CN113347659B - Flow prediction method and device - Google Patents
Flow prediction method and device Download PDFInfo
- Publication number
- CN113347659B CN113347659B CN202110610080.2A CN202110610080A CN113347659B CN 113347659 B CN113347659 B CN 113347659B CN 202110610080 A CN202110610080 A CN 202110610080A CN 113347659 B CN113347659 B CN 113347659B
- Authority
- CN
- China
- Prior art keywords
- target
- parameter
- sampling
- base station
- signaling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 37
- 230000011664 signaling Effects 0.000 claims abstract description 243
- 238000004458 analytical method Methods 0.000 claims abstract description 53
- 238000005206 flow analysis Methods 0.000 claims abstract description 48
- 238000012549 training Methods 0.000 claims abstract description 21
- 238000005070 sampling Methods 0.000 claims description 173
- 238000013527 convolutional neural network Methods 0.000 claims description 52
- 230000000875 corresponding effect Effects 0.000 claims description 35
- 230000015654 memory Effects 0.000 claims description 24
- 238000011176 pooling Methods 0.000 claims description 10
- 230000002596 correlated effect Effects 0.000 claims description 7
- 230000008014 freezing Effects 0.000 claims description 7
- 238000007710 freezing Methods 0.000 claims description 7
- 230000007787 long-term memory Effects 0.000 claims description 3
- 230000006403 short-term memory Effects 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000010363 phase shift Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The embodiment of the disclosure discloses a flow prediction method and a device, wherein the flow prediction method comprises the following steps: acquiring historical downlink flow parameters of a target base station and at least one type of historical signaling parameters of the target base station; acquiring a correlation coefficient between a historical signaling parameter and a historical downlink flow parameter, and determining at least one type of target signaling parameter in the at least one type of historical signaling parameter, wherein the correlation coefficient between the target signaling parameter and the historical downlink flow parameter is greater than a first correlation coefficient threshold value; determining a target traffic analysis model obtained by pre-training according to the target signaling parameters and the correlation coefficients of the target signaling parameters and the historical downlink traffic parameters; and inputting the target signaling parameter and the historical downlink flow parameter into the target flow analysis model to obtain a predicted flow value of the target base station. The traffic prediction method can predict the future traffic of the base station, so that the predicted future traffic of the base station is more accurate.
Description
Technical Field
The present disclosure relates to the field of communications technologies, and in particular, to a traffic prediction method and apparatus.
Background
With the development of wireless network technology in recent years, more and more users begin to use mobile terminal devices (e.g. smart phones, tablet computers, etc.) capable of accessing a wireless network, and usually, a base station that can be deployed by a telecom operator provides network access services for the mobile terminal devices of nearby users. In actual use, due to factors such as changes in user behaviors and upgrading of mobile terminal devices, telecommunication operators need to correspondingly adjust base stations (for example, increase or decrease the number of base stations, change devices of base stations, etc.) according to changes in traffic of the base stations so as to adapt to different traffic demands. It should be noted that such adjustment is a specific use case that needs to comprehensively consider future traffic of the base station, because if the traffic flow exceeds the load of the base station, so that the network service experience of the user is greatly reduced, then it is too late to adjust the relevant base station. Therefore, the future traffic of the base station needs to be predicted before the base station is adjusted accordingly. However, in the course of research and practice on the prior art, the inventors of the embodiments of the present application found that the prior art cannot accurately predict the future traffic of the base station.
Disclosure of Invention
The embodiment of the disclosure provides a traffic prediction method and a traffic prediction device.
In a first aspect, a traffic prediction method is provided in the embodiments of the present disclosure.
The method comprises the following steps:
acquiring historical downlink flow parameters of a target base station and at least one type of historical signaling parameters of the target base station;
obtaining a correlation coefficient between a historical signaling parameter and a historical downlink traffic parameter, and determining at least one type of target signaling parameter in the at least one type of historical signaling parameter, wherein the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than a first correlation coefficient threshold value;
determining a target traffic analysis model obtained by pre-training according to the target signaling parameter and the correlation coefficient of the target signaling parameter and the historical downlink traffic parameter;
and inputting the target signaling parameter and the historical downlink flow parameter into the target flow analysis model to obtain a predicted flow value of the target base station.
With reference to the first aspect, in a first implementation manner of the first aspect, the target traffic analysis model is composed of a target convolutional neural network model and a target long-short term memory network model;
inputting a target signaling parameter and a historical downlink flow parameter into a target flow analysis model to obtain a predicted flow value of a target base station, wherein the target signaling parameter and the historical downlink flow parameter are used as input, and the method comprises the following steps:
inputting a target signaling parameter and a historical downlink flow parameter as input, and inputting the target signaling parameter and the historical downlink flow parameter into a target convolutional neural network model to obtain target characteristics of the target signaling parameter and the historical downlink flow parameter;
and inputting the target characteristics as input into the target long-term and short-term memory network model to obtain a predicted flow value.
With reference to the first implementation manner of the first aspect, in a second implementation manner of the first aspect, the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is positively correlated to the convolutional layer convolutional kernel size corresponding to the target signaling parameter in the target convolutional neural network model.
With reference to the first implementation manner of the first aspect, in a third implementation manner of the first aspect, the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is inversely correlated with the number of convolutional layer convolutional kernels corresponding to the target signaling parameter in the target convolutional neural network model.
With reference to the first implementation manner of the first aspect, in a fourth implementation manner of the first aspect, when a correlation coefficient between a target signaling parameter and a historical downlink traffic parameter is greater than or equal to a second correlation coefficient threshold, the target convolutional neural network model includes a pooling layer corresponding to the target signaling parameter;
and when the correlation coefficient of the target signaling parameter and the historical downlink flow parameter is smaller than the second correlation coefficient threshold value, the target convolutional neural network model does not comprise the pooling layer corresponding to the target signaling parameter.
