Weather estimation device

Abstract

A weather estimation device 10 comprises: an acquisition unit 112 that acquires first information relating to the quality of communication between a terminal 30 and a base station 40; and an estimation unit 115 that estimates, on the basis of the first information, weather information relating to the weather in a first region R[1] which includes the terminal 30 and the base station 40.

Description

Weather estimation apparatus

[0001] The present invention relates to a weather estimation apparatus.

[0002] Conventionally, there is a technique for predicting weather information based on observation data observed by a weather radar.

[0003] For example, Patent Document 1 discloses a weather prediction apparatus comprising: a weather radar that measures a rainfall and snowfall area; a movement vector calculation unit that calculates a movement vector of the rainfall and snowfall area based on parameters required for weather prediction; and a predicted image generation unit that generates a radar image at a future time using the movement vector and coefficients at an arbitrary time.

[0004] Japanese Unexamined Patent Publication No. 09-090057

[0005] However, in the measurement of atmospheric moisture over a wide area by weather radar, since measurement is performed by radiating radio waves near the 10 GHz band to the sky, there is a problem that measurement of distant conditions cannot be performed because the radio waves are blocked by nearby rain. On the other hand, point measurement using a small radar or Automated Meteorological Data Acquisition System has a problem that it is difficult to deploy to a wide area.

[0006] Accordingly, an object of the present invention is to provide a weather estimation apparatus using a wide-area atmospheric moisture measurement method capable of measuring distant conditions, as compared with the prior art.

[0007] A weather estimation apparatus according to a preferred aspect of the present invention comprises: an acquisition unit that acquires first information related to communication quality between a terminal and a base station; and an estimation unit that estimates weather information related to weather in a first area including the terminal and the base station based on the first information.

[0008] According to the present invention, it is possible to provide a weather estimation apparatus using a wide-area atmospheric moisture measurement method capable of measuring distant conditions, as compared with the prior art.

[0009] Block diagram showing an example of the overall configuration of weather estimation system 1. Block diagram showing an example of the overall configuration of weather estimation system 1. Block diagram showing an example of the configuration of machine learning device 20. Table showing an example of the first data DS1. Table showing an example of the second data DS2. Table showing an example of the first training data TD1. Table showing an example of the third data DS3. Table showing an example of the second training data TD2. Table showing an example of the fourth data DS4. Table showing an example of the third training data TD3. Table showing an example of the fifth data DS5. Table showing an example of the fourth training data TD4. Table showing an example of the sixth data DS6. Table showing an example of the seventh data DS7. Table showing an example of the fifth training data TD5. Block diagram showing an example of the configuration of weather estimation device 10. Table showing an example of the eighth data DS8. Table showing an example of the ninth data DS9. Table showing an example of the tenth data DS10. Table showing an example of the second input data ID2. Table showing an example of the third input data ID3. Table showing an example of the fourth input data ID4. Table showing an example of the fifth input data ID5. A flowchart showing an example of the operation of the weather estimation device 10.

[0010] 1: The first embodiment of the weather estimation system 1 will be described below with reference to Figures 1 to 24.

[0011] 1-1: Configuration of the First Embodiment 1-1-1: Overall Configuration Figures 1 and 2 are block diagrams showing an example of the overall configuration of the weather estimation system 1 according to this embodiment. As shown in Figure 1, the weather estimation system 1 comprises a weather estimation device 10, a machine learning device 20, and terminals 30[1] to 30[n]. The weather estimation device 10, the machine learning device 20, and terminals 30[1] to 30[n] are connected to each other via a communication network NET1 so that they can communicate with one another. n is an integer of 1 or more.

[0012] In the following explanation, terminals 30[1] to 30[n] may be collectively referred to as "terminal 30". In Figure 1, user U uses terminal 30. Also, user U[1] uses terminal 30[1]. User U[2] uses terminal 30[2]. User U[k] uses terminal 30[k]. User U[n] uses terminal 30[n]. k is an integer between 1 and n, inclusive. Furthermore, in the following explanation, user U[k] may be used as a representative example of user U, and terminal 30[k] may be used as a representative example of terminal 30.

[0013] Furthermore, as shown in Figure 2, the weather estimation system 1 includes, in addition to the weather estimation device 10, machine learning device 20, and terminals 30[1] to 30[n] described above, base stations 40[1] to 40[z]. The weather estimation device 10, the machine learning device 20, and base stations 40[1] to 40[z] are connected to each other via the communication network NET2 so that they can communicate with one another. z is an integer of 1 or more. Also, x is an integer between 1 and z, inclusive. Note that communication network NET1 and communication network NET2 may be the same communication network or may be different communication networks. In the following description, base stations 40[1] to 40[z] may be collectively referred to as "base station 40". Also, base station 40[x] may be used as a representative example of base station 40.

[0014] The weather estimation device 10 estimates weather information for a region R that includes one or more base stations 40, based on first information regarding the quality of communication obtained from base stations 40[1] to base stations 40[z].

[0015] The machine learning device 20 determines the learning model LM that the weather estimation device 10 will use to estimate weather information. The learning model LM determined by the machine learning device 20 is transmitted from the machine learning device 20 to the weather estimation device 10.

[0016] Terminal 30 is a communication terminal device used by user U. Terminal 30 is, for example, a mobile phone, a smartphone, or a tablet. Terminal 30 also receives weather information estimated by the weather estimation device 10 from the weather estimation device 10. Terminal 30 displays the weather information received from the weather estimation device 10 on a display device installed in Terminal 30.

[0017] Base station 40 is a land-based radio station. Base station 40 conducts wireless communication with terminal 30. Base station 40 is at the end of the telephone network and relays calls and communications between it and terminal 30. By relaying calls and communications between base station 40 and terminal 30, calls and communications become possible between terminal 30 itself. Base stations 40 are connected to each other by wired, wireless, or satellite link.

[0018] In the example shown in Figure 2, the base station 40[1] and terminals 30[1] to 30[p] are located within region R[1], where p is an integer greater than or equal to 1. Region R[1] is an example of a "first region". The base station 40[1] communicates wirelessly with each of terminals 30[1] to 30[p]. The base station 40[1] also transmits information regarding the quality of communication between each of terminals 30[1] to 30[p] to the weather estimation device 10 and the machine learning device 20 via the communication network NET2.

[0019] In the example shown in Figure 2, for the sake of simplicity, the illustration of the terminal 30 communicating with base station 40[2] in region R[2] is omitted. Similarly, the illustration of the terminal 30 communicating with base station 40[x] in region R[x] and the terminal 30 communicating with base station 40[z] in region R[z] is omitted.

[0020] Furthermore, in the example shown in Figure 2, there is one base station 40[1] and p terminals 30[1] to 30[p] within region R[1]. However, the number of base stations 40 present in region R[1] is any number greater than or equal to 1. Similarly, the number of terminals 30 present in region R[1] is any number greater than or equal to 1. The same applies to regions R[2] to R[z]. Specifically, the number of base stations 40 and the number of terminals 30 present in each region R[2] to R[z] are any number greater than or equal to 1.

[0021] In Figure 2, the frequency bands of the radio waves used for communication by base stations 40[1] to 40[z] may be the same or different. More specifically, between any two base stations 40 included in base stations 40[1] to 40[z], the frequency bands of the radio waves used for communication by each base station 40 may be the same or different.

[0022] In Figure 2, the weather estimation device 10 acquires information regarding the quality of communication between one or more terminals 30 and each of the base stations 40[1] to 40[z] from each of the base stations 40[1] to 40[z]. Similarly, the machine learning device 20 also acquires information regarding the quality of communication between one or more terminals 30 and each of the base stations 40[1] to 40[z] from each of the base stations 40[1] to 40[z], in the same manner as the weather estimation device 10.

[0023] More specifically, during machine learning, the machine learning device 20 acquires information regarding the quality of communication between each of the learning base stations 40[1] to 40[z] and one or more learning terminals 30 that communicate with each of the learning base stations 40[1] to 40[z]. Similarly, during weather estimation, the weather estimation device 10 acquires information regarding the quality of communication between each of the learning base stations 40[1] to 40[z] and one or more learning terminals 30 that communicate with each of the learning base stations 40[1] to 40[z]. The information acquired by the weather estimation device 10 is an example of "first information".

