CN107171848A - A kind of method for predicting and device - Google Patents

A kind of method for predicting and device Download PDF

Info

Publication number
CN107171848A
CN107171848A CN201710391210.1A CN201710391210A CN107171848A CN 107171848 A CN107171848 A CN 107171848A CN 201710391210 A CN201710391210 A CN 201710391210A CN 107171848 A CN107171848 A CN 107171848A
Authority
CN
China
Prior art keywords
cell
history
target
predicted
tree hierarchy
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.)
Granted
Application number
CN201710391210.1A
Other languages
Chinese (zh)
Other versions
CN107171848B (en
Inventor
张迪
黄琛灿
肖冬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN201710391210.1A priority Critical patent/CN107171848B/en
Publication of CN107171848A publication Critical patent/CN107171848A/en
Application granted granted Critical
Publication of CN107171848B publication Critical patent/CN107171848B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/14—Network analysis or design
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W24/00—Supervisory, monitoring or testing arrangements
    • H04W24/06—Testing, supervising or monitoring using simulated traffic

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The embodiment of the present application discloses a kind of method for predicting, the predicated error of the use flow for effectively reducing cell.The embodiment of the present application method includes：Obtain in presetting range, the history of all cells always uses data on flows；The history of all cells is always used data on flows as input, the total of all cells in the presetting range is predicted using flow to obtain target prediction value using the first forecast model；The use traffic prediction value for obtaining the Target cell is predicted to the use flow of Target cell according to the target prediction value, wherein, the Target cell is any one cell in the presetting range.The embodiment of the present application also discloses a kind of volume forecasting device, the predicated error of the use flow for effectively reducing cell.

