CN113365357A

CN113365357A - Image recognition model training method, carrier wave adjusting method, device and medium

Info

Publication number: CN113365357A
Application number: CN202110638972.3A
Authority: CN
Inventors: 盛莉莉; 任飞; 顾伟; 周奕昕; 谷俊江
Original assignee: China United Network Communications Group Co Ltd
Current assignee: China United Network Communications Group Co Ltd
Priority date: 2021-06-08
Filing date: 2021-06-08
Publication date: 2021-09-07
Anticipated expiration: 2041-06-08
Also published as: CN113365357B

Abstract

The application provides an image recognition model training method, a carrier wave adjusting method, a device and a medium. The method comprises the following steps: acquiring a network element of a carrier to be adjusted; according to a first target parameter trend graph of a network element, using a trained image recognition model to obtain a group to which the network element belongs and a network element group matched with the network element; acquiring a network element list of License resources to be adjusted according to the group to which the network element belongs and the network element group matched with the network element; the network element list includes: a network element sublist of License resources to be added and a network element sublist of License resources to be reduced; and adjusting the License resources of the carrier waves of the network elements in the network element list. The carrier wave can be automatically adjusted.

Description

Image recognition model training method, carrier wave adjusting method, device and medium

Technical Field

The present application relates to communications technologies, and in particular, to a method, an apparatus, and a medium for training an image recognition model.

Background

Due to the directional flow of people and the like, network traffic in certain areas can have peak periods and valley periods, namely, network tidal effects exist. For example, during the working hours of a working day, the network traffic of an office building can increase greatly, and a traffic peak occurs; accordingly, a traffic trough may occur in the office building network during the off-hours. The proliferation of traffic may result in insufficient network capacity in the area and network congestion.

The carrier expansion of the network element with a service peak period is a commonly used means for solving the network congestion problem at present. However, if each network element with a traffic peak is configured with static capacity-expansion License resources, the License resources are wasted. Therefore, at present, the above problem is solved by dynamically adjusting the carrier, that is, performing License resource rollback on the network element with traffic in the low valley period, and then loading the returned License resource to the network element with traffic in the peak period in the period.

The current dynamic carrier adjustment is performed manually, that is, an operator needs to match a pair of network elements capable of participating in License allocation manually, and then manually issues instructions such as expansion or capacity reduction on an OMC, and further manually performs License resource allocation operation on a License server of a main equipment manufacturer for many times. However, due to the large number of network elements, it is difficult for an operator to manually match all network elements; and for network elements with short network tidal effect periods, the operator is difficult to frequently perform License resource rollback or loading work. Therefore, the manual carrier adjustment method has the defects of low efficiency, incapability of accurately adjusting the carrier, and the like.

Disclosure of Invention

The application provides an image recognition model training method, a carrier wave adjusting method, a device and a medium, which are used for solving the problems of low efficiency and incapability of accurately adjusting carrier waves in a manual carrier wave adjusting mode.

In a first aspect, the present application provides a method for training an image recognition model, the method comprising:

acquiring a training sample set; wherein the training sample set comprises: the marked intermediate sample network element set and a first target parameter trend graph of each sample network element in the intermediate sample network element set; the labeled intermediate sample network element set comprises at least one group of single-carrier sample network elements and at least one group of double-carrier sample network elements, the single-carrier sample network elements belonging to the same group have the same first label and pairing identification, and the first label comprises: the frequency of occurrence of a peak in the first target parameter trend graph of the single carrier sample network element and the time of occurrence of the peak; the dual carrier sample network elements belonging to the same group have the same second label and pairing identification, and the second label includes: the number of troughs of two carriers appearing simultaneously in the first target parameter trend graph of the dual-carrier sample network element and the time of the troughs appearing simultaneously; the pairing identification is used for representing a network element group paired with the group of sample network elements, the paired sample network element group comprises a group of single-carrier sample network elements and a group of double-carrier sample network elements, the peak and the trough of the first target parameter trend graph of the paired single-carrier sample network elements and the double-carrier sample network elements are complementary, and the first target parameter value can represent the service condition of License resources of the sample network elements;

training an image recognition model by using the training sample set to obtain a trained image recognition model; the trained image recognition model is used for acquiring a group to which the network element belongs and a network element group matched with the network element according to a first target parameter trend graph of the network element.

Optionally, the obtaining a training sample set includes:

acquiring an intermediate sample network element set; the middle sample network element set comprises a single-carrier sample network element and a double-carrier sample network element;

selecting the first target parameter as a clustering index, and performing clustering operation on the intermediate sample network element set to obtain at least one group of single carrier sample network elements and at least one group of double carrier sample network elements;

drawing a first target parameter trend graph of each sample network element according to a first target parameter value set of the carrier of each sample network element in the middle sample network element set within a preset time length;

pairing at least one group of single carrier sample network elements and at least one group of double carrier sample network elements; the paired sample network element group comprises a group of single-carrier sample network elements and a group of double-carrier sample network elements, and the peaks and the troughs of the first target parameter trend graph of the single-carrier sample network elements and the double-carrier sample network elements are complementary;

labeling the sample network elements in the middle sample network element set according to the grouping of the sample network elements;

and taking the marked intermediate sample network element set and the first target parameter trend graph of each sample network element as the training sample set.

Optionally, the obtaining an intermediate sample primitive set includes:

acquiring an initial sample network element set and a first target parameter value set of a carrier of each sample network element within a preset time length; the initial sample set of network elements comprises: a plurality of single carrier sample network elements and a plurality of dual carrier sample network elements;

removing the sample network elements meeting a first preset condition from the initial sample network element set to obtain an intermediate sample network element set; the first preset condition includes: the maximum value of the first target parameter value set of the single-carrier sample network element in the preset time length is smaller than or equal to a preset threshold value, and the minimum value of the first target parameter value set of at least one carrier in the double-carrier sample network element in the preset time length is larger than the preset threshold value.

In a second aspect, the present application provides a carrier adjustment method, including:

acquiring a network element of a carrier to be adjusted;

according to a first target parameter trend graph of a network element, using a trained image recognition model to obtain a group to which the network element belongs and a network element group matched with the network element; the trained image recognition model is obtained by adopting the training method of any one of the first aspect; the first target parameter can represent the use condition of License resources of the network element;

acquiring a network element list of License resources to be adjusted according to the group to which the network element belongs and the network element group matched with the network element; the network element list includes: a network element sublist of License resources to be added and a network element sublist of License resources to be reduced;

and adjusting the License resources of the carrier waves of the network elements in the network element list.

