Information processing device

Abstract

To accurately estimate the radio wave environment. [Solution] The information processing device 20 includes an image data generation unit 62 that determines a predetermined range on a map including the estimated location, which is the location where the radio wave environment is to be estimated, as the estimation range, and generates estimation image data 182, which is image data in which one or more types of building information relating to buildings are associated as color information, for each block within the estimation range based on map data 54 of the estimation range, and an estimation unit 68 that estimates the radio wave environment at the estimated location by inputting the estimation image data 182 including the estimated location into a trained machine learning model 58 that outputs an estimation result of the radio wave environment in response to the input of the estimation image data 182.

Description

Technical Field

[0002]

[0001] The present invention relates to an information processing apparatus that performs processing for estimating a radio wave environment.

Background Art

[0002] For example, Patent Document 1 discloses a wireless communication system including a server machine and a plurality of wireless machines (wireless communication terminals) capable of multi-hop wireless communication with the server machine. In such Patent Document 1, connection possibility data indicating an index as to whether communication from the wireless machine to the server machine is possible is predicted based on the position data of the wireless machine.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] For example, at a position where a wireless communication terminal is not actually installed, it is difficult to acquire the actual radio wave environment by the wireless communication terminal, and it has not been possible to appropriately grasp whether the radio wave environment is sufficient. Therefore, for example, when newly installing a wireless communication terminal at a desired position or when newly installing a concentrator (relay device), it is desired to be able to appropriately estimate the radio wave environment at the desired position.

[0005] In view of such problems, an object of the present invention is to provide an information processing apparatus capable of appropriately estimating a radio wave environment.

Means for Solving the Problems

[0006] To solve the above problems, the information processing device of the present invention comprises: an image data generation unit that determines a predetermined range on a map including an estimated position, which is a location where radio wave environment estimation is desired, as an estimation range, and generates estimation image data, which is image data in which one or more types of building information relating to buildings are associated as color information, for each block within the estimation range based on map data of the estimation range; and an estimation unit that estimates the radio wave environment at an estimated position by inputting estimation image data including the estimated position into a trained machine learning model that outputs an estimation result of the radio wave environment in response to the input of estimation image data.

[0007] Furthermore, the image data generation unit may determine the estimation range such that the transmission position of the radio waves subject to estimation of the radio wave environment and the estimated position, which is the reception position of the radio waves subject to estimation of the radio wave environment, are each placed in predetermined positions within the figure that forms the estimation range.

[0008] Furthermore, the information processing device may further include a model generation unit for generating machine learning models. The image data generation unit determines a predetermined range on a map, including the locations of the installed first wireless communication terminal and the installed second wireless communication terminal, as a training range. Based on the map data of the training range, it generates training image data, which is image data in which one or more types of building information are associated as color information for each block within the training range. The model generation unit may then perform machine learning using the training image data and data of the actual radio wave environment at the location of the second wireless communication terminal when the radio waves transmitted from the first wireless communication terminal are received by the second wireless communication terminal, as training data, and generate a machine learning model.

[0009] Furthermore, the information processing device further includes a numerical data generation unit that generates estimation numerical data, which is numerical data correlated with the radio wave environment in the estimation range, based on map data of the estimation range. The trained machine learning model outputs an estimation result of the radio wave environment for input of estimation image data and estimation numerical data. The estimation unit may also estimate the radio wave environment at the estimated location by inputting estimation image data including the estimated location and estimation numerical data in the estimation range into the trained machine learning model.

[0010] Furthermore, the information processing device may further include a model generation unit for generating machine learning models, wherein the image data generation unit determines a predetermined range on a map including the locations of the installed first wireless communication terminal and the installed second wireless communication terminal as the training range, and generates training image data, which is image data in which one or more types of building information relating to a building are associated as color information for each block within the training range based on the map data of the training range, and the numerical data generation unit generates training numerical data, which is numerical data correlated with the radio wave environment in the training range based on the map data of the training range, and the model generation unit may perform machine learning using the training image data, the training numerical data, and data on the actual radio wave environment at the location of the second wireless communication terminal when the radio waves transmitted from the first wireless communication terminal are received by the second wireless communication terminal as training data to generate a machine learning model.

[0011] Furthermore, the estimation range includes the transmission location of the radio waves subject to estimation of the radio wave environment and the estimated location, which is the reception location of the radio waves subject to estimation of the radio wave environment. The estimation numerical data may include at least one of the following: the number of buildings included within the estimation range, the number of buildings within the estimation range that are in contact with the line segment connecting the transmission location and the estimated location, and the ratio of the length of the part of the line segment that is in contact with a building to the length of the line segment.

[0012] Furthermore, one or more types of building information may include at least one of the following: building height, building area, and building type.

[0013] Furthermore, the information processing device may also include a first network design unit, in which multiple combinations of estimated locations, which are the transmission locations of radio waves subject to radio wave environment estimation and the reception locations of radio waves subject to radio wave environment estimation, are specified in a predetermined area where network design is desired. The estimation unit estimates the radio wave environment at the estimated location for each of the specified multiple combinations using a trained machine learning model, and for each of the specified multiple combinations, determines whether or not communication can be established at the estimated location according to the estimation result of the radio wave environment at the estimated location. The first network design unit may extract the combinations from the specified multiple combinations that are determined to be capable of establishing communication, and group the estimated locations in the extracted combinations so that they belong to a single network.

[0014] Furthermore, the information processing device may further include a second network design unit, and the image data generation unit determines an estimation range that includes transmission locations of radio waves subject to estimation of the radio wave environment, which belong to the first network, and estimated locations that are reception locations of radio waves subject to estimation of the radio wave environment, which belong to the second network, generates estimation image data based on map data of the estimation range, and the estimation unit estimates the radio wave environment at the estimated location by inputting the estimation image data including the estimated location into a machine learning model, determines whether or not communication can be established at the estimated location according to the estimation result, and if it is determined that communication can be established at the estimated location, the second network design unit may combine the first network and the second network into a single network. [Effects of the Invention]

[0015] According to the present invention, it becomes possible to appropriately estimate the radio wave environment. [Brief explanation of the drawing]