With reference to the first aspect, or any one of the first implementation manner of the first aspect to the fourth implementation manner of the first aspect, in a fifth implementation manner of the first aspect, the method further includes:
acquiring a sampling base station data set, wherein the sampling base station data set comprises sampling downlink flow parameters of a sampling base station and at least one type of sampling signaling parameters of the sampling base station, and the correlation coefficient of the sampling signaling parameters and the sampling downlink flow parameters is greater than a first correlation coefficient threshold;
determining an initial flow analysis model according to the sampling signaling parameters and the correlation coefficients of the sampling signaling parameters and the sampling downlink flow parameters;
and taking the sampling signaling parameters of a first time period and the sampling downlink flow parameters of the first time period in the sampling base station data set as input, and taking the sampling downlink flow parameters of a second time period in the sampling base station data set as output to train the initial downlink flow analysis model to obtain a target flow analysis model, wherein the first time period is earlier than the second time period.
With reference to the fifth implementation manner of the first aspect, in a sixth implementation manner of the first aspect, the initial traffic analysis model is composed of an initial convolutional neural network model and an initial long-short term memory network model;
taking the sampling signaling parameters of the first time period and the sampling downlink flow parameters of the first time period in the sampling base station data set as inputs, taking the sampling downlink flow parameters of the second time period in the sampling base station data set as outputs to train the initial downlink flow analysis model, and obtaining a target flow analysis model, wherein the method comprises the following steps:
determining a plurality of first time periods with the time length of 24 hours and a second time period corresponding to each first time period by taking 1 hour as a time interval;
taking a sampling signaling parameter of a first time period in a sampling base station data set and a sampling downlink flow parameter of the first time period as the input of an initial convolutional neural network model, taking the output of the initial convolutional neural network model as the input of an initial long-short term memory network model, taking a sampling downlink flow parameter of a second time period corresponding to the first time period in the sampling base station data set as the output to train the initial downlink flow analysis model, and obtaining a target flow analysis model, wherein the target flow analysis model consists of a target convolutional neural network model and a target long-short term memory network model
With reference to the fifth implementation manner of the first aspect, in a seventh implementation manner of the first aspect, the sampling base station includes a target base station and a similar base station of the target base station;
taking a sampling signaling parameter of a first time period and a sampling downlink flow parameter of the first time period in a sampling base station data set as input, taking a sampling downlink flow parameter of a second time period in the sampling base station data set as output to train an initial downlink flow analysis model, and obtaining a target flow analysis model, wherein the method comprises the following steps:
taking a sampling signaling parameter of a similar base station in a sampling base station data set in a first time interval and a sampling downlink flow parameter of the similar base station in the first time interval as inputs, taking the sampling downlink flow parameter of the similar base station in the sampling base station data set in a second time interval as an output to train an initial downlink flow analysis model, and obtaining a pre-training downlink flow analysis model;
freezing a convolution layer in a pre-training downlink flow analysis model;
and taking the sampling signaling parameter of the target base station in the sampling base station data set in the first time period and the sampling downlink flow parameter of the target base station in the first time period as inputs, and taking the sampling downlink flow parameter of the target base station in the sampling base station data set in the second time period as an output to train the pre-training downlink flow analysis model after freezing the convolutional layer, thereby obtaining the target flow analysis model.
In a second aspect, the present disclosure provides a flow prediction apparatus;
specifically, the flow rate prediction device includes:
a base station parameter obtaining module configured to obtain historical downlink traffic parameters of a target base station and at least one type of historical signaling parameters of the target base station;
the target parameter determining module is configured to acquire a correlation coefficient between the historical signaling parameter and the historical downlink traffic parameter, and determine at least one type of target signaling parameter from the at least one type of historical signaling parameter, wherein the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than a first correlation coefficient threshold;
the analysis model determining module is configured to determine a pre-trained target traffic analysis model according to the target signaling parameter and a correlation coefficient between the target signaling parameter and the historical downlink traffic parameter;
and the flow parameter prediction module is configured to input the target signaling parameter and the historical downlink flow parameter into the target flow analysis model to obtain a predicted flow value of the target base station.
In a third aspect, an embodiment of the present disclosure provides an electronic device, including a memory and at least one processor; wherein the memory is configured to store one or more computer instructions that are executable by the at least one processor to implement the method steps of any one of the first aspect, the first implementation of the first aspect, to the seventh implementation of the first aspect.
The technical scheme provided by the embodiment of the disclosure can have the following beneficial effects:
according to the technical scheme, the historical downlink flow parameters of the target base station and at least one type of historical signaling parameters of the target base station are obtained, and the correlation coefficient between the historical signaling parameters and the historical downlink flow parameters is obtained, wherein the correlation coefficient is used for indicating the correlation between the signaling parameters and the historical downlink flow parameters, and the stronger the correlation between the signaling parameters and the historical downlink flow parameters is, the higher the influence degree of the signaling parameters of the base station on the downlink flow of the base station is. And determining at least one type of target signaling parameters of which the correlation coefficient with the historical downlink traffic parameters is greater than a first correlation coefficient threshold value from the at least one type of historical signaling parameters, wherein the correlation coefficient between the target signaling parameters and the historical downlink traffic parameters is greater than the first correlation coefficient threshold value, namely the target signaling parameters are parameters with higher influence degree on the downlink traffic of the base station in the historical signaling parameters. And then, determining a target traffic analysis model obtained by pre-training according to the target signaling parameter and the correlation coefficient of the target signaling parameter and the historical downlink traffic parameter, so that the target traffic analysis model can analyze the future traffic of the target base station based on the target signaling parameter and the historical downlink traffic parameter. And then, inputting the target signaling parameter and the historical downlink flow parameter as input into a target flow analysis model to obtain a predicted flow value of the target base station. Therefore, the technical scheme of the application can obtain the predicted flow value of the target base station based on the signaling parameter with higher influence degree on the downlink flow of the base station and the historical downlink flow parameter of the base station, namely, the future flow of the base station is predicted, so that the predicted future flow of the base station is more accurate.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
Other features, objects, and advantages of the present disclosure will become more apparent from the following detailed description of non-limiting embodiments when taken in conjunction with the accompanying drawings. In the drawings:
FIG. 1 illustrates a flow diagram of a flow prediction method according to an embodiment of the present disclosure;
FIG. 2 illustrates a flow chart of a traffic prediction method according to an embodiment of the present disclosure;
FIG. 3 shows a block diagram of a flow prediction device according to an embodiment of the present disclosure;
FIG. 4 shows a block diagram of an electronic device according to an embodiment of the present disclosure;
FIG. 5 is a schematic block diagram of a computer system suitable for implementing a flow prediction method according to an embodiment of the present disclosure.
