KR101779965B1 - Method for generating a fingerprint map at any position - Google Patents

Abstract

 The present invention relates to a construction method for generating a fingerprint map for location recognition using the reception sensitivity of a radio signal of a wireless communication infrastructure collected at an arbitrary location, Dividing the collection area into a plurality of segments according to size and density of the collection points, collecting signals at each collection position corresponding to each of the divided segments, selecting a representative point of the segment, A representative pattern detection step of detecting a representative signal pattern and a map construction step of constructing a radio frequency fingerprint map using the detected representative signal pattern for each segment.

Description

METHOD FOR GENERATING A FINGERPRINT MAP AT ANY POSITION BACKGROUND OF THE INVENTION [0001]

 The present invention relates to a construction method for generating a fingerprint map for location recognition using the radio signal reception sensitivity of a wireless communication infrastructure collected at an arbitrary position.

Recently, the location based service (LBS) market has been expanded through popularization of GPS technology and car navigation. In recent years, interest in the location based service market has been intensified due to the development of infrastructure and the emergence of smart phones, The development of related industries / technologies is also accelerating. As a result, many researches are being conducted. Especially, as a mobile device such as a smart phone is popularized, a location-based service technology in a room is being studied. However, due to the poor radio environment in an indoor environment, The location recognition technique using fingerprint is getting popular.

In order to recognize the position using the radio wave fingerprint, the radio frequency signal reception sensitivity pattern of the wireless communication infrastructure is measured at various points in the indoor specific space, and a radio wave fingerprint map is constructed. Then, the user requesting the position recognition enters the collection area, The user analyzes the pattern of the radio signal received by the user and maps the position of the user to the point showing the most similar pattern.

In this case, in order to construct a radio-frequency fingerprint map, a method of calculating a representative pattern by a point is generally used, averaging the data collected several times while staying at a central point of the segments divided by the collection area, This is because the received signal strength (or signal sensitivity) of the collected wireless communication infrastructure continuously changes even in a situation where the propagation environment is extremely stable.

However, since the RFID fingerprint map construction method must accurately move to the center point of each segment in the collection area and wait for a few minutes at the corresponding point and collect a plurality of data, the collection process is very inefficient and the collection space is wide There is a problem that the time, effort, and cost for constructing the radiofrequency map are excessively generated. In addition, since the collected sensitivity pattern is sensitive to changes in environment such as surrounding walls, it is necessary to perform a re-collection and update process from time to time, thereby increasing the inefficiency.

SUMMARY OF THE INVENTION In order to solve the problems described above, it is an object of the present invention to provide a radio frequency identification method and a radio wave communication method, in which, even in the case of collecting a radio wave pattern at an arbitrary position, And to provide a method for constructing an accurate radio wave fingerprint map.

In order to achieve the above object, the present invention provides a method of measuring a position of a gastric acid signal, comprising the steps of: dividing the collection area into a plurality of segments according to the size of the collection zone of the gastric acid signal and the density of the collection point; A representative pattern detection step of detecting a representative signal pattern at the representative point by collecting a signal at a representative point of the segment and selecting a representative point of the segment, and constructing a map for constructing a radio frequency fingerprint map using the detected representative signal pattern for each segment The method comprising the steps of: constructing a radio wave fingerprint map at an arbitrary position including a step of constructing a radio wave fingerprint map;

According to the present invention, it is possible to construct and generate a relatively accurate radiofrequency fingerprint map more efficiently and conveniently in the room, thereby achieving a dramatic reduction in the cost of building a radiofrequency map in a wide area.

When a Wi-Fi AP signal is used as a signal for position measurement, which is used as an indoor location-based service infrastructure, AP is periodically transmitted with a beacon signal, It is possible to construct an efficient radio wave fingerprint map.

In addition, when a device capable of recognizing the position of a collector is also used by using dead reckoning (DR) or an image-based position recognition method, it is possible to collect the reception sensitivity pattern while the collector is moving. In this case, it becomes possible to construct an automated and highly efficient radio wave fingerprint map, and it is possible to utilize it to recognize the position of equipment capable of receiving the wireless communication signal in the future.

FIG. 1 is a flowchart illustrating a method of constructing a radio frequency fingerprint map using random position reception sensitivity information according to an embodiment of the present invention.
FIG. 2 is a plan view showing distribution of collection points in the corresponding region in FIG. 1,
FIG. 3 is a plan view showing a state where a corresponding region is divided into segments in FIG. 1,
FIG. 4 is a flowchart showing a step of detecting a representative signal pattern of a segment in FIG. 1,
FIG. 5 is a plan view showing a state in which each segment is divided into sub segments in FIG. 4,
FIG. 6 is a flowchart showing a step of detecting a segment representative pattern using the representative pattern of the sub-segment in FIG.

Hereinafter, a method for constructing a radio-frequency fingerprint map at an arbitrary position according to an embodiment of the present invention will be described in detail with reference to the drawings. In the following description, the positional relationship of each component is principally described based on the drawings. The drawings may be simplified for simplicity of the description or exaggerated when necessary. Therefore, the present invention is not limited thereto, and other steps may be added, modified or omitted.

