Network resource intelligent scheduling method and device based on cloud computing
By building user service usage sequences and performance data sequences, setting QoS thresholds, dividing hot services and cold services, and adjusting network resources in combination with service level agreements, the problem of low adaptability and accuracy of network resource scheduling in the prior art is solved. , realizes the precise allocation and efficient utilization of resources.
Patent Information
- Application Number
- CN202410911728.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-09
- Publication Date
- 2024-11-29
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, network resource scheduling has low adaptability and accuracy, making it difficult to meet dynamically changing service needs, resulting in resource waste and performance bottlenecks.
By collecting data related to users' historical service, building service usage sequences and performance data sequences, setting QoS thresholds, dividing hot and cold services, determining initial network resources in combination with service level agreements, monitoring actual usage, and adjusting resource allocation to match targets Resources required.
It improves the accuracy and adaptability of network resource scheduling, ensures reasonable allocation and adjustment of resources, reduces resource waste, and adapts to the service needs of different users.
Smart Images

Figure CN119052336A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of network resource management, and more specifically, to a method and device for intelligent scheduling of network resources based on cloud computing. Background Art
[0002] With the rapid development of cloud computing technology, users' demand for network resources is growing. How to efficiently manage and schedule network resources has become a key challenge. Traditional manual resource allocation methods are difficult to adapt to dynamically changing service needs, resulting in resource waste and performance bottlenecks. Moreover, different users have different service usage habits and different resource needs.
[0003] Therefore, how to improve the accuracy and adaptability of network resource scheduling is a technical problem that needs to be solved. Summary of the invention
[0004] The present invention provides a network resource intelligent scheduling method based on cloud computing, which is used to solve the technical problems of low adaptability and accuracy of network resource scheduling in the prior art. The method comprises:
[0005] Collect each user's historical service-related data, and build each user's service usage sequence and performance data sequence through the historical service-related data. According to the service usage sequence and performance data sequence, build each user's standard service usage chain in multiple time periods, and set the QoS threshold in each time period.
[0006] Determine the popularity of each service through historical service-related data of users, and divide the services into hot services and cold services according to the popularity. Mark the hot services and cold services in the standard service usage chain, determine the service level agreement of each service, and combine the service level agreement and the standard service usage chain to determine the initial network resources required for each service in different time periods;
[0007] Allocate initial network resources for services, monitor actual service usage by users over a period of time, and determine the QoS of the current service. Determine whether to adjust initial network resources based on the QoS and QoS threshold of the current service.
[0008] If the initial network resources are adjusted, the target resources required for the service are evaluated and resource scheduling is performed based on the target resources required.
[0009] In some embodiments of the present application, a standard service usage chain for each user in multiple time periods is constructed according to the service usage sequence and the performance data sequence, including:
[0010] The service usage sequence includes the service usage sequence and the service usage data sequence. The service usage sequence is the service usage record over time. The service usage data sequence is the service usage related parameters over time. The performance data sequence is the service performance related parameters over time.
[0011] Perform time alignment processing on the service usage sequence, service usage data sequence and performance data sequence, and plan the service usage sequence of the user in the same period according to the service usage times of multiple service usage sequences in the same period;
[0012] Determine the standard range of each type of data by using the characteristic values in multiple service usage data sequences and performance data sequences in the same period, and mark the standard range of each type of data under the corresponding service in the service usage sequence;
[0013] The service usage sequences marked with standard ranges in different time periods are connected in chronological order, thereby constructing a standard service usage chain for each user in multiple time periods.
[0014] In some embodiments of the present application, the standard range of each type of data is determined by characteristic values in multiple service usage data sequences and performance data sequences in the same period, including:
[0015] The eigenvalues include the most frequent minimum value, the most frequent maximum value, and the most frequent median value;
[0016] The intersection of the values of multiple service usage data series and multiple performance data series is taken as the intersection range, and the distances from the two ends of the intersection range to the minimum value with the highest frequency and the maximum value with the highest frequency are calculated respectively, which are recorded as the first distance and the second distance;
[0017] Calculate the distance from the median with the highest frequency to the two ends of the intersection range respectively, and record them as the third distance and the fourth distance;
[0018] The extension ratios of the two ends of the intersection range are respectively mapped by the third distance and the fourth distance, and the two ends of the intersection range are extended by the extension ratio, the first distance and the second distance to obtain the standard range of each data.
