Edge computing power prediction method, computing power scheduling method and system

Abstract

The present invention relates to the field of Internet of Things, and provides a computing power prediction method, computing power orchestration method and system for edge computing. The computing power prediction method includes: counting the arrival rates of different types of business flows, setting the distribution function of business flow arrival according to the arrival rates of different types of business flows; calculating the system efficiency of edge computing according to the distribution function of business flow arrival, and calculating the system efficiency ratio of each type of business flow; under preset constraints, predicting the computing power value required for each type of business flow according to the system efficiency ratio of each type of business flow. The computing power prediction method of the present invention can estimate the computing power resource requirements of edge computing services in a refined manner. The present invention adopts a computing power prediction method to predict the computing power values ​​required for various types of business flows of edge computing, and maps various types of business flows to the application programs of corresponding containers according to the predicted computing power values, so as to provide a flexible and efficient computing power allocation scheme for edge computing and improve the utilization rate of computing power resources.

Description

Technical Field

[0001] The present invention relates to the field of Internet of Things, and specifically to a computing power prediction method for edge computing, a computing power orchestration method for edge computing, a computing power orchestration system for edge computing, and a storage medium. Background Art

[0002] 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. Edge computing, as a new computing mode, can enable data to be processed promptly and effectively in the edge network near the source. In the edge computing mode, the terminal offloads tasks to nearby edge servers for execution, effectively alleviating the network congestion and high latency problems caused by transmitting all data to the data center in traditional cloud computing. The bottleneck of edge computing is mainly limited computing, communication, storage and other resources. When the number of terminals and the scale of data increase further, the edge nodes tend to be saturated, resulting in a serious decline in service quality.

[0003] The number of terminals, data, and services in the new business ecosystem of the power Internet of Things continues to increase, and massive amounts of power data need to be analyzed and calculated on edge computing devices to achieve efficient and flexible business processing and decision-making. Due to the dynamic nature of new power system services, the edge computing resource scheduling method has changed from a fixed configuration mode based on determined business objects to an elastic configuration mode based on dynamic business objects. At present, the dynamic matching between the growing business processing needs and limited computing resources is a difficult problem for edge computing. Summary of the invention

[0004] The purpose of the present invention is to provide a computing power prediction method and a computing power orchestration method for edge computing, so as to provide a flexible and efficient computing power allocation solution for edge computing and improve the utilization rate of computing power resources.

[0005] In order to achieve the above object, the first aspect of the present invention provides a method for predicting computing power of edge computing, the method comprising:

[0006] Count the arrival rates of different types of business flows, and set the distribution function of business flow arrival according to the arrival rates of different types of business flows;

[0007] Calculate the system efficiency of edge computing based on the distribution function of the business flow arrival, and calculate the system efficiency ratio of each business flow;

[0008] Under the preset constraints, the computing power required for each type of business flow is predicted based on the system efficiency ratio of each type of business flow.

[0009] Furthermore, the distribution function of the service flow arrival is:

[0010]

[0011] Among them, λ1 is the periodic service arrival rate, λ2 is the random service arrival rate, δ is the impulse function, and t is the time;

[0012] The periodic services and the random services are divided according to the service type characteristics of edge computing.

[0013] Furthermore, the method of calculating the system efficiency of edge computing according to the distribution function of the business flow arrival includes: performing a Laplace transform on the distribution function of the business flow arrival to obtain a system efficiency factor; and calculating the system efficiency of edge computing according to the system efficiency factor.

[0014] Furthermore, the calculation of the system efficiency ratio of each type of business flow includes:

[0015] Count the proportions of different types of business flows, multiply the proportions of each type of business flow by the system efficiency, and get the system efficiency proportions of each type of business flow.