[0016] [Figure 1]FIG. 1 is a schematic diagram showing an example of the configuration of a wireless communication system according to the first embodiment. [Figure 2] FIG. 2 is a schematic diagram showing an example of a meter and a wireless communication terminal. [Figure 3] FIG. 3 is a schematic diagram for explaining the configuration of an information processing apparatus according to the first embodiment. [Figure 4] FIG. 4 is a flowchart showing an example of the processing flow at the time of generating a machine learning model. [Figure 5] FIG. 5 is a diagram for explaining an example of a teacher range. [Figure 6] FIG. 6 is a diagram for explaining an example of teacher image data. [Figure 7] FIG. 7 is a diagram for explaining an example of teacher numerical data. [Figure 8] FIG. 8 is a conceptual diagram for explaining the outline of estimation using a machine learning model. [Figure 9] FIG. 9 is a flowchart showing an example of the processing flow at the time of estimating a radio wave environment. [Figure 10] FIG. 10 is a conceptual diagram for explaining the outline of estimation using a machine learning model in the information processing apparatus of the second embodiment. [Figure 11] FIG. 11 is a schematic diagram for explaining the configuration of an information processing apparatus according to the third embodiment. [Figure 12] FIG. 12 is a conceptual diagram for explaining the concept of processing executed by the information processing apparatus according to the third embodiment. [Figure 13] FIG. 13 is a flowchart showing an example of the processing flow at the time of estimating a radio wave environment in the third embodiment. [Figure 14] FIG. 14 is a schematic diagram for explaining the configuration of an information processing apparatus according to the fourth embodiment. [Figure 15] FIG. 15 is a conceptual diagram for explaining the concept of processing executed by the information processing apparatus according to the fourth embodiment. [Figure 16] FIG. 16 is a flowchart showing an example of the processing flow at the time of estimating a radio wave environment in the fourth embodiment. [Modes for carrying out the invention]

[0017] Preferred embodiments of the present invention will be described in detail below with reference to the attached drawings. The dimensions, materials, and other specific numerical values ​​shown in these embodiments are merely illustrative to facilitate understanding of the invention and do not limit the present invention unless otherwise specified. In this specification and drawings, elements having substantially the same function and configuration are denoted by the same reference numerals to avoid redundant explanations, and elements not directly related to the present invention are omitted from the illustration.

[0018] (First Embodiment) Figure 1 is a schematic diagram showing an example of the configuration of a wireless communication system 1 according to the first embodiment. The wireless communication system 1 includes a plurality of meters 10, a plurality of wireless communication terminals 12, a plurality of concentrators 14, a base station 16, and an information processing device 20.

[0019] Meter 10 is, for example, a smart meter and is installed on a per-customer basis. Specifically, meter 10 is a gas meter that automatically reads the amount of gas used by a customer when a gas company supplies gas to the customer. Alternatively, the meter may be an electricity meter that automatically reads the amount of electricity used by a customer when an electricity company supplies electricity to the customer.

[0020] Each wireless communication terminal 12 is installed in a one-to-one correspondence with each meter 10. Each wireless communication terminal 12 can communicate wirelessly with other wireless communication terminals 12. The concentrator 14 is installed in correspondence with one of the multiple wireless communication terminals 12 and can communicate via wired connection with the corresponding wireless communication terminal 12. The concentrator 14 can communicate wirelessly with the information processing device 20 via the base station 16. In other words, the concentrator 14 is a relay device that relays communication between the wireless communication terminals 12 and the information processing device 20.

[0021] The wireless communication terminal 12 is powered, for example, by a battery. To conserve battery power, the wireless communication terminal 12 performs wireless communication over relatively short distances. As a result, not all wireless communication terminals 12 can directly communicate with the wireless communication terminal 12 associated with the concentrator 14. Therefore, the wireless communication terminal 12 hops through one or more other wireless communication terminals 12 that are capable of wireless communication to communicate with the wireless communication terminal 12 associated with the concentrator 14. The wireless communication terminal 12 associated with the concentrator 14 then communicates with the concentrator 14 via a wired connection, and the concentrator 14 communicates with the information processing device 20 via wireless communication through the base station 16. In this way, each wireless communication terminal 12 is capable of multi-hop wireless communication with the information processing device 20.

[0022] As described above, the wireless communication terminal 12 is designed for relatively short-range wireless communication. Therefore, depending on the installation location of the wireless communication terminal 12 or the surrounding environment, wireless communication between the wireless communication terminals 12 may not function effectively, and as a result, wireless communication between the wireless communication terminal 12 and the information processing device 20 may not be established. For example, if the wireless communication terminals 12 are too far apart, or if there are obstacles between the wireless communication terminals 12 that interfere with communication, wireless communication may not function effectively. In addition, although the wireless communication terminal 12 is installed together with the meter 10 in accordance with the meter's expiration timing, if it is installed in a location isolated from other wireless communication terminals 12, wireless communication between it and the information processing device 20 may not function effectively.

[0023] As will be described in detail later, the information processing device 20 of the wireless communication system 1 in the first embodiment estimates the radio wave environment at a desired location, such as the location where the wireless communication terminal 12 will be newly installed.

[0024] Figure 2 is a schematic diagram showing an example of a meter 10 and a wireless communication terminal 12. The meter 10 includes a measurement unit 30 and a storage unit 32. The measurement unit 30 reads the meter of the object to be measured in the meter 10. The object to be measured is gas usage if the meter 10 is a gas meter, and electricity usage if the meter 10 is an electricity meter. The measurement unit 30 stores the meter reading data showing the meter reading results in the storage unit 32.

[0025] The wireless communication terminal 12 includes a communication unit 34. As described above, the communication unit 34 establishes wireless communication with the information processing device 20 through other wireless communication terminals 12, concentrators 14, and base stations 16. The communication unit 34 transmits the meter reading data stored in the memory unit 32 of the meter 10 to the information processing device 20 at a predetermined timing. For example, the communication unit 34 may transmit the meter reading data to the information processing device 20 at a predetermined time within a 24-hour period.

[0026] Furthermore, the communication unit 34 is not limited to transmitting meter reading data once a day; for example, it may transmit meter reading data when the meter reading value is updated. The communication unit 34 may also transmit meter reading data in response to a transmission command for meter reading data received from the information processing device 20. In addition, the communication unit 34 and the information processing device 20 may transmit and receive various types of data, not limited to meter reading data, such as an identifier that uniquely identifies the wireless communication terminal 12, location data indicating the location of the wireless communication terminal 12, and the signal strength of the radio waves received by the wireless communication terminal 12. Therefore, the information processing device 20 can acquire information about each meter 10 and each wireless communication terminal 12 through communication with each wireless communication terminal 12.

[0027] Figure 3 is a schematic diagram illustrating the configuration of the information processing device 20 according to the first embodiment. The information processing device 20 is composed of a computer and the like, and belongs to the administrator side of the wireless communication system 1. The information processing device 20 includes a communication unit 40, an input / output device 42, storage 44, a processor 46, and memory 48.