Detailed Description
Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. Also, for the sake of clarity, parts not relevant to the description of the exemplary embodiments are omitted in the drawings.
In the present disclosure, it is to be understood that terms such as "including" or "having," etc., are intended to indicate the presence of the disclosed features, numerals, steps, actions, components, parts, or combinations thereof in the specification, and are not intended to preclude the possibility that one or more other features, numerals, steps, actions, components, parts, or combinations thereof are present or added.
It should also be noted that the embodiments and features of the embodiments in the present disclosure may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
As mentioned above, with the development of scientific technology and network technology, more and more users are beginning to use mobile terminal devices (e.g. smart phones, tablet computers, etc.) capable of accessing wireless networks. In general, a base station deployed by a telecommunications carrier is used for providing a network access service for a mobile terminal device used by a user, and the base station belongs to a wireless network system deployed by the telecommunications carrier.
In actual use, due to factors such as changes in user behaviors and upgrading of mobile terminal devices, telecommunication operators need to correspondingly adjust base stations (for example, increase or decrease the number of base stations, change devices of base stations, etc.) according to changes in traffic of the base stations so as to adapt to different traffic demands. It should be noted that this adjustment is a specific usage scenario that requires a comprehensive consideration of future traffic of the base station, because if the relevant base station is adjusted again when the traffic flow exceeds the base station load, it is already late. Therefore, the future traffic of the base station needs to be predicted before the base station is adjusted accordingly.
In the related art, when predicting the future traffic of the base station, the historical traffic parameters of the base station can be obtained, and the future traffic of the base station can be predicted according to the historical traffic parameters of the base station. However, in practice, the inventors of the embodiments of the present application have found that this solution cannot accurately predict the future traffic of the base station.
In view of the above drawbacks, in this embodiment, a traffic prediction method is provided, and in this embodiment, a historical downlink traffic parameter of a target base station and at least one type of historical signaling parameter of the target base station are obtained, and a correlation coefficient between the historical signaling parameter and the historical downlink traffic parameter is obtained, where the correlation coefficient is used to indicate a correlation between the signaling parameter and the historical downlink traffic parameter, and the stronger the correlation between the signaling parameter and the historical downlink traffic parameter is, the higher the influence degree of the signaling parameter of the base station on the downlink traffic of the base station is. And determining at least one type of target signaling parameters of which the correlation coefficient with the historical downlink traffic parameters is greater than a first correlation coefficient threshold value from the at least one type of historical signaling parameters, wherein the correlation coefficient between the target signaling parameters and the historical downlink traffic parameters is greater than the first correlation coefficient threshold value, namely the target signaling parameters are parameters with higher influence degree on the downlink traffic of the base station in the historical signaling parameters. And then, determining a target traffic analysis model obtained by pre-training according to the target signaling parameter and the correlation coefficient of the target signaling parameter and the historical downlink traffic parameter, so that the target traffic analysis model can be ensured to analyze the future traffic of the target base station based on the target signaling parameter and the historical downlink traffic parameter. And then, inputting the target signaling parameter and the historical downlink flow parameter as input into a target flow analysis model to obtain a predicted flow value of the target base station. Therefore, the technical scheme of the application can obtain the predicted flow value of the target base station based on the signaling parameter with higher influence degree on the downlink flow of the base station and the historical downlink flow parameter of the base station, namely, the future flow of the base station is predicted, so that the predicted future flow of the base station is more accurate.
Fig. 1 shows a flow chart of a flow prediction method according to an embodiment of the present disclosure, as shown in fig. 1, the flow prediction method includes the following steps S101-S104:
in step S101, historical downlink traffic parameters of the target base station and at least one type of historical signaling parameters of the target base station are obtained.
In step S102, a correlation coefficient between the historical signaling parameter and the historical downlink traffic parameter is obtained, and at least one type of target signaling parameter is determined in the at least one type of historical signaling parameter, where the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than a first correlation coefficient threshold.
In step S103, a pre-trained target traffic analysis model is determined according to the target signaling parameter and the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter.
In step S104, the target signaling parameter and the historical downlink traffic parameter are input into the target traffic analysis model to obtain the predicted traffic value of the target base station.
In an embodiment of the present disclosure, the target base station is a base station that needs to predict its future traffic.
In an embodiment of the present disclosure, the historical downlink traffic parameter refers to a parameter that is acquired within a preset historical time period and used for indicating downlink traffic of a target base station. The historical signaling parameters of the target base station refer to signaling parameters of the target base station acquired in the same preset historical time period, for example, the historical signaling parameters of the target base station may include connection type parameters, channel quality type parameters, service type parameters, and modulation mode parameters. The connection type parameters are parameters for reflecting the number of times of establishing user connection, and include the number of times of connecting Evolved Radio Access Bearer (E-RAB), the number of times of connecting Radio Resource Control (RRC), and the number of times of paging response. The Channel Quality class parameter is a parameter for reflecting a distance between a user and a base station, and includes a Channel Quality Indicator (CQI) value. The service type parameter is a parameter for reflecting a service type of the user, and the service type parameter includes a quality of service Class Identifier (QCI, qoS Class Identifier) value. The Modulation scheme parameters are parameters for reflecting a transmission Modulation scheme, and include a Quadrature Phase Shift Keying (QPSK) parameter, a Quadrature Amplitude Modulation (16 QAM) parameter, and a Quadrature Amplitude Modulation (64 QAM) parameter.
In an embodiment of the present disclosure, a correlation coefficient between a historical signaling parameter and a historical downlink traffic parameter is used to measure an influence degree of the historical signaling parameter on downlink traffic of a base station, where the larger an absolute value of the correlation coefficient is, the stronger a correlation between the historical signaling parameter and the historical downlink traffic parameter is, the stronger an interconnection exists between the historical signaling parameter and the historical downlink traffic parameter, and the larger an influence degree of the historical signaling parameter on downlink traffic of the base station is. For example, r may be calculated by the following pearson correlation coefficient calculation formula, and a correlation coefficient | r | between the historical signaling parameter and the historical downlink traffic parameter is obtained according to the calculated r:
wherein X i For the ith entry in the historical downstream traffic parameter input sequence,
average value of input sequence for historical downlink flow parameter, Y i For the ith entry in the historical signaling parameter input sequence,
the value range of r is (-1, + 1) for the average value of the historical signaling parameter input sequence, when r is greater than 0, the historical signaling parameter is positively correlated with the historical downlink traffic parameter, when r is less than or equal to 0, the historical signaling parameter is negatively correlated with the historical downlink traffic parameter, the closer to 1, the stronger the correlation is, | r | is, and the first correlation coefficient threshold value can be 0.75.
