urgency of the tasks (such as CPU_Cycles and D_max), and generates a preference list, with tasks with higher preference values ​​ranked first.

[0050] For the edge server ES, its preference for task_i, U_server(ES,Task_i), can be defined as: Here, CPU_Cycles_i represents the number of CPU cycles required for Task_i, D_max_i represents the maximum tolerable latency for Task_i, CPU_Cycles_i / D_max_i represents the urgency of the task (tasks with high computational demands and tight deadlines may have higher rewards or need to be prioritized), Load_ES is the current computational load of the edge server ES, Energy_Remaining_ES represents the percentage of remaining energy of the edge server ES; the more remaining energy, the stronger the server's sustainability in executing tasks, Energy_Cost is the estimated energy consumption for executing the task, and θ1, θ2, and θ3 are all weighting coefficients. The server ES sorts the tasks in descending order according to U_server(ES,Task_i) to generate a preference list.

[0051] The distributed delayed acceptance matching algorithm is a multi-round iterative proposal-accept or rejection process, as follows: Figure 2 As shown, the algorithm begins by initializing all tasks as "unmatched". In each iteration, each unmatched task makes a request to the current most preferred path-server pair. Each path-server pair temporarily accepts the most preferred request and rejects other requests based on its own preference list. The rejected task is removed from its preference list. It is then determined whether there are still unmatched tasks and whether the preference list is not empty. If so, the next iteration is performed; otherwise, the algorithm ends and the optimal matching result is output. In another exemplary embodiment of this application, a distributed delayed acceptance matching algorithm is used to perform three-dimensional matching based on the preference lists of each unloaded task, the preference lists of each communication path, and the preference lists of each edge server to obtain the optimal matching object for each unloaded task. Specifically, in the current iteration number, for any unassigned unloaded task in the current iteration number, the unloaded task sends a request to the most preferred matching object in the preference list of the unloaded task in the current iteration number.

[0052] The communication path in the most preferred matching object determines whether the task to be uninstalled is the most preferred in the preference list under the current iteration number of the communication path to obtain a first judgment result, and the edge server in the most preferred matching object determines whether the task to be uninstalled is the most preferred in the preference list under the current iteration number of the edge server to obtain a second judgment result.

[0053] If either the first judgment result or the second judgment result is negative, then the most preferred matching object in the preference list of the task to be uninstalled at the current iteration number is deleted from the preference list of the task to be uninstalled at the current iteration number, and the preference list of the task to be uninstalled at the next iteration number is obtained.

[0054] If both the first and second judgment results are yes, then the most preferred matching object in the preference list of the task to be uninstalled under the current iteration number is determined as the optimal matching object for the task to be uninstalled, and the task to be uninstalled is deleted from the unassigned tasks under the current iteration number to obtain the unassigned tasks under the next iteration number.

[0055] Update the iteration count and proceed to the next iteration. The iteration continues until all unassigned tasks have been assigned or the preference list for the next iteration of all unassigned tasks is empty, at which point the iteration stops.

[0056] In practical applications, task data is transmitted to the optimal edge server for execution along the optimal communication path.

[0057] In another exemplary embodiment of this application, the process of unloading each task from the edge computing task based on the optimal matching object of each task to be unloaded also includes: a dynamic re-matching mechanism, specifically: real-time monitoring of the status of each communication path and the status of each edge server.

[0058] If the status of any communication path is abnormal, the preference list of the communication path is updated based on the preferences of each task to be uninstalled.

[0059] The preference list of each first target task to be uninstalled is updated according to the preference of each matching object for each first target task to be uninstalled; the optimal matching object for the first target task to be uninstalled includes the communication path.

[0060] A distributed delayed acceptance matching algorithm is used to process the updated preference lists of each first target unloaded task, the updated preference list of the communication path, and the preference lists of each first target edge server to obtain the optimal matching object for each first target unloaded task; the optimal matching object for each first target unloaded task includes the first target edge server.

[0061] If any edge server is in an abnormal state, the edge server's preference list is updated based on the edge server's preference for each task to be uninstalled.

[0062] The preference list of each second target task to be uninstalled is updated according to the preference of each matching object for each second target task to be uninstalled; the optimal matching object for the second target task to be uninstalled includes the edge server.