Description

A kind of method for predicting and device
Technical field
The application is related to the communications field, more particularly to a kind of method for predicting and device.
Background technology
Radio Network System, refer to that telecom operators, in a certain regional widespread deployment, provide the user data transport service Communication system.Generally, Radio Network System includes two big parts, base station sub-system and network subsystem.Wherein, base Stand subsystem and directly provide network insertion service for the terminal device of nearby users.For a telecom operators, base station Quantity up to up to ten thousand or even hundreds thousand of, be dispersedly deployed in regional, be that region is covered comprehensively.Network subsystem is then Base station is connected by cable, data transmit-receive service is provided for the terminal device of access network.
In the planning and adjustment of Radio Network System, due to the factors such as change, the updating terminal device of the user stream of people, electricity The change that letter operator can use according to network traffics, the increase and decrease that the quantity to base station, cell is adapted to is different to adapt to Traffic demand.And this increase and decrease adjustment is the particular condition in use for needing to consider following possible network traffics, because working as When network traffics are arrived using flood peak, then the base station of correlation is gone to purchase and disposes, it is late.Therefore, in order to adapt to The particular condition in use of following possible network traffics is stated, according to concrete condition, it is necessary to different grain size (such as in preset range Cell portion or whole cell) be predicted in the use flows of different cycles (such as 1 or 6, or 12 months after).
In the prior art there is provided a kind of method being predicted based on packet mode, in presetting range, before this according to The history of cell is grouped using the close principle of flow to cell, when the use flow to some cell is predicted, The history of all cells is entered using flow as reference data to the use flow of certain cell using under the packet of some cell Row prediction.It can be seen that forecast model complexity can be suppressed based on packet mode, but to some cell under the packet Use flow when being predicted, then do not consider other correlations that the cell portion is grouped with other, only considered history makes With this close factor of flow so that when being predicted to cell, the information of all related cells is not fully taken into account, is caused The predicted value error of the use flow of final cell is larger.
The content of the invention
The embodiment of the present application provides a kind of method for predicting, for solving in the prior art, to the use stream of cell Amount predicts the problem of predicted value application condition come is big when being predicted.
In order to solve the above problems, the embodiment of the present application provides following technical scheme：
In a first aspect, the embodiment of the present application provides a kind of method for predicting, first obtain in presetting range, all cells History always use data on flows, the history always refers to some history in presetting range before all cells using data on flows The data constituted using the summation of flow, then the history of all cells to acquire always use data on flows as defeated Enter, the total of all cells in presetting range be predicted using flow to obtain target prediction value using the first forecast model, The use flow of any one cell in all cells in Target cell, i.e. presetting range is entered finally according to target prediction value Row prediction obtains the use traffic prediction value of Target cell.
As can be seen here, for any one cell to be predicted, when being predicted to each cell to be predicted, according to The target prediction value of the use flow of all cells is predicted under the presetting range calculated, it is suppressed that pre- to cell Forecast model complexity during survey, and the target prediction value take into account what all cell predictions in presetting range were obtained, not be Cell is grouped, the number of prediction reference cell is confined to cell portion, the use flow of cell is effectively reduced Predicted value error.
In a kind of possible realization, the use flow of Target cell is predicted according to the target prediction value and obtained The use traffic prediction value of the Target cell, including：According in tree hierarchy structure, the use of all cells under first node The history of all cells always uses data on flows under traffic prediction value and Section Point, using the second forecast model to second section The total of all cells is predicted using flow under point, and Section Point is the downstream site of first node, wherein, work as first node During for root node, the total of all cells uses traffic prediction value to be target prediction value under first node；When the 3rd node is determined Under all cells use traffic prediction value after, it is small according to the use traffic prediction value and target of all cells under the 3rd node The history in area always uses data on flows, is predicted that to obtain target small to the use flow of Target cell using the second forecast model The use traffic prediction value in area, Target cell is the cell under the 3rd node.It can thus be seen that in the present implementation, for tree For top mode in shape hierarchical structure, i.e. root node, it is only necessary to train a model, using all small under root node The history in area predicts the use flow of all cells under the root node using total flow；And in tree hierarchy structure Intermediate node, and bottom layer node, it is only necessary to all cells under this layer, and upper layer node use traffic prediction value as defeated Enter, be predicted, rather than the history of upper layer node uses flow as input, and parameter is effectively compressed, actual measurement Effect more preferably, can make it that the use traffic prediction value of final Target cell is more accurate.
In a kind of possible realization, the cell in presetting range is layered can have to obtain tree hierarchy structure A variety of acquisition patterns, one way in which is first to obtain cellular engineering parameter；The cellular engineering parameter obtained is recycled to determine pre- The network topology structure of cell in the range of putting, finally regard network topology structure as the tree hierarchy knot in the embodiment of the present application Structure.I.e. in the present implementation, it is proposed that a kind of mode for obtaining tree hierarchy structure, the exploitativeness of scheme is improved.
In a kind of possible realization, the cell in presetting range is layered can be with to obtain tree hierarchy structure Obtain in the following manner：First obtain cellular engineering parameter；Each is determined in presetting range using the cellular engineering parameter of acquisition The plan position of cell, i.e., draft a position ginseng again using the cellular engineering parameter of acquisition for each cell in presetting range Number, is defined as intending position in the embodiment of the present application, all cells of presetting range is entered finally according to the plan position of each cell Row is layered to obtain tree-like hierarchical structure.As can be seen here, the mode of another acquisition tree hierarchy structure is proposed in this realization, Improve the diversity of scheme.Exemplary, in a kind of possible realization, each cell is determined according to cellular engineering parameter Intend position, the plan position (x of each cell can be determined according to following formulacell、ycell)：
xcell=xsite*(λ*ptrx)*h*sin(α)*cos(θ)；
ycell=ysite*(λ*ptrx)*h*sin(α)*sin(θ)；
Wherein, xsite、ysiteThe respectively longitude of each cell respective base station, dimension；xcell、ycellRespectively each is small The longitude in area, dimension, λ are preset data, ptrxFor the transmission power of base station, h is the antenna height of each cell, and α, θ are respectively The corresponding Downtilt of each cell, deflection.
In a kind of possible realization, according to the plan position of each cell all cells in presetting range are layered with Hierarchical clustering can be carried out by the plan position to each cell by obtaining tree hierarchy structure, so as to obtain institute in presetting range There is the tree hierarchy structure of cell.
In a kind of possible realization, the plan position of each cell is carried out hierarchical clustering to obtain tree hierarchy knot Structure, can use K-mean cluster (also referred to as K mean cluster) modes the plan position of each cell is carried out hierarchical clustering with Obtain tree hierarchy structure.Certainly, merely just so that K-mean is clustered as an example, in practical application, it can rule of thumb be used His hierarchical clustering mode is obtained, and is not limited here.
In a kind of possible realization, the cell in presetting range is layered according to ticket writing to obtain tree-like layer Level knot, including：Obtain ticket writing；Training sample is obtained according to ticket writing；Training sample is expanded into target root vector；It is right Target root vector carries out hierarchical clustering to obtain tree hierarchy structure.In the present implementation, specifically propose how one kind utilizes Ticket writing obtains the mode of the tree hierarchy structure of cell in presetting range, improves the exploitativeness of scheme.
It is described that training sample is obtained according to the ticket writing in a kind of possible realization, including：Inquire about the ticket Record is obtained in each cell in the preset range, the discharge record information of all terminal devices；The flow is remembered Record information is used as the training sample.
In a kind of possible realization, target root vector is carried out hierarchical clustering to obtain tree hierarchy structure, including： Cluster (also referred to as double focusing class) mode to target root vector progress hierarchical clustering to obtain tree hierarchy using bi-clustering Structure.
In a kind of possible realization, the method for predicting that the embodiment of the present application is provided includes：It is tree-like based on first Hierarchical structure, two tree hierarchy structures and the 3rd tree-like hierarchical structure are predicted to distinguish to the use flow of Target cell The first predicted value, the second predicted value and the 3rd predicted value are obtained, wherein, the first tree hierarchy structure is according to network topology structure The tree hierarchy structure of acquisition, the second tree hierarchy structure is according to the tree hierarchy structure for intending position acquisition, the 3rd tree-like layer Level structure is the tree hierarchy structure obtained according to ticket writing；According to the first predicted value, the second predicted value and the 3rd predicted value Obtain the use traffic prediction value of Target cell.I.e. in the present implementation, when the tree hierarchy acquired in the above-mentioned first aspect of utilization When structure prediction obtains predicted value, it is possible to use algorithm, can to determine the optimal predicted value of the use flow of Target cell To effectively improve prediction accuracy of the embodiment of the present application to the use flow of Target cell.
In a kind of possible realization, Target cell is worth to according to the first predicted value, the second predicted value and the 3rd prediction Use traffic prediction value before, this method also includes：Determine that the history of Target cell in preset period of time uses data on flows Weighted average, then using first predicted value, the second predicted value, the 3rd predicted value and the cellular engineering parameter to be defeated Enter, the weighted average is label, is merged based on random forests algorithm, the use flow for obtaining the Target cell is pre- Measured value.In the present implementation, the use traffic prediction value of Target cell resulting under to different tree form hierarchical structure carries out comprehensive Close and judge to determine the use traffic prediction value of final Target cell, i.e., also to have considered cellular engineering parameter, because, Predicted value of the above-mentioned three kinds of tree hierarchy structures on different network formats, the network of different frequency range be probably it is different, because This, when being merged, it is considered to specific cellular engineering parameter, the use volume forecasting of final Target cell can be caused to imitate Fruit is more preferably.
In a kind of possible realization, determine that the history of Target cell in preset period of time uses the weighted average of data on flows Value, can be carried out in the following manner：N number of history to the Target cell in preset period of time is abnormal using the progress of data on flows point Detection, N is positive integer, and N is more than or equal to 2；By being obtained after abnormality detection in preset period of time, history uses flow number strong point N number of abnormal coefficient；Determine that N number of history uses the corresponding weight in flow number strong point according to N number of abnormal coefficient；Made according to N number of history Determine that history of the Target cell in preset period of time uses the weighted average of flow with the corresponding weight in flow number strong point.