[0024] Furthermore, during machine learning, the machine learning device 20 acquires meteorological information for each of the regions R[1] to R[z]. One example of this meteorological information is the amount of precipitation in each of the regions R[1] to R[z]. Another example of this meteorological information is the amount of moisture in the atmosphere in each of the regions R[1] to R[z].

[0025] Other information acquired by the weather estimation device 10 and the machine learning device 20 will be described later with reference to Figures 4 to 15 and 17 to 23.

[0026] 1-1-2: Configuration of the machine learning device. Figure 3 is a block diagram showing an example configuration of the machine learning device 20. As shown in Figure 3, the machine learning device 20 comprises a processing unit 21, a storage device 22, an input device 23, and a communication device 24. Each element of the machine learning device 20 is interconnected by one or more buses for communicating information.

[0027] The processing unit 21 is a processor that controls the entire machine learning device 20. The processing unit 21 is configured using, for example, one or more chips. The processing unit 21 is also configured using a central processing unit (CPU) that includes, for example, interfaces with peripheral devices, arithmetic units, and registers. Some or all of the functions of the processing unit 21 may be implemented by hardware such as a DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), PLD (Programmable Logic Device), and FPGA (Field Programmable Gate Array). The processing unit 21 executes various processes in parallel or sequentially.

[0028] The storage device 22 is a recording medium that can be read from and written to by the processing device 21. The storage device 22 includes, for example, non-volatile memory and volatile memory. Non-volatile memory includes, for example, ROM (Read Only Memory), EPROM (Erasable Programmable Read Only Memory), and EEPROM (Electrically Erasable Programmable Read Only Memory). Volatile memory is, for example, RAM. The storage device 22 stores the control program PR2 executed by the processing device 21. The storage device 22 functions as a work area for the processing device 21.

[0029] Furthermore, the memory device 22 stores the first training data TD1 to the fifth training data TD5.

[0030] The first training data TD1 is the training data TD used by the first model determination unit 214, described later, in machine learning to determine the first learning model LM1 used by the weather estimation device 10. The first learning model LM1 is a learning model LM that has been trained by machine learning to understand the relationship between information on the quality of communication between the learning terminal 30 and the learning base station 40 and information on the weather in region R, which includes the learning terminal 30 and the learning base station 40. The first training data TD1 is the training data TD used in the machine learning, and is data consisting of a pair of information on the quality of communication between the learning terminal 30 and the learning base station 40 and information on the weather in region R, which includes the learning base station 40 and the learning terminal 30.

[0031] Figure 4 is a table showing an example of the first data DS1, which is part of the first training data TD1 and includes information regarding the quality of communication between the learning terminal 30 and the learning base station 40. The first data DS1 has multiple first data blocks DB1. Each first data block DB1 has the following items: "Base Station ID", "Terminal ID", "Signal Strength", and "Date and Time". "Base Station ID" is the identifier of the base station 40. "Terminal ID" is the identifier of the terminal 30. "Signal Strength" is the signal strength of the radio waves received by the base station 40, indicated by the "Base Station ID", when the radio waves transmitted from the terminal 30, indicated by the "Terminal ID", are received by the base station 40, indicated by the "Base Station ID". In the example shown in Figure 4, this "Signal Strength" corresponds to the "Information Regarding Communication Quality" mentioned above. "Date and Time" is the date and time when the numerical value of the signal strength, indicated by the "Signal Strength" item, was measured.

[0032] As an example, the first data DS1 shown in Figure 4 stores the first data block DB1, which indicates that the "base station ID" is "A0", the "terminal ID" is "090XXX", the "signal strength" is "-91 dBm", and the "date and time" when the "signal strength" was measured is "October 1, 2024, 12:00:00".

[0033] Furthermore, in the "Base Station ID" field of the first data DS1 shown in Figure 4, an ID starting with "A" indicates that a base station 40 having that "Base Station ID" is located within area R[A]. Also, in the "Base Station ID" field, an ID starting with "B" indicates that a base station 40 having that "Base Station ID" is located within area R[B]. In other words, in the first data DS1 illustrated in Figure 4, it is shown that within area R[A], there is a base station 40 with "Base Station ID" "A0" and a base station 40 with "Base Station ID" "A1". Also, it is shown that within area R[B], there is a base station 40 with "Base Station ID" "B0". Furthermore, in the first data DS1 illustrated in Figure 4, the terminal 30 with "Terminal ID" "090XXX" communicates with both the base station 40 with "Base Station ID" "A0" and the base station 40 with "Base Station ID" "A1". Furthermore, terminal 30 with "Terminal ID" "080YYY" communicates with base station 40 with "Base Station ID" "A1". Also, terminal 30 with "Terminal ID" "060ZZZ" communicates with base station 40 with "Base Station ID" "B0". In other words, base station 40 with "Base Station ID" "A1" communicates with both terminal 30 with "Terminal ID" "090XXX" and terminal 30 with "Terminal ID" "080YYY".

[0034] Figure 5 is a table showing an example of the second data DS2, which is part of the first training data TD1 and contains weather information for region R, including the training terminal 30 and the training base station 40. The second data DS2 has multiple second data block DB2. Each second data block DB2 has the items "Base Station ID", "Precipitation", and "Date and Time". "Base Station ID" is the identifier of base station 40. "Precipitation" is the precipitation for region R, which includes base station 40 indicated by the "Base Station ID". In Figure 5, this "Precipitation" corresponds to the "Weather Information" mentioned above. "Date and Time" is the date and time when the precipitation value indicated by the "Precipitation" item was measured.

[0035] As an example, the second data DS2 shown in Figure 5 stores a second data block DB2 that indicates the "base station ID" is "A0", the "precipitation amount" is "15 mm", and the "date and time" when the "precipitation amount" was measured is "October 1, 2024, 12:00:00".

[0036] Furthermore, since the precipitation in region R[A] is "15 mm", in the second data DS2 shown in Figure 5, the "precipitation" value in the second data block DB2 with "base station ID" "A0" corresponding to the base station 40 included in region R[A] and the "precipitation" value in the second data block DB2 with "base station ID" "A1" are both "15 mm". On the other hand, since the precipitation in region R[B] is "0 mm", in the second data DS2 shown in Figure 5, the "precipitation" value in the second data block DB2 with "base station ID" "B0" corresponding to the base station 40 included in region R[B] is "0 mm".

[0037] The data determination unit 213, described later, determines the first training data TD1 by combining the first data DS1 and the second data DS2.

[0038] Figure 6 is a table showing an example of the first training data TD1. More specifically, the data determination unit 213 determines the first training data TD1 by merging the first data DS1 and the second data DS2 based on the items common to both the first data DS1 and the second data DS2, namely "base station ID" and "date and time". If the value of the "date and time" item in the first data DS1 and the value of the "date and time" item in the second data DS2 do not match, even though they correspond to the same "base station ID", the data determination unit 213 considers them to be the same if the difference between the two is less than or equal to a predetermined value, and merges the first data DS1 and the second data DS2.

[0039] For the sake of explanation, the first training data TD1 shown in Figure 6 includes a "Date and Time" field, but this field is not mandatory.

[0040] In Figure 3, the second training data TD2 is the training data TD used by the second model determination unit 215, described later, in machine learning to determine the second learning model LM2 used by the weather estimation device 10. The second learning model LM2 is a learning model LM that has been trained by machine learning to understand the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information specifying the frequency band of radio waves used for communication between the learning terminal 30 and the learning base station 40, and weather information in region R including the learning terminal 30 and the learning base station 40. The second training data TD2 is the training data TD used in this machine learning. The second training data TD2 is data consisting of a set of information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information specifying the frequency band of radio waves used for communication between the learning terminal 30 and the learning base station 40, and weather information in region R including the learning terminal 30 and the learning base station 40.

[0041] Figure 7 is a table showing an example of third data DS3, which is part of the first training data TD1 and contains information specifying the frequency band of radio waves used for communication between the learning terminal 30 and the learning base station 40. The third data DS3 has multiple third data blocks DB3. Each third data block DB3 has the items "Base Station ID" and "Frequency Band". The "Base Station ID" is the identifier of the base station 40. The "Frequency Band" is the frequency band of radio waves used by the base station 40, indicated by the "Base Station ID", for communication with the terminal 30.

[0042] As an example, the third data DS3 shown in Figure 7 stores a third data block DB3 that indicates that the "base station ID" is "A0" and the "frequency band" is "30 GHz".

