CN101453747B - Telephone traffic prediction method and apparatus - Google Patents
Telephone traffic prediction method and apparatus Download PDFInfo
- Publication number
- CN101453747B CN101453747B CN2008102253945A CN200810225394A CN101453747B CN 101453747 B CN101453747 B CN 101453747B CN 2008102253945 A CN2008102253945 A CN 2008102253945A CN 200810225394 A CN200810225394 A CN 200810225394A CN 101453747 B CN101453747 B CN 101453747B
- Authority
- CN
- China
- Prior art keywords
- grid
- adjacent
- reports
- district
- time period
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/02—Traffic management, e.g. flow control or congestion control
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a method and a device for predicting traffic, which predicting the traffic based on a mass measurement report and comprises: determining the grid in which a measurement report MR is positioned during reporting according to the report position of the measurement report received, wherein the grid is obtained by dividing the geographic area of the traffic to be tested in advanced; counting the MRs in all time periods of each grid according to the report time and the grid of the MR; acquiring the number of MRs in the M adjacent time periods of a current time period in each grid and calculating a predicted value of the traffic in the current time period in each grid by using a time sequence prediction algorithm. The method and device realize real-time, fine-granularity and high-precision traffic prediction.
Description
Technical field
The present invention relates to moving communicating field, refer to a kind of especially based on measurement report (Measurement Report, telephone traffic prediction method MR) and device.
Background technology
Because the mobility of terminal in the mobile communications network, the distribution of the customer service of mobile communications network is two-dimentional, just free and two dimensions in space, and the telephone traffic distribution of prediction mobile communications network is a current mobile communication network planning and a key issue of guarantee.
Present stage, the traffic forecast for mobile communications network mainly was the prediction of time dimension, and the data foundation is operation and maintenance center (Operations﹠amp; Maintenance Center, the OMC) count value of Tong Ji counter, these count values are stored according to the seasonal effect in time series form.
And the spatial granularity of these count values sub-district (cell) granularity normally, even sometimes be base station controller (Base Station Controller, BSC) level other, such as talk times, switching times etc., the area coverage of sub-district has several square kilometres usually, even tens square kilometres, so the fine granulation in its space is not enough.Simultaneously, the area coverage of a cell is very irregular, and the difference in size of its area is very big.And the space identification of this cell and actual geographical position can not very accurate correspondences, though a cell has longitude and latitude, but because the scrambling of the shape in master control zone and the uncertainty of area, thereby cause specifically to reflect the telephone traffic situation of a geographic area.
In addition, the time granularity of these count values is also bigger usually, generally is half an hour, or even one hour.And other time granularity of hour level for the emergency case/incident of telephone traffic can not be very sensitive reflect.
In the count value with the OMC statistics is on the basis of data foundation, forecasting tool also is the seasonal effect in time series forecasting tool, such as analysis of history, long-term numerical value time series tendency, the tendency of the telephone traffic of prediction next stage, on the basis of cell granularity, according to hour gathering, as the traffic forecast value in a bigger zone.When carrying out traffic forecast, these data are according to being to want level and smooth filtration of elapsed time, and therefore this algorithm is difficult to reflect for the variation of the network traffic situation of burst; And since its data according to be the cell level other, therefore cause precision of prediction lower, can only predicted level greater than the telephone traffic situation of change of the network element of cell.
The prediction algorithm of prior mobile network can only use as the forecasting tool of an off-line, some the telephone traffic of predicting tomorrow such as today, can not as one in real time (less than one hour, in addition reach minute wait more fine-grained) the forecasting tool use.
Summary of the invention
The embodiment of the invention provides a kind of telephone traffic prediction method and device, has realized in real time, fine granularity, high-precision traffic forecast.
A kind of telephone traffic prediction method comprises:
According to the position that reports of the measurement report MR that receives, the grid at the place of determining to give the correct time on the described MR; Described grid is to obtain after in advance the geographic area of telephone traffic to be predicted being divided;
According to calling time on the described MR and the grid at place, count the MR quantity that reports in each time period in each grid;
Obtain the MR quantity that adjacent preceding M the time period of current slot reports in each grid, calculate the traffic forecast value of each grid current slot; Described M is the positive integer greater than 1.
According to said method of the present invention, the measurement report MR that described basis receives reports the position, and the grid at the place of determining to give the correct time on the described MR comprises:
According to described MR, queries static allocation list, the described latitude and longitude value that reports the position at the place of determining to give the correct time on each described MR;
Inquire about the grid at described latitude and longitude value place, determine the grid at the place of giving the correct time on each described MR.
According to said method of the present invention, the latitude and longitude value that reports the position at the place of determining to give the correct time on the MR specifically comprises:
According to each adjacent Cell Broadcast CB control channel BCCH and base station identity code BSIC of comprising among the MR, LAC and CI that Location Area Code LAC that is complementary with BCCH and BSIC that the queries static allocation list is obtained and cell ID CI are defined as each adjacent sub-district;
According to the descending level of the adjacent sub-district that comprises among the MR, the algorithm of adopt setting is inquired about described static configuration table, calculates this MR and reports the position to arrive distance apart from each adjacent cell base station;
According to giving the correct time on this MR apart from the longitude and latitude of each adjacent cell base station position that the distance of each adjacent cell base station and LAC and CI by described each adjacent sub-district check in, by the triangle location algorithm, the latitude and longitude value that reports the position at the place that obtains giving the correct time on each MR.
According to said method of the present invention, LAC that is complementary with BCCH and BSIC that obtains when the queries static allocation list and CI quantity are during greater than the adjacent number of cells of reality, according to the LAC and the CI of the main plot that comprises among the MR, determine the position wherein LAC of close main plot and LAC and the CI that CI is each adjacent sub-district.
According to said method of the present invention, described descending level according to the adjacent sub-district that comprises among the MR adopts the algorithm of setting, and inquires about described static configuration table, calculates this MR and reports the position to arrive the distance of each adjacent cell base station, specifically comprises:
LAC and CI according to described each adjacent sub-district, the queries static allocation list, obtain the transmitting power of each adjacent antenna in cell, the absolute value of the transmitting power of calculating each adjacent antenna in cell and the difference of corresponding described descending level obtains the path loss corresponding with each adjacent sub-district;
According to described path loss, the queries static allocation list determines that this MR reports the position to arrive the distance of each adjacent cell base station.