Optionally, the adjusting License resources of carriers of the network elements in the network element list includes:

matching the types of the License resources of all network elements in the network element sub-list of the License resources to be increased with the types of the License resources of all network elements in the network element sub-list of the License resources to be decreased, and determining whether the amount to be decreased of the License resources of the same type in the network element sub-list of the License resources to be decreased is larger than or equal to the amount to be increased of the License resources of the same type in the network element sub-list of the License resources to be increased;

if the to-be-reduced amount is larger than or equal to the to-be-increased amount, writing the network element of the to-be-increased License resource corresponding to the type of the License resource, the to-be-increased amount of the License resource of the network element of the to-be-increased License resource, the network element of the to-be-reduced License resource, the to-be-reduced amount of the License resource of the network element of the to-be-reduced License resource and the type of the License resource into a License stock adjustment list;

and adjusting the License resources of the carrier waves of the network elements in the License stock adjustment list according to the License stock adjustment list.

Optionally, the method further comprises:

and if the to-be-reduced amount is smaller than the to-be-reduced amount, acquiring the License resource corresponding to the difference value from the existing network stock License list for the network element sub-list of the License resource to be added according to the difference value between the to-be-reduced amount and the to-be-reduced amount.

Optionally, the adjusting, according to the License stock adjustment list, License resources of carriers of network elements in the License stock adjustment list includes:

acquiring a License adjustment file executable by the network element management platform according to the License stock adjustment list;

and sending the License adjustment file to the network element management platform so that the network element management platform adjusts the License resources of the carrier waves of the network elements in the License stock adjustment list.

Optionally, the obtaining, according to the License stock adjustment list, a License adjustment file executable by the network element management platform includes:

acquiring a target License stock adjustment template based on a preset License stock adjustment template and the License stock adjustment list;

sending the target License stock adjusting template to a License platform;

and receiving a License adjusting file returned by the License platform based on the target License stock adjusting template.

Optionally, before the acquiring the target License inventory adjustment template, the method further includes:

and acquiring the preset License stock adjusting template from the License platform.

Optionally, before obtaining the network element list of the License resource to be adjusted according to the group to which the network element belongs and the network element group paired with the network element, the method further includes:

determining whether to perform carrier adjustment on the network element according to the second target parameter value of the network element; the second target parameter can represent the use condition of License resources of the network element, and the first target parameter is different from the second target parameter.

In a third aspect, the present application provides an image recognition model training apparatus, including:

the acquisition module is used for acquiring a training sample set; wherein the training sample set comprises: the marked intermediate sample network element set and a first target parameter trend graph of each sample network element in the intermediate sample network element set; the labeled intermediate sample network element set comprises at least one group of single-carrier sample network elements and at least one group of double-carrier sample network elements, the single-carrier sample network elements belonging to the same group have the same first label and pairing identification, and the first label comprises: the frequency of occurrence of a peak in the first target parameter trend graph of the single carrier sample network element and the time of occurrence of the peak; the dual carrier sample network elements belonging to the same group have the same second label and pairing identification, and the second label includes: the number of troughs of two carriers appearing simultaneously in the first target parameter trend graph of the dual-carrier sample network element and the time of the troughs appearing simultaneously; the pairing identification is used for representing a network element group paired with the group of sample network elements, the paired sample network element group comprises a group of single-carrier sample network elements and a group of double-carrier sample network elements, the peak and the trough of the first target parameter trend graph of the paired single-carrier sample network elements and the double-carrier sample network elements are complementary, and the first target parameter value can represent the service condition of License resources of the sample network elements;

the training module is used for training an image recognition model by using the training sample set to obtain a trained image recognition model; the trained image recognition model is used for acquiring a group to which the network element belongs and a network element group matched with the network element according to a first target parameter trend graph of the network element.

Optionally, the obtaining module is specifically configured to:

performing clustering operation on the sample network elements in the middle sample network element set according to a first target parameter value set of the carrier of each sample network element in the middle sample network element set within a preset time length to obtain at least one group of single-carrier sample network elements and at least one group of double-carrier sample network elements; the distance between the first target parameters of the sample network elements belonging to the same group at the same moment is smaller than or equal to a preset distance;

drawing a first target parameter trend graph of each group of sample network elements according to a first target parameter value set of the carrier of each sample network element in the middle sample network element set within a preset time length;

pairing the at least one group of single carrier sample network elements and the at least one group of double carrier sample network elements; the paired sample network element group comprises a group of single-carrier sample network elements and a group of double-carrier sample network elements, and the peaks and the troughs of the first target parameter trend graph of the single-carrier sample network elements and the double-carrier sample network elements are complementary;

Optionally, the obtaining module is specifically configured to:

In a fourth aspect, the present application provides a carrier adjustment apparatus, including:

the first acquisition module is used for acquiring a network element of a carrier to be adjusted;

the second acquisition module is used for acquiring the group to which the network element belongs and the network element group matched with the network element by using the trained image recognition model according to the first target parameter trend graph of the network element; the trained image recognition model is obtained by adopting the training method of any one of the first aspect; the first target parameter can represent the use condition of License resources of the network element;

a third obtaining module, configured to obtain a network element list of License resources to be adjusted according to the group to which the network element belongs and a network element group paired with the network element; the network element list includes: a network element sublist of License resources to be added and a network element sublist of License resources to be reduced;

and the adjusting module is used for adjusting the License resources of the carrier waves of the network elements in the network element list.

Optionally, the adjusting module is specifically configured to:

Optionally, the adjusting module is specifically further configured to:

Optionally, the adjusting module is specifically configured to:

sending the target License stock adjusting template to a License platform;

Optionally, the apparatus further comprises: a fourth obtaining module;

the fourth obtaining module is specifically configured to: and acquiring the preset License stock adjusting template from the License platform before the adjusting module acquires a target License stock adjusting template based on a preset License stock adjusting template and the License stock adjusting list.

Optionally, the third obtaining module is further configured to: before a network element list of License resources to be adjusted is acquired according to the group to which the network element belongs and the network element group matched with the network element, whether carrier adjustment is carried out on the network element is determined according to a second target parameter value of the network element; the second target parameter can represent the use condition of License resources of the network element, and the first target parameter is different from the second target parameter.

In a fifth aspect, the present application provides an electronic device, comprising: at least one processor and memory;

the memory stores computer-executable instructions;

the at least one processor executing the computer-executable instructions stored by the memory causes the electronic device to perform the method of any of the first aspects or the method of any of the second aspects.

In a sixth aspect, the present application provides a computer-readable storage medium having stored thereon computer-executable instructions for performing the method of any one of the first aspect or the method of any one of the second aspect when executed by a processor.

In a seventh aspect, the present application provides a computer program product comprising a computer program that, when executed by a processor, performs the method of any one of the first aspect or the method of any one of the second aspect.

