CN115515171A

CN115515171A - Load prediction method and device of SA network and electronic equipment

Info

Publication number: CN115515171A
Application number: CN202110688324.9A
Authority: CN
Inventors: 袁姣红; 戴晓群
Original assignee: China Mobile Communications Group Co Ltd; China Mobile Group Henan Co Ltd
Current assignee: China Mobile Communications Group Co Ltd; China Mobile Group Henan Co Ltd
Priority date: 2021-06-21
Filing date: 2021-06-21
Publication date: 2022-12-23

Abstract

The embodiment of the application provides a load prediction method and device of an SA network and electronic equipment, target sample data in the SA network is obtained, a corresponding load index is determined aiming at the SA network based on the target sample data, the index correlation prediction value of the index correlation at a future specified time is predicted by a first prediction model, the first index value of the load index at the future specified time is predicted by a second prediction model, and early warning is carried out on the SA network according to the first index value. And predicting the first index value of the load index at a future specified time by combining the first prediction model and the second prediction model with target sample data. Therefore, when the predicted load index in the SA network is abnormal, the load index can be processed in time to avoid the load abnormality of the SA network, the internet surfing experience of a 5G client is guaranteed, and the experience of the user on the 5G network is improved.

Description

Load prediction method and device of SA network and electronic equipment

Technical Field

The present application relates to the field of information communication technologies, and in particular, to a load prediction method and apparatus for an SA network, and an electronic device.

Background

The fifth generation mobile communication technology (5thgenerationmobilecommunications technology, 5g) is a new generation broadband mobile communication technology with high rate, low latency, and large connectivity. The 5G is divided into two networking types, namely Non-stand alone (NSA) and Stand Alone (SA). The SA is to create a new network including a new base station, a backhaul link, and a core network, thereby implementing all features and functions of the 5G network and providing a high-rate internet experience for a user. For the SA network, the load situation in the SA network is related to the internet access rate of the 5G client. How to predict the load in the SA network to ensure the internet surfing experience of the 5G customer is a technical problem to be solved by those skilled in the art.

Disclosure of Invention

The embodiment of the application aims to provide a load prediction method and device for an SA network and electronic equipment, so that the load in the SA network is predicted, and the internet surfing experience of a 5G client is guaranteed.

In order to solve the above technical problem, the embodiment of the present application is implemented as follows:

in a first aspect, an embodiment of the present application provides a load prediction method for an SA network, including:

acquiring target sample data in the SA network;

determining a corresponding load index in the SA network based on the target sample data;

at least one part of the target sample data is used as an input variable of a first prediction model, and an index correlation quantity predicted value of the index correlation quantity at a future specified time is predicted;

predicting a first index value of the load index at the future designated time by taking the index correlation quantity predicted value as an input variable of a second prediction model;

and early warning the SA network according to the first index value.

In a second aspect, an embodiment of the present application provides a load prediction apparatus for an SA network, including:

the acquisition module is used for acquiring target sample data in the SA network;

the determining module is used for determining a corresponding load index for the SA network based on the target sample data;

the first prediction module is used for predicting the index correlation quantity predicted value of the index correlation quantity at a future specified time by taking at least one part of the target sample data as an input variable of a first prediction model;

the second prediction module is used for taking the index correlation quantity predicted value as an input variable of a second prediction model and predicting a first index value of the load index at the future specified time;

and the early warning module is used for early warning the SA network according to the first index value.

In a third aspect, an embodiment of the present application provides an electronic device, including a processor, a communication interface, a memory, and a communication bus; the processor, the communication interface and the memory complete mutual communication through a bus; the memory is used for storing a computer program; the processor is configured to execute the program stored in the memory to implement the steps of the load prediction method for the SA network according to the first aspect.

In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the load prediction method of the SA network according to the first aspect are implemented.

According to the technical scheme provided by the embodiment of the application, the target sample data in the SA network is obtained, the corresponding load index is determined for the SA network based on the target sample data, at least one part of the target sample data is used as an input variable of a first prediction model, the index correlation prediction value of the load index at a future specified time is predicted, the index correlation prediction value is used as an input variable of a second prediction model, the first index value of the load index at the future specified time is predicted, and the SA network is warned according to the first index value. And predicting a first index value of the load index at a future specified time by combining the first prediction model and the second prediction model with target sample data, and then early warning the SA network according to the first index value. Therefore, when the predicted load index in the SA network is abnormal, early warning can be sent out, so that maintenance personnel can process the load index in time, the abnormal load condition of the SA network is avoided, the Internet surfing experience of 5G clients is guaranteed, and the experience of users on the 5G network is improved.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the description below are only some embodiments described in the present application, and for those skilled in the art, other drawings may be obtained according to these drawings without creative efforts.