Said method of the present invention also comprises: the traffic forecast value that calculates each grid current slot is optimized;
Optimize the traffic forecast value of current slot in the grid, comprising:
The previous time period of determining described current slot is gone up the degree of correlation of each adjacent cells and optimised grid; The described degree of correlation is determined according to the MR that reports in each grid involved main plot and neighbor cell;
The variable quantity of the MR quantity that the previous time period of determining current slot described in each adjacent cells reported with respect to previous time period again;
According to the described degree of correlation and variable quantity the traffic forecast value of described optimised grid is optimized the traffic forecast value after being optimized.
According to said method of the present invention, the described previous time period of determining described current slot is gone up the degree of correlation of each adjacent cells and optimised grid, specifically comprises:
Add up the described previous time period and go up the involved sub-district of MR that reports in optimised grid and each adjacent cells, described sub-district comprises main plot and adjacent sub-district;
Calculate the similarity of sub-district involved described in sub-district involved described in each adjacent cells and the optimised grid respectively, obtain the previous time period to go up the degree of correlation of each adjacent cells and optimised grid.
According to said method of the present invention, described according to the described degree of correlation and variable quantity, the traffic forecast value of described optimised grid is optimized, the traffic forecast value after being optimized specifically comprises:
Calculate the product of variable quantity of the described MR quantity of the degree of correlation of each adjacent cells correspondence and this grid respectively, and the algebraical sum of all products that calculate;
With described algebraical sum and the addition of described traffic forecast value, the traffic forecast value after being optimized.
According to said method of the present invention, after the grid at the place of giving the correct time on described definite described MR, also comprise:
Generate the information record of all MR that each time period reports in each grid; Comprise main plot, neighbor cell that gives the correct time on each MR and the latitude and longitude value that reports the position in the described information record.
A kind of traffic forecast device comprises:
The grid determination module is used for the position that reports according to the measurement report MR that receives, the grid at the place of determining to give the correct time on the described MR; Described grid is to obtain after in advance the geographic area of telephone traffic to be predicted being divided;
The quantity statistics module is used for counting the MR quantity that reports in each time period in each grid according to calling time on the described MR and the grid at place;
Prediction module is used for obtaining before each grid current slot adjacent the MR quantity that M time period reports, and employing time series forecasting algorithm calculates the traffic forecast value of each grid current slot.
According to said apparatus of the present invention, described grid determination module comprises:
The adjacent area determining unit, each that is used for comprising adjacent Cell Broadcast CB control channel BCCH and base station identity code BSIC according to MR, LAC and CI that Location Area Code LAC that is complementary with BCCH and BSIC that the queries static allocation list is obtained and cell ID CI are defined as each adjacent sub-district;
Metrics calculation unit is used for the descending level of the adjacent sub-district that comprises according to MR, adopts the algorithm of setting, and inquires about described static configuration table, calculates this MR and reports the position to arrive distance apart from each adjacent cell base station;
The longitude and latitude determining unit, be used for according to giving the correct time on this MR apart from the longitude and latitude of each adjacent cell base station position that the distance of each adjacent cell base station and LAC and CI by described each adjacent sub-district check in, by the triangle location algorithm, the latitude and longitude value that reports the position at the place that obtains giving the correct time on each MR;
The grid determining unit is used to inquire about the described grid that reports the latitude and longitude value place of position, determines the grid at the place of giving the correct time on described each MR.
Said apparatus of the present invention also comprises:
Optimal module is used for the traffic forecast value that calculates each grid current slot is optimized.
According to said apparatus of the present invention, described optimal module comprises:
Degree of correlation determining unit, the previous time period that is used for determining described current slot is gone up the degree of correlation of each adjacent cells and optimised grid; The described degree of correlation is determined according to the MR that reports in each grid involved main plot and neighbor cell;
The variable quantity determining unit, the variable quantity of the MR quantity that the previous time period that is used for determining current slot described in each adjacent cells reported with respect to previous time period again;
Optimize the unit, be used for described traffic forecast value being optimized the traffic forecast value after being optimized according to the described degree of correlation and variable quantity.
Said apparatus of the present invention also comprises: the record generation module is used for generating the information record of all MR that each time period of each grid reports; Comprise main plot, neighbor cell that gives the correct time on each MR and the latitude and longitude value that reports the position in the described information record.
Telephone traffic prediction method that the embodiment of the invention provides and device by telephone traffic prediction method and the device that provides in the invention process, by the magnanimity MR that receives is carried out analyzing and processing, determine that it reports position and affiliated grid; Count the MR quantity that reports in each time period in each grid; And, predict the traffic forecast value of each grid current slot by the MR quantity that M time period before current slot known in each grid adjacent reports.The inventive method can be applied in the system of online network load monitoring in real time, and real-time estimate next stage telephone traffic distributes, and has realized more fine granulation, real-time call prediction more accurately by defining geographical grid and time division section.
Description of drawings
Fig. 1 is the flow chart of telephone traffic prediction method in the embodiment of the invention;
Fig. 2 reports the flow chart of position for the place of giving the correct time on definite MR in the embodiment of the invention;
Fig. 3 a is the principle schematic of embodiment of the invention intermediate cam location algorithm;
Fig. 3 b is the location diagram of embodiment of the invention intermediate cam location algorithm;
Optimize the flow chart of predicted value in the telephone traffic prediction method that Fig. 4 provides for the embodiment of the invention;
Fig. 5 is the principle schematic of embodiment of the invention intermediate cam location algorithm;
Fig. 6 is the schematic diagram of MR number change amount in the adjacent cells in the embodiment of the invention;
Fig. 7 is the structural representation of traffic forecast device in the embodiment of the invention.