It is described that institute is determined using the corresponding weight in flow number strong point according to N number of history in a kind of possible realization The weighted average that history of the Target cell in preset period of time uses flow is stated, including：N number of history is used into flow number The corresponding weight in strong point is weighted averagely uses data on flows to obtain history of the Target cell in preset period of time Weighted average.
Second aspect, the embodiment of the present application, which provides a kind of volume forecasting device, includes acquisition module, the first prediction module And second prediction module.Wherein, acquisition module, for obtaining in presetting range, the history of all cells always uses flow number According to；First prediction module, for always using the history of all cells data on flows as input, uses the first forecast model pair The total of all cells is predicted to obtain target prediction value using flow in the presetting range that acquisition module is obtained；Second prediction Module, the target prediction value for being predicted according to the first prediction module is predicted to the use flow of Target cell obtains target The use traffic prediction value of cell, wherein, Target cell is any one cell in presetting range.
In the second aspect of the application, the comprising modules of above-mentioned volume forecasting device can also carry out aforementioned first aspect In described in various possible implementations the step of, refer to foregoing in various possible implementations in first aspect Illustrate, specifically no longer repeat herein.
The third aspect, the embodiment of the present application also provides another volume forecasting device, and the volume forecasting device, which has, to be realized The function of the behavior of volume forecasting device in the above method, above-mentioned functions can be realized by hardware, can also be held by hardware The corresponding software of row is realized.Hardware or software include one or more modules corresponding with above-mentioned functions.It is possible in one kind In design, the structure of volume forecasting device includes including memory and processor, and stores on a memory and can be The computer program run on processor, is realized during the computing device computer program in above-mentioned first aspect/first aspect Various possible implementations, are repeated no more here.
Fourth aspect, this application provides a kind of computer-readable recording medium, is stored in computer-readable recording medium There is instruction, when run on a computer so that computer performs above-mentioned each side/described method of each realization, specifically It see described above, be not discussed here.
As can be seen from the above technical solutions, for any one cell to be predicted, to each cell to be predicted When being predicted, it is predicted i.e. according to the target prediction value of the use flow of all cells under the presetting range calculated Can, it is suppressed that forecast model complexity when predicting cell, and the target prediction value take into account it is all small in presetting range Area's prediction is obtained, and is not to be grouped cell, the number of prediction reference cell is confined into cell portion, efficiently reduced The predicted value error of the use flow of cell.
Brief description of the drawings
In order to illustrate more clearly of the technical scheme in the embodiment of the present application, make required in being described below to embodiment Accompanying drawing is briefly described.
Fig. 1 is an a kind of system framework schematic diagram of method for predicting of the embodiment of the present application；
Fig. 2 is a kind of one embodiment schematic flow sheet of method for predicting of the embodiment of the present application；
Fig. 3 is a kind of cellular engineering parameter schematic diagram of the embodiment of the present application；
Fig. 4 is a kind of tree hierarchy structural representation in a kind of method for predicting of the embodiment of the present application；
Fig. 5 is a kind of another embodiment schematic flow sheet of method for predicting of the embodiment of the present application；
Fig. 6 is a kind of volume forecasting device one embodiment structural representation of the embodiment of the present application；
Fig. 7 is a kind of another example structure schematic diagram of volume forecasting device of the embodiment of the present application；
Fig. 8 is a kind of another example structure schematic diagram of volume forecasting device of the embodiment of the present application；
Fig. 9 is a kind of computer equipment one embodiment structural representation of the embodiment of the present application.
Embodiment
The embodiment of the present application provides a kind of method for predicting, the use volume forecasting for effectively reducing cell The predicated error of value.
In order that those skilled in the art more fully understand application scheme, below in conjunction with the embodiment of the present application Technical scheme in accompanying drawing, description the embodiment of the present application.Obviously, described embodiment is only the implementation of the application part Example, rather than whole embodiments.Based on the embodiment in the application, it is every other that those of ordinary skill in the art are obtained Embodiment, should all belong to the scope of the application protection.
It should be noted that term " first " in the description and claims of this application and above-mentioned accompanying drawing, " Two ", the (if present) such as " the 3rd ", " the 4th " be for distinguishing similar object, without for describe specific order or Precedence.It should be appreciated that the data so used can be exchanged in the appropriate case, so that the embodiments described herein can Implemented with the order in addition to the content for illustrating or describing herein.In addition, term " comprising " and and their any change Shape, it is intended that covering is non-exclusive to be included, for example, containing the process of series of steps or unit, method, system, product Or equipment is not necessarily limited to those steps or the unit clearly listed, but may include not list clearly or for these The intrinsic other steps of process, method, product or equipment or unit.
The method for predicting that the embodiment of the present application is provided can apply in various communication systems, exemplary , it can apply to global system for mobile communications (global system for mobile communication, GSM), code Divide multiple access (code division multiple access, CDMA) system, WCDMA (wideband code Division multiple access, WCDMA) system, GPRS (general packet radio Service, GPRS), UMTS (universal mobile telecommunications system, UMTS), time-division Long Term Evolution (time division long term evolution, TD-LTE), frequency division Long Term Evolution (the 5th third-generation mobile communication technology in frequency division long term evolution, FDD-LTE and future In various communication systems such as (5rd-generation mobile communication technology, 5G), tool Body is not limited herein.
First, a kind of system framework of the method for predicting first provided the embodiment of the present application carries out an introduction, Referring to Fig. 1, Fig. 1 be a kind of the embodiment of the present application one system framework schematic diagram of method for predicting, including terminal device and Base station, can realize the transmitting-receiving of data or signaling by air interface between terminal device base station and terminal device.The terminal is set It is standby be referred to alternatively as access terminal, user equipment (user equipment, UE), subscriber unit, subscriber station, movement station, mobile station, The devices such as Distant Station, remote terminal mobile device, are not limited specifically herein.System framework figure shown in Fig. 1 is at this In be merely illustrative, to the embodiment of the present application constitute limit.
Base station refers to the network equipment communicated with terminal device, and the base station can be the base station (base of gsm system Transceiver station, BTS) or WCDMA system in base station (node B, NB), can also refer to LTE systems Evolved base station (evolution node B, eNB or eNodeB) in system, or can refer in the communication systems such as following 5G Base station equipment, also do not limit herein specifically.
Base station can correspond to one or more cells, during terminal device is realized by cell and base station and interacted, Each cell can production and application flow, propose the side that a kind of use flow to cell is predicted in the embodiment of the present application Method, its core concept is, obtains in presetting range, and the history of all cells always uses data on flows；By all cells History always use data on flows as input, all cells in the presetting range are always used using the first forecast model Flow is predicted to obtain target prediction value, and the use flow of Target cell is predicted according to the target prediction value To the use traffic prediction value of the Target cell, wherein, the Target cell is in all cell under the presetting range Any one cell.In the embodiment of the present application, the use flow preset value of Target cell is carried out using target prediction value During prediction, there are various ways to be predicted, embodiment as described below.
Above-mentioned core concept is now directed to, a kind of method for predicting of the embodiment of the present application is described in detail.
Referring to Fig. 2, Fig. 2 is a kind of method for predicting one embodiment schematic flow sheet of the embodiment of the present application, including：
Step 101, the cell in presetting range is layered to obtain tree hierarchy structure.
Wherein, above-mentioned presetting range can be selected according to practical situations, certainly, in the specific implementation, can be with Being drawn a circle to approve according to some prediction experiences includes Target cell to be predicted in above-mentioned preset range, the presetting range.
In addition, in the embodiment of the present application, can there is all cells of a variety of preset layered modes to above-mentioned presetting range It is layered, is not limited here, below by some exemplary provided the embodiment of the present application to small in presetting range The mode that area is layered is described：
In some embodiments of the present application, the cell in the presetting range is layered to obtain tree hierarchy knot Structure, including：Obtain cellular engineering parameter；The network topology of cell in the presetting range is determined according to the cellular engineering parameter Structure；It regard the network topology structure as the tree hierarchy structure.
It should be understood that preserving various cellular engineering parameter informations in cellular engineering parameter list, include the net of cell The specific delamination of network topological structure, wherein, cellular engineering parameter includes region code (area), Location Area Code (location area code, lac), base station identity code (site_id) and cell ID (cell_id), according to above-mentioned Cell work ginseng parameter can correspond to the network topology structure for obtaining cell in presetting range.In order to make it easy to understand, with a reality Example exemplified by illustrate, referring particularly to shown in Fig. 3, Fig. 3 is one schematic diagram of a cellular engineering parameter, it can be seen that It is 3031 and 3,033 two bands of position that the region that region code is ARBSC01, which is divided into Location Area Code, is divided into again for 3031 times ARU001's and ARU002, divide specific cell again under ARU001 and ARU002.Schematic diagram according to Fig. 3, it may be determined that Go out the network topology structure of all cells shown in Fig. 3, it should be appreciated that tree hierarchy relation is presented in the network topology structure of cell, In the application, it may be determined that the network topology structure of cell be used as the tree hierarchy structure in the embodiment of the present application.Need in addition It is noted that Fig. 3 is merely illustrative herein, the application is not limited.
In addition to above-mentioned layered mode, in some embodiments of the present application, the cell in the presetting range is carried out It is layered to obtain tree hierarchy structure, including：Obtain cellular engineering parameter；Determined according to the cellular engineering parameter described preset In the range of each cell plan position；All cells of the presetting range are layered according to the plan position of each cell To obtain the tree-like hierarchical structure.
In this programme, in order that must be more reasonable to the layering result of cell, in some implementations of the embodiment of the present application Example, it is proposed that another layered mode, is each cell definitions one in presetting range with specific reference to cellular engineering parameter Intend position, utilize the tree hierarchy structure for intending position generation cell of each cell of definition.
Wherein, in some embodiments of the present application, it is determined that the plan position of each cell, including：
Plan position (the x of each cell is determined according to below equationcell、ycell)：
xcell=xsite*(λ*ptrx)*h*sin(α)*cos(θ)；
ycell=ysite*(λ*ptrx)*h*sin(α)*sin(θ)；
