CN105744561A - Indoor and outdoor separation method for multi-dimension measurement report - Google Patents

Abstract

The invention discloses an indoor and outdoor separation method for a multi-dimension measurement report. The indoor and outdoor separation method is used for effectively compensating the limitation of an MR grid evaluation method. A multi-dimension indoor and outdoor MR separation method is formed based on an indoor source signal separation method, an outdoor test feature separation method and a user mobility separation method for distinguishing indoor and outdoor MR data, in order to form indoor and outdoor MR grids for evaluating the quality of wireless networks in residential buildings, office buildings, hotels and other indoor scenes and various outdoor road scenes.

Description

Various dimensions measurement report indoor and outdoor separation method

Technical field

The present invention relates to a kind of various dimensions measurement report indoor and outdoor separation method.

Background technology

Measurement report MR is that user initiates business and the measurement report of generation in the communication process of base station, it is radio environment measurements data round-the-clock, all the period of time, it it is the foundation of wireless resource scheduling, direct reaction quality of service and user's perception, it possesses mass data feature, and sample variance is little, and accuracy is high, and the advantage such as procurement cost is low, therefore MR is the effective means instructing operator's planning, building and optimize.Traditional appraisal procedure have (1) community and MR evaluation measures, namely by adding up the coverage rate of community, matter rate, up-downgoing receives the index such as unbalanced power ratio and evaluates the quality of community.(2) by MR grid ground physics and chemistry, the quality of wireless network of each grid is geographically evaluated with less granularity.

(1) the MR appraisal procedure of cell-level can only determine the quality of wireless network of current area, and the coverage of Dan Yige community is very wide, it is impossible to orients which position signalling concrete bad.

(2) although MR rasterizing appraisal procedure can evaluate the geographical position quality of wireless network more refined, but have certain error owing to current MR specifically positions longitude and latitude, so carry out the quality of wireless network in the degree of depth or street in agent's room with concrete grid can there is inaccurate situation.

Summary of the invention

It is an object of the invention to overcome the deficiencies in the prior art, a kind of various dimensions measurement report indoor and outdoor separation method is provided, by based on sub-signal source, room partition method, based on outdoor test feature partition method, based on user mobility partition method, form various dimensions indoor and outdoor MR separation method in order to MR data inside and outside divided chamber, and then form indoor and outdoor MR grid, in order to assess the quality of wireless network of the indoor scenes such as residential building, office building, hotel and various outdoor road scene.

It is an object of the invention to be achieved through the following technical solutions: various dimensions measurement report indoor and outdoor separation method, it includes following sub-step:

S1: data acquisition, divides basic data and GIS data by acquisition interface collection MR data, drive test data, room from server；

S2: MR data be carried out, removes invalid data and noise；

S3: MR data carrying out room divided data and separates calculating, the room calculating MR data to be detected splits reliability a, it is judged that room splits the relation of reliability a and room point threshold value A, if room splits reliability a less than room point threshold value A, then separate and terminate, jump procedure S6, if not then jump procedure S4；

S4: MR data are carried out user mobility and separates calculating, calculate the mobility confidence level b of MR data to be detected, judge the relation of indoor and outdoor separation confidence level b and threshold value thre, if indoor and outdoor separation confidence level b is higher than threshold value thre1, and less than threshold value thre2, then separate and terminate, jump procedure S6, otherwise jump procedure S5；

S5: MR data are carried out outdoor test feature and separates calculating, calculate the test feature confidence level of MR data to be detected, test feature confidence level c input neural network grader is judged, then separate and terminate, jump procedure S6, if can not be easily separated, it is impossible to the MR data of differentiation discard；

S6: export current MR separating resulting, namely output MR belongs to the result of indoor or outdoors；Described threshold value A value is 432, and the value of described threshold value thre1 is 1.2, and the value of described threshold value thre2 is 0.8.