Embodiment
The telephone traffic prediction method that the embodiment of the invention provides based on to magnanimity measurement report MR analyzing and processing, is predicted the telephone traffic on the following space-time (or other key indexs).Wherein, the definition of space-time is the geographical grid of the time series and the two dimension of one dimension.MR is the measurement data of the channel quality of mobile terminal reporting, and (Slow Associated Control Channel SACCH) transmits, as the judgement foundation that network switches and power is controlled by slow associated control channel.MR comprises the measurement data of main Serving cell (Serving Cell is called for short the main plot) and adjacent sub-district (Neighbour Cell).The acquiescence cycle of uploading of measurement report is 0.48S, because base transceiver station (Base Transceiver Station, BTS) with base station controller (BaseStation Controller, BSC) communication interface between---Abis interface is not a standard interface, the MR data mode difference of therefore different equipment producers comprises: complete transmission, the average or sampling etc. according to a plurality of cycles.
The flow process of the inventive method as shown in Figure 1, its execution in step is as follows:
Step S10: a large amount of measurement report MR that receive are handled, determine the position that reports of giving the correct time on each MR.Be specially: according to the various parameters of main plot that comprises in described MR and the static configuration table and adjacent sub-district, the latitude and longitude value that reports the position at the place of determining to give the correct time on each MR.
Determine to give the correct time on each MR process that reports the position at place as shown in Figure 2, comprises the following steps:
S101, obtain MR one by one.Wherein, the data structure of MR is as follows:
typedef struct
{
time timestamp；
Int LAC; The LAC of // Serving cell
Int CI; The CI of // Serving cell
Int DL_LEV_ServingCell; The descending level of // Serving cell
Int DL_QUAL_ServingCell; The downlink quality of // Serving cell
Int UL_LEV_ServingCell; The last line level of // Serving cell
Int UL_QUAL_ServingCell; The up quality of // Serving cell
Int TA; The timing advance of // Serving cell
Int BSIC_NB[6]; // six BSIC of strong adjacent sub-district
Int BCCH_NB[6]; // six BCCH of strong adjacent sub-district
Int DL_LEV_NB[6]; // six the most descending level of strong adjacent sub-district
}_MR；
Wherein: LAC be Location Area Code (Location Area Code, LAC);
CI be cell ID (Cell Identity, CI);
BSIC be base station identity code (Base Station Identity Code, BSIC);
BCCH be Broadcast Control Channel (Broadcast Control Channel, BCCH).
S102, determine the LAC and the CI of the adjacent sub-district of giving the correct time on the MR.Be specially:
According to each adjacent sub-district BCCH and BSIC of comprising among the MR, LAC and CI that LAC that is complementary with BCCH and BSIC that the queries static allocation list is obtained and CI are defined as each adjacent sub-district.
Owing to do not comprise the LAC and the CI information of adjacent sub-district in the portable terminal measurement report, only comprise BCCH and BSI.And BCCH and BSIC are recursive in network, can not a certain sub-district of unique identification, and LAC and CI are only the unique identification of sub-district.Therefore need be according to the BCCH and the BSIC of adjacent sub-district, the queries static allocation list determines to meet the adjacent sub-district BCCH and the LAC of BSIC and LAC and the CI that CI is adjacent sub-district that comprise among the MR.
Especially, LAC that is complementary with BCCH and BSIC that obtains when the queries static allocation list and CI quantity are during greater than the adjacent number of cells of reality, according to the LAC and the CI of the main plot that comprises among the MR, determine the position wherein LAC of close main plot and LAC and the CI that CI is each adjacent sub-district.
Wherein, the data structure of the static configuration table of base station is as follows:
typedef struct
{
Int LAC; The LAC of // sub-district
Int CI; The CI of // sub-district
Double Longtitude; The longitude of // sub-district
Double Latitude; The latitude of // sub-district
Double POW; The antenna transmitting power of // sub-district
Double Height; The antenna height of // sub-district
Double Dir; The antenna directional angle of // sub-district
Double TILT; The Downtilt of // sub-district
Double Antenna_Corr; The antenna directive gain factor of // sub-district
Int Frequency_BAND; The frequency range of // sub-district
Int Geography_TYPE; The geographical pattern of // sub-district, city/suburb or the like
}_CELLINFO；
S103, according to the descending level of the adjacent sub-district that comprises among the MR, adopt the algorithm of setting, the queries static allocation list calculates this MR and reports the position to arrive distance apart from each adjacent cell base station.
At first, obtain the descending level of the adjacent sub-district that comprises among the transmitting power of each adjacent antenna in cell and the MR, determine the path loss corresponding with each adjacent sub-district.
Wherein, according to the LAC and the CI of each adjacent sub-district, the queries static allocation list can obtain the antenna transmitting power of each adjacent sub-district; Calculate the absolute value of difference of the descending level of the corresponding adjacent sub-district that comprises among antenna transmitting power and the MR, obtain the value of the path loss corresponding with each adjacent sub-district.
Also will consider of the influence of the factors such as plan loss of the diversity gain of gain, portable terminal reception antenna of direction, the portable terminal reception antenna of antenna for base station and dual polarized antenna especially, in practice to path loss.Consider that the path loss calculation formula after the above-mentioned factor is as follows:
Lpdown＝PoutBTS+Cori+GaMS+GdMS-LslantBTS-PinMS
Wherein, LPdown is a path loss;
PinMS is the power (being descending level) that portable terminal receives;
PoutBTS is the transmitting power (should comprise mixer, feeder line equal loss and antenna gain) of BTS antenna;
Cori is the direction coefficient of antenna for base station;
GaMS is the gain of portable terminal reception antenna;
GdMS is the diversity gain of portable terminal reception antenna;
LslantBTS is the polarization loss of dual polarized antenna.
Then, according to the path loss of determining, the queries static allocation list adopts village Okumura radio wave propagation decay computation schema difficult to understand, arrives the distance of each adjacent cell base station when determining mobile terminal reporting MR.Be specially:
LAC and CI according to each adjacent sub-district, from the static configuration table, check in parameters needed in village difficult to understand (Okumura) the radio wave propagation decay computation schema, for example: the antenna height of the antenna height of each adjacent cell base station, the antenna gain factor, operating frequency and portable terminal etc., to calculate again with each adjacent sub-district path loss substitution village (Okumura) difficult to understand radio wave propagation decay computation schema, the distance of each adjacent cell base station of determining to give the correct time on the MR.