[0019] In some embodiments of the present application, setting the QoS threshold for each time period includes:
[0020] De-noising, cleaning and normalizing each type of data required for service quality in the service usage data series and performance data series;
[0021] By service type, each type of data required for service quality is weighted and summed to obtain a comprehensive index, and the QoS threshold is set based on the comprehensive index of multiple services;
[0022] ;
[0024] Among them, QoS 阈 is the QoS threshold, n1 is the number of services currently used, n2 is the number of data types required for the i-th service quality, W i is the comprehensive indicator of the i-th service, QoS 总 The service quality is the sum of the comprehensive indicators of all services currently used, Indicates the service quality mapped by the sum of comprehensive indicators of all currently used services.
[0025] In some embodiments of the present application, the popularity of each service is determined by using the user's historical service-related data, including:
[0026] De-noising, cleaning and normalizing each type of data required for hot and cold situations in service usage data series and performance data series;
[0027]
[0028] Among them, C j is the hotness and coldness of the jth service, m is the number of each type of data required for the hotness and coldness of the jth service, α a The hot and cold weights of the data required for the a-th hot and cold situation of the j-th service, E a The data size required for the a-th hot and cold situation of the j-th service, R1 j , R2 j 、R3 j are the number of users, number of user visits, and duration of user use of the jth service, respectively. f(R1 j , R2 j , R3 j 0 is the user usage degree obtained by the number of users of the jth service, the number of user visits, and the user usage time, and k is a preset constant.
[0029] In some embodiments of the present application, the service level agreement and the standard service usage chain are combined to determine the initial network resources required for each service in different time periods, including:
[0030] Determine the total initial network resource interval for each service through a service level agreement, and select a portion of the total initial network resource interval as a first resource interval through a data standard range of the service in the standard service usage chain;
[0031] Selecting a portion of the total initial network resource interval as the second resource interval by using the hot and cold types and hot and cold degrees of the services in the standard service usage chain;
[0032] If the first resource interval and the second resource interval have an intersection, the intersection interval is used as the initial network resource range required for the service;
[0033] Otherwise, for the hot service, determine the average value point between the first resource interval and the second resource interval, and determine the median point of the larger one of the first resource interval and the second resource interval, and use the average value point and the median point as two endpoints of the required initial network resource range;
[0034] For cold services, the average value point between the first resource interval and the second resource interval is determined, and the median point of the smaller one of the first resource interval and the second resource interval is determined, and the average value point and the median point are used as the two endpoints of the required initial network resource range.
[0035] In some embodiments of the present application, judging whether to adjust the initial network resources by the QoS and QoS threshold of the current service includes:
[0036] If the QoS of the current service is greater than the QoS threshold, no adjustment is made to the initial network resources;
[0037] Otherwise, the initial network resources are adjusted.
[0038] In some embodiments of the present application, if the initial network resources are adjusted, the target resources required for the service are evaluated, and resource scheduling is performed according to the target resources required, including:
[0039] The target resources required for the service are evaluated by the matching degree between the current service and the standard service usage chain, and the resources are scheduled, scaled and migrated according to the target resources required.
[0040] Correspondingly, the present application also provides a network resource intelligent scheduling device based on cloud computing, including:
[0041] A construction module is used to collect historical service-related data of each user, and to construct a service usage sequence and performance data sequence of each user through the historical service-related data, to construct a standard service usage chain for each user in multiple time periods according to the service usage sequence and performance data sequence, and to set a QoS threshold for each time period;
[0042] A determination module is used to determine the hotness and coldness of each service through the user's historical service-related data, and to classify the services into two types: hot services and cold services according to the hotness and coldness, and to mark the hot services and cold services in the standard service usage chain, and to determine the service level agreement of each service, and to determine the initial network resources required for each service in different time periods by combining the service level agreement and the standard service usage chain;
[0043] The monitoring module is used to allocate the initial network resources of the service, monitor the actual service usage of the user over a period of time, and determine the QoS of the current service. It determines whether to adjust the initial network resources based on the QoS and QoS threshold of the current service.