Fig. 1 is a first flowchart illustrating a load prediction method of an SA network according to an embodiment of the present application;

FIG. 2 is an architecture diagram of a first predictive model provided by an embodiment of the present application;

fig. 3 is a second flowchart illustrating a load prediction method of an SA network according to an embodiment of the present application;

fig. 4 is a third flowchart illustrating a load prediction method of an SA network according to an embodiment of the present application;

fig. 5 is a schematic block diagram illustrating a load prediction apparatus of an SA network according to an embodiment of the present application;

fig. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

The embodiment of the application provides a load prediction method and device of an SA network and electronic equipment, so that the load in the SA network is predicted, and the internet surfing experience of a 5G client is guaranteed.

In order to make those skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

In some scenarios, the SA Network architecture includes a 5G virtualization Function (VNF), and the VNF may be a virtualized Network element, which implements functions in the SA Network. For example, the Access and Mobility Management Function (AMF) may correspond to a Mobility Management Network element, which is responsible for Mobility of users and Access Management, session Management Function (SMF), may correspond to a Session Management Network element, which is responsible for Session Management of users, policy Control Function (PCF), which may correspond to a Policy Control Network element, which is responsible for Policy Control, user Plane Function (UPF), which may correspond to a User Plane Network element, which is responsible for routing and forwarding Function of User Plane, unified Data Management Function (UDM), which may correspond to a Unified Data Management Network element, which is responsible for Unified processing of Data, including User identification, user subscription Data and authentication Data, and charging Function (CHF), the Network Slice Selection Function (NSSF) may correspond to a charging Network element, be responsible for online charging, offline charging, and fusion charging, may correspond to a Network Slice Selection Network element, be responsible for managing information related to Network slices, be responsible for a Network open Function (NEF), be responsible for a Network open Network element, be responsible for managing externally open Network Data, be responsible for a Network storage Function (NRF), be responsible for a Network storage Network element, be responsible for registering and managing each Network Function, be responsible for an Application Function (AF), be responsible for an Application Network element, be responsible for various services of an Application layer, user Equipment (UE), a Data Network (Data Network, DN), be responsible for a Data Network element, be responsible for an operator service Function, internet Access services or third party services, radio Access Networks (RAN), such as a gbnodeb base station, etc.

The UE is connected with the AMF through an N1 interface, the RAN is connected with the AMF through an N2 interface, the RAN is connected with the UPF through an N3 interface, the UPF is connected with the SMF through an N4 interface, the UPF is connected with the DN through an N6 interface, and the UPF can be connected through N9 interfaces (such as N9a and N9c interfaces). The PCF is connected with the AF through an N5 interface.

For SA networks, the SA network load situation is related to the internet speed of 5G clients. How to predict the load in the SA network to ensure the internet surfing experience of 5G clients is a technical problem to be solved by those skilled in the art.

Therefore, the embodiments of the present application provide a technical solution that can solve the above problems, and refer to the following contents specifically.

For example, as shown in fig. 1, an execution subject of the method may be a server, where the server may be an independent server or a server cluster composed of a plurality of servers, and the server may be a server capable of load prediction of an SA network.

The load prediction method of the SA network may specifically include the steps of:

in S101, target sample data in the SA network is acquired.

Specifically, target sample data in the SA network includes, but is not limited to: 5G terminal data (e.g. total amount of 5G terminals, 5G terminal promotion plan and user change machine rhythm, 5G soft switch on duty ratio of 5G terminals, version upgrade progress of SA network system priority of 5G terminals, etc.), gbodeb data (e.g. number of gbodeb base stations and coverage area, construction plan of gbodeb base stations, etc.), traffic of 5G VNF network elements (e.g. number of AMF registered users, AMF hooks, number of 5G base stations, number of SMF peak sessions, number of PCF signed users, number of SMF/UPF peak Quality of Service (QoS) flows, N3/N6 interface flow, maximum allocated address number, N3/N9a/N9c/N6 interface receiving a set of General Packet Radio Service (GPRS) tunneling Protocol (GPRS tunneling Protocol, GTP) Packet, number of bytes per CHF time unit and 5G on UDM), number of signaling users and number of signaling users (e.g. signaling users, number of signaling users, and number of load, and number of network resources (e.g. CPU resource usage), etc.), and usage of CPU resource usage (e.g. CPU usage rate).