Can revise the correction factor in the computing formula according to factors such as concrete communication environments and orographic conditions especially, in actual applications.
S104, according to giving the correct time on the MR apart from the distance of each adjacent cell base station and the longitude and latitude of each adjacent cell base station position, by the triangle location algorithm, the latitude and longitude value that reports the position at the place that obtains giving the correct time on each MR.
Wherein, the longitude and latitude of each adjacent cell base station position, LAC and CI queries static allocation list by each adjacent sub-district obtain.
Especially, when the number of adjacent sub-district is many, also can use all adjacent sub-districts, select three in the adjacent sub-district that also can from measurement report, relate to, arrive the distance of these three adjacent cell base stations during by mobile terminal reporting MR, the reporting position information when determining the mobile terminal reporting measurement report.
Especially, selecting three zones, can be that any selection also can be selected the highest sub-district of correlation.
Arrive the distance of these three adjacent cell base stations during according to mobile terminal reporting MR, and the actual range between per two base stations, adopt the triangle location algorithm to carry out multiple location, the reporting position information in the time of can determining the mobile terminal reporting measurement report.The principle of triangle location algorithm is shown in Fig. 3 a and Fig. 3 b.Wherein, d1-d3, d1-d2 and d2-d3 represent two distances between the base station respectively.For being that the position in the center of circle concern schematic diagram with the base station, dash area is for being to determine the regional location that gives the correct time on the MR among Fig. 3 b.
Wherein, the actual range between per two base stations can pass through the queries static allocation list, and the latitude and longitude value of the LAC of the adjacent sub-district of each that look into and the pairing base station location of CI can calculate the actual range between per two base stations.
What triangle determined that algorithm determines may be a point accurately, also may be a zone, is one when regional when what determine, determines to report position, i.e. longitude and latitude when this regional mid point is reporting measurement reports.
Give the correct time on the MR longitude and latitude position at place of S105, output.
For example, step S10 specifically can pass through array function realization down in practical operation:
_Located_MR Get_MR_Position(MR input，_CELLINFO data)
{
_Located_MR output
return output；
}
And the MR data structure behind the location is as follows:
typedef struct
{
int MR_longtitude；
int MR_latitude；
time timestamp；
Int LAC; The LAC of // Serving cell
Int CI; The CI of // Serving cell
Int DL_LEV_ServingCell; The descending level of // Serving cell
Int DL_QUAL_ServingCell; The downlink quality of // Serving cell
Int UL_LEV_ServingCell; The last line level of // Serving cell
Int UL_QUAL_ServingCell; The up quality of ∥ Serving cell
Int TA; The timing advance of // Serving cell
Int BSIC_NB[6]; // six BSIC of strong adjacent sub-district
Int BCCH_NB[6]; // six BCCH of strong adjacent sub-district
Int DL_LEV_NB[6]; // six the most descending level of strong adjacent sub-district
}_Located_MR；
Step S11: according to the position that reports of each MR, the grid at the place of determining to give the correct time on it.
Wherein, grid is to obtain after in advance the geographic area of telephone traffic to be predicted being divided, and its size can be set as required, and defines by longitude and latitude, and the size that for example can set grid is 100 meters
*100 meters etc.
By determining the position that reports of each MR, for example: latitude and longitude information, the grid at inquiry latitude and longitude value place promptly can be determined the grid at each MR place.
Step S12: call time on each MR and on the give the correct time grid at place, generate the information record of all MR that each time period reports in each grid.
Because single MR is not had clear meaning for space-time traffic forecast, need carry out analyzing and processing to the MR of magnanimity, obtain the quantizating index of MR on macroeconomic significance.In order to handle, in the MR data input database under needing to determine behind the grid to magnanimity MR.Can pass through the ODBC universal data interface, the MR behind the location is saved to database, in order to using.
According to calling time on give the correct time on each MR that determines residing space and corresponding this MR, the time period at the place of more than calling time is a foundation, generates the data model (be information record sheet) of magnanimity MR on time and two-dimensional space, and deposits in the database.
Wherein, comprise the information such as main plot, neighbor cell and longitude and latitude of giving the correct time on each MR in the information record.
The form of information record can be as shown in the table
| Time T ime | Main plot Cell | |
|
Adjacent sub-district 3 NB3 | |
|
Adjacent sub-district 6 NB6 | Longitude Lat | | Attribute | 1 | Attribute | 2 Attribute2 | …… |
| 9:10 | Sub-district 1 | Sub-district 3 | Sub-district 9 | Sub-district 5 | Sub-district 7 | Sub-district 6 | Sub-district 2 | XX | YY | —— | —— | …… | ||
| 9:10 | Sub-district 2 | Sub-district 3 | Sub-district 6 | Sub-district 5 | Sub-district 7 | Sub-district 1 | Sub-district 8 | XX | YY | —— | —— | …… | ||
| …… | …… | …… | …… | …… | …… | …… | …… | …… | …… | …… | …… | …… |
For example: in the practical operation, during with the MR input database, specifically can pass through array function realization down:
Int Store_MR_to_DB(_Located_MR data)
{
}
Step S13:, count the MR quantity that reports in each time period in each grid according to calling time on the described MR and the grid at place.
Step S14: obtain the MR quantity that adjacent preceding M the time period of current slot reports in each grid, adopt the time series forecasting algorithm, calculate the traffic forecast value of each grid current slot.
At each grid, its telephone traffic all can have the statistical value sequential recording of a MR quantity on time series.For example when the length of each time period was 1 minute, then this statistical value sequence was a sequence of values of storing according to minute granularity.
The MR quantity that the current slot that obtains and will predict reported in adjacent preceding M time period, the telephone traffic of prediction current slot.Wherein M is the positive integer greater than 1, specifically can calculate by following formula:
Traffic forecast value=radix+∑ variation tendency * time granularity+shake factor
Wherein, the quantity of the MR that reports for adjacent with current slot previous time period of radix;
Variation tendency is in described preceding M the time period, the number change value of the MR that the relative time period previous with it of the MR that each time period reports reports;
Time granularity is a short length of time;
The shake factor is the mean value of the difference of adjacent twice number change value.