[0044] The scheduling module is used to evaluate the target resources required for the service if the initial network resources are adjusted, and to schedule resources according to the target resources required.
[0045] By applying the above technical solution, historical service-related data of each user is collected, and the service usage sequence and performance data sequence of each user are constructed through the historical service-related data. The standard service usage chain of each user in multiple time periods is constructed according to the service usage sequence and performance data sequence, and the QoS threshold in each time period is set; the hotness of each service is determined through the historical service-related data of the user, and the service is divided into two types of hot services and cold services according to the hotness, and the hot services and cold services are marked in the standard service usage chain, and the service level agreement of each service is determined. The initial network resources required for each service in different time periods are determined in combination with the service level agreement and the standard service usage chain; the initial network resources of the service are allocated, the actual service usage of the user over a period of time is monitored, and the QoS of the current service is determined, and the QoS and QoS threshold of the current service are used to determine whether to adjust the initial network resources; if the initial network resources are adjusted, the target resources required for the service are evaluated, and the resources are scheduled according to the target resources required. This application constructs a standard service usage chain for each user in multiple time periods to determine the typical needs and usage patterns of each user in different time periods, thereby helping to accurately allocate the subsequent initial network resources. The QoS and QoS threshold of the current service are used to determine whether to adjust the initial network resources, thereby improving the accuracy and adaptability of network resource scheduling and ensuring the reasonable allocation and adjustment of network resources. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.
[0047] Figure 1 A schematic diagram of a process of a network resource intelligent scheduling method based on cloud computing proposed in an embodiment of the present invention is shown;
[0048] Figure 2 A schematic diagram of the structure of a network resource intelligent scheduling device based on cloud computing proposed in an embodiment of the present invention is shown. DETAILED DESCRIPTION
[0049] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0050] The present application embodiment provides a network resource intelligent scheduling method based on cloud computing, such as Figure 1 As shown, the method comprises the following steps:
[0051] Step S101, collect historical service-related data of each user, and build each user's service usage sequence and performance data sequence through the historical service-related data, build a standard service usage chain for each user in multiple time periods based on the service usage sequence and performance data sequence, and set the QoS threshold for each time period.
[0052] In this embodiment, historical service related data includes service records, service usage data and service performance data, the service usage data includes access frequency, data transmission volume, response time, etc., and the service performance data includes CPU utilization, memory usage, network throughput, etc. The standard service usage chain is used to represent the typical needs and usage patterns of users for services in different time periods, that is, multiple services in chronological order and related parameters of the services (standard parameter range).
[0053] In some embodiments of the present application, a standard service usage chain for each user in multiple time periods is constructed according to the service usage sequence and the performance data sequence, including:
[0054] The service usage sequence includes the service usage sequence and the service usage data sequence. The service usage sequence is the service usage record over time. The service usage data sequence is the service usage related parameters over time. The performance data sequence is the service performance related parameters over time.
[0055] Perform time alignment processing on the service usage sequence, service usage data sequence and performance data sequence, and plan the service usage sequence of the user in the same period according to the service usage times of multiple service usage sequences in the same period;
[0056] Determine the standard range of each type of data by using the characteristic values in multiple service usage data sequences and performance data sequences in the same period, and mark the standard range of each type of data under the corresponding service in the service usage sequence;
[0057] The service usage sequences marked with standard ranges in different time periods are connected in chronological order, thereby constructing a standard service usage chain for each user in multiple time periods.
[0058] In this embodiment, the service usage sequence of the user in the same period is planned according to the service usage times of multiple service usage sequences in the same period, that is, the service usage sequence of the user is planned according to the historical service usage times. For example, in the period t1, the number of times the sequence of services x1-x2 is used is 31 times, and the number of times the sequence of services x3-x4 is used is 2 times, then services x1-x2 are used as the service sequence of the user in the period t1, and the services with more times and the corresponding sequence are used as the service usage sequence.
[0059] In this embodiment, the standard range of each data is determined by using characteristic values in multiple service usage data sequences and performance data sequences in the same time period. The user's demand for services in different time periods may be different, so a standard range for each data is specified. The standard range is the typical data range of these services for the user in the time period.