In a possible implementation manner, after obtaining target sample data in the SA network, before determining a corresponding load indicator for the SA network based on the target sample data, the method further includes the following steps:

and preprocessing target sample data, wherein the preprocessing comprises data cleaning processing and null value interpolation processing.

Specifically, data cleaning is carried out on target sample data to remove data which do not accord with the actual situation of the current network so as to improve the accuracy of the target sample data and further improve the prediction precision of the prediction model. For example, data with daily ring ratio fluctuation range of working users exceeding a first threshold (e.g., 10%) or flow fluctuation range exceeding a second threshold (e.g., 15%), data with daily ring ratio fluctuation range of users on holidays exceeding a third threshold (e.g., 20%) or data with flow fluctuation range exceeding a fourth threshold (e.g., 30%) are removed. And performing null interpolation processing on the target data sample, namely performing interpolation through median, mean or mode and the like.

In S102, based on the target sample data, a corresponding load index is determined for the SA network.

Specifically, the core network element in the SA network plays an important role in the internet speed of the user, and in one possible implementation, when calculating the load index in the SA network, the load index of the core network element in the SA network may be used as the load index in the SA network. Specifically, based on the types of the target sample data, capacity type indexes such as an AMF registration capacity utilization rate, an AMF base station management capacity utilization rate, an SMF session capacity peak utilization rate, an SMF peak QoS flow capacity utilization rate, an UPF peak QoS flow capacity utilization rate, an UDM user capacity utilization rate, a PCF user capacity utilization rate and the like, and load indexes such as an address pool peak utilization rate, an UPF data throughput capacity utilization rate, a traffic bandwidth utilization rate (for example, an N3/N6 interface traffic bandwidth utilization rate), a signaling load, a telecommunications cloud resource utilization rate and the like can be calculated.

The AMF registration capacity utilization rate, the AMF base station management capacity utilization rate, the SMF session capacity peak utilization rate, the SMF peak QoS flow capacity utilization rate, the UPF peak QoS flow capacity utilization rate, the UDM user capacity utilization rate, the CHF ticket processing capacity utilization rate, and the PCF user capacity utilization rate may be collectively referred to as a capacity utilization rate.

The AMF registration capacity utilization rate = AMF registered user number/AMF registration capacity License, the AMF registration capacity utilization rate is a key capacity index of the AMF, and if the AMF registration capacity utilization rate is too high, a user may not be successfully registered in the 5G SA network.

The AMF base station management capacity utilization ratio = AMF number of hooked 5G base stations/maximum number of supported base stations, and the number of base stations for AMF support connection is usually limited, so the capacity managed by the AMF base station needs to be considered when a new nodeb base station is created.

The SMF session capacity peak utilization = peak PDU session number/maximum PDU session number, and the SMF session capacity peak utilization is a core capacity load index of the SMF, which is a content that must be paid attention to in network planning to know whether a traffic overflow situation will occur. In the 5G network, information interaction between the UE and the network is performed through a PDU session, which provides data connectivity between the 5G UE and the DN, all data traffic of the UE must be carried through the PDU session, and the PDU session should be established before the UE sends the data traffic.

SMF peak QoS flow capacity utilization = peak QoS flow number/maximum QoS flow number.

UPF peak QoS flow capacity utilization = peak QoS flow number/maximum QoS flow number, one PDU session may include up to 64 QoS flows, each QoS flow having different TFT parameters, the SMF peak QoS flow capacity utilization and the UPF peak QoS flow capacity utilization being mainly used to determine whether there is a traffic overflow situation.

The UDM user capacity utilization rate =5G user Number/maximum storable Mobile Subscriber Number (MSISDN) Number, which is used for knowing the change trend of the UDM storage capacity utilization rate and preparing network capacity expansion in advance.

The PCF user capacity utilization rate = PCF signing user number/PCF maximum signing user number, the ratio of the signing users actually stored by the UDR of the PCF to the capacity of the signing users, and the index is used for knowing whether the service limitation condition exists.

