System capacity prediction method, device and equipment of internet of things platform, and storage medium

Abstract

The application provides a system capacity prediction method, device and equipment of an Internet of Things platform and a storage medium. The method comprises the following steps: obtaining running indexes of a plurality of service layers of the Internet of Things platform and the business volumes of the running indexes at different time points in a set time period, the set time being a continuous time period before the current time point; generating a data pair set of the service health degrees and the business volumes at each time point in the set time period according to the running indexes and the business volumes of the running indexes at different time points; training a nonlinear regression prediction model of the business volumes and the health degrees according to the data pair set; and predicting the system capacity of the Internet of Things platform according to the nonlinear regression prediction model of the business volumes and the health degrees. The method improves the accuracy of the system capacity prediction of the Internet of Things platform.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a system capacity prediction method, apparatus, device and storage medium for an Internet of Things (IoT) platform. Background Technology

[0002] With the continuous development of IoT technology and the expanding application areas, the number of IoT platform connections and the number of core network devices they interface with are rapidly increasing. This has led to increasingly complex IoT platform software architectures and a growing number of scenarios supported by IoT platforms, ranging from high-bandwidth 5G devices to medium-bandwidth Cat1 devices and low-bandwidth NB-IoT devices. These different scenarios result in significant differences in API requests, billing requests, call detail record (CDR) storage, and core network commands, leading to varying system capacity requirements. Furthermore, to cope with short-term business surges during special holidays such as Singles' Day (11.11), billing cycles, and the start of the school season, and to plan host resources accordingly, accurate prediction of the IoT platform's system capacity is necessary.

[0003] Existing technology monitors system operation data and hardware operation data, determines system data capacity based on the system operation data, further obtains capacity correlation parameters between system operation data and hardware operation data, and determines system physical capacity based on these parameters. Finally, the system capacity is determined based on both system data capacity and system physical capacity.

[0004] However, the method of determining the current system capacity solely through system operation data and hardware operation data cannot determine the system capacity of the IoT platform in a timely manner during the transition between different scenarios, resulting in an inaccurate prediction method for the system capacity of the IoT platform. Summary of the Invention

[0005] This application provides a system capacity prediction method, apparatus, device, and storage medium for an Internet of Things (IoT) platform to solve the problem of inaccurate system capacity prediction for IoT platforms.

[0006] In a first aspect, this application provides a system capacity prediction method for an Internet of Things (IoT) platform, comprising:

[0007] The system acquires the operational metrics of multiple service layers of an IoT platform within a set time period and their business volume at different points in time. The set time period includes a continuous time period prior to the current time. The IoT platform includes multiple service layers, and each service layer includes multiple operational metrics.

[0008] The operational indicators are normalized based on their base threshold and maximum threshold to obtain the health data of each operational indicator.

[0009] The health data of each operational indicator are weighted and summed to obtain the service health at each time point within a set time period.

[0010] Based on the service health level and its business volume at different time points, generate a set of data pairs of service health level and business volume at each time point within the set time period;

[0011] The nonlinear regression prediction model for business volume and health status of the dataset is trained based on the data.

[0012] The system capacity of the IoT platform is predicted based on the nonlinear regression prediction model of the business volume and health status.

[0013] Secondly, this application provides a system capacity prediction device for an Internet of Things (IoT) platform, comprising:

[0014] The acquisition module is used to acquire the operating metrics of multiple service layers of the IoT platform within a set time period and their business volume at different time points. The set time period is a continuous time period including the current time point. The IoT platform includes multiple service layers, and each service layer includes multiple operating metrics.

[0015] The generation module is used to normalize the operation indicators according to the basic threshold and the highest threshold of each operation indicator to obtain the health data of each operation indicator; perform weighted summation on the health data of each operation indicator to obtain the service health at each time point within a set time period; and generate a set of data pairs of service health and business volume at each time point within the set time period based on the service health and its business volume at different time points.

[0016] The training module is used to train a nonlinear regression prediction model for business volume and health status of the set based on the data.

[0017] The prediction module is used to predict the system capacity of the IoT platform based on a nonlinear regression prediction model of the business volume and health status.