[0028] The communication unit 40 can communicate wirelessly with each wireless communication terminal 12 through the base station 16 and the concentrator 14. The input / output device 42 includes an output device such as a display device and an input device such as a touch panel or keyboard.

[0029] The storage 44 is composed of non-volatile memory elements. The storage 44 stores, for example, customer information 50, terminal information 52, map data 54, building information 56, and a machine learning model 58. The building information 56 is associated with the map data 54 and may, for example, be included in the map data 54.

[0030] The customer information 50 includes, for example, information about the customer to whom the meter 10 is installed, and location data indicating the location of the wireless communication terminal 12 installed on the meter 10. The location data of the wireless communication terminal 12 may be stored, for example, when the wireless communication terminal 12 is installed, and updated each time the number of wireless communication terminals 12 increases or decreases. Since the wireless communication terminal 12 is installed on the meter 10, the location data of the wireless communication terminal 12 may be substantially the same as the meter location data indicating the installation location of the meter 10 to which the wireless communication terminal 12 is installed. The location data of the wireless communication terminal 12 is registered, for example, when the meter 10 to which the wireless communication terminal 12 is attached is installed at a customer's location.

[0031] Terminal information 52 may include communication history information representing the communication history of the wireless communication terminal 12, and radio wave environment information representing the radio wave environment at the location of the wireless communication terminal 12. The radio wave environment information may include, for example, the received strength of radio waves transmitted from another wireless communication terminal 12 when such radio waves are received. The radio wave environment information may be updated, for example, when the wireless communication terminal 12 communicates with another wireless communication terminal 12.

[0032] The map data 54 includes graphic data of buildings (structures) that represent their location, size, and extent. The map data 54 includes graphic data of buildings within a predetermined range that includes at least the location of the wireless communication terminal 12. The map data 54 may also be aerial or satellite imagery. The map data may be updated periodically, for example, every year.

[0033] Building information 56 is information that represents the characteristics of individual buildings in the map data 54. Building information 56 may include, for example, the height of the building, the area of ​​the building, and the type of building. Building information 56 may also include the site (land) area. The type of building may include, for example, the material, such as whether it is made of wood or reinforced concrete. The type of building may include the type of housing, such as whether it is an apartment building or a detached house. The type of building may include information on the number of floors, such as how many floors it has. The type of building may also include information on its use, such as whether it is a commercial building or a general-use building. Note that building information 56 is not limited to the types of information exemplified above and may include various types of information that represent the characteristics of the building.

[0034] The information processing device 20 of the first embodiment estimates the radio wave environment at a desired location using a trained machine learning model 58. The machine learning model 58 stored in the storage 44 is a machine learning model 58 trained to estimate the radio wave environment. The training of the machine learning model 58 and the estimation of the radio wave environment will be described in detail later.

[0035] The memory 48 is connected to the processor 46 and includes ROM for storing programs and RAM as a work area. The processor 46 executes various processes in the information processing device 20 by executing programs. For example, by executing programs, the processor 46 functions as a data acquisition unit 60, an image data generation unit 62, a numerical data generation unit 64, a model generation unit 66, and an estimation unit 68.

[0036] The data acquisition unit 60 may establish wireless communication with the communication unit 40 and acquire various types of data from each wireless communication terminal 12, such as meter reading data, customer information, terminal information, map data, and building data. The data acquisition unit 60 stores the data acquired from the wireless communication terminals 12 in the storage 44.

[0037] Below, we will first explain the training of machine learning model 58, and then explain estimation using the trained machine learning model 58.

[0038] Figure 4 is a flowchart showing an example of the processing flow when generating a machine learning model 58. First, the image data generation unit 62 determines the location of the installed first wireless communication terminal among the multiple wireless communication terminals 12 as the transmission position and the location of the installed second wireless communication terminal as the reception position (S10).

[0039] Next, the image data generation unit 62 refers to the map data 54 and determines a predetermined area on the map, including the transmission location (location of the installed first wireless communication terminal) and the reception location (location of the installed second wireless communication terminal), as the training area (S11). The training area can be divided into multiple blocks of predetermined size.

[0040] Figure 5 illustrates an example of a teacher's area. In Figure 5, frame 100 shows an example of a teacher's area. For example, as illustrated by frame 100, the teacher's area may be formed as a rectangle on the map.

[0041] In Figure 5, circle 110 indicates an example of the location of the installed first wireless communication terminal, i.e., the transmission location. In Figure 5, circle 112 indicates an example of the location of the installed second wireless communication terminal, i.e., the reception location.

[0042] As shown in Figure 5, the image data generation unit 62 determines the training range such that the location of the installed first wireless communication terminal (transmission location) and the location of the installed second wireless communication terminal (reception location) are included within the training range.

[0043] More specifically, the image data generation unit 62 determines the training area such that the location of the installed first wireless communication terminal (transmission position) and the location of the installed second wireless communication terminal (reception position) are positioned at predetermined locations within the figure that forms the training area.

[0044] For example, as shown in Figure 5, the image data generation unit 62 may determine the training range such that the transmission position, exemplified by the circle 110, is located at the lower left corner of the rectangle, and the reception position, exemplified by the circle 112, is located at the upper right corner of the rectangle.

[0045] Furthermore, the shape of the teacher's area is not limited to a quadrilateral; it may be any shape, such as a polygon other than a quadrilateral. Also, the method for determining the teacher's area is not limited to the example described above; various methods may be adopted under predetermined rules. For example, the teacher's area may be determined such that the transmission position is located in the center of the bottom edge of the quadrilateral and the reception position is located in the center of the top edge of the quadrilateral.

[0046] Let's return to Figure 4 for explanation. The image data generation unit 62 generates training image data (S12) based on the map data 54 of the determined training area.

[0047] The teacher image data is image data in which one or more types of building information 56 relating to a building are associated as color information for each block within the teacher range. The one or more types of building information 56 may include, for example, at least one of the following: building height, building area, and building type.

[0048] Figure 6 illustrates an example of teacher image data. Figures 6(A) and 6(D) show examples of map color image data 120. Map color image data 120 is image data in which one or more types of building information 56 relating to buildings are associated as color information with each block of map data 54. Because the characteristics of each building differ, the image of map color image data 120 is represented, for example, as a mottled pattern containing multiple colors.

[0049] Figures 6(A), 6(B), and 6(C) illustrate a first example of generating training image data, while Figures 6(D), 6(E), and 6(F) illustrate a second example of generating training image data.