The utilization rate of the processing capacity of the CHF call ticket = CHF unit time call ticket volume/maximum call ticket processing capacity, the call ticket is an income source of a telecommunication operator, the correct processing of the call ticket is important, and the overstocked call ticket may cause the limitation of user services and income loss, so that the utilization rate of the processing capacity of the call ticket of the CHF call ticket needs to be concerned, and the overload operation is avoided.

The address pool peak utilization = number of maximum assigned addresses/total number of IP addresses contained in the address pool, a ratio of the maximum number of addresses actually used (assigned) by the address pool to the total number of addresses contained in the address pool, and the index is used to know whether a traffic limitation condition exists.

The UPF data throughput capacity utilization rate = (the number of GTP packet bytes received by an N3 interface + the number of GTP packet bytes received by an N9a interface + the number of GTP packet bytes received by an N9c interface + the number of GTP packet bytes received by an N6 interface) × 8/(the maximum data throughput rate × 1000), and the index is used for knowing whether the data throughput capacity of the UPF complete machine has a bottleneck.

In S103, at least a part of the target sample data is used as an input variable of the first prediction model, and a predicted value of the index correlation value at a future specified time is predicted.

Specifically, the first prediction model is a prediction model that has been trained, and the first prediction model may be a BP neural network model. The BP neural network model adopts a three-layer structure, and the structure is shown in figure 2, wherein x _i Respectively, input variables, i =1,2,3, \8230, and n, wherein n is the number of the input variables. For example, when n is 7, the input variables of the BP neural network model may be the total amount of 5G terminals at a certain time point, the promotion plan and user switching rhythm of the 5G terminals, the 5G soft switch on duty ratio of the 5G terminals, the version upgrade progress with priority of the SA network system of the 5G terminals, the coverage condition of the gbodeb, and the AMF

The number of the registered users and the N6 interface flow. Namely, the input layer of the BP neural network model has 7 neurons, and Y is the output of the BP neural network model, which is the index related quantity (such as the predicted value of the number of 5G users and the predicted value of 5G flow) at a specified time in the future. The number of neurons in the middle layer of the BP neural network model can be 14, a sigmoid function is adopted in a hidden layer, and y = x is adopted in an output layer function.

The input and output relationships of the BP neural network model can be expressed by the following equation:

the weight from the mth neuron to the neuron of the intermediate layer to the neuron of the output layer, tansig represents a sigmoid function,

the weights of the neuron corresponding to the ith value to the mth neuron in the middle layer are input,

is the threshold of the mth neuron in the middle layer, b ^o Is the threshold for output layer neurons.

And a predicted value representing the predicted index-related quantity.

In one possible implementation, with at least a part of the target sample data as an input variable of the first prediction model, predicting the index correlation prediction value of the index correlation at a specified future time includes:

and taking 5G terminal data (the total amount of 5G terminals, a 5G terminal promotion plan and user switching rhythm, the 5G soft switch opening ratio of the 5G terminals, the version upgrading progress of the 5G terminal SA network system priority), gNodeB coverage data, AMF registered user number and N6 interface flow as input variables of the BP neural network model.

And predicting the number of 5G users and the predicted value of 5G flow at the specified time in the future.

And taking the index correlation quantity predicted value as an input variable of the second prediction model, and predicting a first index value of the load index at a future specified time comprises the following steps:

and (3) taking the predicted value of the 5G user number and the predicted value of the 5G flow as input variables of a binary linear regression model, and predicting a first index value of the load index at a future designated time.

In a possible implementation manner, in order to reduce the prediction error of the BP neural network model and improve the prediction accuracy, the BP neural network model reversely transmits data from the output layer to the input layer, and the value of the weight w is readjusted until the prediction error of the BP neural network model reaches the minimum, and then the training is stopped, so that the training of the BP neural network model is completed. After predicting the index correlation magnitude for the load index at a specified time in the future, the method further comprises:

and calculating the precision score of the first prediction model based on the predicted value of the index correlation quantity and the actual value of the index correlation quantity in the target sample data.

Specifically, the accuracy score of the first prediction model may be calculated based on the following formula:

wherein R is ² A precision score representing the first predictive model,

an index correlation quantity predicted value representing an output of the first prediction model,

actual value, N, representing the index-related quantity _S Representing the size of the sample size input into the first predictive model.

And under the condition that the precision score does not meet the preset requirement, the index correlation quantity at the specified time in the future is predicted again.

In S104, a first index value of the load index at a future designated time is predicted using the index correlation value as an input variable of the second prediction model.

