A task offloading method and device based on frequent application flow scheduling

Abstract

The present application provides a task offloading method and device based on frequent application set traffic scheduling. The method is based on a multi-user MEC system composed of MEC servers, base stations, and terminal users. First, a partial offloading model is designed based on a real application usage data set, and the application is regarded as the smallest unit in the task offloading process, and the association rules between applications are mined; the association rules are used together with the terminal user node status information to formulate MEC data pricing strategies, regulate the terminal user partial offloading strategies and evenly allocate system resources; under the premise of considering the system capacity constraints, the MEC system profit maximization problem is formulated as a nonlinear programming problem; and the relationship between MEC servers and terminal nodes is simulated through a Stackelberg game model to maximize the profit of the MEC server while preventing network congestion problems.

Description

Technical Field

[0001] The present application relates to the field of Internet of Things big data technology, and in particular, to a task offloading method and device based on frequent application set traffic scheduling. Background Art

[0002] The principle of task offloading in MEC (Mobile Edge Computing) is to extend cloud computing capabilities to the edge of cellular networks, which will minimize network congestion and improve resource optimization, user experience, and overall network performance. In MEC, task offloading is considered an effective method to ensure user quality of service by offloading compute-intensive or latency-sensitive tasks to edge devices or nearby edge servers. The main purpose of offloading is to reduce service response latency and improve service quality. In addition, when edge nodes lack processing power, computation can be migrated to edge servers or cloud data centers to improve overall system performance. To make task offloading decisions, many aspects can be considered, such as maximizing performance and minimizing energy consumption. Therefore, in the task offloading process, it is necessary to introduce and implement QoE requirements and formulate reasonable offloading strategies.

[0003] In MEC, the core issue lies in how to make offloading decisions, which are influenced by multiple factors, including task characteristics, network conditions, and platform differences. For example, the diversity and complexity of tasks may reduce the potential benefits of offloading. However, while task offloading solves the problems of computational latency and energy consumption in MEC, it also increases the frequency of data exchange in the IoT, leading to increasing network load. IoT applications require processing massive amounts of data. With the development and application of MEC task offloading technology, additional data exchange (including task data, node information, and so on) is generated between users and between users and servers. This rapid increase in data exchange exacerbates traffic congestion during peak hours and in hotspots. Therefore, balancing user service quality and network traffic data supply and demand in MEC systems and alleviating network communication pressure have become important challenges that network service providers urgently need to address.

[0004] In practical applications, the demand for traffic data in the IoT is influenced by many dynamic factors. From a spatiotemporal perspective, IoT devices have varying traffic data requirements at different locations and times. From a user perspective, in edge computing systems with task offloading, traffic data requirements are closely related to the types of tasks performed by the end devices. Due to these dynamic factors, system loads during peak hours can be significantly higher than during off-peak hours, making it unrealistic for MEC servers to simultaneously meet all user requests.

[0005] To characterize data traffic and manage traffic congestion, some work has proposed economic mechanisms that consider both time and application perspectives. However, existing task offloading methods that incorporate economic mechanisms do not consider the relationship between user application usage sequences and user satisfaction. For example, a proposed two-dimensional data pricing scheme based on time and application category awareness divides applications into categories based on their core functions, addressing network congestion and optimizing MEC system revenue. However, it does not consider the different impacts of dynamic factors on different users and ignores the impact of users' task usage habits, resulting in lower MEC system revenue and capacity utilization. Summary of the Invention

[0006] In view of the problems of low MEC system revenue and capacity utilization caused by user satisfaction and usage habits mentioned above, the present application provides a task offloading method and device based on frequent application set traffic scheduling, which maximizes MEC server revenue and balances network load by scheduling network traffic while ensuring capacity constraints and user satisfaction.