[0018] Thirdly, this application provides a system capacity prediction device for an Internet of Things (IoT) platform, comprising:

[0019] Processor, memory, communication interface;

[0020] The memory is used to store the executable instructions of the processor;

[0021] The processor is configured to execute the system capacity prediction method for the IoT platform as described in the first aspect above by executing the executable instructions.

[0022] Fourthly, this application provides a readable storage medium, comprising: storing thereon a computer program, which, when executed by a processor, implements the system capacity prediction method for the Internet of Things platform as described in the first aspect above.

[0023] The system capacity prediction method, apparatus, device, and storage medium for the Internet of Things (IoT) platform provided in this application acquire the operating indicators of multiple service layers of the IoT platform within a set time period and their service volume at different time points. Based on the operating indicators and their service volume at different time points, a set of data pairs of service health and service volume at each time point within the set time period is generated. The service health at different time points is generated using the operating indicators of the IoT platform for predicting the current system capacity of the IoT platform, thus improving the accuracy of the system capacity prediction method. Furthermore, a nonlinear regression prediction model for service volume and health is trained based on the set of data pairs, and the system capacity of the IoT platform is predicted based on this nonlinear regression prediction model. The combination of service health and service volume at each time point to generate a set of data pairs for training the nonlinear regression prediction model further improves the accuracy of the system capacity prediction method. Attached Figure Description

[0024] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0025] Figure 1 A flowchart illustrating the system capacity prediction method for the IoT platform provided in this application embodiment;

[0026] Figure 2 This is a flowchart illustrating the process of generating a set of data pairs of service health and business volume at different points in time within a set time period, based on operational indicators and their business volume at different points in time, as provided in the embodiments of this application.

[0027] Figure 3 A flowchart illustrating the process of training a nonlinear regression prediction model for business volume and health based on data set according to an embodiment of this application;

[0028] Figure 4 This is a schematic diagram illustrating the process of updating the nonlinear regression prediction model for business volume and health status provided in the embodiments of this application.

[0029] Figure 5 A flowchart illustrating another method for predicting the system capacity of an IoT platform provided in this application embodiment;

[0030] Figure 6A schematic diagram illustrating the process of predicting the system capacity of an IoT platform using a nonlinear regression prediction model based on business volume and health status, as provided in this application embodiment.

[0031] Figure 7 A schematic diagram of the structure of a system capacity prediction device for an Internet of Things platform provided in this application embodiment;

[0032] Figure 8 This is a schematic diagram of the structure of a system capacity prediction device for an Internet of Things platform provided in an embodiment of this application.

[0033] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0034] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0035] Existing technologies monitor system and hardware operation data, determine system data capacity based on the system operation data, further obtain capacity correlation parameters between the system and hardware operation data, and determine the system physical capacity based on these parameters. Finally, the system capacity is determined based on the system data capacity and the system physical capacity. However, this method of determining the current system capacity solely through system and hardware operation data cannot timely determine the system capacity of the IoT platform during various scenario transitions, resulting in inaccurate system capacity prediction methods for IoT platforms.

[0036] This application obtains operational metrics of multiple service layers of an IoT platform within a set time period and their business volume at different time points. Based on the operational metrics and their business volume at different time points, it generates a set of data pairs of service health and business volume at each time point within the set time period. Specifically, by generating service health at different time points using the operational metrics of the IoT platform, it is used to predict the current system capacity of the IoT platform, thus improving the accuracy of the system capacity prediction method for the IoT platform. Furthermore, it trains a nonlinear regression prediction model for business volume and health based on the data pair set, and then uses this model to predict the system capacity of the IoT platform. Specifically, by combining the service health and business volume at each time point to generate a set of data pairs for training the nonlinear regression prediction model for business volume and health, it further improves the accuracy of the system capacity prediction method for the IoT platform.

[0037] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0038] Figure 1 This is a flowchart illustrating the system capacity prediction method for the Internet of Things platform provided in the first embodiment of this application.

[0039] like Figure 1 As shown, the system capacity prediction method for the IoT platform in this embodiment may include the following steps:

[0040] Step S101: Obtain the operating metrics of multiple service layers of the IoT platform within a set time period and their business volume at different time points. The set time period is a continuous time period including the current time point.