Specifically, as the index correlation amount, there is a significant linear correlation with the load index, for example, there is a significant linear correlation between the 5G user amount and the 5G flow rate and each load index.

In a possible implementation mode, a classical binary linear regression model is used as a second prediction model to reflect a linear relation, namely after the BP neural network model is adopted to obtain the predicted values of the number of 5G users and the 5G flow at the future designated time, the predicted values of all load indexes are further calculated based on the binary linear regression model. The method comprises the following specific steps:

setting dependent variable Y (each load index) and independent variable X (index-related quantity predicted values, e.g. X for each index) ₁ And X ₂ Representing the predicted value of the 5G user quantity and the predicted value of the 5G flow) satisfies the following formula:

Y＝β ₀ +β ₁ X ₁ +β ₂ X ₂ +ε

wherein, beta ₀ 、β ₁ And beta ₂ Is a binary regression parameter, epsilon is a statistical error, and follows a normal distribution with a mean value of zero, i.e., epsilon: N (0, sigma) ² )。

For (X) ₁ ,X ₂ Y) for n experiments, a sample with a capacity of n and a finite sample model can be obtained, as follows:

least squares estimation of beta from least squares

As shown in the following formula:

the above formula is a binary regression prediction model of load indexes such as capacity utilization rate, address pool utilization rate, traffic bandwidth utilization rate, signaling load, data throughput rate, telecommunication cloud resource utilization rate and the like. Least squares estimation by the binary regression prediction model

X is a known quantity, and the first index value of each load index can be predicted through the binary regression model.

In S105, the SA network is warned according to the first index value.

Specifically, whether the load index corresponding to each first index value is abnormal or not is judged according to the magnitude of the first index value, and when the load index is abnormal, early warning is given.

In one possible implementation manner, as shown in fig. 3, S105 specifically includes the following steps:

in S1051, for the SA network, the corresponding load level is determined from the first index value.

In particular, in one possible implementation, the load level may be divided into five levels. The first load level may be represented as a total resource shortage of the SA network, and in the first load level, index parameters such as a capacity utilization rate, an address pool peak utilization rate, a traffic bandwidth utilization rate, a signaling load, a UPF data throughput capacity utilization rate, and a telecommunications cloud resource utilization rate are all at a higher level (i.e., exceed respective corresponding early warning thresholds). The second load level may be represented as a network capability (e.g., a core network element) of the SA network, and can satisfy user experience, where the capacity utilization rate, the address pool peak utilization rate, the traffic bandwidth utilization rate, the signaling load, and the UPF data throughput capacity utilization rate are at a lower level (all do not exceed their respective corresponding early warning thresholds), and the telecommunications cloud resource utilization rate is at a higher level (exceeds a corresponding early warning threshold).

The third load level may be expressed as software capacity License or address pool resource shortage (e.g. core network element) of the SA network, which may result in 5G traffic limitation. In the third load level, at least one load index in the capacity utilization rate and/or the address pool peak utilization rate load indexes exceeds an early warning threshold, but the traffic bandwidth utilization rate, the signaling load, the UPF data throughput capacity and the telecommunication cloud resource utilization rate are all in a lower level (all do not exceed the respective corresponding early warning thresholds).

The fourth load level may indicate that traffic bandwidth or data throughput or forwarding processing capacity in the SA network is insufficient (e.g., a core network element), and in the fourth load level, resources of a telecommunications cloud resource utilization rate, a software capacity License and an address pool are sufficient, but there is a case where at least one load index exceeds an early warning threshold in an N3/N6 interface traffic bandwidth utilization rate, a signaling load or a UPF data throughput capacity utilization rate.

A fifth load level, in which all load indicators are below the early warning threshold, may be expressed as a resource surplus in the SA network.

In S1052, an early warning work order corresponding to the load level is issued.

And after the corresponding load grades are determined, sending out early warning work orders corresponding to the load grades, wherein the early warning work orders carry abnormal load indexes in the corresponding load grades. According to the level of the load level, the load level and the early warning work order can be marked, the importance degree of each load level is marked in sequence according to the level of the load level, and the lower the load level is, the higher the importance degree of the early warning work order is, so that maintenance personnel can process the early warning work order preferentially. For example, the early warning work order corresponding to the first load level is marked with a numeral 1 to represent a response level, for example, a first-level response indicates that the first-level response is the most important early warning work order and needs urgent processing, and correspondingly, the second load level to the fifth load level are marked with 2 to 5 in sequence, and the represented response levels are a second-level response, a third-level response, a fourth-level response and a fifth-level response respectively. When the early warning work order is marked as 5, the resources of the SA network are rich, and the early warning work order can not be sent out.