[0007] In order to achieve the above objectives, this application provides the following technical solutions:

[0008] A task offloading method based on frequent application set traffic scheduling is based on a multi-user MEC system, wherein the system comprises a MEC server, a base station, and at least two end users. The method comprises:

[0009] Collect the node information of the end user and determine the set of tasks to be offloaded. Each end user has only one task to be offloaded at each time t.

[0010] Establishing a partial offloading model in the MEC server and obtaining a set of users participating in task offloading;

[0011] Considering applications as the smallest unit in the task offloading process, we mine application-related association rules from application usage records in the application usage behavior modeling and prediction dataset.

[0012] Determining the terminal user's application uninstallation expectation based on the application's usage frequency, the association rules, and the terminal user satisfaction function;

[0013] The relationship between the MEC server and each end user is simulated based on the Stackelberg game model, and the benefit function that maximizes the end user's benefit is calculated using the reverse induction method to obtain the scheduling decision;

[0014] Summing the service revenue at time t, calculating the maximum revenue function of the MEC server while ensuring the traffic constraints of the multi-user MEC system, and solving to obtain the local optimal offloading strategy and the local optimal pricing strategy;

[0015] Solving the local optimal unloading strategy and the local optimal pricing strategy using an iterative method and a preset stopping criterion to determine a final pricing strategy;

[0016] The MEC server sends the final pricing policy to the corresponding end user, and the end user formulates a corresponding offloading policy according to a preset benefit function and determines the offloading ratio.

[0017] Furthermore, the set of tasks to be unloaded is represented as Where: I is the set of applications called by the task, θ is the set of task input data volume, τ is the set of offloading ratios, It is a set of uninstall expectations, which represents the highest price that the end user can accept.

[0018] Furthermore, the set of users participating in task offloading is N=1,2,...,n, where n∈N represents a terminal user in the set; at each time t, the set I=1,2,...,i,I is the task sequence executed by terminal user n, where i∈I represents a task in the set. is the amount of input data when the terminal user n performs the i-th task at time t, is the traffic price of the application used by the i-th task uninstalled by end user n at time t; Indicates the uninstall ratio.

[0019] Furthermore, the method of treating the application as the smallest unit in the task offloading process and mining application-related association rules from the application usage records in the application usage behavior modeling and prediction dataset includes:

[0020] Extract the application usage records of each terminal user, and use the Cspade algorithm in the time series pattern mining algorithm to calculate the frequent application set that meets the support, confidence and lift constraints, which is the association rule that meets the requirements.

[0021] Furthermore, the application uninstallation expectation of the terminal user is the uninstallation expectation of the terminal user n for the application task i within time t. in, is the expected offloading of unit traffic data, and f is the end-user satisfaction function.

[0022] Furthermore, the benefit function is in The traffic price of the application is determined by the pricing policy formulated by the MEC server; Indicates the amount of task offloaded, that is, the part offloaded to the MEC server for execution.

[0023] A task offloading device based on frequent application set traffic scheduling is based on a multi-user MEC system. The system consists of a MEC server, a base station, and at least two terminal users. The device includes:

[0024] The first processing unit is used to collect node information of the terminal user and determine a set of tasks to be offloaded. Each terminal user has only one task to be offloaded at each time t;

[0025] A second processing unit is configured to establish a partial offloading model in the MEC server and obtain a set of users participating in task offloading;

[0026] a third processing unit, configured to regard an application as the smallest unit in the task offloading process and mine association rules related to the application from application usage records in the application usage behavior modeling and prediction dataset;

[0027] a fourth processing unit, configured to determine the terminal user's application uninstallation expectation based on the application's usage frequency, the association rule, and the terminal user satisfaction function;

[0028] A fifth processing unit is configured to simulate the relationship between the MEC server and each end user based on a Stackelberg game model, and use a backward induction method to calculate a benefit function that maximizes the benefit of the end user to obtain a scheduling decision;

[0029] a sixth processing unit, configured to sum the service revenue at time t, calculate the maximum revenue function of the MEC server while ensuring the traffic constraint of the multi-user MEC system, and solve for a local optimal offloading strategy and a local optimal pricing strategy;

[0030] a seventh processing unit, configured to solve the local optimal unloading strategy and the local optimal pricing strategy using an iterative method and a preset stopping criterion, and determine a final pricing strategy;

[0031] The eighth processing unit is used for the MEC server to send the final pricing policy to the corresponding terminal user, and the terminal user formulates a corresponding offloading policy according to a preset benefit function and determines the offloading ratio.