Wherein, the xsite、ysiteThe longitude of each respectively described cell respective base station, dimension；The xcell、ycell The longitude of each respectively described cell, dimension, the λ are preset data, the ptrxFor the transmission power of the base station, institute The antenna height that h is each cell is stated, described α, θ are respectively the Downtilt of each cell respective base station, side To angle.It should be understood that above-mentioned preset data is empirical data, the empirical data specifically obtained by practical situations is determined, this In do not limit.
By above-mentioned formula, it may be determined that go out the plan position of each cell presetting range Nei.
In some embodiments of the present application, the plan position of each cell described in the basis is in the presetting range Cell is layered to obtain the tree hierarchy structure, including：Hierarchical clustering is carried out to the plan position of each cell To obtain the tree hierarchy structure.When being layered in the plan position to each cell, some levels can be specifically utilized Change clustering algorithm the plan position of each cell of generation is layered to obtain tree hierarchy structure.It should be noted that can There are a variety of hierarchical clustering algorithms to be layered to obtain tree hierarchy structure the plan position of each cell, do not limit here It is fixed.
Exemplary, in some embodiments of the present application, K-mean can be used to cluster mode to each described cell Plan position carry out hierarchical clustering to obtain the tree hierarchy structure.It is exemplary, can be with by taking a concrete instance as an example Follow the steps below cluster：
1st, the number of plies that can first set the tree hierarchy structure of needs is L and packet count is N, wherein, L and N are to be just whole Number, clusters mode first by K-mean and the plan position of cell is divided into N groups；
2nd, each group in the N groups obtained for above-mentioned steps, is continuing with K-mean cluster modes and is divided into N groups, wherein, If the number of cell is less than N in certain group, no longer divided；
3rd, repeat step 2, until reaching that total number of plies L obtains the tree hierarchy structure of cell in presetting range.
In some embodiments of the present application, spectral clustering mode can also be used to carry out the plan position of each cell Layering is not limited specifically, also not repeated to obtain the tree hierarchy structure herein.
Two kinds set forth above are layered to be set according to preset layered mode to the cell in the presetting range The mode of shape hierarchical structure, one of which is the mode that the position relationship based on cell is layered, in the embodiment of the present application, It can also be layered by following layered mode：
All cells in the presetting range are layered according to ticket writing to obtain the tree hierarchy structure.
In some embodiments of the present application, all cells in the presetting range are divided according to ticket writing record Layer to obtain the tree hierarchy knot, including：Obtain ticket writing；Training sample is obtained according to the ticket writing；Will be described Training sample expands into target root vector；The target root vector is carried out hierarchical clustering to obtain the tree hierarchy knot Structure.Illustrate, it should be appreciated that be recorded as shape such as in ticket writing<Time, user, cell, traffic>Tuple, this Apply in embodiment, shape can be polymerized to such as<User, cell, traffic>Record, wherein, the traffic fields after polymerization Represent within certain time, a certain terminal device always uses flow in the history that a certain cell occurs.Here it is possible to polymerize acquisition Presetting range in each cell<User, cell, traffic>Record as training sample, in this application, can be right The training sample of acquisition pre-process obtaining final training sample.Wherein, above-mentioned pretreatment can include following at least one Plant processing mode：1st, delete in the training sample that polymerization is obtained, training sample too low traffic is exemplary, can remove Traffic is less than 1MB (million) training sample, does not limit specifically；2nd, in the training sample obtained to polymerization, to each training The traffic of sample takes the logarithm processing again after adding 1.After training sample is obtained, the training sample is expanded into target root Vector<Cell, traffic_of_user_1, traffic_of_user_2 ..., traffic_of_user_n>, then to obtaining Target root vector carry out hierarchical clustering to obtain the tree hierarchy structure.
It is described described tree-like to obtain to target root vector progress hierarchical clustering in some examples of the application Hierarchical structure, including：Mode is clustered using bi-clustering, and hierarchical clustering is carried out to the target root vector to obtain State tree hierarchy structure.It should be noted that clustered based on target root vector here, it is mainly used for presetting range It is interior, cell and terminal device are divided into some clusters, and the cluster between both, in one-to-one relationship.Illustrate, it is assumed that There is A terminal device clusters, A cell clusters, A terminal devices cluster and A cell clusters are one-to-one relationship, then illustrate A terminal devices Terminal device under cluster can be frequently occurred in the cell under the A cell clusters.
It is further to note that in the embodiment of the present application, can have a variety of hierarchical clustering algorithms to target root to Amount be layered to obtain tree hierarchy structure, do not limit here, for example, it is also possible to using spectral clustering mode to target root to Amount is layered to obtain tree hierarchy structure.
Step 102, determine in the tree hierarchy structure, the history of all cells always uses data on flows under root node.
In the embodiment of the present application, the cell in presetting range is carried out according to the layered mode described by step 101 Layering is obtained after tree hierarchy structure, it may be determined that in the tree hierarchy structure, the history of all cells is always used under root node Flow.It is appreciated that determining in the tree hierarchy structure, the history of all cells always uses data on flows under root node, actual Just it is to determine that the history of all cells in presetting range always uses data on flows.
The history for needing exist for all cells under explanation, root node as described herein is always referred to using data on flows In previous certain section of preset time, the history of all cells uses flow aggregate data under root node.Exemplary, when this is default Between can be or before 3 months, or before 6 months, or the year before, not limit herein specifically before one month.In practical application In, it can determine to need the history of all cells under the root node needed for obtaining to use data on flows according to demand.
Step 103, data on flows always used according to the history of all cells under the root node, use the first forecast model The total of all cells under the root node is predicted using flow to obtain target prediction value.
After obtaining and always using flow according to the history of all cells under the root node, the first prediction mould can be used Type is predicted to obtain target prediction value to the total of all cells under the root node using flow.
Wherein, the first prediction module can be that time series models are predicted, and common time series models have to return certainly Return model (autoregressive, AR) and Vector Autoression Models (vector autoregressive, VAR), by above-mentioned Time series models can speculate future with the correlation between numerical value before and after hunting time sequence from the past, so as to The flow service condition in future is deduced according to flow service condition before.Therefore, in the embodiment of the present application, it can use Above-mentioned forecast model, which is predicted, obtains target prediction value.Certainly, except use time series model, there can also be other return Return model, do not limit herein specifically.
Step 104, the use flow of Target cell is predicted according to the target prediction value obtains Target cell Using traffic prediction value, wherein, the Target cell is any one cell in all cells under the root node.
Wherein, in embodiments herein, the use flow of Target cell is carried out according to the target prediction value pre- The use traffic prediction value of Target cell is measured, including：
According in the tree hierarchy structure, the total of all cells uses traffic prediction value and second section under first node The history of all cells always uses data on flows under point, and all cells under Section Point are always used using the second forecast model Flow is predicted, and the Section Point is the downstream site of the first node, needs exist for explanation, mentioned here First node, refers to all superior nodes on Section Point, including the every first nodes being connected with Section Point, for the ease of Understand, the relation to first node and Section Point by taking a tree hierarchy structure as an example is illustrated here, refers to Fig. 4 institutes Show.Wherein, when the first node is the root node, all cells always uses traffic prediction value under the first node For the target prediction value；When after always using traffic prediction value of all cells under the 3rd node is determined, according to the described 3rd All cells under each node in total superior node using traffic prediction value and the 3rd node of all cells under node It is total always use data on flows using traffic prediction value and the history of the Target cell, use second forecast model pair The use flow of the Target cell is predicted to obtain the use traffic prediction value of Target cell, and the Target cell is institute State the cell under the 3rd node.
Exemplary, the second forecast model can use time series models, for example, it may be AR forecast models, VAR are pre- Model is surveyed, is not limited here.Certainly, except use time series model, there can also be other regression models, specifically herein Do not limit.
For the ease of understanding above-mentioned prediction process, below by taking the tree hierarchy structure obtained according to network topology as an example, with The example of one reality, according to the target prediction value (use after 1 month is set to above-mentioned to the use flow of Target cell Flow) process that is predicted is described：Assuming that existing cell ARU001A, its corresponding base station is ARU001, Location Area Code For 3031, region code is ARBSC01.First since top layer (root node), the history of all cells based on ARBSC01 uses stream Measure (assuming that there are 10000 cells under ARBSC01), the use streams of ARU001A after one month are predicted using time series models Measure target prediction value.After the completion of, based on all cell flows (assuming that there are 500 cells under 3031) under 3031, then add On the ARBSC01 target prediction value that has just obtained, predict 3031 use flow after one month.After the completion of, it is based on The history of all cells under ARU001 uses flow (assuming that having 10), all cells under plus target prediction value, 3031 Use flow predicted value, use time series model prediction ARU001 under the use flow of all cells after one month, After the completion of, the history based on cell ARU001A uses flow, along with all cells under target prediction value, 3031 and ARU001 Use flow after one month predicted value, then reuse use stream of the time series models again to cell ARU001A Amount is predicted, and obtains the use traffic prediction values of cell ARU001A after one month.
As can be seen here, for any one cell to be predicted, when being predicted to each cell to be predicted, according to The target prediction value of the use flow of all cells is predicted under the presetting range calculated, it is suppressed that pre- to cell Forecast model complexity during survey, and the target prediction value take into account what all cell predictions in presetting range were obtained, not be Cell is grouped, the number of prediction reference cell is confined to cell portion, the use flow of cell is effectively reduced Predicted value error.
It should be noted that in the embodiment of the present application, another method for predicting is additionally provided, referring to Fig. 5, should Method includes：
Step 201, the cell in the presetting range is layered to obtain the first tree hierarchy structure, second respectively Tree hierarchy structure and the 3rd tree-like hierarchical structure.
Wherein, above-mentioned first tree hierarchy structure, the second tree hierarchy structure and the 3rd tree-like hierarchical structure are respectively Obtained according to different layered modes.In some embodiments of the present application, above-mentioned first tree hierarchy structure, the second tree-like layer The preset layered mode that level structure and the 3rd tree-like hierarchical structure can describe for prenex embodiment respectively is in presetting range Cell is layered to obtain.The description of previous embodiment can be specifically referred to, no longer repeats to repeat here one by one.It is exemplary , here it is assumed for convenience of description that the first tree hierarchy structure is the network topology structure according to the cell in presetting range Obtained tree hierarchy structure, the second tree hierarchy structure be according to the plan position of each cell in presetting range obtain it is tree-like Hierarchical structure, and the 3rd tree-like hierarchical structure is the tree hierarchy structure obtained according to ticket writing.
