Edge computing computing power arrangement method, device, equipment, medium and program product

Abstract

Embodiments of the present application disclose a kind of edge computing computing power arrangement method, device, equipment, medium and program product.It includes: obtaining the to-be-processed service flow information of each to-be-processed service flow;According to the service flow characteristics in each to-be-processed service flow information, each to-be-processed service flow is mapped to corresponding real-time service subqueue and non-real-time service subqueue;According to the to-be-processed data volume of each real-time service subqueue and computing power reservation error correction coefficient, determine real-time computing power resource reservation quantity, and determine real-time computing power arrangement scheme according to real-time computing power resource reservation quantity;According to the to-be-processed data volume of each non-real-time service subqueue, current remaining computing power resource and real-time computing power resource reservation quantity, determine non-real-time computing power arrangement scheme;Wherein, computing power reservation error correction coefficient is determined according to the last actual computing power resource occupation and the last real-time computing power resource reservation quantity in the last business cycle corresponding to the current business cycle in the last business cycle.It improves the computing power resource utilization of edge computing device.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things, and in particular to an edge computing computing power arrangement method and device, equipment, medium and program product. BACKGROUND

[0002] The Internet of Things is an extension application and network extension of communication networks and the Internet. It uses sensing technology and intelligent equipment to identify the physical world, transmits information through the network, and performs calculation, processing and knowledge mining. The purpose is to realize information interaction and seamless connection between people and things, and between things, and to implement control, accurate management and scientific decision-making of the physical world. The Internet of Things has been widely applied in the power field and formed the power Internet of Things.

[0003] As a new computing mode, edge computing can enable data to be processed in a timely and effective manner in the edge network near the source. With the rapid development of the Internet of Things, the explosive growth of data volume far exceeds the bearing capacity of network bandwidth, and the emergence of a large number of new intelligent applications has put forward higher requirements for system latency performance. In the edge computing mode, the terminal offloads tasks to the nearby edge server for execution, effectively alleviating the network congestion and high delay problems caused by transmitting all data to the data center in traditional cloud computing.

[0004] However, the computing, communication and storage resources of edge computing devices are limited. In order to realize efficient and flexible business processing and decision-making, the terminals, data and businesses under the new business ecology of the power Internet of Things are increasing, and massive power data need to be analyzed and calculated on edge computing devices, which will make the resources of edge computing devices tend to be saturated, resulting in a serious decline in service quality. At the same time, due to the dynamic change characteristics of new power system businesses, the edge computing resource scheduling mode has changed from fixed configuration mode according to determined business objects to elastic configuration mode according to dynamic business objects. Therefore, the contradiction between the increasing business processing demand and the limited computing power resources is increasingly prominent, which reduces the business processing capacity of the power Internet of Things network. SUMMARY

[0005] The present application provides an edge computing computing power arrangement method, device, equipment, medium and program product. Different computing power arrangement schemes are formulated for real-time businesses and non-real-time businesses in different computing power arrangement modes according to different types of business flows, so that the dynamic change of business demand and actual business processing state can be fully considered for real-time businesses and non-real-time businesses, and a flexible and efficient computing power arrangement scheme suitable for edge computing devices is obtained, and the utilization rate of computing power resources of edge computing devices is improved.

[0006] In a first aspect, an edge computing computing power arrangement method is provided, comprising:

[0007] obtain to-be-processed service flow information of each to-be-processed service flow in a current service period;

[0008] map each to-be-processed service flow into a corresponding real-time service sub-queue and a non-real-time service sub-queue according to a service flow feature in each to-be-processed service flow information;

[0009] determine a real-time computing resource reservation amount according to a to-be-processed data amount of each real-time service sub-queue and a computing resource reservation error correction coefficient, and determine a real-time computing arrangement scheme according to the real-time computing resource reservation amount;

[0010] determine a non-real-time computing arrangement scheme according to a to-be-processed data amount of each non-real-time service sub-queue, a current residual computing resource, and the real-time computing resource reservation amount;

[0011] The computing resource reservation error correction coefficient is determined according to an actual computing resource occupation amount in a previous service period corresponding to the current service period and a previous real-time computing resource reservation amount.

[0012] In a second aspect, an embodiment of the present application further provides an edge computing computing arrangement device, comprising:

[0013] an information obtaining module, configured to obtain to-be-processed service flow information of each to-be-processed service flow in a current service period;

[0014] a queue mapping module, configured to map each to-be-processed service flow into a corresponding real-time service sub-queue and a non-real-time service sub-queue according to a service flow feature in each to-be-processed service flow information;

[0015] a first scheme determining module, configured to determine a real-time computing resource reservation amount according to a to-be-processed data amount of each real-time service sub-queue and a computing resource reservation error correction coefficient, and determine a real-time computing arrangement scheme according to the real-time computing resource reservation amount;

[0016] a second scheme determining module, configured to determine a non-real-time computing arrangement scheme according to a to-be-processed data amount of each non-real-time service sub-queue, a current residual computing resource, and the real-time computing resource reservation amount;

[0017] The computing resource reservation error correction coefficient is determined according to an actual computing resource occupation amount in a previous service period corresponding to the current service period and a previous real-time computing resource reservation amount.

[0018] In a third aspect, an embodiment of the present application further provides an edge computing computing arrangement device, comprising:

[0019] at least one processor; and a memory connected with the at least one processor in communication;

[0020] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to implement the edge computing computing power arrangement method of any one of the embodiments of the application.

[0021] In a fourth aspect, the embodiments of the application further provide a storage medium containing computer executable instructions for executing the edge computing computing power arrangement method of any one of the embodiments of the application when executed by a computer processor.

[0022] In a fifth aspect, the embodiments of the application further provide a computer program product comprising a computer program for executing the edge computing computing power arrangement method of any one of the embodiments of the application when executed by a processor.

[0023] The edge computing computing power arrangement method, device, equipment, medium and program product provided by the embodiment of the application, by obtaining the to-be-processed service flow information of each to-be-processed service flow in the current service period; according to the service flow characteristics in each to-be-processed service flow information, each to-be-processed service flow is mapped to the corresponding real-time service sub-queue and non-real-time service sub-queue; according to the to-be-processed data amount of each real-time service sub-queue and the computing power reservation error correction coefficient, the real-time computing power resource reservation amount is determined, and the real-time computing power arrangement scheme is determined according to the real-time computing power resource reservation amount; according to the to-be-processed data amount of each non-real-time service sub-queue, the current remaining computing power resource and the real-time computing power resource reservation amount, the non-real-time computing power arrangement scheme is determined; wherein the computing power reservation error correction coefficient is determined according to the last actual computing power resource occupation amount and the last real-time computing power resource reservation amount in the last service period corresponding to the current service period. By adopting the above technical scheme, when it is necessary to provide corresponding computing power resources for each to-be-processed service flow in the current service period, first, whether the different to-be-processed service flows belong to real-time type services is distinguished and service queue mapping is performed according to the characteristics of the different to-be-processed service flows, and different computing power arrangement methods are used for real-time type services and non-real-time type services to determine the computing power arrangement scheme. Among them, in order to ensure that the real-time type service can have enough computing power resources for execution, when the computing power arrangement scheme is determined, the real-time computing power arrangement scheme will be determined first, so as to complete the reservation of the real-time computing power resource reservation amount required for the real-time service in the remaining computing power resources in the edge computing device according to the determined real-time computing power arrangement scheme; and for non-real-time type services, the current remaining available computing power resources in the edge computing device and the real-time computing power reservation resource amount are used for determination, which ensures that the real-time task can be executed in time, and at the same time, the remaining computing power resources in the edge computing device can be fully utilized. At the same time, the computing power reservation error correction coefficient determined according to the actual service execution in the last service period is used to assist the determination of the real-time computing power arrangement scheme, so that the final obtained real-time computing power arrangement scheme can fully consider the dynamically changing service demand and service actual processing state, and then the non-real-time computing power arrangement scheme determined based on the real-time computing power resource reservation amount can also fully consider the dynamically changing service demand and service actual processing state, so that a more flexible and efficient computing power arrangement scheme suitable for the edge computing device is finally obtained, and the computing power resource utilization rate of the edge computing device is improved.