In the foregoing embodiment, a correlation coefficient between a historical downlink traffic parameter and the historical downlink traffic parameter is obtained by obtaining the historical downlink traffic parameter of a target base station and at least one type of historical signaling parameter of the target base station, where the correlation coefficient is used to indicate a correlation between the signaling parameter and the historical downlink traffic parameter, and the stronger the correlation between the signaling parameter and the historical downlink traffic parameter, the higher the influence degree of the signaling parameter of the base station on the downlink traffic of the base station is. And determining at least one type of target signaling parameters of which the correlation coefficient with the historical downlink traffic parameters is greater than a first correlation coefficient threshold value from the at least one type of historical signaling parameters, wherein the correlation coefficient between the target signaling parameters and the historical downlink traffic parameters is greater than the first correlation coefficient threshold value, namely the target signaling parameters are parameters with higher influence degree on the downlink traffic of the base station in the historical signaling parameters. And then, determining a target traffic analysis model obtained by pre-training according to the target signaling parameter and the correlation coefficient of the target signaling parameter and the historical downlink traffic parameter, so that the target traffic analysis model can analyze the future traffic of the target base station based on the target signaling parameter and the historical downlink traffic parameter. And then, inputting the target signaling parameter and the historical downlink flow parameter as input into a target flow analysis model to obtain a predicted flow value of the target base station. Therefore, the technical scheme of the application can obtain the predicted flow value of the target base station based on the signaling parameter with higher influence degree on the downlink flow of the base station and the historical downlink flow parameter of the base station, namely, the future flow of the base station is predicted, so that the predicted future flow of the base station is more accurate.
In one embodiment of the present disclosure, the target traffic analysis model is composed of a target convolutional neural network model and a target long-short term memory network model.
Step S104, inputting the target signaling parameter and the historical downlink traffic parameter as inputs into the target traffic analysis model to obtain the predicted traffic value of the target base station, which may include the following steps:
and inputting the target signaling parameter and the historical downlink flow parameter into the target convolutional neural network model to obtain the target signaling parameter and the target characteristics of the historical downlink flow parameter.
And inputting the target characteristics as input into the target long-short term memory network model to obtain a predicted flow value.
The target convolutional neural network model is used for extracting respective characteristics of each type of target signaling parameter and the historical downlink traffic parameter, for example, the target convolutional neural network model may be a multi-head convolutional neural network model, a convolutional layer corresponding to each type of target signaling parameter and the historical downlink traffic parameter is provided in the multi-head convolutional neural network model to obtain the target signaling parameter characteristics of each type of target signaling parameter and the historical downlink traffic characteristics of the historical downlink traffic parameter, and the target characteristics are formed by splicing the target signaling parameter characteristics and the historical downlink traffic characteristics. For another example, the target convolutional neural network model may be composed of a plurality of convolutional neural network models corresponding to each type of target signaling parameter one to one, and a convolutional neural network model corresponding to the historical downlink traffic parameter, where the plurality of convolutional neural network models corresponding to each type of target signaling parameter one to one are used to obtain the target signaling characteristics of the corresponding target signaling parameter, and the convolutional neural network model corresponding to the historical downlink traffic parameter is used to obtain the historical downlink traffic characteristics of the historical downlink traffic parameter, where the target characteristics are formed by splicing the target signaling parameter characteristics and the historical downlink traffic characteristics.
For example, the target signaling parameter and the historical downlink traffic parameter input into the target convolutional neural network model may be the target signaling parameter and the historical downlink traffic parameter within a certain time period, so the target feature output by the target convolutional neural network model is used to represent the target signaling parameter and the feature of the historical downlink traffic parameter within the certain time period, and the target feature is input into the target long-short term memory network model, so that the time correlation of the target signaling parameter and the historical downlink traffic parameter within the certain time period may be captured.
The target signaling parameters and the historical downlink flow parameters are used as input, the target convolutional neural network model is input to obtain target characteristics of the target signaling parameters and the historical downlink flow parameters, the target characteristics are used as input, the target long-term and short-term memory network model is input to obtain predicted flow values, the respective characteristics of each type of target signaling parameters and the historical downlink flow parameters can be extracted to the maximum extent, the time correlation of each type of target signaling parameters and the historical downlink flow parameters on a time sequence can be captured, and the accuracy of the obtained predicted flow values is improved.
In an embodiment of the present disclosure, a correlation coefficient between a target signaling parameter and a historical downlink traffic parameter is positively correlated with a convolutional layer convolutional kernel size corresponding to the target signaling parameter in a target convolutional neural network model.
For example, if the target signaling parameter includes a modulation mode parameter, it can be known that the correlation coefficient is 0.9 by obtaining the correlation coefficient between the modulation mode parameter and the historical downlink traffic parameter, and the size of the convolutional layer convolutional kernel corresponding to the modulation mode parameter in the determined target convolutional neural network model is 6; by obtaining the correlation coefficient of the modulation mode parameter and the historical downlink flow parameter, it can be known that the correlation coefficient is 0.8, and the size of the convolutional layer convolutional kernel corresponding to the modulation mode parameter in the determined target convolutional neural network model is 3.
When the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is larger, the target signaling parameter has more redundant information, the convolutional layer convolutional kernel corresponding to the target signaling parameter is larger, and the characteristics of the target signaling parameter are extracted through the convolutional layer convolutional kernel, so that the influence of the redundant information can be reduced, and the accuracy of the obtained predicted traffic value is improved.
In an embodiment of the present disclosure, a correlation coefficient between a target signaling parameter and a historical downlink traffic parameter is inversely correlated with a number of convolutional layer convolutional kernels corresponding to the target signaling parameter in a target convolutional neural network model.
For example, if the target signaling parameter includes a modulation mode parameter, it can be known that the correlation coefficient is 0.9 by obtaining the correlation coefficient between the modulation mode parameter and the historical downlink traffic parameter, and the number of convolutional layers corresponding to the modulation mode parameter in the determined target convolutional neural network model is 32; by obtaining the correlation coefficient of the modulation mode parameter and the historical downlink flow parameter, it can be known that the correlation coefficient is 0.8, and the number of convolutional layer convolutional kernels corresponding to the modulation mode parameter in the determined target convolutional neural network model is 64.