FIG. 1 is a flowchart illustrating a method of constructing a radio wave fingerprint map using random position reception sensitivity information according to an embodiment of the present invention. As shown in FIG. 1, a method of constructing a radio-frequency fingerprint map according to the present embodiment includes analyzing a distribution point distribution S100, dividing a region into a plurality of segments S200, (S300), and constructing a radio wave fingerprint map (S400).

First, in step S100 of analyzing the distribution of the collection points, as shown in FIG. 2, a method of displaying the position of the collection points on a top view of the corresponding area is performed. At this time, the position of the collection zone corresponds to an arbitrary position, which may be a position where signal collection for estimation is performed. As shown in FIG. 2, it can be seen that the distribution of the collection point is higher in the upper region (reference plane) in the region, and the distribution of the collection point in the lower region is relatively lower.

On the other hand, if analysis of the collection point distribution is completed, a step of dividing the corresponding area into a plurality of segments can be performed (S200). At this time, it is possible to divide the segment into a plurality of segments (S) in consideration of the size of the area where the radio frequency fingerprint map is to be constructed and the density of the collection point. If the size of the segment is too small, there is no collected data in the segment and the representative pattern of the segment can not be found. If the size of the segment is too large, the resolution of the position recognition deteriorates.

Therefore, in this embodiment, as shown in FIG. 3, the upper portion having the higher density of the collection point divides the space into smaller segments, and the lower portion having the lower density of the collection points divides into larger segments, It is possible to divide the zone into a plurality of segments in consideration of the presence or absence of the collected data in the segment.

As an example, this segment segmentation step may be performed by first setting a minimum segment that can divide the collection zone satisfying the corresponding resolution constraint, if there is a minimum resolution that the location service provider should meet, It is also possible to proceed with this step by reducing the size of the initially set segment. That is, when the collection density of the segment is high in the segment set in the first place, the segment is further divided into segments, and in the region where the density of the collection point is low and the data is judged to be insufficient, the segmented segment can be maintained primarily.

However, this is merely an example for segment segmentation, and it is also possible to proceed to segmenting into segments according to various methods.

Meanwhile, when the vertical segment is segmented, the signal sensitivity patterns collected for each segment are analyzed to calculate a representative pattern (S300). Here, a plurality of collection points may exist in one segment, and each collection point may exist at various positions within the segment. Therefore, the representative position can be determined within the segment, and the representative pattern at the representative position can be extracted.

At this time, in order to provide a service considering user's convenience, it is preferable to uniformize the representative position selection criteria in each segment (for example, the center of the segment). Otherwise, if a different representative point is determined for each segment, the position of the user may be perceived as bouncing at a certain point, and this phenomenon may be more serious when the size of the segment becomes larger. Therefore, a specific point such as the center of each segment can be designated as a representative point, and a collected data analysis and an average value derivation process for calculating a representative pattern of the reception sensitivity per segment at the representative point can be performed.

To this end, in the present embodiment, the representative pattern detection step for each segment can be divided into the following steps. FIG. 4 is a flowchart showing a step of detecting a representative signal pattern of a segment in FIG. 1, and FIG. 5 is a plan view showing a state in which each segment is divided into sub segments in FIG.

As shown in FIG. 4, in this embodiment, the representative pattern detection step for each segment is divided into a sub segment segmentation step S310, a sub segment-based representative pattern detection step S320, and a segment representative pattern calculation step using a sub- (S330). In this case, it is possible to construct an accurate radio wave fingerprint map by taking the distribution point distribution into consideration.

5, each segment is divided into a plurality of sub-segments. In this embodiment, four sub-segments (s) having the same shape as a quadrant of a general xy coordinate system are used. (S310).

Then, the reception sensitivity pattern collected for each sub segment may be averaged to detect a representative pattern of the sub segment (S320). The step of detecting the representative pattern of the sub-segment can be detected using the received sensitivity measured at the collection point included in the corresponding sub-segment. In this case, it is possible to approximate the distribution of the collection points according to the presence or absence of each representative pattern by dividing the sub-segments.

On the other hand, when a representative pattern of a sub segment is detected, a step of calculating a representative pattern of the segment is performed. At this time, as shown in Fig. 5, the distribution of the collection points is different for each segment, and the distribution of the collection points of such segments is determined by the number of sub-segments in which the representative pattern exists in the segment (i.e., (The number indicated in the circle in FIG. 5 represents the number of sub-segments having a representative pattern in the corresponding segment). The number of sub-segments having the representative pattern in the segment is represented by the number of sub-segments. Accordingly, in this step, a representative pattern of the corresponding segment may be calculated using a representative pattern of the sub-segment according to whether a representative pattern is detected in each of the sub-segments in the corresponding segment (S330).

The representative pattern calculation step (S330) of this segment will be described in order according to the number of sub-segments having a representative pattern with reference to FIG.

a. When the number of sub-segments having a representative pattern is one

First, a case in which the number of sub-segments having a representative pattern is one in FIG. 6 will be described. In this case, since it is difficult to judge that the collected data is accurately collected in the middle of the segment, it is confirmed whether or not there is data of the adjacent segment, and if there is data, correction is performed using collected data of the adjacent segment to calculate a more accurate segment representative pattern (S331).

On the other hand, if there is no adjacent segment or data is not collected in the adjacent segment, further correction and estimation can not be performed. Therefore, the detected sub-segment representative pattern is directly approximated to the segment representative pattern and mapped (S332).

b. When the number of sub-segments having a representative pattern is two

In this case, the representative pattern can be calculated in consideration of where the two sub-segments having the representative pattern are relatively located in the segment. First, in step S333, it is determined whether the two sub-segments where the data are collected are diagonal sub-segments. If the sub-segments are diagonal sub-segments, the representative patterns of the two sub-segments are averaged to calculate a segment representative pattern.

However, if it is not located diagonally, it is determined whether or not there is a representative pattern of the adjacent segment in the symmetric position, and the corresponding segment representative pattern is generated by using the representative pattern of the adjacent segment in the symmetric position (S334).

If there is no segment adjacent to the symmetric position or there is no data to be used for correction when the segment is not collected at the symmetric position, the average of the representative patterns of the two sub-segments is approximated to the segment representative pattern (S335)

c. When the number of sub-segments having a representative pattern is 3

In this case, the difference between the sub-segment representative patterns in the three sub-segments having the representative pattern among the four sub-segments can be analyzed and the sub-segment representative value can be generated using the difference. If the difference of the sub-segment representative pattern is not large, it is determined that the change of the infrastructure information is not large in the segment. In this case, the infrastructure representative information is simply averaged to generate the segment representative pattern (S336). The threshold value of the difference between the sub-segment representative patterns is selected through experiments.

However, if the difference between the sub-segment representative patterns exceeds the threshold value, it means that the variation of the received sensitivity pattern collected between the sub-segments is large. In this case, in order to generate a correct segment representative pattern, The segment representative pattern can be selected through the correction using the segment representative pattern.

If the representative pattern of the surrounding segment is present, the segment representative pattern is calculated by using the data of the surrounding segment (S337). If there is no collected data of the surrounding segment to be used for correction, the average of the sub- The representative pattern is approximated and used (S338).

d. If you have a representative pattern in all sub-segments

If the collection is done in all four sub-segments and a sub-segment representative pattern is generated, this means that a relatively uniform distribution has been collected in the segment, and thus a segment representative pattern is generated with a simple average of the sub- .

In this way, a radio wave fingerprint map can be constructed by detecting representative patterns of segments and mapping them (S400). Here, there is a possibility that the collected data does not exist at a specific point in the segmentation process due to the arbitrary characteristic of the collected position. In this case, for the continuity of service, it is necessary to construct a more complete propagation map by estimating using the infrastructure information in the adjacent segment. In this case, the same method as the interpolation method used in general can be used. However, if collecting is not carried out for a specific part of the collection area as a whole, no propagation map is made through estimation. This is because, in the case of indoor, there is a space where generation of a radio wave map can not be made, such as a structure having a void structure in which a specific floor is opened from the upper part to the upper part, and the radio wave fingerprint map is not constructed in such a place .

S: segment s: subsegment

Claims (7)

A dividing step of dividing the collection area into a plurality of segments according to a collection area size of a signal for position measurement and a density of collection points;
A representative pattern detection step of collecting a signal at each of the collected positions corresponding to each of the segmented segments, and detecting a representative signal pattern at the representative point by selecting representative points of the segment; And,
And constructing a radio wave fingerprint map using the detected representative signal pattern for each segment.
The method according to claim 1,
Wherein the dividing step includes dividing the space into a small segment where the density of the collection point is higher than the critical density and dividing the portion where the density of the collection point is lower than the critical density into segments larger than the smallest segment. Of the fingerprint image. The method according to claim 1,
Wherein the dividing step comprises the steps of: if the resolution limitation is set in advance, setting a minimum segment capable of dividing the collection area while satisfying the resolution limitation; And
And reducing the size of the first set segment based on the density of the collection point. The method of claim 3,
Wherein the step of dividing further comprises the step of maintaining the size of the initially set segment in a region where the density of the collection point is low and the data is judged to be insufficient, . The method according to claim 1,
Wherein the selection criteria of the representative points in each segment are unified. 6. The method of claim 5,
Wherein the representative point is the center of each segment. ≪ RTI ID = 0.0 > 11. < / RTI > The method according to claim 1,
The representative pattern detection step may include: dividing each segment into four sub-segments corresponding to a quadrant;
Detecting a representative pattern for each sub-segment; And
And deriving a representative pattern of the segment using the representative pattern of each of the sub-segments.

Images