[0016] Furthermore, under the preset constraints, the computing power required for each type of business flow is predicted according to the system efficiency ratio of each type of business flow, including:

[0017] Under the constraints of business loss rate and business processing delay, the value range of the computing power required for each business flow is calculated according to the system efficiency ratio of each business flow;

[0018] Under the condition of a preset average business processing rate, the optimal computing power value within the value range of the computing power value is calculated, and the optimal computing power value is used as the computing power value required for the business flow.

[0019] A second aspect of the present invention provides a method for orchestrating computing power for edge computing, the method comprising:

[0020] Obtain multiple service flows for various terminal devices;

[0021] The edge computing computing power prediction method according to claim 1 is used to predict the computing power values ​​required for various types of edge computing business flows;

[0022] According to the computing power required by each business flow, each business flow is mapped to the application of the corresponding container.

[0023] Furthermore, mapping various business flows to the application programs of corresponding containers according to the computing power values ​​required by various business flows includes:

[0024] Determine the application to which each type of business flow belongs, and group the business flows belonging to the same application;

[0025] According to the priority attribute of the application, the set formed by the business flows belonging to the application is sorted, and the weight value is marked on the set;

[0026] The business flows in each set are mapped to the applications of the corresponding containers according to the weight values ​​of the sets.

[0027] Furthermore, mapping various business flows to the application programs of corresponding containers according to the computing power values ​​required by various business flows also includes:

[0028] According to the real-time requirements of various business flows, the weight values ​​of the corresponding sets are modified;

[0029] The business flows in the set are mapped to the application programs of the corresponding containers according to the modified weight values ​​of the set.

[0030] Furthermore, the method further comprises:

[0031] Count the total computing power of applications corresponding to the same type of business flows;

[0032] The applications corresponding to the same type of business flows are grouped into the same container, and computing resources matching the total computing power value of the applications are configured for the container.

[0033] Furthermore, the method further comprises:

[0034] After acquiring multiple service flows of multiple terminal devices, the multiple service flows are cached to form a service flow mixed queue.

[0035] A third aspect of the present invention provides an edge computing computing power orchestration system, the system comprising:

[0036] The computing power prediction module is used to predict the computing power required for various business flows based on the characteristics of different business flows of edge computing;

[0037] The computing power mapping module is used to map various business flows to the corresponding container applications according to the computing power values ​​required by each business flow.

[0038] Furthermore, the computing power required for each type of business flow is predicted based on the characteristics of different business flows of edge computing, including:

[0039] Count the arrival rates of different types of business flows, and set the distribution function of business flow arrival according to the arrival rates of different types of business flows;

[0040] Calculate the system efficiency of edge computing based on the distribution function of the business flow arrival, and calculate the system efficiency ratio of each business flow;

[0041] Under the preset constraints, the computing power required for each type of business flow is predicted based on the system efficiency ratio of each type of business flow.

[0042] Furthermore, the distribution function of the service flow arrival is:

[0043]

[0044] Among them, λ1 is the periodic service arrival rate, λ2 is the random service arrival rate, δ is the impulse function, and t is the time;

[0045] The periodic services and the random services are divided according to the service type characteristics of edge computing.

[0046] Further, the calculating the system efficiency of edge computing according to the distribution function of the business flow arrival includes: performing Laplace transform on the distribution function of the business flow arrival to obtain a system efficiency factor;

[0047] The system efficiency of edge computing is calculated based on the system efficiency factor.

[0048] Furthermore, under the preset constraints, the computing power required for each type of business flow is predicted according to the system efficiency ratio of each type of business flow, including:

[0049] Under the constraints of business loss rate and business processing delay, the value range of the computing power required for each business flow is calculated according to the system efficiency ratio of each business flow;

[0050] Under the condition of a preset average business processing rate, the optimal computing power value within the value range of the computing power value is calculated, and the optimal computing power value is used as the computing power value required for the business flow.

[0051] Furthermore, mapping various business flows to the application programs of corresponding containers according to the computing power values ​​required by various business flows includes:

[0052] Determine the application to which each type of business flow belongs, and group the business flows belonging to the same application;

[0053] According to the priority attribute of the application, the set formed by the business flows belonging to the application is sorted, and the weight value is marked on the set;

[0054] The business flows in each set are mapped to the applications of the corresponding containers according to the weight values ​​of the sets.