Room divided data separation calculating in described step S3 includes following sub-step:

S31: the MR data collected are carried out segmentation by the session of IMSI；

S32: obtain main Serving cell Cell1 and each adjacent area { Cell2, the Cell3 ... CellN} that detect according to current MR data；

S33: will set { Cell1, Cell2 ... CellN} and divided data storehouse, room match, if CellN is branch website, room, then put into set { Celli1, Celli2 ... CelliN}；

S34: obtain longitude and latitude track { { lon1, lat1}, { lon2, the lat2} ... { lonn, latn}} of single user according to step S21；Obtain longitude and latitude { { hlon1, hlat1}, { hlon2, the hlat2} ... { hlon, hlatm}} of branch website, each room, divided data storehouse, room；

S35: according to step S34, obtain user trajectory and nearest branch website, room distance set d1, d2 ... dn}, then calculate average distance avgdist=(d1+d2+ ... the dn)/n of user and branch website, room；

S36: if average distance avgdist is less than a thresholding A, and the set of step S33 Celli1, Celli2 ... CelliN} is not empty set, then active user's session produce MR be positioned at indoor, otherwise enter step S4 determine whether.

User mobility separation calculating in described step S4 includes following sub-step:

S41: the MR data collected are divided into grid by N*N rice in morning in the evening, render in different grids by longitude and latitude；

S42: longitude and latitude track { { lon1, the lat1}, { lon2 of each user in each grid in traversal step S41, lat2} ... { lonn, latn}}, obtain the speed set { vk1 of each user, vk2 ... vkn} and variance, in this, as each grid mobility in evening feature, as grader；

S43: by collect daytime MR data by IMSI and user conversation segmentation, as detection object；

S44: obtain longitude and latitude track { { lon1, la1}, { lon2, the la2} ... { lonn, lan}} calculate its speed { v1, v2 ... vk} and variance of single user according to S43；

S45: the speed average in S44 and variance are inputted in the grader in S42, exports the indoor and outdoor separation confidence level b of each user；

S46: if indoor and outdoor separation confidence level b is more than a thresholding thre1, then judge that the MR of active user's session generation on daytime is as indoor, if indoor and outdoor separation confidence level b less than a thresholding thre2, then judges that MR that user conversation produces is as outdoor, otherwise carries out the judgement of step S5.

Outdoor test feature separation calculating in described step S5 includes following sub-step:

S51: by N*N rice, the DT/CQT data of outdoor test are divided into grid, renders in different grids by longitude and latitude, forms M grid altogether, and each grid determines its center longitude { { lon1, la1}, { lon2, la2} ... { lonm, lam}}

S52: choose BP artificial neural network algorithm and set up disaggregated model, using the first six strong RSCP, EC/IO value of each sampled point of each DT/CQT grid as input layer parameter, indoor and outdoors is represented as output layer parameter with 1 or 0, each DT/CQT grid is trained to a grader, total M grader；

S53: by collect daytime MR data by IMSI and user conversation segmentation, as detection object；

S54: by N*N rice, the MR of each session of each user is divided into grid, renders in different grids by longitude and latitude, forms K grid altogether, and each grid determines its center longitude { mrlon1, mrla1}, { mrlon2, mrla2} ... { mrlonk, mrlak}；

S55: find out the neural network classifier that in step S54, each MR grid is corresponding according to minimum distance；

S56: travel through the MR grid in each S54, extracts every MR in MR grid, obtains the first six strong RSCP and EC/NO 12 input parameters as grader；

S57: 12 parameters be input in the neural network classifier that current MR grid is corresponding, obtains every MR indoor and outdoor attribute after computing.

