KR20100026215A - Method to extract and represent terrain obstacles for radio network optimization from measured drive-test data - Google Patents

Abstract

PURPOSE: A method for extracting geographical information for optimizing wireless network using a measured data is provided to automatically optimize the wireless network while improving accuracy of interpretation for radio wave. CONSTITUTION: A base station data and a measurement data are obtained(111). The measured data about a first base station is classified according to the influence to a geographical feature(112). LOS(Line-Of-Sight) and NLOS(Non-LOS) are classified by comparing a measured signal strength with a predicted signal strength. The measured data is determined as LOS when the measured signal intensity is greater than the predicted signal intensity. A boundary point between the classified data is determined(113). The geographical feature information is extracted(114) using the determined boundary point.

Description

Method for extracting feature information for wireless network optimization using measurement data {METHOD TO EXTRACT AND REPRESENT TERRAIN OBSTACLES FOR RADIO NETWORK OPTIMIZATION FROM MEASURED DRIVE-TEST DATA}

The present invention relates to a feature information extraction method for optimizing a wireless network and a computer-readable recording medium recording a program for realizing the method. More particularly, the feature information is extracted based on base station data and measurement data. After constructing the feature data, the constructed feature data is again used for radio wave analysis and wireless network optimization to increase the accuracy of radio wave analysis and automatically perform wireless network optimization. A method for extracting feature information for optimization and a computer-readable recording medium having recorded thereon a program for realizing the method.

In general, radio wave analysis tools use Geographic Information System (GIS) data (terrain data) to simulate radio wave propagation as accurately as possible.

The terrain data used in the radio wave analysis tool may be divided into digital terrain data for radio wave analysis and an image for a user. The terrain data is again made up of terrain (Terrain / Clutter), terrain characteristics (Morphology), artificial features (Vector) data. Meanwhile, the image for the user may use a picture taken by a satellite or an aircraft, or scan a map when the picture is not available.

And the result of radio wave analysis may get completely different result according to resolution of GIS data used. For example, the topographical data that is the basis of radio wave analysis can be classified into full plane, low resolution, medium resolution, and high resolution according to the resolution.

Next, the low resolution data, the medium resolution data, and the high resolution data will be described in detail.

First, low resolution data refers to data representing the terrain at 300 meters or higher resolution. Low resolution data is used to develop an approximate business plan at the start of a wireless business. Low-resolution data represents the entire city in dozens of pixels, making it unsuitable for accurate cell design in urban areas.

On the other hand, the medium resolution data refers to data representing the terrain at a resolution of 100 to 200 meters or more. The medium resolution data includes the following two pieces of data.

The first data is digital terrain data, which refers to data (DEM) in which the terrain is composed of pixels having altitude values.

The second data is terrain feature data, and refers to data (Morphology) indicating the feature of the terrain or the type of use.

The terrain feature data includes the height of the object on the ground, diffraction coefficients, and propagation attenuation coefficients.

As can be seen in the examples below, the medium resolution data does not include building data such as the actual boundaries or the actual height of each building.

Using this medium resolution data, two types of propagation analysis can be performed.

Deterministic Model

Empirical Model

The deterministic model is a method of calculating the intensity of radio waves reaching a specific point using electromagnetic wave propagation theory. At this time, in order to use the deterministic model, there must be three-dimensional (3D) terrain data. However, since the medium resolution data does not represent the actual height of each building, but only the statistical height for each feature of the terrain, that is, it does not include three-dimensional (3D) terrain data, a deterministic model is generally used for the medium resolution data. It is impossible to apply it, but the deterministic model can be applied to the medium resolution data only when calculating the propagation in the high frequency band.

And the empirical model ignores the effect of the building and assumes the results produced by the analysis of the actual measurement results and assumptions about the results (building density, spacing between buildings, average building height, average road width). Etc.), and perform radio wave analysis. Therefore, the empirical model is not significantly affected by the precision of the terrain data. This empirical model is generally somewhat inaccurate but requires the least amount of computation. And some propagation models can determine parameters assuming building density or average building height. Although these propagation models have been sufficiently verified, in practice, it is difficult to obtain values such as building density and average building height, and because the density or height is not constant, in order to obtain accurate results, tuning the propagation model parameters using measurement data. This has the necessary disadvantages.

High resolution data, on the other hand, is used to provide the radio wave environment of the downtown as accurately as possible. At this time, all the data affecting the radio waves are modeled.