[0055] Furthermore, the system further comprises:

[0056] The service flow cache module is used to cache multiple service flows after acquiring multiple service flows of multiple terminal devices to form a service flow mixed queue.

[0057] Furthermore, the system further comprises:

[0058] The service interface module is used to receive multiple service flows from multiple terminal devices through multiple communication ports.

[0059] On the other hand, the present invention also provides a storage medium on which computer program instructions are stored. When the computer program instructions are executed, the above-mentioned edge computing power prediction method or the above-mentioned edge computing power orchestration method is implemented.

[0060] The edge computing computing power prediction method of the present invention sets the distribution function of the arrival of business flows according to the arrival rates of different types of business flows, calculates the system efficiency ratio of various types of business flows of edge computing through the distribution function, and predicts the computing power value required for various types of business flows under constraints. This method can finely estimate the computing power resource requirements of edge computing services, so as to facilitate the dynamic arrangement of computing power resources.

[0061] The computing power orchestration method and system of the present invention adopt a computing power prediction method to predict the computing power values ​​required for various business flows of edge computing, and map various business flows to the applications of the corresponding containers according to the predicted computing power values, so as to configure computing power resources that match the total computing power value of the application, provide a flexible and efficient computing power allocation solution for edge computing, and improve the utilization rate of computing power resources.

[0062] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0064] Figure 1 It is a flow chart of a method for predicting computing power of edge computing provided by one embodiment of the present invention;

[0065] Figure 2 It is a flow chart of a computing power orchestration method for edge computing provided by an embodiment of the present invention;

[0066] Figure 3 It is a block diagram of an edge computing computing power orchestration system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0067] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0068] Figure 1 FIG. 1 is a flow chart of a method for predicting edge computing power provided by an embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides a method for predicting computing power of edge computing, and the method comprises the following steps:

[0069] S21. Count the arrival rates of different types of business flows, and set the distribution function of business flow arrival according to the arrival rates of different types of business flows.

[0070] According to the characteristics of edge computing business types, IoT edge computing services are divided into different types, and the characteristics of each type of business are independent of each other, for example, they are divided into periodic services and random services.

[0071] The service arrival rate refers to the number of service units arriving per second. Assuming that the periodic service arrival rate is λ1 and the random service arrival rate is λ2, the distribution function of service flow arrival is set to G(t).

[0072]

[0073] Among them, λ1 is the periodic service arrival rate, λ2 is the random service arrival rate, δ is the impulse function, and t is the time.

[0074] S22. Calculate the system efficiency of edge computing based on the distribution function of the business flow arrival, and calculate the system efficiency ratio of each type of business flow.

[0075] Perform Laplace transform on the distribution function G(t) of the traffic flow arrival to obtain the system efficiency factor α:

[0076] α=G * (μ-μα);

[0077] Among them, the G*() function represents the Laplace transform of the distribution function G(t), α is the system efficiency factor, and μ is the predicted computing power value.

[0078] Substitute the system efficiency factor α into the following formula:

[0079]

[0080] The system efficiency η of edge computing is calculated.

[0081] The proportion of different types of business flows is counted, and the proportion of each type of business flow is multiplied by the system efficiency η to obtain the system efficiency proportion η' of each type of business flow.

[0082] S23. Under the preset constraints, predict the computing power required for each type of business flow based on the system efficiency ratio of each type of business flow.

[0083] Specifically, under the constraints of business loss rate and business processing delay, the range of computing power required for each type of business flow is calculated according to the system efficiency ratio of each type of business flow; under the condition of a pre-set average business processing rate, the optimal computing power value within the range of computing power values ​​is calculated, and the optimal computing power value is used as the computing power required for the business flow.