[0041] Specifically, the operating metrics of the current IoT platform can be obtained within a set time period. The set time period includes a continuous period of time prior to the current time point, and the length of the continuous period of time can be preset according to user needs.

[0042] Specifically, the IoT platform includes multiple service layers, so it is possible to obtain the operating metrics of multiple service layers of the IoT platform within a set time period.

[0043] Optionally, in addition to obtaining the operating metrics of multiple service layers of the IoT platform within a set time period, the operating metrics of the core network elements corresponding to the IoT platform can also be obtained.

[0044] Specifically, it is possible to obtain the business volume of the IoT platform at different points in time within the aforementioned set time period.

[0045] Optionally, the operating metrics of multiple service layers of the IoT platform within the specified time period and their business volume at different points in time can be stored in the memory of the system capacity prediction device of the IoT platform provided in this application.

[0046] Step S102: Based on the operational indicators and their business volume at different time points, generate a set of data pairs of service health and business volume at each time point within a set time period.

[0047] Specifically, based on the current operating indicators of the IoT platform within the set time period obtained in step S101, and the business volume of the IoT platform at different time points within the set time period, a set of data pairs of service health and business volume at each time point within the set time period can be generated.

[0048] Optionally, the operational metrics of the current IoT platform within the specified time period can be used to obtain the service health at different points in time within the specified time period through methods such as normalization and weighted summation. This health can then be combined with the business volume of the IoT platform at different points in time within the specified time period to generate a set of data pairs of service health and business volume at each point in time within the specified time period. This set can then be used to train a nonlinear regression prediction model for business volume and health.

[0049] Optionally, the service health at different points in time within the specified time period can be obtained by determining a health model. This application does not limit the method of determining the health model.

[0050] Step S103: Train the nonlinear regression prediction model for business volume and health status based on the data set.

[0051] Specifically, the data set of service health and business volume at each time point within the set time period generated in step S103 can be used as a training set to train the nonlinear regression prediction model of business volume and health.

[0052] Optionally, multiple nonlinear regression models, such as univariate polynomial nonlinear regression models, can be selected, and the regression errors of each nonlinear regression model can be combined to establish a nonlinear regression prediction model for business volume and health.

[0053] Step S104: Predict the system capacity of the IoT platform based on the nonlinear regression prediction model of business volume and health status.

[0054] Specifically, the system capacity of the current IoT platform can be predicted based on the nonlinear regression prediction model of business volume and health generated in step S103.

[0055] The nonlinear regression prediction model for service volume and health can calculate the correlation between the health of each service and the system capacity. Optionally, the nonlinear regression prediction model for service volume and health can be combined with a preset health threshold to predict the system capacity of the current IoT platform.

[0056] The system capacity prediction method for an IoT platform provided in this embodiment obtains the operating indicators of multiple service layers of the IoT platform within a set time period and their service volume at different time points. Based on the operating indicators and their service volume at different time points, a set of data pairs of service health and service volume at each time point within the set time period is generated. The service health at different time points is generated using the operating indicators of the IoT platform and used for predicting the current system capacity of the IoT platform, thus improving the accuracy of the system capacity prediction method. Furthermore, a nonlinear regression prediction model for service volume and health is trained based on the set of data pairs, and the system capacity of the IoT platform is predicted based on this nonlinear regression prediction model. The combination of service health and service volume at each time point to generate a set of data pairs is used for training the nonlinear regression prediction model, further improving the accuracy of the system capacity prediction method.

[0057] Figure 2 This is a flowchart illustrating the process of generating a set of service health and traffic data pairs for each time point within a set time period based on operational indicators and their traffic volume at different time points, as provided in the second embodiment of this application. Figure 1 Based on the illustrated embodiment, this embodiment elaborates on the process of generating a data set of service health and business volume at each time point within a set time period, according to operational indicators and their business volume at different time points.

[0058] like Figure 2 As shown, in this embodiment, generating a set of data pairs of service health and business volume at each time point within a set time period based on operational indicators and their business volume at different time points may include the following steps:

[0059] Step S201: Normalize the operating indicators based on the basic threshold and the highest threshold of each operating indicator to obtain the health data of each operating indicator.

