Bandwidth and computing power distribution method and system for edge computing

Abstract

The invention relates to the technical field of communication resource allocation, and discloses a bandwidth and computing power allocation method and system for edge computing, and the method comprises the steps: obtaining real-time task data of an edge computing system; constructing a bandwidth and computing power allocation model by taking maximization of system resource efficiency and minimization of the sum of task flow average waiting delay and task loss rate as optimization objectives, and setting constraint conditions; real-time task data are input into the bandwidth and computing power distribution model for iterative solution, and an optimal bandwidth and computing power distribution scheme meeting constraint conditions is obtained. Through accurate bandwidth and computing power distribution, idle or excessive occupation of resources can be avoided, the overall resource utilization efficiency is improved, and the resource utilization rate is increased. The technical problem of how to improve the bandwidth and computing power resource dynamic allocation utilization rate in the edge computing environment is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of communication resource allocation, and in particular to a method and system for allocating bandwidth and computing power for edge computing. Background Art

[0002] With the rapid development of the Internet of Things (IoT), 5G communications, and real-time intelligent applications, edge computing has become a key technology for addressing the high latency and bandwidth bottlenecks of centralized cloud computing. By moving computing tasks to edge nodes close to data sources, edge computing significantly reduces network latency and improves real-time performance.

[0003] Currently, bandwidth and computing power allocation methods in edge computing environments typically use simple prediction algorithms to adjust resources. However, these algorithms lack high prediction accuracy and cannot adjust in real time, making them difficult to meet the multi-task coordination requirements in complex network environments. Overall, existing technologies still have significant room for improvement in terms of dynamic resource adjustment and efficient allocation. Summary of the Invention

[0004] The present invention provides a bandwidth and computing power allocation method and system for edge computing, which solves the technical problem of how to improve the dynamic allocation utilization of bandwidth and computing power resources in an edge computing environment.

[0005] A first aspect of the present invention provides a method for allocating bandwidth and computing power for edge computing, comprising:

[0006] Obtain real-time task data from edge computing systems;

[0007] With the optimization goals of maximizing system resource efficiency and minimizing the sum of the average waiting delay of the task flow and the task loss rate, a bandwidth and computing power allocation model is constructed and constraints are set;

[0008] The real-time task data is input into the bandwidth and computing power allocation model for iterative solution to obtain an optimal bandwidth and computing power allocation solution that meets the constraint conditions.

[0009] Optionally, the constraint conditions include bandwidth overflow probability constraint, task flow delay constraint and task loss rate constraint.

[0010] Optionally, the using the real-time task data to input the bandwidth and computing power allocation model for iterative solution includes:

[0011] The real-time task data is input into the bandwidth and computing power allocation model, and an iterative solution is performed using the Lagrange multiplier method or reinforcement learning.

[0012] Optionally, it also includes:

[0013] When the iterative solution result does not meet the constraint conditions, a priority factor is introduced or a preset upper limit threshold of each constraint is adjusted based on the constraint conditions;

[0014] Jump to the step of using the real-time task data to input the bandwidth and computing power allocation model for iterative solution until the iterative solution result meets the constraint condition.

[0015] Optionally, the priority factor is obtained in the following manner:

[0016] Determining a comprehensive task arrival rate using the real-time task data;

[0017] Constructing a task flow arrival distribution function according to the comprehensive arrival rate of the tasks;

[0018] Performing Laplace transform on the task flow arrival distribution function to obtain a transformed task flow arrival distribution function;

[0019] Performing an integration operation on the transformed task flow arrival distribution function using the method of integration by parts to obtain a first integral;

[0020] performing an integral operation on the transformed task flow arrival distribution function to obtain a second integral;

[0021] A priority factor is obtained by performing a ratio operation on the first integral and the second integral.

[0022] Optionally, the real-time task data includes an average arrival rate of high-priority tasks and an average arrival rate of low-priority tasks, and determining the comprehensive task arrival rate using the real-time task data includes:

[0023] The average arrival rate of the high-priority tasks and the average arrival rate of the low-priority tasks are used to perform a sum operation to obtain a comprehensive task arrival rate.