[0050] As shown in Figures 6(A) and 6(D), the image data generation unit 62 generates map color image data 120 based on the map data 54. In the map color image data 120, the image of each block of the map data 54 is represented by a color corresponding to the characteristics of the buildings in that block.

[0051] For example, in the map color image data 120, the colors of a one-story building and a two-story building may be different. Also, for example, if the height of a building is used as building information 56, the image of the location of the building may be displayed in a color corresponding to the height of that building.

[0052] The color information corresponding to the building information may be color information represented by the three primary colors (RGB (red, green, blue)) or the three attributes of color (hue, lightness, saturation), or it may be grayscale.

[0053] If the building information 56 includes characteristics of multiple types of buildings, the color information corresponding to the building information 56 may be identified by a combination of the multiple types of building information 56. For example, the height of a building may be associated with the "R (red)" element, the area of ​​a building with the "G (green)" element, and the type of building with the "B (blue)" element, and the building information 56 and the color information may be associated by a combination of these building heights, building areas, and building types.

[0054] As shown in Figures 6(A) and 6(D), the image data generation unit 62 generates pre-image data for training, as shown in Figures 6(B) and 6(E), by extracting the map color image data 120 of the training area exemplified by the frame 100 from the map color image data 120.

[0055] In essence, the image data generation unit 62 generates pre-image data for training based on the map data 54 of the training area, in which one or more types of building information 56 relating to buildings are associated as color information for each block within the training area.

[0056] For example, as shown in Figure 6(B), if the shape of the training area exemplified by frame 100 is a vertically elongated rectangle, training pre-image data 130 of that vertically elongated rectangle is generated. Also, for example, as shown in Figure 6(E), if the shape of the training area exemplified by frame 100 is a horizontally elongated rectangle, training pre-image data 130 of that horizontally elongated rectangle is generated.

[0057] As shown in Figures 6(C) and 6(F), the image data generation unit 62 generates teacher image data by normalizing the teacher pre-image data 130 into a square image data 132 of a specified size. In the normalized teacher image data, the shape and dimensions of the teacher range are normalized, as are the shape and dimensions of the color images within the teacher range.

[0058] As shown in Figures 6(B) and 6(E), even if the shapes and dimensions of the figures in the training pre-image data 130 differ, as shown in Figures 6(C) and 6(F), normalization can generate training image data with substantially the same shapes and dimensions. As a result, the information processing device 20 can perform training using multiple training image data with substantially identical shapes and dimensions, thereby improving the accuracy of the training.

[0059] In this example, we have described how to generate teacher image data by normalizing the teacher pre-image data 130. However, the normalization process can be omitted, and the teacher pre-image data 130 itself can be used as the teacher image data. However, the method of generating teacher image data by normalizing the teacher pre-image data 130 is more preferable.

[0060] Furthermore, this example describes how to generate map color image data 120 from map data 54 and extract the map color image data 120 for the training area from the map color image data 120 to generate pre-image data 130 for the training area. However, the image data generation unit 62 may also extract the map data 54 for the training area from the map data 54 and generate pre-image data 130 for the training area from the extracted map data 54 for the training area.

[0061] Let's return to Figure 4 for explanation. After generating the training image data, the numerical data generation unit 64 generates training numerical data based on the training area map data 54 from the map data 54 (S13). The training numerical data is numerical data that correlates with the radio wave environment in the training area.

[0062] Figure 7 illustrates an example of teacher data. Figure 7 shows an example of a map represented by map data 54 within the teacher range exemplified by frame 100. The hatching 140 in Figure 7 shows an example of the location, size, and extent of buildings on the map.

[0063] In Figure 7, circle 110 indicates an example of the location of the installed first wireless communication terminal, i.e., the transmission location. In Figure 7, circle 112 indicates an example of the location of the installed second wireless communication terminal, i.e., the reception location. In Figure 7, the line 142 indicates the line segment connecting the transmission location of circle 110 and the reception location of circle 112.

[0064] Here, the following three indicators (1) to (3) are specific examples of indicators that correlate with the radio wave environment within the scope of teacher training. (1) The number of buildings included in the teacher's area. (2) The number of buildings within the teacher's area that are in contact with the line segment (i.e., straight line 142) connecting the location of the first wireless communication terminal (i.e., the transmission location marked with a circle 110) and the location of the second wireless communication terminal (i.e., the receiving location marked with a circle 112). (3) The ratio of the length of the part of the line segment (i.e., straight line 142) that is in contact with the building to the length of the line segment.

[0065] In the example in Figure 7, (1) above corresponds to the number of hatched shapes 140 contained within frame 100, which is, for example, 16. Note that the number of buildings included in the teacher's area may be calculated as one even if it is only a part of a building.

[0066] Furthermore, in the example in Figure 7, (2) above corresponds to the number of hatching figures 140 that are in contact with the straight line 142, and is, for example, 4.

[0067] Furthermore, in the example in Figure 7, (3) above corresponds to the value obtained by dividing the length of the part of the line 142 that is in contact with the hatching 140 by the length of the line 142, which is, for example, approximately 40%.

[0068] The numerical data for teachers may include at least one of the three indicators (1) to (3) above.

[0069] Furthermore, the numerical data for teachers may include not only the indicators (1) to (3) above, but also any numerical data that correlates with the radio wave environment within the teacher's scope.

[0070] Let's return to Figure 4 for explanation. After generating the numerical data for training, the model generation unit 66 acquires data on the actual radio wave environment at the position of the second wireless communication terminal (reception position) when the radio waves transmitted from the first wireless communication terminal are received by the second wireless communication terminal (S14).

[0071] For the sake of explanation, actual radio wave environment data is sometimes referred to as measured radio wave environment data. Measured radio wave environment data may include, for example, the actual radio wave intensity value at the location (reception position) of the second wireless communication terminal. The radio wave intensity value may be a numerical value, or it may be a rank value obtained by dividing the numerical value into multiple rank levels.

[0072] Next, the information processing device 20 determines whether a sufficient number of sets of training image data, training numerical data, and radio wave environment measurement data have been prepared for the training to be performed (S15). If it determines that the number of sets is insufficient (NO in S15), the information processing device 20 returns to step S10, changes the combination of transmission and reception positions, and repeats the processing from S11 onwards.

[0073] If the information processing device 20 determines that the number of sets is sufficient (YES in S15), the model generation unit 66 generates a machine learning model 58 (S16). More specifically, the model generation unit 66 performs machine learning using training image data within the training range, training numerical data within the said training range, and data on the actual radio wave environment at the location (receiving position) of the second wireless communication terminal (measured radio wave environment data) as training data to generate a machine learning model 58. The type of machine learning can be any known algorithm, such as deep learning.