[0060] In some embodiments of the present application, the standard range of each type of data is determined by characteristic values in multiple service usage data sequences and performance data sequences in the same period, including:
[0061] The eigenvalues include the most frequent minimum value, the most frequent maximum value, and the most frequent median value;
[0062] The intersection of the values of multiple service usage data series and multiple performance data series is taken as the intersection range, and the distances from the two ends of the intersection range to the minimum value with the highest frequency and the maximum value with the highest frequency are calculated respectively, which are recorded as the first distance and the second distance;
[0063] Calculate the distance from the median with the highest frequency to the two ends of the intersection range respectively, and record them as the third distance and the fourth distance;
[0064] The extension ratios of the two ends of the intersection range are respectively mapped by the third distance and the fourth distance, and the two ends of the intersection range are extended by the extension ratio, the first distance and the second distance to obtain the standard range of each data.
[0065] In this embodiment, the extension ratios of the two ends of the intersection range are obtained by mapping the third distance and the fourth distance respectively. Different third distances and fourth distances correspond to different extension ratios, and the extension ratio is the extension ratio (percentage) of the first distance and the second distance. The two ends of the intersection range are extended by the extension ratio, the first distance and the second distance. For example, the extension ratios of the two ends are z1 and z2, which correspond to the first distance b1 and the second distance b2 respectively. z1*b1 and z2*b2 are the extension distances of the two ends of the intersection range, respectively, so as to obtain the standard range of the data.
[0066] In some embodiments of the present application, setting the QoS threshold for each time period includes:
[0067] De-noising, cleaning and normalizing each type of data required for service quality in the service usage data series and performance data series;
[0068] By service type, each type of data required for service quality is weighted and summed to obtain a comprehensive index, and the QoS threshold is set based on the comprehensive index of multiple services;
[0069]
[0070] Among them, QoS 阈 is the QoS threshold, n1 is the number of services currently used, n2 is the number of data types required for the i-th service quality, W i is the comprehensive indicator of the i-th service, QoS 总 The service quality is the sum of the comprehensive indicators of all services currently used, Indicates the service quality mapped by the sum of comprehensive indicators of all currently used services.
[0071] In this embodiment, multiple services are used simultaneously and may affect each other. The QoS threshold here is the QoS threshold (service quality threshold) of multiple services in the current period, and the QoS threshold is represented by an average value, which is larger than the general average value.
[0072] Step S102, determine the hotness or coldness of each service through the user's historical service-related data, and divide the services into two types: hot services and cold services according to the hotness or coldness, mark the hot services and cold services in the standard service usage chain, determine the service level agreement for each service, and combine the service level agreement and the standard service usage chain to determine the initial network resources required for each service in different time periods.
[0073] In this embodiment, the network resource requirements of popular services and unpopular services may be different. Here, we first define the degree of popularity and divide hot services into cold services. Service level agreement. Service providers usually provide service specifications, including expected performance and resource requirements. Analyze the service level agreement (SLA) to understand the performance goals and resource constraints of the service. The adjustment or scheduling of resources is carried out under this premise and will not exceed this scope.
[0074] In some embodiments of the present application, the popularity of each service is determined by using the user's historical service-related data, including:
[0075] De-noising, cleaning and normalizing each type of data required for hot and cold situations in service usage data series and performance data series;
[0076]
[0077] Among them, C j is the hotness and coldness of the jth service, m is the number of each type of data required for the hotness and coldness of the jth service, α a The hot and cold weights of the data required for the a-th hot and cold situation of the j-th service, E a The data size required for the a-th hot and cold situation of the j-th service, R1 j , R2 j 、R3 j are the number of users, number of user visits, and duration of user use of the jth service, respectively. f(R1 j , R2 j , R3 j ) is the user usage degree obtained by the number of users of the j-th service, the number of user visits, and the user usage time, and k is a preset constant.
[0078] In this embodiment, It means that each data required for the hot and cold conditions is modified by user usage, so as to obtain a hot and cold degree that can characterize the popularity of the service.