For example: get the record of the MR quantity that reports in preceding 10 time periods, the traffic forecast value of prediction current slot.
Above-mentioned is that the telephone traffic in each grid is carried out prediction on the time series, after the prediction of carrying out on the time series, can also be according to two-dimensional space statistics variations rule, the traffic forecast value that calculates each grid current slot is optimized further revises and optimize, make it more approach actual value.Following step is exactly to optimize the process of the traffic forecast value on the time series, describes with the traffic forecast value of optimizing current slot in the grid (optimised grid), as shown in Figure 4, comprises the following steps:
Step S20: the previous time period of determining current slot is gone up the degree of correlation of each adjacent cells and optimised grid.
Wherein, the degree of correlation is determined according to the MR that reports in each grid involved main plot and neighbor cell, is specifically comprised:
(1) according to the information record, add up the previous time period and go up the involved sub-district of MR that reports in optimised grid and each adjacent cells, wherein the sub-district comprises main plot and adjacent sub-district.
(2) calculate the similarity of sub-district involved in sub-district involved in each adjacent cells and the optimised grid respectively, obtain the previous time period to go up the degree of correlation of each adjacent cells and optimised grid.
Wherein, the many more degrees of correlation in identical sub-district are big more, if all identical, then the degree of correlation is 100%, if one also inequality, then are 0.
The degree of correlation shows with the form of space vector, and after the sign grid was determined, in the two-dimensional space of grid, the similitude of the sub-district that relates between grid and the grid reflected that with this two telephone traffics between the grid flow and the quantized value that intersects.
At each time period, according to the involved sub-district of MR that reports in each grid of information record statistics, for example: Grid (I, j), Cell (cell1, cell2, cell3 ...); I, j represents longitude and latitude.
Calculate the degree of correlation of the sub-district that relates in each grid, the span of the degree of correlation is 0-100%.For example, can be the degree of correlation of each adjacent cells and intermediate grid as shown in Figure 5.The last form of expression of space vector is a sparse matrix, has defined the contact amount between the grid, in other words the telephone traffic infiltration capacity.
Step S21: the MR quantity that each time period reports in each grid that counts according to step S12, the variable quantity of the MR quantity that the previous time period of determining current slot described in each adjacent cells reported with respect to previous time period again, wherein, variable quantity can be on the occasion of, negative value or zero.
From database, obtain statistics, determine above-mentioned two time periods, the situation of change of the MR quantity that reports in each adjacent cells.
For example: the variable quantity of the MR quantity that reports with respect to previous time period again at previous time period of each adjacent cells current slot can be as shown in Figure 6.
For example: in the practical operation, can pass through array function realization down:
Int Store_MRVector_to_DB(int Vector_X，int Vector_Y)
{
}
Step S22:, the traffic forecast value of optimised grid is optimized the traffic forecast value after being optimized according to the above-mentioned degree of correlation and the variable quantity that obtain.
Calculate the product of the MR number change amount of the degree of correlation of each adjacent cells correspondence and this grid respectively, and the algebraical sum of all products that calculate.
With the traffic forecast value addition of this optimised grid on the time series that obtains among the algebraical sum that calculates and the step S14, the traffic forecast value after being optimized.Instant empty telephone traffic predicted value.
According to above-mentioned telephone traffic prediction method of the present invention, can make up a kind of traffic forecast device, as shown in Figure 7, comprising: grid determination module 101, quantity statistics module 102 and prediction module 103.
Preferable, grid determination module 101 further can comprise: adjacent area determining unit 1011, metrics calculation unit 1012, longitude and latitude determining unit 1013 and grid determining unit 1014.
Adjacent area determining unit 1011, each that is used for comprising according to MR adjacent Cell Broadcast CB control channel BCCH and base station identity code BSIC obtain LAC and the CI that the Location Area Code LAC that is complementary with BCCH and BSIC and cell ID CI are defined as each adjacent sub-district with the queries static allocation list.
Longitude and latitude determining unit 1013, be used for according to giving the correct time on this MR apart from the longitude and latitude of each adjacent cell base station position that the distance of each adjacent cell base station and LAC and CI by each adjacent sub-district check in, by the triangle location algorithm, the latitude and longitude value that reports the position at the place that obtains giving the correct time on each MR.
Above-mentioned traffic forecast device also comprises: optimal module 104 is used for the traffic forecast value that calculates each grid current slot is optimized.
Preferable, optimal module 104 further can comprise: degree of correlation determining unit 1041, variable quantity determining unit 1042 and optimization unit 1043.
Degree of correlation determining unit 1041, the previous time period that is used for determining current slot is gone up the degree of correlation of each adjacent cells and optimised grid; Wherein, the degree of correlation is determined according to the MR that reports in each grid involved main plot and neighbor cell.
Variable quantity determining unit 1042, the variable quantity of the MR quantity that the previous time period that is used for determining each adjacent cells current slot reported with respect to previous time period again.
Above-mentioned traffic forecast device also comprises: record generation module 105 is used for generating the information record of all MR that each time period of each grid reports; Wherein, comprise main plot, neighbor cell that gives the correct time on each MR and the latitude and longitude value that reports the position in the information record.
Telephone traffic prediction method that provides in the invention process and device by the magnanimity MR that receives is carried out analyzing and processing, determine that it reports position and affiliated grid; Count the MR quantity that reports in each time period in each grid; And, predict the traffic forecast value of each grid current slot by the MR quantity that M time period before current slot known in each grid adjacent reports.The inventive method can be applied in the system of online network load monitoring in real time, and real-time estimate next stage telephone traffic distributes, and has realized more fine granulation, real-time call prediction more accurately on the time series by defining geographical grid and time division section.For example: the zone of 100m * 100m, minute being time granularity of unit etc.