[0060] The IoT platform described in step S101 includes multiple service layers, each of which includes multiple operational metrics. Optionally, the service layers of the IoT platform may include: Infrastructure as a Service (IaaS) layer, Platform as a Service (PaaS) layer, and Software as a Service (SaaS) layer.

[0061] Specifically, after obtaining the operational indicators described in step S101, the operational indicators can be normalized based on their basic thresholds (i.e., the minimum threshold and the maximum threshold) to obtain the health data of each operational indicator. The calculation formula for the health data of each operational indicator is as follows:

[0062]

[0063] Where, r i For the i-th operating metric, h i This is the normalized result of the i-th operational indicator, i.e., the health data of the i-th operational indicator.

[0064] Specifically, the above formula can be used to normalize the operational indicators at each time point within a set time period, resulting in a health dataset (h1, h2, ... h1) corresponding to the operational indicators at each time point. i , ..., h n ), where n is the number of operating indicators.

[0065] Step S202: Perform weighted summation on the health data of each operational indicator to obtain the service health status at each time point within the set time period.

[0066] Specifically, the health data of each operational indicator obtained in step S201 can be weighted and summed according to the weight values ​​corresponding to each operational indicator to obtain the service health at each time point within a set time period. The weight values ​​corresponding to each operational indicator can be preset according to user needs. The calculation formula for the service health at each time point within the set time period is as follows:

[0067]

[0068] Where, λ i For the i-th operating indicator r i The weight value is H, where H represents the service health level.

[0069] Specifically, the service health of the IoT platform at each point in time within a set time period can be calculated using the formula described above, thus obtaining the set of service health of the current IoT platform within the preset time period.

[0070] Step S203: Based on the service health level and its business volume at different time points, generate a set of data pairs of service health level and business volume at each time point within a set time period.

[0071] Specifically, based on the service health level and corresponding business volume at each time point within the set time period obtained in step S202, a set of data pairs of service health level and business volume at each time point within the set time period can be generated. The data content in each data pair set is <service health level, business volume>.

[0072] This embodiment provides a process for generating a set of service health and business volume data pairs for each time point within a set time period based on operational indicators and their business volume at different time points. Through normalization and weighted summation, the service health of the current IoT platform at each time point within the set time period is generated, improving the accuracy of the generated service health and further enhancing the accuracy of the IoT platform's system capacity prediction method.

[0073] Figure 3 This is a flowchart illustrating the process of training a nonlinear regression prediction model for business volume and health based on a data set, as provided in the third embodiment of this application. Figure 1 or Figure 2 Based on the illustrated embodiment, this embodiment elaborates on the process of training a nonlinear regression prediction model for business volume and health status based on a data set.

[0074] like Figure 3 As shown, training the nonlinear regression prediction model for business volume and health based on the data set in this embodiment may include the following steps:

[0075] Step S301: Obtain multiple univariate polynomial nonlinear regression models.

[0076] Specifically, multiple univariate polynomial nonlinear regression models can be obtained, and this application does not impose specific limitations on the obtained univariate polynomial nonlinear regression models.

[0077] Step S302: Based on the data set, perform machine learning on multiple univariate polynomial nonlinear regression models respectively to obtain nonlinear regression prediction models for business volume and health corresponding to the univariate polynomial nonlinear regression models.

[0078] Specifically, based on the set of data pairs of service health and business volume at each time point within the set time period generated in step S102 or step S203, machine learning can be performed on the multiple univariate polynomial nonlinear regression models obtained in step S301 to obtain nonlinear regression prediction models of business volume and health corresponding to the univariate polynomial nonlinear regression models.

[0079] Step S303: Calculate the regression error of the nonlinear regression prediction model for each business volume and health status, and select the nonlinear regression prediction model for business volume and health status with the smallest regression error as the final nonlinear regression prediction model for business volume and health status.

[0080] Specifically, after obtaining the nonlinear regression prediction models for business volume and health corresponding to each univariate polynomial nonlinear regression model, the regression error of each nonlinear regression prediction model for business volume and health can be calculated. Finally, the nonlinear regression prediction model for business volume and health with the smallest regression error is selected as the final nonlinear regression prediction model for business volume and health.