[0043] The data determination unit 213, described later, determines the second training data TD2 by combining the first training data DS1, the second training data DS2, and the third training data DS3. In other words, the data determination unit 213 determines the second training data TD2 by combining the first training data TD1 and the third training data DS3.

[0044] FIG. 8 is a table showing an example of second teacher data TD2. More specifically, the data determination unit 213 merges the first teacher data TD1 and the third data DS3 based on "base station ID", which is an item common to the first teacher data TD1 and the third data DS3, thereby determining the second teacher data TD2.

[0045] For convenience of explanation, the second teacher data TD2 shown in FIG. 8 includes an item of "date and time", but the "date and time" item is not essential.

[0046] In FIG. 8, the information of the "frequency band" item is an example of "information specifying a frequency band of radio waves used for communication between the learning terminal 30 and the learning base station 40". In the overall configuration shown in FIG. 2, when each of base stations 40[1] to 40[z] uses only one frequency band for radio waves used for communication, the "base station ID" in the first teacher data TD1 shown in FIG. 6 is also an example of "information specifying a frequency band of radio waves used for communication between the learning terminal 30 and the learning base station 40".

[0047] In FIG. 3, third teacher data TD3 is teacher data TD used by a third model determination unit 216 described later in machine learning for determining a third learning model LM3 used by the weather estimation device 10. The third learning model LM3 is a learned learning model LM obtained by machine learning of the relationship among: information on communication quality between the learning terminal 30 and the learning base station 40; information on an environment where the learning base station 40 is installed; and information on weather in a region R including the learning terminal 30 and the learning base station 40. The third teacher data TD3 is teacher data TD used for said machine learning. The third teacher data TD3 is data in which information on communication quality between the learning terminal 30 and the learning base station 40, information on an environment where the learning base station 40 is installed, and information on weather in the region R including the learning terminal 30 and the learning base station 40 are combined as a set.

[0048] FIG. 9 is a table showing an example of fourth data DS4 including information related to an environment where a learning base station 40 is installed, among third teacher data TD3. The fourth data DS4 includes a plurality of fourth data blocks DB4. Each fourth data block DB4 has items of "base station ID" and "environment". The "base station ID" is an identifier of the base station 40. The "environment" is information indicating whether the base station 40 indicated by the "base station ID" is installed indoors or outdoors.

[0049] As an example, the fourth data DS4 shown in FIG. 9 stores a fourth data block DB4 indicating that the "base station ID" is "A0" and the "environment" is "outdoor".

[0050] A data determination unit 213, which will be described later, determines the third teacher data TD3 by combining first data DS1, second data DS2, and the fourth data DS4. In other words, the data determination unit 213 determines the third teacher data TD3 by combining first teacher data TD1 and the fourth data DS4.

[0051] FIG. 10 is a table showing an example of the third teacher data TD3. More specifically, the data determination unit 213 determines the third teacher data TD3 by merging the first teacher data TD1 and the fourth data DS4 based on the "base station ID", which is an item common to the first teacher data TD1 and the fourth data DS4.

[0052] For convenience of description, although the third teacher data TD3 shown in FIG. 10 has an item of "date and time", the item of "date and time" is not essential.

[0053] In FIG. 10, the information of the "environment" item is an example of "information related to the environment where the learning base station 40 is installed".

[0054] In Figure 3, the fourth training data TD4 is the training data TD used by the fourth model determination unit 217, described later, in machine learning to determine the fourth learning model LM4 used by the weather estimation device 10. The fourth learning model LM4 is a learning model LM that has been trained by machine learning to understand the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information regarding the model of the learning terminal 30, and information regarding the weather in region R including the learning terminal 30 and the learning base station 40. The fourth training data TD4 is the training data TD used in this machine learning. The fourth training data TD4 is data consisting of a set of information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information regarding the model of the learning terminal 30, and information regarding the weather in region R including the learning terminal 30 and the learning base station 40.

[0055] Figure 11 is a table showing an example of the fifth data DS5, which is part of the fourth training data TD4 and contains information about the model of the learning terminal 30. The fifth data DS5 has multiple fifth data blocks DB5. Each fifth data block DB5 has the fields "Terminal ID" and "Model". The "Terminal ID" is the identifier of the terminal 30. The "Model" is information indicating the model of the terminal 30 indicated by the "Terminal ID".

[0056] As an example, the fifth data DS5 shown in Figure 11 stores the fifth data block DB5, which indicates that the "terminal ID" is "090XXX" and the "model" is "Android". Note that in Figure 11, "Android" and "iOS" are both registered trademarks.

[0057] The data determination unit 213, described later, determines the fourth training data TD4 by combining the first training data DS1, the second training data DS2, and the fifth training data DS5. In other words, the data determination unit 213 determines the fourth training data TD4 by combining the first training data TD1 and the fifth training data DS5.

[0058] Figure 12 is a table showing an example of the fourth training data TD4. More specifically, the data determination unit 213 determines the fourth training data TD4 by merging the first training data TD1 and the fifth data DS5 based on the "terminal ID," which is a common item in both the first training data TD1 and the fifth data DS5.

[0059] For the sake of explanation, the fourth training data TD4 shown in Figure 12 includes a "Date and Time" field, but this field is not mandatory.

[0060] In Figure 3, the fifth training data TD5 is the training data TD used by the fifth model determination unit 218, described later, in machine learning to determine the fifth learning model LM5 used by the weather estimation device 10. The fifth learning model LM5 is a learning model LM that has been trained by machine learning to understand the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, the location information of the learning terminal 30, the location information of the learning base station 40, and weather information in region R including the learning terminal 30 and the learning base station 40. The fifth training data TD5 is the training data TD used in this machine learning. The fifth training data TD5 is data consisting of a set of information regarding the quality of communication between the learning terminal 30 and the learning base station 40, the location information of the learning terminal 30, the location information of the learning base station 40, and weather information in region R including the learning terminal 30 and the learning base station 40.

[0061] Figure 13 is a table showing an example of the sixth data DS6, which includes location information of the training terminal 30, from the fifth training data TD5. The sixth data DS6 has multiple sixth data blocks DB6. Each sixth data block DB6 has the following items: "Terminal ID", "Terminal Latitude", "Terminal Longitude", and "Date and Time". "Terminal ID" is the identifier of terminal 30. "Terminal Latitude" is information indicating the latitude where terminal 30, indicated by the "Terminal ID", is located. "Terminal Longitude" is information indicating the longitude where terminal 30, indicated by the "Terminal ID", is located. "Date and Time" is the date and time when the location information indicated by the "Terminal Latitude" and "Terminal Longitude" items was measured.

[0062] As an example, the sixth data DS6 shown in Figure 13 stores the sixth data block DB6, which indicates that the "terminal ID" is "090XXX", the "terminal latitude" is "35.692378", the "terminal longitude" is "139.704726", and the "date and time" when these "terminal latitude" and "terminal longitude" were measured is "October 1, 2024, 12:00:00".

[0063] Figure 14 is a table showing an example of the seventh data DS7, which includes location information of the training base station 40, from the fifth training data TD5. The seventh data DS7 has multiple seventh data block DB7. Each seventh data block DB7 has the items "Base Station ID", "Base Station Latitude", and "Base Station Longitude". "Base Station ID" is the identifier of the base station 40. "Base Station Latitude" is information indicating the latitude where the base station 40, indicated by the "Base Station ID", is located. "Base Station Longitude" is information indicating the longitude where the base station 40, indicated by the "Base Station ID", is located.

[0064] As an example, the seventh data DS7 shown in Figure 14 stores the seventh data block DB7, which indicates that the "base station ID" is "A0", the "base station latitude" is "35.692354", and the "base station longitude" is "139.704797".

[0065] The data determination unit 213, described later, determines the fifth training data TD5 by combining the first training data DS1, the second training data DS2, the sixth training data DS6, and the seventh training data DS7. In other words, the data determination unit 213 determines the fifth training data TD5 by combining the first training data TD1, the sixth training data DS6, and the seventh training data DS7.