[0032] Furthermore, the third processing unit is specifically used to extract application usage records of each terminal user, and use the Cspade algorithm in the time series pattern mining algorithm to calculate the frequent application set that meets the support, confidence and lift constraints as the association rules that meet the requirements.

[0033] A storage medium includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the task offloading method based on frequent application set traffic scheduling as described above.

[0034] An electronic device comprising at least one processor, and at least one memory and bus connected to the processor; wherein the processor and the memory communicate with each other via the bus; and the processor is used to call program instructions in the memory to execute the task offloading method based on frequent application set traffic scheduling as described above.

[0035] The task offloading method and device based on frequent application set traffic scheduling described in this application are based on a multi-user MEC system composed of MEC servers, base stations, and terminal users. First, a partial offloading model is designed based on a real application usage data set, and the application is regarded as the smallest unit in the task offloading process, and the association rules between applications are mined; the association rules are used together with the terminal user node status information to formulate MEC data pricing strategies, regulate the terminal user partial offloading strategies and evenly allocate system resources; under the premise of considering the system capacity constraints, the MEC system profit maximization problem is formulated as a nonlinear programming problem; the relationship between MEC servers and terminal nodes is simulated through the Stackelberg game model to maximize the profit of the MEC server while preventing network congestion problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0037] Figure 1 This is a schematic diagram of the structure of a multi-user MEC system disclosed in an embodiment of the present application;

[0038] Figure 2 This is a flow chart of a task offloading method based on frequent application set flow scheduling disclosed in an embodiment of the present application;

[0039] Figure 3 This is a structural diagram of a task offloading device based on frequent application set flow scheduling disclosed in an embodiment of the present application;

[0040] Figure 4 This is a schematic structural diagram of an electronic device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0041] This application provides a task offloading method and device based on frequent application set flow scheduling. Figure 1The multi-user MEC system consists of a MEC server 11, a base station 12, and an end user 13. It uses a time series pattern mining algorithm to mine application usage association rules to improve the revenue of the MEC server while ensuring system traffic constraints. At the same time, it uses a comprehensive consideration of single application usage expectations, frequent application set usage expectations, and user satisfaction to model a user service evaluation model, thereby achieving the effect of fine-grained adjustment of task offloading ratios through data pricing, and realizing the goal of balancing network load and optimizing MEC server revenue.

[0042] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0043] Please see the attached Figure 2 , is a flowchart of a task offloading method based on frequent application set flow scheduling provided by an embodiment of the present application. Figure 2 As shown, the embodiment of the present application provides a task offloading method based on frequent application set traffic scheduling, based on a multi-user MEC system, the system consists of a MEC server, a base station, and at least two terminal users, and the method includes the following steps:

[0044] S201: Collect node information of end users and determine a set of tasks to be offloaded. Each end user has only one task to be offloaded at each time t.

[0045] In the embodiment of the present application, at each time t, the MEC server collects the terminal node status information, obtains the task information of each terminal node, and represents a computing task or a task offloading behavior as Where: I is the set of applications called by the task, θ is the set of task input data volume, τ is the set of offloading ratios, It is a set of uninstall expectations, which represents the highest price that the end user can accept.

