CN102209385A - Method for calculating position of base station based on spatial outlier data mining algorithm - Google Patents
Method for calculating position of base station based on spatial outlier data mining algorithm Download PDFInfo
- Publication number
- CN102209385A CN102209385A CN2011101370784A CN201110137078A CN102209385A CN 102209385 A CN102209385 A CN 102209385A CN 2011101370784 A CN2011101370784 A CN 2011101370784A CN 201110137078 A CN201110137078 A CN 201110137078A CN 102209385 A CN102209385 A CN 102209385A
- Authority
- CN
- China
- Prior art keywords
- base station
- data
- point
- algorithm
- distance
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides a method for calculating the position of a base station based on a spatial outlier data mining algorithm. The method comprises the following steps of: firstly, calculating a data set center for a certain data set T to be processed, wherein the geometric center, the center of gravity and the like of the data set can be selected during determination of the data center; secondly, expressing a distance from an ith point to the data set center with Di and recording D as an average value of Di; and finally, recording the points of which the absolute value of D(k) minus P(k) exceeds a certain threshold value as outlier points. Compared with a K-nearest neighbor algorithm, the algorithm has the advantages that: only the distance from each point to the data set center is calculated, and the step of scanning the entire data set every time when a point is calculated is saved; and the algorithm is more suitable for extracting the outlier points in the data set by a base station. The method disclosed by the invention is more suitable for the mining algorithm of outlier data points of the base station, so that the base station is more accurately and efficiently positioned. Only the distance from each point to the data set center is calculated, and the step of scanning the entire data set every time when a point is calculated is saved, so that the algorithm is more suitable for acquiring the outlier points in the data set by the extraction base station.
Description
Technical field
The present invention relates to the base station positioning field, relate generally to a kind of based on space outlier data digging algorithm and according to the method for calculation base station position, base station data collection point.
Background technology
Location-based service (Location Based Service, LBS), it is the positional information (geographical coordinate that obtains mobile phone users by the radio communication network of telecommunications mobile operator (as GSM net, CDMA net) or outside locate mode (as GPS), or geodetic coordinates), at GIS-Geographic Information System (GIS, Geographic Information System) under the support of platform, provides a kind of value-added service of respective service for the user.Therefore wireless location technology is key technology among the LBS, also is simultaneously to solve enterprise to move the effective means of using as comprehensive location requirement (comprising indoor, outdoor) in the mobile crm system.
In the mobile communication network navigation system, be to position according to the base station location that terminal inserts.Therefore in the position fixing process base station location to obtain be the basis.
The purpose of space outlier data digging is to find out and lies in abnormal data sparse and isolated relatively in the mass data, and promptly non-space attribute and other objects have visibly different spatial object in the spatial neighborhood; And they are local instabilities, even for not unusual on the whole, but other contiguous objects are had extreme value.The also definition generally adopted of neither one of outlier at present, Hawkins has disclosed the essence of outlier in a sense to the definition that peels off: " outlier is so different with other points, to such an extent as to allow the people suspect that they are to be produced by the different mechanism of another one ".
Spatial data has space attribute and two kinds of attributes of non-space attribute; Shekhar etc. are in the definition of space outlier; define neighborhood relationships with space attribute; with non-space attribute definition distance function; this definition meets the general thinking of GIS; but occurring with similar non-space attribute through regular meeting in geographical phenomenon is the situation of adjacency; therefore can define and the excavated space outlier from an opposite angle; promptly define neighborhood relationships, define distance function with space attribute with the non-space attribute.In other words, the space outlier be with its non-space attribute neighborhood in the very significant spatial object of other spatial objects difference on the locus.
In recent years, the researcher has proposed a large amount of detection algorithms that peels off, and roughly can reduce following a few class to them: based on the method for statistics, based on the method for density, based on the method for the degree of depth, based on the method for distance with based on the method that departs from.This paper mainly adopts based on the method for distance and removes the data that peel off in the base station data, utilizes remaining data calculation base station position then.
