Information processing device, communication system, information processing method, and program
The system addresses inaccurate radio quality predictions by comparing real-time measurements with past data and adapting control strategies, ensuring stable wireless communication by retraining models or switching prediction methods when necessary.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- NT T INC
- Filing Date
- 2024-11-26
- Publication Date
- 2026-06-04
AI Technical Summary
Existing radio quality prediction methods for autonomous driving and robot remote operation are inaccurate due to environmental changes during operation, leading to potential failures in wireless communication control.
A system that compares real-time measured wireless quality values with past performance data and adjusts control strategies based on the comparison results, including model retraining and switching to more accurate prediction methods when discrepancies exceed a threshold.
Ensures accurate and timely control of wireless communication by accounting for environmental changes, preventing degradation and maintaining stable communication links.
Smart Images
Figure JP2024041795_04062026_PF_FP_ABST
Abstract
Description
Information Processing Apparatus, Communication System, Information Processing Method, and Program
[0001] The present invention relates to a technique for predicting radio quality.
[0002] In order to operate autonomous driving or robot remote operation, stable communication using radio is required. For stable use of wireless communication, there is a method of predicting radio quality (which may also be called wireless communication quality) and performing control based on the predicted value. Radio quality prediction can be performed, for example, based on past performance such as past radio quality. However, accurate prediction is difficult without considering not only past performance but also environmental changes during operation, and quality prediction / control taking into account changes from past performance is required.
[0003] As a conventional technique related to radio quality prediction, for example, there are techniques disclosed in Non-Patent Documents 1 and 2. In these conventional techniques, a machine learning model (e.g., neural network) learned using pre-measured performance data is used to predict the radio quality at the location of the position information or the future position from the position information and environmental information, and utilize it for control. However, data measured in the past is used as the data for quality prediction. Therefore, due to changes in the radio propagation environment during operation compared to the past radio propagation environment, the prediction result may be significantly different from the actual one.
[0004] Kenichi Kawamura, et al,. "Predictive QoS of wireless communication for autonomous driving vehicles by fine tuning" PIRMC2023, Toronto, ON, Canada, pp. 1-6, Sep 2023. Naoki Shibuya, et al. "A study on a method for judging the risk of radio quality degradation using multiple predictions". General Assembly 2023, May 2023.
[0005] To predict wireless quality, historical wireless quality data from various locations and times is collected as training data for pre-training, and this data is used to train the model. Therefore, if the radio wave propagation environment during operation changes from the radio wave propagation environment in the training data, the predicted value may be significantly inaccurate, potentially preventing proper control for wireless communication. To improve prediction accuracy, it is necessary to consider changes in the radio wave propagation environment between the pre-training period and the operation period when performing predictions and control.
[0006] This invention has been made in view of the above points, and aims to provide a technology that enables appropriate control for wireless communication.
[0007] According to the disclosed technology, an information processing device is provided which includes a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality and notifies the terminal's control unit of the comparison result, and a prediction calculation unit that calculates a predicted value for future wireless quality and notifies the control unit of the predicted value, and in the comparison result, if the wireless quality indicated by the measured value has deteriorated by a threshold or more compared to the past performance, the control unit is provided to execute control to avoid deterioration.
[0008] The disclosed technology provides a technology that enables proper control for wireless communication.
[0009] This is a diagram illustrating the problem. This is a diagram of the communication system configuration in Example 1. This is a diagram showing the processing flow of the communication system in Example 1. This is a diagram of the communication system configuration in Example 2. This is a diagram showing the processing flow of the communication system in Example 2. This is a detailed diagram of the communication system in Example 2. This is a diagram of the communication system configuration in Example 3. This is a diagram showing the processing flow of the communication system in Example 3. This is a diagram showing an example of the hardware configuration of the device.
[0010] Hereinafter, embodiments of the present invention (this embodiment) will be described with reference to the drawings. The embodiments described below are merely examples, and the embodiments to which the present invention is applied are not limited to the embodiments described below.
[0011] First, the prior art and its problems will be described in more detail below, and then the technology related to this embodiment will be described.
[0012] (Conventional Technology and its Challenges) Figure 1 shows an example of a system configuration in conventional technology. In the system shown in Figure 1, terminal 1, autonomous vehicle 2, and wireless quality prediction server 3 are shown. Measurement data (actual data) is transmitted from multiple terminals 1 to the wireless quality prediction server 3.
[0013] The wireless quality prediction server 3 trains a neural network model using historical data from numerous terminals 1. Using the trained model, the wireless quality prediction server 3 predicts, for example, the wireless quality of an autonomous vehicle 2 at a location several seconds in the future.
[0014] However, as mentioned above, the data used for quality prediction is based on data measured in the past. Therefore, the radio wave propagation environment during operation is different from the radio wave propagation environment at the time of measurement, and the prediction results may differ significantly from the actual quality. As a result, for example, there may be cases where the terminal is unable to properly perform the control necessary to continue wireless communication.
[0015] (Outline of Embodiments) Below, Embodiments 1 to 3 will be described as technologies to solve the above problems. In each embodiment, the configuration and operation of the terminal 100 that performs wireless communication and the server 200 that predicts wireless quality will be described. The terminal 100 may be a smartphone or other device held by a person, a vehicle such as an autonomous vehicle (or a terminal installed in a vehicle), or any other device. Furthermore, the configurations of the terminal 100 and server 200 shown below basically only show the configurations according to this embodiment, and general functions such as wireless communication functions and data communication functions are not shown.
[0016] (Example 1) First, Example 1 will be described.
[0017] <Example 1: Overall System Configuration> Figure 2 shows the configuration of the communication system in Example 1. As shown in Figure 2, this communication system has a terminal 100 and a server 200. Communication is possible between the terminal 100 and the server 200 via a network including a wireless section. Note that both the server 200 and the terminal 100 may be called information processing devices. Also, "terminal 100 and server 200", terminal 100, and server 200 may each be called a communication system.
[0018] <Example 1: Configuration and Operation of Terminal 100> As shown in Figure 2, terminal 100 has a measuring unit 110 and a control unit 120. Terminal 100 includes general functions for wireless communication. The operation of each part shown in Figure 2 is as follows.
[0019] The measurement unit 110 acquires "received power (RSRP (Reference Signal Received Power), RSSI (Received Signal Strength Indicator), etc.)" from the terminal 100, or "wireless parameters related to wireless quality such as RB (Resource Block) and MCS (Modulation Coding Scheme)" from the terminal 100. The values acquired by the terminal 100 are all examples of "measured values related to wireless quality."
[0020] Based on the comparison results and predicted values (future wireless quality) notified by the server 200, the control unit 120 performs actions such as switching the communication path used by the terminal 100, controlling the transmission rate during video transmission, and controlling the behavior of the terminal 100, including its speed and path.
[0021] <Example 1: Configuration and Operation of Server 200> As shown in Figure 2, Server 200 has a past performance DB (database) 210, a comparison unit 220 between past performance and measured values (referred to as the comparison unit 220), and a prediction calculation unit 230. The operation of each unit is as follows.
[0022] The historical performance DB210 stores data such as past measured values related to wireless quality acquired for pre-training, and measured values measured during operation. The stored data may include, for example, past measured values, information on the location where those measured values were taken, and information on the time when those measured values were taken. Both measured values and past measured values include, for example, received power, RB, RSRQ (Reference Signal Received Quality), throughput, etc. The historical performance DB210 can notify each functional unit of the stored data as needed.
[0023] The prediction calculation unit 230 has a pre-trained model (e.g., a machine learning model such as a neural network) necessary for calculating a future predicted value of wireless quality. In other words, the model has been pre-trained. Using this model, the prediction calculation unit 230 can calculate, for example, the future wireless quality (predicted value) at a specified location.
[0024] In addition to predictions using machine learning models, the prediction calculation unit 230 can also perform estimations (predictions) using prior radio wave propagation simulations or estimations (predictions) using estimation formulas based on measured values.
[0025] The comparison unit 220 compares past performance data from the past performance DB 210 with actual measured values (which may also be called measured values) from the measurement unit 110, and can notify the control unit 120 of the comparison results. The values used for comparison (past performance data and actual measured values) are not limited to specific ones, but for example, the following can be used as the values to be compared: RSRP, RSSI, MCS, RB, throughput, SINR (Signal-to-Noise Ratio), RSRQ, etc.
[0026] Furthermore, the past performance data used for comparison by the comparison unit 220 is not limited to individual values. The comparison unit 220 may also compare statistical data from the area surrounding the current location of the terminal 100 with the measured values, or compare past simulation results for that area with the measured values, and so on.
[0027] When the control unit 120 of terminal 100 detects that the comparison result between past performance and measured values indicates that "measured values have been measured that show a significant deterioration in wireless quality compared to past performance," it performs control based on that result (i.e., that wireless quality has deteriorated). This control may include, for example, network switching or video rate control to avoid deterioration. Furthermore, "significant deterioration" means, for example, that the difference between past performance and measured values is greater than or equal to a threshold.
[0028] Furthermore, if the control unit 120 detects that the comparison result indicates that "the measured value is equivalent to past performance," it will perform control using the predicted value notified by the prediction calculation unit 230.
[0029] Alternatively, the server 200 may decide which value to use for control and notify the terminal 100 of the decision result.
[0030] Furthermore, in Embodiment 1, the functional divisions of the terminal 100 and the server 200 are as shown in Figure 2, but the functional divisions are not limited to those shown in Figure 2. For example, one, two, or all of the past performance DB 210, comparison unit 220, and prediction calculation unit 230 may be provided in the terminal 100.
[0031] <Example 1: Processing Flow> The processing flow of the communication system in Example 1 will be explained with reference to Figure 3.
[0032] In S101, the measurement unit 110 performs a measurement and transmits the measured value related to wireless quality to the comparison unit 220.
[0033] In S102, the comparison unit 220 compares the measured value with past performance data obtained from the past performance DB 210. The past performance data to be compared may be, for example, past performance data from a past time (time within a day) at the same location as the terminal 100 where the measured value was taken, or past measurement values from past measurement points near the location of the terminal 100 where the measured value was taken. Furthermore, the past performance data to be compared is not limited to the measurement result of a single point, but may also be the average value or median of multiple points within a certain range, or other values obtained by performing statistical processing. In addition, multiple conditions may be set, such as being near the current location (coordinates) and the time being close to the current time.
[0034] For example, the comparison unit 220 calculates the relationship between past performance and the measured value, as well as the degree of deterioration, which indicates how much the actual value has deteriorated compared to past performance. The degree of deterioration may be expressed as the difference between the actual value and past performance, or as a ratio between the actual value and past performance.
[0035] In S103, the comparison unit 220 determines whether the wireless quality shown by the measured value has deteriorated by a threshold or more compared to past performance. If the determination result is Yes, the process proceeds to S104; otherwise, the process proceeds to S106.
[0036] In S104, the comparison unit 220 notifies the control unit 120 that there is a high probability that the wireless quality has deteriorated. In S105, based on this notification, the control unit 120 selects a control method that takes into account the deterioration of the wireless quality and performs the control. Specifically, for example, it switches the connected network to avoid deterioration. Alternatively, the comparison unit 220 may notify the control unit 120 of the comparison results (such as the difference between the actual value and past performance), and the control unit 120 may determine that there is a high probability that the wireless quality has deteriorated.
[0037] In S106, which proceeds if there is no degradation above a threshold, the prediction calculation unit 230 notifies the control unit 120 of the predicted value of the wireless quality. In S107, the control unit 120 selects a control method based on the predicted value notified by the prediction calculation unit 230 and performs the control. In this example, the terminal 100 performs the control, but the server 200 may also perform the control.
[0038] As described above, in Example 1, immediately before control based on predicted values is implemented, the comparison unit 220 compares the measured value with past performance and calculates the magnitude relationship between the measured value and past performance, as well as the magnitude of the error between the measured value and past performance.
[0039] If the calculated error is in the direction of degradation and is greater than or equal to a threshold, the control unit 120 performs control to avoid degradation before using the predicted value. The threshold can be set arbitrarily. Control to avoid degradation may include, for example, "communication path switching, transmission rate control, speed / path control, etc." performed by the terminal 100, or control performed by the server 200.
[0040] <Effects of Example 1> The technology according to Example 1 makes it possible to implement control measures to avoid degradation even when the operating environment of the system changes significantly compared to the environment of past performance, resulting in a deterioration of prediction accuracy.
[0041] (Example 2) Next, Example 2 will be described. It is assumed that the function of Example 2 is an additional function to the function of Example 1. However, the function of Example 2 may not include the function of Example 1.
[0042] <Example 2: Overall System Configuration> Figure 4 shows the configuration of the communication system in Example 2. As shown in Figure 4, this communication system has a terminal 100 and a server 200. Communication is possible between the terminal 100 and the server 200 via a network including a wireless section. Note that both the server 200 and the terminal 100 may be called information processing devices. Also, "terminal 100 and server 200", terminal 100, and server 200 may each be called a communication system.
[0043] <Example 2: Configuration and Operations of Terminal 100> As shown in FIG. 4, terminal 100 includes a measurement unit 110 and a control unit 120. Note that terminal 100 includes general functions for performing wireless communication. The operations of each unit shown in FIG. 4 are as follows.
[0044] The measurement unit 110 acquires "received power (RSRP, RSSI, etc.)" in terminal 100 or "radio parameters related to radio quality such as RB and MCS" in terminal 100. Any value acquired by terminal 100 is an example of an "actually measured value related to radio quality".
[0045] The control unit 120 performs operations such as switching the communication path through which terminal 100 is communicating, controlling the transmission rate during video transmission, and controlling the behavior of terminal 100 such as speed and path, based on the comparison result and predicted value (future radio quality) notified from server 200.
[0046] <Example 2: Configuration and Operations of Server 200> As shown in FIG. 4, server 200 includes a past performance DB 210, a comparison unit for past performance and actually measured values (referred to as comparison unit 220), and a prediction calculation unit 230. The operations of each unit are as follows.
[0047] The past performance DB 210 stores data such as actually measured values related to radio quality acquired for pre-learning and actually measured values measured during operation. The data to be stored may include, for example, actually measured values, information on the location where the actually measured value was measured, and information on the time when the actually measured value was measured. Both the actually measured value and the past actually measured value are, for example, received power, RB, RSRQ (Reference Signal Received Quality), throughput, etc. The past performance DB 210 can notify the stored data to each functional unit as needed.
[0048] The prediction calculation unit 230 has a pre-trained model (e.g., a machine learning model such as a neural network) necessary for calculating a future predicted value of radio quality. That is, it is assumed that pre-learning of the model has been performed. The prediction calculation unit 230 can use the model to calculate, for example, the future radio quality (predicted value) at a specified location.
[0049] In addition to predictions using machine learning models, the prediction calculation unit 230 can also perform estimations (predictions) using prior radio wave propagation simulations or estimations (predictions) using estimation formulas based on measured values.
[0050] The comparison unit 220 compares past performance data from the past performance DB 210 with actual measured values from the measurement unit 110 and can notify the control unit 220 of the comparison results. The values used for comparison (past performance values and actual measured values) are the same as in Example 1.
[0051] Furthermore, the past performance data used for comparison by the comparison unit 220 is not limited to individual values. The comparison unit 220 may also compare statistical data from the area surrounding the current location of the terminal 100 with the measured values, or compare past simulation results for that area with the measured values, and so on.
[0052] In Example 2, if the comparison unit 220 finds a discrepancy between past performance and actual values that exceeds a threshold range, time, or number of data points, it sends the actual data and a notification (instruction) to the prediction calculation unit 230 for retraining. The prediction calculation unit 230 then retrains the model using the actual data based on this notification.
[0053] The "range" mentioned above may refer to a range of location or a range of time. "Confirming a discrepancy" means, for example, that the magnitude of the difference between past performance and the measured value exceeds a threshold.
[0054] Regarding the retraining performed by the prediction calculation unit 230, it is performed in both cases: when the measured value shows good wireless quality compared to past performance, and when the measured value shows degraded wireless quality compared to past performance. Furthermore, the timing of the actual retraining is not limited to a specific timing.
[0055] Furthermore, during model retraining, for example, the prediction calculation unit 230 inputs data (e.g., location and time) to the model and adjusts the model's parameters using methods such as backpropagation so that the error between the model's output for that data and the correct data (e.g., wireless quality at that location in the future (a few seconds later)) is minimized.
[0056] Furthermore, in Embodiment 2, the functional divisions of the terminal 100 and the server 200 are as shown in Figure 4, but the functional divisions are not limited to those shown in Figure 4. For example, one, two, or all of the past performance DB 210, comparison unit 220, and prediction calculation unit 230 may be provided in the terminal 100.
[0057] <Example 2: Processing Flow> The processing flow of the communication system in Example 2 will be explained with reference to Figure 5.
[0058] In S201, the measurement unit 110 performs a measurement and transmits the measured value related to wireless quality to the comparison unit 220.
[0059] In S202, the comparison unit 220 compares the measured value with past performance data obtained from the past performance DB 210. The past performance data to be compared may be, for example, past performance data from a past time (time within a day) at the same location as the terminal 100 where the measured value was taken, or past measurement values from past measurement points near the location of the terminal 100 where the measured value was taken. Furthermore, the past performance data to be compared may not be limited to the measurement result of a single point, but may also be the average value or median of multiple points within a certain range, or other values obtained by performing statistical processing. In addition, multiple conditions may be set, such as being near the current location (coordinates) and the time being close to the current time.
[0060] For example, the comparison unit 220 calculates the magnitude relationship (difference) between past performance and the measured value, as well as the degree of deterioration, which indicates how much the actual value has deteriorated compared to past performance. The degree of deterioration may be expressed as the difference between the actual value and past performance, or as a ratio between the actual value and past performance.
[0061] In S203, the comparison unit 220 determines whether there is a discrepancy between the measured value and past performance, for example, over a period exceeding a threshold. If the determination result is Yes, the process proceeds to S204; otherwise, it proceeds to S206.
[0062] In S204, the comparison unit 220 determines that the actual environment has changed from the environment in past performance and notifies the prediction calculation unit to retrain the model. In S205, the prediction calculation unit 230 retrains the model and updates the model.
[0063] In S206, if it is determined that there is no deviation exceeding the threshold, the same process as in Example 1 is performed.
[0064] In other words, in Embodiment 2, immediately before control based on predicted values is implemented, the comparison unit 220 compares the actual value with past performance and calculates the magnitude relationship between the measured value and past performance, and the magnitude of the error between the measured value and past performance. If the calculation results show that the error is greater than or equal to a threshold, or occurs over a period greater than or equal to a threshold, the prediction calculation unit 230 is instructed to relearn. The magnitude of the threshold can be set arbitrarily.
[0065] <Detailed System Configuration Example in Example 2 (Including Example 1)> Figure 6 shows a detailed configuration example of the communication system in Example 2 (Including Example 1). As shown in Figure 6, the terminal 100 has a measurement unit 110, a control unit 120, and a location information acquisition unit 130. The server 200 has a past performance DB 210, a comparison unit 220, a prediction calculation unit 230, a location information collection unit 240, a measurement information collection unit 250, and a control instruction unit 260.
[0066] The measurement unit 110, control unit 120, past performance DB 210, comparison unit 220, and prediction calculation unit 230 each have the functions already described.
[0067] In the examples above, the location information acquisition unit 130 has been described as being included in the measurement unit 110. The location information acquisition unit 130 acquires the location information of the terminal 100 and transmits the location information to the server 200.
[0068] In previous examples, the location information collection unit 240 and the measurement information collection unit 250 are assumed to be included in, for example, the comparison unit 220 or the past performance DB 210. Also, the control instruction unit 260 is assumed to be included in the comparison unit 220 or the prediction calculation unit 230.
[0069] The location information collection unit 240 collects location information from the terminal 100. The measurement information collection unit 250 collects measurement information from the terminal 100. The control instruction unit 260 issues control instructions to the terminal 100.
[0070] <Effects of Example 2> With the technology of Example 2, even if the operating environment of the system changes significantly compared to the environment of past performance, and as a result the accuracy of prediction deteriorates, the machine learning model used for prediction can be updated to operate a model that follows the environmental changes. As a result, appropriate control can be implemented.
[0071] (Example 3) Next, Example 3 will be described. Note that Example 1 and Example 3 may be combined, or Example 2 and Example 3 may be combined.
[0072] <Example 3: Overall System Configuration> Figure 7 shows the configuration of the communication system in Example 3. As shown in Figure 7, this communication system has a terminal 100 and a server 200. Communication is possible between the terminal 100 and the server 200 via a network including a wireless section. Note that both the server 200 and the terminal 100 may be called information processing devices. Also, "terminal 100 and server 200", terminal 100, and server 200 may each be called a communication system.
[0073] <Example 2: Configuration and Operation of Terminal 100> As shown in Figure 7, terminal 100 has a measuring unit 110 and a control unit 120. Terminal 100 includes general functions for wireless communication. The operation of each part shown in Figure 7 is as follows.
[0074] The measurement unit 110 acquires the "received power (RSRP, RSSI, etc.)" from the terminal 100, or "wireless parameters related to wireless quality such as RB and MCS" from the terminal 100. The values acquired by the terminal 100 are all examples of "measured values related to wireless quality".
[0075] Based on the predicted values (future wireless quality) notified by the server 200, the control unit 120 performs actions such as switching the communication path used by the terminal 100, controlling the transmission rate during video transmission, and controlling the behavior of the terminal 100, including its speed and path.
[0076] <Example 3: Configuration and Operation of Server 200> As shown in Figure 7, Server 200 has a comparison unit 225 (referred to as the comparison unit 225) for comparing predicted and measured values, and a prediction calculation unit 230. The operation of each unit is as follows.
[0077] The prediction calculation unit 230 has a pre-trained model (e.g., a machine learning model such as a neural network) necessary for calculating a predicted value of wireless quality. In other words, the model is assumed to have been pre-trained. The prediction calculation unit 230 uses this model to calculate, for example, the future wireless quality (predicted value) at a specified location. In addition to prediction using a machine learning model, the prediction calculation unit 230 can also perform estimation (prediction) using prior radio wave propagation simulations or estimation (prediction) using estimation formulas from measured values. Furthermore, the prediction calculation unit 230 can also calculate the current wireless quality as a predicted value.
[0078] The comparison unit 225 compares the current predicted value from the prediction calculation unit 230 (e.g., predicted value at the current time and current location) with the measured value from the measurement unit 110 and calculates the extent of the discrepancy. The discrepancy is, for example, the magnitude of the difference between the predicted value and the measured value.
[0079] The values used for comparison (predicted and measured values) are not limited to specific ones, but for example, the following can be used as values: RSRP, RSSI, MCS, RB, throughput, SINR, RSRQ, etc.
[0080] If the comparison unit 225 finds a discrepancy between the predicted value and the actual value within a range exceeding a threshold, within a time period exceeding a threshold, or within a number of data points exceeding a threshold, it receives predicted values from other prediction methods from the prediction calculation unit 230, selects the prediction method with the smallest discrepancy, and notifies the prediction calculation unit 230 of that prediction method. The prediction calculation unit 230 uses the selected prediction method to calculate future predicted values and notifies the control unit 120 of the terminal 100 of these predicted values.
[0081] The "range" mentioned above may refer to a range of location or a range of time. "Confirming a deviation" means, for example, that the magnitude of the difference between the current predicted value and the actual value exceeds a threshold.
[0082] For example, let's call a prediction method using a machine learning model Prediction Method A, a prediction method based on radio wave propagation simulation Prediction Method B, and a prediction method using an estimation formula Prediction Method C.
[0083] The comparison unit 225 detects a discrepancy between the predicted value (current) and the actual value obtained by prediction method A for a period exceeding a threshold. If the comparison unit 225 detects that the discrepancy between the predicted value (current) and the actual value obtained by prediction method B is the smallest among prediction methods A to C, it instructs the prediction calculation unit 230 to use prediction method B.
[0084] In Embodiment 3, the functional divisions of the terminal 100 and the server 200 are as shown in Figure 7, but the functional divisions are not limited to those shown in Figure 7. For example, one or all of the comparison unit 225 and the prediction calculation unit 230 may be provided in the terminal 100.
[0085] <Example 3: Processing Flow> The processing flow of the system in Example 3 will be explained with reference to the flowchart in Figure 8.
[0086] In S301, the measurement unit 110 performs a measurement and transmits the measured value related to wireless quality to the comparison unit 225.
[0087] In S302, the comparison unit 225 calculates the discrepancy (e.g., difference) between the predicted value and the measured value. The predicted value to be used for comparison is, for example, the predicted value at the same location as the terminal 100 where the measured value was taken, and at the same time as the measured value was taken.
[0088] In S303, the comparison unit 225 determines whether there is a discrepancy of more than a threshold between the measured value and the predicted value. If the determination result is Yes, the process proceeds to S304; otherwise, it proceeds to S306.
[0089] In S304, the comparison unit 225 obtains the predicted values for each of the multiple prediction methods from the prediction calculation unit 230 and compares each predicted value with the measured value. In S305, the comparison unit 225 selects the prediction method that yielded the prediction value closest to the measured value and instructs the prediction calculation unit 230 to use that prediction method. From this point onward, the prediction calculation unit 230 uses that prediction method.
[0090] In S306, the prediction calculation unit 230 notifies the control unit 120 of future predicted values, and the control unit 120 performs control based on the predicted values from the prediction calculation unit 230.
[0091] In other words, in Embodiment 3, immediately before control based on the predicted value is implemented, the comparison unit 220 compares the predicted value with the measured value and calculates the magnitude relationship and the magnitude of the error. If the calculated result shows that the error exceeds a threshold for the prediction method currently in use, the error is also calculated for the predicted values of another prediction method, and the prediction method with the smallest error is selected. The magnitude of the threshold can be set arbitrarily.
[0092] <Effects of Example 3> With the technology of Example 3, even if the operating environment of the system changes significantly compared to the environment of past performance, and as a result the accuracy of a particular prediction method deteriorates, the optimal method can be selected by switching to another prediction method, and as a result, correct control can be performed.
[0093] (Example Hardware Configuration) Any of the devices described in this embodiment (terminals, servers, information processing devices, or communication systems, etc.) can be realized, for example, by having a computer run a program. This computer may be a physical computer or a virtual machine on the cloud.
[0094] In other words, the device can be realized by using hardware resources such as the CPU and memory built into a computer to execute a program corresponding to the processing performed by the device. The program can be recorded on a computer-readable recording medium (such as portable memory), saved, and distributed. It can also be provided via a network, such as the Internet or email.
[0095] Figure 9 shows an example of the hardware configuration of the computer described above. The computer in Figure 9 has a drive device 1000, an auxiliary storage device 1002, a memory device 1003, a CPU 1004, an interface device 1005, a display device 1006, an input device 1007, an output device 1008, etc., all of which are interconnected by bus B. The computer may also be equipped with a GPU.
[0096] The program that enables processing on the computer is provided on a recording medium 1001, such as a CD-ROM or memory card. When the recording medium 1001 containing the program is set in the drive device 1000, the program is installed from the recording medium 1001 to the auxiliary storage device 1002 via the drive device 1000. However, the program does not necessarily have to be installed from the recording medium 1001; it may also be downloaded from another computer via a network. The auxiliary storage device 1002 stores the installed program as well as necessary files and data.
[0097] The memory device 1003 reads and stores a program from the auxiliary storage device 1002 when a program startup command is received. The CPU 1004 implements the functions related to the memory device 1003 according to the program stored in the memory device 1003. The interface device 1005 is used as an interface for connecting to a network, etc. The display device 1006 displays a GUI (Graphical User Interface) etc., based on a program. The input device 1007 consists of a keyboard and mouse, buttons, or a touch panel, etc., and is used to input various operation commands. The output device 1008 outputs the calculation results.
[0098] (Effects of the Embodiment) As described above, the technology described in this embodiment provides a technology that enables appropriate control for wireless communication.
[0099] The following additional information is disclosed regarding the embodiments described above.
[0100] <Notes> (Note 1) An information processing device comprising: a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality and notifies the terminal's control unit of the comparison result; and a prediction calculation unit that calculates a predicted value for future wireless quality and notifies the control unit of the predicted value, wherein, in the comparison result, the wireless quality indicated by the measured value has deteriorated by more than a threshold compared to the past performance, the control unit executes control to avoid deterioration. (Note 2) An information processing device comprising: a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality; and a prediction calculation unit that calculates a predicted value for future wireless quality using a trained model and notifies the control unit of the terminal of the predicted value, wherein, when the comparison unit confirms a discrepancy between the measured value and the past performance, the prediction calculation unit retrains the model using the measured value related to wireless quality. (Note 3) Information processing device comprising: a prediction calculation unit that calculates future predicted values related to wireless quality using a specific prediction method from among multiple prediction methods and notifies the terminal's control unit of the predicted values; and a comparison unit that compares measured values related to wireless quality with current predicted values calculated by the prediction calculation unit using the specific prediction method, wherein the comparison unit confirms a discrepancy between the measured values and the current predicted values, and the prediction calculation unit changes the prediction method used from the specific prediction method to another prediction method. (Note 4) Communication system comprising: a control unit; a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality and notifies the control unit of the comparison result; and a prediction calculation unit that calculates future predicted values for wireless quality and notifies the control unit of the predicted values, wherein in the comparison result, if the wireless quality indicated by the measured values has deteriorated by more than a threshold compared to the past performance, the control unit executes control to avoid deterioration.(Note 5) A communication system comprising: a control unit; a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality; and a prediction calculation unit that calculates a predicted value for future wireless quality using a trained model and notifies the control unit of the predicted value, wherein if the comparison unit confirms a discrepancy between the measured value and the past performance, the prediction calculation unit retrains the model using the measured value related to wireless quality. (Note 6) A communication system comprising: a control unit; a prediction calculation unit that calculates a predicted value for future wireless quality using a specific prediction method from among a plurality of prediction methods and notifies the control unit of the predicted value; and a comparison unit that compares measured values related to wireless quality with the current predicted value calculated by the prediction calculation unit using the specific prediction method, wherein if the comparison unit confirms a discrepancy between the measured value and the current predicted value, the prediction calculation unit changes the prediction method used from the specific prediction method to another prediction method. (Appendix 7) An information processing method to be performed by an information processing device, comprising: a comparison step of comparing measured values related to wireless quality with past performance related to wireless quality and notifying the control unit of a terminal of the comparison result; and a prediction calculation step of calculating a predicted value for future wireless quality and notifying the control unit of the predicted value, wherein, in the comparison result, the wireless quality indicated by the measured value has deteriorated by a threshold or more compared to the past performance, the control unit executes control to avoid deterioration. (Appendix 8) A non-temporary storage medium storing a program for causing a computer to function as an information processing device described in any one of Appendix 1 to 3.
[0101] Although this embodiment has been described above, the present invention is not limited to this specific embodiment, and various modifications and changes are possible within the scope of the gist of the invention as described in the claims.
[0102] 100 Terminal 110 Measurement unit 120 Control unit 130 Location information acquisition unit 200 Server 210 Past performance DB 220, 225 Comparison unit 230 Prediction calculation unit 240 Location information collection unit 250 Measurement information collection unit 260 Control instruction unit 1000 Drive device 1001 Recording medium 1002 Auxiliary storage device 1003 Memory device 1004 CPU 1005 Interface device 1006 Display device 1007 Input device 1008 Output device
Claims
1. An information processing device comprising: a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality and notifies the control unit of the terminal of the comparison result; and a prediction calculation unit that calculates a predicted value for future wireless quality and notifies the control unit of the predicted value, wherein, in the comparison result, if the wireless quality indicated by the measured value has deteriorated by a threshold or more compared to the past performance, the control unit executes control to avoid deterioration.
2. An information processing device comprising: a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality; and a prediction calculation unit that calculates a predicted value for future wireless quality using a trained model and notifies the control unit of the terminal of the predicted value, wherein if the comparison unit confirms a discrepancy between the measured value and the past performance, the prediction calculation unit retrains the model using the measured value related to wireless quality.
3. An information processing device comprising: a prediction calculation unit that calculates future predicted values related to wireless quality using a specific prediction method from among multiple prediction methods and notifies the terminal's control unit of the predicted values; and a comparison unit that compares measured values related to wireless quality with current predicted values calculated by the prediction calculation unit using the specific prediction method, wherein if the comparison unit confirms a discrepancy between the measured values and the current predicted values, the prediction calculation unit changes the prediction method used from the specific prediction method to another prediction method.
4. A communication system comprising: a control unit; a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality and notifies the control unit of the comparison results; and a prediction calculation unit that calculates predicted values for future wireless quality and notifies the control unit of the predicted values, wherein, in the comparison results, if the wireless quality indicated by the measured values has deteriorated by more than a threshold compared to the past performance, the control unit executes control to avoid deterioration.
5. A communication system comprising: a control unit; a comparison unit that compares measured values related to wireless quality with past performance related to wireless quality; and a prediction calculation unit that calculates predicted values for future wireless quality using a trained model and notifies the control unit of the predicted values, wherein if the comparison unit confirms a discrepancy between the measured values and the past performance, the prediction calculation unit retrains the model using the measured values related to wireless quality.
6. A communication system comprising: a control unit; a prediction calculation unit that calculates future predicted values related to wireless quality using a specific prediction method from among a plurality of prediction methods and notifies the control unit of said predicted values; and a comparison unit that compares measured values related to wireless quality with current predicted values calculated by the prediction calculation unit using the specific prediction method, wherein if the comparison unit confirms a discrepancy between the measured values and the current predicted values, the prediction calculation unit changes the prediction method used from the specific prediction method to another prediction method.
7. An information processing method executed by an information processing device, comprising: a comparison step of comparing measured values related to wireless quality with past performance related to wireless quality and notifying the control unit of a terminal of the comparison result; and a prediction calculation step of calculating a predicted value for future wireless quality and notifying the control unit of the predicted value, wherein, in the comparison result, the wireless quality indicated by the measured value has deteriorated by a threshold or more compared to the past performance, the control unit executes control to avoid deterioration.
8. A program for causing a computer to function as an information processing device according to any one of claims 1 to 3.