According to the technical scheme provided by the embodiment of the application, the first index value of the load index at the future designated time is predicted by combining the first prediction model and the second prediction model with the target sample data, so that the SA network is pre-warned according to the first index value. Therefore, when the predicted load index of the SA network is abnormal, early warning can be sent out, so that maintenance personnel can timely process the load index, the abnormal load condition of the SA network is avoided, the Internet surfing experience of a 5G client is guaranteed, and the experience of a user on the 5G network is improved.

For example, as shown in fig. 4, an execution subject of the method may be a server, where the server may be an independent server or a server cluster composed of a plurality of servers, and the server may be a server capable of load prediction of the SA network.

in S401, target sample data in the SA network is acquired.

In S402, based on the target sample data, a corresponding load index is determined for the SA network.

In S403, for the target sample data, a second index value of the load index at the acquisition time of the target sample data is calculated.

Specifically, the target sample data is acquired at the acquisition time, and then the second index value of each load index at the acquisition time is calculated according to the calculation mode of each load index in S102.

In a possible implementation manner, after S403, the following steps are further included:

and marking the load index corresponding to the second index value exceeding the threshold value.

And storing the marked load index and the load index corresponding to the second index value which does not exceed the threshold value.

In particular, the load indicator with a second indicator value exceeding the threshold may be marked with the marker "Abnormal". Therefore, maintenance personnel can conveniently and directly search the abnormal load index, and in addition, the second index which does not exceed the threshold value can be marked by adopting a marking word 'normal' so that the maintenance personnel can distinguish the abnormal load index from the normal load index.

In S404, for the SA network, a corresponding load level is determined according to the second index value.

In S405, an early warning corresponding to the load level is performed.

In S406, at least a part of the target sample data is used as an input variable of the first prediction model, and a predicted value of the index correlation value at a future specified time is predicted.

In S407, the first index value of the load index at the future designated time is predicted using the index correlation value as the input variable of the second prediction model.

In S408, the SA network is warned according to the first index value.

It is to be noted that the implementations of S401 to S402 and S406 to S408 are the same as or similar to the implementations of S101 to S105, which can be referred to each other, and the details of the embodiments of the present application are not repeated herein.

According to the technical scheme provided by the embodiment of the application, the first index value of the load index at the future designated time is predicted by combining the first prediction model and the second prediction model with the target sample data, so that the SA network is pre-warned according to the first index value. Therefore, when the predicted load index of the SA network is abnormal, the load index can be processed in time to avoid the load abnormality of the SA network, the internet surfing experience of a 5G client is guaranteed, and the experience of the user on the 5G network is improved.

In addition, the index value of the load index at the acquisition time of the target sample data can be calculated, when the load index of the SA network calculated at the acquisition time is abnormal, the load index can be processed in time, the problem of load index abnormality can be solved quickly, the internet surfing experience of a 5G client is guaranteed, and the experience of a user on the 5G network is improved.

Corresponding to the load prediction method of the SA network provided in the foregoing embodiment, based on the same technical concept, the embodiment of the present application further provides a load prediction apparatus of the SA network, fig. 5 is a schematic diagram of module components of the load prediction apparatus of the SA network provided in the embodiment of the present application, the load prediction apparatus of the SA network is configured to execute the load prediction method of the SA network described in fig. 1 to fig. 4, and as shown in fig. 5, the load prediction apparatus of the SA network includes: the system comprises an acquisition module 501, a determination module 502, a first prediction module 503, a second prediction module 504 and an early warning module 505.

An obtaining module 501, configured to obtain target sample data in the SA network.

A determining module 502, configured to determine a corresponding load indicator for the SA network based on the target sample data.

The first prediction module 503 is configured to predict an index correlation value of the index correlation at a future specified time, using at least a portion of the target sample data as an input variable of the first prediction model.

And a second prediction module 504, configured to predict a first index value of the load index at a future specified time by using the index correlation value prediction value as an input variable of a second prediction model.

And the early warning module 505 is configured to perform early warning on the SA network according to the first index value.