Outlier based on distance is proposed by Knorr and Ng the earliest, they regard record as in the higher dimensional space point, outlier is defined as distance between data centralization and the great majority points all greater than the point of certain threshold value, usually be described to DB (pct, dmin), a record O is called outlier among the data set T, has the distance of pct partial data and O at least greater than dmin among and if only if the data set T.Change a kind of angle and consider, note M=N * (1-pct), peel off to detect and whether promptly judge with some O distance less than the point of dmin more than M.If then O is not an outlier, otherwise O is an outlier.
Rastogi ﹠amp; Ramaswamy is in the above on the basis based on the definition of the outlier of distance, proposes improved k-arest neighbors (k-NN) based on the distance detection algorithm that peels off.Distance with k the nearest neighbor point point of Dk (P) expression point P at first calculates the k-nearest neighbor distance of being had a few among the data set T, according to value big or small then descending, and the algorithm handle comes top n0 point and is labeled as n0 outlier.A major defect of this algorithm is to calculate the Dk (P) that is had a few, and the Dk (P) of a point of every calculating will the run-down data set, and for large data sets, its I/O time number usually makes that the computational efficiency of algorithm is very low.
Summary of the invention
In order to address the above problem, the present invention is according to the characteristics that collect same base station data, k-arest neighbors (k-NN) detection algorithm that peels off is improved, a kind of method based on space outlier data digging algorithm computation base station location has been proposed, be more suitable for the mining algorithm of base station data point that peels off, thereby make the location precise and high efficiency more of base station.
Core concept of the present invention is: for a certain pending data set T, at first, calculate the data set center.The geometric center of determining to choose data set of data center, center of gravity etc.Secondly, use D
iRepresent that i is put the distance at data set center, note D is D
iAverage, last, with those | D (k)-P (k) | the point that surpasses certain threshold value is designated as outlier.With respect to the K-nearest neighbor algorithm, this algorithm only calculates each distance of putting the data set center, has saved the step that point of every calculating is all wanted the whole data set of run-down, and this algorithm is more suitable in extracting the outlier that base stations is concentrated.
Technical scheme of the present invention is: a kind of method based on space outlier data digging algorithm computation base station location may further comprise the steps:
Step 1. is gathered base station data, sets up the set of base station data collection point, execution in step 2, step 3;
Step 2. goes out central point C according to the data computation of all collection points in the set of base station data collection point
0, as the initial position C of base station
0, execution in step 4;
Step 3. is divided into several base stations data acquisition group, execution in step 4 according to the signal strength signal intensity of terminal equipment with base station data collection point set tier;
Each point i is to base station C in the same base station data collection group of step 4. calculating
0Distance D
i, ask all D again
iAverage distance D, execution in step 5;
Step 5. is obtained the each point i of same base station data collection group to base station C
0Distance D
iDeviation D with average distance D
I0, to all D in same group
I0Descending descending sort;
Step 6. is judged D
I0Whether greater than setting threshold X, if then execution in step 7, otherwise execution in step 8;
Step 7. is removed this i from this group, repeating step 4, step 5, step 6 and step 7 all dispose until the data of all base station data collection groups;
Step 8. keeps these i data in this group, the data of the some i of reservation in all groups are reformulated new set;
Step 9. according in the new set have the data of an i to calculate the final position of base station in conjunction with the signal strength signal intensity of terminal equipment.
Further, the central point C described in the step 2
0Computational methods be: geometric average value-based algorithm or arithmetic mean value-based algorithm.Because the base station location of this moment is an initial value, be for judging the data set center of the reference that outlier provides, simply being averaged algorithm can meet the demands.
Further, the average distance D described in step 4, the step 5 arrives base station C for each point i
0Distance D
iThe arithmetic mean value.Because the average distance D of this moment also is a reference value, be for judging the data of the reference that the outlier departure degree provides, simply being averaged algorithm can meet the demands.
Further, the value of the setting threshold X described in the step 6 is according to being: D
I0/ D〉0.3.Determining of threshold value is a key, can miss some outlier if the value of threshold value is too small, excessively then can filter normal point, gets D
I0/ D〉basis of design of 0.3 setting threshold, can judge outlier and normal point more accurately.