[0079] In some embodiments of the present application, the service level agreement and the standard service usage chain are combined to determine the initial network resources required for each service in different time periods, including:
[0080] Determine the total initial network resource interval for each service through a service level agreement, and select a portion of the total initial network resource interval as a first resource interval through a data standard range of the service in the standard service usage chain;
[0081] Selecting a portion of the total initial network resource interval as the second resource interval by using the hot and cold types and hot and cold degrees of the services in the standard service usage chain;
[0082] If the first resource interval and the second resource interval have an intersection, the intersection interval is used as the initial network resource range required for the service;
[0083] Otherwise, for the hot service, determine the average value point between the first resource interval and the second resource interval, and determine the median point of the larger one of the first resource interval and the second resource interval, and use the average value point and the median point as two endpoints of the required initial network resource range;
[0084] For cold services, the average value point between the first resource interval and the second resource interval is determined, and the median point of the smaller one of the first resource interval and the second resource interval is determined, and the average value point and the median point are used as the two endpoints of the required initial network resource range.
[0085] In this embodiment, the first resource interval is determined by considering the data standard range of the service, where the resource interval refers to the resource size range. When the first resource interval and the second resource interval have an intersection, the intersection interval is used as the network resource range. When the two do not have an intersection, different median points are selected for hot services and cold services to construct the endpoints of the network resource range. Hot services often require large fluctuations in resources and require more resources, while cold services are the opposite.
[0086] Step S103, allocating initial network resources for the service, monitoring the actual service usage of the user over a period of time, and determining the QoS of the current service, and determining whether to adjust the initial network resources based on the QoS of the current service and the QoS threshold.
[0087] In some embodiments of the present application, judging whether to adjust the initial network resources by the QoS and QoS threshold of the current service includes:
[0088] If the QoS of the current service is greater than the QoS threshold, no adjustment is made to the initial network resources;
[0089] Otherwise, the initial network resources are adjusted.
[0090] In this embodiment, the current service here refers to all services in the current period, and the average value of the QoS sum is taken as the QoS of the current service.
[0091] Step S104: if the initial network resources are adjusted, the target resources required for the service are evaluated, and resource scheduling is performed according to the target resources required.
[0092] In some embodiments of the present application, if the initial network resources are adjusted, the target resources required for the service are evaluated, and resource scheduling is performed according to the target resources required, including:
[0093] The target resources required for the service are evaluated by the matching degree between the current service and the standard service usage chain, and the resources are scheduled, scaled and migrated according to the target resources required.
[0094] In this embodiment, the specific target resources required for each service are evaluated, and the matching degree between the current service and the standard service usage chain is determined based on the data size range of the current service and the data standard range of the standard service usage chain, so as to determine the specific target resources required for each service. Scheduling, scaling and migration of resources, using a load balancer to distribute traffic to different service instances to balance the load. If a service has too many resources, the resource allocation of the service may be reduced to avoid wasting resources. If a service has insufficient resources, the resource allocation of the service may be increased to improve service performance. In some cases, if one service has too many resources and another service has insufficient resources, it may be considered to transfer resources from the service with too many resources to the service with fewer resources.
[0095] It should be noted that the user's standard service usage chain can also predict the user's demand for services in different time periods, so as to prepare resources in advance and avoid insufficient or excessive resources. As a benchmark for service quality evaluation, it helps cloud service providers understand the service quality level expected by users and adjust service strategies accordingly.
[0096] By applying the above technical solution, historical service-related data of each user is collected, and the service usage sequence and performance data sequence of each user are constructed through the historical service-related data. The standard service usage chain of each user in multiple time periods is constructed according to the service usage sequence and performance data sequence, and the QoS threshold in each time period is set; the hotness of each service is determined through the historical service-related data of the user, and the service is divided into two types of hot services and cold services according to the hotness, and the hot services and cold services are marked in the standard service usage chain, and the service level agreement of each service is determined. The initial network resources required for each service in different time periods are determined in combination with the service level agreement and the standard service usage chain; the initial network resources of the service are allocated, the actual service usage of the user over a period of time is monitored, and the QoS of the current service is determined, and the QoS and QoS threshold of the current service are used to determine whether to adjust the initial network resources; if the initial network resources are adjusted, the target resources required for the service are evaluated, and the resources are scheduled according to the target resources required. This application constructs a standard service usage chain for each user in multiple time periods to determine the typical needs and usage patterns of each user in different time periods, thereby helping to accurately allocate the subsequent initial network resources. The QoS and QoS threshold of the current service are used to determine whether to adjust the initial network resources, thereby improving the accuracy and adaptability of network resource scheduling and ensuring the reasonable allocation and adjustment of network resources.