[0024] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the application, nor is it used to limit the scope of the application. Other features of the application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to make the technical solutions in the embodiments of the present application clearer, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0026] Figure 1 A flow chart of an edge computing computing power arrangement method provided for the first embodiment of the present application is shown in FIG. 1.

[0027] Figure 2 A flow chart of an edge computing computing power arrangement method provided for the second embodiment of the present application is shown in FIG. 2.

[0028] Figure 3 A structural schematic diagram of an edge computing computing power arrangement device provided for the third embodiment of the present application is shown in FIG. 3.

[0029] Figure 4 A structural schematic diagram of an edge computing computing power arrangement device provided for the fourth embodiment of the present application is shown in FIG. 4. DETAILED DESCRIPTION

[0030] In order to make the technical solutions in the embodiments of the present application clearer, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort on the basis of these drawings.

[0031] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] Embodiment one

[0033] Figure 1A flowchart of an edge computing computing power arrangement method provided for the first embodiment of the present application. The first embodiment of the present application can be applicable to the case of planning and allocating computing power resources for edge computing devices in the power Internet of Things. The method can be executed by an edge computing computing power arrangement device, which can be realized by software and / or hardware and can be configured in an edge computing computing power arrangement device. Optionally, the edge computing computing power arrangement device can be an electronic device, which can be a notebook, a desktop computer, a smart tablet, an edge computing device in the power Internet of Things, etc. The present application does not limit this.

[0034] As shown in Figure 1 , the edge computing computing power arrangement method provided by the present application specifically includes the following steps:

[0035] S101, obtaining the to-be-processed service flow information of each to-be-processed service flow in the current service period.

[0036] In the present embodiment, the service period can be specifically understood as the time period required from the beginning to the completion of a service, or one time period when the complete time required from the beginning to the completion of a service is divided into multiple time periods according to certain rules. The current service period can be specifically understood as the service period in which the current time is located. The to-be-processed service flow can be specifically understood as the service data input to the edge computing device through the service interface and needing to be processed by the computing power resources in the edge computing device. The to-be-processed service flow information can be specifically understood as the information representing the related characteristics of the to-be-processed service flow.

[0037] Specifically, in order to make full use of the computing power resources of the edge computing device in the edge computing process, the corresponding computing power resources are allocated to the to-be-processed service before the service flow to be executed is input to the edge computing device. The multiple to-be-processed service flows accessed to the edge computing device through each service interface in the current service period can be counted. Each to-be-processed service flow can be generated by a power terminal device accessing the edge computing device, so that the to-be-processed service flow information representing the related characteristics of the to-be-processed service flow is obtained at the same time.

[0038] S102, mapping each to-be-processed service flow to the corresponding real-time service sub-queue and non-real-time service sub-queue according to the service flow characteristics in each to-be-processed service flow information.

[0039] In the embodiment, the service flow feature can be understood as specific or potential meaningful information in a certain aspect of the service flow. Optionally, for the service flow in the power Internet of Things, the service flow feature corresponding thereto can be represented by a plurality of feature parameters, for example, the feature parameters can include maximum packet, minimum packet, average packet size, average arrival time, interval time mean, service flow size, service flow duration, and protocol flag bit, etc., and the embodiment of the application does not limit this. The real-time service sub-queue can be understood as a service queue containing real-time type to-be-processed service flows. The non-real-time service sub-queue can be understood as a service queue containing non-real-time type to-be-processed service flows. It can be understood that a plurality of application software contained in the edge computing device can be divided into real-time type application software and non-real-time type application software according to the service type that can be processed, each application software can correspond to a service sub-queue, and the service flow mapped to the service sub-queue is the service required to be processed by the application software; that is, the service sub-queue corresponding to the real-time type application software is the real-time service sub-queue, and the service sub-queue corresponding to the non-real-time type application software is the non-real-time service sub-queue; the edge computing device can contain a plurality of real-time service sub-queues and a plurality of non-real-time service sub-queues at the same time according to the number of application software; since a plurality of to-be-processed service flows can be processed by the same application software in the edge computing device, the plurality of to-be-processed service flows can be mapped to the same real-time service sub-queue or non-real-time service sub-queue, and the order of each to-be-processed service flow in the corresponding service sub-queue is determined according to the mapping order.

[0040] Specifically, the service flow feature corresponding to each to-be-processed service flow is extracted from each to-be-processed service flow information, and whether the to-be-processed service flow needs real-time processing or not and which software in the edge computing device the to-be-processed service flow should belong to can be determined based on the service flow feature, and then the to-be-processed service flow can be mapped to the service sub-queue corresponding to the software after the software to which the to-be-processed service flow should be allocated is determined, the service sub-queue corresponding to the software for real-time processing is determined as the real-time service sub-queue, and the service sub-queue corresponding to the software for non-real-time processing is determined as the non-real-time service sub-queue.

[0041] S103, determining a real-time computing resource reservation quantity according to the to-be-processed data quantity of each real-time service sub-queue and a computing resource reservation error correction coefficient, and determining a real-time computing arrangement scheme according to the real-time computing resource reservation quantity.

[0042] The computing resource reservation error correction coefficient is determined according to the last actual computing resource occupation quantity and the last real-time computing resource reservation quantity in the last service period corresponding to the current service period.

[0043] In the embodiment, the to-be-processed data amount can be specifically understood as the data amount in the same service sub-queue that needs to be processed by the corresponding software in the edge computing device, that is, the data amount that needs to be processed by using the computing power resources in the edge computing device. The computing power reservation error correction coefficient can be specifically understood as a correction value for adjusting the estimated required reservation amount of computing power resources for real-time type services in the current period based on the computing power occupation in the processing of real-time type services in the last service period. The real-time computing power resource reservation amount can be specifically understood as the amount of computing power resources that the edge computing device needs to reserve for the to-be-processed service flow that needs real-time processing in the current service period. The real-time computing power arrangement scheme can be specifically understood as a scheme for arranging and allocating the required computing power resources for the software corresponding to each real-time service sub-queue in the current service period.

[0044] In the embodiment, the last service period can be specifically understood as a service period before the current service period. It can be understood that the length of the last service period can be consistent with or inconsistent with the length of the current service period, and the embodiment of the application does not limit this. The last actual computing power resource occupation amount can be specifically understood as the amount of computing power resources actually consumed by the real-time type software or each real-time service sub-queue in the process of service processing in the last service period. The last real-time computing power resource reservation amount can be specifically understood as the real-time computing power resource reservation amount determined when processing the last service period.

[0045] Specifically, the to-be-processed data amount of each real-time service sub-queue is comprehensively counted and converted into the amount of computing power resources that the edge computing device needs to provide for processing the data amount contained in the to-be-processed service flow contained in each real-time service sub-queue. Then, the error correction of the determined computing power resource amount is performed based on the computing power reservation error correction coefficient, and the real-time computing power resource reservation amount that needs to be reserved by the edge computing device for real-time type services in the current service period is finally obtained. Then, the real-time computing power resource reservation amount is divided based on the demand of each real-time service sub-queue for computing power resources, and the real-time computing power arrangement scheme for the software corresponding to each real-time service sub-queue is obtained.

[0046] S104, determining a non-real-time computing power arrangement scheme according to the to-be-processed data amount of each non-real-time service sub-queue, the current remaining computing power resource, and the real-time computing power resource reservation amount.