[0066] Figure 15 is a table showing an example of the fifth training data TD5. More specifically, the data determination unit 213 merges the first training data TD1 and the sixth data DS6 based on the items common to both the first training data TD1 and the sixth data DS6, namely "Terminal ID" and "Date and Time". Furthermore, the data determination unit 213 determines the fifth training data TD5 by merging the first training data TD1 and the seventh data DS7 based on the item common to both the first training data TD1 and the seventh data DS7, namely "Base Station ID". Note that if the value of the "Date and Time" item in the first training data TD1 and the value of the "Date and Time" item in the sixth data DS6 correspond to the same "Terminal ID" but do not match, the data determination unit 213 considers them to be the same if the difference between the two is less than or equal to a predetermined value, and merges the first training data TD1 and the sixth data DS6.

[0067] For the sake of explanation, the fifth training data TD5 shown in Figure 15 includes a "Date and Time" field, but this field is not mandatory.

[0068] In Figure 3, the input device 23 is a device that receives operations from the administrator of the machine learning device 20. For example, the input device 23 is configured to include a keyboard, touchpad, touch panel, or pointing device such as a mouse. Here, if the input device 23 is configured to include a touch panel, it may also function as a display device.

[0069] The communication device 24 is hardware that acts as a transmitting and receiving device for communicating with other devices. The communication device 24 is also called, for example, a network device, a network controller, a network card, or a communication module. The communication device 24 may be equipped with a connector for wired connection and an interface circuit corresponding to the connector. The communication device 24 may also be equipped with a wireless communication interface. Examples of connectors and interface circuits for wired connection include products compliant with wired LAN, IEEE 1394, and USB. Examples of wireless communication interfaces include products compliant with wireless LAN and Bluetooth®.

[0070] The processing unit 21 functions as a communication control unit 211, an acquisition unit 212, a data determination unit 213, a first model determination unit 214, a second model determination unit 215, a third model determination unit 216, a fourth model determination unit 217, and a fifth model determination unit 218, for example, by reading and executing the control program PR2 from the storage device 22.

[0071] The communication control unit 211 causes the communication device 24 to send and receive various information, data, and signals with other devices.

[0072] The acquisition unit 212 acquires information regarding the quality of communication between the learning terminal 30 and the learning base station 40 from the learning base station 40. This information corresponds to the first data DS1 illustrated in Figure 4.

[0073] Furthermore, the acquisition unit 212 acquires information from the learning base station 40 that specifies the frequency band of the radio waves used for communication between the learning terminal 30 and the learning base station 40. This information corresponds to the third data DS3 illustrated in Figure 7.

[0074] Furthermore, the acquisition unit 212 acquires information from the learning base station 40 regarding the environment in which the learning base station 40 is installed. This information corresponds to the fourth data DS4 illustrated in Figure 9.

[0075] Furthermore, the acquisition unit 212 acquires information from the learning base station 40 regarding the model of the terminal 30 that communicates with the learning base station 40. This information corresponds to the fifth data DS5 illustrated in Figure 11.

[0076] Furthermore, the acquisition unit 212 acquires location information of the terminal 30 that communicates with the learning base station 40, and location information of the learning base station 40, from the learning base station 40. The location information of the terminal 30 that communicates with the learning base station 40 corresponds to the sixth data DS6 illustrated in Figure 13. The location information of the learning base station 40 corresponds to the seventh data DS7 illustrated in Figure 14.

[0077] As an example, the learning base station 40 obtains the location information of the learning terminal 30 from an application that uses location information and is installed on the terminal 30.

[0078] Furthermore, the acquisition unit 212 acquires weather information for the region R, which includes the learning terminal 30 and the learning base station 40, from an external device. This information corresponds to the second data DS2 illustrated in Figure 5.

[0079] As described above, the data determination unit 213 uses the various information acquired by the acquisition unit 212 to determine the first training data TD1 to the fifth training data TD5 to be stored in the storage device 22.

[0080] The first model determination unit 214 determines the first learned model LM1 by performing machine learning using the first training data TD1.

[0081] The second model determination unit 215 determines the second learning model LM2 by performing machine learning using the second training data TD2.

[0082] The third model determination unit 216 determines the third learning model LM3 by performing machine learning using the third training data TD3.

[0083] Furthermore, it is preferable for the third model determination unit 216 to determine the third learning model LM3 by weighting the data contained in the third training data TD3 with information regarding the communication quality with the outdoor base station 40.

[0084] The fourth model determination unit 217 determines the fourth learning model LM4 by performing machine learning using the fourth training data TD4.

[0085] Furthermore, it is preferable for the fourth model determination unit 217 to determine the fourth learning model LM4 by weighting information regarding the communication quality between a terminal 30 of a specific model and the base station 40.

[0086] The fifth model determination unit 218 determines the fifth learning model LM5 by performing machine learning using the fifth training data TD5.

[0087] 1-1-3: Diagram 16 of the weather estimation device configuration is a block diagram showing an example configuration of the weather estimation device 10. As shown in Figure 16, the weather estimation device 10 comprises a processing unit 11, a storage device 12, an input device 13, and a communication device 14. Each element of the weather estimation device 10 is interconnected by one or more buses for communicating information.

[0088] The processing unit 11 is a processor that controls the entire weather estimation device 10. The processing unit 11 is configured using, for example, one or more chips. The processing unit 11 is also configured using, for example, a central processing unit (CPU) that includes interfaces with peripheral devices, an arithmetic unit, and registers. Some or all of the functions of the processing unit 11 may be implemented by hardware such as a DSP, ASIC, PLD, and FPGA. The processing unit 21 executes various processes in parallel or sequentially.

[0089] The storage device 12 is a recording medium that can be read from and written to by the processing device 11. The storage device 12 includes, for example, non-volatile memory and volatile memory. The non-volatile memory is, for example, ROM, EPROM, and EEPROM. The volatile memory is, for example, RAM. The storage device 12 stores the control program PR1 that the processing device 11 executes. The storage device 12 functions as a work area for the processing device 11.

[0090] Furthermore, the memory device 12 stores the first learning model LM1 to the fifth learning model LM5. The first learning model LM1 to the fifth learning model LM5 are the first learning model LM1 to the fifth learning model LM5 acquired from the machine learning device 20 by the acquisition unit 112, which will be described later.

[0091] The input device 13 is a device that receives operations from the administrator of the weather estimation device 10. For example, the input device 13 is configured to include a keyboard, touchpad, touch panel, or pointing device such as a mouse. If the input device 13 is configured to include a touch panel, it may also function as a display device.

[0092] The communication device 14 is hardware that acts as a transmitting and receiving device for communicating with other devices. The communication device 14 is also called, for example, a network device, a network controller, a network card, or a communication module. The communication device 14 may be equipped with a connector for wired connection and an interface circuit corresponding to the connector. The communication device 14 may also be equipped with a wireless communication interface. Examples of connectors and interface circuits for wired connection include products compliant with wired LAN, IEEE 1394, and USB. Examples of wireless communication interfaces include products compliant with wireless LAN and Bluetooth®.

[0093] The processing unit 11 functions as a communication control unit 111, an acquisition unit 112, a reception unit 113, a data determination unit 114, an estimation unit 115, and a display control unit 116, for example, by reading and executing the control program PR1 from the storage device 12.

[0094] The communication control unit 111 causes the communication device 14 to send and receive various information, data, and signals with other devices.

[0095] The acquisition unit 112 acquires the first learning model LM1 to the fifth learning model LM5 from the machine learning device 20. The acquisition unit 112 stores the acquired first learning model LM1 to the fifth learning model LM5 in the storage device 12.

[0096] Furthermore, the acquisition unit 112 acquires information from the base station 40 regarding the quality of communication between the terminal 30 communicating with the base station 40 and the base station 40. This information is an example of "first information".

[0097] Figure 17 is a table showing an example of an eighth data DS8 containing the first information. The eighth data DS8 has multiple eighth data block DB8s. Each eighth data block DB8 has the items "Base Station ID", "Terminal ID", "Signal Strength", and "Date and Time". "Base Station ID" is the identifier of base station 40. "Terminal ID" is the identifier of terminal 30. "Signal Strength" is the signal strength of the radio waves received when base station 40, indicated by the "Base Station ID", receives radio waves transmitted from terminal 30, indicated by the "Terminal ID". In the example shown in Figure 17, this "Signal Strength" corresponds to the "Information Regarding Communication Quality" and "First Information" mentioned above. "Date and Time" is the date and time when the numerical value of the signal strength indicated by the "Signal Strength" item was measured.