In addition, above-mentioned presetting range can be selected according to practical situations, certainly, in the specific implementation, can be with Being drawn a circle to approve according to some prediction experiences includes Target cell to be predicted in above-mentioned preset range, the presetting range.
Step 202, based on the first tree hierarchy structure, the second tree hierarchy structure and the 3rd tree-like hierarchical structure The use flow of Target cell is predicted to obtain the first predicted value, the second predicted value and the 3rd predicted value respectively, it is described Target cell is any one cell in the presetting range.
It is specific in the first tree hierarchy structure, the second tree hierarchy structure and the 3rd tree-like hierarchical structure, often Under kind of tree hierarchy structure, how the process being predicted to the use flow of target can be refering to described above, here no longer Repeat.
203rd, the use volume forecasting of Target cell is worth to according to the first predicted value, the second predicted value and the 3rd prediction Value.
In the embodiment of the present application, respectively according to the above-mentioned predicted value of tree hierarchy structure first, the second predicted value and After this 3 predicted values of 3rd predicted value, this three can be merged, it is pre- with the use flow for obtaining final Target cell Measured value, it is therefore intended that select the minimum predicted value of a predicated error as the use traffic prediction value of Target cell.Here, may be used To there is a variety of determination modes, as long as making it possible to select the minimum predicted value of a predicated error as the use flow of Target cell Predicted value, specific mode is not limited here.
It is pre- according to the use flow that first predicted value, the second predicted value and the 3rd prediction are worth to the Target cell Before measured value, methods described also includes：
Determine that the history of the Target cell in preset period of time uses the weighted average of data on flows；
Wherein, preset period of time can be selected according to actual conditions, exemplary, can be before one month, or 3 Before month, or before 6 months, or the year before, do not limit herein specifically.
Determine that the history of the Target cell in preset period of time uses the weighted average of data on flows, including：
It is described using first predicted value, the second predicted value, the 3rd predicted value and the cellular engineering parameter as input Weighted average is label, is merged based on random forests algorithm, obtains the use traffic prediction value of the Target cell.Its In, exemplary, when being merged based on random forests algorithm, cellular engineering parameter can take frequency point information as input, tool Body is not limited herein.
In some embodiments of the embodiment of the present application, the history for determining the Target cell in preset period of time is used The weighted average of data on flows, including：N number of history to the Target cell in the preset period of time is clicked through using data on flows Row abnormality detection, the N is positive integer, and the N is more than or equal to 2；By obtaining the preset period of time after the abnormality detection Interior, the history uses N number of abnormal coefficient at flow number strong point；Determine that N number of history is used according to N number of abnormal coefficient The corresponding weight in flow number strong point；The Target cell is determined using the corresponding weight in flow number strong point according to N number of history History in preset period of time uses the weighted average of flow.
Wherein, in some embodiments of the present application, by being obtained after the abnormality detection in the preset period of time, each History is referred to every using the size progress abnormality detection at flow number strong point to each history using the abnormal coefficient at flow number strong point Individual history uses the abnormal coefficient e at flow number strong point.Abnormal coefficient e, represents that how different the history has using the size of data point Often (such as remote from average), more abnormal, then abnormal coefficient is also bigger.
In some embodiments of the present application, determined according to each history using the abnormal coefficient at flow number strong point every Individual history can be determined using the weight at flow number strong point by below equation：Weight=exception coefficient e, or weight=exception system Ratio r can be explained in number e*.
Wherein, this can be explained ratio and can obtain in the following manner：The transmission at flow number strong point is used with each history Speed is input, carries out linear regression, correspondence obtains the percentage error f that each history uses flow number strong point, if 1-f is every Individual history uses the explanation ratio r that the scope at flow number strong point is between 0-1.As can be seen here, soluble ratio defines history Changed using the exception and transmission rate at flow number strong point either with or without relation, more have relation, then should more pay close attention to, what correspondence was calculated Weight is then bigger.The weight that each history uses data point, the weighted average to be generated here are calculated by the way that ratio r can be explained Value, the actual weighted average for flow value in preset period of time.W defines the distribution of weight.If be directly averaged, during presetting Exemplified by Duan Weiyi months, then daily weight is 1/30；If directly taking maximum, only some day is 1, and other are all It is 0.Here be the compromise of two kinds of ways by the way that the way of ratio can be explained.
It is described corresponding using flow number strong point according to N number of history in some embodiments of the embodiment of the present application Weight determines that history of the Target cell in preset period of time uses the weighted average of flow, including：By N number of history It is weighted and is averagely used with obtaining history of the Target cell in preset period of time using the corresponding weight in flow number strong point The weighted average of data on flows.
As can be seen from the above technical solutions, it is also proposed that a kind of Target cell according to different tree form hierarchical structure makes The method that best predictor is determined with traffic prediction value, more efficiently reduces the predicated error of the use flow of cell.
It should be noted that for foregoing each method embodiment, in order to be briefly described, therefore it is all expressed as a series of Combination of actions, but those skilled in the art should know, the application is not limited by described sequence of movement because According to the application, some steps can be carried out sequentially or simultaneously using other.Secondly, those skilled in the art should also know Know, embodiment described in this description belongs to exemplary embodiment, involved action and module not necessarily this Shen Please be necessary.
For ease of preferably implementing the scheme described in the above embodiments of the present application, it is also provided below for implementing above-mentioned The relevant apparatus of scheme.
Referring to Fig. 6, the embodiment of the present application provides a kind of volume forecasting device, including acquisition module 101, first is predicted The prediction module 103 of module 102 and second.
Wherein, acquisition module 101, for obtaining in presetting range, the history of all cells always uses data on flows；
First prediction module 102, for always using the history of all cells data on flows as input, uses In the presetting range that one forecast model is obtained to the acquisition module 101 all cells it is total using flow be predicted with Obtain target prediction value；
Second prediction module 103, for the target prediction value predicted according to first prediction module 102 to target The use flow of cell is predicted the use traffic prediction value for obtaining the Target cell, wherein, the Target cell is institute State any one cell in presetting range.
As can be seen from the above technical solutions, for any one cell to be predicted, to each cell to be predicted When being predicted, it is predicted i.e. according to the target prediction value of the use flow of all cells under the presetting range calculated Can, it is suppressed that forecast model complexity when predicting cell, and the target prediction value take into account it is all small in presetting range Area's prediction is obtained, and is not to be grouped cell, the number of prediction reference cell is confined into cell portion, efficiently reduced The predicted value error of the use flow of cell.
Referring to Fig. 7, in some embodiments of the embodiment of the present application, described device also includes：
Hierarchical block 104, for being layered to the cell in the presetting range to obtain tree hierarchy structure；
Second prediction module 103 is used for the target prediction value predicted according to first prediction module 102 to target The use flow of cell is predicted the use traffic prediction value for obtaining the Target cell, including：
Second prediction module 103, for according in the tree hierarchy structure, all cells to be total under first node Data on flows is always used using the history of all cells under traffic prediction value and Section Point, using the second forecast model to institute State the total of all cells under Section Point to be predicted using flow, the Section Point saves for the subordinate of the first node Point, wherein, when the first node is the root node, all cells always uses traffic prediction value under the first node The target prediction value predicted for first prediction module 102；
After the use traffic prediction value of all cells under the 3rd node is determined, for according to institute under the 3rd node There is in total superior node using traffic prediction value and the 3rd node of cell always using for all cells under each node Traffic prediction value and the history of the Target cell always use data on flows, using second forecast model to the target The use flow of cell is predicted the use traffic prediction value for obtaining the Target cell, and the Target cell is the described 3rd Cell under node.
In some embodiments of the present application, the hierarchical block 104 is used to carry out the cell in the presetting range Layering is included with obtaining tree hierarchy structure：
The hierarchical block 104 is used for：
Obtain cellular engineering parameter；
The network topology structure of cell in the presetting range is determined according to the cellular engineering parameter；
It regard the network topology structure as the tree hierarchy structure.
In some embodiments of the present application, the hierarchical block 104 is used to carry out the cell in the presetting range Layering is included with obtaining tree hierarchy structure：
The hierarchical block 104 is used for：
Plan position (the x of each cell presetting range Nei is determined according to below equationcell、ycell)：
xcell=xsite*(λ*ptrx)*h*sin(α)*cos(θ)；
ycell=ysite*(λ*ptrx)*h*sin(α)*sin(θ)；
Wherein, the xsite、ysiteThe longitude of each respectively described cell respective base station, dimension；The xcell、ycell The longitude of each respectively described cell, dimension, the λ are preset data, the ptrxFor the transmission power of the base station, institute The antenna height that h is each cell is stated, described α, θ are respectively the Downtilt of each cell respective base station, side To angle；
All cells in the presetting range are layered according to the plan position of each cell to obtain the tree Shape hierarchical structure.
In some embodiments of the present application, the hierarchical block 104 is used for the plan position pair according to each cell All cells are layered to obtain the tree hierarchy structure in the presetting range, including：
The hierarchical block 104, carries out hierarchical clustering to obtain the tree for the plan position to each cell Shape hierarchical structure.
In some embodiments of the present application, the hierarchical block 104 is used to carry out the plan position of each cell Hierarchical clustering to obtain the tree hierarchy structure, including：
The hierarchical block 104, level is carried out for clustering mode using K-mean to the plan position of each cell Change cluster to obtain the tree hierarchy structure.
In some embodiments of the present application, the hierarchical block 104 is used for according to preset layered mode to described preset In the range of cell be layered and included with obtaining tree hierarchy structure：
The hierarchical block 104 specifically for：
Obtain ticket writing；
Training sample is obtained according to the ticket writing；
The training sample is expanded into target root vector；
The target root vector is carried out hierarchical clustering to obtain the tree hierarchy structure.
In some embodiments of the present application, the hierarchical block 104 is used to obtain training sample according to the ticket writing This, including：
The hierarchical block 104 is used for：
Inquire about in each cell that the ticket writing is obtained in the preset range, the flow note of all terminal devices Record information；
It regard the discharge record information as the training sample.