According to the image recognition model training method, the carrier adjustment method, the device and the medium, after the network elements needing carrier adjustment are obtained, the group to which each network element belongs and the network element group matched with the network element can be obtained based on the first target parameter trend graphs of the network elements, so that the information can be utilized to obtain the network element list comprising the network element sub-list of the License resources to be increased and the network element sub-list of the License resources to be reduced, and the License resources of the carriers of the network elements can be dynamically adjusted by using the list. The method can automatically match a pair of network elements which can participate in License resource allocation, and automatically adjust without manual participation and processing, thereby improving the efficiency and accuracy of carrier adjustment.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and together with the description, serve to explain the principles of the application.

Fig. 1 is a schematic diagram of a communication system architecture to which a carrier adjustment method according to an embodiment of the present application is applied;

fig. 2 is a flowchart illustrating a carrier adjustment method according to an embodiment of the present application;

fig. 3 is a PRB utilization trend graph of a set of dual-carrier network elements in 24 hours according to an embodiment of the present application;

fig. 4 is a PRB utilization trend graph in 24 hours for a set of single carrier network elements provided in an embodiment of the present application;

FIG. 5 is a flowchart illustrating an image recognition model training method according to an embodiment of the present disclosure;

fig. 6 is a flowchart illustrating a specific example of a carrier adjustment method according to an embodiment of the present application;

FIG. 7 is a flowchart illustrating an example of a training method for an image recognition model according to an embodiment of the present disclosure;

FIG. 8 is a schematic structural diagram of an image recognition model training apparatus according to an embodiment of the present disclosure;

fig. 9 is a schematic structural diagram of a carrier adjustment apparatus according to an embodiment of the present application;

fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

With the above figures, there are shown specific embodiments of the present application, which will be described in more detail below. These drawings and written description are not intended to limit the scope of the inventive concepts in any manner, but rather to illustrate the inventive concepts to those skilled in the art by reference to specific embodiments.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

The terms referred to in this application are explained first:

the network element related to the present application may be a Remote Radio Unit (RRU), and the network element has at least one carrier. Wherein one carrier covers one cell. When the network element covers a cell, the network element can be called a single carrier network element; when the network element covers two cells, it can be called a dual carrier network element.

Wherein, each network element can be configured with License resource for realizing the communication of the cell. It should be understood that the more License resources of a network element, the larger the network capacity of the network element, the more traffic can be carried. Compared with a single-carrier network element, the dual-carrier network element is configured with more License resources.

When the capacity expansion of the single carrier network element is needed, the capacity expansion can be changed into a double carrier network element by increasing a certain number of License resources. Similarly, when the capacity reduction of the dual-carrier network element is needed, the capacity reduction can be performed to the single-carrier network element by reducing a certain number of License resources.

At present, when the traffic of a cell covered by a network element changes, the License resource configuration of the network element may be adjusted, so that the network element switches between a single-carrier network element and a dual-carrier network element, thereby adapting the License resource configuration to the change of the traffic. For example, when the traffic of the cell covered by the single carrier network element is at the peak, the License resource of the network element may be increased to switch to the dual carrier network element. This act of increasing License resources of a network element may also be referred to as License resource loading. On the contrary, when the traffic of the cell covered by the dual-carrier network element is at the valley value, the License resource of the network element can be reduced to switch the network element into the single-carrier network element. This act of reducing License resources of a network element may also be referred to as License resource fallback.

The License resource adjustment method for switching the network element to be adjusted between the single carrier network element and the dual carrier network element by backspacing or loading the License resource of the network element to be adjusted is also called dynamic adjustment of the carrier. The network element of the carrier to be adjusted comprises a single carrier network element to be adjusted and a double carrier network element to be adjusted, namely, the single carrier network element needs to increase License resources to realize the purpose of capacity expansion under the condition that License resources are in short supply during the peak time period of network traffic, and the double carrier network element can reduce License resources to realize the purpose of capacity reduction under the condition that License resources are excessive during the valley time period of network traffic.

At present, dynamic carrier adjustment is performed manually, that is, an operator needs to match a pair of network elements capable of participating in License allocation manually, then manually issue instructions such as expansion or capacity reduction on an Operation and Maintenance Center (OMC), and further manually perform License resource allocation Operation on a License server of a main equipment manufacturer for multiple times. However, due to the large number of network elements, it is difficult for an operator to manually match all network elements; and for network elements with short network tidal effect periods (namely, the time length of the peak value and/or the valley value of the traffic is short), the operator is difficult to frequently perform License resource rollback or loading work. Therefore, the manual carrier adjustment method has the defects of low efficiency, incapability of accurately adjusting the carrier, and the like.

In view of the above problems, the present application provides a carrier adjustment method, which can automatically adjust License resources of a carrier of a network element.

The carrier adjustment method provided by the present application can be applied to the schematic diagram of the communication system architecture shown in fig. 1. As shown in fig. 1, the communication system includes: the system comprises a plurality of network elements, a network element management platform, electronic equipment, a License platform and a License resource pool. The network element management platform is used for providing data information of a plurality of network elements, and the data information comprises types, numbers and the like of License resources configured on the plurality of network elements. The License platform is used for providing License resources in the License resource pool for the network elements and controlling the binding or unbinding process of the network elements and the License resources. The electronic device can acquire data information of a plurality of network elements from the network element management platform, and automatically select a plurality of pairs of network elements which can participate in carrier adjustment through a data processing process. The electronic equipment can also interact with a License platform to acquire the authority of loading License resources unbound with the original binding network element onto other network elements.

The electronic device may be, for example, any terminal or server having processing capabilities. The terminal referred to herein may also be referred to as a terminal device, a User Equipment (UE), a Mobile Station (MS), a Mobile Terminal (MT), and the like, and may be, for example, a mobile phone (mobile phone), a tablet computer (pad), or a computer with a wireless transceiving function.

The Network element Management platform may be any platform capable of managing Network elements, such as an OMC, an Network Element Management System (EMS), or a Network Management System (NMS). Fig. 1 is a schematic diagram of an example of an element management platform as an OMC.

The following describes the technical solutions of the present application and how to solve the above technical problems with specific embodiments. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.

Fig. 2 is a flowchart illustrating a carrier adjustment method according to an embodiment of the present application. As shown in fig. 2, the method of the present application may include:

s101, obtaining a network element of a carrier to be adjusted.

The network element of the carrier to be adjusted referred to herein may be, for example, a network element belonging to a region, including a single-carrier network element that needs to increase License resources and a dual-carrier network element that needs to decrease License resources.

The method for acquiring the network element of the carrier to be adjusted is not limited in the present application, and for example, the network element may be obtained by screening the network element in the current network based on the first preset condition. For example, the first preset condition includes: the maximum value of the first target parameter value set of the single carrier network element in the preset time length is smaller than or equal to the preset threshold value, and the minimum value of the first target parameter value set of at least one carrier in the double carrier network element in the preset time length is larger than the preset threshold value.