[0098] As an example, the eighth data DS8 shown in Figure 17 stores the eighth data block DB8, which indicates that the "base station ID" is "A0", the "terminal ID" is "090AAA", the "signal strength" is "-101 dBm", and the "date and time" when the "signal strength" was measured is "October 1, 2024, 13:00:00".

[0099] Furthermore, in the "Base Station ID" field of the eighth data DS8 shown in Figure 17, an ID starting with "A" indicates that a base station 40 having that "Base Station ID" is located within area R[A]. Also, in the "Base Station ID" field, an ID starting with "B" indicates that a base station 40 having that "Base Station ID" is located within area R[B]. In other words, in the eighth data DS8 exemplified in Figure 17, it is shown that within area R[A], there is a base station 40 with "Base Station ID" "A0" and a base station 40 with "Base Station ID" "A1". Also, in the eighth data DS8 exemplified in Figure 17, it is shown that within area R[B], there is a base station 40 with "Base Station ID" "B0". Furthermore, in the eighth data DS8 illustrated in Figure 17, terminal 30 with "Terminal ID" "090AAA" communicates with both base station 40 with "Base Station ID" "A0" and base station 40 with "Base Station ID" "A1". Also, terminal 30 with "Terminal ID" "080BBB" communicates with base station 40 with "Base Station ID" "A1". Also, terminal 30 with "Terminal ID" "060CCC" communicates with base station 40 with "Base Station ID" "B0". In other words, base station 40 with "Base Station ID" "A1" communicates with both terminal 30 with "Terminal ID" "090AAA" and terminal 30 with "Terminal ID" "080BBB".

[0100] Furthermore, in Figure 16, the acquisition unit 112 acquires information from the base station 40 that specifies the frequency band of radio waves that the base station 40 uses for communication with the terminal 30. This information is an example of "second information".

[0101] Referring to Figure 7, the third data DS3 described above includes the second information.

[0102] Furthermore, in Figure 16, the acquisition unit 112 acquires information from the base station 40 regarding the environment in which the base station 40 is installed. This information is an example of "third information".

[0103] Referring to Figure 9, the fourth data DS4 described above includes the third information.

[0104] Furthermore, in Figure 16, the acquisition unit 112 acquires information from the base station 40 regarding the model of the terminal 30 that communicates with the base station 40. This information is an example of "fourth information".

[0105] Figure 18 is a table showing an example of the ninth data DS9 containing the fourth information. The ninth data DS9 has multiple ninth data block DB9. Each ninth data block DB9 has the fields "Terminal ID" and "Model". The "Terminal ID" is the identifier of terminal 30. The "Model" is information indicating the model of terminal 30 indicated by the "Terminal ID".

[0106] As an example, the ninth data DS9 shown in Figure 18 stores the ninth data block DB9, which indicates that the "terminal ID" is "090AAA" and the "model" is "Android". Note that in Figure 18, both "Android" and "iOS" are registered trademarks.

[0107] Furthermore, in Figure 16, the acquisition unit 112 acquires location information of the model of the terminal 30 that communicates with the base station 40, as well as location information of the base station 40, from the base station 40.

[0108] For example, the base station 40 obtains the location information of the terminal 30 from an application that uses location information and is installed on the terminal 30.

[0109] Figure 19 is a table showing an example of a 10th data DS10 containing location information of terminal 30. The 10th data DS10 has a plurality of 10th data block DB10. Each 10th data block DB10 has the items "Terminal ID", "Terminal Latitude", "Terminal Longitude", and "Date and Time". "Terminal ID" is the identifier of terminal 30. "Terminal Latitude" is information indicating the latitude where terminal 30, indicated by the "Terminal ID", is located. "Terminal Longitude" is information indicating the longitude where terminal 30, indicated by the "Terminal ID", is located. "Date and Time" is the date and time when the location information indicated by the "Terminal Latitude" and "Terminal Longitude" items was measured.

[0110] As an example, the tenth data DS10 shown in Figure 19 stores a tenth data block DB10 that indicates the "terminal ID" is "090AAA", the "terminal latitude" is "35.692415", the "terminal longitude" is "139.704831", and the "date and time" when these "terminal latitude" and "terminal longitude" were measured is "October 1, 2024, 13:00:00".

[0111] Referring to Figure 14, the seventh data DS7 described above includes location information of the base station 40.

[0112] In Figure 16, the reception unit 113 receives a learning model LM from the input device 13 or terminal 30[k] to be used by the estimation unit 115 (described later) when estimating weather information. Specifically, the reception unit 113 receives the result of selecting one learning model LM from among the first learning model LM1 to the fifth learning model LM5 stored in the storage device 12.

[0113] The data determination unit 114 determines the input data ID to be input to the learning model LM.

[0114] If the learning model LM received by the reception unit 113 is the first learning model LM1, the data determination unit 114 determines the first input data ID1. The first input data ID1 is the same data as the eighth data DS8 illustrated in Figure 17.

[0115] For the sake of explanation, the first input data ID1 shown in Figure 17 includes a "Date and Time" field, but the "Date and Time" field is not mandatory.

[0116] If the learning model LM received by the reception unit 113 is the second learning model LM2, the data determination unit 114 determines the second input data ID2.

[0117] The data determination unit 114 determines the second input data ID2 by combining the eighth data DS8 and the third data DS3. In other words, the data determination unit 114 determines the second input data ID2 by combining the first input data ID1 and the third data DS3.

[0118] Figure 20 is a table showing an example of the second input data ID 2. More specifically, the data determination unit 213 determines the second input data ID 2 by merging the first input data ID 1 and the third data DS3 based on the "base station ID," which is an item common to both the first input data ID 1 and the third data DS3.

[0119] For the sake of explanation, the second input data ID2 shown in Figure 19 has a "Date and Time" field, but the "Date and Time" field is not mandatory.

[0120] In Figure 20, the information in the "Frequency Band" field is an example of "Second Information." Furthermore, in the overall configuration shown in Figure 2, if each of the base stations 40[1] to 40[z] uses only one radio frequency band for communication, then the "Base Station ID" in the first input data ID1 shown in Figure 17 is also an example of "Second Information."

[0121] In Figure 16, if the learning model LM received by the reception unit 113 is the third learning model LM3, the data determination unit 114 determines the third input data ID3.

[0122] The data determination unit 114 determines the third input data ID 3 by combining the eighth data DS8 and the fourth data DS4. In other words, the data determination unit 114 determines the third input data ID 3 by combining the first input data ID 1 and the fourth data DS4.

[0123] Figure 21 is a table showing an example of the third input data ID 3. More specifically, the data determination unit 213 determines the third input data ID 3 by merging the first input data ID 1 and the fourth data DS4 based on the "base station ID," which is an item common to both the first input data ID 1 and the fourth data DS4.

[0124] For the sake of explanation, the third input data ID3 shown in Figure 21 includes a "Date and Time" field, but the "Date and Time" field is not mandatory.

[0125] In Figure 16, if the learning model LM received by the reception unit 113 is the fourth learning model LM4, the data determination unit 114 determines the fourth input data ID4.

[0126] The data determination unit 114 determines the fourth input data ID 4 by combining the eighth data DS8 and the ninth data DS9. In other words, the data determination unit 114 determines the fourth input data ID 4 by combining the first input data ID 1 and the ninth data DS9.

[0127] Figure 22 is a table showing an example of the fourth input data ID 4. More specifically, the data determination unit 114 determines the fourth input data ID 4 by merging the first input data ID 1 and the ninth data DS9 based on the "terminal ID," which is an item common to both the first input data ID 1 and the ninth data DS9.

[0128] For the sake of explanation, the fourth input data ID 4 shown in Figure 22 has a "Date and Time" field, but the "Date and Time" field is not mandatory.

[0129] In Figure 16, if the learning model LM received by the reception unit 113 is the fifth learning model LM5, the data determination unit 114 determines the fifth input data ID5.

[0130] The data determination unit 114 determines the fifth input data ID 5 by combining the eighth data DS8, the tenth data DS10, and the seventh data DS7. In other words, the data determination unit 114 determines the fifth input data ID 5 by combining the first input data ID 1, the tenth data DS10, and the seventh data DS7.