[0097] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.
[0098] In order to further explain the technical idea of the present invention, the technical solution of the present invention is now described in combination with specific application scenarios.
[0099] Correspondingly, the present application also provides a network resource intelligent scheduling device based on cloud computing, such as Figure 2 As shown, including:
[0100] A construction module is used to collect historical service-related data of each user, and to construct a service usage sequence and performance data sequence of each user through the historical service-related data, to construct a standard service usage chain for each user in multiple time periods according to the service usage sequence and performance data sequence, and to set a QoS threshold for each time period;
[0101] A determination module is used to determine the hotness and coldness of each service through the user's historical service-related data, and to classify the services into two types: hot services and cold services according to the hotness and coldness, and to mark the hot services and cold services in the standard service usage chain, and to determine the service level agreement of each service, and to determine the initial network resources required for each service in different time periods by combining the service level agreement and the standard service usage chain;
[0102] The monitoring module is used to allocate the initial network resources of the service, monitor the actual service usage of the user over a period of time, and determine the QoS of the current service. It determines whether to adjust the initial network resources based on the QoS and QoS threshold of the current service.
[0103] The scheduling module is used to evaluate the target resources required for the service if the initial network resources are adjusted, and to schedule resources according to the target resources required.
[0104] Those skilled in the art will appreciate that the modules in the device in the implementation scenario can be distributed in the device in the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more devices different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple sub-modules.
[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.
Claims
1. A network resource intelligent scheduling method based on cloud computing, characterized in that: include: Collect each user's historical service-related data, and build each user's service usage sequence and performance data sequence through the historical service-related data. According to the service usage sequence and performance data sequence, build each user's standard service usage chain in multiple time periods, and set the QoS threshold in each time period. Determine the popularity of each service through historical service-related data of users, and divide the services into hot services and cold services according to the popularity. Mark the hot services and cold services in the standard service usage chain, determine the service level agreement of each service, and combine the service level agreement and the standard service usage chain to determine the initial network resources required for each service in different time periods; Allocate initial network resources for services, monitor actual service usage by users over a period of time, and determine the QoS of the current service. Determine whether to adjust initial network resources based on the QoS and QoS threshold of the current service. If the initial network resources are adjusted, the target resources required for the service are evaluated and resource scheduling is performed based on the target resources required.
2. A method for intelligently scheduling network resources based on cloud computing as claimed in claim 1, characterized in that: Based on the service usage sequence and performance data sequence, a standard service usage chain for each user in multiple time periods is constructed, including: The service usage sequence includes the service usage sequence and the service usage data sequence. The service usage sequence is the service usage record over time. The service usage data sequence is the service usage related parameters over time. The performance data sequence is the service performance related parameters over time. Perform time alignment processing on the service usage sequence, service usage data sequence and performance data sequence, and plan the service usage sequence of the user in the same period according to the service usage times of multiple service usage sequences in the same period; Determine the standard range of each type of data by using the characteristic values in multiple service usage data sequences and performance data sequences in the same period, and mark the standard range of each type of data under the corresponding service in the service usage sequence; The service usage sequences marked with standard ranges in different time periods are connected in chronological order, thereby constructing a standard service usage chain for each user in multiple time periods.
3. A method for intelligently scheduling network resources based on cloud computing as claimed in claim 2, characterized in that: The standard range of each type of data is determined by using the characteristic values of multiple service usage data series and performance data series in the same period, including: The eigenvalues include the most frequent minimum value, the most frequent maximum value, and the most frequent median value; The intersection of the values of multiple service usage data series and multiple performance data series is used as the intersection range, and the distances from the two ends of the intersection range to the minimum value with the highest frequency and the maximum value with the highest frequency are calculated respectively, which are recorded as the first distance and the second distance; Calculate the distance from the median with the highest frequency to the two ends of the intersection range respectively, and record them as the third distance and the fourth distance; The extension ratios of the two ends of the intersection range are respectively mapped by the third distance and the fourth distance, and the two ends of the intersection range are extended by the extension ratio, the first distance and the second distance to obtain the standard range of each data.