In described step S35, the computational methods of the average distance of user and branch website, room include: set the longitude and latitude of Target cell A as (LonA, LatA), the longitude and latitude of Target cell B is (LonB, LatB), benchmark according to 0 degree of warp, east longitude degree of learning from else's experience on the occasion of, west longitude degree of learning from else's experience negative value, north latitude takes 90-latitude value, south latitude takes 90+ latitude value, then 2 after above-mentioned process are counted as (MLonA, MLatA) and (MLonB, MLatB)；Derive according to triangle, it is possible to obtain calculating the equation below of 2 distances:

C=sin (MLatA) * sin (MLatB) * cos (MLonA-MLonB)+cos (MLatA) * cos (MLatB)

Dist=R*Accros (C) * Pi/180

Dist is the distance of Target cell and abutting subdistrict；

The method calculating user trajectory and branch website, arbitrary room average distance:

$avgdisti=\underset{1}{\overset{n}{\Σ}}(R*Accros(Ci)*Pi/180)/n$

Ci=sin (lati) * sin (hlatk) * cos (loni-hlonk)+cos (lati) * cos (hlatk) (1≤i≤n)

In above formula, k value represents kth room and divides, and Ci represents the distance of ith sample point and room minute k in user trajectory, and avgdist represents the average distance of whole user trajectory and branch website, kth room；

According to the distance that user trajectory and arbitrary room are divided, then can obtain minimum distance is:

Minavgdist=min (avgidst1, avgdist2......avgdistk).

In described step S51, DT/CQT test sample point grid is thrown in and is included following sub-step:

S511: longitude and latitude is uniformly processed, retains 5 decimals, if figure place is more than 5 after the longitude and latitude arithmetic point collected, then dispenses the figure place after the 5th, if less than 5, then mending 0 below；

S512: determine the grid identifier of every DT/CQT data, for 50 meters of * 50 meters of grids, after intercepting longitude and latitude arithmetic point, the 4th span being assumed to be a, a is 0≤a≤9, if a≤5, it is 0 that a unifies value；If a > 5, then a value is 5；By first 7 of longitude, first 6 be added with latitude are added the grid identifier namely obtaining every DT/CQT with a with a；

S513: determine the grid identifier of each DT/CQT sampled point, then determines that each sampled point is corresponding with which grid again, and the grid namely completing DT/CQT test sample point is thrown in.

In described step S52, BP artificial neural network algorithm is set up the method for disaggregated model and is included following sub-step:

S521: data normalization processes, and normalization algorithm is:

Y=(x-min)/(max-min) (0≤y≤1)

X in above formula is RSCP or the Ec/No value of concrete MR sampled point N pilot tone, and min is the minimum value of RSCP or Ec/No, respectively-112 and-24；Max is the maximum occurrences of RSCP or Ec/No, respectively-40 and-1, and y is the value after normalization, is mapped as the span of 0 to 1；

S522: initializing input layer to the weight array iptHidWeights [12] [6] of hidden layer and hidden layer to the weight array hidOptWeights [6] [2] of output layer, initial value utilizes random function to generate；

S523: obtain input layer and propagate to the weighted value of hidden layer and hidden layer to the weighted value of output layer, such as following formula:

$neti=\underset{j=1}{\overset{n}{\Σ}}wijxij-\mathrm{\θ}$

Yi=f (neti)

$f\left(x\right)=\frac{1}{1+e-ax}$

In above formula, wij is input layer to hidden layer and hidden layer to the weighted value of output layer, θ represents that a threshold value acquiescence takes 0, xij is i-th input layer to the input value of jth hidden layer or i-th hidden layer to the input value of jth output layer, and yi is the output valve of neuron i；F (x) is transfer function；The indoor and outdoor attribute output of each MR can be obtained by above-mentioned three formulas；

S524: according to the weight array iptHidWeights [12] [6] of the actual output in theoretical output calibration step S423 and hidOptWeights [6] [2], continuous iterative learning.

The invention has the beneficial effects as follows: the invention provides a kind of various dimensions measurement report indoor and outdoor separation method, effectively compensate for the limitation of MR rasterizing appraisal procedure, by based on sub-signal source, room partition method, based on outdoor test feature partition method, based on user mobility partition method, form various dimensions indoor and outdoor MR separation method in order to MR data inside and outside divided chamber, and then form indoor and outdoor MR grid, in order to assess the quality of wireless network of the indoor scenes such as residential building, office building, hotel and various outdoor road scene.