[0131] Figure 23 is a table showing an example of the fifth input data ID 5. More specifically, the data determination unit 114 merges the first input data ID 1 and the tenth data DS 10 based on the items common to both the first input data ID 1 and the tenth data DS 10, namely "Terminal ID" and "Date and Time". Furthermore, the data determination unit 114 determines the fifth input data ID 5 by merging the first input data ID 1 and the seventh data DS 7 based on "Base Station ID". Note that if the value of the "Date and Time" item in the first input data ID 1 and the value of the "Date and Time" item in the tenth data DS 10 do not match, even though they correspond to the same "Terminal ID", the data determination unit 114 considers them to be the same if the difference between the two is less than or equal to a predetermined value, and merges the first input data ID 1 and the tenth data DS 10.

[0132] For the sake of explanation, the fifth input data ID 5 shown in Figure 23 has a "Date and Time" field, but the "Date and Time" field is not mandatory.

[0133] In Figure 16, the estimation unit 115 estimates weather information using one of the first learning model LM1 to the fifth learning model LM5.

[0134] More specifically, if the selection result received by the reception unit 113 is the first learning model LM1, the estimation unit 115 inputs the first input data ID1 to the first learning model LM1, and then uses the output result from the first learning model LM1 as the weather information estimation result.

[0135] In other words, the estimation unit 115 estimates meteorological information regarding the weather in region R by inputting the above-mentioned first information into a first learning model LM1 that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40 and information regarding the weather in region R including the learning terminal 30 and the learning base station 40. Region R is an example of the "first region".

[0136] As a result, the weather estimation device 10 can estimate weather information using a wide-area atmospheric moisture measurement method that allows for measurement of conditions at a distance, compared to conventional technology.

[0137] Specifically, the weather estimation device 10 constructs a first learning model LM1 that estimates precipitation based on radio wave intensity information between base stations 40 and terminals 30, which is collected in real time from all over Japan, and then becomes capable of estimating precipitation in real time on a nationwide scale.

[0138] More specifically, if, for example, rainfall of 10 mm / h occurs in a certain region R, the machine learning device 20 constructs a first learning model LM1 using training data TD, which is an event in which the radio wave intensity used for communication with one terminal 30 in the 28 GHz band of a base station 40 located in that region R decreases by 10 dB compared to normal. The weather estimation system 1, which aggregates information on communication quality, can obtain values ​​for the amount of moisture in the atmosphere in real time across the entire area where the mobile wireless network is installed.

[0139] For example, comparing the past radio wave strength included in the first training data TD1 illustrated in Figure 6 with the current radio wave strength included in the first input data ID1 illustrated in Figure 17, in region R[A] which includes base station 40 with base station ID "A0" and base station 40 with base station ID "A1", the current radio wave strength has uniformly decreased compared to the past radio wave strength one hour ago. The radio wave strength used for communication by two different terminals 30 communicating with base station 40 with base station ID "A1", specifically terminal 30 with terminal ID "090AAA" and terminal 30 with terminal ID "080BBB", has both decreased. Furthermore, the signal strength of the radio waves used for communication with terminal 30, terminal ID 090AAA, is decreasing at both base stations 40, specifically base station 40 with base station ID A0 and base station 40 with base station ID A1. These phenomena enable the weather estimation device 10 to observe changes in atmospheric conditions over a wide area.

[0140] Furthermore, by combining the above-mentioned weather information estimation method with existing methods using weather radar, the weather estimation system 1 can estimate rainfall conditions with greater accuracy. On the other hand, by using the weather information estimation results from the weather estimation system 1, businesses can optimize the wireless access area in their mobile phone networks, taking weather conditions into account. This optimization includes, as an example, controlling frequency selection while taking rainfall conditions into account.

[0141] Furthermore, if the selection result received by the reception unit 113 is the second learning model LM2, the estimation unit 115 inputs the second input data ID2 to the second learning model LM2, and then uses the output result from the second learning model LM2 as the weather information estimation result.

[0142] In other words, the estimation unit 115 estimates weather information by inputting the above-mentioned first information and the above-mentioned second information into a second learning model LM2 that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information specifying the frequency band of radio waves used for communication between the learning terminal 30 and the learning base station 40, and weather information in the region R including the learning terminal 30 and the learning base station 40.

[0143] Multiple frequency bands may be used during communication between terminal 30 and base station 40. In the high-frequency bands of several tens of GHz or higher used in 5G, the degree of attenuation due to moisture is greater compared to lower-frequency bands. The estimation unit 115 can improve the accuracy of weather information estimation based on changes in radio wave intensity in multiple frequency bands by using the second learning model LM2. For example, the machine learning device 20 determines the second learning model LM2 based on the fact that if attenuation occurs only on the high-frequency side and the degree of attenuation is large, the attenuation is caused by an obstacle, while if attenuation occurs only on the high-frequency side and the degree of attenuation is small, the attenuation is caused by moisture in the atmosphere.

[0144] Furthermore, if the selection result received by the reception unit 113 is the third learning model LM3, the estimation unit 115 inputs the third input data ID3 to the third learning model LM3, and then uses the output result from the third learning model LM3 as the meteorological information estimation result.

[0145] In other words, the estimation unit 115 estimates weather information by inputting the above-mentioned first information and the above-mentioned third information into a third learning model LM3 that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information regarding the environment in which the learning base station 40 is installed, and weather information in region R including the learning terminal 30 and the learning base station 40.

[0146] As described above, when the machine learning device 20 determines the third learning model LM3 by weighting information regarding the communication quality with the outdoor base station 40, the accuracy of the estimation unit 115 in estimating weather information increases.

[0147] Furthermore, if the selection result received by the reception unit 113 is the fourth learning model LM4, the estimation unit 115 inputs the fourth input data ID4 to the fourth learning model LM4, and then uses the output result from the fourth learning model LM4 as the weather information estimation result.

[0148] In other words, the estimation unit 115 estimates weather information by inputting the above-mentioned first information and the above-mentioned fourth information into a fourth learning model LM4 that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information regarding the model of the learning terminal 30, and weather information in the region R including the learning terminal 30 and the learning base station 40.

[0149] For example, depending on the type of OS (Operating System) installed on terminal 30, terminal 30 may select a frequency with high attenuation due to the amount of moisture in the atmosphere, making communication easier. The estimation unit 115 can estimate weather information based on this phenomenon.

[0150] As described above, when the machine learning device 20 determines the fourth learning model LM4 by weighting information regarding the communication quality between a specific type of terminal 30 and the base station 40, the accuracy of the estimation unit 115 in estimating weather information increases.

[0151] Furthermore, if the selection result received by the reception unit 113 is the fifth learning model LM5, the estimation unit 115 inputs the fifth input data ID 5 to the fifth learning model LM5, and uses the output result from the fifth learning model LM5 as the meteorological information estimation result.

[0152] In other words, the estimation unit 115 estimates weather information by inputting the above-mentioned first information, the location information of the terminal 30, and the location information of the base station 40 into a fifth learning model LM5 that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, the location information of the learning terminal 30, the location information of the learning base station 40, and weather information in the region R including the learning terminal 30 and the learning base station 40.

[0153] Based on the location information of terminal 30 and base station 40, the distance between terminal 30 and base station 40 becomes clear. The estimation unit 115 can estimate weather information based on this distance. In other words, the estimation unit 115 can estimate weather information that takes into account whether the attenuation of radio wave intensity is due to moisture in the atmosphere or distance.

[0154] The display control unit 116 causes the weather information estimated by the estimation unit 115 to be displayed on an external display device. For example, the display device may be the terminal 30[k] itself or a display device installed in the terminal 30[k].

[0155] 1-2: The operation diagram 24 of the first embodiment is a flowchart showing an example of the operation of the weather estimation device 10 according to the first embodiment.

[0156] In step S1, the processing unit 11 of the weather estimation device 10 functions as an acquisition unit 112. The processing unit 11 acquires first information from the base station 40 regarding the quality of communication between the terminal 30 and the base station 40.

[0157] In step S2, the processing unit 11 functions as a receiving unit 113. The processing unit 11 receives the selection result of the learning model LM from the input device 13 or terminal 30[k]. Here, as an example, it is assumed that the processing unit 11 receives the first learning model LM1 as the selection result of the learning model LM.

[0158] In step S3, the processing unit 11 functions as a data determination unit 114. The processing unit 11 determines the input data ID to be input to the learning model LM received in step S2. This input data ID includes the first information obtained in step S1. Here, as an example, it is assumed that the processing unit 11 determines the first input data ID 1 to be input to the first learning model LM1.

[0159] In step S4, the processing unit 11 functions as an estimation unit 115. Based on the first information acquired in step S1, the processing unit 11 estimates weather information for the first region including the terminal 30 and the base station 40. More specifically, the processing unit 11 inputs the input data ID determined in step S3 to the learning model LM received in step S2, and uses the output result output from the learning model LM as the weather information estimation result. Here, as an example, it is assumed that the processing unit 11 inputs the first input data ID 1 to the first learning model LM1, and then uses the output result output from the first learning model LM1 as the weather information estimation result.

[0160] In step S5, the processing unit 11 functions as a display control unit 116. The processing unit 11 causes the weather information estimated in step S4 to be displayed on a display device. The display device is, for example, the terminal 30[k] itself, or a display device provided in the terminal 30[k].

[0161] 1-3: Effects of the First Embodiment The weather estimation device 10 according to this embodiment comprises an acquisition unit 112 and an estimation unit 115. The acquisition unit 112 acquires first information relating to the quality of communication between the terminal 30 and the base station 40. The estimation unit 115 estimates weather information relating to the weather in a region R[1] including the terminal 30 and the base station 40 based on the first information.

[0162] By having the above configuration, the weather estimation device 10 can estimate weather information using a wide-area atmospheric moisture measurement method that allows for measurement of conditions at a distance, compared to conventional technology.

[0163] Furthermore, in the weather estimation device 10 according to this embodiment, the estimation unit 115 estimates weather information by inputting first information to a learned model LM that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40 and information regarding weather in the region R including the learning terminal 30 and the learning base station 40.

[0164] By having the above configuration, the weather estimation device 10 can construct a first learning model LM1 that estimates precipitation based on radio wave strength information between base stations 40 and terminals 30, which is collected in real time from all over Japan, and then perform precipitation estimation on a nationwide scale in real time.

[0165] Furthermore, in the weather estimation device 10 according to this embodiment, the acquisition unit 112 acquires second information that specifies the frequency band of radio waves used for communication. The estimation unit 115 estimates weather information by inputting the first information and the second information to a learning model LM that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information specifying the frequency band of radio waves used for communication between the learning terminal 30 and the learning base station 40, and weather information in the region R including the learning terminal 30 and the learning base station 40.

[0166] By having the above configuration, the weather estimation device 10 can improve the accuracy of weather information estimation based on changes in radio wave intensity across multiple frequency bands.

[0167] Furthermore, in the weather estimation device 10 according to this embodiment, the acquisition unit 112 acquires third information relating to the environment in which the base station 40 is installed. The estimation unit 115 estimates weather information by inputting the first information and the third information to a learning model LM that has already learned the relationship between information relating to the quality of communication between the learning terminal 30 and the learning base station 40, information relating to the environment in which the learning base station 40 is installed, and weather information in the region R including the learning terminal 30 and the learning base station 40.

[0168] By having the above configuration, the weather estimation device 10 improves the accuracy of its weather estimation when the machine learning device 20 determines the third learning model LM3 by weighting information regarding the communication quality with the outdoor base station 40.

[0169] Furthermore, in the weather estimation device 10 according to this embodiment, the acquisition unit 112 acquires fourth information regarding the model of the terminal 30. The estimation unit 115 estimates weather information by inputting the first information and the fourth information to a learning model LM that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, information regarding the model of the learning terminal 30, and weather information in the region R including the learning terminal 30 and the learning base station 40.

[0170] The weather estimation device 10, with the above configuration, can estimate weather information that takes into account events that make it easier for the terminal 30 to communicate by selecting a frequency with high attenuation due to the amount of moisture in the atmosphere, depending on the type of OS installed on the terminal 30. As a result, the accuracy of the weather estimation device 10's weather estimation is improved.

[0171] Furthermore, in the weather estimation device 10 according to this embodiment, the acquisition unit 112 acquires location information of the terminal 30 and location information of the base station 40. The estimation unit 115 estimates weather information by inputting the first information, the location information of the terminal 30, and the location information of the base station 40 into a learning model LM that has already learned the relationship between information regarding the quality of communication between the learning terminal 30 and the learning base station 40, the location information of the learning terminal 30, the location information of the learning base station 40, and weather information in the region R including the learning terminal 30 and the learning base station 40.

[0172] The weather estimation device 10, with the above configuration, can determine the distance between terminal 30 and base station 40 based on the location information of terminal 30 and base station 40. Furthermore, the weather estimation device 10 can estimate weather information based on this distance. In other words, the estimation unit 115 can estimate weather information that takes into account whether the attenuation of radio wave intensity is due to moisture in the atmosphere or distance.

[0173] Furthermore, the weather estimation device 10 according to this embodiment further comprises a display control unit 116. The display control unit 116 causes weather information to be displayed on a display device.

[0174] By having the above configuration, the weather estimation device 10 allows, for example, a user U on terminal 30 to visually view weather information.

[0175] Furthermore, in the weather estimation device 10 according to this embodiment, the weather information is information indicating the amount of precipitation in region R[1].

[0176] By having the above configuration, the weather estimation device 10 can recognize the estimated precipitation amount as a result of weather information estimation.

[0177] Furthermore, in the weather estimation device 10 according to this embodiment, the first information is the signal strength of the radio waves received when the base station 40 receives radio waves transmitted from the terminal 30.

[0178] By having the above configuration, the weather estimation device 10 can estimate weather information based on the signal strength of the radio waves used for communication between the terminal 30 and the base station 40.

[0179] 2. Modifications The present disclosure is not limited to the embodiments illustrated above. Specific examples of modifications are given below.

[0180] 2-1: Modification 1 In the above embodiment, it was assumed that each base station 40 uses one frequency band of radio waves for communication with the terminal 30. However, each base station 40 may use multiple frequency bands of radio waves for communication with the terminal 30.

[0181] 2-2: Modification 2 In the above embodiment, the first training data TD1 to the fifth training data TD5 include the item "precipitation". However, the first training data TD1 to the fifth training data TD5 may include "amount of moisture in the atmosphere" or "humidity" instead of the item "precipitation". These "amount of moisture in the atmosphere" or "humidity" are examples of "weather-related meteorological information".

[0182] 2-3: Modification 3 In the above embodiment, the first training data TD1 to the fifth training data TD5 include the item "signal strength" as information regarding the quality of communication between the learning terminal 30 and the learning base station 40. However, the first training data TD1 to the fifth training data TD5 may include other items instead of the item "signal strength" as information regarding the quality of communication between the learning terminal 30 and the learning base station 40. For example, the first training data TD1 to the fifth training data TD5 may include the item "amount of error detected in communication" instead of the item "signal strength".

[0183] In this case, the acquisition unit 112 of the weather estimation device 10 acquires the amount of error detection as the "first information" mentioned above.

[0184] 2-4: Modification 4 In the above embodiment, the data determination unit 213 in the machine learning device 20 determines the second training data TD2 by combining the first training data TD1 and the third data DS3. The data determination unit 213 also determines the third training data TD3 by combining the first training data TD1 and the fourth data DS4. The data determination unit 213 also determines the fourth training data TD4 by combining the first training data TD1 and the fifth data DS5. The data determination unit 213 also determines the fifth training data TD5 by combining the first training data TD1, the sixth data DS6, and the seventh data DS7.

[0185] However, the data determination unit 213 may determine the training data TD by combining any number of data DS from the third data DS3 to the seventh data DS7 with the first training data TD1. For example, the data determination unit 213 may determine the training data TD by combining all of the third data DS3 to the seventh data DS7 with the first training data TD1.

[0186] In this case, the acquisition unit 112 of the weather estimation device 10 acquires data for all items of the training data TD except for "rainfall amount". The data determination unit 114 determines an input data ID that includes all items of the training data TD except for "rainfall amount".

[0187] 3. Other (1) In the embodiments described above, ROM and RAM were given as examples for the storage device 12 and storage device 22, but other suitable storage media include flexible disks, magneto-optical disks (e.g., compact disks, digital multipurpose disks, Blu-ray® disks), smart cards, flash memory devices (e.g., cards, sticks, key drives), CD-ROMs (Compact Disc-ROMs), registers, removable disks, hard disks, floppy® disks, magnetic strips, databases, servers, and other appropriate storage media.