4. The method for intelligently scheduling network resources based on cloud computing according to claim 3, characterized in that: Set the QoS threshold for each time period, including: De-noising, cleaning and normalizing each type of data required for service quality in the service usage data series and performance data series; By service type, each type of data required for service quality is weighted and summed to obtain a comprehensive index, and the QoS threshold is set based on the comprehensive index of multiple services; Among them, QoS 阈 is the QoS threshold, n1 is the number of services currently used, n2 is the number of data types required for the i-th service quality, W i is the comprehensive indicator of the i-th service, QoS 总 The service quality is the sum of the comprehensive indicators of all services currently used, Indicates the service quality mapped by the sum of comprehensive indicators of all currently used services.
5. The method for intelligently scheduling network resources based on cloud computing according to claim 2, characterized in that: Determine the popularity of each service through the user's historical service-related data, including: De-noising, cleaning and normalizing each type of data required for hot and cold situations in service usage data series and performance data series; Among them, C j is the hotness and coldness of the jth service, m is the number of each type of data required for the hotness and coldness of the jth service, α a The hot and cold weights of the data required for the a-th hot and cold situation of the j-th service, E a The data size required for the a-th hot and cold situation of the j-th service, R1 j , R2 j 、R3 j are the number of users, number of user visits, and duration of user use of the jth service, respectively. f(R1 j , R2 j , R3 j ) is the user usage degree obtained by the number of users of the j-th service, the number of user visits, and the user usage time, and k is a preset constant.
6. The method for intelligently scheduling network resources based on cloud computing according to claim 2, characterized in that: Combine the service level agreement and standard service usage chain to determine the initial network resources required for each service at different time periods, including: Determine the total initial network resource interval for each service through a service level agreement, and select a portion of the total initial network resource interval as a first resource interval through a data standard range of the service in the standard service usage chain; Selecting a portion of the total initial network resource interval as the second resource interval by using the hot and cold types and hot and cold degrees of the services in the standard service usage chain; If the first resource interval and the second resource interval have an intersection, the intersection interval is used as the initial network resource range required for the service; Otherwise, for the hot service, determine the average value point between the first resource interval and the second resource interval, and determine the median point of the larger one of the first resource interval and the second resource interval, and use the average value point and the median point as two endpoints of the required initial network resource range; For cold services, the average value point between the first resource interval and the second resource interval is determined, and the median point of the smaller one of the first resource interval and the second resource interval is determined, and the average value point and the median point are used as the two endpoints of the required initial network resource range.
7. The method for intelligently scheduling network resources based on cloud computing according to claim 1, characterized in that: The QoS and QoS thresholds of the current service are used to determine whether to adjust the initial network resources, including: If the QoS of the current service is greater than the QoS threshold, no adjustment is made to the initial network resources; Otherwise, the initial network resources are adjusted.
8. The method for intelligently scheduling network resources based on cloud computing according to claim 1, characterized in that: If the initial network resources are adjusted, the target resources required for the service are evaluated and resource scheduling is performed based on the target resources required, including: The target resources required for the service are evaluated by the matching degree between the current service and the standard service usage chain, and the resources are scheduled, scaled and migrated according to the target resources required.
9. A network resource intelligent scheduling device based on cloud computing, characterized in that: include: A construction module is used to collect historical service-related data of each user, and to construct a service usage sequence and performance data sequence of each user through the historical service-related data, to construct a standard service usage chain for each user in multiple time periods according to the service usage sequence and performance data sequence, and to set a QoS threshold for each time period; The determination module is used to determine the hotness and coldness of each service through the user's historical service-related data, and divide the services into two types: hot services and cold services according to the hotness and coldness, mark the hot services and cold services in the standard service usage chain, determine the service level agreement of each service, and determine the initial network resources required for each service in different time periods by combining the service level agreement and the standard service usage chain; The monitoring module is used to allocate the initial network resources of the service, monitor the actual service usage of the user over a period of time, and determine the QoS of the current service. It determines whether to adjust the initial network resources based on the QoS and QoS threshold of the current service. The scheduling module is used to evaluate the target resources required for the service if the initial network resources are adjusted, and to schedule resources according to the target resources required.