FIG. 1 is a diagram illustrating a comparison of radio wave analysis results of a deterministic model and an empirical model for the same terrain.

In the propagation analysis tool, as shown in FIG. 1, terrain features are uploaded as vector data on top of the terrain data, and each feature has an accurate height value with the corresponding characteristic data.

However, this high resolution data is expensive to construct, and the three-dimensional (3D) propagation analysis that considers each feature has very high complexity, which makes it difficult to apply it to the present technology.

FIG. 2 is a diagram illustrating terrain data including a feature using a vector, and FIG. 3 is a diagram illustrating terrain data including a feature.

The radio wave analysis tool also supports a method of using the features as a digital surface model (DEM) included in the digital elevation model (DEM) as shown in FIG. 2. This method has the advantage of low construction cost because the height information extracted from satellite or aerial data can be used without manual operation, but the accuracy of the building height is lowered according to the resolution of the terrain data, and the material of the building (concrete, Steel frame, etc.) can not be specified. In addition, this method also has the disadvantage that it is not possible to consider the wireless communication terminal at the bottom because the same treatment of the building and the terrain.

Therefore, in general, after conducting radio wave analysis using empirical model to determine coverage, drive test is performed to adjust parameters of empirical model using methods such as minimum mean spare (MMSE). To increase the accuracy. However, this method has a disadvantage in that the possibility of failing to predict the coverage hole caused by the feature is quite high due to the limitation of not considering the feature.

In conclusion, the existing radio wave analysis method has a problem of inaccurate propagation simulation without expensive building data, and due to this problem, a person must actually perform a wireless network optimization task.

In other words, the conventional radio wave analysis method has a problem that the radio wave simulation results are inaccurate without expensive building data, and accordingly, radio wave analysis and radio network optimization are performed to reduce radio wave accuracy and to automatically optimize the radio network. There is a disadvantage that cannot be done.

As described above, the prior art as described above has a problem that the radio wave simulation results are inaccurate, and thus, the accuracy of the radio wave analysis is inferior and the wireless network optimization cannot be automatically performed. Is the task.

Therefore, the present invention extracts feature information based on base station data and measured data to construct feature data, and then uses the constructed feature data again for radio wave analysis and wireless network optimization to increase the accuracy of radio wave analysis and automatically. An object of the present invention is to provide a method for extracting feature information for optimizing a wireless network using measurement data and a computer-readable recording medium having recorded thereon a program for realizing the method.

The objects of the present invention are not limited to the above-mentioned objects, and other objects and advantages of the present invention which are not mentioned above can be understood by the following description, and will be more clearly understood by the embodiments of the present invention. Also, it will be readily appreciated that the objects and advantages of the present invention may be realized by the means and combinations thereof indicated in the claims.

According to an aspect of the present invention, there is provided a method of extracting feature information, comprising: a data obtaining step of obtaining base station data and measurement data; A measurement data classification step of classifying the measurement data for the first base station according to the influence of the feature; A determination step of determining boundary points between the divided measurement data; And extracting feature information using the determined boundary point to extract feature information.

The method may further include extracting feature information for each base station by repeating the measurement data classification step to extracting the feature information for all base stations other than the first base station. do.

On the other hand, the present invention, a feature information extraction system having a processor, the data acquisition function for acquiring base station data and measurement data; A measurement data classification function for classifying measurement data for the first base station according to whether the feature is affected by the feature; A determination function for determining a boundary point between the divided measurement data; And a computer readable recording medium having recorded thereon a program for realizing a feature information extraction function for extracting feature information using the determined boundary point.

In addition, the present invention, to further implement the function of extracting feature information for each base station by repeatedly performing the measurement data classification function to the feature information extraction function for all base stations other than the first base station A computer readable recording medium having recorded a program is provided.

As described above, the present invention has an effect of extracting and storing information on a feature of a feature much more simply than a conventional radio wave analysis method.

In other words, in order to solve the problem of inaccurate propagation simulation and wireless network optimization without expensive building data, the present invention is more cost effective than conventional radio wave analysis and wireless network optimization methods. Accurate propagation strength prediction is possible.

In other words, in order to solve the problem that humans had to perform the radio propagation simulation in practice because the present invention had to perform inaccurate propagation simulation without expensive building data, all base station data and measurement data were selected. After loading the information extraction system (e.g., a computer), the feature information is extracted based on the base station data and the measurement data to construct the feature data, and then the feature data is again used for radio wave analysis and wireless network optimization. This improves the accuracy of radio wave analysis and enables automatic network optimization.