[0074] The model generation unit 66 stores the generated machine learning model 58 in the storage 44 (S17), and then terminates the series of processes shown in Figure 4.

[0075] Figure 8 is a conceptual diagram illustrating the general outline of estimation using the machine learning model 58. As shown in Figure 8, the model generation unit 66 generates the machine learning model 58 using the training image data 172, training numerical data 174, and measured radio wave environment data 176 as training data 170.

[0076] As shown in Figure 8, the estimation unit 68 inputs the estimation image data 182 and estimation numerical data 184 as input data 180 to the machine learning model 58, and outputs the radio wave environment estimation data 192 as output data 190.

[0077] The estimation image data 182 is image data generated for estimation purposes using the same method as the training image data 172. The estimation numerical data 184 is numerical data generated for estimation purposes using the same method as the training numerical data 174. The radio wave environment estimation data 192 is the estimation result of the radio wave environment at the estimation location, which is the location where the estimation of the radio wave environment is desired. The estimation location is the reception location of the radio waves that are the subject of the radio wave environment estimation.

[0078] In other words, the trained machine learning model 58 outputs an estimated result of the radio wave environment in response to the inputs of estimation image data 182 and estimation numerical data 184.

[0079] Figure 9 is a flowchart showing an example of the processing flow when estimating the radio wave environment. First, the transmission location of the radio wave to be estimated and the estimated location, which is the reception location of the radio wave, are specified (S30). The specification of the transmission location and estimated location may be done by an administrator managing the information processing device 20 by inputting the transmission location and estimated location into the information processing device 20, or it may be done by the information processing device 20 according to a predetermined program.

[0080] For example, the location of an existing first wireless communication terminal among multiple wireless communication terminals may be designated as the transmission location, and the desired location for a new second wireless communication terminal to receive radio waves from the first wireless communication terminal may be designated as the estimated location.

[0081] Furthermore, for example, when both a first wireless communication terminal that transmits radio waves and a second wireless communication terminal that receives those radio waves are newly installed, the location where the first wireless communication terminal is newly installed may be designated as the transmission location, and the location where the second wireless communication terminal is newly installed may be designated as the estimated location.

[0082] Next, the image data generation unit 62 refers to the map data 54 and determines a predetermined range on the map that includes the specified transmission location and estimated location as the estimation range (S31). The estimation range can be divided into multiple blocks of a predetermined size.

[0083] Since the estimated location is usually different from the location of the installed first wireless communication terminal or the installed second wireless communication terminal when the teacher's range was determined, the location of the estimation range on the map may be different from the location of the teacher's range on the map.

[0084] If the teacher's area is a quadrilateral, the estimation area is also assumed to be a quadrilateral. In this case, the aspect ratio of the teacher's area quadrilateral may differ from the aspect ratio of the estimation area quadrilateral. The area of ​​the teacher's area may also differ from the area of ​​the estimation area.

[0085] Furthermore, the image data generation unit 62 determines the estimation range such that the transmission position of the radio waves to be estimated and the estimated position, which is the reception position of the radio waves to be estimated, are each placed in predetermined positions within the figure that forms the estimation range. More specifically, the image data generation unit 62 determines the estimation range such that the transmission position in the figure of the training range corresponds to the transmission position in the figure of the estimation range, and the reception position in the figure of the training range corresponds to the estimated position in the figure of the estimation range.

[0086] For example, when determining the teacher's range, if the teacher's range is determined such that the transmission position is located at the lower left corner of the rectangle and the reception position is located at the upper right corner of the rectangle, then when determining the estimation range, the estimation range will also be determined such that the transmission position is located at the lower left corner of the rectangle and the estimation position is located at the upper right corner of the rectangle.

[0087] Next, the image data generation unit 62 generates estimation image data 182 based on the map data 54 of the determined estimation range from the map data 54 (S32).

[0088] The estimation image data 182, like the training image data 172, is image data in which one or more types of building information 56 related to a building are associated as color information with each block within the estimation range.

[0089] The one or more types of building information 56 in the estimation image data 182 may include, for example, at least one of the following: building height, building area, and building type, similar to the one or more types of building information 56 in the training image data 172.

[0090] Furthermore, the association between building information 56 and color information in the estimation image data 182 may be substantially the same as the association between building information 56 and color information in the training image data. For example, in the estimation image data, the height of the building may be associated with the "R (red)" element, the area of ​​the building with the "G (green)" element, and the type of building with the "B (blue)" element, and the building information 56 and color information may be associated with these combinations of building height, building area, and building type.

[0091] More specifically, the image data generation unit 62 generates pre-estimation image data based on the map data 54 of the estimation area, in which one or more types of building information 56 relating to buildings are associated as color information for each block within the estimation area.

[0092] The image data generation unit 62 may generate map color image data 120 from the map data 54 and extract the map color image data 120 for the estimation range from the map color image data 120 to generate pre-estimated image data for the estimation range. Alternatively, the image data generation unit 62 may extract the map data 54 for the estimation range from the map data 54 and generate pre-estimated image data for the estimation range from the extracted map data 54 for the estimation range.

[0093] The image data generation unit 62 generates estimation image data 182 by normalizing the estimation pre-image data into square image data of a predetermined size. The predetermined size of the square may be substantially the same as the predetermined size of the square used when normalizing the teacher pre-image data.

[0094] Furthermore, similar to the method for generating training image data, when generating estimation image data, the normalization process may be omitted, and the estimation pre-image data itself may be used as the estimation image data.

[0095] Next, the numerical data generation unit 64 generates estimated numerical data 184 based on the map data 54 for the estimation range (S33). The estimated numerical data 184 is numerical data that correlates with the radio wave environment in the estimation range.

[0096] The estimation numerical data 184, like the training numerical data 174, may include at least one of the following: the number of buildings included in the estimation range, the number of buildings within the estimation range that are in contact with the line segment connecting the transmission position and the estimation position (reception position), and the ratio of the length of the part of the line segment that is in contact with a building to the length of the line segment.

[0097] Next, the estimation unit 68 inputs estimation image data 182 including the estimated location and estimation numerical data 184 within the estimation range into the trained machine learning model 58 to derive radio wave environment estimation data 192 (S34). The radio wave environment estimation data 192 represents an estimated value of the radio wave environment (e.g., radio wave intensity) at the estimated location. Therefore, the estimation unit 68 can estimate the radio wave environment at a desired location using machine learning.