Accompanying drawing explanation

Fig. 1 is various dimensions measurement report indoor and outdoor separation method schematic diagrams；

Fig. 2 is room divided data separation computational methods schematic diagram；

Fig. 3 is user mobility separation computational methods schematic diagram；

Fig. 4 is outdoor test feature separation computational methods schematic diagram.

Detailed description of the invention

Below in conjunction with accompanying drawing, technical scheme is described in further detail, but protection scope of the present invention is not limited to the following stated.

As it is shown in figure 1, various dimensions measurement report indoor and outdoor separation method, it includes following sub-step:

S1: data acquisition, divides basic data and GIS data by acquisition interface collection MR data, drive test data, room from server；The MR data on daytime are collected as detection；Using evening MR data, drive test data, room divided data is as training set；

S2: MR data be carried out, removes invalid data and noise；

S3: MR data carrying out room divided data and separates calculating, the room calculating MR data to be detected splits reliability a, it is judged that room splits the relation of reliability a and room point threshold value A, if room splits reliability a less than room point threshold value A, then separate and terminate, jump procedure S6, if not then jump procedure S4；

S4: MR data are carried out user mobility and separates calculating, calculate the mobility confidence level b of MR data to be detected, judge the relation of indoor and outdoor separation confidence level b and threshold value thre, if indoor and outdoor separation confidence level b is higher than threshold value thre1, and less than threshold value thre2, then separate and terminate, jump procedure S6, otherwise jump procedure S5；

S5: MR data are carried out outdoor test feature and separates calculating, calculate the test feature confidence level of MR data to be detected, test feature confidence level c input neural network grader is judged, then separate and terminate, jump procedure S6, if can not be easily separated, it is impossible to the MR data of differentiation discard；

S6: export current MR separating resulting, namely output MR belongs to the result of indoor or outdoors；

Described threshold value A value is 432, and the value of described threshold value thre1 is 1.2, and the value of described threshold value thre2 is 0.8.

As in figure 2 it is shown, the room divided data separation calculating in described step S3 includes following sub-step:

S31: the MR data collected are carried out segmentation by the session of IMSI；

S32: obtain main Serving cell Cell1 and each adjacent area { Cell2, the Cell3 ... CellN} that detect according to current MR data；

S33: will set { Cell1, Cell2 ... CellN} and divided data storehouse, room match, if CellN is branch website, room, then put into set { Celli1, Celli2 ... CelliN}；

S34: obtain longitude and latitude track { { lon1, lat1}, { lon2, the lat2} ... { lonn, latn}} of single user according to step S21；Obtain longitude and latitude { { hlon1, hlat1}, { hlon2, the hlat2} ... { hlon, hlatm}} of branch website, each room, divided data storehouse, room；

S35: according to step S34, obtain user trajectory and nearest branch website, room distance set d1, d2 ... dn}, then calculate average distance avgdist=(d1+d2+ ... the dn)/n of user and branch website, room；

S36: if average distance avgdist is less than a thresholding A, and the set of step S33 Celli1, Celli2 ... CelliN} is not empty set, then active user's session produce MR be positioned at indoor, otherwise enter step S4 determine whether.

As it is shown on figure 3, the user mobility separation calculating in described step S4 includes following sub-step:

S41: the MR data collected are divided into grid by N*N rice in morning in the evening, render in different grids by longitude and latitude；

S42: longitude and latitude track { { lon1, the lat1}, { lon2 of each user in each grid in traversal step S41, lat2} ... { lonn, latn}}, obtain the speed set { vk1 of each user, vk2 ... vkn} and variance, in this, as each grid mobility in evening feature, as grader；

S43: by collect daytime MR data by IMSI and user conversation segmentation, as detection object；

S44: obtain longitude and latitude track { { lon1, la1}, { lon2, the la2} ... { lonn, lan}} calculate its speed { v1, v2 ... vk} and variance of single user according to S43；

S45: the speed average in S44 and variance are inputted in the grader in S42, exports the indoor and outdoor separation confidence level b of each user；

S46: if indoor and outdoor separation confidence level b is more than a thresholding thre1, then judge that the MR of active user's session generation on daytime is as indoor, if indoor and outdoor separation confidence level b less than a thresholding thre2, then judges that MR that user conversation produces is as outdoor, otherwise carries out the judgement of step S5.