[0046] S202: Establishing a partial offloading model in the MEC server and obtaining a set of users participating in task offloading;

[0047] In the embodiment of the present application, the set of users participating in task offloading is N=1,2,...,n, where n∈N represents a terminal user in the set; at each time t, the set I=1,2,...,i,I is a sequence of tasks executed by terminal user n, where i∈I represents a task in the set. is the amount of input data when the terminal user n performs the i-th task at time t, is the traffic price of the application used by the i-th task uninstalled by the terminal user n at time t; Indicates the uninstall rate of each user. The specific formula is:

[0048]

[0049] in, represents the amount of uninstallation, then, It is the actual traffic cost of the application. The local execution part does not consume network traffic data.

[0050] S203: Considering the application as the smallest unit in the task offloading process, mining association rules related to the application from the application usage records in the application usage behavior modeling and prediction dataset;

[0051] In the embodiment of the present application, the application is considered as the smallest unit in the task offloading process, and association rules related to the application are mined from the application usage records in the application usage behavior modeling and prediction dataset, including:

[0052] Extract the application usage records of each terminal user, and use the Cspade algorithm in the time series pattern mining algorithm to calculate the frequent application set that meets the support, confidence and lift constraints, which is the association rule that meets the requirements.

[0053] In a specific embodiment, the MEC server collects and maintains an application usage database to save each user's application usage records, divides a day into four time periods, and calculates the user's single application usage frequency σ in each time period based on the user's usage records. ni And three indicators of association rules, including support Confidence and lift Support, confidence, and lift are metrics used to evaluate the strength of association rules. For any two tasks, A and B, support represents the frequency of itemsets in which A is executed before B, defined in set theory as S(A→B) = P(A→B). Confidence represents the frequency of itemsets in which A is executed before B, defined as C(A→B) = P(B|A). Lift reflects the correlation between executions A and B, defined as L(A→B) = P(B|A) / P(B). The Cspade algorithm, a time series pattern mining algorithm, is used to mine association rules. Compared to frequent pattern mining algorithms (such as Apriori and FPGrowth), the Cspade algorithm minimizes I / O costs while taking into account the order of transactions. It is suitable for extracting frequent patterns from computationally intensive tasks involving multiple applications.

[0054] S204: Determine the terminal user's application uninstallation expectation based on the application's usage frequency, the association rule, and the terminal user satisfaction function;

[0055] In the embodiment of the present application, the application uninstallation expectation of the terminal user is calculated. The uninstallation expectation includes two parts: the unit traffic uninstallation expectation of the application and the user satisfaction function. The application uninstallation expectation of the terminal user is the uninstallation expectation of the terminal user n for the application task i at time t. in, is the expected offloading of unit traffic data, and f is the end-user satisfaction function.

[0056] Unit traffic unloading expectation It depends on the usage frequency of the application itself and the association rules related to the application. The specific formula is:

[0057]

[0058] Among them, ρ1 is a weighting factor used to determine the frequency of use of the application itself and the impact of association rules related to the application on expectations, σ ni is the usage frequency of application i by user n, is the support of user n for association rule m of application i.

[0059] The calculation formula of satisfaction function is:

[0060]

[0061]

[0062] Among them, ρ2 is the weighting factor, β i It represents the traffic consumption speed of application i and is positively correlated with the computational complexity of the application itself. The average traffic consumption of application i in the task offloading dataset can be calculated based on the The average traffic consumption of all applications The ratio of . represents the correlation factor, where is a binary indicator variable used to indicate whether there is an association rule m between the current application i and the application that has been used in the previous moment. is the confidence of user n on the association rule m of application i, is the improvement of user n on association rule m of application i.

[0063] S205: Simulating the relationship between the MEC server and each end user based on the Stackelberg game model, and using the reverse induction method to calculate the benefit function that maximizes the end user's benefit to obtain a scheduling decision;

[0064] S206: Sum the service revenue at time t, calculate the maximum revenue function of the MEC server while ensuring the traffic constraint of the multi-user MEC system, and solve to obtain the local optimal offloading strategy and the local optimal pricing strategy;

[0065] In the embodiment of the present application, the uninstallation strategy that maximizes the user's benefit is calculated. The benefit function is:

[0066]

[0067] in, Is the application uninstall expectation, It is the unit traffic price multiplied by the unloading volume, that is, the actual price. The traffic price of the application is determined by the pricing policy formulated by the MEC server; Indicates the amount of task offloaded, that is, the part offloaded to the MEC server for execution.