[0098] The estimation unit 68 notifies the administrator of the estimation results by displaying the derived radio wave environment estimation data 192 on a display device (S35), and then terminates the series of processes shown in Figure 9.

[0099] As described above, the information processing device 20 of the first embodiment determines a predetermined range on a map that includes the estimated location, which is the location where the radio wave environment is to be estimated, as the estimation range, and generates estimation image data 182, which is image data in which one or more types of building information 56 relating to buildings are associated as color information for each block within the estimation range, based on the map data 54 of the estimation range. The information processing device 20 of the first embodiment estimates the radio wave environment at the estimated location by inputting the estimation image data 182, which includes the estimated location, into a trained machine learning model 58 that outputs an estimation result of the radio wave environment in response to the input of the estimation image data 182.

[0100] In other words, in the information processing device 20 of the first embodiment, the radio wave environment at the estimated location is estimated using estimation image data 182, in which one or more types of building information 56 relating to a building are associated as color information. As a result, in the information processing device 20 of the first embodiment, compared to an embodiment that uses a normal map in which buildings are represented as mere shapes, the characteristics of the building are reflected in the estimation image data 182 in more detail by the use of color information, and a lot of information about the building is indirectly referenced as a basis for estimating the radio wave environment.

[0101] Therefore, the information processing device 20 of the first embodiment can appropriately estimate the radio wave environment and improve the accuracy of the radio wave environment estimation.

[0102] Furthermore, in the information processing device 20 of the first embodiment, in addition to the estimation image data 182, estimation numerical data 184, which is numerical data correlated with the radio wave environment in the estimation range, is also used to estimate the radio wave environment at the estimated location.

[0103] As a result, in the information processing device 20 of the first embodiment, compared to the example that uses only the estimation image data 182, the amount of information available for estimating the radio wave environment is further increased, making it possible to estimate the radio wave environment more appropriately and further improve the accuracy of the radio wave environment estimation.

[0104] (Second Embodiment) Figure 10 is a conceptual diagram illustrating the general outline of estimation using the machine learning model 58 in the information processing device 220 of the second embodiment.

[0105] In the first embodiment described above, training image data 172, training numerical data 174, and measured radio wave environment data 176 were used to generate the machine learning model 58. In the first embodiment described above, estimated radio wave environment data 192 was output by inputting estimation image data 182 and estimation numerical data 184 into the machine learning model 58 (see Figure 8).

[0106] In contrast, as shown in Figure 10, the model generation unit 66 of the second embodiment generates a machine learning model 58 based on the training image data 172 and the measured radio wave environment data 176, without using the training numerical data 174. Then, as shown in Figure 10, the estimation unit 68 of the second embodiment outputs estimated radio wave environment data 192 by inputting the estimated image data 182 into the machine learning model 58, without using the estimation numerical data 184.

[0107] Thus, the information processing device 220 of the second embodiment estimates the radio wave environment at the estimated location by inputting the estimation image data 182, which includes the estimated location, into a trained machine learning model 58 that outputs an estimation result of the radio wave environment in response to the input of estimation image data 182.

[0108] In other words, the information processing device 220 of the second embodiment, similar to the information processing device 20 of the first embodiment, uses estimation image data 182 in which one or more types of building information 56 relating to a building are associated as color information to estimate the radio wave environment at the estimated location.

[0109] Therefore, the information processing device 220 of the second embodiment can appropriately estimate the radio wave environment, similar to the information processing device 20 of the first embodiment, and can improve the accuracy of the radio wave environment estimation.

[0110] (Third embodiment) Figure 11 is a schematic diagram illustrating the configuration of the information processing device 320 according to the third embodiment. As shown in Figure 11, the processor 46 of the information processing device 320 in the third embodiment also functions as a first network design unit 330 by executing a program, compared to the first embodiment.

[0111] Figure 12 is a conceptual diagram illustrating the concept of processing performed by the information processing device 320 according to the third embodiment. In the wireless communication system 1, one network is formed by one concentrator 14 and multiple wireless communication terminals 12 that can communicate directly or indirectly with it.

[0112] Here, for example, when designing the installation location of the concentrator 14, one concentrator 14 is installed for one network consisting of multiple wireless communication terminals 12 that can communicate with each other. In this case, suppose that one of the multiple wireless communication terminals 12, wireless communication terminal 12A, is not installed. As shown by the dashed arrow 350 in Figure 12, if the uninstalled wireless communication terminal 12A can establish communication with another wireless communication terminal 12 (for example, wireless communication terminal 12B in Figure 12), then, as shown by the dashed line 352, the uninstalled wireless communication terminal 12A can be included in one network. In this case, one concentrator 14 can be installed for one network that includes the uninstalled wireless communication terminal 12A.

[0113] Therefore, the information processing device 320 of the third embodiment estimates the radio wave environment at the location of the wireless communication terminal 12A in accordance with the first or second embodiment described above, and determines whether communication can be established based on the estimation result. Then, the information processing device 320 of the third embodiment designs a network that groups the multiple wireless communication terminals 12, including the wireless communication terminal 12A for which communication has been determined to be possible, as a single network, as shown by the dashed line 352.

[0114] Figure 13 is a flowchart showing an example of the processing flow when estimating the radio wave environment in the third embodiment. Here, we will explain the differences from the flowchart in Figure 9 of the first embodiment, and for convenience, we will omit explanations of the common points.

[0115] First, in a predetermined area where a network consisting of wireless communication terminals 12 is desired, the transmission location and estimated location are specified (S40). The specification of the transmission location and estimated location may be done by an administrator managing the information processing device 20 or a network designer inputting the transmission location and estimated location into the information processing device 20, or it may be done by the information processing device 20 according to a predetermined program.

[0116] Next, the image data generation unit 62 determines an estimation range that includes the specified transmission location and estimated location (S31). Based on the map data 54 of the determined estimation range, the image data generation unit 62 generates estimation image data 182 (S32). Based on the map data 54 of the determined estimation range, the numerical data generation unit 64 generates estimation numerical data 184 (S33). The estimation unit 68 inputs the estimation image data 182 including the estimated location and the estimation numerical data 184 in the estimation range into the trained machine learning model 58 to derive radio wave environment estimation data 192 (S34).

[0117] In accordance with the second embodiment, the estimation unit 68 may derive radio wave environment estimation data 192 by inputting estimation image data 182 including the estimated position into a trained machine learning model 58.

[0118] The estimation unit 68 determines whether or not communication can be established at the estimated location according to the derived radio wave environment estimation data 192 (i.e., the estimation result of the radio wave environment at the estimated location) (S41). For example, the estimation unit 68 may determine that communication can be established if the estimated value of the radio wave intensity included in the derived radio wave environment estimation data 192 is equal to or greater than a predetermined threshold, and determine that communication cannot be established if the estimated value of the radio wave intensity is less than the predetermined threshold.

[0119] Next, the information processing device 320 determines whether a combination of transmission location and estimated location has been specified for all candidate combinations of transmission location and estimated location in a predetermined region where network design is desired (S42). If it determines that there are unspecified combinations (NO in S42), the information processing device 320 returns to step S40, specifies the transmission location and estimated location for the unspecified combination, and repeats the processing from step S31 onwards.

[0120] In other words, the information processing device 320 specifies multiple combinations of a transmission location of radio waves subject to radio wave environment estimation and an estimated location, which is the reception location of radio waves subject to radio wave environment estimation, in a predetermined area where network design is desired. For each of the specified multiple combinations, the estimation unit 68 estimates the radio wave environment at the estimated location using a trained machine learning model 58. For each of the specified multiple combinations, the estimation unit 68 determines whether or not communication can be established at the estimated location according to the estimation result of the radio wave environment at the estimated location.

[0121] If it is determined that a combination of transmission location and estimated location has been specified for all candidate combinations (YES in S42), the first network design unit 330 extracts the combinations from the specified multiple combinations that the estimation unit 68 has determined to be able to establish communication (S43).

[0122] The first network design unit 330 groups the estimated positions in the extracted combinations so that they belong to one network (S44).

[0123] The first network design unit 330 notifies the administrator or designer of the network design results by displaying the grouped estimated positions on a display device (S45), and then terminates the series of processes shown in Figure 13.

[0124] Thus, in the information processing device 320 of the third embodiment, multiple estimated locations for which communication can be established are grouped together.

[0125] Therefore, in the third embodiment of the information processing device 320, by installing wireless communication terminals 12 at a group of estimated locations, the network composed of wireless communication terminals 12 can be easily designed. As a result, in the third embodiment of the information processing device 320, it is easy to design a system in which one concentrator 14 is installed in one network composed of wireless communication terminals 12.

[0126] (Fourth Embodiment) Figure 14 is a schematic diagram illustrating the configuration of the information processing device 420 according to the fourth embodiment. As shown in Figure 14, the processor 46 of the information processing device 420 in the fourth embodiment also functions as a second network design unit 430 by executing a program, compared to the third embodiment.

[0127] Figure 15 is a conceptual diagram illustrating the concept of processing performed by the information processing device 420 according to the fourth embodiment. As described above, in the wireless communication system 1, one network is formed by one concentrator 14 and multiple wireless communication terminals 12 that can communicate directly or indirectly.

[0128] Here, for example, suppose that the first network illustrated by the dashed line 450 in Figure 15 and the second network illustrated by the double-dashed line 452 in Figure 15 are adjacent to each other. In such an example, if, for example, a wireless communication terminal 12D belonging to the second network can establish communication with a wireless communication terminal 12C belonging to the first network, as shown by the dashed arrow 454 in Figure 15, then the first network and the second network can be combined into a single network, as shown by the dashed line 456. In this case, since only one concentrator 14 is needed for each network, if there is a concentrator 14 in both the first network and the second network, one of those concentrators 14 can be omitted.

[0129] Therefore, the information processing device 420 of the fourth embodiment, in accordance with the first or second embodiment described above, takes the position of the wireless communication terminal 12C as the transmission position, estimates the radio wave environment at the position of the wireless communication terminal 12D, and determines whether communication can be established based on the estimation result. If the information processing device 420 of the fourth embodiment determines that communication can be established, it combines the first network (dotted line 450) and the second network (double-dotted line 452) into a single network (dashed line 456).

[0130] Figure 16 is a flowchart showing an example of the processing flow when estimating the radio wave environment in the fourth embodiment. Here, we will explain the differences from the flowchart in Figure 9 of the first embodiment, and for convenience, we will omit explanations of the common points.

[0131] First, the transmission location of the radio waves subject to estimation of the radio wave environment, which belongs to the first network (for example, the location of wireless communication terminal 12C in Figure 15), and the estimated reception location of the radio waves subject to estimation of the radio wave environment, which belongs to the second network (for example, the location of wireless communication terminal 12D in Figure 15), are specified (S50). The specification of the transmission location and estimated location may be performed by an administrator managing the information processing device 20 or a network designer inputting the transmission location and estimated location into the information processing device 20, or it may be performed by the information processing device 20 according to a predetermined program.

[0132] Next, the image data generation unit 62 determines an estimation range that includes the specified transmission location and estimated location (S31). Based on the map data 54 of the determined estimation range, the image data generation unit 62 generates estimation image data 182 (S32). Based on the map data 54 of the determined estimation range, the numerical data generation unit 64 generates estimation numerical data 184 (S33). The estimation unit 68 inputs the estimation image data 182 including the estimated location and the estimation numerical data 184 in the estimation range into the trained machine learning model 58 to derive radio wave environment estimation data 192 (S34).

[0133] In accordance with the second embodiment, the estimation unit 68 may derive radio wave environment estimation data 192 by inputting estimation image data 182 including the estimated position into a trained machine learning model 58.

[0134] The estimation unit 68 determines whether or not communication can be established at the estimated location according to the derived radio wave environment estimation data 192 (i.e., the estimation result of the radio wave environment at the estimated location) (S51). For example, the estimation unit 68 may determine that communication can be established if the estimated value of the radio wave intensity included in the derived radio wave environment estimation data 192 is equal to or greater than a predetermined threshold, and determine that communication cannot be established if the estimated value of the radio wave intensity is less than the predetermined threshold.

[0135] If it is determined that communication can be established at the estimated location (YES in S51), the second network design unit 430 combines the first network to which the transmission location belongs and the second network to which the estimated location belongs into a single network (S52).

[0136] On the other hand, if it is determined that communication cannot be established at the estimated location (NO in S51), the second network design unit 430 separates the first network to which the transmission location belongs from the second network to which the estimated location belongs (S53).

[0137] After step S52 or step S53, the second network design unit 430 notifies the administrator or designer of the network design results by displaying them on a display device (S54), and then terminates the series of processes shown in Figure 16.