[0084] The edge computing device receives multiple service flows from multiple terminal devices through multiple communication ports. After obtaining multiple service flows from multiple terminal devices, it caches multiple service flows to form a mixed service flow queue.

[0085] In this implementation, the computing power value is predicted by constructing a computing power prediction model of the edge computing node. The formula of the computing power prediction model is as follows:

[0086]

[0087] Among them, η' is the system efficiency ratio of the business flow, ω is the pre-set average business processing rate, μ is the predicted computing power value, and P over is the overflow probability of the cache queue, P loss is the service loss rate constraint, W s is the average waiting delay of the service flow, T s It is the service processing delay constraint condition.

[0088] The execution process of the computing power prediction model is as follows:

[0089] Step 1: Initialize the basic variables of the computing power prediction model. The periodic service arrival rate is λ1, and the random service arrival rate is λ2.

[0090] Step 2: Extract statistical parameters such as the service arrival time interval in the service flow cache queue and calculate the comprehensive service arrival rate service λ C ;

[0091] Step 3: Calculate the system efficiency factor α using the distribution function G(t) of the traffic flow arrival;

[0092] Step 4: Set the overflow probability P of the cache queue over , Service loss rate constraint P loss , business processing delay constraint T s etc. constraints;

[0093] Step 5: Use the enumeration method to calculate the service loss rate constraint condition P under the condition of the preset service average processing rate ω loss , business processing delay constraint T s The value range of the computing power value μ;

[0094] Step 6: Calculate the optimal solution of maxη'=ω / μ. The μ obtained at this time is the final predicted computing power value.

[0095] The edge computing computing power prediction method of this embodiment sets the distribution function of the service flow arrival according to the arrival rate of different types of service flows, calculates the system efficiency ratio of various types of service flows of edge computing through the distribution function, and predicts the computing power value required for various types of service flows under constraints. This method can finely estimate the computing power resource requirements of edge computing services, so as to facilitate the dynamic arrangement of computing power resources.

[0096] An embodiment of the present invention also provides a computing power orchestration method for edge computing, which maps computing power resources based on the above-mentioned computing power prediction method, and maps various business flows to different business containers and corresponding application programs APP.

[0097] Figure 2 FIG. 1 is a flow chart of a computing power arrangement method for edge computing provided by an embodiment of the present invention. Figure 2 As shown, this embodiment provides a computing power arrangement method for edge computing, and the method includes the following steps:

[0098] S10. Acquire multiple service flows of multiple terminal devices.

[0099] The edge computing device receives multiple business flows from multiple terminal devices through multiple communication ports. After obtaining the multiple business flows from the multiple terminal devices, the multiple business flows are cached to form a mixed business flow queue.

[0100] S20. Use the above-mentioned edge computing computing power prediction method to predict the computing power values ​​required for various edge computing business flows.

[0101] The mixed queue of business flows formed by multiple business flows is input into the computing power prediction module, and the computing power prediction module executes the above-mentioned computing power prediction method steps to calculate the computing power values ​​required for various business flows of edge computing.

[0102] S30. Map each type of business flow to the application program of the corresponding container according to the computing power value required by each type of business flow.

[0103] In one embodiment, the application to which each type of business flow belongs is determined, and the business flows belonging to the same application are grouped; the groups formed by the business flows belonging to the application are sorted according to the priority attribute of the application, and the groups are labeled with weight values. The business flows in each group are mapped to the application of the corresponding container according to the weight value of the group.

[0104] In another embodiment, mapping various types of business flows to applications corresponding to containers further includes: modifying the weight value of the corresponding set according to the real-time requirements of the various types of business flows; and mapping the business flows in the set to the applications corresponding to the container according to the modified set weight value.

[0105] In addition, the computing power orchestration method of this embodiment also includes: counting the total computing power value of applications corresponding to the same type of business flows; merging the applications corresponding to the same type of business flows into the same container, and configuring computing power resources for the container that match the total computing power value of the applications.