Is inaccurate in this dimension of time series to the pulling factor of prediction data, needs comprehensive a plurality of dimensional space contact pulling factors, finally exports the more accurate prediction value of each grid telephone traffic load.The present invention further provides based on analytic induction mass data two-dimensional space Changing Pattern, optimize the traffic forecast value that obtains by from the time series prediction, realized the traffic forecast that two dimensions in time and space combine, carry out tentative prediction by the given data on the actual sequence, be optimized by the Changing Pattern on the space, realized prediction more accurate, reasonable, science.
And the inventive method both can be provided in the line prediction, promptly can be transplanted in MR collection and the navigation system, and according to the latest network situation, the telephone traffic distribution situation of next time period of real-time calculating is analyzed data characteristics, and the school is inclined to one side automatically; The scene that can also be used for the off-line prediction, highly versatile.
Existing technology usually realizes is spatially cell level (several square kilometres) and BSC rank (tens square kilometres), hour rank traffic forecast in time, and precision can not satisfy the user demand of reality; And can not adjust the data source of forecast model and the parameter of model automatically according to current up-to-date situation.The present invention has overcome above-mentioned shortcoming, realized on the basis of the collection of MR data and location, two dimensions in time and space are fine granularity (space 100m * 100m more, confidence level is greater than 70%, time can arrive 1 minute, precision of prediction is greater than 80%) space-time traffic forecast, and be, thereby bring mobile communications network traffic forecast into a new stage according to the telephone traffic of real-time, the online prediction network of latest network situation of change load.
The above; only be the preferable embodiment of the present invention; but protection scope of the present invention is not limited thereto; anyly be familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily, replace or be applied to other similar devices, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claims.
Claims (14)
1. a telephone traffic prediction method is characterized in that, comprising:
According to the position that reports of the measurement report MR that receives, the grid at the place of determining to give the correct time on the described MR; Described grid is to obtain after in advance the geographic area of telephone traffic to be predicted being divided;
According to calling time on the described MR and the grid at place, count the MR quantity that reports in each time period in each grid;
Obtain the MR quantity that adjacent preceding M the time period of current slot reports in each grid, calculate the traffic forecast value of each grid current slot; Described M is the positive integer greater than 1.
2. the method for claim 1 is characterized in that, the measurement report MR that described basis receives reports the position, and the grid at the place of determining to give the correct time on the described MR comprises:
According to described MR, queries static allocation list, the described latitude and longitude value that reports the position at the place of determining to give the correct time on each described MR;
Inquire about the grid at described latitude and longitude value place, determine the grid at the place of giving the correct time on each described MR.
3. method as claimed in claim 2 is characterized in that, the latitude and longitude value that reports the position at the place of determining to give the correct time on the MR specifically comprises:
According to each adjacent Cell Broadcast CB control channel BCCH and base station identity code BSIC of comprising among the MR, LAC and CI that Location Area Code LAC that is complementary with BCCH and BSIC that the queries static allocation list is obtained and cell ID CI are defined as each adjacent sub-district;
According to the descending level of the adjacent sub-district that comprises among the MR, the algorithm of adopt setting is inquired about described static configuration table, calculates this MR and reports the position to arrive distance apart from each adjacent cell base station;
According to giving the correct time on this MR apart from the longitude and latitude of each adjacent cell base station position that the distance of each adjacent cell base station and LAC and CI by described each adjacent sub-district check in, by the triangle location algorithm, the latitude and longitude value that reports the position at the place that obtains giving the correct time on each MR.
4. method as claimed in claim 3, it is characterized in that, LAC that is complementary with BCCH and BSIC that obtains when the queries static allocation list and CI quantity are during greater than the adjacent number of cells of reality, according to the LAC and the CI of the main plot that comprises among the MR, determine the position wherein LAC of close main plot and LAC and the CI that CI is each adjacent sub-district.
5. method as claimed in claim 4 is characterized in that, described descending level according to the adjacent sub-district that comprises among the MR adopts the algorithm of setting, and inquires about described static configuration table, calculates this MR and reports the position to arrive the distance of each adjacent cell base station, specifically comprises:
LAC and CI according to described each adjacent sub-district, the queries static allocation list, obtain the transmitting power of each adjacent antenna in cell, the absolute value of the transmitting power of calculating each adjacent antenna in cell and the difference of corresponding described descending level obtains the path loss corresponding with each adjacent sub-district;
According to described path loss, the queries static allocation list determines that this MR reports the position to arrive the distance of each adjacent cell base station.
6. the method for claim 1 is characterized in that, also comprises: the traffic forecast value that calculates each grid current slot is optimized;
Optimize the traffic forecast value of current slot in the grid, comprising:
The previous time period of determining described current slot is gone up the degree of correlation of each adjacent cells and optimised grid; The described degree of correlation is determined according to the MR that reports in each grid involved main plot and neighbor cell;
The variable quantity of the MR quantity that the previous time period of determining current slot described in each adjacent cells reported with respect to previous time period again;
According to the described degree of correlation and variable quantity the traffic forecast value of described optimised grid is optimized the traffic forecast value after being optimized.
7. method as claimed in claim 6 is characterized in that, the described previous time period of determining described current slot is gone up the degree of correlation of each adjacent cells and optimised grid, specifically comprises:
Add up the described previous time period and go up the involved sub-district of MR that reports in optimised grid and each adjacent cells, described sub-district comprises main plot and adjacent sub-district;
Calculate the similarity of sub-district involved described in sub-district involved described in each adjacent cells and the optimised grid respectively, obtain the previous time period to go up the degree of correlation of each adjacent cells and optimised grid.
8. method as claimed in claim 6 is characterized in that, and is described according to the described degree of correlation and variable quantity, and the traffic forecast value of described optimised grid is optimized, and the traffic forecast value after being optimized specifically comprises:
Calculate the product of variable quantity of the described MR quantity of the degree of correlation of each adjacent cells correspondence and this grid respectively, and the algebraical sum of all products that calculate;
With described algebraical sum and the addition of described traffic forecast value, the traffic forecast value after being optimized.