The above objects, features, and advantages will become more apparent from the detailed description given hereinafter with reference to the accompanying drawings, and accordingly, those skilled in the art to which the present invention pertains may share the technical idea of the present invention. It will be easy to implement. In addition, in describing the present invention, when it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted. Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Here, prior to the description of a specific embodiment of the present invention, it will be described with respect to the propagation analysis method in the case of LOS (Line-Of-Sight) and NLOS (Non-LOS).

In general, propagation analysis differs depending on the case of LOS (Line-Of-Sight) and NLOS (Non-LOS).

First, the propagation attenuation calculation method in the case of LOS (Line-Of-Sight) is as follows.

Line-Of-Sight (LOS) is a case where the receiver Rx and the transmitter Tx are directly visible. However, in the radio wave analysis, it means that not only the receiver Rx and the transmitter Tx are directly visible, but also a Fresnel Zone that is not hidden between the receiver Rx and the transmitter Tx should be established.

If this condition is satisfied, propagation attenuation between the receiving end Rx and the transmitting end Tx may be calculated using Equation 1 below using the "ITU-R 525 model".

Here, L _fsd denotes free space loss, and d denotes the distance between the receiver Rx and the transmitter Tx.

Next, the radio wave attenuation calculation method in the case of NLOS (Non-LOS) is as follows.

In the case of NLOS, the diffraction loss must be added to the aforementioned free space loss.

Fresnel theory states that propagation attenuation caused by a single knife-edge obstruction on a straight line between the receiving end (Rx) and the transmitting end (Tx) can be obtained by Fresnel Integral, Since it is not practical to apply, the diffraction loss due to such knife edge is obtained as shown in Equation 2 below.

Here, h / r is also called "clearance ratio", and the variable v means the square root of "clearance ratio" as defined in [Equation 2]. Here, r is the radius of the Fresnel ellipsoid at distance d from the transmitter Tx, and h is the height of the obstacle corresponding to the upper part of the line-of-sight (LOS). This can be easily understood by looking at the method of determining v shown in FIG. 4. In this case, since only one obstacle is present, generalizing Equation 2 may be expressed as Equation 3 below. In Equation 3, i is the number of obstacles between the transmitting end Tx and the receiving end Rx.

Therefore, the signal attenuation in the case of NLOS (Non-LOS) can be represented by the following formula (4).

In addition, sub-path attenuation and reflection can be considered for more accurate calculations. However, in one embodiment of the present invention, for convenience of description, the sub path attenuation and reflection will not be considered. However, it will be possible to those skilled in the art to implement in consideration of secondary path attenuation and reflection in actual implementation.

5A and 5B are diagrams illustrating an embodiment of a feature information extraction system to which the present invention is applied.

As shown in FIG. 5A, the feature information extraction system to which the present invention is applied includes a keyboard 51 which is an input device for inputting data such as a computer 51 performing a calculation necessary for the present invention, a location and power of a base station, and the like. 52, a mouse 53, and a printer 54, which is an output device for outputting calculation results.

As shown in FIG. 5B, the computer 51 of the feature information extraction system to which the present invention is applied includes a central processing unit 55, a main memory unit 56 connected to the central processing unit 55, and An auxiliary memory device 57 connected to the main memory device 56 and a peripheral device 58 connected to the main memory device 56 are provided.

As described above, the feature information extraction system to which the present invention is applied includes a central processing unit 55 that controls and manages the overall operation of a computer, a program executed by the central processing unit 55, and is used while performing a task or A main memory 56 and a secondary memory 57 for storing various data generated while performing a task, input / output devices 52 to 54 for inputting / outputting data to and from a user, and a peripheral for a communication interface, etc. Device 58.

In addition, the auxiliary memory device 57 stores a large amount of data, and the input / output devices 52 to 54 may include a general keyboard 52, a mouse 53, a display device, a printer 54, and the like. It includes.

However, since the computer hardware environment having the configuration as described above is only a technique well known in the art, detailed description thereof will be omitted herein. However, the main memory device 56 of the hardware system as described above extracts feature information based on base station data and measurement data, constructs the feature data, and then reconstructs the constructed feature data for radio wave analysis and wireless network optimization. By using the feature, a feature extraction algorithm for storing radio information is improved and the wireless network optimization can be automatically performed, which is performed under the control of the central processing unit 55.