[0106] In a more specific embodiment, mapping various types of business flows to the application APP of the corresponding container includes the following process:

[0107] 1) Determine the APP to which each business flow belongs. Business flows belonging to the same APP form a set;

[0108] 2) Sort the priority attributes of each APP and mark the weight value of the service flow set respectively;

[0109] 3) Count the real-time requirements of each service flow, sort them according to the remaining time of the service flow arrival, increase the priority weight of the service flow whose remaining time is close to the threshold, and modify its weight value;

[0110] 4) According to the corrected weight value, the business flow is mapped to the APP of the corresponding business in order;

[0111] 5) Count the total predicted computing power values ​​of similar business apps, merge such businesses into the same container, and configure matching computing power resources for the container.

[0112] The computing power orchestration method of this embodiment adopts a computing power prediction method to predict the computing power values ​​required for various business flows of edge computing, and maps various business flows to the applications of the corresponding containers according to the predicted computing power values, so as to configure computing power resources that match the total computing power value of the application, provide a flexible and efficient computing power allocation solution for edge computing, and improve the utilization rate of computing power resources.

[0113] Figure 3 FIG. 1 is a block diagram of an edge computing computing power orchestration system provided by an embodiment of the present invention. Figure 3As shown, this embodiment provides a computing power orchestration system for edge computing, and the system includes: a computing power prediction module, a computing power mapping module, a service cache module and a service interface module. The computing power prediction module is used to predict the computing power values ​​required for various types of business flows according to the characteristics of different business flows of edge computing. The computing power mapping module is used to map various types of business flows to the application programs of the corresponding containers according to the computing power values ​​required for various types of business flows. The business cache module is used to cache multiple business flows after acquiring multiple business flows of multiple terminal devices to form a mixed queue of business flows. The business interface module is used to receive multiple business flows from multiple terminal devices from multiple communication ports.

[0114] The computing power prediction module counts the arrival rates of different types of business flows, sets the distribution function of business flow arrival according to the arrival rates of different types of business flows; calculates the system efficiency of edge computing according to the distribution function of business flow arrival, and calculates the system efficiency proportion of each type of business flow; under preset constraints, predicts the computing power required for each type of business flow according to the system efficiency proportion of each type of business flow.

[0115] Specifically, the computing power prediction module divides the IoT edge computing services into different types according to the characteristics of the edge computing services, and the characteristics of each type of services are independent of each other, for example, they are divided into periodic services and random services. Assuming that the arrival rate of periodic services is λ1 and the arrival rate of random services is λ2, the distribution function of service flow arrival is set to G(t).

[0116]

[0117] Among them, λ1 is the periodic service arrival rate, λ2 is the random service arrival rate, δ is the impulse function, and t is the time.

[0118] Perform Laplace transform on the distribution function G(t) of the traffic flow arrival to obtain the system efficiency factor α:

[0119] α=G * (μ-μα);

[0120] Among them, the G*() function represents the Laplace transform of the distribution function G(t), α is the system efficiency factor, and μ is the predicted computing power value.

[0121] Substitute the system efficiency factor α into the following formula:

[0122]

[0123] The system efficiency η of edge computing is calculated.

[0124] The proportion of different types of business flows is counted, and the proportion of each type of business flow is multiplied by the system efficiency η to obtain the system efficiency proportion η' of each type of business flow.

[0125] Under the constraints of business loss rate and business processing delay, the range of computing power required for each type of business flow is calculated according to the system efficiency ratio of each type of business flow; under the condition of a pre-set average business processing rate, the optimal computing power value within the range of computing power values ​​is calculated, and the optimal computing power value is used as the computing power required for the business flow.

[0126] The computing power mapping module determines the applications to which various types of business flows belong, and forms a set of business flows belonging to the same application; sorts the sets formed by the business flows belonging to the application according to the priority attributes of the application, and marks the sets with weight values; and maps the business flows in each set to the application of the corresponding container according to the weight values ​​of the sets.