Further, in the step 9, the institute in the new set has the data of an i to go out the final position of base station in conjunction with the algorithm computation of the signal strength signal intensity employing weighted average of terminal equipment.After the base stations outlier data digging was handled, the spatial distribution of the signal strength signal intensity of residue collection point was in normal condition.Same base station mobile handset signal intensity profile rule is that mobile phone signal intensity can be along with the increase of the distance of distance base station and weaken.Adopt the position of coming calculation base station based on the weighted average algorithm of mobile phone signal intensity based on this rule.
By a kind of method of the present invention based on space outlier data digging algorithm computation base station location, can improve k-arest neighbors (k-NN) detection algorithm that peels off, a kind of method based on space outlier data digging algorithm computation base station location has been proposed, be more suitable for the mining algorithm of base station data point that peels off, thereby make the location precise and high efficiency more of base station.With respect to the K-nearest neighbor algorithm, the present invention only calculates each distance of putting the data set center, has saved the step that point of every calculating is all wanted the whole data set of run-down, is more suitable in extracting the outlier that base stations is concentrated.
Description of drawings
Fig. 1 is the schematic diagram in the outlier zone of the present invention's one most preferred embodiment.
Fig. 2 is the flow chart of the present invention's one most preferred embodiment.
Embodiment
Now the present invention is further described with embodiment in conjunction with the accompanying drawings.
As shown in Figure 1, the data in the white portion are the outlier data, and its method of judging outlier is: certain data point is to the distance D at data set center
I0With all data points to the ratio of the average distance D at data set center greater than 0.3, i.e. D
I0/ D〉0.3 o'clock, this point is outlier.
In conjunction with Fig. 2, the flow process of a most preferred embodiment of the present invention is further specified;
Step 1. is gathered base station data, sets up the set of base station data collection point, execution in step 2, step 3;
Step 2. goes out central point C according to the The data geometrical mean algorithm computation of all collection points in the set of base station data collection point
0, as the initial position C of base station
0, execution in step 4;
Step 3. is divided into several base stations data acquisition group, execution in step 4 according to the signal strength signal intensity of terminal equipment with base station data collection point set tier;
Each point i is to base station C in the same base station data collection group of step 4. calculating
0Distance D
i, adopt the arithmetic mean value to calculate all D again
iAverage distance D, execution in step 5;
Step 5. is obtained the each point i of same base station data collection group to base station C
0Distance D
iDeviation D with average distance D
I0, to all D in same group
I0Descending descending sort;
Step 6. is judged D
I0Whether satisfy D
I0/ D〉0.3, if then execution in step 7, otherwise execution in step 8;
Step 7. is removed this i from this group, repeating step 4, step 5, step 6 and step 7 all dispose until the data of all base station data collection groups;
Step 8. keeps these i data in this group, the data of the some i of reservation in all groups are reformulated new set;
Step 9. has the data of an i to go out the final position of base station in conjunction with the algorithm computation of the signal strength signal intensity employing weighted average of terminal equipment according to the institute in the new set.
The present invention considers to use the non-space attribute to define neighborhood relationships from opposite angle, define the thought of distance function with space attribute, then the signal strength signal intensity of terminal equipment can be used as the standard of dividing data collection, and each is put the distance at data set center as distance function.Simultaneously this packet mode also is for fear of the interference of data set between the different mobile phone signal intensity, reduces the complexity of data processing.
Although specifically show and introduced the present invention in conjunction with preferred embodiment; but the those skilled in the art should be understood that; in the spirit and scope of the present invention that do not break away from appended claims and limited; can make various variations to the present invention in the form and details, be protection scope of the present invention.