[0068] For the above The derivation shows that the optimal uninstallation decision for a single user must exist and is the feasible downward direction of traffic price, and r is the coefficient of satisfaction function f. The benefit function is a function of the user itself, which is related to the unit price and the amount of offloads. The MEC server revenue function is the sum of the unit price and the amount of offloads for all users in the entire system for different applications.

[0069] S207: Solving the local optimal unloading strategy and the local optimal pricing strategy using an iterative method and a preset stopping criterion to determine a final pricing strategy;

[0070] Calculate the maximum revenue function of the MEC server at time t while satisfying the traffic constraint μ. In order to ensure that both peak and off-peak periods are taken into account, μ is taken as a value between the highest and lowest traffic consumption values.

[0071] The MEC server revenue maximization function is expressed as:

[0072]

[0073] Substitution The optimal unloading strategy is derived and the optimal pricing strategy is calculated. It should be noted that a nonlinear programming problem can be solved using an efficient nonlinear programming solver. However, it should be noted that solving such constrained problems generally only yields a local optimal solution, requiring an iterative approach to approximate the optimal solution.

[0074] S208: The MEC server sends the final pricing policy to the corresponding terminal user, and the terminal user formulates a corresponding offloading policy according to a preset benefit function and determines the offloading ratio.

[0075] An embodiment of the present application provides a task offloading method based on frequent application set traffic scheduling. The method is based on a multi-user MEC system composed of an MEC server, a base station, and terminal users. First, a partial offloading model is designed based on a real application usage data set, and the application is regarded as the smallest unit in the task offloading process, and the association rules between applications are mined; the association rules are used together with the terminal user node status information to formulate an MEC data pricing strategy, regulate the terminal user partial offloading strategy and evenly allocate system resources; under the premise of considering the system capacity constraint, the MEC system profit maximization problem is formulated as a nonlinear programming problem; the relationship between the MEC server and the terminal node is simulated through the Stackelberg game model to maximize the profit of the MEC server while preventing network congestion problems.

[0076] See also Figure 3 Based on the task offloading method based on frequent application set traffic scheduling disclosed in the above embodiment, this embodiment correspondingly discloses a task offloading device based on frequent application set traffic scheduling, based on a multi-user MEC system, the system comprising a MEC server, a base station, and at least two terminal users, the device including:

[0077] The first processing unit 301 is used to collect node information of the terminal user and determine a set of tasks to be offloaded. Each terminal user has only one task to be offloaded at each time t;

[0078] The second processing unit 302 is configured to establish a partial offloading model in the MEC server and obtain a set of users participating in task offloading;

[0079] The third processing unit 303 is configured to regard the application as the smallest unit in the task offloading process and mine association rules related to the application from the application usage records in the application usage behavior modeling and prediction dataset;

[0080] The fourth processing unit 304 is configured to determine the terminal user's application uninstallation expectation based on the application's usage frequency, the association rule, and the terminal user satisfaction function;

[0081] A fifth processing unit 305 is configured to simulate the relationship between the MEC server and each end user based on a Stackelberg game model, and use a backward induction method to calculate a benefit function that maximizes the benefit of the end user to obtain a scheduling decision;

[0082] The sixth processing unit 306 is configured to sum the service revenue at time t, calculate the maximum revenue function of the MEC server while ensuring the traffic constraint of the multi-user MEC system, and solve to obtain a local optimal offloading strategy and a local optimal pricing strategy;

[0083] The seventh processing unit 307 is configured to solve the local optimal unloading strategy and the local optimal pricing strategy using an iterative method and a preset stopping criterion to determine a final pricing strategy;

[0084] The eighth processing unit 308 is configured for the MEC server to send the final pricing policy to the corresponding terminal user, and the terminal user formulates a corresponding offloading policy according to a preset benefit function and determines the offloading ratio.

[0085] Furthermore, the third processing unit 303 is specifically configured to extract application usage records of each terminal user, and use the Cspade algorithm in the time series pattern mining algorithm to calculate the frequent application sets that meet the support, confidence and lift constraints as the required association rules.

[0086] The task unloading device based on frequent application set flow scheduling includes a processor and a memory. The above-mentioned first processing unit, second processing unit and third processing unit are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0087] The processor contains a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set up, and by adjusting kernel parameters to schedule network traffic, the MEC server revenue is maximized and the network load is balanced under the premise of ensuring capacity constraints and user satisfaction.

[0088] An embodiment of the present application provides a storage medium having a program stored thereon, which, when executed by a processor, implements the task offloading method based on frequent application set traffic scheduling.

[0089] An embodiment of the present application provides a processor, which is used to run a program, wherein the task offloading method based on frequent application set traffic scheduling is executed when the program is running.

[0090] The present application embodiment provides an electronic device, such as Figure 4 As shown, the electronic device 40 includes at least one processor 401, and at least one memory 402 and a bus 403 connected to the processor; wherein the processor 401 and the memory 402 communicate with each other through the bus 403; the processor 401 is used to call the program instructions in the memory 402 to execute the above-mentioned task unloading method based on frequent application set traffic scheduling.

[0091] The electronic devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0092] The present application also provides a computer program product which, when executed on a data processing device, is adapted to execute a program for initializing the following method steps:

[0093] Collect the node information of the end user and determine the set of tasks to be offloaded. Each end user has only one task to be offloaded at each time t.

[0094] Establishing a partial offloading model in the MEC server and obtaining a set of users participating in task offloading;

[0095] Considering applications as the smallest unit in the task offloading process, we mine application-related association rules from application usage records in the application usage behavior modeling and prediction dataset.

[0096] Determining the terminal user's application uninstallation expectation based on the application's usage frequency, the association rules, and the terminal user satisfaction function;

[0097] The relationship between the MEC server and each end user is simulated based on the Stackelberg game model, and the benefit function that maximizes the end user's benefit is calculated using the reverse induction method to obtain the scheduling decision;

[0098] Summing the service revenue at time t, calculating the maximum revenue function of the MEC server while ensuring the traffic constraints of the multi-user MEC system, and solving to obtain the local optimal offloading strategy and the local optimal pricing strategy;

[0099] Solving the local optimal unloading strategy and the local optimal pricing strategy using an iterative method and a preset stopping criterion to determine a final pricing strategy;

[0100] The MEC server sends the final pricing policy to the corresponding end user, and the end user formulates a corresponding offloading policy according to a preset benefit function and determines the offloading ratio.

[0101] Furthermore, the set of tasks to be unloaded is represented as Where: I is the set of applications called by the task, θ is the set of task input data volume, τ is the set of offloading ratios, It is a set of uninstall expectations, which represents the highest price that the end user can accept.

[0102] Furthermore, the set of users participating in task offloading is N=1,2,...,n, where n∈N represents a terminal user in the set; at each time t, the set I=1,2,...,i,I is the task sequence executed by terminal user n, where i∈I represents a task in the set. is the amount of input data when the terminal user n performs the i-th task at time t, is the traffic price of the application used by the i-th task uninstalled by end user n at time t; Indicates the uninstall ratio.

[0103] Furthermore, the method of treating the application as the smallest unit in the task offloading process and mining application-related association rules from the application usage records in the application usage behavior modeling and prediction dataset includes:

[0104] Extract the application usage records of each terminal user, and use the Cspade algorithm in the time series pattern mining algorithm to calculate the frequent application set that meets the support, confidence and lift constraints, which is the association rule that meets the requirements.