[0047] In the embodiment, the current remaining computing power resource can be specifically understood as the amount of remaining available computing power resources in the edge computing device when entering the current service period. The non-real-time computing power arrangement scheme can be specifically understood as a scheme for allocating the required computing power resources for the software corresponding to each non-real-time service sub-queue in the current service period.

[0048] Specifically, the amount of computing resource that can be provided for non-real-time type service processing in the current service period can be determined according to the current remaining computing resource and the real-time computing resource reservation amount. Further, the amount of computing resource required by the non-real-time type service flow in the current service period can be determined based on the amount of data to be processed of each non-real-time service sub-queue. Since the execution timeliness requirement of the non-real-time service is not high, the amount of computing resource required by the non-real-time type service flow can not be provided at one time. At this time, the processing of the amount of computing resource required by the non-real-time type service flow in the current service period can be completed in batches based on the amount of computing resource that can be provided for non-real-time type service processing in the current service period, and the corresponding non-real-time computing resource arrangement scheme is obtained.

[0049] The technical scheme of the embodiment comprises the following steps: obtaining to-be-processed service flow information of each to-be-processed service flow in a current service period; mapping each to-be-processed service flow into a corresponding real-time service sub-queue and a non-real-time service sub-queue according to service flow characteristics in each to-be-processed service flow information; determining a real-time computing resource reservation amount according to a to-be-processed data amount of each real-time service sub-queue and a computing resource reservation error correction coefficient, and determining a real-time computing arrangement scheme according to the real-time computing resource reservation amount; determining a non-real-time computing arrangement scheme according to a to-be-processed data amount of each non-real-time service sub-queue, a current residual computing resource and the real-time computing resource reservation amount; wherein the computing resource reservation error correction coefficient is determined according to an actual computing resource occupation amount in a previous service period corresponding to the current service period and a previous real-time computing resource reservation amount. By using the above technical scheme, when computing resources need to be provided for each to-be-processed service flow in the current service period, first, whether the different to-be-processed service flows belong to real-time type services is distinguished and service queue mapping is performed according to the characteristics of the different to-be-processed service flows, and different computing arrangement methods are used to determine the computing arrangement scheme for real-time type services and non-real-time type services. In order to ensure that real-time type services can have sufficient computing resources for execution, when the computing arrangement scheme is determined, the real-time computing arrangement scheme is determined first, so that the real-time computing resource reservation amount required for real-time services can be reserved in priority in the residual computing resources in the edge computing device according to the determined real-time computing arrangement scheme; and the non-real-time type services are determined according to the current residual available computing resources in the edge computing device and the real-time computing resource reservation amount, so that the residual computing resources in the edge computing device can be fully utilized while ensuring that real-time tasks can be executed in time. At the same time, the computing resource reservation error correction coefficient determined according to the actual service execution in the previous service period is used to assist the determination of the real-time computing arrangement scheme, so that the final obtained real-time computing arrangement scheme can fully consider the dynamically changing service demand and service actual processing state, and thus the non-real-time computing arrangement scheme determined based on the real-time computing resource reservation amount can also fully consider the dynamically changing service demand and service actual processing state, and finally a more flexible and efficient computing arrangement scheme suitable for the edge computing device is obtained, and the computing resource utilization rate of the edge computing device is improved.

[0050] Embodiment two

[0051] Figure 2A flowchart of an edge computing computing power arrangement method provided for the second embodiment of the present application is shown in the figure. Based on the above-mentioned optional technical solutions, the second embodiment of the present application is further optimized. After the mapping of each to-be-processed service flow to the corresponding service sub-queue is completed, the total real-time computing power demand is obtained by statistically converting the to-be-processed data flow of each real-time service sub-queue. The corresponding real-time computing power resource reservation amount is obtained by correcting the computing power reservation error correction coefficient. Then, the real-time computing power arrangement scheme is determined according to the proportion of the data amount corresponding to different real-time service sub-queues in the total real-time demand data amount. After the determination of the real-time computing power resource reservation amount is completed, the non-real-time available computing power resource available for non-real-time type services can be determined according to the current remaining computing power resource and the real-time computing power resource reservation amount. On the basis of ensuring that the real-time type services can be processed with sufficient computing power resources, the non-real-time available computing power resource is fully utilized according to the non-real-time delay threshold required by the non-real-time type services combined with data aggregation processing, and the non-real-time computing power arrangement scheme is obtained. The current remaining computing power resource in the current service period can be fully utilized. And at the end of each service period, the computing power reservation error correction coefficient can be re-determined, and the determined computing power reservation error correction coefficient is updated according to the delay during the processing of the real-time type services in the current period, so that the finally determined computing power reservation error correction coefficient can better reflect the dynamically changing business demand and business actual processing state, and thus the computing power arrangement schemes determined according to the computing power reservation error correction coefficient can be more suitable for the processing of edge computing devices, and the computing power resource utilization rate of edge computing devices is improved.

[0052] As shown in the figure, the edge computing computing power arrangement method provided by the embodiment of the present application specifically includes the following steps: Figure 2

[0053] S201, obtaining to-be-processed service flow information of each to-be-processed service flow in a current service period.

[0054] S202, inputting each to-be-processed service flow information into a pre-trained service type determination model to extract service flow features of each to-be-processed service flow information through the service type determination model, and performing service classification operation according to each service flow feature to determine the service type corresponding to each to-be-processed service flow.

[0055] Among them, the service type determination model is a lightweight long short-term memory network (Naive Long Short-Term Memory, Naive-LSTM) model.

[0056] ​In the embodiment, the business type determination model can be understood as a neural network model used to classify the business type of the business logistics information according to the business characteristics in the input business logistics information. For example, the business type can include a real-time business type and a non-real-time business type.

[0057] Specifically, the business flow information corresponding to each to-be-processed business flow is input into the pre-trained business type determination model. The business type determination model first extracts features from the to-be-processed business flow information, and then analyzes the to-be-processed business flow corresponding to the to-be-processed business flow information according to the extracted features, to determine whether the to-be-processed business flow needs real-time processing, and classifies the to-be-processed business flow according to the determination result. The to-be-processed business flow is divided into a real-time business type and a non-real-time business type. If it is determined that the to-be-processed business flow is a real-time business type, step S203 is performed; if it is determined that the to-be-processed business flow is a non-real-time business type, step S204 is performed.

[0058] For example, the business type determination model can be a Naive-LSTM model composed of an input layer, an LSTM layer, and an output layer. The input layer is used to input the business flow features extracted from the to-be-processed business flow information. The type of the business flow features can be set to m classes. The LSTM layer can be composed of m*n Naive-LSTM units, that is, an LSTM layer has m characteristic values*n Naive-LSTM units. The output layer is used to process the data output by the LSTM layer via a softmax classification function to obtain the business type corresponding to the to-be-processed business flow information.

[0059] Since the LSTM layer simplifies the internal structure and internal parameters of the LSTM unit, the amount of data processing is reduced. The structure formula of Naive-LSTM is as follows:

[0060]

[0061] wherein x k is the input vector of the kth time; y k is the output vector of the kth time; z k is the input signal; i k is the input gate; c k is the state unit; o k is the output gate; W z is the input weight matrix in z k ; U z , U i , and U o are the recurrent weight matrices in z k , i k , and o k , respectively; and bz bias matrix; σ() is a sigmoid activation function; g() is a tanh activation function. k bias matrix; σ() is a sigmoid activation function; g() is a tanh activation function.

[0062] S203, mapping the to-be-processed service flow of the real-time service type to the corresponding real-time service sub-queue, and performing S205.