[0127] Specifically, the process of the computing power mapping module mapping the business flow is as follows:

[0128] 1) Determine the APP to which each business flow belongs. Business flows belonging to the same APP form a set;

[0129] 2) Sort the priority attributes of each APP and mark the weight value of the service flow set respectively;

[0130] 3) Count the real-time requirements of each service flow, sort them according to the remaining time of the service flow arrival, increase the priority weight of the service flow whose remaining time is close to the threshold, and modify its weight value;

[0131] 4) According to the corrected weight value, the business flow is mapped to the APP of the corresponding business in order;

[0132] 5) Count the total predicted computing power values ​​of similar business apps, merge such businesses into the same container, and configure matching computing power resources for the container.

[0133] The computing power orchestration system of this embodiment builds a computing power prediction model for multiple businesses. The computing power prediction module predicts the computing power values ​​required for various business flows of edge computing. The computing power mapping module maps various business flows to the application programs of the corresponding containers according to the predicted computing power values, so as to configure computing power resources that match the total computing power value of the application programs, provide a flexible and efficient computing power allocation solution for edge computing, and improve the utilization rate of computing power resources.

[0134] An embodiment of the present invention also provides a storage medium on which computer program instructions are stored. When the computer program instructions are executed, the above-mentioned edge computing power prediction method or the above-mentioned edge computing power orchestration method is implemented.

[0135] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0136] The present application is described with reference to the flowcharts and / or block diagrams of the methods, systems and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0137] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0138] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0139] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for predicting computing power of edge computing, characterized in that: The method comprises: Receive multiple service flows from multiple terminal devices through edge computing devices, and cache the received multiple service flows to form a service flow cache queue; Count the arrival rates of different types of business flows in the business flow cache queue, and set the distribution function of business flow arrival according to the arrival rates of different types of business flows; Calculate the system efficiency of edge computing based on the distribution function of the business flow arrival, and calculate the system efficiency ratio of each business flow; Under preset constraints, the computing power required for each type of business flow is predicted according to the system efficiency ratio of each type of business flow, including: under the condition that the overflow probability of the business flow cache queue is less than or equal to the business loss rate constraint condition, and the average waiting delay of the business flow is less than or equal to the business processing delay constraint condition, the value range of the computing power required for each type of business flow is calculated according to the system efficiency ratio of each type of business flow, under the condition of a preset average business processing rate, the optimal computing power value within the value range of the computing power value is calculated, and the optimal computing power value is used as the computing power required for the business flow; Determine the application to which each type of business flow belongs based on the computing power required by each type of business flow, group the business flows belonging to the same application, map the business flows in the group to the application of the corresponding container based on the weight value of the group, and configure matching computing power resources for the container.

2. The edge computing power prediction method according to claim 1, characterized in that: The distribution function of the service flow arrival is: Among them, λ1 is the periodic service arrival rate, λ2 is the random service arrival rate, δ is the impulse function, and t is the time; The periodic services and the random services are divided according to the service type characteristics of edge computing.

3. The edge computing power prediction method according to claim 2, characterized in that: The system efficiency of edge computing is calculated according to the distribution function of the service flow arrival, including: Perform Laplace transform on the distribution function of the traffic flow arrival to obtain the system efficiency factor; The system efficiency of edge computing is calculated based on the system efficiency factor.

4. The edge computing power prediction method according to claim 1, characterized in that: The calculation of the system efficiency ratio of each type of business flow includes: Count the proportions of different types of business flows, multiply the proportions of each type of business flow by the system efficiency, and get the system efficiency proportions of each type of business flow.