[0081] This embodiment provides a process for training a nonlinear regression prediction model for business volume and health based on a data set. This involves acquiring multiple univariate polynomial nonlinear regression models, then performing machine learning on each of these models based on the data set to obtain the corresponding nonlinear regression prediction models for business volume and health. Finally, the regression error of each nonlinear regression prediction model for business volume and health is calculated, and the model with the smallest regression error is selected as the final nonlinear regression prediction model for business volume and health. By selecting univariate polynomial nonlinear regression models and combining the regression error values ​​of each model for training the nonlinear regression prediction model for business volume and health, the accuracy of the nonlinear regression prediction model for business volume and health is improved, further enhancing the accuracy of the system capacity prediction method for the IoT platform.

[0082] Figure 4 This is a schematic diagram illustrating the process of updating the nonlinear regression prediction model for business volume and health status provided in the fourth embodiment of this application. Figure 3 Based on the illustrated embodiment, since the IoT platform is a platform with rapidly growing business, the nonlinear regression prediction model with fixed business volume and health status cannot accurately predict the current system capacity of the IoT platform. Therefore, it is necessary to update the nonlinear regression prediction model with business volume and health status. This embodiment describes in detail the process of updating the nonlinear regression prediction model with business volume and health status.

[0083] like Figure 4As shown, updating the nonlinear regression prediction model for business volume and health in this embodiment may include the following steps:

[0084] Step S401: Set the update time for the nonlinear regression prediction model of business volume and health status.

[0085] Specifically, the update time of the nonlinear regression prediction model for business volume and health status can be preset according to user needs.

[0086] Step S402: If the current time exceeds the update time, re-acquire the operating indicators of multiple service layers of the IoT platform within the set time period and their business volume at different time points.

[0087] Specifically, if the current time exceeds the preset update time in step S401, the operating indicators of multiple service layers of the IoT platform within the set time period and their business volume at different time points can be re-acquired. The process of acquiring the operating indicators of multiple service layers of the IoT platform within the set time period and their business volume at different time points can be referred to the description in step S101 above, and will not be repeated in this embodiment.

[0088] Step S403: Based on the multiple re-acquired operational metrics and their business volume at different time points, obtain the updated set of data pairs of service health and business volume at each time point within the set time period.

[0089] Specifically, based on the multiple operating indicators re-acquired in step S402 and their business volume at different time points, an updated set of data pairs of service health and business volume at each time point within a set time period can be obtained. The process of obtaining the set of data pairs of service health and business volume at each time point within a set time period can be referred to the description in step S102 above, and will not be repeated in this embodiment.

[0090] Step S404: Train the nonlinear regression prediction model for the business volume and health of the dataset based on the updated data to obtain the updated nonlinear regression prediction model for the business volume and health.

[0091] Specifically, the nonlinear regression prediction model for the set of business volume and health can be trained based on the updated data obtained in step S403 to obtain the updated nonlinear regression prediction model for business volume and health. The process of obtaining the nonlinear regression prediction model for business volume and health can be referred to the description in step S103, which will not be repeated in this embodiment.

[0092] Optionally, the system capacity of the current IoT platform can be predicted based on the updated nonlinear regression prediction model of business volume and health.

[0093] Figure 5 This is a flowchart illustrating another method for predicting the system capacity of an IoT platform provided in an embodiment of this application.

[0094] The process of updating the nonlinear regression prediction model for business volume and health provided in this embodiment involves re-acquiring the training set of the nonlinear regression prediction model for business volume and health at a preset update time, and training the nonlinear regression prediction model for business volume and health based on the re-acquired training set to obtain the updated nonlinear regression prediction model for business volume and health. This ensures the real-time performance of the nonlinear regression prediction model for business volume and health, and further improves the accuracy of the system capacity prediction method of the Internet of Things platform.