The first target parameter refers to a parameter capable of characterizing the use condition of License resources of the network element, for example: physical Resource Block (PRB) utilization, user rate, and the like. Taking the PRB utilization rate as the first target parameter as an example, the higher the PRB utilization rate of a network element in a time period is, the more License resources the network element uses in the time period is.

Taking the PRB utilization rate as a first target parameter, the preset duration is 24 hours, and the preset threshold is 60%, for a single carrier network element, if the maximum value of the PRB utilization rate in 24 hours is less than or equal to 60%, it may be considered that the network element does not need to be expanded, and does not participate in the subsequent steps; for a dual-carrier network element, if the lowest value of either the first carrier PRB utilization rate or the second carrier PRB utilization rate within 24 hours is greater than 60%, it may be considered that the dual-carrier network element does not need to perform volume reduction and does not participate in subsequent steps.

For example, the network element of the carrier to be adjusted may be obtained by the electronic device by screening using a first preset condition based on the data of the network element of the current network, which is acquired from the network element management platform. Or, the network element of the carrier to be adjusted may be input to the electronic device after being screened by the operator based on the first preset condition, or may be sent to the electronic device by another device, and the like.

S102, according to a first target parameter trend graph of the network element, a group to which the network element belongs and a network element group matched with the network element are obtained by using a trained image recognition model.

The first target parameter trend graph of the network element refers to a numerical variation graph of a first target parameter (such as PRB utilization) of the network element within a preset time duration (such as 24 hours). Taking a first target parameter as an example of PRB utilization, fig. 3 is a graph of a trend of PRB utilization of a set of dual-carrier network elements in 24 hours, which is provided in this embodiment of the present application; fig. 4 is a PRB utilization trend graph in 24 hours for a set of single carrier network elements provided in an embodiment of the present application. As shown in fig. 3 and 4, the first target parameter trend graph may be, for example: a graph of PRB utilization trend over 24 hours for a set of dual-carrier network elements as shown in fig. 3, or a graph of PRB utilization trend over 24 hours for a set of single-carrier network elements as shown in fig. 4. Wherein, the first 00 in fig. 3: 00-23: the curve in the range of 00 points represents the curve of the value change of the PRB utilization rate of the first carrier of the dual-carrier tuple in 24 hours, for example, the second 00 in fig. 3: 00-23: the curve in the 00-point range represents the curve of the PRB utilization value of the second carrier of the dual-carrier tuple in the 24 hours. It should be understood that these two 00: 00-23: point 00 may be 00 on the same day: 00-23: 00, or 00: 00-23: point 00.

Taking fig. 3 as an example, the ratio of the total weight of the particles is 13: 00-17: in the time period of 00, if the first carrier curve and the second carrier curve of the dual-carrier network element group are concave, it can be considered that a trough appears in the time period; taking fig. 4 as an example, the ratio of (21: 00) - (02): 00, if the curve of the single-carrier network element group appears convex upwards, it can be considered that a peak appears in the network element in the time period.

The trained image recognition model can classify the network element based on the first target parameter trend graph of the network element to obtain the classification group to which the network element belongs. In addition, the model can also obtain a network element group paired with the network element. The group to which the network element belongs and the network element group paired with the network element, which are obtained by the image recognition model, may be only the group number to which the network element belongs and the network element group number paired with the network element. For example, the content acquired by the image recognition model may be: the network element belongs to a single carrier network element A group, and the network element can be paired with the network elements belonging to a double carrier network element B group. Alternatively, the content acquired by the image recognition model may be: the network element belongs to a double-carrier network element C group, and the network element can be paired with the network elements belonging to a single-carrier network element D group.

If the trough time period in the first target parameter trend graph of the dual-carrier network element group can cover the peak time period in the first target parameter trend graph of the single-carrier network element group, the two network element groups are said to be paired, that is, in the peak time period in the first target parameter trend graph of the single-carrier network element group, the network element in the dual-carrier network element group can provide surplus License resources for the network element in the single-carrier network element group for capacity expansion. For example: as shown in fig. 3 and 4, the time period of 21: 00-02: 00 points of the single-carrier network element group in fig. 4 is a peak time period, and the time period of the double-carrier network element group in fig. 3 is a trough time period, the two network element groups can be paired.

The image recognition model may be obtained by training using a neural network, or may be obtained by training using another network model that can realize classification, for example. For how to train the image recognition model, refer to the detailed description of the following embodiments.

S103, acquiring a network element list of License resources to be adjusted according to the group to which the network element belongs and the network element group matched with the network element; the network element list includes: and the network element sub-list of the License resources to be increased and the network element sub-list of the License resources to be reduced.

The network element to be added with License resources is a single carrier network element, and the network element to be reduced with License resources is a double carrier network element. The network element sub-list to which the License resource is to be added and the network element sub-list to which the License resource is to be reduced may include License resource information of each network element, for example. The License resource information of each network element comprises the type and the number of the License resources of the network element, the identity of the network element, the pairing information of the network element and the like. The identity of the network element may be, for example, any identifier capable of uniquely identifying the network element, such as an Electronic Serial Number (ESN) code of the network element, an ID of the network element in the current network, or a cell ID corresponding to the network element. The pairing information of a network element as referred to herein is a pairing identification of a network element group paired with the network element.

The information of the network element in each sub-list may be obtained from the network element management platform after the network element list of the sub-list is obtained, or may be screened from the information of the network element of the current network, which is obtained from the network element management platform before.

Taking the example that after the network element list of the sub-list is obtained, the information of the network elements in the sub-lists is obtained from the OMC:

the electronic equipment can input the identity of the single carrier network element to be adjusted and the identity of the double carrier network element to be adjusted into the command generating component, generate a License resource present network configuration command for inquiring the double carrier network element to be adjusted and a License resource configuration consistency command for checking the single carrier network element to be adjusted, then use the command issuing component to issue the commands to the OMC, and download a result file generated after the OMC executes the commands. The electronic device can analyze the result file, and obtain the network element sub-list of the License resources to be increased and the network element sub-list of the License resources to be decreased through data arrangement.

The License resource present network configuration instruction for inquiring the dual-carrier network element to be adjusted is used for inquiring and acquiring the type and the quantity of License resources corresponding to the dual-carrier network element to be adjusted, and information such as the identity of the network element. And the License resource configuration consistency command for checking the single carrier network element to be adjusted is used for inquiring and acquiring the type and the quantity of License resources corresponding to the single carrier network element to be adjusted, the identity identification of the network element and other information.

Optionally, before performing step S103, it may further be determined whether to perform carrier adjustment on the network element according to the second target parameter value of each network element. The second target parameter is a parameter that can characterize the use condition of License resources of the network element besides the first target parameter. Taking the first target parameter as the utilization rate of the PRB as an example, the second target parameter may be, for example, a user rate or a Quality of Service Identifier (QCI) value. Alternatively, for example, if the first target parameter is the user rate, the second target parameter may be, for example, PRB utilization rate or QCI value. Alternatively, for example, if the first target parameter is a QCI value, the second target parameter may be, for example, PRB utilization rate or user rate.