As shown in Figure 4, the outdoor test feature separation calculating in described step S5 includes following sub-step:

S51: by N*N rice, the DT/CQT data of outdoor test are divided into grid, renders in different grids by longitude and latitude, forms M grid altogether, and each grid determines its center longitude { { lon1, la1}, { lon2, la2} ... { lonm, lam}}

S52: choose BP artificial neural network algorithm and set up disaggregated model, using the first six strong RSCP, EC/IO value of each sampled point of each DT/CQT grid as input layer parameter, indoor and outdoors is represented as output layer parameter with 1 or 0, each DT/CQT grid is trained to a grader, total M grader；

S53: by collect daytime MR data by IMSI and user conversation segmentation, as detection object；

S54: by N*N rice, the MR of each session of each user is divided into grid, renders in different grids by longitude and latitude, forms K grid altogether, and each grid determines its center longitude { mrlon1, mrla1}, { mrlon2, mrla2} ... { mrlonk, mrlak}；

S55: find out the neural network classifier that in step S54, each MR grid is corresponding according to minimum distance；

S56: travel through the MR grid in each S54, extracts every MR in MR grid, obtains the first six strong RSCP and EC/NO 12 input parameters as grader；

S57: 12 parameters be input in the neural network classifier that current MR grid is corresponding, obtains every MR indoor and outdoor attribute after computing.

The earth is the spheroid of an intimate standard, and its equatorial radius is 6378.140 kms, and polar radius is 6356.755 kms, mean radius 6371.004 km.If with 0 degree of warp for benchmark, then just can calculate the surface distance (ignoring the error that calculating is brought by earth surface landform here, be only theoretic estimated value) of this point-to-point transmission according to the longitude and latitude of earth surface any two points.In described step S35, the computational methods of the average distance of user and branch website, room include: set the longitude and latitude of Target cell A as (LonA, LatA), the longitude and latitude of Target cell B is (LonB, LatB), benchmark according to 0 degree of warp, east longitude degree of learning from else's experience on the occasion of, west longitude degree of learning from else's experience negative value, north latitude takes 90-latitude value, south latitude takes 90+ latitude value, then 2 after above-mentioned process are counted as (MLonA, MLatA) and (MLonB, MLatB)；Derive according to triangle, it is possible to obtain calculating the equation below of 2 distances:

C=sin (MLatA) * sin (MLatB) * cos (MLonA-MLonB)+cos (MLatA) * cos (MLatB)

Dist=R*Accros (C) * Pi/180

Dist is the distance of Target cell and abutting subdistrict；

The method calculating user trajectory and branch website, arbitrary room average distance:

$avgdisti=\underset{1}{\overset{n}{\Σ}}(R*Accros(Ci)*Pi/180)/n$

Ci=sin (lati) * sin (hlatk) * cos (loni-hlonk)+cos (lati) * cos (hlatk) (1≤i≤n)

In above formula, k value represents kth room and divides, and Ci represents the distance of ith sample point and room minute k in user trajectory, and avgdist represents the average distance of whole user trajectory and branch website, kth room；

According to the distance that user trajectory and arbitrary room are divided, then can obtain minimum distance is:

Minavgdist=min (avgidst1, avgdist2......avgdistk).

In described step S51, DT/CQT test sample point grid is thrown in and is included following sub-step:

S511: longitude and latitude is uniformly processed, retains 5 decimals, if figure place is more than 5 after the longitude and latitude arithmetic point collected, then dispenses the figure place after the 5th, if less than 5, then mending 0 below；

S512: determine the grid identifier of every DT/CQT data, for 50 meters of * 50 meters of grids, after intercepting longitude and latitude arithmetic point, the 4th span being assumed to be a, a is 0≤a≤9, if a≤5, it is 0 that a unifies value；If a > 5, then a value is 5；By first 7 of longitude, first 6 be added with latitude are added the grid identifier namely obtaining every DT/CQT with a with a, for instance: the grid identifier of 50 meters * 50 meters of (106.34562,29.38127) is (106.3455,29.3810).