[0063] A distributed delayed acceptance matching algorithm is used to process the updated preference lists of each second target unloading task, the updated preference list of the edge server, and the preference lists of each second communication path to obtain the optimal matching object for each second target unloading task; the optimal matching object for the second target unloading task includes the second communication path.

[0064] In practical applications, the status of the communication path can be the link quality (link packet loss rate), and the status of the edge server can be the server load. When the detected status change exceeds the preset threshold, it is judged as an abnormal status (e.g., if the server load is >90%, it is judged as an abnormal edge server status, and if the link packet loss rate is >20%, it is judged as an abnormal communication path status).

[0065] The system continuously monitors the network status (the status of communication paths and edge servers). When the detected status change exceeds a preset threshold, a local re-matching mechanism is triggered. Only the affected tasks and resources are re-executed with the preference list generation and distributed delayed acceptance matching algorithm. When the network status changes significantly, the system is locally optimized to ensure that the system remains in a better state.

[0066] This application models the edge computing task offloading problem in industrial wireless networks as a three-way matching game problem involving "task-communication path-edge server". In cloud-edge-device collaborative industrial wireless networks, efficient and reliable task offloading is achieved through multi-dimensional dynamic matching.

[0067] This application integrates the three dimensions of task, communication path, and edge server into a matching game framework, achieving deep collaboration and joint optimization of communication and computing resources, improving resource utilization efficiency at the system level, and realizing three-dimensional collaborative optimization.

[0068] The distributed delayed reception matching algorithm proposed in this application does not require global optimization by a central controller, has a fast response speed, low computational overhead, and can well adapt to the dynamic characteristics of random arrival of tasks and time-varying network status in industrial wireless networks, thus achieving high dynamic adaptability.

[0069] The distributed delayed acceptance matching algorithm of this application promotes the matching results to tend to be stable and balanced, avoiding overload of some links or servers. At the same time, through the transmission reliability of the path and the dynamic rematching mechanism, it significantly improves the reliability and real-time performance of the task's successful execution, and achieves balance and reliability assurance.

[0070] This application also provides a simulation setup to verify the effectiveness of the industrial wireless network edge computing task offloading method provided in the above embodiments. Specifically, the network topology is as follows: 50 terminal devices and 4 edge servers are deployed in a 500m×500m factory area. The terminal devices are connected to the edge servers through a 3-hop to 5-hop wireless mesh network.

[0071] Task model: Tasks arrive randomly and follow a Poisson process. Priority ∈ [1,5], with higher values ​​indicating higher priority; Data_Size ∈ [0.5,5] MB; CPU_Cycles ∈ [0.5,3] Giga Cycles; D_max ∈ [50,300] ms.

[0072] Communication model: The wireless channel bandwidth is 20MHz, and the path loss model adopts the COST-231Hata model, considering Rayleigh fading. The link transmission rate and bit error rate are dynamically calculated based on the signal-to-noise ratio.

[0073] Computational Model: The edge server has a computing power of [5,15] GHz and uses an M / M / 1 queuing model to simulate the task processing queue.

[0074] Comparison Algorithm: Greedy Algorithm: The task always selects the available server with the strongest computing power and transmits through the shortest path.

[0075] Random Unloading Algorithm: The task randomly selects an available edge server and path.

[0076] Simulation Results and Analysis: Task Success Rate: As network load increases, the task success rate (the proportion of tasks completed within the deadline) decreases for all methods. However, the method proposed in this application maintains the highest success rate under different loads because it comprehensively considers latency and reliability in its decision-making.

[0077] Load balancing: The standard deviation of server load in this application is much lower than that of the greedy algorithm and the random offload algorithm. The greedy algorithm causes servers with high computing power to become overloaded quickly, while this application naturally achieves a balanced distribution of load among different servers through matching game theory, thereby improving system stability.

[0078] Based on the same inventive concept, this application also provides an industrial wireless network edge computing task offloading device for implementing the above-mentioned industrial wireless network edge computing task offloading method. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations of one or more embodiments of the industrial wireless network edge computing task offloading device provided below can be found in the limitations of the industrial wireless network edge computing task offloading method described above, and will not be repeated here.

[0079] In one exemplary embodiment, such as Figure 4 As shown, an industrial wireless network edge computing task offloading device is provided, including: an industrial wireless network model construction module, used to construct an industrial wireless network model based on the topology between terminal devices, routing nodes, cloud computing centers and multiple edge servers within the target industrial wireless network.

[0080] The preference list construction module is used to construct a preference list for any task to be unloaded based on the task's preferences for each matching object; the matching object is a combination scheme of communication path and edge server constructed based on the industrial wireless network model; for any communication path, a preference list for the communication path is constructed based on the communication path's preferences for each task to be unloaded; the preferences of the communication path for the task to be unloaded are determined based on the task parameters of the task to be unloaded; for any edge server, a preference list for the edge server is constructed based on the edge server's preferences for each task to be unloaded; the preferences of the edge server for the task to be unloaded are determined based on the task parameters of the task to be unloaded.

[0081] The optimal matching object determination module is used to perform three-dimensional matching based on the preference lists of each task to be uninstalled, the preference lists of each communication path, and the preference lists of each edge server using a distributed delayed acceptance matching algorithm to obtain the optimal matching object for each task to be uninstalled.

[0082] The task unloading module is used to perform edge computing task unloading on each task based on the optimal matching object for each task to be unloaded.

[0083] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 5As shown, the computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores offloading data for industrial wireless network edge computing tasks. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for offloading tasks from industrial wireless network edge computing tasks.

[0084] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0085] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described method embodiments.

[0086] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the above-described method embodiments.

[0087] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described method embodiments.

[0088] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0089] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0090] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, etc., and are not limited to these.

[0091] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0092] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for offloading edge computing tasks in an industrial wireless network, characterized in that, The industrial wireless network edge computing task offloading method includes: Based on the topology of terminal devices, routing nodes, cloud computing centers, and multiple edge servers within the target industrial wireless network, construct an industrial wireless network model. For any task to be unloaded, a preference list for the task to be unloaded is constructed based on the task's preference for each matching object; the matching object is a combination scheme of communication path and edge server constructed based on the industrial wireless network model. For any given communication path, a preference list for the communication path is constructed based on the preferences of the communication path for each task to be uninstalled; the preferences of the communication path for the tasks to be uninstalled are determined based on the task parameters of the tasks to be uninstalled. For any given edge server, a preference list for each task to be uninstalled is constructed based on the edge server's preference for each task to be uninstalled; the edge server's preference for each task to be uninstalled is determined based on the task parameters of the task to be uninstalled. A distributed delayed acceptance matching algorithm is adopted to perform three-dimensional matching based on the preference lists of each task to be uninstalled, the preference lists of each communication path, and the preference lists of each edge server, so as to obtain the optimal matching object for each task to be uninstalled. Edge computing task uninstallation is performed on each task based on the optimal matching object for each task to be uninstalled.

2. The industrial wireless network edge computing task offloading method according to claim 1, characterized in that, The calculation process for the preference of the task to be uninstalled for each matching object is as follows: Multiple communication paths are constructed based on an industrial wireless network model; For any task to be unloaded, multiple matching objects corresponding to the task to be unloaded are constructed based on the industrial wireless network model and each communication path. Based on the transmission reliability and latency of each communication path, the unloading cost of the task to be unloaded, and the waiting latency and computation latency of each edge server, the utility of the task to be unloaded in selecting each matching object is calculated, and the preference of the task to be unloaded for each matching object is obtained.

3. The industrial wireless network edge computing task offloading method according to claim 1, characterized in that, The calculation process for the communication path's preference for each task to be unloaded is as follows: Obtain the task parameters of each task to be uninstalled, as well as the available bandwidth and load status of each communication path; For any given communication path, the preference of the communication path for each task to be unloaded is obtained based on the available bandwidth and load status of the communication path and the task parameters of each task to be unloaded.

4. The industrial wireless network edge computing task offloading method according to claim 1, characterized in that, The calculation process for the edge server's preference for each task to be unloaded is as follows: Obtain the task parameters of each task to be unloaded, as well as the computing load and remaining energy of each edge server; For any edge server, the preference of the communication path for each task to be unloaded is obtained based on the edge server's computing load and remaining energy, as well as the task parameters of each task to be unloaded.