[0063] Specifically, after determining the service type corresponding to each to-be-processed service flow, if the service type of the to-be-processed service flow is a real-time service type, the software capable of processing the to-be-processed service flow can be selected from the software capable of processing real-time services contained in the edge computing device, and the to-be-processed service flow is written into the real-time service sub-queue corresponding to the software by mapping. After all to-be-processed service flows are mapped, S205 is performed.

[0064] It can be understood that the to-be-processed service flows of each service type are mapped into the corresponding real-time service sub-queue according to the output order of the service type determination model, but since the to-be-processed service flows in each real-time service sub-queue of the same service cycle are processed in the same batch, the mapping order has little effect on the processing of each to-be-processed service flow.

[0065] S204, mapping the to-be-processed service flow of the non-real-time service type to the corresponding non-real-time service sub-queue, and performing S205.

[0066] Specifically, after determining the service type corresponding to each to-be-processed service flow, if the service type of the to-be-processed service flow is a non-real-time service type, the software capable of processing the to-be-processed service flow can be selected from the software capable of processing non-real-time services contained in the edge computing device, and the to-be-processed service flow is written into the non-real-time service sub-queue corresponding to the software by mapping. After all to-be-processed service flows are mapped, S205 is performed.

[0067] S205, determining the total real-time demand data amount according to the to-be-processed data amount of each real-time service sub-queue.

[0068] Specifically, for each real-time service sub-queue, the to-be-processed data amount of the real-time service sub-queue can be determined according to the queue length, queue capacity, and queue processing speed of the real-time service sub-queue. The to-be-processed data amounts of all real-time service sub-queues are summed up, and the total data amount required for processing the real-time service type service in the current service cycle is obtained. The total data amount is determined as the total real-time demand data amount.

[0069] S206, determining the product of the total real-time demand data amount and the preset computing power conversion factor as the total real-time computing power demand amount.

[0070] In the embodiment, the preset computing power conversion factor can be specifically understood as a parameter determined in advance according to actual conditions, which is used to convert the corresponding data amount in the required processing service flow and the computing power resources that can be provided by the edge computing device. The total real-time computing power demand amount can be specifically understood as the amount of computing power resources that the edge computing device needs to provide in total for the to-be-processed service flow of the real-time service type in the current service period.

[0071] Specifically, in order to determine the amount of computing power resources provided by the edge computing device for the to-be-processed service flow of the real-time service type, after the total real-time demand data amount of the to-be-processed data flow of the real-time service type is determined, the total real-time demand data amount can be converted into a representation of the amount of computing power resources according to the preset computing power conversion factor provided in advance. At this time, the total real-time demand data amount can be multiplied by the preset amount conversion factor, and the product is determined as the total real-time computing power demand amount.

[0072] S207, determining the difference between the total real-time computing power demand amount and the computing power reservation error correction coefficient as the real-time computing power resource reservation amount.

[0073] Specifically, in the actual data processing process, the actual occupied computing power resource amount will fluctuate due to various factors. In order to make full use of the computing power resources in the edge computing device, the total real-time computing power demand amount determined for the current service period can be adjusted by using the computing power reservation error correction coefficient determined based on the last service period. In the embodiment, the difference between the total real-time computing power demand amount and the computing power reservation error correction coefficient is obtained, and the difference is determined as the real-time computing power resource reservation amount reserved and allocated for the real-time service type service flow in the current service period.

[0074] S208, distributing the real-time computing power resource reservation amount to the real-time service containers corresponding to each real-time service sub-queue according to the proportion of the to-be-processed data amount of each real-time service sub-queue in the total real-time demand data amount, and determining a real-time computing power arrangement scheme.

[0075] Specifically, since the edge computing device usually has sufficient computing power resources for the real-time service type service, the real-time computing power resource reservation amount is usually less than or equal to the available computing power resource amount of the edge device in the current service period, that is, the to-be-processed service flow in each real-time service sub-queue can be allocated sufficient computing power resources for processing. At this time, the real-time computing power resource reservation amount can be directly divided according to the proportion of the to-be-processed data amount of each real-time service sub-queue in the total real-time demand data amount, and each allocated computing power resource can be allocated to the real-time service container corresponding to each real-time service sub-queue, so as to determine the real-time computing power arrangement scheme.

[0076] It can be understood that in the case that the real-time computing resource reservation amount is greater than the available computing resource amount of the edge device in the current service period, the edge computing device will also give priority to guarantee the processing of the to-be-processed service flow of the real-time service type, at this time, the available computing resource amount in the current service period can be taken as the real-time computing resource reservation amount, and the determination of the real-time computing arrangement scheme is completed through the same division method as described above.

[0077] S209, determining the difference between the current remaining computing resource and the real-time computing resource reservation amount as the non-real-time available computing resource.

[0078] Specifically, since the edge computing device needs to give priority to guarantee the processing of the real-time service type in the service processing process. Therefore, in the current service period, the computing resource required by the real-time service type needs to be reserved first, and then the remaining computing resource is determined as the computing resource available for executing the non-real-time service type service, that is, the difference between the current remaining computing resource in the current service period and the real-time computing resource reservation amount is determined as the non-real-time available computing resource.

[0079] S210, determining the total non-real-time computing demand amount according to the to-be-processed data amount of each non-real-time service sub-queue and a preset computing conversion factor.

[0080] Specifically, for each non-real-time service sub-queue, the to-be-processed data amount of the non-real-time service sub-queue can be determined according to the queue length, queue capacity and queue processing speed and other queue characteristic parameters corresponding to the non-real-time service sub-queue. Then the to-be-processed data amounts of all non-real-time service sub-queues can be summed up, that is, the total data amount required for processing the non-real-time service type service in the current service period can be obtained. In order to determine the computing resource amount required by the edge computing device for the to-be-processed service flow of the non-real-time service type, the preset computing conversion factor can be used to convert it into a representation method of computing resource amount, that is, the total data amount required for processing the non-real-time service type service is multiplied by the preset computing conversion factor, and the product is determined as the total non-real-time computing demand amount.

[0081] S211, whether the total non-real-time computing demand amount is greater than the non-real-time available computing resource, if yes, performing S212; if no, performing S213.

[0082] Specifically, although the non-real-time service type service does not require real-time execution, real-time execution is still the optimal execution mode. Therefore, after the total non-real-time computing power demand is determined, the total non-real-time computing power demand can be compared with the non-real-time available computing power resources. If the comparison result is that the total non-real-time computing power demand is less than or equal to the non-real-time available computing power resources, it can be considered that the edge computing device still has enough computing power resources for the non-real-time service type service after providing computing power resources for the real-time service type service, that is, the real-time processing of the non-real-time service type service can be realized, and S213 is executed. If the comparison result is that the total non-real-time computing power demand is greater than the non-real-time available computing power resources, it can be considered that the remaining computing power resources are insufficient for the real-time processing of the non-real-time service type service after the edge computing device provides computing power resources for the real-time service type service, and S212 is executed.

[0083] S212, aggregate each non-real-time service sub-queue to generate an aggregated service block, and determine a non-real-time computing power arrangement scheme according to an aggregated data amount of the aggregated service block, the non-real-time available computing power resources, and a preset non-real-time delay threshold.

[0084] In the embodiment, the aggregated service block can be understood as a data block whose data amount is the sum of the data amounts corresponding to each aggregated to-be-processed service stream after the to-be-processed service streams in different non-real-time service sub-queues are aggregated. The aggregated data amount can be understood as the data amount contained in the aggregated database. The preset non-real-time delay threshold can be understood as a time threshold that is set in advance according to actual conditions and is used to determine whether the non-real-time service type service can be executed within the required time.