The algorithm to achieve in the present invention, that is, the algorithm for extracting feature information for optimizing the wireless network using measurement data, obtains base station data and measurement data, and determines whether the measurement data for the first base station is affected by the feature. The algorithm determines the boundary points between the divided measurement data and extracts feature information using the determined boundary points for all base stations.

FIG. 6 is a diagram illustrating an example of a building position in a downtown area and a measurement signal according thereto, and FIG. 7 is a diagram showing data available to an actual engineer in an example of a building position in a downtown area and a measurement signal accordingly. 8 is a diagram showing a case where a line separating the boundary between LOS and NLOS is determined from data available by a real engineer in an example of a building position in an urban area and a measurement signal accordingly, and FIG. In the example of the measurement signal according to the drawing showing an example of the building information extracted by separating the LOS and NLOS from the data available to the actual engineer, Figure 10 is a real estate in the example of the building position and the resulting measurement signal in the urban area The ratio of the building information extracted from the available data to LOS and NLOS and the actual building location. And Figure 11 is a flow diagram of one embodiment a radio network optimization feature has information for extracting the measurement data using the method according to the invention.

First, a technical gist of a method for extracting feature information for optimizing a wireless network using measured data according to the present invention will be described.

In order to provide the present invention, after loading (or inputting) all base station data and measurement data into a feature information extraction system (for example, a computer), feature information is extracted by extracting feature information based on the base station data and measurement data. After constructing, by using the constructed feature data again for radio wave analysis and radio network optimization, the accuracy of radio wave analysis is increased, and radio network optimization can be automatically performed.

In general, if the radio wave is not affected by the feature, the radio wave emitted from the base station proceeds through the space according to the free-space propagation model. Therefore, if the radio wave emitted from the base station is not affected by the feature, the value of the radio wave intensity that attenuates the intensity of the radio wave emitted from the base station by applying the free-space propagation model is approximately equal to the measured radio wave intensity. It will have a value.

However, if the radio wave emitted from the base station is affected by the feature, the value of the radio wave intensity that attenuates the intensity of the radio wave emitted from the base station by applying the free-space propagation model is greater than the measured radio wave intensity. It will have a value.

In the present invention, the signal strength of the measured data is determined by using the above characteristics (a result of comparing the measured value of the propagation intensity with the value of the propagation intensity attenuated by the free space propagation model by applying the free space propagation model). The measurement data is displayed as LOS and NLOS. That is, the measurement data is divided into measurement data not affected by the feature and measurement data affected by the feature.

In this case, the signal strength from the specific base station (hereinafter referred to as "measurement signal strength") measured at any one of several measurement points at which a signal from one specific base station among several base stations is received is measured in a downtown area. If it is greater than or equal to the signal strength (hereinafter, referred to as "predicted signal strength") predicted by the propagation model in consideration of additional loss (attenuation), it is determined as LOS, and if it is small, it is determined as NLOS.

At this time, the measured signal strength of the measurement point of the known LOS position, which is the reference signal strength, should satisfy the following [Equation 1], and since all values except for the δ value can be seen in [Equation 1] below, From Equation 1 below, δ values corresponding to the propagation environment of the region can be obtained.

Here, the meaning of each parameter of the above [Equation 1] is as follows.

SS_indBm _LOS : Signal strength in dBm in case of LOS

TxPow_indBm: signal transmit power in dBm

-AntennaGain_indB: Antenna gain in dB

PathLoss_indB (): Path loss function in dB

d _SH : distance between the base station and the terminal

f: frequency of the signal

d: Distance between the transmitting point and the receiving point of the radio wave

보다 작거나 또는 일정 차이 이상으로 작은 경우라면, 해당 측정 지점을 NLOS로 판단하게 된다.Therefore, the measurement propagation intensity at the measurement point is the predicted propagation intensity for that measurement point.

If it is smaller or smaller than a certain difference, the measurement point is determined as NLOS.

6 shows an example of a building position in a downtown area and a measurement signal in this case. In this case, the data available to the actual engineer is as shown in the example shown in FIG. In this case, as shown in FIG. 7, when the measured signal strength of the measurement point is greater than or equal to the predicted signal strength, it is determined as LOS, and when it is small, it is determined as NLOS to distinguish between LOS and NLOS. The case shown in red is shown.

As described above, the measurement data for the specific base station is divided into LOS and NLOS, and then displayed, and as shown in FIG. 8, the boundary point of the LOS and NLOS is found to determine a line from the corresponding base station to the boundary point.

Next, in order to indicate the location of the feature, an arc is determined based on the line determined in FIG. 8 to extract the feature information. The radius of the arc is then the distance from the base station to the nearest NLOS starting point in the arc. 9 shows an example of the circular arc thus determined.

The above process should be performed repeatedly for all base stations, and the extracted feature information is used for radio wave analysis and wireless network optimization for the base stations.

Next, referring to FIG. 11, the operation flow of the feature information extraction method for optimizing the wireless network using the measurement data according to the present invention is as follows.

First, all base station data and measurement data are loaded (or input) into a feature information extraction system (eg, a computer) (111).

Subsequently, for all measurement points in which a signal transmitted from a specific base station is received, the LOS and the NLOS are distinguished by comparing the measurement signal strength and the prediction signal strength (112). That is, by comparing the measured signal strength and the predicted signal strength, when the measured signal strength is greater than or equal to the predicted signal strength, it is determined as LOS, and when it is small, it is determined as NLOS to distinguish between LOS and NLOS. In other words, the measurement data is divided into measurement data not affected by the feature and the measurement data affected by the feature.

Subsequently, the boundary point between the separated LOS and NLOS is found to determine a line from the corresponding specific base station to the boundary point (113). That is, after searching for the boundary point between the area where the measurement data is not affected by the feature and the area where the measurement data is affected by the feature, the line from the specific base station to the searched boundary point is determined.

Thereafter, in order to indicate the location of the feature, an arc is determined based on the determined line to extract the feature information (114). That is, the feature information is extracted by determining an arc that can represent the NLOS region based on the determined line. At this time, the radius of the arc is the distance from the specific base station to the nearest NLOS starting point in the arc.

This process of "111" to "114" should be performed repeatedly for all base stations, and thus the extracted feature information of each is used for radio wave analysis and wireless network optimization for the base stations.

By repeating each process as described above, it is possible to infer and install features between each base station and features, and after that, when performing radio wave analysis and wireless network optimization using these features, the LOS region and NLOS region are determined. This makes it possible to predict more accurate propagation intensity values.

By using this method, feature information can be extracted simply, and feature information is represented by an arc from a base station. Therefore, feature information can be represented and managed by the following simple data structure.

{Start angle, end angle, radius}

At this time, the feature information is extracted for each base station, the start angle is to start the measurement data at each base station at the true north (0 degrees) position and search the measurement data by angle in the clockwise direction, the measurement data from LOS to NLOS It means the point of change. The end angle starts at the start angle, and is the point where the NLOS changes to LOS by searching the measured data clockwise by angle. The radius is the line segment connecting the start and end angles around the base station, and the distance between the base station and the base station is the two of the two points where the LOS and NLOS states change.

In this case, each {start angle, end angle, radius} is defined for each base station, and one pair of start angle and end angle may overlap with another pair of start angle and end angle. For example, there may be feature information of {0 degrees, 60 degrees, 30 meters} (A), {30 degrees, 50 degrees, 20 meters} (B) for one base station. This may occur when there are measurement data of multiple paths, as shown in the lower right of FIG. 8, when there are measurement points where the LOS changes to NLOS after continuing from clockwise to LOS after NLOS change. have. In this case, before the feature information A of one {start angle, end angle, radius} is completed, another feature information B of {start angle, end angle, radius} begins to be formed. In this case, since the feature close to the base station makes NLOS all paths behind it, the feature information (B) always has a smaller radius than the feature information (A), and the feature information (B) is always the feature. It has less end angle than feature information (A). Of course, the start angle of the feature information (B) is larger than the feature information (A).

As described above, in order to solve the problem that humans had to perform the radio propagation work in practice because the present invention had to perform inaccurate propagation simulation without expensive building data, all without expensive building data. After loading the base station data and the measurement data into the feature information extraction system (for example, a computer), the feature information is extracted based on the base station data and the measurement data, and the feature data is reconstructed. By using the feature data on the radio wave analysis system and the radio network optimization system, the radio wave analysis accuracy is improved and the radio network optimization is performed automatically.

On the other hand, the method of the present invention as described above can be written in a computer program. And the code and code segments constituting the program can be easily inferred by a computer programmer in the art. In addition, the written program is stored in a computer-readable recording medium (information storage medium), and read and executed by a computer to implement the method of the present invention. The recording medium may include any type of computer readable recording medium.

The present invention described above is capable of various substitutions, modifications, and changes without departing from the technical spirit of the present invention for those skilled in the art to which the present invention pertains. It is not limited by the drawings.

The present invention can be used for radio wave analysis and radio network optimization accordingly.

1 is a diagram showing a comparison of radio wave analysis results of a deterministic model and an empirical model for the same terrain;

2 is a diagram illustrating terrain data including a feature using a vector;

3 is a diagram showing terrain data including features;

4 is an explanatory diagram for a determination method of v,

5a and 5b is a configuration diagram of an embodiment of a feature information extraction system to which the present invention is applied;

6 is a view showing an example of a building position and a measurement signal according to the urban area;

7 shows data available to an actual engineer in an example of a building location in the urban area and thus a measurement signal;

8 is a view showing a case where a line separating the boundary between LOS and NLOS is determined from data available to a real engineer in an example of a building position in a downtown area and a measurement signal accordingly;

9 is a view showing an example of the building information extracted by separating the LOS and NLOS from the data available to the actual engineer in the example of the building location and the measurement signal according to the urban area;

FIG. 10 is a comparison diagram between building information extracted from LOS and NLOS extracted from data available to a real engineer in an example of a building location and a measurement signal according to a building location, and a real building location;

11 is a flowchart illustrating a method for extracting feature information for optimizing a wireless network using measured data according to the present invention.

* Explanation of symbols for the main parts of the drawings

51: Computer 52: Keyboard

53: mouse 54: printer

55: central processing unit 56: main memory unit

57: auxiliary storage device 58: peripheral device

Claims (12)

In the feature information extraction method, A data acquisition step of acquiring base station data and measurement data; A measurement data classification step of classifying the measurement data for the first base station according to the influence of the feature; A determination step of determining boundary points between the divided measurement data; And Feature information extraction step of extracting feature information using the determined boundary point Feature information extraction method comprising a. The method of claim 1, Extracting feature information for each base station by repeating the measurement data separating step to the feature information extracting step for all base stations other than the first base station; Feature information extraction method further comprising a. The method according to claim 1 or 2, The measurement data classification step, A feature information extraction method for distinguishing line-of-sight (LOS) and non-LOS (NLOS) by comparing measured signal strength and predicted signal strength for all measurement points. The method of claim 3, wherein The measurement data classification step, Comparing measured signal strength and predicted signal strength for all the measurement points; Determining, by the LOS, if the measured signal strength is equal to or greater than the predicted signal strength; And If the measured signal strength is less than the predicted signal strength, determining by NLOS; Feature information extraction method comprising a. The method of claim 3, wherein The determining step, Searching for boundary points between the divided measurement data; And Determining a line from the first base station to the searched boundary point Feature information extraction method comprising a. The method of claim 5, wherein The determining step, And detecting a boundary point between the divided LOS and NLOS to determine a line from the first base station to the searched boundary point. The method of claim 6, The feature extraction step, Feature information extraction method for extracting feature information by determining an arc based on the determined line. The method of claim 7, wherein The feature extraction step, Based on the determined line, the feature information extraction method characterized in that to extract the feature information by determining the arc that can represent the NLOS region. The method of claim 8, The extracted feature information, Feature information extraction method having a structure of {start angle, end angle, radius}. The method of claim 9, The start angle indicates a point where the measurement data is changed from LOS to NLOS by searching the measurement data for each angle in the clockwise direction, starting from the position of true north (0 degree) at the first base station. The end angle starts at the start angle and indicates a point where the NLOS is changed to LOS by searching the measurement data clockwise for each angle. The radius is obtained by obtaining a line segment connecting the start angle and the end angle with respect to the first base station, and between the point closest to the first base station and the base station among the two points where the LOS and NLOS states change at the two line segments. Feature information extraction method characterized in that it represents a distance. In a feature information extraction system having a processor, A data acquisition function for acquiring base station data and measurement data; A measurement data classification function for classifying measurement data for the first base station according to whether the feature is affected by the feature; A determination function for determining a boundary point between the divided measurement data; And Feature information extraction function for extracting feature information using the determined boundary point A computer-readable recording medium having recorded thereon a program for realizing this. The method of claim 11, Function to extract feature information for each base station by repeating the measurement data classification function to the feature information extraction function for all base stations other than the first base station A computer-readable recording medium that records a program for further realization.

Images