[0095] Figure 6 This is a flowchart illustrating the process of predicting the system capacity of an IoT platform using a nonlinear regression prediction model based on business volume and health status, as provided in the fifth embodiment of this application. Figure 4 Based on the illustrated embodiment, this embodiment elaborates on the process of predicting the system capacity of the Internet of Things platform using a nonlinear regression prediction model based on business volume and health status.

[0096] like Figure 6 As shown, the method for predicting the system capacity of an IoT platform using a nonlinear regression prediction model based on business volume and health status in this embodiment may include the following steps:

[0097] Step S601: Based on the nonlinear regression prediction model of business volume and health, calculate the system capacity threshold corresponding to the service health of the IoT platform when the health threshold is reached.

[0098] The higher the service health of an IoT platform, the healthier the IoT platform, and the more capable its system capacity is of handling the current business volume. Conversely, the lower the service health of an IoT platform, the less healthy the IoT platform is, and the less capable its system capacity is of handling the current business volume.

[0099] Specifically, based on the nonlinear regression prediction model of the business volume and health obtained in step S404, the system capacity threshold corresponding to the service health of the IoT platform when it reaches the health threshold can be calculated. The health threshold is the minimum health level for judging the IoT platform as unhealthy, and can be preset according to user needs. The system capacity threshold is the maximum business volume of the IoT platform corresponding to the health threshold.

[0100] Step S602: Use the system capacity threshold as the system capacity of the IoT platform.

[0101] Specifically, the system capacity threshold calculated in step S601 can be used as the system capacity of the IoT platform.

[0102] Optionally, expansion timelines and the required system capacity over a certain period can be preset based on business growth trends. Specifically, based on future business development, it can be determined whether the current IoT platform's predicted system capacity can meet future business needs. Expansion timelines and the required system capacity over a certain period can be preset based on future business development. If the required system capacity over a certain period is less than the system capacity threshold, no expansion will be performed on the IoT platform. If the required system capacity over a certain period exceeds the system capacity threshold, the IoT platform will be expanded at the preset expansion timelines. This ensures that the IoT platform maintains high service health and that the system's hardware and software resources operate at a reasonable level of high utilization. The specific expansion values ​​can be referenced from the preset system capacity required over a certain period.

[0103] This embodiment provides a process for predicting the system capacity of an IoT platform using a nonlinear regression prediction model based on service volume and health status. By using the nonlinear regression prediction model based on service volume and health status, the system capacity threshold corresponding to the service health status of the IoT platform is calculated. This system capacity threshold is then used as the system capacity of the IoT platform. By using the system capacity threshold corresponding to the service health status as the health status threshold as the system capacity of the IoT platform, the accuracy of the system capacity prediction method for the IoT platform is improved.

[0104] Figure 7 This is a schematic diagram of the structure of a system capacity prediction device for an Internet of Things platform provided in the sixth embodiment of this application.

[0105] like Figure 7 As shown, the system capacity prediction device 70 of the IoT platform in this embodiment includes an acquisition module 71, a generation module 72, a training module 73, and a prediction module 74.

[0106] The acquisition module 71 is used to acquire the operating indicators of multiple service layers of the IoT platform within a set time period and their business volume at different time points. The set time period is a continuous time period including the current time point.

[0107] The generation module 72 is used to generate a set of data pairs of service health and business volume at each time point within a set time period, based on the operating indicators and their business volume at different time points.

[0108] Training module 73 is used to train a nonlinear regression prediction model for business volume and health status based on the data set.

[0109] The prediction module 74 is used to predict the system capacity of the IoT platform based on a nonlinear regression prediction model of business volume and health status.

[0110] The apparatus provided in this embodiment can be used to execute the above-described method embodiments. Figures 1 to 6 The technical solution is similar in principle and effect, and will not be described again in this embodiment.

[0111] Figure 8 This is a schematic diagram of the structure of a system capacity prediction device for an Internet of Things platform provided in the seventh embodiment of this application.

[0112] like Figure 8 As shown, the system capacity prediction device 80 of the IoT platform in this embodiment includes: a processor 81, a memory 82, and a communication interface 83.

[0113] Memory 82 is used to store the processor's executable instructions.

[0114] The processor 81 is configured to execute the above method embodiments by executing executable instructions. Figures 1 to 6 Any method for predicting the system capacity of an IoT platform.