9. as the arbitrary described method of claim 1-8, it is characterized in that, after the grid at the place of giving the correct time on described definite described MR, also comprise:
Generate the information record of all MR that each time period reports in each grid; Comprise main plot, neighbor cell that gives the correct time on each MR and the latitude and longitude value that reports the position in the described information record.
10. a traffic forecast device is characterized in that, comprising:
The grid determination module is used for the position that reports according to the measurement report MR that receives, the grid at the place of determining to give the correct time on the described MR; Described grid is to obtain after in advance the geographic area of telephone traffic to be predicted being divided;
The quantity statistics module is used for counting the MR quantity that reports in each time period in each grid according to calling time on the described MR and the grid at place;
Prediction module is used for obtaining before each grid current slot adjacent the MR quantity that M time period reports, and employing time series forecasting algorithm calculates the traffic forecast value of each grid current slot.
11. device as claimed in claim 10 is characterized in that, described grid determination module comprises:
The adjacent area determining unit, each that is used for comprising adjacent Cell Broadcast CB control channel BCCH and base station identity code BSIC according to MR, LAC and CI that Location Area Code LAC that is complementary with BCCH and BSIC that the queries static allocation list is obtained and cell ID CI are defined as each adjacent sub-district;
Metrics calculation unit is used for the descending level of the adjacent sub-district that comprises according to MR, adopts the algorithm of setting, and inquires about described static configuration table, calculates this MR and reports the position to arrive distance apart from each adjacent cell base station;
The longitude and latitude determining unit, be used for according to giving the correct time on this MR apart from the longitude and latitude of each adjacent cell base station position that the distance of each adjacent cell base station and LAC and CI by described each adjacent sub-district check in, by the triangle location algorithm, the latitude and longitude value that reports the position at the place that obtains giving the correct time on each MR;
The grid determining unit is used to inquire about the described grid that reports the latitude and longitude value place of position, determines the grid at the place of giving the correct time on described each MR.
12. device as claimed in claim 10 is characterized in that, also comprises:
Optimal module is used for the traffic forecast value that calculates each grid current slot is optimized.
13. device as claimed in claim 12 is characterized in that, described optimal module comprises:
Degree of correlation determining unit, the previous time period that is used for determining described current slot is gone up the degree of correlation of each adjacent cells and optimised grid; The described degree of correlation is determined according to the MR that reports in each grid involved main plot and neighbor cell;
The variable quantity determining unit, the variable quantity of the MR quantity that the previous time period that is used for determining current slot described in each adjacent cells reported with respect to previous time period again;
Optimize the unit, be used for described traffic forecast value being optimized the traffic forecast value after being optimized according to the described degree of correlation and variable quantity.
14. as the arbitrary described device of claim 10-13, it is characterized in that, also comprise:
The record generation module is used for generating the information record of all MR that each time period of each grid reports; Comprise main plot, neighbor cell that gives the correct time on each MR and the latitude and longitude value that reports the position in the described information record.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102253945A CN101453747B (en) | 2008-10-31 | 2008-10-31 | Telephone traffic prediction method and apparatus |
| PCT/CN2009/001173 WO2010048780A1 (en) | 2008-10-31 | 2009-10-23 | Method and apparatus for traffic prediction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2008102253945A CN101453747B (en) | 2008-10-31 | 2008-10-31 | Telephone traffic prediction method and apparatus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101453747A CN101453747A (en) | 2009-06-10 |
| CN101453747B true CN101453747B (en) | 2010-09-08 |
Family
ID=40735711
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2008102253945A Active CN101453747B (en) | 2008-10-31 | 2008-10-31 | Telephone traffic prediction method and apparatus |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN101453747B (en) |
| WO (1) | WO2010048780A1 (en) |
Families Citing this family (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101453747B (en) * | 2008-10-31 | 2010-09-08 | ***通信集团北京有限公司 | Telephone traffic prediction method and apparatus |
| CN101888640A (en) * | 2010-07-09 | 2010-11-17 | 广州杰赛科技股份有限公司 | Positioning method of city mobile station |
| CN102655654A (en) * | 2011-03-04 | 2012-09-05 | 中兴通讯股份有限公司 | Telephone traffic statistics method and device as well as networking method and device |
| CN103179583B (en) * | 2011-12-26 | 2015-07-29 | ***通信集团设计院有限公司 | A kind of network simulation method and device |
| CN102612071B (en) | 2012-02-24 | 2014-11-05 | 华为技术有限公司 | Method and device for presenting indoor telephone traffic map |
| CN102625351A (en) * | 2012-02-28 | 2012-08-01 | 北京拓明科技有限公司 | Location method of simulation base station based on A interface signal |
| CN103379510B (en) * | 2012-04-23 | 2018-09-25 | 中兴通讯股份有限公司 | A kind of method and device carrying out plot planning using MR data |
| CN103458434B (en) * | 2012-05-30 | 2017-04-12 | 华为技术服务有限公司 | Method and device for determining antenna feeder parameters |
| CN102883352B (en) * | 2012-09-10 | 2015-09-09 | 北京拓明科技有限公司 | Based on the GSM cell parameter optimization method of traffic modeling and traffic forecast |
| CN103906121B (en) * | 2012-12-28 | 2017-11-21 | ***通信集团内蒙古有限公司 | A kind of method and device of localized external interference |
| CN103916887B (en) * | 2012-12-31 | 2017-11-03 | ***通信集团浙江有限公司 | The adjacent area leakage localization method positioned based on grid |