[0024] A second aspect of the present invention provides a bandwidth and computing power allocation system for edge computing, including:

[0025] Acquisition module, used to obtain real-time task data of edge computing system;

[0026] A construction module is used to build a bandwidth and computing power allocation model and set constraints with the optimization goals of maximizing system resource efficiency and minimizing the sum of the average waiting delay and task loss rate of the task flow;

[0027] The solution module is used to use the real-time task data to input the bandwidth and computing power allocation model for iterative solution to obtain the optimal bandwidth and computing power allocation solution that meets the constraint conditions.

[0028] The third aspect of the present invention provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory. When the computer program is executed by the processor, the processor executes the steps of the edge computing bandwidth and computing power allocation method as described in any one of the above items.

[0029] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed, implements the bandwidth and computing power allocation method for edge computing as described in any one of the above items.

[0030] A fifth aspect of the present invention provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the edge computing bandwidth and computing power allocation method as described in any one of the above items.

[0031] It can be seen from the above technical solutions that the present invention has the following advantages:

[0032] In the present invention, maximizing system resource efficiency can achieve refined allocation and intensive utilization of resources, and minimizing the sum of the average waiting delay of task flows and the task loss rate can improve the service quality and reliability of the edge computing system. A bandwidth and computing power allocation model is constructed with maximizing system resource efficiency and minimizing the sum of the average waiting delay of task flows and the task loss rate as the dual optimization goals, and constraints are set. The feasible domain of resource allocation is limited from three dimensions: bandwidth overflow probability constraint, task flow delay constraint, and task loss rate constraint. After the real-time task data collected in real time is input into the bandwidth and computing power allocation model, a preset iterative algorithm is used to solve it. In each iteration, the algorithm adjusts the bandwidth and computing power allocation parameters according to the quality of the current solution, and gradually approaches the optimal bandwidth and computing power allocation scheme that meets all constraints, so that resources can be reasonably allocated in a multi-task environment, ensuring that the bandwidth and computing power requirements of different tasks are met. The present invention can avoid idle or excessive resource occupation through precise bandwidth and computing power allocation, improve overall resource utilization efficiency, and solve the technical problem of how to improve the dynamic allocation utilization rate of bandwidth and computing power resources in an edge computing environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention 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 invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0034] Figure 1A flowchart of a method for allocating bandwidth and computing power for edge computing provided in Example 1 of the present invention;

[0035] Figure 2 A flowchart of a method for allocating bandwidth and computing power for edge computing provided in the second embodiment of the present invention;

[0036] Figure 3 This is a block diagram of a bandwidth and computing power allocation system for edge computing provided in Example 3 of the present invention;

[0037] Figure 4 This is a structural block diagram of a computer device provided in Example 4 of the present invention. DETAILED DESCRIPTION

[0038] Embodiments of the present invention provide a bandwidth and computing power allocation method and system for edge computing, which are used to solve the technical problem of how to improve the dynamic allocation utilization of bandwidth and computing power resources in an edge computing environment.

[0039] In order to make the purpose, features, and advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0040] Existing methods for allocating bandwidth and computing power in edge computing environments primarily include static allocation and priority-based allocation. Static allocation, which pre-determines resource quotas for each task, is suitable for scenarios with relatively stable resource demands. However, it lacks flexibility when task demands fluctuate significantly, resulting in low resource utilization. Priority-based allocation prioritizes resources for high-priority tasks based on their importance or latency requirements. This improves the response speed of high-priority tasks to a certain extent, but still cannot adapt to the complex demands of edge environments.

[0041] See also Figure 1 , Figure 1 This is a flowchart of the steps of a method for allocating bandwidth and computing power for edge computing provided in Example 1 of the present invention.

[0042] The present invention provides a method for allocating bandwidth and computing power for edge computing, including:

[0043] Step 101: Obtain real-time task data of the edge computing system.

[0044] Edge computing system refers to a distributed computing architecture that is specifically used to deploy data processing, storage, and applications at the edge of the network close to the data source or terminal device, and perform tasks such as data processing and analysis at the edge of the network.

[0045] Real-time task data refers to the immediate, time-varying information related to tasks during the operation of edge computing systems. This includes task arrival time, task type, task priority (e.g., high, medium, or low), task data size, required computing power, and task sensitivity to latency. It reflects the current status and requirements of tasks and serves as a crucial basis for building and optimizing bandwidth and computing power allocation models.

[0046] Tasks refer to the various types of workloads that edge computing systems need to handle, including but not limited to data processing tasks for analyzing, calculating, converting, and performing other operations on raw data (such as video stream data collected by smart cameras, which requires image recognition, target detection, and other processing), computationally intensive tasks such as complex simulation calculations in industrial production and numerical calculations in scientific research, communication-related tasks for data transmission processes and related network configuration, data packaging and unpacking, and other operations (for example, in the Internet of Things environment, data collected by numerous sensor nodes needs to be transmitted to the edge server), and control tasks such as issuing control instructions and adjusting the operating status of equipment.

[0047] In an embodiment of the present invention, real-time task data related to tasks and dynamically changing over time is obtained during the operation of the edge computing system.

[0048] Step 102: Build a bandwidth and computing power allocation model and set constraints with the optimization goal of maximizing system resource efficiency and minimizing the sum of the average waiting delay of the task flow and the task loss rate.

[0049] System resource efficiency refers to a comprehensive indicator of the edge computing system's utilization of bandwidth and computing power resources and output effects. It is used to measure how much effective task processing can be generated by data transmitted through bandwidth under unit computing power consumption.

[0050] The average waiting delay of a task flow refers to the average time a task waits in the queue to be processed.

[0051] The task loss rate refers to the probability that a task is discarded due to system overload or insufficient resources.

[0052] The sum of the average waiting delay of the task flow and the task loss rate refers to a comprehensive service quality indicator used to measure the system's ability to process tasks. The smaller the indicator, the more timely the system processes tasks and the lower the risk of packet loss.

[0053] The bandwidth and computing power allocation model refers to an optimization model used to determine the optimal allocation of bandwidth and computing power in edge computing systems. The model finds the optimal combination of bandwidth and computing power by balancing resource efficiency and service quality.

[0054] Constraints refer to the restrictions that must be met in the optimization model to ensure the feasibility and rationality of the allocation plan, including but not limited to bandwidth overflow probability constraints, task flow delay constraints, and task loss rate constraints.

[0055] In an embodiment of the present invention, a bandwidth and computing power allocation model with multiple optimization objectives is constructed with the optimization goals of maximizing system resource efficiency and minimizing the sum of the average waiting delay of task flows and the task loss rate, and the constraints of the model are set to ensure that the model solves the optimal bandwidth and computing power allocation solution.

[0056] Step 103: Use the real-time task data input bandwidth and computing power allocation model to perform iterative solution to obtain the optimal bandwidth and computing power allocation solution that meets the constraints.

[0057] In an embodiment of the present invention, real-time task data is input into a bandwidth and computing power allocation model, and a preset linear solution algorithm is used to perform an iterative solution, thereby obtaining an optimal bandwidth and computing power allocation solution that meets the constraint conditions.

[0058] In the present invention, maximizing system resource efficiency can achieve refined allocation and intensive utilization of resources, and minimizing the sum of the average waiting delay of task flows and the task loss rate can improve the service quality and reliability of the edge computing system. A bandwidth and computing power allocation model is constructed with maximizing system resource efficiency and minimizing the sum of the average waiting delay of task flows and the task loss rate as the dual optimization goals, and constraints are set. The feasible domain of resource allocation is limited from three dimensions: bandwidth overflow probability constraint, task flow delay constraint, and task loss rate constraint. After the real-time task data collected in real time is input into the bandwidth and computing power allocation model, a preset iterative algorithm is used to solve it. In each iteration, the algorithm adjusts the bandwidth and computing power allocation parameters according to the quality of the current solution, and gradually approaches the optimal bandwidth and computing power allocation scheme that meets all constraints, so that resources can be reasonably allocated in a multi-task environment, ensuring that the bandwidth and computing power requirements of different tasks are met. The present invention can avoid idle or excessive resource occupation through precise bandwidth and computing power allocation, improve overall resource utilization efficiency, and solve the technical problem of how to improve the dynamic allocation utilization rate of bandwidth and computing power resources in an edge computing environment.

[0059] See also Figure 2 , Figure 2 A flowchart of the steps of a bandwidth and computing power allocation method for edge computing provided in Example 2 of the present invention.

[0060] The present invention provides a method for allocating bandwidth and computing power for edge computing, including:

[0061] Step 201: Obtain real-time task data of the edge computing system.

[0062] In the embodiment of the present invention, the specific implementation process of step 201 is similar to that of step 101 and will not be repeated here.

[0063] Step 202: With the optimization goal of maximizing system resource efficiency and minimizing the sum of the average waiting delay of the task flow and the task loss rate, a bandwidth and computing power allocation model is constructed, and constraints are set.

[0064] Furthermore, the constraints include bandwidth overflow probability constraint, task flow delay constraint, and task loss rate constraint;

[0065] It should be noted that the bandwidth overflow probability constraint corresponds to the dimension of network resource reliability. By limiting the probability that the allocated bandwidth exceeds the actual available bandwidth, it avoids network congestion or data packet loss caused by insufficient bandwidth, thereby ensuring network transmission stability.

[0066] The task flow delay constraint corresponds to the dimension of task processing timeliness. By limiting the maximum allowable time from task receipt to completion of processing, it ensures that tasks with high real-time requirements (such as autonomous driving instructions and medical emergency data) are processed within the delay tolerated by the business.

[0067] The task loss rate constraint corresponds to the service quality reliability dimension. By controlling the proportion of tasks that cannot be processed and are lost due to insufficient resources (such as bandwidth or computing power exhaustion), the system's task carrying capacity and service reliability are improved.

[0068] The bandwidth overflow probability constraint is specifically that the bandwidth overflow probability is less than or equal to a preset bandwidth overflow probability upper limit threshold;

[0069] The task flow delay constraint includes the task flow average waiting delay constraint and the task flow processing delay constraint. Specifically, the task flow average waiting delay is less than or equal to the preset task flow average waiting delay upper threshold, and the task flow processing delay is less than or equal to the preset task flow processing delay upper threshold.

[0070] The task loss rate constraint is specifically that the task loss rate is less than or equal to the preset task loss rate upper limit threshold.

[0071] In the specific implementation, the objective function of the bandwidth and computing power allocation model is:

[0072]

[0073] Where, Indicates the system resource efficiency, Indicates the bandwidth allocation ratio. Indicates the actual allocated computing resources, represents the average task processing rate, represents the average waiting delay of the task flow, Represents the task loss rate.

[0074] The specific constraints are:

[0075]

[0076] Where, represents the probability of bandwidth resource overflow, Indicates the preset bandwidth overflow probability upper limit threshold, Indicates the upper threshold of the average waiting delay of the preset task flow. Indicates the upper threshold of the preset task loss rate. Indicates the task flow processing delay, Indicates the upper threshold of the preset task flow processing delay.

[0077] Step 203: Use the real-time task data input bandwidth and computing power allocation model to perform iterative solution to obtain the optimal bandwidth and computing power allocation solution that meets the constraints.

[0078] It should be noted that before using the Lagrange multiplier method or reinforcement learning for iterative solution, since the constraints within the constraint conditions directly or indirectly affect the parameters within the objective function, it is necessary to convert the above constraints into dynamic constraint expressions. By inputting real-time task data into the bandwidth and computing power allocation model to update the constraint boundaries, the dynamic constraint expressions are made to correspond to the final iterative solution variables, as follows:

[0079] Bandwidth overflow probability constraint:

[0080] Real-time task data includes the total bandwidth and available bandwidth of the task flow.

[0081] Bandwidth overflow probability Bandwidth allocation ratio , comprehensive arrival rate of tasks , when bandwidth allocated to high priority tasks (Total bandwidth of task flow ) is insufficient to handle its traffic, the probability of bandwidth overflow increases.

[0082] when When, it means , and transform the inequality to obtain the dynamic constraint expression of bandwidth overflow probability constraint:

[0083]

[0084] Where, Indicates available bandwidth.

[0085] Task flow average waiting delay constraint:

[0086] Real-time task data includes task processing rate.

[0087] According to the M / M / 1 queuing model, the average waiting delay of high-priority tasks , is the average arrival rate of tasks, is the task processing rate, and ,when When, it means , and transform the inequality to obtain the dynamic constraint expression of the task flow average waiting delay constraint:

[0088]

[0089] Task flow processing delay constraints:

[0090] Real-time task data includes task data volume.

[0091] Task flow processing delay is related to task processing rate , and because , that is ,when When, it means , and transform the inequality to obtain the dynamic constraint expression of the task flow processing delay constraint:

[0092]

[0093] Task loss rate constraint:

[0094] Real-time task data includes the average arrival rate of high-priority tasks, the average arrival rate of low-priority tasks, the loss rate caused by network transmission packet loss, and the total task processing rate.

[0095] It should be noted that the task loss rate consists of two parts: one is the loss rate caused by packet loss during network transmission, and the other is the probability of backlog in the system and eventually being discarded due to insufficient task processing capacity (task arrival rate is greater than processing rate).

[0096] Comprehensive arrival rate of tasks , is the average arrival rate of high priority tasks, is the average arrival rate of low-priority tasks, and the loss rate due to insufficient processing capacity is , therefore, the total task loss rate , and satisfies .

[0097] Furthermore, a real-time task data input bandwidth and computing power allocation model is adopted, and the Lagrange multiplier method or reinforcement learning is used for iterative solution.

[0098] In the embodiment of the present invention, the bandwidth and computing power allocation model of real-time task data input is adopted, and the Lagrange multiplier method or reinforcement learning is used for iterative solution to iteratively adjust the bandwidth allocation ratio. and represents the actual allocated computing resources , until satisfied Maximize the average waiting time of the task flow , task loss rate Once the indicators are met, the bandwidth allocation ratio that meets the indicators will be and represents the actual allocated computing resources As the optimal bandwidth and computing power allocation solution.

[0099] It should be noted that the Lagrange multiplier method or reinforcement learning is a conventional algorithm and will not be described in detail here.

[0100] Step 204: When the iterative solution result does not satisfy the constraint conditions, a priority factor is introduced or a preset upper threshold of each constraint is adjusted based on the constraint conditions.

[0101] The priority factor refers to a key parameter used to measure the weight relationship between high-priority and low-priority tasks in resource allocation. Its essence is the resource allocation tendency coefficient calculated based on the task flow distribution characteristics. It reflects the priority of high-priority tasks relative to low-priority tasks in bandwidth and computing power allocation. The larger the priority factor, the higher the resource allocation weight obtained by the high-priority task.

[0102] Furthermore, the priority factor is obtained as follows:

[0103] Use real-time task data to determine the overall task arrival rate;

[0104] Furthermore, the real-time task data includes an average arrival rate of high-priority tasks and an average arrival rate of low-priority tasks. The real-time task data is used to determine the comprehensive arrival rate of tasks, including:

[0105] The average arrival rate of high-priority tasks and the average arrival rate of low-priority tasks are summed to obtain the comprehensive arrival rate of tasks.

[0106] The comprehensive arrival rate of tasks is:

[0107]

[0108] Construct the task flow arrival distribution function based on the comprehensive task arrival rate;

[0109] The task flow arrival distribution function is specifically:

[0110]

[0111] Where, Represents the task flow arrival distribution function.

[0112] Perform Laplace transform on the task flow arrival distribution function to obtain the transformed task flow arrival distribution function ;

[0113] It should be noted that the task flow distribution function describes relevant characteristics such as the time interval between task arrivals. The Laplace transform is performed on it in order to analyze and calculate the task flow characteristics in a space that is more convenient to process, such as the frequency domain.

[0114] The transformed task flow arrival distribution function is integrated using the method of integration by parts to obtain the first integral;

[0115] In the embodiment of the present invention, the method of integration by parts is used to perform an integral operation on the transformed task flow arrival distribution function through mathematical tools, and the first integral .

[0116] Performing an integral operation on the transformed task flow arrival distribution function to obtain a second integral;

[0117] In the embodiment of the present invention, the second integral can be obtained by performing an integral operation on the transformed task flow arrival distribution function through mathematical tools. .

[0118] A priority factor is obtained by performing a ratio operation on the first integral and the second integral.

[0119] In the embodiment of the present invention, the first integral and the second integral are used to perform a ratio operation to obtain the priority factor .

[0120] In an embodiment of the present invention, when the iterative solution result does not meet the constraint condition, a priority factor may be introduced into the dynamic constraint expression of the bandwidth overflow probability constraint. The dynamic constraint expression introducing the priority factor is specifically:

[0121] Divide the bandwidth allocation ratio into high priority task bandwidth allocation ratio and low-priority task bandwidth allocation ratio , that is ;

[0122] Based on the priority factor, the average arrival rate of high-priority tasks and the average arrival rate of low-priority tasks, a bandwidth allocation ratio function for high-priority tasks and a bandwidth allocation ratio function for low-priority tasks are obtained;

[0123] The bandwidth allocation ratio function of high-priority tasks is specifically:

[0124]

[0125] This means that the larger the priority factor, the higher the proportion of the total bandwidth allocation that high-priority tasks receive.

[0126] The bandwidth allocation ratio function of low-priority tasks is specifically:

[0127]

[0128] A bandwidth allocation ratio function for low-priority tasks is set to limit the bandwidth upper limit of low-priority tasks.

[0129] Combined with the dynamic constraint expression without introducing the priority factor, and satisfying , , then the dynamic constraint expression that introduces the priority factor is specifically:

[0130]

[0131] It is worth mentioning that by introducing the bandwidth overflow probability constraint after the priority factor, while considering the bandwidth overflow probability constraint, the priority factor is used to achieve differentiation in bandwidth allocation between high-priority and low-priority tasks, ensuring that high-priority tasks can obtain bandwidth resources that match their priority.

[0132] Alternatively, the preset upper thresholds of each constraint are adjusted, including the preset bandwidth overflow probability upper threshold, the preset task flow average waiting delay upper threshold, the preset task flow processing delay upper threshold, and the preset task loss rate upper threshold.

[0133] Step 205 : Jump to the step of performing iterative solution using the real-time task data input bandwidth and computing power allocation model until the iterative solution result meets the constraint conditions.

[0134] In an embodiment of the present invention, the step of performing iterative solution using the real-time task data input bandwidth and computing power allocation model is jumped to execution until the iterative solution result meets the constraint condition.

[0135] In the present invention, maximizing system resource efficiency can achieve refined allocation and intensive utilization of resources, and minimizing the sum of the average waiting delay of task flows and the task loss rate can improve the service quality and reliability of the edge computing system. A bandwidth and computing power allocation model is constructed with maximizing system resource efficiency and minimizing the sum of the average waiting delay of task flows and the task loss rate as the dual optimization goals, and constraints are set. The feasible domain of resource allocation is limited from three dimensions: bandwidth overflow probability constraint, task flow delay constraint, and task loss rate constraint. After the real-time task data collected in real time is input into the bandwidth and computing power allocation model, a preset iterative algorithm is used to solve it. In each iteration, the algorithm adjusts the bandwidth and computing power allocation parameters according to the quality of the current solution, and gradually approaches the optimal bandwidth and computing power allocation scheme that meets all constraints, so that resources can be reasonably allocated in a multi-task environment, ensuring that the bandwidth and computing power requirements of different tasks are met. The present invention can avoid idle or excessive resource occupation through precise bandwidth and computing power allocation, improve overall resource utilization efficiency, and solve the technical problem of how to improve the dynamic allocation utilization rate of bandwidth and computing power resources in an edge computing environment.

[0136] See also Figure 3 , Figure 3 This is a structural block diagram of a bandwidth and computing power allocation system for edge computing provided in Example 3 of the present invention.

[0137] The present invention provides an edge computing bandwidth and computing power allocation system, comprising:

[0138] Acquisition module 301, used to obtain real-time task data of the edge computing system;

[0139] A construction module 302 is configured to construct a bandwidth and computing power allocation model and set constraints with the optimization goal of maximizing system resource efficiency and minimizing the sum of the average waiting delay of the task flow and the task loss rate;

[0140] The solution module 303 is used to use the real-time task data input bandwidth and computing power allocation model to perform iterative solution to obtain the optimal bandwidth and computing power allocation solution that meets the constraint conditions.

[0141] Furthermore, the constraints include bandwidth overflow probability constraint, task flow delay constraint and task loss rate constraint.

[0142] Furthermore, the solution module 303 includes:

[0143] The algorithm solving submodule is used to use the real-time task data input bandwidth and computing power allocation model, and use the Lagrange multiplier method or reinforcement learning for iterative solution.

[0144] Furthermore, it also includes:

[0145] An adjustment module is used to introduce a priority factor or adjust a preset upper threshold of each constraint based on the constraint conditions when the iterative solution result fails to meet the constraint conditions;

[0146] The jump module is used to jump to the step of iteratively solving the problem using the real-time task data input bandwidth and computing power allocation model until the iterative solution result meets the constraint conditions.

[0147] Furthermore, the priority factor is obtained as follows:

[0148] Use real-time task data to determine the overall task arrival rate;

[0149] Construct the task flow arrival distribution function based on the comprehensive task arrival rate;

[0150] Performing Laplace transform on the task flow arrival distribution function to obtain the transformed task flow arrival distribution function;

[0151] The transformed task flow arrival distribution function is integrated using the method of integration by parts to obtain the first integral;

[0152] Performing an integral operation on the transformed task flow arrival distribution function to obtain a second integral;

[0153] A priority factor is obtained by performing a ratio operation on the first integral and the second integral.

[0154] Furthermore, the real-time task data includes an average arrival rate of high-priority tasks and an average arrival rate of low-priority tasks. The real-time task data is used to determine the comprehensive arrival rate of tasks, including:

[0155] The average arrival rate of high-priority tasks and the average arrival rate of low-priority tasks are summed to obtain the comprehensive arrival rate of tasks.

[0156] In the present invention, maximizing system resource efficiency can achieve refined allocation and intensive utilization of resources, and minimizing the sum of the average waiting delay of task flows and the task loss rate can improve the service quality and reliability of the edge computing system. A bandwidth and computing power allocation model is constructed with maximizing system resource efficiency and minimizing the sum of the average waiting delay of task flows and the task loss rate as the dual optimization goals, and constraints are set. The feasible domain of resource allocation is limited from three dimensions: bandwidth overflow probability constraint, task flow delay constraint, and task loss rate constraint. After the real-time task data collected in real time is input into the bandwidth and computing power allocation model, a preset iterative algorithm is used to solve it. In each iteration, the algorithm adjusts the bandwidth and computing power allocation parameters according to the quality of the current solution, and gradually approaches the optimal bandwidth and computing power allocation scheme that meets all constraints, so that resources can be reasonably allocated in a multi-task environment, ensuring that the bandwidth and computing power requirements of different tasks are met. The present invention can avoid idle or excessive resource occupation through precise bandwidth and computing power allocation, improve overall resource utilization efficiency, and solve the technical problem of how to improve the dynamic allocation utilization rate of bandwidth and computing power resources in an edge computing environment.

[0157] See also Figure 4 , Figure 4 This is a structural block diagram of a computer device provided in Example 4 of the present invention.

[0158] An electronic device according to an embodiment of the present invention includes: a memory 401 and a processor 402, wherein the memory 401 stores a computer program; when the computer program is executed by the processor 402, the processor 402 executes the bandwidth and computing power allocation method for edge computing as in any of the above embodiments.