[0105] Furthermore, the application uninstallation expectation of the terminal user is the uninstallation expectation of the terminal user n for the application task i within time t. in, is the expected offloading of unit traffic data, and f is the end-user satisfaction function.

[0106] Furthermore, the benefit function is in The traffic price of the application is determined by the pricing policy formulated by the MEC server; Indicates the amount of task offloaded, that is, the part offloaded to the MEC server for execution.

[0107] The present application is described in terms of flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0108] In a typical configuration, the device includes one or more processors (CPUs), memory, and a bus. The device may also include input / output interfaces, network interfaces, and the like.

[0109] Memory may include non-permanent memory in a computer-readable medium, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory includes at least one memory chip. Memory is an example of a computer-readable medium.

[0110] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0111] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0112] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0113] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A task offloading method based on frequent application set flow scheduling, characterized in that: Based on a multi-user MEC system, the system comprises a MEC server, a base station, and at least two end users. The method includes: Collect the node information of the end user and determine the set of tasks to be offloaded. Each end user has only one task to be offloaded at each time t. Establishing a partial offloading model in the MEC server and obtaining a set of users participating in task offloading; Considering applications as the smallest unit in the task offloading process, we mine application-related association rules from application usage records in the application usage behavior modeling and prediction dataset. Determining the application uninstallation expectation of the terminal user based on the usage frequency of the application itself, the association rule, and the terminal user satisfaction function; The relationship between the MEC server and each end user is simulated based on the Stackelberg game model, and the benefit function that maximizes the end user's benefit is calculated using the reverse induction method to obtain the scheduling decision; Summing the service revenue at time t, calculating the maximum revenue function of the MEC server while ensuring the traffic constraints of the multi-user MEC system, and solving for the local optimal offloading strategy and the local optimal pricing strategy; Solving the local optimal unloading strategy and the local optimal pricing strategy using an iterative method and a preset stopping criterion to determine a final pricing strategy; The MEC server sends the final pricing policy to the corresponding end user, and the end user formulates a corresponding offloading policy based on a preset benefit function and determines the offloading ratio; The application uninstallation expectation includes two parts: the application unit traffic uninstallation expectation and the user satisfaction function. The application uninstallation expectation of the terminal user is the uninstallation expectation of the terminal user n for the application task i at time t. in, For unit traffic data offloading expectations, is the amount of input data when the end user n performs the i-th task at time t, and f is the end user satisfaction function; Unit traffic unloading expectation It depends on the usage frequency of the application itself and the association rules related to the application. The specific formula is: Among them, ρ1 is a weighting factor used to determine the frequency of use of the application itself and the impact of association rules related to the application on expectations, σ ni is the usage frequency of application i by user n, is the support degree of user n for association rule m of application i; The calculation formula of satisfaction function is: in, represents the uninstall ratio of each user, ρ2 is the weighting factor; β i It represents the traffic consumption speed of application i and is positively correlated with the computational complexity of the application itself. According to the average traffic consumption of application i in the task offloading dataset, The average traffic consumption of all applications is the highest The ratio of represents the correlation factor, where is a binary indicator variable used to indicate whether there is an association rule m between the current application i and the application that has been used in the previous moment; is the confidence of user n in association rule m of application i; is the improvement of user n on association rule m of application i.

2. The method according to claim 1, characterized in that The set of tasks to be offloaded is represented as Where: I is the set of applications called by the task, θ is the set of task input data volume, τ is the set of offloading ratios, It is a set of uninstall expectations, which represents the highest price that the end user can accept.

3. The method according to claim 1, characterized in that The user set participating in task offloading is N=1,2,...,n, where n∈N represents a terminal user in the set; at each time t, the set I=1,2,...,i, where I is the task sequence executed by terminal user n, where i∈I represents a task in the set. The specific formula is: in, Indicates the amount of uninstalls.