By using the second target parameter to further screen the network element, the accuracy of the determined network element to be adjusted can be further improved.

S104, adjusting the License resources of the carrier waves of the network elements in the network element list.

As a possible implementation manner, the electronic device matches the types of License resources of each network element in the network element sub-list of the License resources to be increased with the types of License resources of each network element in the network element sub-list of the License resources to be decreased, and determines whether the amount to be decreased of the License resources of the same type in the network element sub-list of the License resources to be decreased is greater than or equal to the amount to be increased of the License resources of the same type in the network element sub-list of the License resources to be increased.

If the to-be-reduced amount is larger than or equal to the to-be-increased amount, it is indicated that the to-be-reduced amount of the License resource of the type in the network element of the current to-be-adjusted carrier can meet the amount required to be increased. The electronic device may write the network element to which the License resource is to be added, the License resource to be incremented of the network element to which the License resource is to be added, the network element to which the License resource is to be decreased, and the License resource to be decremented of the network element to which the License resource is to be decreased, which correspond to the type of the License resource, into the License stock adjustment list.

If the to-be-reduced amount is smaller than the to-be-increased amount, it is indicated that the to-be-reduced amount of the License resource of the type in the network element of the current to-be-adjusted carrier cannot meet the amount required to be increased. The electronic device may write the network element of the License resource to be added, the network element of the License resource to be decreased, and the network element of the License resource to be decreased, which correspond to the type of the License resource, into the License stock adjustment list. For example, if the License resource to be decremented corresponding to the type of the License resource is 1 in number and the License resource to be incremented is 2 in number, the electronic device may write the License resource to be incremented corresponding to the License resource to be incremented being 1 in number, the License resource to be incremented being 1 in the network element to which the License resource is to be added, the License resource to be decremented being 1 in the network element to which the License resource is to be decreased, and the type of the License resource into the License stock adjustment list.

In this implementation, for the part of the network element to be incremented corresponding to the difference between the number 2 and the number 1 and the part of the network element to be incremented corresponding to the License resource to be added, the License resource may not be adjusted this time.

Or, the electronic device may obtain, for the network element sub-list to which the License resource is to be added, the License resource corresponding to the difference from the existing network stock License list according to the difference between the to-be-incremented amount and the to-be-decremented amount. The network element data in the existing network stock License list can be downloaded from the network element management platform. Or, the network element data in the existing network stock License list may also be input by a user, or the network element data in the existing network stock License list may also be obtained by other devices from a platform for managing network elements and then sent to the electronic device.

Optionally, after the network element sub-list for adding the License resource acquires the License resource corresponding to the difference value from the existing network stock License list, the License resource corresponding to the difference value may be written into the License stock adjustment list together.

Finally, the electronic device can adjust the License resources of the carrier waves of the network elements in the License stock adjustment list according to the License stock adjustment list obtained in the above manner.

For example, taking the network element management platform to execute License resource adjustment as an example, the electronic device may obtain a License adjustment file executable by the network element management platform according to the License stock adjustment list. And then, the electronic equipment sends a License adjustment file to the network element management platform so that the network element management platform adjusts the License resources of the carrier waves of the network elements in the License stock adjustment list. Optionally, if the electronic device integrates a function of the network element management platform, the electronic device may adjust License resources of carriers of network elements in a License stock adjustment list based on the License adjustment file.

For how to adjust the License resources of the carrier waves of the network elements in the License stock adjustment list based on the License adjustment file, reference may be made to a manner in which the network element management platform adjusts the License resources of the carrier waves of the network elements in the License stock adjustment list in the prior art, which is not described herein again.

The electronic device obtains the License adjustment file executable by the network element management platform according to the License stock adjustment list by adopting the following method:

for example, the electronic device obtains a target License stock adjustment template according to the License stock adjustment list, a preset License stock adjustment template and the License stock adjustment list; then, the electronic device may send the target License stock quantity adjusting template to a License platform, and receive a License adjusting file returned by the License platform based on the target License stock quantity adjusting template.

Or if the electronic device integrates the functions of the License platform, the electronic device may obtain the target License stock adjustment template according to the License stock adjustment list, the preset License stock adjustment template and the License stock adjustment list; then, the electronic device may generate a License adjustment file executable by the network element management platform based on the target License inventory adjustment template.

Exemplarily, taking the implementation of the step by the interaction of the electronic device with the OMC and License platform as an example:

first, the electronic device may combine the generated License stock adjustment list with the ESN code of the device electronic serial number to generate a License stock preparation template.

Second, the electronic device may interact with the License platform for the first time. That is, the electronic device may automatically open a Web page of the License platform, automatically input information such as a user name and a password to log in the platform, and upload the License stock preparation template to the License platform. And the License platform provides a preset License stock adjusting template corresponding to the License resource needing to be adjusted based on the information in the License stock preparing template. And the electronic equipment downloads the preset License stock adjusting template.

And thirdly, the electronic equipment can fill the to-be-increased amount or the to-be-decreased amount of the type of the corresponding License resource in the downloaded preset License stock adjustment template according to the License failure code table, the License stock adjustment list and other data, so as to generate the target License stock adjustment template.

Finally, the electronic device may interact with the License platform a second time. Namely, the electronic device can automatically open a Web page of the License platform, automatically input information such as a user name and a password to log in the platform, and upload a target License stock adjustment template to the License platform. After the steps are completed, the License platform can allocate License resources according to the target License stock adjustment template and generate a License adjustment file containing License resource adjustment information. The electronic device may download and parse the License adjustment file.

The electronic equipment can generate an instruction for executing the License adjusting file through the instruction generating component, and the instruction is issued to the OMC through the instruction issuing component. After the OMC loads the License adjustment file, the corresponding network element can complete the action of taking the License resource into effect or losing the License resource, and finally the License resource of the carrier wave of the network element in the License stock adjustment list is automatically adjusted.

Illustratively, the License failure code table may be obtained by way of interaction between the electronic device and the network element management platform. Taking the network element management platform as the OMC as an example, the electronic device may input the identities of the single carrier network element to be adjusted and the dual carrier network element to be adjusted into the instruction generating component, generate an instruction for querying the network element License failure code, issue the instruction to the OMC by using the instruction issuing component, and download a result file generated after the OMC executes the instruction. The electronic equipment can analyze the result file and obtain the License failure code table through data arrangement.

According to the carrier adjustment method provided by the application, after the network elements needing carrier adjustment are obtained, the group to which each network element belongs and the network element group matched with the network element can be obtained based on the first target parameter trend graphs of the network elements, so that the network element list comprising the network element sub-list of the License resources to be increased and the network element sub-list of the License resources to be reduced can be obtained by using the information, and the License resources of the carriers of the network elements can be dynamically adjusted by using the list. The method can automatically match a pair of network elements which can participate in License allocation, and automatically adjust without manual participation and processing, thereby improving the efficiency and accuracy of carrier adjustment.

Although the above embodiments have been described and illustrated with the electronic device interacting with the network element management platform and the License platform. However, it can be understood by those skilled in the art that the electronic device may also integrate the functions of the network element management platform and/or the License platform, and when the electronic device integrates the functions of the network element management platform and/or the License platform, the aforementioned actions of the electronic device interacting with the electronic device may be completed and implemented by the electronic device itself, which is not described herein again.

The following describes and explains a training method of the image recognition model. It should be understood that an execution subject of the model training method provided in this embodiment may be the electronic device that executes the carrier adjustment method, or may be another device different from the electronic device, for example, a server, and the like. The following embodiments are illustrated by taking a server as an example.

Fig. 5 is a schematic flowchart of an image recognition model training method according to an embodiment of the present application.

As shown in fig. 5, the method of the present application may include:

s201, a training sample set is obtained.

Wherein the training sample set comprises: the marked intermediate sample network element set and a first target parameter trend graph of each sample network element in the intermediate sample network element set; the labeled intermediate sample network element set comprises at least one group of single-carrier sample network elements and at least one group of double-carrier sample network elements, the single-carrier sample network elements belonging to the same group have the same first label and pairing identification, and the first label comprises: the frequency of occurrence of a peak in the first target parameter trend graph of the single carrier sample network element and the time of occurrence of the peak; the dual carrier sample network elements belonging to the same group have the same second label and pairing identification, and the second label includes: the number of troughs of two carriers appearing simultaneously in the first target parameter trend graph of the dual-carrier sample network element and the time of the troughs appearing simultaneously; the pairing identification is used for representing a network element group paired with the group of sample network elements, the paired sample network element group comprises a group of single-carrier sample network elements and a group of double-carrier sample network elements, peaks and troughs of the first target parameter trend graph of the paired single-carrier sample network elements and the double-carrier sample network elements are complementary, and the first target parameter value can represent the service condition of License resources of the sample network elements. Exemplarily, PRB utilization may be chosen as the first target parameter.

And the intermediate sample network element set is obtained by removing the sample network elements meeting the first preset condition from the initial sample network element set.

The initial sample network element set includes network elements of carriers to be adjusted. The initial sample network element set may be downloaded from the OMC, or may be downloaded from other network element management platforms, for example: EMS or NMS, etc. any platform capable of managing network elements. Or, the network element data in the existing network stock License list may also be input by a user, or the network element data in the existing network stock License list may also be obtained by other devices from a platform for managing network elements and then sent to the electronic device.

The first preset condition includes: the maximum value of the first target parameter value set of the single-carrier sample network element in the preset time length is smaller than or equal to a preset threshold value, and the minimum value of the first target parameter value set of at least one carrier in the double-carrier sample network element in the preset time length is larger than the preset threshold value. Taking the PRB utilization rate as a first target parameter, the preset duration is 24 hours, and the preset threshold is 60%, for a single carrier network element, if the maximum value of the PRB utilization rate in 24 hours is less than or equal to 60%, it may be considered that the network element has no need of capacity expansion, and does not participate in the subsequent steps; for a dual-carrier network element, if the lowest value of the PRB utilization rate of any one of the first carrier or the second carrier within 24 hours is greater than 60%, it may be considered that the dual-carrier network element has no necessity of volume reduction, and does not participate in subsequent steps.

The single-carrier sample network element group and the double-carrier sample network element group are obtained through clustering operation, and the specific clustering mode is as follows:

performing clustering operation on the sample network elements in the middle sample network element set according to a first target parameter value set of the carrier of each sample network element in the middle sample network element set within a preset time length to obtain at least one group of single-carrier sample network elements and at least one group of double-carrier sample network elements; the distance between the first target parameters of the sample network elements belonging to the same group at the same time is less than or equal to a preset distance. The preset distance can be determined according to the clustering requirement.

The training sample set is obtained by matching the single-carrier sample network element group and the double-carrier sample network element group obtained by the clustering operation, and the specific matching mode is as follows:

drawing a first target parameter trend graph of each group of sample network elements according to a first target parameter value set of a carrier of each sample network element in a single-carrier sample network element group and a double-carrier sample network element group obtained by clustering operation within a preset time length;

labeling the labels of the sample network elements according to the trend of the first target parameter trend graph of the sample network elements;

illustratively, the first label of the single-carrier sample network element group and the second label of the dual-carrier sample network element group may be obtained by: and manually labeling by an operator after observing the first target parameter trend graph of one sample network element group.

The pairing identifier refers to a mark capable of representing the pairing relationship between two network elements. Illustratively, the pairing identification may be obtained in the following manner: the single carrier sample network element group and the double carrier sample network element group which can be paired are marked with the same group number, namely, the single carrier sample network element group is respectively named as a single carrier sample network element 1 group, a single carrier sample network element 2 group, … and a single carrier sample network element M group, the double carrier sample network element group is respectively named as a double carrier sample network element 1 group, a double carrier sample network element 2 group, … and a double carrier sample network element N group, and two groups of samples (such as the single carrier sample network element) with the same group number are marked with the same group numberThe local network element group 1 and the dual carrier sample network element group 1) have a pairing relationship. When M is less than or equal to N, the double-carrier sample network element group is considered to meet the capacity expansion requirement of the single-carrier sample network element group, and the subsequent steps can be continued; when M > N, then the following two approaches can be taken: removing redundant single carrier sample network element group (M-N group) data without participating in subsequent steps, or expanding the number of double carrier sample network element groups to search N₁A new dual carrier sample network element capable of pairing with redundant single carrier sample network elements (M-N group) until M is less than or equal to N + N₁。

S202, training an image recognition model by using the training sample set to obtain a trained image recognition model; the trained image recognition model is used for acquiring a group to which the network element belongs and a network element group matched with the network element according to a first target parameter trend graph of the network element.

The image recognition model is trained using a network model that enables classification, such as a convolutional neural network. The trained image recognition model can classify the network element based on the first target parameter trend graph of the network element to obtain a classification group to which the network element belongs, and the classified network element can directly obtain information of a network element group matched with the network element according to a matching identifier of the group to which the network element belongs.

For how to train the model by using the training sample set, reference may be made to a model training method in the prior art, which is not limited in this application.

According to the training method of the image recognition model, the image recognition model can be automatically classified based on the first target trend graph of the network element through the training mode, and information of the network element group matched with the network element can be accurately acquired according to the matching identification. The method can automatically match a pair of network elements which can participate in License allocation without manual participation and processing, and improves the efficiency and accuracy of network element classification and matching.

The carrier adjustment method and the image recognition model training method are described below by way of specific examples.

Fig. 6 is a flowchart illustrating a specific example of a carrier adjustment method according to an embodiment of the present application. The specific example is described by taking an example of the carrier adjustment achieved by interaction between the electronic device and the network element management platform and the License platform, and selecting the OMC as the network element management platform, the PRB utilization rate as the first target parameter, the preset time duration of 24 hours, and the preset threshold of 60%. As shown in fig. 6, a specific example of the carrier adjustment method provided in the embodiment of the present application may include:

s301, the electronic equipment downloads PRB utilization rate data of single carrier network elements and PRB utilization rate data of double carrier network elements in the past 24 hours in the existing network from the OMC.

S302, the electronic equipment removes network element data of which the maximum value of PRB utilization rate within 24 hours in the single carrier network element is less than or equal to 60% and network element data of which the minimum value of either PRB utilization rate of the first carrier or PRB utilization rate of the second carrier within 24 hours in the double carrier network element is greater than 60%, and the network element of the carrier to be adjusted is obtained.

And S303, the electronic equipment draws a PRB utilization rate trend graph of the network element of each carrier to be adjusted within the 24 hours, and obtains the group to which the network element of each carrier to be adjusted belongs and the network element group matched with the network element by using a trained image recognition model according to the PRB utilization rate trend graph.

S304, the electronic equipment automatically generates an instruction and sends the instruction to the OMC, and the information such as the types and the number of License resources of the network elements of the carriers to be adjusted, the ESN codes of the equipment electronic serial numbers of the network elements, the License failure codes of the network elements and the like is obtained by inquiring and downloading the information from the OMC.

S305, the electronic equipment generates a network element sub-list of License resources to be increased and a network element sub-list of License resources to be reduced according to the group to which the network element of each carrier to be adjusted belongs, the network element group paired with the network element, the type and the number of the License resources of the network element and other information.

S306, the electronic equipment compares the License resource to-be-increased and to-be-decreased of the same License resource type in the two sub-lists, and removes the network element data corresponding to the License resource type of which the to-be-decreased amount is smaller than the to-be-increased amount to obtain a License stock adjustment list.

And S307, merging the generated License stock adjustment list and the ESN code of the equipment by the electronic equipment to generate a License stock preparation template, uploading the License stock preparation template to a License platform, and downloading the preset License stock adjustment template.

And S308, the electronic equipment fills the increment or decrement of each License resource type and the License failure code into a preset License stock adjusting template to obtain a target License stock adjusting template, and uploads the target License stock adjusting template to the License platform.

And S309, the License platform redistributes License resources according to the target License stock adjustment template and generates a License adjustment file containing License resource adjustment information.

And S310, the License platform sends the License adjusting file to the electronic equipment.

S311, the electronic equipment analyzes the License adjustment file and uploads the analyzed result to the OMC.

And S312, the OMC executes the License adjusting file to complete one-time carrier adjustment.

After obtaining the network element that needs to be subjected to carrier adjustment from the OMC, the electronic device provided in this embodiment may obtain, based on the first target parameter trend chart of the network element, a group to which the network element belongs and a network element group paired with the network element. The electronic equipment can generate a License stock adjustment list by using the information, further interact with a License platform to obtain a License adjustment file based on the list, and upload an analysis result of the License adjustment file to the OMC to complete one-time carrier adjustment.

The carrier adjustment method provided by the embodiment can automatically match a pair of network elements which can participate in License deployment, and automatically adjust without manual participation and processing, thereby improving the efficiency and accuracy of carrier adjustment.

Fig. 7 is a flowchart illustrating a specific example of an image recognition model training method according to an embodiment of the present application. The specific example is described by taking an example of an electronic device interacting with a network element management platform to implement image recognition model training, and selecting an OMC as the network element management platform, a PRB utilization rate as a first target parameter, a preset duration of 24 hours, and a preset threshold of 60%. This specific example is illustrated by using a convolutional neural network for image recognition model training. As shown in fig. 7, a specific example of the image recognition model training method provided in the embodiment of the present application may include:

s401, the electronic equipment downloads PRB utilization rate data of single carrier network elements and PRB utilization rate data of double carrier network elements in the past 24 hours in the existing network from the OMC to obtain an initial sample network element set.

S402, the electronic equipment removes network element data of which the maximum value of PRB utilization rate within 24 hours in a single carrier network element is less than or equal to 60% and network element data of which the minimum value of either PRB utilization rate of a first carrier or PRB utilization rate of a second carrier within 24 hours in a double carrier network element is greater than 60%, and obtains an intermediate sample network element set.

And S403, the electronic equipment performs clustering operation on each sample network element in the middle sample network element set according to the PRB utilization rate value set of the carrier of each sample network element in the middle sample network element set within 24 hours.

S404, the electronic equipment draws a PRB utilization rate trend graph of each group of sample network elements obtained by clustering operation within the 24 hours.

S405, the electronic device displays the PRB utilization rate trend graph of each group of sample network elements so as to obtain a first label and a second label corresponding to the PRB utilization rate trend graph of each group of sample network elements, which is input by the operator based on the PRB utilization rate trend graph of each group of sample network elements.

S406, the electronic device labels the PRB utilization rate trend graphs of the sample network elements in each group based on the first labels and the second labels corresponding to the PRB utilization rate trend graphs of the sample network elements in each group.

S407, the electronic device pairs each group of sample network elements with the first label and the second label, and labels the pairing identifier to obtain a labeled intermediate sample network element set.

S408, the electronic equipment takes the marked middle sample network element set and the PRB utilization rate trend graph of each sample network element in the middle sample network element set as training sample sets, and trains the image recognition model to obtain the trained image recognition model.

The trained image recognition model can automatically classify the network elements based on PRB utilization rate trend graphs of the network elements, and can accurately acquire information of network elements matched with the network elements according to matching identifiers.

The image recognition model training method provided by the embodiment can automatically match a pair of network elements capable of participating in License deployment without manual participation and processing, and improves the efficiency and accuracy of network element classification and pairing.

Fig. 8 is a schematic structural diagram of an image recognition model training apparatus according to an embodiment of the present application.

As shown in fig. 8, the apparatus includes an acquisition module 21 and a training module 22. Wherein:

an obtaining module 21, configured to obtain a training sample set; wherein the training sample set comprises: the marked intermediate sample network element set and a first target parameter trend graph of each sample network element in the intermediate sample network element set; the labeled intermediate sample network element set comprises at least one group of single-carrier sample network elements and at least one group of double-carrier sample network elements, the single-carrier sample network elements belonging to the same group have the same first label and pairing identification, and the first label comprises: the frequency of occurrence of a peak in the first target parameter trend graph of the single carrier sample network element and the time of occurrence of the peak; the dual carrier sample network elements belonging to the same group have the same second label and pairing identification, and the second label includes: the number of troughs of two carriers appearing simultaneously in the first target parameter trend graph of the dual-carrier sample network element and the time of the troughs appearing simultaneously; the pairing identification is used for representing a network element group paired with the group of sample network elements, the paired sample network element group comprises a group of single-carrier sample network elements and a group of double-carrier sample network elements, peaks and troughs of the first target parameter trend graph of the paired single-carrier sample network elements and the double-carrier sample network elements are complementary, and the first target parameter value can represent the service condition of License resources of the sample network elements.