In some embodiments of the present application, the hierarchical block 104 is used to carry out stratification to the target root vector Cluster to obtain the tree hierarchy structure, including：
The hierarchical block 104, level is carried out for clustering mode using bi-clustering to the target root vector Change cluster to obtain the tree hierarchy structure.
Referring to Fig. 8, in some embodiments of the present application, second prediction module 103 is used for：
Based on the first tree hierarchy structure, two tree hierarchy structures and the 3rd tree-like hierarchical structure to the Target cell Use flow be predicted to obtain the first predicted value, the second predicted value and the 3rd predicted value respectively, wherein, first tree Shape hierarchical structure is the tree hierarchy structure obtained according to the network topology structure, according to the second tree hierarchy structure The tree hierarchy structure that the plan position is obtained, the 3rd tree-like hierarchical structure is according to the tree-like of ticket writing acquisition Hierarchical structure；
Described device also includes：
3rd prediction module 105, for predicted according to second prediction module 103 first predicted value, second Predicted value and the 3rd prediction are worth to the use traffic prediction value of the Target cell.
In some embodiments of the present application, described device also includes determining module 106；
The determining module 106, for the institute predicted in the 3rd prediction module according to second prediction module 103 The first predicted value, the second predicted value and the 3rd prediction is stated to be worth to before the use traffic prediction value of the Target cell, it is determined that The history of the Target cell uses the weighted average of data on flows in preset period of time；
First predicted value that 3rd prediction module 105 is used to being predicted according to second prediction module 103, the Two predicted values and the 3rd prediction are worth to the use traffic prediction value of the Target cell, including：
3rd prediction module 105, for first predicted value, the second predicted value, the 3rd predicted value and institute It is input to state cellular engineering parameter, and the weighted average is label, is merged based on random forests algorithm, obtains the mesh Mark the use traffic prediction value of cell.As can be seen from the above technical solutions, it is also proposed that one kind is according to different tree form level knot The predicted value of structure is merged to determine the volume forecasting device of best predictor, more efficiently effectively reduces cell Use flow predicated error.
In some embodiments of the present application, the determining module 106 is used to determine the Target cell in preset period of time History included using the weighted average of data on flows：
The determining module 106 specifically for：
N number of history to the Target cell in the preset period of time uses data on flows point to carry out abnormality detection, and the N is Positive integer, the N is more than or equal to 2；
By being obtained after the abnormality detection in the preset period of time, the history uses N number of exception at flow number strong point Coefficient；
Determine that N number of history uses the corresponding weight in flow number strong point according to N number of abnormal coefficient；
Determine the Target cell in preset period of time using the corresponding weight in flow number strong point according to N number of history History uses the weighted average of flow.
In some embodiments of the present application, the determining module 106 uses flow for described according to N number of history The corresponding weight of data point determines that history of the Target cell in preset period of time uses the weighted average of flow, including：
N number of history is weighted using the corresponding weight in flow number strong point and averagely existed with obtaining the Target cell History in preset period of time uses the weighted average of data on flows.
It should be noted that the content such as information exchange, implementation procedure between each module/unit of said apparatus, due to Embodiment of the method in the embodiment of the present application is based on same design, its technique effect brought and the application embodiment of the method phase Together, particular content can be found in the narration in the foregoing shown embodiment of the method for the application, and here is omitted.
The embodiment of the present application also provides a kind of computer-readable storage medium, wherein, the computer-readable storage medium has program stored therein, The program can realize the part or all of step described in the above method embodiment when being computer-executed.
Referring to Fig. 9, the embodiment of the present application provides a kind of computer equipment, the computer equipment can be used as above-mentioned flow Prediction meanss, the method for realizing above-mentioned volume forecasting device, computer equipment 300 includes：
Processor 301, memory 302 and storage are on a memory and computer program (its that can run on a processor In, the quantity of the processor 301 in computer equipment 300 can be with one or more, in Fig. 9 by taking a processor as an example).At this In some embodiments of application, the computer equipment can also include COM1 303, wherein, COM1 303, processor 301 and memory 302 can be connected by bus or other manner, do not limit herein specifically, wherein, to pass through bus in Fig. 8 Illustrated exemplified by connection.
Memory 302 can also include read-only storage and random access memory, and to processor 301 provide instruction and Data.The a part of of memory 302 can also include nonvolatile RAM (non-volatile random Access memory, NVRAM).Memory 302 is stored with operating system and operational order, executable module or data knot Structure, either their subset or their superset, wherein, operational order may include various operational orders, each for realizing Plant operation.Operating system may include various system programs, for realizing various basic businesses and handling hardware based task.
The operation of the control computer equipment 300 of processor 301, processor 301 can also be referred to as CPU (central processing unit, CPU).In specific application, each component of computer equipment passes through bus system coupling It is combined, wherein bus system can also include power bus, controlling bus and status signal in addition to including data/address bus Bus etc..But for the sake of clear explanation, various buses are referred to as bus system in figure.
The method that above-mentioned the embodiment of the present application is disclosed can apply in processor 301, or be realized by processor 301. Processor 301 can be a kind of IC chip, the disposal ability with signal.In implementation process, the above method it is each Step can be completed by the computer program in processor 301.Above-mentioned processor 301 can be general processor, numeral Signal processor (digital signal processing, DSP), application specific integrated circuit (application-specific Integrated circuit, ASIC), field programmable gate array (field-programmable gate array, ) or other PLDs, discrete gate or transistor logic, discrete hardware components FPGA.Can realize or Person performs disclosed each method, step and logic diagram in the embodiment of the present application.General processor can be microprocessor or Person's processor can also be any conventional processor etc..The step of method with reference to disclosed in the embodiment of the present application, can be straight Connect and be presented as that hardware decoding processor performs completion, or performed with the hardware in decoding processor and software module combination Into.Software module can be positioned at random access memory, flash memory, read-only storage, and programmable read only memory or electrically-erasable can In the ripe storage medium in this areas such as programmable memory, register.The storage medium is located at memory 302, and processor 301 is read Computer program in access to memory 302, the step of completing the above method with reference to its hardware.
COM1 303 can be used for receiving or sending signaling/information, such as receiving cellular engineering parameter.
In the embodiment of the present application, processor 301, for performing aforementioned flow Forecasting Methodology.
It should also be noted that, in the above-described embodiments, can wholly or partly by software, hardware, firmware or It is combined to realize.When implemented in software, it can realize in the form of a computer program product whole or in part.
The computer program product includes one or more computer instructions.Load and perform on computers the meter During calculation machine programmed instruction, produce whole or in part according to the flow or function described in the embodiment of the present invention.The computer can To be all-purpose computer, special-purpose computer, computer network or other programmable devices.The computer instruction can be deposited Store up in a computer-readable storage medium, or from a computer-readable recording medium to another computer-readable recording medium Transmission, for example, the computer instruction can pass through wired (example from web-site, computer, server or data center Such as coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or wireless (such as infrared, wireless, microwave) mode to another website Website, computer, server or data center are transmitted.The computer-readable recording medium can be that computer can be deposited Any usable medium of storage is either set comprising data storages such as one or more usable mediums integrated server, data centers It is standby.The usable medium can be magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or partly lead Body medium, such as solid state hard disc (solid state disk, SSD) etc..
In addition it should be noted that, device embodiment described above be only it is schematical, wherein it is described as separation The unit of part description can be or may not be it is physically separate, the part shown as unit can be or It can not be physical location, you can with positioned at a place, or can also be distributed on multiple NEs.Can be according to reality Some or all of module therein is selected to realize the purpose of this embodiment scheme the need for border.In addition, what the application was provided In device embodiment accompanying drawing, the annexation between module represents there is communication connection between them, specifically can be implemented as one Bar or a plurality of communication bus or signal wire.
, can be with several embodiments provided herein, it should be understood that disclosed system, module and method Realize by another way.For example, device embodiment described above is only schematical, for example, the module Divide, only a kind of division of logic function there can be other dividing mode when actually realizing, such as multiple units or component Another system can be combined or be desirably integrated into, or some features can be ignored, or do not perform.It is another, it is shown or The coupling each other discussed or direct-coupling or communication connection can be the indirect couplings of device or unit by some interfaces Close or communicate to connect, can be electrical, machinery or other forms.
The unit illustrated as separating component can be or may not be it is physically separate, it is aobvious as unit The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple On NE.Some or all of unit therein can be selected to realize the mesh of this embodiment scheme according to the actual needs 's.
Through the above description of the embodiments, it is apparent to those skilled in the art that the application can be borrowed Software is helped to add the mode of required common hardware to realize, naturally it is also possible to include application specific integrated circuit, specially by specialized hardware Realized with CPU, private memory, special components and parts etc..Generally, all functions of being completed by computer program can Easily realized with corresponding hardware, moreover, can also be a variety of many for the particular hardware structure for realizing same function Sample, such as analog circuit, digital circuit or special circuit.But, it is more for purposes of this application in the case of software program it is real It is now more preferably embodiment.Understood based on such, the technical scheme of the application is substantially made to prior art in other words The part of contribution can be embodied in the form of software product, and the computer software product is stored in the storage medium that can be read In, floppy disk, USB flash disk, mobile hard disk, read-only storage, random access memory, magnetic disc or the CD of such as computer, including Some instructions are to cause a computer equipment (can be personal computer, server, or network equipment etc.) to perform sheet Apply for the method described in each embodiment.
Described above, above example is only to the technical scheme for illustrating the application, rather than its limitations；Although with reference to before Embodiment is stated the application is described in detail, it will be understood by those within the art that：It still can be to preceding State the technical scheme described in each embodiment to modify, or equivalent substitution is carried out to which part technical characteristic；And these Modification is replaced, and the essence of appropriate technical solution is departed from the spirit and scope of each embodiment technical scheme of the application.