[0085] Specifically, in the case where it is determined that the non-real-time available computing power resources cannot be directly used to process the to-be-processed service streams corresponding to the non-real-time service sub-queues, the to-be-processed service streams in each non-real-time service sub-queue required to be processed can be aggregated to obtain an aggregated service block, so that each non-real-time service type to-be-processed service stream can be processed as a whole. Then, whether the unified execution of the aggregated service block by the non-real-time available computing power resources can meet the execution time requirement of the non-real-time service type can be determined according to the aggregated data amount of the aggregated service block, the non-real-time available computing power resources in the current service period, and the preset non-real-time delay threshold. If not, the aggregated service block can be split or adjusted in other ways, so that the execution time of one non-real-time service type service can meet the requirement of the preset non-real-time delay threshold, and finally the non-real-time computing power arrangement scheme is obtained.

[0086] Optionally, the non-real-time computing power arrangement scheme is determined according to the aggregated data amount of the aggregated service block, the non-real-time available computing power resources, and the preset non-real-time delay threshold, which can be realized by the following way:

[0087] 1) the ratio of the aggregated data amount to the non-real-time available computing resource is determined as the processing delay of the aggregated service block.

[0088] 2) if the processing delay is less than or equal to the preset non-real-time delay threshold, the aggregated service block is determined as a non-real-time computing resource arrangement scheme by processing as a batch.

[0089] 3) if the processing delay is greater than the preset non-real-time delay threshold, the aggregated service block is split, and the split aggregated service block is taken as a new aggregated service block, and step 1) is returned to execute.

[0090] In this embodiment, the processing delay of the aggregated service block can be understood as the time required for the aggregated service block of the aggregated data amount to be processed by the non-real-time available computing resource.

[0091] Specifically, when determining the non-real-time computing resource arrangement scheme, the ratio of the aggregated data amount to the non-real-time available computing resource can be first calculated as the processing delay of the aggregated service block. This processing delay can be understood as the time required for processing all the non-real-time service types of the to-be-processed service stream by the non-real-time available computing resource. Then, the processing delay can be compared with the preset non-real-time delay threshold. If the processing delay is less than or equal to the preset non-real-time delay threshold, it can be considered that the aggregated service block is processed by the non-real-time available computing resource in the edge computing device as a batch, and the business execution requirement can also be met. At this time, the aggregated service block can be directly determined as a non-real-time computing resource arrangement scheme by processing as a batch. If the processing delay is greater than the preset non-real-time delay threshold, it can be considered that the aggregated service block is processed by the non-real-time available computing resource in the edge computing device as a batch, and it is difficult to meet the business execution requirement. At this time, the aggregated service block can be split according to some preset rules, and the split aggregated service block is taken as a new aggregated service block. Each new aggregated service block is processed by the non-real-time available computing resource in the edge computing device as a batch to determine whether each new aggregated service block can meet the business execution requirement. The above steps are repeated until all the aggregated service blocks meeting the requirement are obtained, and each aggregated service block is determined as a non-real-time computing resource arrangement scheme by processing as a batch.

[0092] For example, assuming that D block is the aggregated data amount of the aggregated service block, C free is the non-real-time available computing resource, T Limit is the preset non-real-time delay threshold, the processing delay of the aggregated service block can be expressed as If T i ≤ T Limit , the aggregated service block can be determined as a non-real-time computing resource arrangement scheme by processing as a batch. If T i > T LimitThen the aggregated service block can be adjusted to D block = D block * 0.5, and repeating the above steps until the condition T i ≤ T Limit is met, obtaining the corresponding non-real-time computing power arrangement scheme.

[0093] S213, the non-real-time available computing power resources are allocated according to the ratio between the to-be-processed data amounts of each non-real-time service sub-queue, to determine the non-real-time computing power arrangement scheme.

[0094] Specifically, when the non-real-time available computing power resources can be directly used to process the to-be-processed service streams in each non-real-time service sub-queue, a similar manner to S208 can be used to directly allocate the non-real-time available computing power resources according to the ratio between the to-be-processed data amounts of each non-real-time service sub-queue, or according to the proportion of the to-be-processed data amount of each non-real-time service sub-queue in the total data amount corresponding to all non-real-time service sub-queues, and each allocated computing power resource is allocated to the non-real-time service container corresponding to each non-real-time service sub-queue, to obtain the corresponding non-real-time computing power arrangement scheme.

[0095] Optionally, the embodiment of the application also provides a determination method of a computing power reservation error correction coefficient, comprising:

[0096] substituting the actual computing power resource occupation amount of each time point in the previous service period corresponding to the current service period, and the real-time computing power resource reservation amount of each time point corresponding to the current service period into a preset correction coefficient determination formula, to determine the computing power reservation error correction coefficient corresponding to the current service period;

[0097] wherein the preset correction coefficient determination formula is:

[0098]

[0099] wherein f C is the computing power reservation error correction coefficient corresponding to the current service period; T is the length of the previous service period; T i is the previous service period; T i+1 is the current service period; C real (t) is the actual computing power resource occupation amount at time t; C resv (t) is the real-time computing power resource reservation amount at time t.

[0100] Specifically, in each service period, after the real-time computing power arrangement scheme is determined, the real-time service type to be processed in the service period is processed according to the real-time computing power arrangement scheme. Since the actual computing power resource amount occupied by the real-time service sub-queue in the service period fluctuates during processing, the computing power reservation error correction coefficient required for the next period can be re-determined according to the actual computing power resource amount and the real-time computing power resource reservation amount in each service period, so that the computing power reservation error correction coefficient used in each service period can fully consider the actual service processing in the service period most closely related to it, and the accuracy of determining the real-time computing power resource reservation amount in each service period is improved.

[0101] For example, the above formula takes the last service period of the current service period as an example to illustrate the determination method of the computing power reservation error correction coefficient in the current service period. By dividing the last service period into multiple time points, the difference square of the last real-time computing power resource reservation amount and the last actual computing power resource occupation amount corresponding to each time point is integrated, the square root of the integral value is solved, and the computing power reservation error correction coefficient at the minimum square root is determined as the computing power reservation error correction coefficient corresponding to the current period.

[0102] It can be understood that in the current service period, after the real-time computing power arrangement scheme is determined, the same method as above is used to determine the computing power reservation error correction coefficient required for the next service period of the current service period, and the embodiments of the present application will not be described in detail.

[0103] Optionally, to improve the correction accuracy of the computing power reservation error correction coefficient on the total real-time computing power demand of the real-time service type, the time demand for processing the real-time service type in the last service period according to the real-time computing power arrangement scheme also needs to be considered. Therefore, after the computing power reservation error correction coefficient corresponding to the current service period is determined in the last service period, it further includes:

[0104] According to the last uncompleted data amount corresponding to the last service period, the last real-time demand data amount corresponding to the last real-time computing power resource reservation amount, and the last real-time computing power resource reservation amount, the last processing delay of the last service period is determined. If the last processing delay is less than or equal to the preset real-time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is kept unchanged. If the last processing delay is greater than the preset real-time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is updated by a preset additional step.

[0105] In the embodiment, the last unfinished data amount can be specifically understood as the data amount of the real-time service type service that has not been processed in the edge computing device at the beginning of the last service period. The last processing delay can be specifically understood as the time consumed for processing the real-time service type service according to the real-time computing power arrangement scheme in the last service period. The preset real-time delay threshold can be specifically understood as a time threshold that is set in advance according to actual conditions and is used to determine whether the real-time service type service can be executed within the required time. The preset additional step can be specifically understood as a preset value that is determined in advance according to actual conditions and is used to adjust the computing power reservation error correction coefficient so as to make the computing power reservation error correction coefficient more close to the actual value.

[0106] Specifically, the last unfinished data flow corresponding to the last service period and the last real-time demand data amount corresponding to the last real-time computing power reservation amount are summed up to determine the total data amount that needs to be processed by using the computing power resource of the last real-time computing power reservation amount, and then the total data amount is divided by the last real-time computing power reservation amount, that is, the total time length required for processing the real-time service type service in the last service period can be obtained, which is also the last processing delay. Then, the last processing delay can be compared with the preset real-time delay threshold. If the last processing delay is less than or equal to the preset real-time delay threshold, it can be considered that the computing power reservation error correction coefficient currently determined is consistent with the deviation caused in the actual operation process, and adjustment is not required, that is, the computing power reservation error correction coefficient corresponding to the current service period can be kept unchanged. If the last processing delay is greater than the preset real-time delay threshold, it can be considered that the computing power reservation error correction coefficient currently determined is inconsistent with the deviation caused in the actual operation process, and the computing power reservation error correction coefficient can be increased by a preset additional step to complete the update of the computing power reservation error correction coefficient corresponding to the current service period, that is, the computing power reservation error correction coefficient used in the current service period should be the computing power reservation error correction coefficient increased by the preset additional step.

[0107] The technical scheme of the embodiment is used to complete mapping of each to-be-processed service flow to a corresponding service sub-queue, and then to obtain total real-time computing power demand by performing statistical conversion on the to-be-processed data flow of each real-time service sub-queue, to obtain a corresponding real-time computing power resource reservation amount by correcting the computing power reservation error correction coefficient, and to determine a real-time computing power arrangement scheme according to the proportion of the data amount corresponding to each real-time service sub-queue in the total real-time demand data amount. After the real-time computing power resource reservation amount is determined, the non-real-time available computing power resource available for non-real-time services can be determined according to the current remaining computing power resource and the real-time computing power resource reservation amount, so that the non-real-time available computing power resource is fully utilized according to the non-real-time delay threshold required by the non-real-time services and the data aggregation processing, and a non-real-time computing power arrangement scheme is obtained, so that the current remaining computing power resource in the current service period can be fully utilized. The computing power reservation error correction coefficient can be re-determined at the end of each service period, and the determined computing power reservation error correction coefficient is updated according to the delay in processing the real-time type services in the current period, so that the finally determined computing power reservation error correction coefficient can better reflect the dynamically changing service demand and service actual processing state, and thus the computing power arrangement schemes determined according to the computing power reservation error correction coefficient can be more suitable for processing of the edge computing device, and the computing power resource utilization rate of the edge computing device is improved.

[0108] Embodiment three

[0109] Figure 3 A structural schematic diagram of an edge computing computing power arrangement device provided by the embodiment three of the application is shown in Figure 3 The edge computing computing power arrangement device includes an information acquisition module 31, a queue mapping module 32, a first scheme determination module 33, and a second scheme determination module 34.

[0110] The information acquisition module 31 is used to acquire to-be-processed service flow information of each to-be-processed service flow in a current service period. The queue mapping module 32 is used to map each to-be-processed service flow to a corresponding real-time service sub-queue and a non-real-time service sub-queue according to the service flow characteristics in the to-be-processed service flow information. The first scheme determination module 33 is used to determine a real-time computing power resource reservation amount according to the to-be-processed data amount of each real-time service sub-queue and a computing power reservation error correction coefficient, and to determine a real-time computing power arrangement scheme according to the real-time computing power resource reservation amount. The second scheme determination module 34 is used to determine a non-real-time computing power arrangement scheme according to the to-be-processed data amount of each non-real-time service sub-queue, the current remaining computing power resource, and the real-time computing power resource reservation amount. The computing power reservation error correction coefficient is determined according to the last actual computing power resource occupation amount and the last real-time computing power resource reservation amount in the last service period corresponding to the current service period.

[0111] The technical scheme of the embodiment of the application, when it is necessary to provide corresponding computing resource for each to-be-processed service flow in the current service period, first distinguishes whether each to-be-processed service flow belongs to a real-time type service according to the characteristics of the to-be-processed service flow, and maps the service queue. Different computing arrangement modes are used to determine the computing arrangement scheme for real-time type services and non-real-time type services. In order to ensure that the real-time type service can have sufficient computing resource for execution, when the computing arrangement scheme is determined, the real-time computing arrangement scheme is first determined, so that the real-time computing resource reservation amount required for the real-time service can be preferentially reserved in the remaining computing resource in the edge computing device according to the determined real-time computing arrangement scheme. For non-real-time type services, the remaining computing resource in the edge computing device can be fully utilized while ensuring that the real-time task can be executed in time, by determining the non-real-time computing arrangement scheme according to the current remaining available computing resource in the edge computing device and the real-time computing resource reservation amount. At the same time, the computing resource reservation error correction coefficient determined according to the actual service execution in the last service period is used to assist the determination of the real-time computing arrangement scheme, so that the final obtained real-time computing arrangement scheme can fully consider the dynamically changing service demand and service actual processing state, and thus the non-real-time computing arrangement scheme determined based on the real-time computing resource reservation amount can also fully consider the dynamically changing service demand and service actual processing state, so that a more flexible and efficient computing arrangement scheme suitable for the edge computing device is finally obtained, and the computing resource utilization rate of the edge computing device is improved.

[0112] Optionally, the first scheme determination module 33 is specifically configured to:

[0113] determine a total real-time demand data amount according to the to-be-processed data amount of each real-time service subqueue;

[0114] determine a total real-time computing resource demand amount as the product of the total real-time demand data amount and a preset computing resource conversion factor;

[0115] determine a real-time computing resource reservation amount as the difference between the total real-time computing resource demand amount and the computing resource reservation error correction coefficient;

[0116] distribute the real-time computing resource reservation amount to the real-time service container corresponding to each real-time service subqueue according to the proportion of the to-be-processed data amount of each real-time service subqueue in the total real-time demand data amount, and determine a real-time computing arrangement scheme.

[0117] Optionally, the second scheme determination module 34 is specifically configured to:

[0118] determine a non-real-time available computing resource as the difference between the current remaining computing resource and the real-time computing resource reservation amount;

[0119] determine a total non-real-time computing power demand amount according to the to-be-processed data amounts of the non-real-time service sub-queues and a preset computing power conversion factor;

[0120] If the total non-real-time computing power demand amount is less than or equal to the non-real-time available computing power resource, the non-real-time available computing power resource is allocated according to the ratio between the to-be-processed data amounts of the non-real-time service sub-queues to determine a non-real-time computing power scheduling scheme.

[0121] If the total non-real-time computing power demand amount is greater than the non-real-time available computing power resource, the non-real-time service sub-queues are aggregated to generate an aggregated service block, and a non-real-time computing power scheduling scheme is determined according to the aggregated data amount of the aggregated service block, the non-real-time available computing power resource and a preset non-real-time delay threshold.

[0122] Optionally, determining a non-real-time computing power scheduling scheme according to the aggregated data amount of the aggregated service block, the non-real-time available computing power resource and a preset non-real-time delay threshold comprises:

[0123] determining a processing delay of the aggregated service block according to the ratio between the aggregated data amount and the non-real-time available computing power resource;

[0124] If the processing delay is less than or equal to the preset non-real-time delay threshold, the aggregated service block is processed as a batch to determine a non-real-time computing power scheduling scheme.

[0125] If the processing delay is greater than the preset non-real-time delay threshold, the aggregated service block is split, and the split aggregated service block is taken as a new aggregated service block, and the step of determining a processing delay of the aggregated service block according to the ratio between the aggregated data amount and the non-real-time available computing power resource is performed again.

[0126] Optionally, the queue mapping module 32 is specifically configured to:

[0127] input each to-be-processed service flow information into a pre-trained service type determination model to extract service flow features of each to-be-processed service flow information through the service type determination model, and perform service classification operation according to each service flow feature to determine a service type corresponding to each to-be-processed service flow;

[0128] map a to-be-processed service flow with a real-time service type to a corresponding real-time service sub-queue;

[0129] map a to-be-processed service flow with a non-real-time service type to a corresponding non-real-time service sub-queue;

[0130] The service type determination model is a lightweight long short-term memory network model.

[0131] Optionally, the determination manner of the computing power reservation error correction coefficient comprises:

[0132] substituting the corresponding previous actual computing power resource occupation amount at each time in the previous service period corresponding to the current service period and the corresponding previous real-time computing power resource reservation amount at each time into a preset correction coefficient determination formula to determine the computing power reservation error correction coefficient corresponding to the current service period.

[0133] The preset correction coefficient determination formula is:

[0134]

[0135] The f C The computing power reservation error correction coefficient corresponding to the current service period; T is the length of the previous service period; T i is the previous service period; T i+1 is the current service period; C real (t) is the previous actual computing power resource occupation amount at time t; and C resv (t) is the previous real-time computing power resource reservation amount at time t.

[0136] Optionally, the edge computing computing power arrangement device further comprises a coefficient updating module.

[0137] The coefficient updating module is configured to, after determining the computing power reservation error correction coefficient corresponding to the current service period, determine a previous processing time delay of the previous service period according to the previous uncompleted data amount of the previous service period, the previous real-time demand data amount corresponding to the previous real-time computing power resource reservation amount and the previous real-time computing power resource reservation amount; if the previous processing time delay is less than or equal to a preset real-time time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is kept unchanged; and if the previous processing time delay is greater than the preset real-time time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is updated through a preset additional step length.

[0138] The edge computing computing power arrangement device provided by the embodiments of the present application can execute the edge computing computing power arrangement method provided by any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0139] Embodiment Four

[0140] Figure 4This is a schematic diagram of an edge computing orchestration device according to Embodiment 4 of the present invention. The edge computing orchestration device 40 can represent various forms of digital computers, such as laptops, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. The edge computing orchestration device 40 can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0141] like Figure 4 As shown, the edge computing orchestration device 40 includes at least one processor 41 and a memory, such as a read-only memory (ROM) 42 and a random access memory (RAM) 43, communicatively connected to the at least one processor 41. The memory stores computer programs executable by the at least one processor. The processor 41 can perform various appropriate actions and processes based on the computer program stored in the ROM 42 or loaded from storage unit 48 into the RAM 43. The RAM 43 can also store various programs and data required for the operation of the edge computing orchestration device 40. The processor 41, ROM 42, and RAM 43 are interconnected via a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0142] Multiple components in the edge computing orchestration device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard and mouse; an output unit 47, such as various types of displays and speakers; a storage unit 48, such as a disk and optical disk; and a communication unit 49, such as a network interface card (NIC), modem, or wireless transceiver. The communication unit 49 allows the edge computing orchestration device 40 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0143] Processor 41 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 41 performs the various methods and processes described above, such as edge computing power orchestration methods.

[0144] In some embodiments, the edge computing computing power orchestration method can be implemented as a computer program tangibly embodied in a computer readable storage medium, e.g., storage unit 48. In some embodiments, parts or all of the computer program can be loaded and / or installed onto the computing power orchestration device 40 via, e.g., ROM 42 and / or communication unit 49. When the computer program is loaded into RAM 43 and executed by processor 41, one or more steps of the above-described edge computing computing power orchestration method can be performed. Alternatively, in other embodiments, processor 41 can be configured to perform the edge computing computing power orchestration method by way of other means (e.g., by way of firmware).

[0145] Optionally, the embodiments of the present application further provide a computer program product, comprising a computer program which, when executed by a processor, implements the edge computing computing power orchestration method provided by any of the embodiments of the present application.

[0146] The various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (PLD), a computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0147] Computer programs used to implement the methods of the present application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor of the machine, implements the functions / acts specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0148] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0149] To provide for interaction with a user, the systems and techniques described here can be implemented on a computing device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computing device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0150] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), blockchain network, and the Internet.

[0151] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0152] It should be understood that the various forms of flow shown above can be used to reorder, add or delete steps. For example, each step described in the present application can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, which is not limited herein.

[0153] The above detailed description does not constitute a limitation on the scope of protection of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. An edge computing computing power arrangement method, characterized in that, The method comprises: obtaining to-be-processed service flow information of each to-be-processed service flow in a current service period; mapping each to-be-processed service flow to a corresponding real-time service sub-queue and a non-real-time service sub-queue according to service flow characteristics in each to-be-processed service flow information; determining a real-time computing resource reservation amount according to a to-be-processed data amount of each real-time service sub-queue and a computing resource reservation error correction coefficient, and determining a real-time computing arrangement scheme according to the real-time computing resource reservation amount; determining a non-real-time computing arrangement scheme according to a to-be-processed data amount of each non-real-time service sub-queue, a current remaining computing resource, and the real-time computing resource reservation amount; wherein the computing resource reservation error correction coefficient is determined according to an actual computing resource occupation amount and a previous real-time computing resource reservation amount in a previous service period corresponding to the current service period; wherein the determining of the real-time computing resource reservation amount according to the to-be-processed data amount of each real-time service sub-queue and the computing resource reservation error correction coefficient, and the determining of the real-time computing arrangement scheme according to the real-time computing resource reservation amount, comprises: determining a total real-time demand data amount according to the to-be-processed data amount of each real-time service sub-queue; determining a total real-time computing demand amount as a product of the total real-time demand data amount and a preset computing conversion factor; determining a real-time computing resource reservation amount as a difference between the total real-time computing demand amount and the computing resource reservation error correction coefficient; allocating the real-time computing resource reservation amount to a real-time service container corresponding to each real-time service sub-queue according to a proportion of the to-be-processed data amount of each real-time service sub-queue in the total real-time demand data amount, to determine a real-time computing arrangement scheme.

2. The edge computing computing power arrangement method according to claim 1, characterized in that, The determining of the non-real-time computing arrangement scheme according to the to-be-processed data amount of each non-real-time service sub-queue, the current remaining computing resource, and the real-time computing resource reservation amount, comprises: determining a non-real-time available computing resource as a difference between the current remaining computing resource and the real-time computing resource reservation amount; determining a total non-real-time computing demand amount according to the to-be-processed data amount of each non-real-time service sub-queue and a preset computing conversion factor; if the total non-real-time computing demand amount is less than or equal to the non-real-time available computing resource, allocating the non-real-time available computing resource according to a ratio between the to-be-processed data amounts of each non-real-time service sub-queue to determine a non-real-time computing arrangement scheme; if the total non-real-time computing demand amount is greater than the non-real-time available computing resource, aggregating each non-real-time service sub-queue to generate an aggregated service block, and determining a non-real-time computing arrangement scheme according to an aggregated data amount of the aggregated service block, the non-real-time available computing resource, and a preset non-real-time time delay threshold.

3. The edge computing algorithm power arrangement method of claim 2, wherein, The determining of the non-real-time computing arrangement scheme according to the aggregated data amount of the aggregated service block, the non-real-time available computing resource, and the preset non-real-time time delay threshold, comprises: determining a processing time delay of the aggregated service block as a ratio between the aggregated data amount and the non-real-time available computing resource. If the processing delay is less than or equal to a preset non-real-time delay threshold, the aggregated service block is processed as a batch to determine a non-real-time computing power arrangement scheme; If the processing delay is greater than the preset non-real-time delay threshold, the aggregated service block is split, and the split aggregated service block is taken as a new aggregated service block, and the step of determining the processing delay of the aggregated service block by the ratio of the aggregated data volume to the non-real-time available computing power resource is returned to be executed.