Claims (5)
1. the method based on space outlier data digging algorithm computation base station location is characterized in that, may further comprise the steps:
Step 1. is gathered base station data, sets up the set of base station data collection point, execution in step 2, step 3;
Step 2. goes out central point C according to the data computation of all collection points in the set of base station data collection point
0, as the initial position C of base station
0, execution in step 4;
Step 3. is divided into several base stations data acquisition group, execution in step 4 according to the signal strength signal intensity of terminal equipment with base station data collection point set tier;
Each point i is to base station C in the same base station data collection group of step 4. calculating
0Distance D
i, ask all D again
iAverage distance D, execution in step 5;
Step 5. is obtained the each point i of same base station data collection group to base station C
0Distance D
iDeviation D with average distance D
I0, to all D in same group
I0Descending descending sort, execution in step 6;
Step 6. is judged D
I0Whether greater than setting threshold X, if then execution in step 7, otherwise execution in step 8;
Step 7. is removed this i from this group, repeating step 4, step 5, step 6 and step 7 all dispose until the data of all base station data collection groups;
Step 8. keeps these i data in this group, the data of the some i of reservation in all groups are reformulated new set;
Step 9. according in the new set have the data of an i to calculate the final position of base station in conjunction with the signal strength signal intensity of terminal equipment.
2. a kind of method based on space outlier data digging algorithm computation base station location according to claim 1 is characterized in that the central point C described in the step 2
0Computational methods be: geometric average value-based algorithm or arithmetic mean value-based algorithm.
3. a kind of method based on space outlier data digging algorithm computation base station location according to claim 1 is characterized in that, the average distance D described in step 4, the step 5 arrives base station C for each point i
0Distance D
iThe arithmetic mean value.
4. a kind of method based on space outlier data digging algorithm computation base station location according to claim 1 is characterized in that, the value of the setting threshold X described in the step 6 is according to being: D
I0/ D〉0.3.
5. a kind of method according to claim 1 based on space outlier data digging algorithm computation base station location, it is characterized in that, in the step 9, the institute in the new set has the data of an i to go out the final position of base station in conjunction with the algorithm computation of the signal strength signal intensity employing weighted average of terminal equipment.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110137078 CN102209385B (en) | 2011-05-25 | 2011-05-25 | Method for calculating position of base station based on spatial outlier data mining algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110137078 CN102209385B (en) | 2011-05-25 | 2011-05-25 | Method for calculating position of base station based on spatial outlier data mining algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102209385A true CN102209385A (en) | 2011-10-05 |
CN102209385B CN102209385B (en) | 2013-10-16 |
Family
ID=44698000
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201110137078 Active CN102209385B (en) | 2011-05-25 | 2011-05-25 | Method for calculating position of base station based on spatial outlier data mining algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102209385B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013174013A1 (en) * | 2012-05-25 | 2013-11-28 | 华为技术有限公司 | Method, server and system for determining site |
CN105138650A (en) * | 2015-08-28 | 2015-12-09 | 成都康赛信息技术有限公司 | Hadoop data cleaning method and system based on outlier mining |
CN105678704A (en) * | 2015-11-02 | 2016-06-15 | 安庆师范学院 | Nonlocal median value blind noise reduction method based on visual perception |
CN106162652A (en) * | 2016-08-29 | 2016-11-23 | 杭州电子科技大学 | A kind of base station location localization method based on drive test data |
CN108521628A (en) * | 2018-03-29 | 2018-09-11 | 维沃移动通信有限公司 | A kind of localization method, device and mobile terminal |
CN111639703A (en) * | 2020-05-29 | 2020-09-08 | 国家计算机网络与信息安全管理中心广东分中心 | Method for calculating base station position based on minimum surrounding circle of discrete point set |
CN113466826A (en) * | 2021-05-12 | 2021-10-01 | 武汉中仪物联技术股份有限公司 | Data denoising method, device, equipment and medium for range radar |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1354491B1 (en) * | 2000-12-29 | 2008-02-06 | Ekahau Oy | Location estimation in wireless telecommunication networks |