| CN103916868B (en) * | 2012-12-31 | 2017-10-27 | ***通信集团浙江有限公司 | Position area optimization method and system |
| CN104254083B (en) * | 2013-06-28 | 2017-12-22 | 华为技术有限公司 | Predict the method and device of traffic hotspots |
| CN103987056A (en) * | 2014-05-30 | 2014-08-13 | 南京华苏科技有限公司 | Wireless network telephone traffic prediction method based on big-data statistical model |
| CN107005889A (en) * | 2015-05-18 | 2017-08-01 | 华为技术有限公司 | Communication means and equipment |
| CN106255126B (en) * | 2015-06-09 | 2019-10-25 | ***通信集团湖北有限公司 | A kind of method and device of identification base station location area flower arrangement |
| CN107466043B (en) * | 2016-06-03 | 2020-11-27 | ***通信集团河北有限公司 | Method and equipment for determining azimuth angle of base station antenna |
| CN108243466B (en) * | 2016-12-27 | 2020-10-30 | ***通信集团设计院有限公司 | Method and device for identifying low-speed user equipment residing in high-speed rail private network |
| CN107155172A (en) * | 2017-04-01 | 2017-09-12 | 广州逸信电子科技有限公司 | The alignment system and method for mobile subscriber |
| CN108093432B (en) * | 2017-12-26 | 2021-04-02 | ***通信集团江苏有限公司 | User terminal distinguishing method, device, equipment and medium |
| CN108259097B (en) * | 2017-12-29 | 2021-03-09 | 中国通信建设集团设计院有限公司 | Method and device for correcting wireless propagation model based on MR data |
| CN109996241B (en) * | 2017-12-30 | 2022-04-01 | ***通信集团贵州有限公司 | Method, device, equipment and medium for determining information of unmanned aerial vehicle base station lift-off |
| CN110049501B (en) * | 2018-01-15 | 2022-04-15 | 中兴通讯股份有限公司 | Data acquisition method, data acquisition device and computer-readable storage medium |
| CN108495256A (en) * | 2018-03-09 | 2018-09-04 | 深圳市网信联动通信技术股份有限公司 | A kind of method and system measuring tidal effect based on 4/5G mobile communications networks |
| CN109120463B (en) * | 2018-10-15 | 2022-01-07 | 新华三大数据技术有限公司 | Flow prediction method and device |
| CN111278041B (en) * | 2018-12-05 | 2023-04-07 | ***通信集团甘肃有限公司 | Method and equipment for determining group behavior place |
| TWI697863B (en) * | 2018-12-14 | 2020-07-01 | 中華電信股份有限公司 | System and method for platforms construction order |
| CN111385804B (en) * | 2018-12-28 | 2022-06-14 | 中兴通讯股份有限公司 | Cell cluster dividing method and electronic equipment |
| CN110430328B (en) * | 2019-06-26 | 2021-09-03 | 深圳市跨越新科技有限公司 | Method and system for predicting call center telephone traffic based on LightGBM model |
| CN110809284B (en) * | 2019-09-25 | 2022-09-16 | 福建新大陆软件工程有限公司 | Positioning method, system, device and readable storage medium based on MR data |
| CN114095856B (en) * | 2020-07-31 | 2022-09-23 | 中国电信股份有限公司 | Processing method and processing device for saving energy of base station |
| CN113962468A (en) * | 2021-10-29 | 2022-01-21 | 杭州青橄榄网络技术有限公司 | Energy consumption monitoring and statistics-based energy consumption carbon emission management method and system |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100546405C (en) * | 2006-05-29 | 2009-09-30 | 大唐移动通信设备有限公司 | A kind of method that generates adjacent sub-district |
| CA2677087A1 (en) * | 2007-02-05 | 2008-08-14 | Andrew Corporation | System and method for optimizing location estimate of mobile unit |
| CN101159976B (en) * | 2007-08-24 | 2010-11-17 | ***通信集团设计院有限公司 | Method and device of predicting telephone traffic and channel configuration |
| CN101299867B (en) * | 2007-12-27 | 2012-06-20 | 华为技术有限公司 | Method, apparatus and system for detecting network cover |
| CN101453747B (en) * | 2008-10-31 | 2010-09-08 | ***通信集团北京有限公司 | Telephone traffic prediction method and apparatus |
-
2008
- 2008-10-31 CN CN2008102253945A patent/CN101453747B/en active Active
-
2009
- 2009-10-23 WO PCT/CN2009/001173 patent/WO2010048780A1/en active Application Filing
Also Published As
| Publication number | Publication date |
|---|---|
| WO2010048780A1 (en) | 2010-05-06 |
| CN101453747A (en) | 2009-06-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101453747B (en) | Telephone traffic prediction method and apparatus | |
| EP3890361B1 (en) | Cell longitude and latitude prediction method and device, server, base station, and storage medium | |
| CN101646201B (en) | Method, device and system for determining terminal position | |
| CN102111871B (en) | Terminal location method and device based on cell identifier location technology | |
| EP1872150B1 (en) | Mobile location | |
| US7116987B2 (en) | Location estimation of wireless terminals through pattern matching of deduced and empirical signal-strength measurements | |
| US20050014511A1 (en) | Location estimation of wireless terminals through pattern matching of deduced signal strengths | |
| CN102984745B (en) | Combined estimation method for Wi-Fi AP (wireless fidelity access point) position and path loss model | |
| CN103068035A (en) | Wireless network location method, device and system | |
| CN101390309A (en) | Radio mobile unit location system | |
| CN101964981B (en) | Method and device for pre-evaluating telephone traffic and configuring carrier frequency of new base station | |
| CN103108341B (en) | Method and the device of network uplink interference matrix is built based on measurement report data | |
| CN102685762A (en) | Optimizing method and device of antenna feeder parameters | |
| JP7486659B2 (en) | Energy saving processing method and processing device for base station | |
| CN102377494A (en) | Frequency spectrum sensing method, device and system | |
| CN112867147B (en) | Positioning method and positioning device | |
| CN101304595B (en) | Method and apparatus for acquiring frequency plan | |
| EP3657841A1 (en) | Method and device for obtaining emission probability, transfer probability and sequence positioning | |
| CN103906078A (en) | Relation matrix establishment system and method and PCI intelligent optimization system and method | |
| CN115002795A (en) | Beam forming method and device, electronic equipment and readable storage medium | |
| CN114339913B (en) | Neighbor cell updating method and device of wireless network, medium and electronic equipment | |
| CN111385804A (en) | Cell cluster dividing method and electronic equipment | |
| Jaziri et al. | Tracking traffic peaks in mobile networks using statistics of performance metrics | |
| EP3468235B1 (en) | Positioning method and device | |
| CN110708703B (en) | Equipment model selection method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |