Wireless quality prediction device, wireless quality prediction method, and program

Abstract

In order to make it possible to predict the wireless quality of a wireless terminal after the elapse of a predetermined time without creating a wireless quality distribution, this wireless quality prediction device, which predicts the wireless quality between a wireless base station and a wireless terminal that performs wireless communication with the wireless base station, comprises: a reception power learner configured to predict a predicted value of the reception power of the wireless terminal after the elapse of a predetermined time, on the basis of the terminal position of the wireless terminal and an estimated value of the terminal state after the elapse of the predetermined time; a wireless quality learner configured to predict a predicted value of the wireless quality after the elapse of the predetermined time, on the basis of information about the wireless base station, information about the wireless communication, information about the wireless terminal, and the predicted value of the reception power; and a notification unit configured to notify the wireless terminal of the predicted value of the wireless quality after the elapse of the predetermined time.

Description

[Technical Field]

[0001] The present invention relates to a wireless quality prediction device, a wireless quality prediction method, and a program. [Background technology]

[0002] A method is being studied in which radio wave measurement results in a wireless communication area are measured in advance to create a wireless quality distribution (heat map), and handover control is based on the created wireless quality distribution (see, for example, Non-Patent Document 1). [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Konishi et al., "Effects of Automatic Optimization by Self-Organizing Networks (SON) Technology for LTE / LTE-Advanced Systems," IEICE Transactions on Network Engineering, Vol. J97-B No. 8 pp. 599-610, 2014 Summary of the Invention [Problem to be solved by the invention]

[0004] To create a wireless quality distribution, it is necessary to measure wireless quality, such as throughput, for each prediction area. However, because measuring throughput uses a large amount of communication bandwidth and interferes with other communications, creating a wireless quality distribution has been difficult.

[0005] The embodiment of the present invention has been made in consideration of the above-mentioned problems, and makes it possible to predict the wireless quality of a wireless terminal after a predetermined time has elapsed without creating a wireless quality distribution. [Means for solving the problem]

[0006] In order to solve the above problem, a wireless quality prediction device according to an embodiment of the present invention is a wireless quality prediction device that predicts wireless quality between a wireless base station and a wireless terminal that performs wireless communication with the wireless base station, a first neural network that has been trained in advance using as training data the current terminal position and terminal state of another wireless terminal that wirelessly communicates with the wireless base station, and past received power of the other wireless terminal from a predetermined time ago as feature quantities, and the current received power of the other wireless terminal, and inputs into the first neural network the estimated terminal position and terminal state of the wireless terminal after the predetermined time has elapsed, and the received power of the wireless terminal; a received power learning device configured to predict a predicted value of received power of the wireless terminal after the predetermined time has elapsed; a second neural network that has been trained in advance to predict communication quality of the wireless communication, using information about the wireless base station, information about the wireless communication, information about other wireless terminals that perform wireless communication with the wireless base station, information about interference waves of the other wireless terminals, and received power of the other wireless terminals as feature quantities, and the second neural network is input with the information about the wireless base station, the information about the wireless communication, information about the wireless terminal, the information about the interference waves measured by the wireless terminal, and a predicted value of the received power of the wireless terminal after the predetermined time has elapsed; The wireless communication device includes a wireless quality learning device configured to predict a predicted value of the wireless quality after the predetermined time has elapsed, and a notification unit configured to notify the wireless terminal of the predicted value of the wireless quality after the predetermined time has elapsed. [Effects of the Invention]

[0007] According to the embodiment of the present invention, it becomes possible to predict the wireless quality of a wireless terminal after a predetermined time has elapsed, without creating a wireless quality distribution. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a diagram illustrating an example of a system configuration of a wireless communication system according to an embodiment of the present invention. [Figure 2] FIG. 10 is a diagram for explaining a learning device for received power according to the present embodiment. [Figure 3] FIG. 10 is a diagram for explaining learning of a learning device of received power according to the present embodiment. [Figure 4] FIG. 10 is a diagram for explaining a learning device of wireless quality according to the present embodiment. [Figure 5] FIG. 10 is a diagram for explaining learning of a learning device of wireless quality according to the present embodiment. [Figure 6] 10 is a flowchart illustrating an example of a process for predicting wireless quality according to the present embodiment. [Figure 7] FIG. 1 is a diagram (1) for explaining a wireless quality prediction process according to the present embodiment. [Figure 8] FIG. 10 is a diagram (2) for explaining the wireless quality prediction process according to the present embodiment. [Figure 9] 1 is a diagram illustrating an example of the hardware configuration of a wireless quality prediction device and a wireless base station according to an embodiment of the present invention. [Figure 10] FIG. 2 is a diagram illustrating an example of a hardware configuration of a wireless terminal according to the present embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. The embodiment described below is merely an example, and the embodiment to which the present invention is applied is not limited to the following embodiment.

[0010] <System configuration> 1 is a diagram showing an example of a system configuration of a wireless communication system according to this embodiment. The wireless communication system 1 includes a wireless base station 10, a wireless terminal 20 that performs wireless communication with the wireless base station 10, and a wireless quality prediction device 100 that predicts wireless quality between the wireless base station 10 and the wireless terminal 20.

[0011] The wireless base station 10 provides wireless communication services such as 5G (5th Generation) and LTE (Long Term Evolution) to wireless terminals 20 located within the communication area of ​​the wireless base station 10. The wireless terminals 20 can perform wireless communication with the wireless base station 10 within the communication area of ​​the wireless base station 10.

[0012] The wireless quality predicting device 100 predicts wireless quality between the wireless base station 10 and the wireless terminal 20, and provides the wireless terminal 20 with the predicted value of the wireless quality. For example, the wireless quality predicting device 100 predicts the throughput (an example of wireless quality) of the wireless terminal 20 after a predetermined time has elapsed (n seconds), and notifies the wireless terminal 20 of the predicted value of the throughput.

[0013] The wireless terminal 20 performs control such as handover of the wireless terminal 20 based on the notified predicted value of throughput.

[0014] (About the assignment) In LTE, for example, the received power strength Ms of the connected cell s is compared with the received power strength Mn of the neighboring cell n, and handover is performed to switch the cell to which the connection is made. offset,s,n When the switching condition continues for a certain period (TTT: Time to Trigger) or more, the handover process is executed. offset,s,n is an offset value that is uniquely set between cell s and cell n.

[0015] However, in this method, a handover is determined after the wireless quality has deteriorated, and therefore, for example, if the wireless quality deteriorates suddenly, the communication quality may temporarily deteriorate. This is a problem.

[0016] Therefore, a method has been studied in which radio wave measurement results in a wireless communication area are measured in advance to create a wireless quality distribution (heat map), and handover control is based on the created wireless quality distribution (for example, Non-Patent Document 1).

[0017] However, in order to create a wireless quality distribution, it is necessary to measure wireless quality, such as throughput, for each prediction area. However, since measuring throughput uses a large amount of communication bandwidth and interferes with other communications, creating a wireless quality distribution has been difficult.

[0018] Therefore, in this embodiment, a wireless quality prediction device and a wireless quality prediction method are provided that can predict the wireless quality of a wireless terminal after a predetermined time has elapsed without creating a wireless quality distribution.

[0019] <Functional configuration> (Functional configuration of wireless quality prediction device) The wireless quality predicting device 100 includes one or more computers, and by executing a predetermined program on the one or more computers, realizes, for example, a received power learning device 101, a wireless quality learning device 102, a notification unit 103, and an MCS conversion unit 104. Note that at least a part of the above functional configurations may be realized by hardware.

[0020] The received power learning device 101 executes a received power prediction process to predict a predicted value of the received power of the wireless terminal 20 after a predetermined time has elapsed (for example, after n seconds) based on the estimated values ​​of the terminal position and terminal state of the wireless terminal 20 after a predetermined time has elapsed (for example, after n seconds). The received power learning device 101 also notifies the wireless quality learning device 102 of the predicted value of the received power after n seconds.

[0021] Here, the terminal position is position information (for example, latitude, longitude, etc.) indicating the position of the wireless terminal 20. Furthermore, the terminal state is information such as the orientation and speed of the wireless terminal 20 acquired by devices such as an acceleration sensor, a gyro sensor, or an IMU (Inertial Measurement Unit) provided in the wireless terminal 20. Even for the same wireless terminal 20, differences in received power occur depending on the orientation or speed, so it is desirable to use information on the terminal state when predicting received power. Note that the terminal state may use only the orientation of the wireless terminal 20 or only the speed of the wireless terminal 20, etc.

[0022] The wireless quality learning device 102 executes a wireless quality prediction process to predict the predicted value of wireless quality after a predetermined time has elapsed (n seconds) based on information about the wireless base station, information about the wireless communication, information about the wireless terminal, and the predicted value of the received power predicted by the received power learning device 101.

[0023] Here, the information about the radio base station includes, for example, information such as the manufacturer, type, and communication standard of the radio base station 10, and the number of connected radio terminals 20. The processing time (delay) differs depending on the manufacturer or type of the radio base station 10. The number of connected radio terminals 20 also affects the radio quality (for example, throughput). Therefore, it is desirable to use the information about the radio base station when predicting the radio quality. However, at least a part of the information about the radio base station may be omitted. The information about the radio communication includes, for example, information such as the MCS (Modulation and Coding Scheme) of the radio communication or the maximum throughput value. The MCS indicates a combination of a data modulation scheme and a channel coding rate, and normally, the larger the MCS number, the larger the transport block size, and the higher the throughput that can be achieved.

[0024] The wireless terminal information includes, for example, the type of wireless terminal 20 (such as the manufacturer or model name of the chipset) and the form of the wireless terminal 20 (such as a robot, vehicle, smartphone, or wearable terminal). Wireless quality (such as throughput) may differ depending on the manufacturer and model name of the chipset. Also, the measured received power varies depending on the shape of the wireless terminal 20, etc. Therefore, it is desirable to use information from wireless base stations when predicting wireless quality. However, at least a portion of the wireless terminal information may be omitted.

[0025] The wireless quality of the wireless terminal 20 predicted by the wireless quality learning device 102 includes, for example, the throughput of wireless communication between the wireless base station 10 and the wireless terminal 20. Note that the wireless quality of the wireless terminal 20 may be an index other than the throughput, but the following description will be given assuming that the wireless quality of the wireless terminal 20 is the throughput.

[0026] The notification unit 103 notifies the wireless terminal 20 via the wireless base station 10 of the predicted value of the throughput of the wireless terminal 20 after n seconds, which is predicted by the wireless quality learning device 102.

[0027] The MCS conversion unit 104 performs MCS conversion processing to convert, for example, the MCS notified from the radio base station 10 into a maximum throughput value for wireless communication. For example, the MCS conversion unit 104 has correspondence information that stores a plurality of MCSs and maximum throughput values ​​corresponding to each MCS in association with each other, and uses this correspondence information to convert the MCS into the maximum throughput value. Furthermore, the MCS conversion unit 104 notifies the radio base station information received from the radio base station 10 and the converted maximum throughput, etc. to the wireless quality learning device 102. Note that the function of the MCS conversion unit 104 may be provided in the radio base station 10.

[0028] (Functional configuration of wireless terminal) The wireless terminal 20 implements, for example, a communication control unit 21, a position acquisition unit 22, a status acquisition unit 23, a terminal information transmission unit 24, and a terminal control unit 25 by executing a predetermined program on a computer included in the wireless terminal 20. Note that at least a part of the above functional configurations may be implemented by hardware.

[0029] The communication control unit 21 executes a communication control process to control wireless communication with the wireless base station 10. For example, the communication control unit 21 determines a CQI (Channel Quality Indicator) value indicating the state of the downlink wireless channel from the reception level of a pilot signal received by the antenna 27, and notifies the wireless base station 10 of the determined CQI value. Furthermore, the communication control unit 21 performs wireless communication with the wireless base station 10 via the antenna 27, based on the MCS notified from the wireless base station 10.

[0030] The position acquisition unit 22 acquires the current position (terminal position) of the wireless terminal 20 using, for example, a positioning device such as a GPS (Global Positioning System) device provided in the wireless terminal 20. Furthermore, the position acquisition unit 22 estimates the position (terminal position) of the wireless terminal 20 after a predetermined time has elapsed (n seconds later) based on, for example, the acquired current position of the wireless terminal 20 and the movement of the wireless terminal 20 measured by a device such as an acceleration sensor, a gyro sensor, or an IMU provided in the wireless terminal 20. As an example, the position acquisition unit 22 may estimate the terminal position of the wireless terminal 20 n seconds later using a machine learning model or the like that has been trained in advance using the current position of the wireless terminal 20 and the movement of the wireless terminal 20 as feature quantities and the position of the wireless terminal 20 n seconds later as training data.

[0031] The state acquisition unit 23 acquires the terminal state (orientation, speed, etc.) of the wireless terminal 20 based on sensor data acquired by, for example, an acceleration sensor, a gyro sensor, an IMU, or the like provided in the wireless terminal 20. Furthermore, the state acquisition unit 23 estimates the state (terminal state) of the wireless terminal 20 after a predetermined time has elapsed (n seconds later) based on the sensor data acquired by the acceleration sensor, the gyro sensor, the IMU, etc. For example, the state acquisition unit 23 may estimate the terminal state of the wireless terminal 20 n seconds later using a machine learning model or the like that has been trained in advance using past sensor data of the wireless terminal 20 as feature quantities and the current terminal state of the wireless terminal 20 as training data.

[0032] The terminal information transmitter 24 transmits, for example, information on the wireless terminal, received power, interference wave information (SINR), and information on the terminal position and terminal status after a predetermined time has elapsed (n seconds), to the wireless quality predicting device 100 via the wireless base station 10. Here, the interference wave information (SINR) deteriorates due to the influence of interference waves in the wireless section 2, and if this value is low, for example, an increase in error rate or waiting time (delay) occurs, resulting in a decrease in throughput. Therefore, it is desirable to use the interference wave information (SINR) measured by the wireless terminal 20 when predicting wireless quality. However, the interference wave information (SINR) is optional and not essential.

[0033] Information about the wireless terminal is stored in advance, for example, by the terminal information transmitter 24 or the terminal control unit 25. The received power and interference wave information are acquired, for example, by the terminal information transmitter 24 from the communication control unit 21. The terminal position and terminal state after n seconds are estimated by the position acquisition unit 22 and the state acquisition unit 23.

[0034] The terminal control unit 25 performs overall control of the wireless terminal 20. For example, the terminal control unit 25 performs, for example, handover control based on the throughput of the wireless terminal 20 after a predetermined time (n seconds) has elapsed, which is notified by the wireless quality predicting device 100.

[0035] Furthermore, if the wireless terminal 20 is, for example, a wireless terminal with a movement function, such as a robot or vehicle that holds movement route information and moves along a predetermined route, the terminal control unit 25 controls the movement of the wireless terminal 20. In this case, the terminal control unit 25 may change the movement route of the wireless terminal 20 based on the throughput of the wireless terminal 20 after a predetermined time (n seconds) has elapsed, which is notified by the wireless quality predicting device 100.

[0036] (Functional configuration of wireless base station) The radio base station 10 implements, for example, a communication control unit 11 and a transmission / reception unit 12 by executing a predetermined program on a computer included in the radio base station 10. Note that at least a part of the above functional configurations may be implemented by hardware.

[0037] The communication control unit 11 executes a communication control process to control wireless communication with the wireless terminal 20. For example, the communication control unit 11 determines an MCS based on a CQI value received from the wireless terminal 20, and notifies the wireless terminal 20 of the determined MCS. The communication control unit 11 also performs wireless communication with the wireless terminal 20 via the antenna 13 based on the determined MCS.

[0038] The transmitter / receiver 12 notifies, for example, information about the wireless base station and the MCS determined by the communication control unit 11 to the wireless quality predicting device 100. The transmitter / receiver 12 also receives from the wireless quality predicting device 100 a predicted value of throughput n seconds from now that the wireless quality predicting device 100 notifies to the wireless terminal 20, and transfers it to the wireless terminal 20.

[0039] (Regarding the received power learning module) 2 is a diagram for explaining a received power learning device according to this embodiment. The received power learning device 101 has a first neural network 201 that has been trained to predict the received power of one or more wireless terminals connected to the wireless base station 10 based on current data or past data acquired from the wireless terminals.

[0040] For example, the first neural network 201 uses the current terminal position and terminal state of one or more wireless terminals that communicate wirelessly with the wireless base station 10, and the received power of the wireless terminals a predetermined time ago as feature quantities, and is pre-trained using the current received power of the terminals as training data.

[0041] The received power learning device 101 inputs the estimated terminal position and terminal state of the wireless terminal 20 after a predetermined time has elapsed, the current received power of the wireless terminal 20, etc. into this first neural network 201, and predicts the predicted value of the received power of the wireless terminal 20 after the predetermined time has elapsed.

[0042] Preferably, the received power learner 101 periodically learns the first neural network 201 using the current location and terminal state of one or more wireless terminals that communicate wirelessly with the wireless base station 10, as well as the current received power of the wireless terminals and past received power from a predetermined time ago.

[0043] Fig. 3 is a diagram for explaining learning by the received power learner 101 according to this embodiment. In Fig. 3, it is assumed that a wireless terminal 20-1 located within a communication area 301 of a wireless base station 10-1 is performing wireless communication with the wireless base station 10-1. In this case, the wireless terminal 20-1 sequentially measures the received power received from the wireless base station 10-1, as well as the terminal position and terminal state of the wireless terminal 20-1, and transmits the measurement results to the wireless quality predicting apparatus 100 via the wireless base station 10-1. The received power learner 101 also learns a first neural network 201 corresponding to the wireless base station 10-1 based on the received measurement results.

[0044] Similarly, it is assumed that wireless terminals 20-2 and 20-3 located within the communication area 302 of wireless base station 10-2 are performing wireless communication with wireless base station 10-2. In this case, wireless terminals 20-2 and 20-3 sequentially measure the received power received from wireless base station 10-2, as well as their own terminal positions and terminal states, and transmit the measurement results to wireless quality predicting apparatus 100 via wireless base station 10-2. Furthermore, received power learner 101 learns first neural network 201 corresponding to wireless base station 10-2 based on the received measurement results.

[0045] As another example, the first neural network 201 may be a neural network that has been trained to predict the current received power using the current terminal positions and terminal states of one or more wireless terminals that communicate wirelessly with the wireless base station 10 as features, as shown in FIG. 2.

[0046] In this way, the received power learning device 101 is configured to predict the predicted value of the received power of the wireless terminal 20 after a predetermined time has passed, based on the estimated values ​​of the terminal position and terminal state of the wireless terminal 20 after a predetermined time has passed.

[0047] (About the wireless quality learner) 4 is a diagram for explaining the learning device of wireless quality according to this embodiment. The learning device of wireless quality 102 has a second neural network 401 that has been trained in advance to predict the communication quality of the wireless communication using information on wireless base stations, information on wireless communication, information on wireless terminals, information on interference waves of wireless terminals 20, and received power of the wireless terminals as feature quantities.

[0048] For example, as shown in FIG. 4, the second neural network 401 is trained to predict the throughput of the wireless terminal 20 using information about the wireless base station, the maximum throughput (or MCS), information about the wireless terminal, and the received power of the wireless terminal 20 as features.

[0049] Preferably, the feature quantity further includes interference wave information of the wireless terminal 20. The interference wave information of the wireless terminal 20 may be, for example, a signal-to-noise ratio (SINR) measured by the wireless terminal 20. Note that the interference wave information of the wireless terminal 20 may be information other than the SINR.

[0050] The wireless quality learning device 102 inputs, for example, information about the wireless base station, the maximum throughput, information about the wireless terminal, the SINR measured by the wireless terminal 20, and a predicted value of the received power of the wireless terminal 20 after a predetermined time has elapsed, to the second neural network 401. In this way, the wireless quality learning device 102 can acquire, from the second neural network 401, a predicted value of the throughput of the wireless terminal 20 after a predetermined time has elapsed.

[0051] Preferably, the wireless quality learner 102 trains the second neural network 401 in advance so that it is independent of the communication area provided by the wireless base station 10, location information, and the like.

[0052] 5 is a diagram for explaining learning of the wireless quality learner according to this embodiment. To enhance versatility, the wireless quality learner 102 preferably learns the second neural network 401 by utilizing data not only on the communication area of ​​the wireless base station 10 to be predicted but also on other communication areas.

[0053] 5, the wireless quality learner 102 acquires learning data such as received power of a wireless terminal, SINR, information about the wireless terminal, information about the wireless base station, and MCS (or maximum throughput) in areas A to C provided by a plurality of wireless base stations 10-1 to 10-3. The wireless quality learner 102 also uses the learning data acquired from various areas and various wireless terminals to train the second neural network 401. Note that the wireless quality learner 102 according to this embodiment does not need to measure throughput during learning and create a wireless quality distribution (heat map).

[0054] <Processing flow> Next, the processing flow of the method for predicting wireless quality according to this embodiment will be described.

[0055] 6 is a flowchart showing an example of a wireless quality prediction process according to this embodiment. This process shows a specific example of a wireless quality prediction process in which the wireless quality prediction device 100 predicts a predicted value of the throughput of the wireless terminal 20 after a predetermined time has elapsed (n seconds) when the wireless base station 10 and the wireless terminal 20 are performing wireless communication.

[0056] In step S601, the wireless terminal 20 acquires the received power, SINR, terminal position, and terminal state of the wireless terminal 20. For example, the terminal information transmission unit 24 of the wireless terminal 20 acquires the received power and SINR (an example of interference wave information) of the wireless terminal 20 from the communication control unit 21 or the terminal control unit 25. Furthermore, the position acquisition unit 22 of the wireless terminal 20 acquires the terminal position of the wireless terminal 20 from a GPS device or the like provided in the wireless terminal 20. Furthermore, the state acquisition unit 23 of the wireless terminal 20 acquires the terminal state (orientation, speed, etc.) of the wireless terminal 20 based on sensor data from an acceleration sensor, a gyro sensor, an IMU, or the like provided in the wireless terminal 20.

[0057] In step 602, wireless terminal 20 estimates (or calculates) an estimated value of the terminal position and terminal state n seconds from now based on the current terminal position and terminal state and sensor data from an acceleration sensor, a gyro sensor, an IMU, or the like, as shown in Fig. 7. For example, position acquisition unit 22 of wireless terminal 20 estimates an estimated value of the terminal position n seconds from now based on the acquired current terminal position and sensor data. Furthermore, state acquisition unit 23 of wireless terminal 20 estimates an estimated value of the terminal state n seconds from now based on the acquired current terminal state and sensor data.

[0058] In step S603, the terminal information transmitter 24 of the wireless terminal 20 transmits wireless terminal information, received power, SINR, and estimated values ​​of the terminal position and terminal state after n seconds, etc. to the received power learner 101 of the wireless quality predicting device 100 via the wireless base station 10. Of the information transmitted by the wireless terminal 20, information such as received power and estimated values ​​of the terminal position and terminal state after n seconds is input to the received power learner 101 as shown in Fig. 7. Furthermore, information such as the wireless terminal information and SINR is input to the wireless quality learner 102 as shown in Fig. 7.

[0059] In step S604, for example, in parallel with the processing of steps S601 to S602, the transmitter / receiver 12 of the wireless base station 10 transmits (notifies) information about the wireless base station and information such as the MCS to the wireless quality predicting apparatus 100. As a result, for example, as shown in Fig. 8, the information about the wireless base station and information such as the MCS transmitted by the wireless base station 10 is input to the MCS conversion unit 104 of the wireless quality predicting apparatus 100. Note that the information about the wireless base station includes, for example, information such as the communication standard and manufacturer of the wireless base station 10, or the number of wireless terminals connected to the wireless base station 10.

[0060] In step S605, the MCS conversion unit 104 of the wireless quality predicting apparatus 100 converts the MCS into a maximum throughput value using the correspondence information described above, and notifies the wireless quality learning device 102 of the converted maximum throughput value together with information about the wireless base station.

[0061] In step S606, the received power learning device 101 of the wireless quality predicting device 100 inputs the input received power, the terminal position and terminal state after n seconds, etc. into the first neural network 201, and predicts a predicted value of the received power after n seconds of the wireless terminal 20. The received power learning device 101 also notifies the wireless quality learning device 102 of the predicted value of the received power after n seconds.

[0062] By each of the above processes, for example, as shown in Figure 8, information such as information about the wireless terminal, the SINR of the wireless terminal 20, information about the wireless base station, the maximum throughput of wireless communication, and the predicted value of the received power of the wireless terminal 20 n seconds later is input to the wireless quality learning device 102.

[0063] In step S607, the wireless quality learning device 102 inputs information about the wireless terminal, SINR, information about the wireless base station, maximum throughput, predicted values ​​of received power, etc. into the second neural network 401, and predicts the predicted value of throughput n seconds later.

[0064] In step S608, the notification unit 103 of the wireless quality predicting device 100 notifies (transmits) to the wireless terminal 20 via the wireless base station 10 the predicted value of the throughput of the wireless terminal 20 after n seconds, which is predicted by the wireless quality learning device 102.

[0065] In step S609, the terminal control unit 25 of the wireless terminal 20 determines, for example, handover control or route change control based on the predicted value of the throughput after n seconds received from the wireless quality predicting device 100.

[0066] In step S610, the terminal control unit 25 executes control such as handover control or route change control according to the result of the determination.

[0067] 6, wireless terminal 20 can predict the throughput of wireless terminal 20 after a predetermined time has elapsed (n seconds), and therefore, if the throughput deteriorates after n seconds, it can start processing such as handover at an earlier timing. Note that the predetermined time (n seconds) is set as a setting value or a setting range by an administrator who manages wireless communication system 1 or a designer who designed the wireless quality, depending on the system requirements or prediction accuracy, etc.

[0068] <Hardware configuration example> (Hardware configuration of the prediction device) Fig. 9 is a diagram showing an example of the hardware configuration of a wireless quality prediction device and a wireless base station according to this embodiment. The wireless quality prediction device 100 and the wireless base station have, for example, the configuration of a computer 900 as shown in Fig. 9. In the example of Fig. 9, the computer 900 has a processor 901, a memory 902, a storage device 903, a communication device 904, an input device 905, an output device 906, a bus B, etc.

[0069] The processor 901 is, for example, an arithmetic unit such as a CPU (Central Processing Unit) that executes predetermined programs to realize various functions. The memory 902 is a storage medium readable by the computer 900, and includes, for example, a RAM (Random Access Memory) and a ROM (Read Only Memory). The storage device 903 is a computer-readable storage medium, and may include, for example, a HDD (Hard Disk Drive), an SSD (Solid State Drive), various optical disks, and magneto-optical disks.

[0070] The communication device 904 includes one or more pieces of hardware (communication devices) for communicating with other devices via a wireless or wired network. The input device 905 is an input device (e.g., a keyboard, a mouse, a microphone, a switch, a button, a sensor, etc.) that receives input from the outside. The output device 906 is an output device (e.g., a display, a speaker, an LED lamp, etc.) that outputs to the outside. Note that the input device 905 and the output device 906 may be integrated into one device (e.g., an input / output device such as a touch panel display).

[0071] The bus B is commonly connected to the above components and transmits, for example, address signals, data signals, and various control signals. The processor 901 is not limited to a CPU, and may be, for example, a DSP (Digital Signal Processor), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array).

[0072] (Wireless terminal hardware configuration) 10 is a diagram showing an example of the hardware configuration of a wireless terminal according to this embodiment. The wireless terminal 20 includes a GPS device 1001, a sensor 1002, and the like in addition to the hardware configuration of the computer 900 described in FIG.

[0073] The GPS device 1001 is a positioning device that receives positioning signals transmitted by GPS satellites and outputs position information indicating the current position of the wireless terminal 20. Note that the GPS device 1001 may be a positioning device other than a GPS. The sensor 1002 is a device that detects the movement or attitude of the wireless terminal 20, such as an acceleration sensor, a gyro sensor, or an IMU.

[0074] (supplement) The wireless quality prediction device 100, the wireless terminal 20, and the wireless base station 10 in this embodiment may be realized not only by dedicated devices but also by a general-purpose computer. In this case, a program for realizing this function may be recorded on a computer-readable recording medium, and the program recorded on this recording medium may be read into a computer system and executed to realize the function. Note that the term "computer system" here includes an OS (Operating System) and hardware such as peripheral devices.

[0075] Additionally, "computer-readable recording media" includes various storage devices such as portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as hard disks built into computer systems. Furthermore, "computer-readable recording media" may also include devices that dynamically store programs for a short period of time, such as communication lines when transmitting programs via networks such as the Internet or communication lines such as telephone lines, and devices that store programs for a certain period of time, such as volatile memory within computer systems that serve as servers or clients in such cases.

[0076] Furthermore, the above program may be one that realizes part of the above-mentioned functions, or may be one that can realize the above-mentioned functions in combination with a program already recorded in a computer system, or may be one that is realized using hardware such as a PLD (Programmable Logic Device) or FPGA (Field Programmable Gate Array).

[0077] <Effects of the embodiment> According to this embodiment, it is possible to predict the wireless quality of the wireless terminal 20 after a predetermined time has elapsed without creating a wireless quality distribution. This allows the wireless terminal 20 to start handover before the wireless quality deteriorates, thereby reducing, for example, temporary deterioration of communication quality due to a sudden deterioration of the wireless quality.

[0078] Furthermore, the wireless quality learning device 102 according to this embodiment eliminates the need to measure throughput for each prediction area and create a wireless quality distribution (heat map).

[0079] <Summary of the embodiment> This specification discloses at least the wireless quality prediction device, wireless quality prediction method, and program described in the following sections. (Section 1) A wireless quality prediction device that predicts wireless quality between a wireless base station and a wireless terminal that performs wireless communication with the wireless base station, a received power learning device configured to predict a predicted value of received power of the wireless terminal after a predetermined time has elapsed, based on an estimated value of a terminal position and a terminal state of the wireless terminal after the predetermined time has elapsed; a wireless quality learning device configured to predict a predicted value of the wireless quality after the predetermined time has elapsed, based on information on the wireless base station, information on the wireless communication, information on the wireless terminal, and a predicted value of the received power; a notification unit configured to notify the wireless terminal of the predicted value of the wireless quality after the predetermined time has elapsed; A wireless quality prediction device comprising: (Section 2) The received power learning device is a first neural network that has been trained in advance using current terminal positions and terminal states of other wireless terminals that communicate wirelessly with the wireless base station, and past received power of the other wireless terminals from a predetermined time ago as feature quantities, and the current received power of the other wireless terminals as training data; inputting the estimated terminal position and terminal state of the wireless terminal after the predetermined time has elapsed, and the received power of the wireless terminal, into the first neural network, and predicting a predicted value of the received power of the wireless terminal after the predetermined time has elapsed; 2. The wireless quality prediction device according to claim 1. (Section 3) The wireless quality prediction device described in claim 2, wherein the received power learning device periodically learns the first neural network using the current terminal position and terminal state of other wireless terminals that communicate wirelessly with the wireless base station, the current received power of the other wireless terminals, and the past received power of the wireless terminals up to the predetermined time before. (Section 4) The wireless quality learner is a second neural network that has been trained in advance to predict communication quality of the wireless communication using feature quantities including information about the wireless base station, information about the wireless communication, information about other wireless terminals that perform wireless communication with the wireless base station, interference wave information about the other wireless terminals, and received power of the other wireless terminals; inputting information about the wireless base station, information about the wireless communication, information about the wireless terminal, information about interference waves measured by the wireless terminal, and a predicted value of received power of the wireless terminal after the predetermined time has elapsed into the second neural network, and predicting a predicted value of the wireless quality after the predetermined time has elapsed; 4. A wireless quality prediction device according to any one of claims 1 to 3. (Section 5) the wireless quality includes a throughput value of the wireless communication; the information about the wireless communication includes a maximum throughput value of the wireless communication; the interference wave information of the wireless terminal includes an SINR value of the wireless communication measured by the wireless terminal; The terminal state of the wireless terminal includes the orientation or velocity of the wireless terminal. 2. The wireless quality prediction device according to claim 1. (Section 6) A wireless quality prediction method for predicting wireless quality between a wireless base station and a wireless terminal that performs wireless communication with the wireless base station, comprising: a process in which a learning device for received power predicts a predicted value of received power of the wireless terminal after a predetermined time has elapsed, based on estimated values ​​of a terminal position and a terminal state of the wireless terminal after the predetermined time has elapsed; a process in which a learning device of wireless quality predicts a predicted value of the wireless quality after the predetermined time has elapsed based on information of the wireless base station, information of the wireless communication, information of the wireless terminal, and the predicted value of the received power; a notification unit notifying the wireless terminal of the predicted value of the wireless quality after the predetermined time has elapsed; A method for predicting wireless quality, comprising: (Section 7) a computer that predicts wireless quality between a wireless base station and a wireless terminal that performs wireless communication with the wireless base station, a received power learning device configured to predict a predicted value of received power of the wireless terminal after a predetermined time has elapsed, based on an estimated value of a terminal position and a terminal state of the wireless terminal after the predetermined time has elapsed; a wireless quality learning device configured to predict a predicted value of the wireless quality after the predetermined time has elapsed, based on information on the wireless base station, information on the wireless communication, information on the wireless terminal, and a predicted value of the received power; a notification unit configured to notify the wireless terminal of the predicted value of the wireless quality after the predetermined time has elapsed; A program that functions as a

[0080] Although the present embodiment has been described above, the present invention is not limited to such a specific embodiment, and various modifications and changes are possible within the scope of the gist of the present invention described in the claims. [Explanation of symbols]

[0081] 1. Wireless communication systems 10. Radio base station 20 Wireless terminal 100 Wireless quality prediction device 101 Received power learning module 102 Wireless Quality Learner 103 Notification Department 201 The First Neural Network 401 Second Neural Network 900 Computers

Claims

1. A wireless quality prediction device that predicts wireless quality between a wireless base station and a wireless terminal that performs wireless communication with the wireless base station, a first neural network that has been trained in advance using as training data the current terminal position and terminal state of another wireless terminal that wirelessly communicates with the wireless base station, and past received power of the other wireless terminal from a predetermined time ago as feature quantities, and the current received power of the other wireless terminal, and that is configured to input estimated values ​​of the terminal position and terminal state of the wireless terminal after the predetermined time has elapsed, and the received power of the wireless terminal, into the first neural network, and predict a predicted value of the received power of the wireless terminal after the predetermined time has elapsed; a wireless quality learning device including a second neural network that has been trained in advance to predict communication quality of the wireless communication, using information about the wireless base station, information about the wireless communication, information about other wireless terminals that perform wireless communication with the wireless base station, interference wave information about the other wireless terminals, and received power of the other wireless terminals as feature quantities, and that is configured to input the information about the wireless base station, the information about the wireless communication, information about the wireless terminal, the interference wave information measured by the wireless terminal, and a predicted value of received power of the wireless terminal after the predetermined time has elapsed into the second neural network, and predict the predicted value of the wireless quality after the predetermined time has elapsed; a notification unit configured to notify the wireless terminal of the predicted value of the wireless quality after the predetermined time has elapsed; A wireless quality prediction device comprising:

2. 2. The wireless quality prediction device according to claim 1, wherein the received power learning device periodically learns the first neural network using the current terminal positions and terminal states of other wireless terminals that communicate wirelessly with the wireless base station, the current received power of the other wireless terminals, and the past received power of the wireless terminals up to the predetermined time before.

3. the wireless quality includes a throughput value of the wireless communication; the information about the wireless communication includes a maximum throughput value of the wireless communication; the interference wave information of the wireless terminal includes an SINR value of the wireless communication measured by the wireless terminal, The terminal state of the wireless terminal includes the orientation or velocity of the wireless terminal. The wireless quality prediction device according to claim 1 .

4. a wireless quality prediction device for predicting wireless quality between a wireless base station and a wireless terminal performing wireless communication with the wireless base station, a process of inputting estimated values ​​of the terminal position and terminal state of another wireless terminal that wirelessly communicates with the wireless base station, and past received power of the other wireless terminal from a predetermined time ago as feature quantities, into a first neural network that has been trained in advance using the current received power of the other wireless terminal as training data, and predicting a predicted value of the received power of the wireless terminal from the predetermined time after the predetermined time has elapsed; a process of inputting the information about the wireless base station, the information about the wireless communication, the information about the wireless terminal, the interference wave information measured by the wireless terminal, and a predicted value of the received power of the wireless terminal after the predetermined time has elapsed into a second neural network that has been trained in advance to predict communication quality of the wireless communication, using information about the wireless base station, information about the wireless communication, information about other wireless terminals that wirelessly communicate with the wireless base station, information about interference waves of the other wireless terminals, and received power of the other wireless terminals as feature quantities, and predicting the predicted value of the wireless quality after the predetermined time has elapsed; a notification unit notifying the wireless terminal of the predicted value of the wireless quality after the predetermined time has elapsed; A wireless quality prediction method for performing the above.

5. a computer for predicting wireless quality between a wireless base station and a wireless terminal that performs wireless communication with the wireless base station, a first neural network that has been trained in advance using as training data the current terminal position and terminal state of another wireless terminal that wirelessly communicates with the wireless base station, and past received power of the other wireless terminal from a predetermined time ago as feature quantities, and the current received power of the other wireless terminal, and that is configured to input estimated values ​​of the terminal position and terminal state of the wireless terminal after the predetermined time has elapsed, and the received power of the wireless terminal, into the first neural network, and predict a predicted value of the received power of the wireless terminal after the predetermined time has elapsed; a wireless quality learning device including a second neural network that has been trained in advance to predict communication quality of the wireless communication, using information about the wireless base station, information about the wireless communication, information about other wireless terminals that perform wireless communication with the wireless base station, interference wave information about the other wireless terminals, and received power of the other wireless terminals as feature quantities, and that is configured to input the information about the wireless base station, the information about the wireless communication, information about the wireless terminal, the interference wave information measured by the wireless terminal, and a predicted value of received power of the wireless terminal after the predetermined time has elapsed into the second neural network, and predict the predicted value of the wireless quality after the predetermined time has elapsed; a notification unit configured to notify the wireless terminal of the predicted value of the wireless quality after the predetermined time has elapsed; A program that functions as a

Images