When the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is larger, the target signaling parameter has more redundant information, the number of convolutional layer convolutional kernels corresponding to the target signaling parameter is smaller, and the characteristics of the target signaling parameter are extracted through the convolutional layer convolutional kernels, so that the influence of the redundant information can be reduced, and the accuracy of the acquired predicted traffic value is improved.
In an embodiment of the present disclosure, when a correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than or equal to the second correlation coefficient threshold, the target convolutional neural network model includes a pooling layer corresponding to the target signaling parameter.
And when the correlation coefficient of the target signaling parameter and the historical downlink flow parameter is smaller than the second correlation coefficient threshold value, the target convolutional neural network model does not comprise the pooling layer corresponding to the target signaling parameter.
For example, if the target signaling parameter includes a modulation mode parameter and the second threshold is 0.85, it can be known that the correlation coefficient is 0.8 by obtaining the correlation coefficient between the modulation mode parameter and the historical downlink traffic parameter, and it can be known that the correlation coefficient is smaller than the second threshold, so that the determined target convolutional neural network model does not include the pooling layer corresponding to the modulation mode parameter.
When the correlation coefficient of the target signaling parameter and the historical downlink flow parameter is small, the redundant information in the target signaling parameter is less, the information in the target signaling parameter can be prevented from being wasted by enabling the determined target convolutional neural network model not to include the pooling layer corresponding to the modulation mode parameter, the effective characteristics in the target signaling parameter can be extracted to the maximum extent, and therefore the accuracy degree of the obtained predicted flow value is improved.
In an embodiment of the present disclosure, the target traffic analysis model does not include a pooling layer corresponding to the historical downlink traffic parameters.
Because the historical downlink flow parameters do not include redundant information, the target flow analysis model does not include a pooling layer corresponding to the historical downlink flow parameters, so that the waste of information in the historical downlink flow parameters can be avoided, the effective characteristics in the historical downlink flow parameters can be maximally extracted, and the accuracy of the acquired predicted flow values is improved.
In an embodiment of the present disclosure, fig. 2 shows a flowchart of a flow prediction method according to an embodiment of the present disclosure, and as shown in fig. 2, the flow prediction method further includes the following steps S105 to S107:
in step S105, a sampling base station data set is obtained, where the sampling base station data set includes sampling downlink traffic parameters of a sampling base station and at least one type of sampling signaling parameters of the sampling base station, and a correlation coefficient between the sampling signaling parameters and the sampling downlink traffic parameters is greater than a first correlation coefficient threshold.
In step S106, an initial traffic analysis model is determined according to the sampled signaling parameter and the correlation coefficient between the sampled signaling parameter and the sampled downlink traffic parameter.
In step S107, the sampling signaling parameter of the first time period in the sampling base station data set and the sampling downlink traffic parameter of the first time period are used as inputs, and the sampling downlink traffic parameter of the second time period in the sampling base station data set is used as an output to train the initial downlink traffic analysis model, so as to obtain a target traffic analysis model, where the first time period is earlier than the second time period.
The sampled downlink traffic parameter of the first time period refers to a parameter acquired in the first time period and used for indicating the downlink traffic of the sampled base station. The sampling signaling parameter of the first time period refers to a signaling parameter of the sampling base station acquired in the first time period, for example, the sampling signaling parameter of the sampling base station may include a connection type parameter, a channel quality type parameter, a service type parameter, and a modulation mode parameter. The sampled downlink traffic parameter of the second time period refers to a parameter acquired in the second time period and used for indicating the downlink traffic of the sampled base station.
By acquiring a sampling base station data set and determining an initial flow analysis model according to the sampling signaling parameters and the correlation coefficients of the sampling signaling parameters and the sampling downlink flow parameters, the initial flow analysis model can be ensured to be used for analyzing the future flow of a target base station after being trained. And then, taking the sampling signaling parameters of the first time period earlier than the second time period in the sampling base station data set and the sampling downlink flow parameters of the first time period as inputs, and taking the sampling downlink flow parameters of the second time period in the sampling base station data set as outputs to train the initial downlink flow analysis model to obtain a target flow analysis model, so that the target flow analysis model can predict the future flow of the base station according to the historical signaling parameters and the downlink flow parameters, and the predicted future flow of the base station is more accurate.
In an embodiment of the present disclosure, the initial traffic analysis model is composed of an initial convolutional neural network model and an initial long-short term memory network model, and in step S107, the sampling signaling parameter of a first time period and the sampling downlink traffic parameter of the first time period in the sampling base station data set are used as inputs, and the sampling downlink traffic parameter of a second time period in the sampling base station data set is used as an output to train the initial downlink traffic analysis model, so as to obtain the target traffic analysis model, which may include the following steps:
a plurality of first time periods with the time length of 24 hours and a second time period corresponding to each first time period are determined by taking 1 hour as a time interval.
And training an initial downlink traffic analysis model by taking the sampling signaling parameters of a first time period in the sampling base station data set and the sampling downlink traffic parameters of the first time period as the input of an initial convolutional neural network model, taking the output of the initial convolutional neural network model as the input of an initial long-short term memory network model, and taking the sampling downlink traffic parameters of a second time period corresponding to the first time period in the sampling base station data set as the output to obtain a target traffic analysis model, wherein the target traffic analysis model consists of a target convolutional neural network model and a target long-short term memory network model.
The traffic variation of the base station is strong periodicity expressed on a scale by taking 'day' as a unit, so that a plurality of first time periods with the time length of 24 hours are determined by taking 1 hour as a time interval, sampling signaling parameters of the first time periods and sampling downlink traffic parameters of the first time periods in a sampling base station data set are used as input of an initial convolutional neural network model, output of the initial convolutional neural network model is used as input of an initial long-short term memory network model, and sampling downlink traffic parameters of a second time period corresponding to the first time periods in the sampling base station data set are used as output to train an initial downlink traffic analysis model, so that the trained initial convolutional neural network model, namely a target convolutional neural network model can extract the characteristics of the sampling signaling parameters and the sampling downlink traffic parameters within 24 hours, the trained initial long-short term memory network model, namely the target long-short term memory network model can acquire the time correlation of the sampling signaling parameters and the characteristics of the sampling downlink traffic parameters on a time sequence, and accordingly the accurate prediction of the future traffic of the base station can be realized based on the sampling signaling parameters and the characteristics of the sampling downlink traffic parameters within 24 hours, and the accurate prediction of the future traffic of the base station can be realized.