[0159] Memory 401 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Memory 401 has storage space 403 for program code 413 for executing any of the method steps described above. For example, storage space 403 for program code may include individual program codes 413 for implementing various steps in the method described above. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards, or floppy disks. The program codes may be compressed, for example, in a suitable format. When executed by a processing device, these codes cause the processing device to execute the various steps in the method described above. These program codes may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, compact disks (CDs), memory cards, or floppy disks. The program codes may be compressed, for example, in a suitable format. When these codes are executed by a computing processing device, they cause the computing processing device to execute the various steps in the edge computing bandwidth and computing power allocation method described above.

[0160] Embodiment 5 of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the bandwidth and computing power allocation method for edge computing as in any of the above embodiments is implemented.

[0161] Embodiment 6 of the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the bandwidth and computing power allocation method for edge computing as described in any of the above embodiments.

[0162] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0163] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0164] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0165] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0166] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0167] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for allocating bandwidth and computing power for edge computing, characterized in that: include: Obtain real-time task data from edge computing systems; With the optimization goals of maximizing system resource efficiency and minimizing the sum of the average waiting delay of the task flow and the task loss rate, a bandwidth and computing power allocation model is constructed and constraints are set; The real-time task data is input into the bandwidth and computing power allocation model for iterative solution to obtain an optimal bandwidth and computing power allocation solution that meets the constraint conditions.

2. The edge computing bandwidth and computing power allocation method according to claim 1 is characterized in that: The constraints include bandwidth overflow probability constraint, task flow delay constraint and task loss rate constraint.

3. The bandwidth and computing power allocation method for edge computing according to claim 1, characterized in that: The step of inputting the real-time task data into the bandwidth and computing power allocation model for iterative solution includes: The real-time task data is input into the bandwidth and computing power allocation model, and an iterative solution is performed using the Lagrange multiplier method or reinforcement learning.

4. The bandwidth and computing power allocation method for edge computing according to claim 2, characterized in that: Also includes: When the iterative solution result does not meet the constraint conditions, a priority factor is introduced or a preset upper limit threshold of each constraint is adjusted based on the constraint conditions; Jump to the step of using the real-time task data to input the bandwidth and computing power allocation model for iterative solution until the iterative solution result meets the constraint condition.

5. The edge computing bandwidth and computing power allocation method according to claim 4 is characterized in that: The priority factor is obtained in the following way: Determining a comprehensive task arrival rate using the real-time task data; Constructing a task flow arrival distribution function according to the comprehensive arrival rate of the tasks; Performing Laplace transform on the task flow arrival distribution function to obtain a transformed task flow arrival distribution function; Performing an integration operation on the transformed task flow arrival distribution function using the method of integration by parts to obtain a first integral; performing an integral operation on the transformed task flow arrival distribution function to obtain a second integral; A priority factor is obtained by performing a ratio operation on the first integral and the second integral.

6. The edge computing bandwidth and computing power allocation method according to claim 1, characterized in that: The real-time task data includes an average arrival rate of high-priority tasks and an average arrival rate of low-priority tasks. Determining the comprehensive arrival rate of tasks using the real-time task data includes: The average arrival rate of the high-priority tasks and the average arrival rate of the low-priority tasks are used to perform a sum operation to obtain a comprehensive task arrival rate.

7. A bandwidth and computing power allocation system for edge computing, characterized in that: include: Acquisition module, used to obtain real-time task data of edge computing system; A construction module is used to build a bandwidth and computing power allocation model and set constraints with the optimization goals of maximizing system resource efficiency and minimizing the sum of the average waiting delay and task loss rate of the task flow; The solution module is used to use the real-time task data to input the bandwidth and computing power allocation model for iterative solution to obtain the optimal bandwidth and computing power allocation solution that meets the constraint conditions.

8. An electronic device, characterized in that: It includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the processor executes the steps of the bandwidth and computing power allocation method for edge computing as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the bandwidth and computing power allocation method for edge computing as described in any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the edge computing bandwidth and computing power allocation method as described in any one of claims 1-6.

Images