S513: determine the grid identifier of each DT/CQT sampled point, then determines that each sampled point is corresponding with which grid again, and the grid namely completing DT/CQT test sample point is thrown in.

In described step S52, BP artificial neural network algorithm is set up the method for disaggregated model and is included following sub-step:

S521: data normalization processes, and normalization algorithm is:

Y=(x-min)/(max-min) (0≤y≤1)

X in above formula is RSCP or the Ec/No value of concrete MR sampled point N pilot tone, and min is the minimum value of RSCP or Ec/No, respectively-112 and-24；Max is the maximum occurrences of RSCP or Ec/No, respectively-40 and-1, and y is the value after normalization, is mapped as the span of 0 to 1；

S522: initializing input layer to the weight array iptHidWeights [12] [6] of hidden layer and hidden layer to the weight array hidOptWeights [6] [2] of output layer, initial value utilizes random function to generate；

S523: obtain input layer and propagate to the weighted value of hidden layer and hidden layer to the weighted value of output layer, such as following formula:

$neti=\underset{j=1}{\overset{n}{\Σ}}wijxij-\mathrm{\θ}$

Yi=f (neti)

$f\left(x\right)=\frac{1}{1+e-ax}$

In above formula, wij is input layer to hidden layer and hidden layer to the weighted value of output layer, θ represents that a threshold value acquiescence takes 0, xij is i-th input layer to the input value of jth hidden layer or i-th hidden layer to the input value of jth output layer, and yi is the output valve of neuron i；F (x) is transfer function；The indoor and outdoor attribute output of each MR can be obtained by above-mentioned three formulas；

S524: according to the weight array iptHidWeights [12] [6] of the actual output in theoretical output calibration step S423 and hidOptWeights [6] [2], continuous iterative learning.

The present invention is by based on sub-signal source, room partition method, based on outdoor test feature partition method, based on user mobility partition method, form various dimensions indoor and outdoor MR separation method in order to MR data inside and outside divided chamber, and then form indoor and outdoor MR grid, in order to assess the quality of wireless network of the indoor scenes such as residential building, office building, hotel and various outdoor road scene.

Claims (7)

1. various dimensions measurement report indoor and outdoor separation method, it is characterised in that: it includes following sub-step:

S1: data acquisition, divides basic data and GIS data by acquisition interface collection MR data, drive test data, room from server；

S2: MR data be carried out, removes invalid data and noise；

S3: MR data carrying out room divided data and separates calculating, the room calculating MR data to be detected splits reliability a, it is judged that room splits the relation of reliability a and room point threshold value A, if room splits reliability a less than room point threshold value A, then separate and terminate, jump procedure S6, if not then jump procedure S4；

S4: MR data are carried out user mobility and separates calculating, calculate the mobility confidence level b of MR data to be detected, judge the relation of indoor and outdoor separation confidence level b and threshold value thre, if indoor and outdoor separation confidence level b is higher than threshold value thre1, and less than threshold value thre2, then separate and terminate, jump procedure S6, otherwise jump procedure S5；

S5: MR data are carried out outdoor test feature and separates calculating, calculate the test feature confidence level of MR data to be detected, test feature confidence level c input neural network grader is judged, then separate and terminate, jump procedure S6, if can not be easily separated, it is impossible to the MR data of differentiation discard；

S6: export current MR separating resulting, namely output MR belongs to the result of indoor or outdoors；

Described threshold value A value is 432, and the value of described threshold value thre1 is 1.2, and the value of described threshold value thre2 is 0.8.

2. various dimensions measurement report indoor and outdoor separation method according to claim 1, it is characterised in that: the room divided data separation calculating in described step S3 includes following sub-step:

S31: the MR data collected are carried out segmentation by the session of IMSI；

S32: obtain main Serving cell Cell1 and each adjacent area { Cell2, the Cell3 ... CellN} that detect according to current MR data；

S33: will set { Cell1, Cell2 ... CellN} and divided data storehouse, room match, if CellN is branch website, room, then put into set { Celli1, Celli2 ... CelliN}；

S34: obtain longitude and latitude track { { lon1, lat1}, { lon2, the lat2} ... { lonn, latn}} of single user according to step S21；Obtain longitude and latitude { { hlon1, hlat1}, { hlon2, the hlat2} ... { hlon, hlatm}} of branch website, each room, divided data storehouse, room；

S35: according to step S34, obtain user trajectory and nearest branch website, room distance set d1, d2 ... dn}, then calculate average distance avgdist=(d1+d2+ ... the dn)/n of user and branch website, room；

S36: if average distance avgdist is less than a thresholding A, and the set of step S33 Celli1, Celli2 ... CelliN} is not empty set, then active user's session produce MR be positioned at indoor, otherwise enter step S4 determine whether.

3. various dimensions measurement report indoor and outdoor separation method according to claim 1, it is characterised in that: the user mobility separation calculating in described step S4 includes following sub-step:

S41: the MR data collected are divided into grid by N*N rice in morning in the evening, render in different grids by longitude and latitude；

S42: longitude and latitude track { { lon1, the lat1}, { lon2 of each user in each grid in traversal step S41, lat2} ... { lonn, latn}}, obtain the speed set { vk1 of each user, vk2 ... vkn} and variance, in this, as each grid mobility in evening feature, as grader；

S43: by collect daytime MR data by IMSI and user conversation segmentation, as detection object；

S44: obtain longitude and latitude track { { lon1, la1}, { lon2, the la2} ... { lonn, lan}} calculate its speed { v1, v2 ... vk} and variance of single user according to S43；

S45: the speed average in S44 and variance are inputted in the grader in S42, exports the indoor and outdoor separation confidence level b of each user；

S46: if indoor and outdoor separation confidence level b is more than a thresholding thre1, then judge that the MR of active user's session generation on daytime is as indoor, if indoor and outdoor separation confidence level b less than a thresholding thre2, then judges that MR that user conversation produces is as outdoor, otherwise carries out the judgement of step S5.

4. various dimensions measurement report indoor and outdoor separation method according to claim 1, it is characterised in that: the outdoor test feature separation calculating in described step S5 includes following sub-step:

S51: by N*N rice, the DT/CQT data of outdoor test are divided into grid, renders in different grids by longitude and latitude, forms M grid altogether, and each grid determines its center longitude { { lon1, la1}, { lon2, la2} ... { lonm, lam}}

S52: choose BP artificial neural network algorithm and set up disaggregated model, using the first six strong RSCP, EC/IO value of each sampled point of each DT/CQT grid as input layer parameter, indoor and outdoors is represented as output layer parameter with 1 or 0, each DT/CQT grid is trained to a grader, total M grader；

S53: by collect daytime MR data by IMSI and user conversation segmentation, as detection object；

S54: by N*N rice, the MR of each session of each user is divided into grid, renders in different grids by longitude and latitude, forms K grid altogether, and each grid determines its center longitude { mrlon1, mrla1}, { mrlon2, mrla2} ... { mrlonk, mrlak}；

S55: find out the neural network classifier that in step S54, each MR grid is corresponding according to minimum distance；

S56: travel through the MR grid in each S54, extracts every MR in MR grid, obtains the first six strong RSCP and EC/NO 12 input parameters as grader；

S57: 12 parameters be input in the neural network classifier that current MR grid is corresponding, obtains every MR indoor and outdoor attribute after computing.

5. various dimensions measurement report indoor and outdoor separation method according to claim 2, it is characterized in that: in described step S35, the computational methods of the average distance of user and branch website, room include: set the longitude and latitude of Target cell A as (LonA, LatA), the longitude and latitude of Target cell B is (LonB, LatB), benchmark according to 0 degree of warp, east longitude degree of learning from else's experience on the occasion of, west longitude degree of learning from else's experience negative value, north latitude takes 90-latitude value, and south latitude takes 90+ latitude value, then 2 after above-mentioned process are counted as (MLonA, and (MLonB, MLatB) MLatA)；Derive according to triangle, it is possible to obtain calculating the equation below of 2 distances:

C=sin (MLatA) * sin (MLatB) * cos (MLonA-MLonB)+cos (MLatA) * cos (MLatB)

Dist=R*Accros (C) * Pi/180

Dist is the distance of Target cell and abutting subdistrict；

The method calculating user trajectory and branch website, arbitrary room average distance:

$avgdisti=\underset{1}{\overset{n}{\Σ}}(R*Accros(Ci)*Pi/180)/n$

Ci=sin (lati) * sin (hlatk) * cos (loni-hlonk)+cos (lati) * cos (hlatk) (1≤i≤n)

In above formula, k value represents kth room and divides, and Ci represents the distance of ith sample point and room minute k in user trajectory, and avgdist represents the average distance of whole user trajectory and branch website, kth room；

According to the distance that user trajectory and arbitrary room are divided, then can obtain minimum distance is:

Minavgdist=min (avgidst1, avgdist2......avgdistk).

6. various dimensions measurement report indoor and outdoor separation method according to claim 4, it is characterised in that: in described step S51, DT/CQT test sample point grid is thrown in and is included following sub-step:

S511: longitude and latitude is uniformly processed, retains 5 decimals, if figure place is more than 5 after the longitude and latitude arithmetic point collected, then dispenses the figure place after the 5th, if less than 5, then mending 0 below；

S512: determine the grid identifier of every DT/CQT data, for 50 meters of * 50 meters of grids, after intercepting longitude and latitude arithmetic point, the 4th span being assumed to be a, a is 0≤a≤9, if a≤5, it is 0 that a unifies value；If a > 5, then a value is 5；By first 7 of longitude, first 6 be added with latitude are added the grid identifier namely obtaining every DT/CQT with a with a；

S513: determine the grid identifier of each DT/CQT sampled point, then determines that each sampled point is corresponding with which grid again, and the grid namely completing DT/CQT test sample point is thrown in.

7. various dimensions measurement report indoor and outdoor separation method according to claim 4, it is characterised in that: in described step S52, BP artificial neural network algorithm is set up the method for disaggregated model and is included following sub-step:

S521: data normalization processes, and normalization algorithm is:

Y=(x-min)/(max-min) (0≤y≤1)

X in above formula is RSCP or the Ec/No value of concrete MR sampled point N pilot tone, and min is the minimum value of RSCP or Ec/No, respectively-112 and-24；Max is the maximum occurrences of RSCP or Ec/No, respectively-40 and-1, and y is the value after normalization, is mapped as the span of 0 to 1；

S522: initializing input layer to the weight array iptHidWeights [12] [6] of hidden layer and hidden layer to the weight array hidOptWeights [6] [2] of output layer, initial value utilizes random function to generate；

S523: obtain input layer and propagate to the weighted value of hidden layer and hidden layer to the weighted value of output layer, such as following formula:

$neti=\underset{j=1}{\overset{n}{\Σ}}wijxij-\mathrm{\θ}$

Yi=f (neti)

$f\left(x\right)=\frac{1}{1+e-ax}$

In above formula, wij is input layer to hidden layer and hidden layer to the weighted value of output layer, θ represents that a threshold value acquiescence takes 0, xij is i-th input layer to the input value of jth hidden layer or i-th hidden layer to the input value of jth output layer, and yi is the output valve of neuron i；F (x) is transfer function；The indoor and outdoor attribute output of each MR can be obtained by above-mentioned three formulas；

S524: according to the weight array iptHidWeights [12] [6] of the actual output in theoretical output calibration step S523 and hidOptWeights [6] [2], continuous iterative learning.