5. A computing power arrangement method for edge computing, characterized in that: The method comprises: Obtain multiple service flows for various terminal devices; The edge computing computing power prediction method according to claim 1 is used to predict the computing power values ​​required for various types of edge computing business flows; According to the computing power required by each type of business flow, each type of business flow is mapped to the application of the corresponding container, including: determining the application to which each type of business flow belongs, and forming a set of business flows belonging to the same application; sorting the set formed by the business flows belonging to the application according to the priority attribute of the application, and marking the weight value of the set; mapping the business flows in each set to the application of the corresponding container according to the weight value of the set.

6. The edge computing computing power arrangement method according to claim 5 is characterized in that: The application program for mapping various business flows to corresponding containers according to the computing power values ​​required by various business flows also includes: According to the real-time requirements of various business flows, the weight values ​​of the corresponding sets are modified; The business flows in the set are mapped to the application programs of the corresponding containers according to the modified weight values ​​of the set.

7. The edge computing computing power arrangement method according to claim 5, characterized in that: The method further comprises: Count the total computing power of applications corresponding to the same type of business flows; The applications corresponding to the same type of business flows are grouped into the same container, and computing resources matching the total computing power value of the applications are configured for the container.

8. The edge computing computing power arrangement method according to claim 5, characterized in that: The method further comprises: After acquiring multiple service flows of multiple terminal devices, the multiple service flows are cached to form a service flow mixed queue.

9. An edge computing computing power orchestration system, characterized in that: The system comprises: The computing power prediction module is used to receive multiple business flows from multiple terminal devices, cache the received multiple business flows to form a business flow cache queue, and predict the computing power required for each type of business flow according to the characteristics of different business flows of edge computing, including: counting the arrival rates of different types of business flows, setting the distribution function of business flow arrival according to the arrival rates of different types of business flows; calculating the system efficiency of edge computing according to the distribution function of business flow arrival, and calculating the system efficiency ratio of each type of business flow; under preset constraints, predicting the computing power required for each type of business flow according to the system efficiency ratio of each type of business flow; The computing power mapping module is used to map various business flows to the applications of the corresponding containers according to the computing power values ​​required by various business flows, specifically including: determining the applications to which various business flows belong, and forming a set of business flows belonging to the same application; sorting the sets formed by the business flows belonging to the application according to the priority attributes of the application, and marking the sets with weight values; mapping the business flows in each set to the applications of the corresponding containers according to the weight values ​​of the sets; The computing power prediction module predicts the computing power required for each type of business flow according to the system efficiency ratio of each type of business flow under preset constraints, specifically including: when the overflow probability of the business flow cache queue is less than or equal to the business loss rate constraint, and the average waiting delay of the business flow is less than or equal to the business processing delay constraint, the value range of the computing power required for each type of business flow is calculated according to the system efficiency ratio of each type of business flow; under the condition of a pre-set average business processing rate, the optimal computing power value within the value range of the computing power value is calculated, and the optimal computing power value is used as the computing power required for the business flow.

10. The edge computing computing power orchestration system according to claim 9, characterized in that: The distribution function of the service flow arrival is: Among them, λ1 is the periodic service arrival rate, λ2 is the random service arrival rate, δ is the impulse function, and t is the time; The periodic services and the random services are divided according to the service type characteristics of edge computing.

11. The edge computing computing power orchestration system according to claim 10, characterized in that: The system efficiency of edge computing is calculated according to the distribution function of the service flow arrival, including: Perform Laplace transform on the distribution function of the traffic flow arrival to obtain the system efficiency factor; The system efficiency of edge computing is calculated based on the system efficiency factor.

12. The edge computing computing power orchestration system according to claim 9, characterized in that: The system further comprises: The service flow cache module is used to cache multiple service flows after acquiring multiple service flows of multiple terminal devices to form a service flow mixed queue.

13. The edge computing computing power orchestration system according to claim 9, characterized in that: The system further comprises: The service interface module is used to receive multiple service flows from multiple terminal devices through multiple communication ports.

14. A storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed, the edge computing computing power prediction method described in any one of claims 1-4, or the edge computing computing power orchestration method described in any one of claims 5-8 is implemented.

Images