4. The edge computing algorithm power arrangement method of claim 1, wherein, The mapping of each of the to-be-processed service flows into the corresponding real-time service sub-queue and non-real-time service sub-queue according to the service flow characteristics in each of the to-be-processed service flow information comprises: inputting each of the to-be-processed service flow information into a pre-trained service type determination model to extract the service flow characteristics of each of the to-be-processed service flow information by the service type determination model, and performing service classification operation according to each of the service flow characteristics to determine the corresponding service type of each of the to-be-processed service flow; mapping the to-be-processed service flow with the real-time service type into the corresponding real-time service sub-queue; mapping the to-be-processed service flow with the non-real-time service type into the corresponding non-real-time service sub-queue; The service type determination model is a lightweight long short-term memory network model.

5. The edge computing algorithm power arrangement method according to any one of claims 1-4, characterized in that, The determination method of the computing power reservation error correction coefficient comprises: substituting the actual computing power resource occupation amount at each time in the previous service period corresponding to the current service period and the previous real-time computing power resource reservation amount corresponding to each time into a preset correction coefficient determination formula to determine the computing power reservation error correction coefficient corresponding to the current service period; The preset correction coefficient determination formula is: ； Wherein, the is the computing power reservation error correction coefficient corresponding to the current business period; the T is the length of the last business period; the is the last business period; the is the current business period; the is the last actual computing power resource occupation at time t; the is the last real-time computing power resource reservation at time t.

6. The edge computing computing power arrangement method according to claim 5, characterized in that, After determining the computing power reservation error correction coefficient corresponding to the current service period, further comprising: determining a previous processing delay of the previous service period according to the previous uncompleted data volume of the previous service period, the previous real-time demand data volume corresponding to the previous real-time computing power resource reservation amount, and the previous real-time computing power resource reservation amount; if the previous processing delay is less than or equal to a preset real-time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is kept unchanged; if the previous processing delay is greater than the preset real-time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is updated by a preset additional step size.

7. An edge computing computing power arrangement device, characterized in that, comprise: an information acquisition module configured to acquire to-be-processed service flow information of each to-be-processed service flow in a current service period; a queue mapping module configured to map each of the to-be-processed service flows into a corresponding real-time service sub-queue and non-real-time service sub-queue according to service flow characteristics in each of the to-be-processed service flow information; a first scheme determination module configured to determine a real-time computing power resource reservation amount according to the to-be-processed data volume of each of the real-time service sub-queues and the computing power reservation error correction coefficient, and determine a real-time computing power arrangement scheme according to the real-time computing power resource reservation amount; The second scheme determination module is configured to determine a non-real-time computing power arrangement scheme according to the amounts of data to be processed of the non-real-time service sub-queues, the current residual computing power resources, and the real-time computing power reservation amount. The computing power reservation error correction coefficient is determined according to an actual computing power resource occupation amount in a previous service period corresponding to the current service period and a previous real-time computing power reservation amount. The first scheme determination module is specifically configured to: determine a total real-time demand data amount according to the amounts of data to be processed of the real-time service sub-queues; determine a total real-time computing power demand amount as a product of the total real-time demand data amount and a preset computing power conversion factor; determine a real-time computing power reservation amount as a difference between the total real-time computing power demand amount and the computing power reservation error correction coefficient; distribute the real-time computing power reservation amount to the real-time service containers corresponding to the real-time service sub-queues according to proportions of the amounts of data to be processed of the real-time service sub-queues in the total real-time demand data amount, to determine a real-time computing power arrangement scheme. 8.The edge computing computing power arrangement apparatus of claim 7, wherein, The second scheme determination module is specifically configured to: determine a non-real-time available computing power resource as a difference between the current residual computing power resources and the real-time computing power reservation amount; determine a total non-real-time computing power demand amount according to the amounts of data to be processed of the non-real-time service sub-queues and the preset computing power conversion factor; if the total non-real-time computing power demand amount is less than or equal to the non-real-time available computing power resource, distribute the non-real-time available computing power resource according to ratios between the amounts of data to be processed of the non-real-time service sub-queues, to determine a non-real-time computing power arrangement scheme; if the total non-real-time computing power demand amount is greater than the non-real-time available computing power resource, aggregate the non-real-time service sub-queues to generate an aggregated service block, and determine a non-real-time computing power arrangement scheme according to an aggregated data amount of the aggregated service block, the non-real-time available computing power resource, and a preset non-real-time time delay threshold. 9.The edge computing computing power arrangement apparatus of claim 8, wherein, The determination of the non-real-time computing power arrangement scheme according to the aggregated data amount of the aggregated service block, the non-real-time available computing power resource, and the preset non-real-time time delay threshold includes: determine a processing time delay of the aggregated service block as a ratio between the aggregated data amount and the non-real-time available computing power resource; if the processing time delay is less than or equal to the preset non-real-time time delay threshold, determine the aggregated service block as a batch for processing to determine a non-real-time computing power arrangement scheme; if the processing time delay is greater than the preset non-real-time time delay threshold, split the aggregated service block, and determine the split aggregated service block as a new aggregated service block, and return to execute the step of determining the processing time delay of the aggregated service block as the ratio between the aggregated data amount and the non-real-time available computing power resource. 10.The edge computing computing power arrangement device according to claim 7, wherein, The queue mapping module is specifically configured to: input each of the to-be-processed service flow information into a pre-trained service type determination model, to extract service flow features of each of the to-be-processed service flow information by the service type determination model, and perform service classification operation according to each of the service flow features, to determine a service type corresponding to each of the to-be-processed service flow. Map the to-be-processed service stream of the service type of the real-time service type to a corresponding real-time service sub-queue; Map the to-be-processed service stream of the service type of the non-real-time service type to a corresponding non-real-time service sub-queue. The service type determination model is a lightweight long short-term memory network model.

11. The edge computing computing power arrangement device according to any one of claims 7-10, characterized in that, The determination manner of the computing power reservation error correction coefficient comprises: Substitute the actual computing power resource occupation amount of each time in the previous service period corresponding to the current service period and the real-time computing power resource reservation amount of each time in the previous service period into a preset correction coefficient determination formula to determine the computing power reservation error correction coefficient corresponding to the current service period. The preset correction coefficient determination formula is: ； Wherein, the is the computing power reservation error correction coefficient corresponding to the current business cycle; the T is the length of the last business cycle; the is the last business cycle; the is the current business cycle; the is the last actual computing power resource occupation at time t; the is the last real-time computing power resource reservation at time t.

12. The edge computing computing power arrangement device according to claim 11, characterized in that, Further comprising: A coefficient updating module; The coefficient updating module is configured to, after determining the computing power reservation error correction coefficient corresponding to the current service period, determine a previous processing delay of the previous service period according to the previous uncompleted data amount of the previous service period, the previous real-time demand data amount corresponding to the previous real-time computing power resource reservation amount, and the previous real-time computing power resource reservation amount. If the previous processing delay is less than or equal to a preset real-time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is kept unchanged. If the previous processing delay is greater than the preset real-time delay threshold, the computing power reservation error correction coefficient corresponding to the current service period is updated by a preset additional step.

13. An edge computing computing power arrangement device, characterized in that, Comprise: At least one processor; And a memory in communication connection with the at least one processor; The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the edge computing computing power arrangement method of any one of claims 1-6.

14. A storage medium containing computer-executable instructions, wherein: The computer executable instructions, when executed by the computer processor, are used to execute the edge computing computing power arrangement method as claimed in any one of claims 1-6.

15. A computer program product comprising a computer program which, when executed by a processor, implements the edge computing computing power arrangement method as claimed in any one of claims 1-6.

15. A computer program product comprising a computer program which, when executed by a processor, implements the edge computing computing power arrangement method as claimed in any one of claims 1-6.

Images