In an embodiment of the present disclosure, the sampling base station includes the target base station and a similar base station of the target base station.
In step S107, taking the sampling signaling parameter of the first time period and the sampling downlink traffic parameter of the first time period in the sampling base station data set as inputs, and taking the sampling downlink traffic parameter of the second time period in the sampling base station data set as an output to train the initial downlink traffic analysis model, so as to obtain the target traffic analysis model, which may include the following steps:
and taking the sampling signaling parameters of the similar base stations in the sampling base station data set in the first time period and the sampling downlink flow parameters of the similar base stations in the first time period as input, and taking the sampling downlink flow parameters of the similar base stations in the sampling base station data set in the second time period as output to train the initial downlink flow analysis model, so as to obtain a pre-training downlink flow analysis model.
Freezing a convolutional layer in the pre-trained downlink traffic analysis model.
And taking the sampling signaling parameter of the target base station in the sampling base station data set in the first time period and the sampling downlink flow parameter of the target base station in the first time period as input, and taking the sampling downlink flow parameter of the target base station in the sampling base station data set in the second time period as output to train the pre-training downlink flow analysis model after freezing the convolutional layer, so as to obtain the target flow analysis model.
The initial downlink traffic analysis model is trained by taking the sampling signaling parameter of the first time period and the sampling downlink traffic parameter of the first time period in the sampling base station data set as input and taking the sampling downlink traffic parameter of the second time period in the sampling base station data set as output, so that the pre-trained downlink traffic analysis model can learn the common information of traffic change. And then freezing a convolution layer in the pre-training downlink traffic analysis model to enable the pre-training downlink traffic analysis model to store the learned common information of traffic variation, taking the sampling signaling parameter of the target base station in the sampling base station data set in the first time period and the sampling downlink traffic parameter of the target base station in the first time period as input, taking the sampling downlink traffic parameter of the target base station in the sampling base station data set in the second time period as output to train the pre-training downlink traffic analysis model after freezing the convolution layer, and enabling the target traffic analysis model to learn the characteristic information of the traffic variation of the target base station on the basis of the learned common information of the traffic variation, so that the predicted traffic value of the target base station, namely the predicted future traffic of the base station, obtained on the basis of the target traffic analysis model is more accurate.
The following are embodiments of the disclosed apparatus that may be used to perform embodiments of the disclosed methods.
Fig. 3 shows a block diagram of a flow prediction apparatus according to an embodiment of the present disclosure, which may be implemented as part or all of an electronic device by software, hardware, or a combination of the two. As shown in fig. 3, the flow rate prediction apparatus includes:
a base station parameter obtaining module 201, configured to obtain a historical downlink traffic parameter of a target base station and at least one type of historical signaling parameter of the target base station.
The target parameter determining module 202 is configured to obtain a correlation coefficient between the historical signaling parameter and the historical downlink traffic parameter, and determine at least one type of target signaling parameter from the at least one type of historical signaling parameter, where the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than a first correlation coefficient threshold.
And the analysis model determination module 203 is configured to determine a pre-trained target traffic analysis model according to the target signaling parameter and the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter.
The traffic parameter prediction module 204 is configured to input the target signaling parameter and the historical downlink traffic parameter into the target traffic analysis model to obtain a predicted traffic value of the target base station.
In the above embodiment, a correlation coefficient between a historical downlink traffic parameter and the historical downlink traffic parameter is obtained by obtaining the historical downlink traffic parameter of a target base station and at least one type of historical signaling parameter of the target base station, where the correlation coefficient is used to indicate a correlation between the signaling parameter and the historical downlink traffic parameter, and the stronger the correlation between the signaling parameter and the historical downlink traffic parameter, the higher the influence degree of the signaling parameter of the base station on the downlink traffic of the base station is. And determining at least one type of target signaling parameters of which the correlation coefficient with the historical downlink traffic parameters is greater than a first correlation coefficient threshold value from the at least one type of historical signaling parameters, wherein the correlation coefficient between the target signaling parameters and the historical downlink traffic parameters is greater than the first correlation coefficient threshold value, namely the target signaling parameters are parameters with higher influence degree on the downlink traffic of the base station in the historical signaling parameters. And then, determining a target traffic analysis model obtained by pre-training according to the target signaling parameter and the correlation coefficient of the target signaling parameter and the historical downlink traffic parameter, so that the target traffic analysis model can be ensured to analyze the future traffic of the target base station based on the target signaling parameter and the historical downlink traffic parameter. And then, inputting the target signaling parameter and the historical downlink flow parameter as input into a target flow analysis model to obtain a predicted flow value of the target base station. Therefore, the technical scheme of the application can obtain the predicted flow value of the target base station based on the signaling parameter with higher influence degree on the downlink flow of the base station and the historical downlink flow parameter of the base station, namely, the future flow of the base station is predicted, so that the predicted future flow of the base station is more accurate.
The present disclosure also discloses an electronic device, fig. 4 shows a block diagram of an electronic device according to an embodiment of the present disclosure, as shown in fig. 4, the electronic device 300 includes a memory 301 and a processor 302; wherein,
the memory 301 is used to store one or more computer instructions, which are executed by the processor 302 to implement the above-described method steps.
FIG. 5 is a schematic block diagram of a computer system suitable for implementing a flow prediction method according to an embodiment of the present disclosure.
As shown in fig. 5, the computer system 400 includes a processing unit 401 that can execute various processes in the above-described embodiments according to a program stored in a Read Only Memory (ROM) 402 or a program loaded from a storage section 408 into a Random Access Memory (RAM) 403. In the RAM403, various programs and data necessary for the operation of the system 400 are also stored. The processing unit 401, the ROM402, and the RAM403 are connected to each other via a bus 404. An input/output (I/O) interface 405 is also connected to bus 404.