[0115] In the above Figure 8 In the illustrated embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0116] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0117] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0118] This application also provides a readable storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the above-described method embodiments. Figures 1 to 6 Any method for predicting the system capacity of an IoT platform.

[0119] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0120] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0121] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A system capacity prediction method for an Internet of Things (IoT) platform, characterized in that, include: The system acquires the operational metrics of multiple service layers of an IoT platform within a set time period and their business volume at different points in time. The set time period includes a continuous time period prior to the current time. The IoT platform includes multiple service layers, and each service layer includes multiple operational metrics. The operational indicators are normalized based on their base threshold and maximum threshold to obtain the health data of each operational indicator. The health data of each operational indicator are weighted and summed to obtain the service health at each time point within a set time period. Based on the service health level and its business volume at different time points, generate a set of data pairs of service health level and business volume at each time point within the set time period; The nonlinear regression prediction model for business volume and health status of the dataset is trained based on the data. The system capacity of the IoT platform is predicted based on the nonlinear regression prediction model of the business volume and health status.

2. The method according to claim 1, characterized in that, The step of training a nonlinear regression prediction model for business volume and health based on the data set includes: Obtain multiple univariate polynomial nonlinear regression models; Based on the data set, machine learning is performed on multiple univariate polynomial nonlinear regression models to obtain nonlinear regression prediction models for business volume and health corresponding to the univariate polynomial nonlinear regression models. Calculate the regression error of each of the nonlinear regression prediction models for business volume and health, and select the nonlinear regression prediction model for business volume and health with the smallest regression error as the final nonlinear regression prediction model for business volume and health.

3. The method according to claim 2, characterized in that, The method further includes: Set the update time for the nonlinear regression prediction model of business volume and health status; If the current time exceeds the update time, re-acquire the operating metrics of multiple service layers of the IoT platform within the set time period and their business volume at different time points; Based on the re-acquired multiple operational metrics and their business volume at different time points, an updated set of data pairs of service health and business volume at each time point within a set time period is obtained. The updated nonlinear regression prediction model for business volume and health is trained on the dataset based on the updated data to obtain the updated nonlinear regression prediction model for business volume and health.

4. The method according to claim 3, characterized in that, The prediction of the system capacity of the IoT platform based on the nonlinear regression prediction model of the business volume and health status includes: Based on the nonlinear regression prediction model of the business volume and health status, calculate the system capacity threshold corresponding to the service health status of the IoT platform when it reaches the health status threshold. The system capacity threshold is used as the system capacity of the IoT platform.

5. The method according to claim 4, characterized in that, The method further includes: If the required system capacity exceeds the system capacity threshold within a certain period of time in the future, the IoT platform will be expanded at a preset expansion time point.

6. The method according to claim 5, characterized in that, The method further includes: Based on business growth trends, the expansion time point and the system capacity required within a certain future period are preset.

7. A system capacity prediction device for an Internet of Things (IoT) platform, characterized in that, include: The acquisition module is used to acquire the operating metrics of multiple service layers of the IoT platform within a set time period and their business volume at different time points. The set time period is a continuous time period including the current time point. The IoT platform includes multiple service layers, and each service layer includes multiple operating metrics. The generation module is used to normalize the operational indicators based on the basic threshold and the highest threshold of each operational indicator to obtain the health data of each operational indicator. The health data of each operational indicator are weighted and summed to obtain the service health at each time point within a set time period. Based on the service health level and its business volume at different time points, generate a set of data pairs of service health level and business volume at each time point within the set time period; The training module is used to train a nonlinear regression prediction model for business volume and health status of the set based on the data. The prediction module is used to predict the system capacity of the IoT platform based on a nonlinear regression prediction model of the business volume and health status.

8. A system capacity prediction device for an Internet of Things (IoT) platform, characterized in that, include: Processor, memory, communication interface; The memory is used to store the executable instructions of the processor; The processor is configured to execute the system capacity prediction method of the Internet of Things platform according to any one of claims 1 to 6 by executing the executable instructions.

9. A readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the system capacity prediction method of the Internet of Things platform according to any one of claims 1 to 6.

Images