Claims (27)

1. a kind of method for predicting, it is characterised in that including：
Obtain in presetting range, the history of all cells always uses data on flows；
The history of all cells is always used to data on flows as input, using the first forecast model to the presetting range The total of interior all cells is predicted to obtain target prediction value using flow；
The use flow for obtaining the Target cell is predicted to the use flow of Target cell according to the target prediction value Predicted value, wherein, the Target cell is any one cell in the presetting range.
2. according to the method described in claim 1, it is characterised in that described that Target cell is made according to the target prediction value The use traffic prediction value for obtaining the Target cell is predicted with flow, including：
Cell in the presetting range is layered to obtain tree hierarchy structure；
According in the tree hierarchy structure, all cells is total using under traffic prediction value and Section Point under first node The history of all cells always uses data on flows, and all cells under the Section Point are always used using the second forecast model Flow is predicted, and the Section Point is the downstream site of the first node, wherein, when the first node is described During node, the total of all cells uses traffic prediction value to be the target prediction value under the first node；
After the use traffic prediction value of all cells under the 3rd node is determined, according to all cells under the 3rd node All cells always use volume forecasting under each node in total superior node using traffic prediction value and the 3rd node Value and the history of the Target cell always use data on flows, and the Target cell is made using second forecast model The use traffic prediction value for obtaining the Target cell is predicted with flow, the Target cell is under the 3rd node Cell.
3. method according to claim 2, it is characterised in that the cell in the presetting range be layered with Tree hierarchy structure is obtained, including：
Obtain cellular engineering parameter；
The network topology structure of cell in the presetting range is determined according to the cellular engineering parameter；
It regard the network topology structure as the tree hierarchy structure.
4. method according to claim 2, it is characterised in that the cell in the presetting range be layered with Tree hierarchy structure is obtained, including：
Plan position (the x of each cell presetting range Nei is determined according to below equationcell、ycell)：
xcell=xsite*(λ*ptrx)*h*sin(α)*cos(θ)；
ycell=ysite*(λ*ptrx)*h*sin(α)*sin(θ)；
Wherein, the xsite、ysiteThe longitude of each respectively described cell respective base station, dimension；The xcell、ycellRespectively The longitude of each cell, dimension, the λ are preset data, the ptrxFor the transmission power of the base station, the h is institute The antenna height of each cell is stated, described α, θ are respectively Downtilt, the deflection of each cell respective base station；
All cells in the presetting range are layered according to the plan position of each cell to obtain the tree-like layer Level structure.
5. method according to claim 4, it is characterised in that the plan position of each cell described in the basis is to described pre- All cells are layered to obtain the tree hierarchy structure in the range of putting, including：
The plan position of each cell is carried out hierarchical clustering to obtain the tree hierarchy structure.
6. method according to claim 5, it is characterised in that the plan position to each cell carries out stratification Cluster to obtain the tree hierarchy structure, including：
Cluster mode to the plan position progress hierarchical clustering of each cell to obtain the tree hierarchy using K-mean Structure.
7. method according to claim 2, it is characterised in that the cell in the presetting range be layered with Tree hierarchy structure is obtained, including：
Obtain ticket writing；
Training sample is obtained according to the ticket writing；
The training sample is expanded into target root vector；
The target root vector is carried out hierarchical clustering to obtain the tree hierarchy structure.
8. method according to claim 7, it is characterised in that described that training sample, bag are obtained according to the ticket writing Include：
Inquire about in each cell that the ticket writing is obtained in the preset range, the discharge record letter of all terminal devices Breath；
It regard the discharge record information as the training sample.
9. method according to claim 7, it is characterised in that it is described the target root vector is carried out hierarchical clustering with The tree hierarchy structure is obtained, including：
Cluster mode to target root vector progress hierarchical clustering to obtain the tree hierarchy using bi-clustering Structure.
10. the method according to claims require 1-9, it is characterised in that methods described includes：
The Target cell is made based on the first tree hierarchy structure, two tree hierarchy structures and the 3rd tree-like hierarchical structure It is predicted with flow to obtain the first predicted value, the second predicted value and the 3rd predicted value respectively, wherein, the first tree-like layer Level structure is the tree hierarchy structure obtained according to the network topology structure, and the second tree hierarchy structure is according to described Intend the tree hierarchy structure that position is obtained, the 3rd tree-like hierarchical structure is the tree hierarchy obtained according to the ticket writing Structure；
The use volume forecasting of the Target cell is worth to according to first predicted value, the second predicted value and the 3rd prediction Value.
11. method according to claim 10, it is characterised in that described according to first predicted value, the second predicted value It is worth to the 3rd prediction before the use traffic prediction value of the Target cell, methods described also includes：
Determine that the history of the Target cell in preset period of time uses the weighted average of data on flows；
It is described to predict that the use flow for being worth to the Target cell is pre- according to first predicted value, the second predicted value and the 3rd Measured value, including：
Using first predicted value, the second predicted value, the 3rd predicted value and the cellular engineering parameter as input, the weighting Average value is label, is merged based on random forests algorithm, obtains the use traffic prediction value of the Target cell.
12. method according to claim 11, it is characterised in that the Target cell goes through in the determination preset period of time History uses the weighted average of data on flows, including：
N number of history to the Target cell in the preset period of time uses data on flows point to carry out abnormality detection, and the N is just whole Number, the N is more than or equal to 2；
By being obtained after the abnormality detection in the preset period of time, the history uses N number of abnormal coefficient at flow number strong point；
Determine that N number of history uses the corresponding weight in flow number strong point according to N number of abnormal coefficient；
History of the Target cell in preset period of time is determined using the corresponding weight in flow number strong point according to N number of history Use the weighted average of flow.
13. method according to claim 12, it is characterised in that described that flow number strong point is used according to N number of history Corresponding weight determines that history of the Target cell in preset period of time uses the weighted average of flow, including：
N number of history is weighted averagely using the corresponding weight in flow number strong point to obtain the Target cell default History in period uses the weighted average of data on flows.
14. a kind of volume forecasting device, it is characterised in that including：
Acquisition module, for obtaining in presetting range, the history of all cells always uses data on flows；
First prediction module, for always using the history of all cells data on flows as input, uses the first prediction The total of all cells is predicted using flow to obtain target in the presetting range that model is obtained to the acquisition module Predicted value；
Second prediction module, the use for the target prediction value predicted according to first prediction module to Target cell Flow is predicted the use traffic prediction value for obtaining the Target cell, wherein, the Target cell is the presetting range Any one interior cell.
15. device according to claim 14, it is characterised in that described device also includes：
Hierarchical block, for being layered to the cell in the presetting range to obtain tree hierarchy structure；
Second prediction module is used for the use to Target cell according to the target prediction value of first prediction module prediction Flow is predicted the use traffic prediction value for obtaining the Target cell, including：
Second prediction module, for according in the tree hierarchy structure, the total of all cells to use stream under first node The history for measuring all cells under predicted value and Section Point always uses data on flows, using the second forecast model to described second The total of all cells is predicted using flow under node, and the Section Point is the downstream site of the first node, wherein, When the first node is the root node, the total of all cells using traffic prediction value is described the under the first node The target prediction value of one prediction module prediction；
After the use traffic prediction value of all cells under the 3rd node is determined, for according to all small under the 3rd node All cells always use flow under each node in total superior node using traffic prediction value and the 3rd node in area Predicted value and the history of the Target cell always use data on flows, using second forecast model to the Target cell Use flow be predicted the use traffic prediction value for obtaining the Target cell, the Target cell is the 3rd node Under cell.
16. device according to claim 15, it is characterised in that the hierarchical block is used in the presetting range Cell is layered to be included with obtaining tree hierarchy structure：
The hierarchical block is used for：
Obtain cellular engineering parameter；
The network topology structure of cell in the presetting range is determined according to the cellular engineering parameter；
It regard the network topology structure as the tree hierarchy structure.
17. device according to claim 15, it is characterised in that the hierarchical block is used in the presetting range Cell is layered to be included with obtaining tree hierarchy structure：
The hierarchical block is used for：
Plan position (the x of each cell presetting range Nei is determined according to below equationcell、ycell)：
xcell=xsite*(λ*ptrx)*h*sin(α)*cos(θ)；
ycell=ysite*(λ*ptrx)*h*sin(α)*sin(θ)；
Wherein, the xsite、ysiteThe longitude of each respectively described cell respective base station, dimension；The xcell、ycellRespectively The longitude of each cell, dimension, the λ are preset data, the ptrxFor the transmission power of the base station, the h is institute The antenna height of each cell is stated, described α, θ are respectively Downtilt, the deflection of each cell respective base station；
All cells in the presetting range are layered according to the plan position of each cell to obtain the tree-like layer Level structure.
18. device according to claim 17, it is characterised in that the hierarchical block is used for according to each cell Intend position all cells in the presetting range are layered to obtain the tree hierarchy structure, including：
The hierarchical block, carries out hierarchical clustering to obtain the tree hierarchy knot for the plan position to each cell Structure.
19. device according to claim 18, it is characterised in that the hierarchical block is used for the plan to each cell Position carries out hierarchical clustering to obtain the tree hierarchy structure, including：
The hierarchical block, for using K-mean cluster mode the plan position of each cell is carried out hierarchical clustering with Obtain the tree hierarchy structure.
20. device according to claim 15, it is characterised in that the hierarchical block is used for according to preset layered mode pair Cell in the presetting range is layered to be included with obtaining tree hierarchy structure：
The hierarchical block specifically for：
Obtain ticket writing；
Training sample is obtained according to the ticket writing；
The training sample is expanded into target root vector；
The target root vector is carried out hierarchical clustering to obtain the tree hierarchy structure.
21. device according to claim 20, it is characterised in that the hierarchical block is used to be obtained according to the ticket writing Training sample is taken, including：
The hierarchical block is used for：
Inquire about in each cell that the ticket writing is obtained in the preset range, the discharge record letter of all terminal devices Breath；
It regard the discharge record information as the training sample.
22. device according to claim 21, it is characterised in that the hierarchical block is used to enter the target root vector Row hierarchical clustering to obtain the tree hierarchy structure, including：
The hierarchical block, hierarchical clustering is carried out for clustering mode using bi-clustering to the target root vector To obtain the tree hierarchy structure.
23. the device according to claim 14-22, it is characterised in that second prediction module is used for：
The Target cell is made based on the first tree hierarchy structure, two tree hierarchy structures and the 3rd tree-like hierarchical structure It is predicted with flow to obtain the first predicted value, the second predicted value and the 3rd predicted value respectively, wherein, the first tree-like layer Level structure is the tree hierarchy structure obtained according to the network topology structure, and the second tree hierarchy structure is according to described Intend the tree hierarchy structure that position is obtained, the 3rd tree-like hierarchical structure is the tree hierarchy obtained according to the ticket writing Structure；
Described device also includes：
3rd prediction module, for first predicted value predicted according to second prediction module, the second predicted value and Three predictions are worth to the use traffic prediction value of the Target cell.
24. device according to claim 23, it is characterised in that described device also includes determining module；
The determining module, for first prediction predicted in the 3rd prediction module according to second prediction module Value, the second predicted value and the 3rd prediction are worth to before the use traffic prediction value of the Target cell, are determined in preset period of time The history of the Target cell uses the weighted average of data on flows；
3rd prediction module be used for according to second prediction module predict first predicted value, the second predicted value and 3rd prediction is worth to the use traffic prediction value of the Target cell, including：
3rd prediction module, for first predicted value, the second predicted value, the 3rd predicted value and the cell work Journey parameter is input, and the weighted average is label, is merged based on random forests algorithm, obtains the Target cell Use traffic prediction value.
25. device according to claim 24, it is characterised in that the determining module is used to determine described in preset period of time The history of Target cell is included using the weighted average of data on flows：
The determining module specifically for：
N number of history to the Target cell in the preset period of time uses data on flows point to carry out abnormality detection, and the N is just whole Number, the N is more than or equal to 2；
By being obtained after the abnormality detection in the preset period of time, the history uses N number of abnormal coefficient at flow number strong point；
Determine that N number of history uses the corresponding weight in flow number strong point according to N number of abnormal coefficient；
History of the Target cell in preset period of time is determined using the corresponding weight in flow number strong point according to N number of history Use the weighted average of flow.
26. device according to claim 25, it is characterised in that the determining module is used to described N number of be gone through according to described History determines that history of the Target cell in preset period of time is put down using the weighting of flow using the corresponding weight in flow number strong point Average, including：
N number of history is weighted averagely using the corresponding weight in flow number strong point to obtain the Target cell default History in period uses the weighted average of data on flows.
27. a kind of computer equipment, including memory, processor and it is stored on the memory and can be in the processing The computer program run on device, is realized such as any one of claim 1-13 institutes described in the computing device during computer program The method stated.
CN201710391210.1A 2017-05-27 2017-05-27 Flow prediction method and device Active CN107171848B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710391210.1A CN107171848B (en) 2017-05-27 2017-05-27 Flow prediction method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710391210.1A CN107171848B (en) 2017-05-27 2017-05-27 Flow prediction method and device