4. The method according to claim 1, wherein The method regards the application as the smallest unit in the task offloading process and mines application-related association rules from application usage records in the application usage behavior modeling and prediction dataset, including: Extract the application usage records of each terminal user, and use the Cspade algorithm in the time series pattern mining algorithm to calculate the frequent application set that meets the support, confidence and lift constraints, which is the association rule that meets the requirements.

5. The method according to claim 1, characterized in that The benefit function is in is the traffic price of the application used by the i-th task uninstalled by end user n at time t, which is determined by the pricing policy formulated by the MEC server; Indicates the amount of task offloaded, that is, the part that is offloaded to the MEC server for execution. The actual traffic cost of the application. The local execution part does not consume network traffic data.

6. A task offloading device based on frequent application flow scheduling, characterized in that: Based on a multi-user MEC system, the system consists of a MEC server, a base station, and at least two end users. The device includes: The first processing unit is used to collect node information of the terminal user and determine a set of tasks to be offloaded. Each terminal user has only one task to be offloaded at each time t; A second processing unit is configured to establish a partial offloading model in the MEC server and obtain a set of users participating in task offloading; a third processing unit, configured to regard an application as the smallest unit in the task offloading process and mine association rules related to the application from application usage records in the application usage behavior modeling and prediction dataset; The fourth processing unit is used to determine the terminal user's application uninstallation expectation based on the application's usage frequency, the association rule, and the terminal user satisfaction function; the application uninstallation expectation includes two parts: the application's unit traffic uninstallation expectation and the user satisfaction function. The terminal user's application uninstallation expectation is the terminal user n's uninstallation expectation for application task i within time t. in, For unit traffic data offloading expectations, is the amount of input data when the end user n performs the i-th task at time t, and f is the end user satisfaction function; Unit traffic unloading expectation It depends on the usage frequency of the application itself and the association rules related to the application. The specific formula is: Among them, ρ1 is a weighting factor used to determine the frequency of use of the application itself and the impact of association rules related to the application on expectations, σ ni is the usage frequency of application i by user n, is the support degree of user n for association rule m of application i; The calculation formula of satisfaction function is: in, represents the uninstall ratio of each user, ρ2 is the weighting factor; β i It represents the traffic consumption speed of application i and is positively correlated with the computational complexity of the application itself. According to the average traffic consumption of application i in the task offloading dataset, The average traffic consumption of all applications is the highest The ratio of represents the correlation factor, where is a binary indicator variable used to indicate whether there is an association rule m between the current application i and the application that has been used in the previous moment; is the confidence of user n in association rule m of application i; is the improvement of user n on association rule m of application i; A fifth processing unit is configured to simulate the relationship between the MEC server and each end user based on a Stackelberg game model, and use a backward induction method to calculate a benefit function that maximizes the benefit of the end user to obtain a scheduling decision; a sixth processing unit, configured to sum the service revenue at time t, calculate the maximum revenue function of the MEC server while ensuring the traffic constraint of the multi-user MEC system, and solve for a local optimal offloading strategy and a local optimal pricing strategy; a seventh processing unit, configured to solve the local optimal unloading strategy and the local optimal pricing strategy using an iterative method and a preset stopping criterion, and determine a final pricing strategy; The eighth processing unit is used for the MEC server to send the final pricing policy to the corresponding terminal user, and the terminal user formulates a corresponding offloading policy according to a preset benefit function and determines the offloading ratio.

7. The device according to claim 6, characterized in that The third processing unit is specifically used to extract the application usage records of each terminal user, and use the Cspade algorithm in the time series pattern mining algorithm to calculate the frequent application set that meets the support, confidence and lift constraints as the association rules that meet the requirements.

8. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the task offloading method based on frequent application set traffic scheduling according to any one of claims 1 to 5.

9. An electronic device, characterized in that: The electronic device includes at least one processor, and at least one memory and bus connected to the processor; wherein the processor and the memory communicate with each other through the bus; the processor is used to call program instructions in the memory to execute the task offloading method based on frequent application set traffic scheduling as described in any one of claims 1 to 5.

Images