[0188] (2) In the embodiments described above, the information, signals, etc. may be represented using any of the various different techniques. For example, the data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

[0189] (3) In the embodiments described above, the input and output information may be stored in a specific location (e.g., memory) or managed using a management table. The input and output information may be overwritten, updated, or appended to. The output information may be deleted. The input information may be transmitted to other devices.

[0190] (4) In the embodiments described above, the determination may be made by a value represented using one bit (0 or 1), by a boolean value (true or false), or by a numerical comparison (for example, a comparison with a predetermined value).

[0191] (5) The processing procedures, sequences, flowcharts, etc., exemplified in the embodiments described above may be rearranged in order, as long as there is no contradiction. For example, the methods described in this disclosure present various step elements using an exemplary order and are not limited to the specific order presented.

[0192] (6) Each function illustrated in Figures 1 to 24 is realized by any combination of at least one of hardware and software. Furthermore, the method of realizing each functional block is not particularly limited. That is, each functional block may be realized using one device that is physically or logically coupled, or it may be realized using two or more physically or logically separated devices that are directly or indirectly connected (for example, using wired or wireless connections). A functional block may also be realized by combining the above one device or the above multiple devices with software.

[0193] (7) The programs illustrated in the embodiments described above should be broadly interpreted to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, execution threads, procedures, functions, etc., whether they are called software, firmware, middleware, microcode, hardware description languages ​​or by other names.

[0194] Furthermore, software, instructions, information, etc., may be transmitted and received via a transmission medium. For example, if software is transmitted from a website, server, or other remote source using at least one of wired technology (such as coaxial cable, fiber optic cable, twisted pair, or digital subscriber line (DSL)) and wireless technology (such as infrared or microwave), then at least one of these wired and wireless technologies is included in the definition of a transmission medium.

[0195] (8) In each of the above-mentioned forms, the terms “system” and “network” shall be used interchangeably.

[0196] (9) The information, parameters, etc. described in this disclosure may be expressed using absolute values, relative values ​​from a given value, or other corresponding information.

[0197] (10) In the embodiments described above, the terminal 30 may be a mobile station (MS). A mobile station may also be referred to by those skilled in the art as a subscriber station, mobile unit, subscriber unit, wireless unit, remote unit, mobile device, wireless device, wireless communication device, remote device, mobile subscriber station, access terminal, mobile terminal, wireless terminal, remote terminal, handset, user agent, mobile client, client, or several other appropriate terms. In this disclosure, terms such as “mobile station,” “user terminal,” “user equipment (UE),” and “terminal” may be used interchangeably.

[0198] (11) In the embodiments described above, the terms “connected,” “coupled,” or any variation thereof, mean any direct or indirect connection or coupling between two or more elements, and may include the presence of one or more intermediate elements between two elements that are “connected” or “coupled” with each other. The coupling or connection between elements may be a physical coupling or connection, a logical coupling or connection, or a combination thereof. For example, “connection” may be reinterpreted as “access.” As used in this disclosure, two elements may be considered to be “connected” or “coupled” with each other using at least one of one or more wires, cables and printed electrical connections, and, in some non-limiting and non-exclusive examples, electromagnetic energy having wavelengths in the radio frequency domain, microwave domain and optical (both visible and invisible) domain.

[0199] (12) In the embodiments described above, the phrase "based on" does not mean "based solely on" unless otherwise specified. In other words, the phrase "based on" means both "based solely on" and "based at least on".

[0200] (13) The terms “determining” and “determining” as used in this disclosure may encompass a wide variety of actions. “Determining” may include, for example, judging, calculating, computing, processing, deriving, investigating, looking up, searching, or inquiring (e.g., searching in a table, database, or other data structure), or ascertaining. “Determining” may also include receiving (e.g., receiving information), transmitting (e.g., sending information), inputting, outputting, or accessing (e.g., accessing data in memory). Furthermore, "judgment" and "decision" can include considering something as having been "judged" or "decided" after resolving, selecting, choosing, establishing, comparing, etc. In other words, "judgment" and "decision" can include considering something as having been "judged" or "decided" after some action. Also, "judgment (decision)" can be reinterpreted as "assuming," "expecting," or "considering."

[0201] (14) In the embodiments described above, where “include,” “including,” and variations thereof are used, these terms are intended to be inclusive, as is the term “comprising.” Furthermore, the term “or” as used in this disclosure is not intended to be exclusive OR.

[0202] (15) In the present disclosure, if articles are added by translation, such as a, an, and the in English, the present disclosure may include the fact that the noun following these articles is plural.

[0203] (16) In this disclosure, the term “A and B are different” may mean “A and B are different from each other.” The term may also mean “A and B are each different from C.” Terms such as “separate” and “combine” may be interpreted in the same way as “different.”

[0204] (17) Each aspect / embodiment described herein may be used individually, in combination, or switched between as needed during implementation. Furthermore, notification of the specified information (e.g., notification that "it is X") is not limited to explicit notification, but may also be implicit (e.g., by not providing such notification).

[0205] Although the present disclosure has been described in detail above, it will be clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the intent and scope of the present disclosure as defined by the claims. Accordingly, the descriptions in the present disclosure are illustrative and not restrictive in any way.

[0206] 1... Weather estimation system, 10... Weather estimation device, 11... Processing unit, 12... Memory device, 13... Input device, 14... Communication device, 20... Machine learning device, 21... Processing unit, 22... Memory device, 23... Input device, 24... Communication device, 30... Terminal, 40... Base station, 111... Communication control unit, 112... Acquisition unit, 113... Reception unit, 114... Data determination unit, 115... Estimation unit, 116... Display control unit, 211... Communication control unit, 212... Acquisition unit, 213...Data determination unit, 214...First model determination unit, 215...Second model determination unit, 216...Third model determination unit, 217...Fourth model determination unit, 218...Fifth model determination unit, DB...Data block, DS...Data, ID...Input data, LM...Learning model, NET1...Communication network, NET2...Communication network, PR1...Control program, PR2...Control program, R...Domain, TD...Training data, U...User

Claims

1. A weather estimation device comprising: an acquisition unit that acquires first information relating to the quality of communication between a terminal and a base station; and an estimation unit that estimates weather information relating to the weather in a first region including the terminal and the base station based on the first information.

2. The weather estimation device according to claim 1, wherein the estimation unit estimates the weather information by inputting the first information to a learning model that has already learned the relationship between information regarding the quality of communication between a learning terminal and a learning base station and information regarding weather in a region including the learning terminal and the learning base station.

3. The weather estimation device according to claim 1, wherein the acquisition unit acquires second information specifying the frequency band of radio waves used for the communication, and the estimation unit estimates the weather information by inputting the first information and the second information to a learning model that has already learned the relationship between information regarding the quality of communication between a learning terminal and a learning base station, information specifying the frequency band of radio waves used for communication between the learning terminal and the learning base station, and information regarding weather in the area including the learning terminal and the learning base station.

4. The precipitation estimation device according to claim 1, wherein the acquisition unit acquires third information relating to the environment in which the base station is installed, and the estimation unit estimates the meteorological information by inputting the first information and the third information to a learning model that has already learned the relationship between information relating to the quality of communication between a learning terminal and a learning base station, information relating to the environment in which the learning base station is installed, and meteorological information in the area including the learning terminal and the learning base station.

5. The precipitation estimation device according to claim 1, wherein the acquisition unit acquires fourth information relating to the model of the terminal, and the estimation unit estimates the weather information by inputting the first information and the fourth information to a learning model that has already learned the relationship between information relating to the quality of communication between the learning terminal and the learning base station, information relating to the model of the learning terminal, and weather information in the area including the learning terminal and the learning base station.

6. The precipitation estimation device according to claim 1, wherein the acquisition unit acquires location information of the terminal and location information of the base station, and the estimation unit estimates the weather information by inputting the first information, the location information of the terminal, and the location information of the base station to a learning model that has learned the relationship between information regarding the quality of communication between the learning terminal and the learning base station, the location information of the learning terminal, the location information of the learning base station, and weather information in the area including the learning terminal and the learning base station.

7. The weather estimation device according to claim 1, further comprising a display control unit for displaying the weather information on a display device.

8. The weather information is information indicating the amount of precipitation in the first region, as described in claim 1.

9. The weather estimation device according to claim 1, wherein the first information is the signal strength of the radio waves received when the base station receives radio waves transmitted from the terminal.

Images