The following components are connected to the I/O interface 405: an input portion 406 including a keyboard, a mouse, and the like; an output section 407 including a display device such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage section 408 including a hard disk and the like; and a communication section 409 including a network interface card such as a LAN card, a modem, or the like. The communication section 409 performs communication processing via a network such as the internet. A driver 410 is also connected to the I/O interface 405 as needed. A removable medium 411 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 410 as necessary, so that a computer program read out therefrom is mounted into the storage section 408 as necessary. The processing unit 401 may be implemented as a CPU, a GPU, a TPU, an FPGA, an NPU, or other processing units.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowcharts or block diagrams may represent a module, a program segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units or modules described in the embodiments of the present disclosure may be implemented by software or hardware. The units or modules described may also be provided in a processor, and the names of the units or modules do not in some cases constitute a limitation on the units or modules themselves.
As another aspect, the present disclosure also provides a computer-readable storage medium, which may be the computer-readable storage medium included in the apparatus in the above-described embodiment; or it may be a separate computer readable storage medium not incorporated into the device. The computer readable storage medium stores one or more programs for use by one or more processors in performing the methods described in the present disclosure.
The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention in the present disclosure is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is made without departing from the inventive concept. For example, the above features and (but not limited to) the features disclosed in this disclosure having similar functions are replaced with each other to form the technical solution.
Claims (9)
1. A method of traffic prediction, comprising:
acquiring historical downlink flow parameters of a target base station and at least one type of historical signaling parameters of the target base station;
obtaining a correlation coefficient between the historical signaling parameter and the historical downlink traffic parameter, and determining at least one type of target signaling parameter in the at least one type of historical signaling parameter, wherein the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than a first correlation coefficient threshold value;
determining a target traffic analysis model obtained by pre-training according to the target signaling parameter and the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter;
inputting the target signaling parameter and the historical downlink flow parameter into the target flow analysis model to obtain a predicted flow value of the target base station;
the method further comprises the following steps:
acquiring a sampling base station data set, wherein the sampling base station data set comprises sampling downlink flow parameters of a sampling base station and at least one type of sampling signaling parameters of the sampling base station, and the correlation coefficient of the sampling signaling parameters and the sampling downlink flow parameters is greater than the first correlation coefficient threshold;
determining an initial flow analysis model according to the sampling signaling parameters and the correlation coefficients of the sampling signaling parameters and the sampling downlink flow parameters;
and taking the sampling signaling parameters of a first time period and the sampling downlink flow parameters of the first time period in the sampling base station data set as inputs, and taking the sampling downlink flow parameters of a second time period in the sampling base station data set as outputs to train an initial downlink flow analysis model to obtain the target flow analysis model, wherein the first time period is earlier than the second time period.
2. The traffic prediction method of claim 1, wherein the target traffic analysis model is composed of a target convolutional neural network model and a target long-short term memory network model;
the step of inputting the target signaling parameter and the historical downlink traffic parameter into the target traffic analysis model to obtain the predicted traffic value of the target base station includes:
inputting the target signaling parameter and the historical downlink flow parameter into the target convolutional neural network model to obtain target characteristics of the target signaling parameter and the historical downlink flow parameter;
and inputting the target characteristics as input into the target long-term and short-term memory network model to obtain the predicted flow value.
3. The traffic prediction method according to claim 2, wherein the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter positively correlates with the convolutional layer convolutional kernel size corresponding to the target signaling parameter in the target convolutional neural network model.
4. The traffic prediction method according to claim 2, wherein a correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is inversely correlated with the number of convolutional layer convolutional kernels corresponding to the target signaling parameter in the target convolutional neural network model.
5. The traffic prediction method according to claim 2, wherein when a correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than or equal to a second correlation coefficient threshold, the target convolutional neural network model includes a pooling layer corresponding to the target signaling parameter;
and when the correlation coefficient of the target signaling parameter and the historical downlink traffic parameter is smaller than the second correlation coefficient threshold value, the target convolutional neural network model does not comprise a pooling layer corresponding to the target signaling parameter.
6. The traffic prediction method of claim 1, wherein the initial traffic analysis model consists of an initial convolutional neural network model and an initial long-short term memory network model;
the training the initial downlink traffic analysis model by taking the sampling signaling parameter of the first time period and the sampling downlink traffic parameter of the first time period in the sampling base station data set as inputs and taking the sampling downlink traffic parameter of the second time period in the sampling base station data set as an output to obtain the target traffic analysis model includes:
determining a plurality of first time periods with the time length of 24 hours and a second time period corresponding to each first time period by taking 1 hour as a time interval;
and taking the sampling signaling parameters of the first time period and the sampling downlink flow parameters of the first time period in the sampling base station data set as the input of the initial convolutional neural network model, taking the output of the initial convolutional neural network model as the input of the initial long-short term memory network model, taking the sampling downlink flow parameters of the second time period corresponding to the first time period in the sampling base station data set as the output to train the initial downlink flow analysis model, and obtaining the target flow analysis model, wherein the target flow analysis model is composed of a target convolutional neural network model and a target long-short term memory network model.
7. The traffic prediction method according to claim 1, wherein the sampling base station includes the target base station and a similar base station of the target base station;
the training the initial downlink traffic analysis model by taking the sampling signaling parameter of the first time period and the sampling downlink traffic parameter of the first time period in the sampling base station data set as inputs and taking the sampling downlink traffic parameter of the second time period in the sampling base station data set as an output to obtain the target traffic analysis model includes:
taking a sampling signaling parameter of the similar base station in the sampling base station data set in the first time period and a sampling downlink flow parameter of the similar base station in the first time period as input, taking the sampling downlink flow parameter of the similar base station in the sampling base station data set in the second time period as output to train the initial downlink flow analysis model, and obtaining a pre-training downlink flow analysis model;
freezing a convolution layer in the pre-training downlink flow analysis model;
and taking the sampling signaling parameter of the target base station in the sampling base station data set in the first time period and the sampling downlink flow parameter of the target base station in the first time period as input, and taking the sampling downlink flow parameter of the target base station in the sampling base station data set in the second time period as output to train the pre-training downlink flow analysis model after the frozen convolutional layer, so as to obtain the target flow analysis model.