The training module 22 is configured to train an image recognition model by using the training sample set to obtain a trained image recognition model; the trained image recognition model is used for acquiring a group to which the network element belongs and a network element group matched with the network element according to a first target parameter trend graph of the network element.

Optionally, the obtaining module 21 is specifically configured to:

The image recognition model training device provided by the application is used for executing the embodiment of the image recognition model training method, the implementation principle and the technical effect are similar, and the details are not repeated.

Fig. 9 is a schematic structural diagram of a carrier adjustment apparatus according to an embodiment of the present application. As shown in fig. 9, the apparatus includes: a first obtaining module 31, a second obtaining module 32, a third obtaining module 33, and an adjusting module 34. Wherein:

a first obtaining module 31, configured to obtain a network element of a carrier to be adjusted.

A second obtaining module 32, configured to obtain, according to a first target parameter trend graph of a network element, a group to which the network element belongs and a network element group paired with the network element by using a trained image recognition model; the trained image recognition model can be obtained by adopting the training method in steps S201 to S202 in the above method embodiment; the first target parameter can represent the use condition of License resources of the network element.

A third obtaining module 33, configured to obtain a network element list of License resources to be adjusted according to the group to which the network element belongs and the network element group paired with the network element; the network element list includes: and the network element sub-list of the License resources to be increased and the network element sub-list of the License resources to be reduced.

An adjusting module 34, configured to adjust a License resource of a carrier of a network element in the network element list.

Optionally, the adjusting module 34 is specifically configured to:

sending the target License stock adjusting template to a License platform;

Optionally, the apparatus may further include: a fourth obtaining module 35, where the fourth obtaining module 35 is configured to: and acquiring the preset License stock adjusting template from the License platform before the adjusting module acquires a target License stock adjusting template based on a preset License stock adjusting template and the License stock adjusting list.

Optionally, the third obtaining module 33 is further configured to: before a network element list of License resources to be adjusted is acquired according to the group to which the network element belongs and the network element group matched with the network element, whether carrier adjustment is carried out on the network element is determined according to a second target parameter value of the network element; the second target parameter can represent the use condition of License resources of the network element, and the first target parameter is different from the second target parameter.

The carrier adjustment device provided by the present application is used for executing the foregoing carrier adjustment embodiments, and the implementation principle and the technical effect thereof are similar, which are not described again.

Fig. 10 is a schematic structural diagram of an electronic device according to an embodiment of the present application. As shown in fig. 10, the electronic device 400 may include: at least one processor 401 and memory 402.

A memory 402 for storing programs. In particular, the program may include program code including computer operating instructions.

The Memory 402 may include a Random Access Memory (RAM) and may also include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory.

The processor 401 is configured to execute computer-executable instructions stored in the memory 402 to implement the image recognition model training method and/or the carrier adjustment method described in the foregoing method embodiments. The processor 401 may be a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits configured to implement the embodiments of the present Application.

Optionally, the electronic device 400 may further include a communication interface 403. In a specific implementation, if the communication interface 403, the memory 402 and the processor 401 are implemented independently, the communication interface 403, the memory 402 and the processor 401 may be connected to each other through a bus and perform communication with each other. The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended ISA (EISA) bus, or the like. Buses may be classified as address buses, data buses, control buses, etc., but do not represent only one bus or type of bus.

Alternatively, in a specific implementation, if the communication interface 403, the memory 402 and the processor 401 are integrated into a single chip, the communication interface 403, the memory 402 and the processor 401 may complete communication through an internal interface.

The present application also provides a computer-readable storage medium, which may include: various media that can store program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a RAM Memory, a magnetic disk, or an optical disk, and in particular, the computer-readable storage medium stores program instructions, and the program instructions are used for the method in the above-mentioned embodiments.

The present application also provides a program product comprising execution instructions stored in a readable storage medium. The at least one processor of the electronic device may read the execution instruction from the readable storage medium, and the execution of the execution instruction by the at least one processor causes the electronic device to implement the image recognition model training method and/or the carrier adjustment method provided by the various embodiments described above.

Finally, it should be noted that: the above embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present application.

Claims

1. An image recognition model training method, characterized in that the method comprises:

2. The method of claim 1, wherein the obtaining a training sample set comprises:

3. The method of claim 2, wherein obtaining the set of intermediate sample cells comprises:

4. A method for carrier adjustment, the method comprising:

acquiring a network element of a carrier to be adjusted;

according to a first target parameter trend graph of a network element, using a trained image recognition model to obtain a group to which the network element belongs and a network element group matched with the network element; the trained image recognition model is obtained by adopting the training method according to any one of claims 1 to 3; the first target parameter can represent the use condition of License resources of the network element;

5. The method of claim 4, wherein the adjusting License resources of carriers of the network elements in the network element list comprises:

6. The method of claim 5, further comprising:

7. The method of claim 5, wherein the adjusting License resources of carriers of network elements in the License stock adjustment list according to the License stock adjustment list comprises:

8. The method of claim 7, wherein the obtaining the License adjustment file executable by the element management platform according to the License inventory adjustment list comprises:

sending the target License stock adjusting template to a License platform;

9. The method of claim 8, wherein before the obtaining the target License inventory adjustment template, based on the preset License inventory adjustment template and the License inventory adjustment manifest, further comprises:

10. The method according to any one of claims 4 to 9, wherein before obtaining the list of network elements for License resources to be adjusted according to the group to which the network element belongs and the network element group paired with the network element, the method further comprises:

11. An image recognition model training apparatus, characterized in that the apparatus comprises:

12. An apparatus for carrier adjustment, the apparatus comprising:

the second acquisition module is used for acquiring the group to which the network element belongs and the network element group matched with the network element by using the trained image recognition model according to the first target parameter trend graph of the network element; the trained image recognition model is obtained by adopting the training method according to any one of claims 1 to 3; the first target parameter can represent the use condition of License resources of the network element;

13. An electronic device, characterized in that the electronic device comprises: at least one processor and memory;

the memory stores computer-executable instructions;

the at least one processor executing the computer-executable instructions stored by the memory causes the electronic device to perform the method of any of claims 1-10.

14. A computer-readable storage medium having computer-executable instructions stored thereon, which when executed by a processor, perform the method of any one of claims 1-10.

15. A computer program product, comprising a computer program which, when executed by a processor, implements the method of any one of claims 1-10.