Publications (2)

Publication Number Publication Date
CN107171848A true CN107171848A (en) 2017-09-15
CN107171848B CN107171848B (en) 2020-07-07

Family

ID=59822172

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710391210.1A Active CN107171848B (en) 2017-05-27 2017-05-27 Flow prediction method and device

Country Status (1)

Country Link
CN (1) CN107171848B (en)

Cited By (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108234198A (en) * 2017-12-19 2018-06-29 清华大学 A kind of base station method for predicting and equipment
CN108399748A (en) * 2018-03-08 2018-08-14 重庆邮电大学 A kind of road travel time forecasting methods based on random forest and clustering algorithm
CN108495341A (en) * 2018-02-08 2018-09-04 西安电子科技大学 A kind of method for obligating resource and system, mobile communication system based on flow direction prediction
CN109120463A (en) * 2018-10-15 2019-01-01 新华三大数据技术有限公司 Method for predicting and device
CN109981393A (en) * 2017-12-28 2019-07-05 中国移动通信集团吉林有限公司 A kind of method and device judging cell flow saturation degree
CN110049501A (en) * 2018-01-15 2019-07-23 中兴通讯股份有限公司 Data capture method, device and computer readable storage medium
CN110175711A (en) * 2019-05-17 2019-08-27 北京市天元网络技术股份有限公司 One kind being based on joint LSTM base station cell method for predicting and device
CN110351752A (en) * 2019-06-27 2019-10-18 珠海云洲智能科技有限公司 A kind of unmanned boat and its network optimized approach, device and storage medium
CN110474808A (en) * 2019-08-20 2019-11-19 中国联合网络通信集团有限公司 A kind of method for predicting and device
CN110769454A (en) * 2018-07-25 2020-02-07 中国移动通信集团浙江有限公司 Flow prediction method and device
CN111030869A (en) * 2019-12-20 2020-04-17 锐捷网络股份有限公司 Network traffic prediction method and prediction device
CN111130950A (en) * 2019-12-31 2020-05-08 中国联合网络通信集团有限公司 Network flow evaluation method and device
CN111130932A (en) * 2019-12-18 2020-05-08 北京浩瀚深度信息技术股份有限公司 Method and device for predicting flow trend based on historical flow and storage medium
CN111162925A (en) * 2018-11-07 2020-05-15 中移(苏州)软件技术有限公司 Network flow prediction method and device, electronic equipment and storage medium
WO2020098016A1 (en) * 2018-11-14 2020-05-22 网宿科技股份有限公司 Network traffic prediction method, device, and electronic device
CN111343006A (en) * 2020-02-12 2020-06-26 咪咕文化科技有限公司 CDN peak flow prediction method, device and storage medium
CN111527734A (en) * 2017-12-29 2020-08-11 华为技术有限公司 Node traffic ratio prediction method and device
CN111913921A (en) * 2019-05-08 2020-11-10 中国移动通信集团福建有限公司 User behavior analysis method and device, equipment and storage medium
CN112492637A (en) * 2020-12-28 2021-03-12 中国联合网络通信集团有限公司 Method and device for predicting cell traffic
CN113038302A (en) * 2019-12-25 2021-06-25 中国电信股份有限公司 Flow prediction method and device and computer storage medium
CN113726557A (en) * 2021-08-09 2021-11-30 国网福建省电力有限公司 Network transmission control optimization method based on flow demand

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101541030A (en) * 2009-05-06 2009-09-23 华为技术有限公司 Method for predicting data based on support vector machine and equipment thereof
CN101541016A (en) * 2009-05-06 2009-09-23 华为技术有限公司 Method for predicting data and equipment
WO2014129945A1 (en) * 2013-02-25 2014-08-28 Telefonaktiebolaget L M Ericsson (Publ) Determination of network parameters in mobile communication networks
CN105634787A (en) * 2014-11-26 2016-06-01 华为技术有限公司 Evaluation method, prediction method and device and system for network key indicator
CN105898762A (en) * 2015-01-26 2016-08-24 华为技术有限公司 Base station optimization and deployment method and device
CN105898761A (en) * 2014-11-07 2016-08-24 南京大学 Heterogeneous cellular wireless network programming scheme

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101541030A (en) * 2009-05-06 2009-09-23 华为技术有限公司 Method for predicting data based on support vector machine and equipment thereof
CN101541016A (en) * 2009-05-06 2009-09-23 华为技术有限公司 Method for predicting data and equipment
WO2014129945A1 (en) * 2013-02-25 2014-08-28 Telefonaktiebolaget L M Ericsson (Publ) Determination of network parameters in mobile communication networks
CN105898761A (en) * 2014-11-07 2016-08-24 南京大学 Heterogeneous cellular wireless network programming scheme
CN105634787A (en) * 2014-11-26 2016-06-01 华为技术有限公司 Evaluation method, prediction method and device and system for network key indicator
CN105898762A (en) * 2015-01-26 2016-08-24 华为技术有限公司 Base station optimization and deployment method and device

Cited By (32)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108234198A (en) * 2017-12-19 2018-06-29 清华大学 A kind of base station method for predicting and equipment
CN108234198B (en) * 2017-12-19 2020-07-07 清华大学 Base station flow prediction method and equipment
CN109981393A (en) * 2017-12-28 2019-07-05 中国移动通信集团吉林有限公司 A kind of method and device judging cell flow saturation degree
CN111527734A (en) * 2017-12-29 2020-08-11 华为技术有限公司 Node traffic ratio prediction method and device
CN111527734B (en) * 2017-12-29 2021-10-26 华为技术有限公司 Node traffic ratio prediction method and device
CN110049501A (en) * 2018-01-15 2019-07-23 中兴通讯股份有限公司 Data capture method, device and computer readable storage medium
CN108495341A (en) * 2018-02-08 2018-09-04 西安电子科技大学 A kind of method for obligating resource and system, mobile communication system based on flow direction prediction
CN108399748B (en) * 2018-03-08 2020-12-22 重庆邮电大学 Road travel time prediction method based on random forest and clustering algorithm
CN108399748A (en) * 2018-03-08 2018-08-14 重庆邮电大学 A kind of road travel time forecasting methods based on random forest and clustering algorithm
CN110769454A (en) * 2018-07-25 2020-02-07 中国移动通信集团浙江有限公司 Flow prediction method and device
CN109120463A (en) * 2018-10-15 2019-01-01 新华三大数据技术有限公司 Method for predicting and device
CN111162925A (en) * 2018-11-07 2020-05-15 中移(苏州)软件技术有限公司 Network flow prediction method and device, electronic equipment and storage medium
CN111162925B (en) * 2018-11-07 2022-06-14 中移(苏州)软件技术有限公司 Network flow prediction method and device, electronic equipment and storage medium
US11252044B2 (en) 2018-11-14 2022-02-15 Wangsu Science & Technology Co., Ltd. Network traffic prediction method, apparatus, and electronic device
WO2020098016A1 (en) * 2018-11-14 2020-05-22 网宿科技股份有限公司 Network traffic prediction method, device, and electronic device
CN111913921A (en) * 2019-05-08 2020-11-10 中国移动通信集团福建有限公司 User behavior analysis method and device, equipment and storage medium
CN110175711A (en) * 2019-05-17 2019-08-27 北京市天元网络技术股份有限公司 One kind being based on joint LSTM base station cell method for predicting and device
CN110351752B (en) * 2019-06-27 2021-11-30 珠海云洲智能科技股份有限公司 Unmanned ship, network optimization method and device thereof, and storage medium
CN110351752A (en) * 2019-06-27 2019-10-18 珠海云洲智能科技有限公司 A kind of unmanned boat and its network optimized approach, device and storage medium
CN110474808A (en) * 2019-08-20 2019-11-19 中国联合网络通信集团有限公司 A kind of method for predicting and device
CN110474808B (en) * 2019-08-20 2022-02-18 中国联合网络通信集团有限公司 Flow prediction method and device
CN111130932B (en) * 2019-12-18 2021-10-19 北京浩瀚深度信息技术股份有限公司 Method and device for predicting flow trend based on historical flow and storage medium
CN111130932A (en) * 2019-12-18 2020-05-08 北京浩瀚深度信息技术股份有限公司 Method and device for predicting flow trend based on historical flow and storage medium
CN111030869A (en) * 2019-12-20 2020-04-17 锐捷网络股份有限公司 Network traffic prediction method and prediction device
CN111030869B (en) * 2019-12-20 2022-09-20 锐捷网络股份有限公司 Network traffic prediction method and prediction device
CN113038302A (en) * 2019-12-25 2021-06-25 中国电信股份有限公司 Flow prediction method and device and computer storage medium
CN113038302B (en) * 2019-12-25 2022-09-30 中国电信股份有限公司 Flow prediction method and device and computer storage medium
CN111130950A (en) * 2019-12-31 2020-05-08 中国联合网络通信集团有限公司 Network flow evaluation method and device
CN111343006A (en) * 2020-02-12 2020-06-26 咪咕文化科技有限公司 CDN peak flow prediction method, device and storage medium
CN112492637A (en) * 2020-12-28 2021-03-12 中国联合网络通信集团有限公司 Method and device for predicting cell traffic
CN112492637B (en) * 2020-12-28 2023-06-27 中国联合网络通信集团有限公司 Method and device for predicting cell traffic
CN113726557A (en) * 2021-08-09 2021-11-30 国网福建省电力有限公司 Network transmission control optimization method based on flow demand

Also Published As

Publication number Publication date
CN107171848B (en) 2020-07-07

Similar Documents

Publication Publication Date Title
CN107171848A (en) A kind of method for predicting and device
CN102567391B (en) Method and device for building classification forecasting mixed model
CN106301984B (en) A kind of mobile communications network capacity prediction methods and device
CN106992904A (en) Network equipment health degree appraisal procedure based on dynamic comprehensive weight
CN107426759A (en) The Forecasting Methodology and system of newly-increased base station data portfolio
CN105634787A (en) Evaluation method, prediction method and device and system for network key indicator
CN105550714A (en) Cluster fusion method for warning information in heterogeneous network environment
CN106487707B (en) A kind of virtual fault tolerant network mapping method of power communication and device
CN105340311B (en) The network equipment, network service prediction meanss and method
CN106843100A (en) Substation secondary device running status level determination method and device
CN103686818A (en) Simulation test method and equipment
CN106454857A (en) Evaluation method and device for network planning
JP7486659B2 (en) Energy saving processing method and processing device for base station
CN108243435A (en) Parameter optimization method and device in a kind of LTE cell scenarios division
CN110363658A (en) Processing method and processing device, storage medium and the electronic device of credit data
Dong et al. Predicting LTE throughput using traffic time series
CN114066184A (en) Area coverage assessment method and device and electronic equipment
CN111797892B (en) Electric power market generator market force monitoring method based on random forest regression
CN105873087A (en) Network index prediction method, device and electronic device
CN104349332B (en) A kind of Topological expansion method and device
CN103178995B (en) A kind of multiple dimensioned Effectiveness Evaluation for Communication Network method of architecture
US8355945B1 (en) Identifying and ranking high-impact churn sectors
CN106993296A (en) The performance estimating method and device of terminal
CN106603294B (en) A kind of synthesis vulnerability assessment method based on power communication web frame and state
Arifin et al. The prediction of mobile data traffic based on the ARIMA model and disruptive formula in industry 4.0: A case study in Jakarta, Indonesia

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