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 |
CN101945325A (en) * | 2010-08-13 | 2011-01-12 | 厦门雅迅网络股份有限公司 | Base station positioning-based friend perception method |
-
2011
- 2011-05-25 CN CN 201110137078 patent/CN102209385B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1354491B1 (en) * | 2000-12-29 | 2008-02-06 | Ekahau Oy | Location estimation in wireless telecommunication networks |
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 |
CN101945325A (en) * | 2010-08-13 | 2011-01-12 | 厦门雅迅网络股份有限公司 | Base station positioning-based friend perception method |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013174013A1 (en) * | 2012-05-25 | 2013-11-28 | 华为技术有限公司 | Method, server and system for determining site |
CN103563448A (en) * | 2012-05-25 | 2014-02-05 | 华为技术有限公司 | Method, server and system for determining site |
CN103563448B (en) * | 2012-05-25 | 2017-12-15 | 华为技术有限公司 | A kind of method, server and system for determining site |
CN105138650A (en) * | 2015-08-28 | 2015-12-09 | 成都康赛信息技术有限公司 | Hadoop data cleaning method and system based on outlier mining |
CN105678704A (en) * | 2015-11-02 | 2016-06-15 | 安庆师范学院 | Nonlocal median value blind noise reduction method based on visual perception |
CN105678704B (en) * | 2015-11-02 | 2018-09-25 | 安庆师范学院 | A kind of non local intermediate value blind landing method for de-noising of view-based access control model perception |
CN106162652A (en) * | 2016-08-29 | 2016-11-23 | 杭州电子科技大学 | A kind of base station location localization method based on drive test data |
CN108521628A (en) * | 2018-03-29 | 2018-09-11 | 维沃移动通信有限公司 | A kind of localization method, device and mobile terminal |
CN111639703A (en) * | 2020-05-29 | 2020-09-08 | 国家计算机网络与信息安全管理中心广东分中心 | Method for calculating base station position based on minimum surrounding circle of discrete point set |
CN111639703B (en) * | 2020-05-29 | 2023-11-14 | 国家计算机网络与信息安全管理中心广东分中心 | Method for calculating position of base station based on minimum bounding circle of discrete point set |
CN113466826A (en) * | 2021-05-12 | 2021-10-01 | 武汉中仪物联技术股份有限公司 | Data denoising method, device, equipment and medium for range radar |
Also Published As
Publication number | Publication date |
---|---|
CN102209385B (en) | 2013-10-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102209385B (en) | Method for calculating position of base station based on spatial outlier data mining algorithm | |
CN110447277B (en) | Fingerprint positioning method and related equipment | |
CN102521973B (en) | A kind of mobile phone switches the road matching method of location | |
CN108181607A (en) | Localization method, device and computer readable storage medium based on fingerprint base | |
CN104335622B (en) | The network planning, optimization method and device based on service distribution | |
CN108419249B (en) | Three-dimensional wireless sensor network clustering covering method, terminal equipment and storage medium | |
CN104320794B (en) | One kind assesses the rational method and device of site deployment | |
CN106465054A (en) | Geo-fencing implementation method and mobile device | |
WO2020024597A1 (en) | Indoor positioning method and apparatus | |
KR20140022695A (en) | Method and apparatus for determining position by using radio environment information | |
JP6358247B2 (en) | Information processing apparatus, information processing method, and program | |
CN114786199B (en) | Method, device, equipment and storage medium for determining network problem point | |
EP2501166B1 (en) | Method and system for selecting candidate neighboring cells | |
WO2016062161A1 (en) | Resource utilization method and device | |
CN108513717B (en) | Prompting method, terminal and server | |
CN107708072B (en) | Trajectory analysis method and device based on mobile communication network | |
CN108574927B (en) | Mobile terminal positioning method and device | |
CN103945433A (en) | Method and device for determining area with weak network coverage | |
KR20110112093A (en) | Method and apparatus for measuring lacation using access point | |
JP6223302B2 (en) | Apparatus, program, and method capable of determining stay of user having portable terminal | |
JP2012039389A (en) | Base station extension position determination method and base station | |
KR20160139796A (en) | Method And Apparatus for Estimating Location by Using Log Data | |
KR20110112094A (en) | Method and apparatus for measuring lacation using access point | |
JP5418059B2 (en) | Neighboring cell design apparatus, neighboring cell designing method and program | |
CN105491574A (en) | Method and device for determining priority of neighbor regions in neighbor region planning |
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 |