According to the technical scheme provided by the embodiment of the application, the target sample data in the SA network is obtained, the corresponding load indexes are determined aiming at the SA network based on the target sample data, at least one part of the target sample data is used as an input variable of a first prediction model, the index correlation prediction value of the load index at the future appointed time is predicted, the index correlation prediction value is used as an input variable of a second prediction model, the first index value of the load index at the future appointed time is predicted, and the SA network is warned according to the first index value. And predicting a first index value of the load index at a future specified time by combining the first prediction model and the second prediction model with target sample data, and then early warning the SA network according to the first index value. Therefore, when the predicted load index in the SA network is abnormal, early warning can be sent out, so that maintenance personnel can timely process the load index, the abnormal load condition of the SA network is avoided, the Internet surfing experience of a 5G client is guaranteed, and the experience of a user on the 5G network is improved.

Optionally, the apparatus for predicting a load of a core network element of the SA network further includes: a calculation module (not shown), a determination module II (not shown), and an early warning module II (not shown).

And the calculation module is used for calculating a second index value of the load index at the acquisition moment of the target sample data according to the target sample data.

And a second determining module, configured to determine, for the SA network, a corresponding load level according to the second index value.

And the early warning module II is used for carrying out early warning corresponding to the load level.

Optionally, the load prediction apparatus of the SA network further includes: a marking module (not shown), and a storage module (not shown).

And the marking module is used for marking the load index corresponding to the second index value exceeding the threshold value.

And the storage module is used for storing the marked load indexes.

Optionally, the load prediction apparatus of the SA network further includes: a pre-processing module (not shown).

And the preprocessing module is used for preprocessing target sample data, and the preprocessing comprises data cleaning processing and null value interpolation processing.

Optionally, the early warning module 505 comprises: a determining unit and an issuing unit.

And the determining unit is used for determining the corresponding load level according to the first index value aiming at the SA network.

And the sending unit is used for sending the early warning work order corresponding to the load level.

Optionally, the first prediction model is a BP neural network model, the second prediction model is a binary linear regression model, and the first prediction module 503 includes:

and the prediction unit is used for predicting a predicted value of the number of 5G users and a predicted value of the 5G user number and the 5G flow of the 5G flow at a future specified time by taking the 5G terminal data, the gNodeB coverage data, the AMF registered user number and the N6 interface flow of a preset time as input variables of the BP neural network model.

The second prediction module 504 includes:

and the prediction unit is used for predicting a first index value of the load index at a future designated time by taking the predicted value of the 5G user quantity and the predicted value of the 5G flow as input variables of the binary linear regression model.

Optionally, the load prediction apparatus of the SA network further includes: a calculation module (not shown) and a third prediction module (not shown).

And the calculation module is used for calculating the precision score of the first prediction model based on the index correlation value predicted value and the actual value of the index correlation value in the target sample data.

And the third prediction module is used for predicting the index correlation quantity of the future specified time again under the condition that the precision score does not meet the preset requirement.

The load prediction apparatus based on the SA network according to the embodiment of the present application can implement each process in the embodiment corresponding to the load prediction method of the SA network, and is not described here again to avoid repetition.

It should be noted that the load prediction apparatus of the SA network provided in the embodiment of the present application and the load prediction method of the SA network provided in the embodiment of the present application are based on the same application concept, so that specific implementation of the embodiment may refer to implementation of the load prediction method of the SA network, and repeated details are not repeated.

Based on the same technical concept, the embodiment of the present application further provides an electronic device for executing the load prediction method of the SA network, and fig. 6 is a schematic structural diagram of an electronic device implementing various embodiments of the present application, as shown in fig. 6. Electronic devices may vary widely in configuration or performance and may include one or more processors 601 and memory 602, where one or more stored applications or data may be stored in memory 602. Wherein the memory 602 may be transient or persistent storage. The application program stored in memory 602 may include one or more modules (not shown), each of which may include a series of computer-executable instructions for the electronic device.

Still further, the processor 601 may be arranged in communication with the memory 602 to execute a series of computer-executable instructions in the memory 602 on the electronic device. The electronic device may also include one or more power supplies 603, one or more wired or wireless network interfaces 604, one or more input-output interfaces 605, one or more keyboards 606.

Specifically, in this embodiment, the electronic device includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface and the memory are communicated with each other through the bus. A memory for storing a computer program. A processor for executing the program stored in the memory, implementing the following method steps:

and acquiring target sample data in the SA network.

Based on the target sample data, a corresponding load index is determined for the SA network.

And predicting the index correlation quantity predicted value of the index correlation quantity at a future specified time by taking at least one part of the target sample data as an input variable of the first prediction model.

And predicting a first index value of the load index at a future designated time by taking the index correlation quantity predicted value as an input variable of the second prediction model.

And early warning the SA network according to the first index value.

According to the technical scheme provided by the embodiment of the application, the target sample data in the SA network is obtained, the corresponding load index is determined for the SA network based on the target sample data, at least one part of the target sample data is used as an input variable of a first prediction model, the index correlation prediction value of the load index at a future specified time is predicted, the index correlation prediction value is used as an input variable of a second prediction model, the first index value of the load index at the future specified time is predicted, and the SA network is warned according to the first index value. And predicting a first index value of the load index at a future specified time by combining the first prediction model and the second prediction model with target sample data, and then early warning the SA network according to the first index value. Therefore, when the predicted load index in the SA network is abnormal, early warning can be sent out, so that maintenance personnel can timely process the load index, the abnormal load condition of the SA network is avoided, the Internet surfing experience of a 5G client is guaranteed, and the experience of a user on the 5G network is improved.

The embodiment further provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the computer program implements the following steps:

and acquiring target sample data in the SA network.

And early warning the SA network according to the first index value.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, apparatus, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In a typical configuration, an electronic device includes one or more processors (CPUs), input/output interfaces, a network interface, and a memory.

The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include transitory computer readable media (transmyedia) such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrases "comprising a," "8230," "8230," or "comprising" does not exclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, apparatus or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The above are merely examples of the present application and are not intended to limit the present application. Various modifications and changes may occur to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for load prediction in an SA network, the method comprising:

acquiring target sample data in the SA network;

determining a corresponding load index for the SA network based on the target sample data;

at least one part of the target sample data is used as an input variable of a first prediction model, and a predicted value of the index correlation quantity at a future specified moment is predicted;

and early warning the SA network according to the first index value.

2. The method according to claim 1, wherein after said determining a corresponding load metric for the SA network based on said target sample data, the method further comprises:

calculating a second index value of the load index at the acquisition moment of the target sample data aiming at the target sample data;

for the SA network, determining a corresponding load level according to the second index value;

and carrying out early warning corresponding to the load level.

3. The method according to claim 2, wherein after said calculating a second index value of said load index at a time of acquisition of said target sample data, said method further comprises:

marking the load index corresponding to the second index value exceeding the threshold value;

and storing the marked load index.

4. The method of claim 1, wherein after said obtaining target sample data in the SA network, the method further comprises:

and preprocessing the target sample data, wherein the preprocessing comprises data cleaning processing and null value interpolation processing.

5. The method according to any one of claims 1-4, wherein the pre-warning of the SA network according to the first index value comprises:

aiming at the SA network, determining a corresponding load level according to the first index value;

and sending an early warning work order corresponding to the load level.

6. The method of claim 1, wherein the first predictive model is a BP neural network model and the second predictive model is a binary linear regression model, and predicting the index-related quantity predicted value at a specified time in the future comprises:

taking 5G terminal data, gNodeB base station coverage data, AMF registered user number and N6 interface flow as input variables of the BP neural network model;

predicting the 5G user number predicted value and the 5G flow predicted value of the 5G user number and the 5G flow at the future designated time;

the step of predicting a first index value of the load index at the future designated time by using the index-related quantity predicted value as an input variable of a second prediction model comprises:

and predicting a first index value of the load index at the future designated time by taking the predicted value of the 5G user number and the predicted value of the 5G flow as input variables of the binary linear regression model.

7. The method of claim 1, wherein after predicting the first indicator value of the load indicator at the future specified time, the method further comprises:

calculating a precision score of the first prediction model based on the predicted value of the index correlation amount and an actual value of the index correlation amount in the target sample data;

8. A load prediction apparatus for an SA network, the apparatus comprising:

the first prediction module is used for predicting the index correlation quantity predicted value of the index correlation quantity at a future specified moment by taking at least one part of the target sample data as an input variable of a first prediction model;

9. An electronic device comprising a processor, a communication interface, a memory, and a communication bus; the processor, the communication interface and the memory complete mutual communication through a bus; the memory is used for storing a computer program; the processor, configured to execute the program stored in the memory, and implement the steps of the load prediction method of the SA network according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when being executed by a processor, carries out the method steps of load prediction of an SA network according to any one of claims 1 to 7.