5. The industrial wireless network edge computing task offloading method according to claim 1, characterized in that, A distributed delayed reception matching algorithm is adopted, which performs three-dimensional matching based on the preference lists of each task to be unloaded, the preference lists of each communication path, and the preference lists of each edge server to obtain the optimal matching object for each task to be unloaded. Specifically, this includes: At the current iteration number, for any unassigned task to be uninstalled at the current iteration number, the task to be uninstalled sends a request to the most preferred matching object in the preference list of the task to be uninstalled at the current iteration number; The communication path in the most preferred matching object determines whether the task to be uninstalled is the most preferred in the preference list under the current iteration number of the communication path to obtain a first judgment result, and the edge server in the most preferred matching object determines whether the task to be uninstalled is the most preferred in the preference list under the current iteration number of the edge server to obtain a second judgment result; If either the first judgment result or the second judgment result is negative, then the most preferred matching object in the preference list of the task to be uninstalled at the current iteration number is deleted from the preference list of the task to be uninstalled at the current iteration number, and the preference list of the task to be uninstalled at the next iteration number is obtained. If both the first and second judgment results are yes, then the most preferred matching object in the preference list of the task to be uninstalled under the current iteration number is determined as the optimal matching object of the task to be uninstalled, and the task to be uninstalled is deleted from the unassigned tasks under the current iteration number to obtain the unassigned tasks under the next iteration number. Update the iteration count and proceed to the next iteration. The iteration continues until all unassigned tasks have been assigned or the preference list for the next iteration of all unassigned tasks is empty, at which point the iteration stops.

6. The industrial wireless network edge computing task offloading method according to claim 1, characterized in that, The process of unloading each task by edge computing based on the optimal matching object also includes: Real-time monitoring of the status of each communication path and each edge server; For any communication path, if the status of the communication path is abnormal, the preference list of the communication path is updated based on the preference of the communication path for each task to be uninstalled. The preference list of each first target task to be uninstalled is updated according to the preference of each matching object for each first target task to be uninstalled; the optimal matching object for the first target task to be uninstalled includes the communication path; A distributed delayed acceptance matching algorithm is used to process the updated preference lists of each first target unloaded task, the updated preference list of the communication path, and the preference lists of each first target edge server to obtain the optimal matching object for each first target unloaded task; the optimal matching object for each first target unloaded task includes the first target edge server. If the state of any edge server is abnormal, the preference list of the edge server is updated based on the edge server's preference for each task to be uninstalled. The preference list of each second target task to be uninstalled is updated according to the preference of each matching object for each second target task to be uninstalled; the optimal matching object for the second target task to be uninstalled includes the edge server; A distributed delayed acceptance matching algorithm is used to process the updated preference lists of each second target unloading task, the updated preference list of the edge server, and the preference lists of each second communication path to obtain the optimal matching object for each second target unloading task; the optimal matching object for the second target unloading task includes the second communication path.

7. An industrial wireless network edge computing task offloading device, characterized in that, The industrial wireless network edge computing task offloading device includes: The industrial wireless network model building module is used to build an industrial wireless network model based on the topology between terminal devices, routing nodes, cloud computing centers, and multiple edge servers within the target industrial wireless network. A preference list construction module is used to construct a preference list for any task to be unloaded based on the task's preferences for each matching object; the matching objects are a combination scheme of communication paths and edge servers constructed based on an industrial wireless network model; for any communication path, a preference list for the communication path is constructed based on the communication path's preferences for each task to be unloaded; the preferences of the communication path for the task to be unloaded are determined based on the task parameters of the task to be unloaded; for any edge server, a preference list for the edge server is constructed based on the edge server's preferences for each task to be unloaded; the preferences of the edge server for the task to be unloaded are determined based on the task parameters of the task to be unloaded. The optimal matching object determination module is used to perform three-dimensional matching based on the preference lists of each task to be uninstalled, the preference lists of each communication path, and the preference lists of each edge server using a distributed delayed acceptance matching algorithm to obtain the optimal matching object for each task to be uninstalled. The task unloading module is used to perform edge computing task unloading on each task based on the optimal matching object for each task to be unloaded.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the industrial wireless network edge computing task offloading method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the industrial wireless network edge computing task offloading method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the industrial wireless network edge computing task offloading method as described in any one of claims 1-6.

Images