[0138] Furthermore, after combining the first network and the second network into a single network, the second network design unit 430, the administrator, or the designer may design the system to install one concentrator 14 on at least one of the wireless communication terminals 12 in either the first network or the second network. In that case, if both the first network and the second network have concentrators 14, the design may be modified to omit one of those concentrators 14.

[0139] Thus, in the information processing device 420 of the fourth embodiment, when it is determined that communication can be established between the wireless communication terminal 12 of the first network and the wireless communication terminal 12 of the second network, the first network and the second network are combined into a single network.

[0140] Therefore, in the information processing device 420 of the fourth embodiment, the network composed of wireless communication terminals 12 can be easily designed. As a result, in the information processing device 420 of the fourth embodiment, it is easy to design a system in which one concentrator 14 is installed in one network composed of wireless communication terminals 12. In addition, in the information processing device 420 of the fourth embodiment, it is possible to suppress the number of concentrators 14 from becoming excessively large, and thus reduce the cost of constructing the wireless communication system 1.

[0141] Preferred embodiments of the present invention have been described above with reference to the attached drawings, but it goes without saying that the present invention is not limited to these embodiments. It is clear to those skilled in the art that various modifications or alterations can be conceived within the scope of the claims, and these will naturally also fall within the technical scope of the present invention.

[0142] For example, programs that enable a computer to function as an information processing device, as well as storage media such as flexible disks, magneto-optical disks, ROMs, CDs, DVDs, and BDs that can be read by the computer and on which such programs are recorded, are also provided. Here, a program refers to a data processing means written in any language or writing method.

[0143] Furthermore, the processes described herein do not necessarily have to be performed chronologically in the order shown in the flowchart; they may include parallel processing or processing using subroutines. [Explanation of symbols]

[0144] 20, 220, 320, 420 Information Processing Devices 54 Map data 56 Building Information 58 Machine Learning Models 62 Image Data Generation Unit 64 Numerical Data Generation Unit 66 Model Generation Unit 68 Estimation part 170 training data 172 Teacher's Image Data 174 Numerical data for teachers 176 Radio wave environment measurement data 182 Image data for estimation 184 Estimation numerical data 192 Radio wave environment estimation data 330 1st Network Design Department 430 Second Network Design Department

Claims

1. An image data generation unit determines a predetermined range on a map that includes an estimated location, which is the location where radio wave environment estimation is desired, as the estimation range, and generates estimation image data for each block within the estimation range, in which one or more types of building information relating to a building are associated as color information. An estimation unit estimates the radio wave environment at the estimated location by inputting the estimation image data, which includes the estimated location, into a trained machine learning model that outputs an estimation result of the radio wave environment in response to the input of the estimation image data. An information processing device equipped with the following features.

2. The information processing apparatus according to claim 1, wherein the image data generation unit determines the estimation range such that the transmission position of the radio waves subject to estimation of the radio wave environment and the estimated position, which is the reception position of the radio waves subject to estimation of the radio wave environment, are each positioned at a predetermined location within the figure forming the estimation range.

3. The system further comprises a model generation unit that generates the aforementioned machine learning model, The image data generation unit determines a predetermined area on a map including the locations of the installed first wireless communication terminal and the installed second wireless communication terminal as the training area, and generates training image data for each block within the training area, in which one or more types of building information relating to a building are associated as color information. The information processing apparatus according to claim 1, wherein the model generation unit performs machine learning using the training image data and data on the actual radio wave environment at the location of the second wireless communication terminal when the second wireless communication terminal receives radio waves transmitted from the first wireless communication terminal as training data to generate the machine learning model.

4. The system further includes a numerical data generation unit that generates estimated numerical data, which is numerical data correlated with the radio wave environment in the estimated range, based on the map data of the estimated range. The trained machine learning model outputs an estimated result of the radio wave environment in response to the input of the estimated image data and the estimated numerical data. The information processing apparatus according to claim 1, wherein the estimation unit estimates the radio wave environment at the estimated location by inputting the estimation image data including the estimated location and the estimation numerical data within the estimation range into the trained machine learning model.

5. The system further comprises a model generation unit that generates the aforementioned machine learning model, The image data generation unit determines a predetermined area on a map including the locations of the installed first wireless communication terminal and the installed second wireless communication terminal as the training area, and generates training image data for each block within the training area, in which one or more types of building information relating to a building are associated as color information. The numerical data generation unit generates training numerical data, which is numerical data correlated with the radio wave environment in the training area, based on the map data of the training area. The information processing apparatus according to claim 4, wherein the model generation unit performs machine learning using the training image data, the training numerical data, and data on the actual radio wave environment at the location of the second wireless communication terminal when the second wireless communication terminal receives radio waves transmitted from the first wireless communication terminal as training data to generate the machine learning model.

6. The estimation range includes the transmission location of the radio waves subject to estimation of the radio wave environment and the estimated location, which is the reception location of the radio waves subject to estimation of the radio wave environment. The information processing apparatus according to claim 4, wherein the estimation numerical data includes at least one of the following: the number of buildings included within the estimation range; the number of buildings within the estimation range that are in contact with the line segment connecting the transmission position and the estimation position; and the ratio of the length of the portion of the line segment in contact with a building to the length of the line segment.

7. The information processing device according to any one of claims 1 to 6, wherein the one or more types of building information include at least one of the building height, building area, and building type.

8. Furthermore, the first network design department is also included. In a designated area where network design is desired, multiple combinations of the transmission location of the radio waves subject to radio wave environment estimation and the estimated location, which is the reception location of the radio waves subject to radio wave environment estimation, are specified. The estimation unit, For each of the specified combinations, the radio wave environment at the estimated location is estimated using the trained machine learning model. For each of the specified combinations, it is determined whether or not communication can be established at the estimated location, according to the estimated result of the radio wave environment at the estimated location. The aforementioned first network design unit, From the specified combinations, extract the combinations that are determined to be able to establish communication, The information processing apparatus according to claim 1, which groups the estimated positions in the extracted combinations so that they belong to one network.

9. Furthermore, a second network design department is provided. The image data generation unit determines the estimation range to include the transmission locations of radio waves subject to estimation of the radio wave environment, which belong to the first network, and the estimated locations, which are the reception locations of radio waves subject to estimation of the radio wave environment, which belong to the second network, and generates the estimation image data based on the map data of the estimation range. The estimation unit inputs the estimation image data, including the estimated location, into the machine learning model to estimate the radio wave environment at the estimated location, and determines whether or not communication can be established at the estimated location according to the estimation result. The information processing apparatus according to claim 1, wherein the second network design unit determines that communication can be established at the estimated location, and combines the first network and the second network to form a single network.

Images