8. A flow prediction device, comprising:
a base station parameter obtaining module configured to obtain historical downlink traffic parameters of a target base station and at least one type of historical signaling parameters of the target base station;
a target parameter determination module configured to obtain a correlation coefficient between the historical signaling parameter and the historical downlink traffic parameter, and determine at least one type of target signaling parameter from the at least one type of historical signaling parameter, where the correlation coefficient between the target signaling parameter and the historical downlink traffic parameter is greater than a first correlation coefficient threshold;
the analysis model determining module is configured to determine a pre-trained target traffic analysis model according to the target signaling parameter and a correlation coefficient between the target signaling parameter and the historical downlink traffic parameter;
a traffic parameter prediction module configured to input the target signaling parameter and the historical downlink traffic parameter into the target traffic analysis model to obtain a predicted traffic value of the target base station;
the flow parameter prediction module is further configured to:
acquiring a sampling base station data set, wherein the sampling base station data set comprises sampling downlink flow parameters of a sampling base station and at least one type of sampling signaling parameters of the sampling base station, and the correlation coefficient of the sampling signaling parameters and the sampling downlink flow parameters is greater than the first correlation coefficient threshold;
determining an initial flow analysis model according to the sampling signaling parameters and the correlation coefficients of the sampling signaling parameters and the sampling downlink flow parameters;
and taking the sampling signaling parameter of the first time period and the sampling downlink flow parameter of the first time period in the sampling base station data set as input, and taking the sampling downlink flow parameter of the second time period in the sampling base station data set as output to train an initial downlink flow analysis model to obtain the target flow analysis model, wherein the first time period is earlier than the second time period.
9. An electronic device comprising a memory and at least one processor; wherein the memory is to store one or more computer instructions, wherein the one or more computer instructions are to be executed by the at least one processor to implement the method steps of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110610080.2A CN113347659B (en) | 2021-06-01 | 2021-06-01 | Flow prediction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110610080.2A CN113347659B (en) | 2021-06-01 | 2021-06-01 | Flow prediction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113347659A CN113347659A (en) | 2021-09-03 |
| CN113347659B true CN113347659B (en) | 2022-12-23 |
Family
ID=77474349
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110610080.2A Active CN113347659B (en) | 2021-06-01 | 2021-06-01 | Flow prediction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113347659B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109995549A (en) * | 2017-12-29 | 2019-07-09 | ***通信集团陕西有限公司 | A kind of method and device for assessing Flow Value |
| CN110896381A (en) * | 2019-11-25 | 2020-03-20 | 中国科学院深圳先进技术研究院 | Deep neural network-based traffic classification method and system and electronic equipment |
| CN111130839A (en) * | 2019-11-04 | 2020-05-08 | 清华大学 | Flow demand matrix prediction method and system |
| CN112399458A (en) * | 2020-11-16 | 2021-02-23 | 北京弘光浩宇科技有限公司 | Big data analysis method for mobile communication network flow |
| CN112468312A (en) * | 2019-09-09 | 2021-03-09 | 中兴通讯股份有限公司 | Network flow prediction method, communication equipment and storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11711310B2 (en) * | 2019-09-18 | 2023-07-25 | Tweenznet Ltd. | System and method for determining a network performance property in at least one network |
| CN112600728B (en) * | 2020-12-07 | 2022-06-14 | 南昌交通学院 | 5G mobile base station flow prediction analysis system based on big data |
-
2021
- 2021-06-01 CN CN202110610080.2A patent/CN113347659B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109995549A (en) * | 2017-12-29 | 2019-07-09 | ***通信集团陕西有限公司 | A kind of method and device for assessing Flow Value |
| CN112468312A (en) * | 2019-09-09 | 2021-03-09 | 中兴通讯股份有限公司 | Network flow prediction method, communication equipment and storage medium |
| CN111130839A (en) * | 2019-11-04 | 2020-05-08 | 清华大学 | Flow demand matrix prediction method and system |
| CN110896381A (en) * | 2019-11-25 | 2020-03-20 | 中国科学院深圳先进技术研究院 | Deep neural network-based traffic classification method and system and electronic equipment |
| CN112399458A (en) * | 2020-11-16 | 2021-02-23 | 北京弘光浩宇科技有限公司 | Big data analysis method for mobile communication network flow |
Non-Patent Citations (1)
| Title |
|---|
| "高校LTE网络流量预测与扩容研究";王帅;《中国优秀硕士学位论文全文数据库信息科技辑(月刊)》;20210215;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113347659A (en) | 2021-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107333232B (en) | Terminal positioning method and network equipment | |
| CN107171831B (en) | Network deployment method and device | |
| CN111277860B (en) | Method, device and equipment for caching video in mobile edge network and readable medium | |
| CN112887905B (en) | Task unloading method based on periodic resource scheduling in Internet of vehicles | |
| CN113408797A (en) | Method for generating flow-traffic prediction multi-time-sequence model, information sending method and device | |
| CN105830415A (en) | Methods, radio communication device and base station device for managing a media stream | |
| CN111147327A (en) | Network quality evaluation method and device | |
| US20210192217A1 (en) | Method and apparatus for processing video | |
| CN114118748B (en) | Service quality prediction method and device, electronic equipment and storage medium | |
| US20160295373A1 (en) | Mobility determination using likelihood estimation | |
| CN114861790A (en) | Method, system and device for optimizing federal learning compression communication | |
| CN113347659B (en) | Flow prediction method and device | |
| CN112969193B (en) | Interference determination method, device and equipment for wireless network | |
| CN117201310A (en) | Network element capacity expansion method and device, electronic equipment and storage medium | |
| CN110673955A (en) | Method, device, system, terminal and storage medium for optimizing memory | |
| CN115086940A (en) | QoS (quality of service) adjusting method, system and device based on 5G and storage medium | |
| CN108293267A (en) | Dynamic back off time based on channel utilization statistics | |
| CN115442832B (en) | Complaint problem positioning method and device and electronic equipment | |
| JP6640067B2 (en) | Delivery control device, delivery control method and program | |
| CN112511702B (en) | Media frame pushing method, server, electronic equipment and storage medium | |
| CN111898061B (en) | Method, apparatus, electronic device and computer readable medium for searching network | |
| CN112423327B (en) | Capacity prediction method and device and storage medium | |
| CN114793453A (en) | Training method, training device and storage medium | |
| JP2021197701A (en) | Information processing device, communication management system, program, and information processing method | |
| CN112312200A (en) | Video cover generation method and device and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |