Wireless use risk determination device, wireless use risk determination method, and program

Abstract

The present invention provides a wireless use risk determination device comprising: a risk determination unit configured to compare a predicted value of quality of wireless communication to a threshold value, thereby determining presence or absence of a risk of deterioration in quality related to the use of the wireless communication; and a threshold value change unit configured to change the threshold value on the basis of actual quality related to the use of the wireless communication and the result of the determination regarding the presence or absence of the risk, thereby increasing flexibility regarding the threshold value for the predicted value of the quality of the wireless communication.

Description

Wireless usage risk determination device, wireless usage risk determination method, and program

[0001] The present invention relates to a wireless usage risk determination device, a wireless usage risk determination method, and a program.

[0002] For the stable use of wireless communication, there are techniques for predicting wireless communication quality. For example, there is a technique that uses machine learning with a neural network to predict wireless communication quality at a current location and future locations based on location information and movement information using previously measured performance data (Non-Patent Document 1). There is also a technique for predicting wireless communication quality using a recurrent neural network (RNN), which performs time-series prediction based on time-series information.

[0003] When using a single quality prediction technology for wireless communication quality, there are issues such as the fact that there are degradation factors that cannot be taken into account. In other words, a single prediction method may not be suitable depending on the actual environmental conditions. For example, prediction methods based on location information have difficulty predicting wireless communication quality degradation caused by moving objects. Prediction methods based on received power have difficulty predicting wireless communication quality degradation caused by base station congestion.

[0004] Therefore, there is a prediction technique that determines the risk of future deterioration of wireless communication quality based on the results of multiple prediction methods (Non-Patent Document 2).

[0005] Naoki Shibuya et al., "A Study on Wireless Communication Quality Prediction Using Machine Learning," IEICE General University, B-6-74, Mar., 2022 Naoki Shibuya et al., "A Study on Wireless Quality Degradation Risk Assessment Method Using Multiple Predictions," IEICE General University, B-6-47, Mar., 2023

[0006] However, in Non-Patent Document 2, the threshold for the predicted value calculated from each of the multiple wireless communication quality prediction methods is uniquely determined for each method, so the threshold may not be appropriate depending on the situation of the terminal using wireless communication.

[0007] The present invention has been made in view of the above points, and has as its object to improve flexibility regarding thresholds for predicted values ​​of the quality of wireless communication.

[0008] In order to solve the above problem, the wireless usage risk assessment device has a risk assessment unit configured to determine whether or not there is a risk of deterioration in the quality of wireless communication usage by comparing a predicted value of the quality of the wireless communication with a threshold, and a threshold change unit configured to change the threshold based on the actual quality of the wireless communication usage and the determination result of whether or not there is the risk.

[0009] It allows for greater flexibility in thresholds for predicted values ​​of wireless communication quality.

[0010] FIG. 1 is a diagram showing an example of the configuration of a wireless communication system according to an embodiment of the present invention. FIG. 2 is a diagram for explaining an overview of an embodiment of the present invention. FIG. 3 is a diagram showing an example of the hardware configuration of a wireless usage risk assessment device 10 according to an embodiment of the present invention. FIG. 4 is a diagram showing an example of the functional configuration of a wireless communication system according to an embodiment of the present invention. FIG. 5 is a sequence diagram for explaining an example of a processing procedure executed in the wireless communication system. FIG. 6 is a diagram showing an example of the configuration of a threshold DB 16 in an initial state. FIG. 7 is a diagram showing an example of the configuration of a threshold DB 16 after a new predicted situation category has been added.

[0011] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will now be described with reference to the accompanying drawings. Fig. 1 is a diagram showing an example of the configuration of a wireless communication system according to an embodiment of the present invention.

[0012] In FIG. 1, one or more terminals 20 are connected to a wireless usage risk assessment device 10 via a network N1 such as the Internet.

[0013] The terminal 20 is a mobile object that uses wireless communication. For example, a robot or a car may be the terminal 20. The terminal 20 may be included in the network N1 via a wireless communication network, for example.

[0014] The wireless usage risk assessment device 10 is one or more computers that assess the risk of degradation of the quality of an application used by a terminal 20 (hereinafter referred to as "application quality") as the terminal 20 moves. If the wireless usage risk assessment device 10 determines that there is degradation in application quality, it notifies the terminal 20 of this. Note that the application used by the terminal 20 refers to an application that uses wireless communication. For an application that handles video, video quality is an example of application quality. Therefore, application quality depends on the quality of wireless communication (hereinafter referred to as "wireless quality"). Application quality is an example of quality related to the use of wireless communication.

[0015] FIG. 2 is a diagram for explaining an outline of an embodiment of the present invention.

[0016] For example, the wireless usage risk assessment device 10 determines whether there is a risk of deterioration in the quality of an application (app quality) on the terminal 20 (hereinafter referred to as "risk assessment") by comparing the predicted value of wireless quality in the future situation (e.g., future location) of the terminal 20, output by a learning device that has learned to predict wireless quality from some input (e.g., location information), with a threshold value (S1).

[0017] The predicted value and the result of the determination of the presence or absence of the risk (hereinafter referred to as the "risk determination result") are stored in the performance DB (S2). The risk determination result may also be notified to the terminal 20 to be controlled (S3). Thereafter, the risk determination result is associated with the actual application quality of the terminal 20 when the actual situation of the terminal 20 becomes the situation assumed in the risk determination (S4).

[0018] If the risk assessment result and the actual application quality are inconsistent in the information stored in the performance DB, the wireless usage risk assessment device 10 corrects the threshold so that the risk assessment result based on the threshold is reversed. Specifically, if the actual application quality deteriorates despite the assessment of "no deterioration," the wireless usage risk assessment device 10 corrects the threshold upward so that it can determine "deteriorated" for similar situations in the future (S5). On the other hand, if the actual application quality does not deteriorate despite the assessment of "deteriorated," the wireless usage risk assessment device 10 corrects the threshold downward so that it can determine "no deterioration" for similar situations in the future (S6).

[0019] The revised threshold is set for each situation for which wireless quality is predicted. When risk assessment is performed for a similar situation in the future, the threshold set for that situation will be used.

[0020] 3 is a diagram showing an example of the hardware configuration of a wireless usage risk assessment device 10 according to an embodiment of the present invention. The wireless usage risk assessment device 10 in FIG. 3 includes a drive device 100, an auxiliary storage device 102, a memory device 103, a processor 104, and an interface device 105, which are all interconnected via a bus B.

[0021] The program that realizes the processing in the wireless usage risk assessment device 10 is provided by a recording medium 101 such as a CD-ROM. When the recording medium 101 storing the program is set in the drive device 100, the program is installed from the recording medium 101 to the auxiliary storage device 102 via the drive device 100. However, the program does not necessarily have to be installed from the recording medium 101, but may be downloaded from another computer via a network. The auxiliary storage device 102 stores the installed program as well as necessary files, data, etc.

[0022] When an instruction to start the program is received, the memory device 103 reads and stores the program from the auxiliary storage device 102. The processor 104 is a CPU or a GPU (Graphics Processing Unit), or a CPU and a GPU, and executes functions related to the wireless usage risk assessment device 10 in accordance with the program stored in the memory device 103. The interface device 105 is used as an interface for connecting to a network.

[0023] The terminal 20 may also have the same hardware configuration as that shown in FIG.

[0024] 4 is a diagram showing an example of the functional configuration of a wireless communication system according to an embodiment of the present invention. In FIG. 4, a terminal 20 includes an information acquisition unit 21, an information transmission unit 22, a control determination unit 23, a control unit 24, and an application unit 25. Each of these units is realized by a process in which one or more programs installed in the terminal 20 are executed by a processor of the terminal 20.

[0025] The wireless usage risk assessment device 10 includes an information receiving unit 11, prediction units 12-1 to 12-n (n≧1) (hereinafter, referred to as the "prediction unit 12" when not distinguishing between them), a risk assessment unit 13, a quality acquisition unit 14, and a threshold change unit 15. Each of these units is realized by the processor 104 executing one or more programs installed in the wireless usage risk assessment device 10. The wireless usage risk assessment device 10 also uses databases (storage units) such as a threshold DB 16 and a performance DB 17. Each of these databases can be realized using, for example, an auxiliary storage device 102 or a storage device connectable to the wireless usage risk assessment device 10 via a network.

[0026] The following describes a processing procedure executed in the wireless communication system: Fig. 5 is a sequence diagram for explaining an example of a processing procedure executed in the wireless communication system.

[0027] For example, when the terminal 20 starts moving, the information acquisition unit 21 of the terminal 20 collects information (hereinafter referred to as "future terminal information") indicating the status of the terminal 20 at a certain time in the future (hereinafter referred to as "future terminal status") (S101). The certain time in the future refers to a time at which wireless quality is predicted (hereinafter referred to as "prediction time"), for example, a time a predetermined number of seconds after the current time. The future terminal status is information including parameter values ​​such as the terminal type, current location, future location, prediction time, application ID, and purpose. The terminal type is the type of the terminal 20 (e.g., robot, automobile, etc.). The current location is information indicating the current location of the terminal 20. The future location is information indicating the location where the terminal 20 is expected to be at the prediction time (hereinafter referred to as "location information"). The future location may be calculated from the moving direction and moving speed of the terminal 20. Note that the location information may be information indicating a pinpoint or information indicating a certain range (e.g., a cell of several meters square). The app ID is identification information for an application that the terminal 20 is scheduled to use at the time of prediction. The use is the use of the terminal 20. For example, if the terminal 20 is a robot, examples of uses include "delivering packages" and "guiding people." If there is a one-to-one relationship between an application and a use, the future terminal information does not need to include the use. Note that the terminal type and use are invariant over time. Note that the parameters by which the terminal status is identified may be changed as appropriate depending on the information required by the prediction unit 12 and the desired granularity of classification of thresholds for values ​​predicted by the prediction unit 12.

[0028] Next, the information transmission unit 22 transmits a request to the wireless usage risk assessment device 10 to determine whether there is a risk of degradation in application quality (S102). The assessment request includes the future terminal information collected in step S101. The assessment request may also include route information indicating a planned movement route of the terminal 20. If the wireless usage risk assessment device 10 knows the movement route of the terminal 20 (for example, if the terminal 20 moves under the control of the wireless usage risk assessment device 10), the assessment request does not need to include the movement route. Furthermore, other information necessary for predicting wireless quality and obtainable by the terminal 20 may also be included in the assessment request.

[0029] When the information receiving unit 11 of the wireless usage risk assessment device 10 receives the assessment request, each prediction unit 12 uses any of the parameters constituting the future terminal information included in the assessment request as input to calculate a predicted value of the future wireless quality of the terminal 20 (S103). When there are multiple prediction units 12 (n>1), the units (scales) of the wireless quality (i.e., the predicted values) may be different from each other.

[0030] For example, the prediction unit 12-1 may receive information indicating the future position and future orientation of the terminal 20 as input, and may predict the future received power by performing propagation estimation from static environmental information such as the building, layout, and obstacles using a building database, CAD information, etc. In this case, the predicted value calculated by the prediction unit 12i is the received power. Note that the future orientation of the terminal 20 may be determined from the movement path of the terminal 20, or may be determined based on the direction from the current position of the terminal 20 to the future position.

[0031] Furthermore, the prediction unit 12-2 may be a machine learning model that learns past performance data (performance data on the relationship between throughput and position, time, etc.) and predicts the throughput at a future position of the terminal 20. In this case, the predicted value calculated by the prediction unit 12-2 is the throughput.

[0032] Next, the risk determination unit 13 performs risk determination (S104) by comparing, for each prediction unit 12, the predicted value by the prediction unit 12 with a threshold stored in the threshold DB 16 that corresponds to the future device status indicated by the prediction unit 12 and the future device information. Here, the future device status refers to a status specified by a combination of parameters such as a future location, a prediction time, an application ID, and a purpose. The threshold corresponding to the prediction unit 12 and the future device status refers to a threshold associated with a category (hereinafter referred to as a "predicted situation category") to which the combination of the prediction unit 12 and the future device status belongs. The predicted situation category is a category for distinguishing thresholds used for risk determination depending on the combination of the prediction unit 12 and the device type, location, time information, application ID, and purpose (the situation assumed by the predicted value). The "location" and "time information" of the predicted situation category are compared with the future location and future time of the future device status. The finer the granularity of the predicted situation category, the more precise the correspondence between the future terminal situation and the threshold value can be, but the more resources (e.g., memory capacity) are required to manage the threshold value. Therefore, the granularity of the predicted situation category may be increased by providing a width (range) to any of the parameters that make up the predicted situation category. For example, the position of the predicted situation category may be greater in granularity than the future position of the future terminal situation. The time information of the predicted situation category may be rounded to a time zone.

[0033] Initially, the predicted situation categories are classified by the prediction unit 12. In other words, the thresholds are not classified according to the combination of "terminal type, location, time information, application ID, and purpose." Therefore, the performance DB 17 in the initial state is, for example, as shown in FIG. 6.

[0034] FIG. 6 is a diagram showing an example of the configuration of the threshold DB 16 in an initial state. In FIG. 6, the threshold DB 16 can store thresholds in association with predicted situation categories, which are combinations of the prediction unit 12, terminal type, location, time information, app ID, and use. However, in the initial state, only records containing values ​​in the prediction unit 12 column are stored, and there is a one-to-one correspondence between the prediction unit 12 and the threshold. Therefore, if there are n prediction units 12, n records are stored in the threshold DB 16. However, an initial threshold value may be set for each unit (scale) of the predicted value. In this case, the "prediction unit" column may be changed to "unit" or the like.

[0035] When the threshold DB 16 has the configuration (initial state) shown in FIG. 6 , the risk determination unit 13 performs risk determination for each predicted value using a threshold corresponding to the prediction unit 12 that outputs the predicted value. In this embodiment, the predicted value is a measure in which the larger the value, the better the application quality. Therefore, if the predicted value is equal to or greater than the threshold, the determination unit determines that there is no risk of application quality deterioration, and if the predicted value is less than the threshold, the determination unit determines that there is a risk of application quality deterioration. Note that if there is a predicted value in which the smaller the value, the better the application quality, the opposite determination may be made.

[0036] Next, the risk determination unit 13 stores the risk determination result for each prediction unit 12 in the performance DB 17 in association with the predicted value by the prediction unit 12 and the future terminal status (S105).

[0037] Next, the risk assessment unit 13 transmits the risk assessment result to the terminal 20 (S106). The risk assessment result transmitted to the terminal 20 may be the risk assessment result for each prediction unit 12, or may be a risk assessment result that combines the risk assessment results for each prediction unit 12. In the latter case, for example, the number or percentage of "risk present" assessment results may be used as the combined assessment result. Furthermore, the predicted value may be transmitted to the terminal 20 together with the risk assessment result. The control determination unit 23 of the terminal 20 transmits a control instruction based on the received risk assessment result to the control unit 24. For example, a control instruction such as connection switching, communication parameter change, or driving route change is transmitted to the control unit 24. The control unit 24 executes control in accordance with the control instruction to ensure a stable communication environment.

[0038] Thereafter, the information transmitting unit 22 of the terminal 20 constantly (e.g., periodically) transmits information indicating the current terminal status (hereinafter referred to as "current terminal status") to the wireless usage risk assessment device 10 (S107). The current terminal status is information including parameter values ​​such as the terminal type, current location, current time, application ID, and purpose. The application ID is identification information for the application currently being used by the terminal 20. The information receiving unit of the wireless usage risk assessment device 10 notifies the quality acquiring unit 14 of the received current terminal status. The quality acquiring unit 14 executes step S108 if the current terminal status (terminal type, current location, current time, application ID, purpose, etc.) matches the terminal type, future location, prediction time, application ID, purpose, etc. of the future terminal status recorded in the performance DB 17 in step S105. In other words, step S108 is executed if the current status of the terminal 20 matches the future status assumed at the prediction time. Note that a certain degree of error may be allowed in determining whether or not the terminal 20 matches.

[0039] In step S108, the quality acquisition unit 14 acquires actual application quality for an application associated with an application ID included in the current terminal status (an application currently being used by the terminal 20). The method of acquiring application quality varies depending on the application. For example, in the case of an application that uploads video captured by the terminal 20 to a server (e.g., an application related to telepresence, a surveillance robot, or video related to in-house and out-of-house use during autonomous driving), the quality of the video received by the server-side application (e.g., VMAF or bit rate) may be acquired as the application quality. If application quality can be measured on the terminal 20, the application quality may be acquired from the application unit 25 of the terminal 20.

[0040] Alternatively, regardless of the application, as long as the indicator is affected by wireless quality, the indicator (e.g., delay amount) may be acquired as the application quality. The quality acquiring unit 14 associates the acquired application quality with a future terminal status corresponding to the current terminal status and stores the association in the performance DB 17. Therefore, the future terminal status is in a state in which the risk assessment result and the actual application quality are associated with each other.

[0041] At any timing thereafter, the threshold modification unit 15 determines whether or not it is necessary to modify each threshold used for each risk determination result (risk determination result for each predicted value) based on the set of the future device status (hereinafter referred to as the "target device status"), the risk determination results, and the actual application quality stored in the performance DB 17 (S109). That is, the threshold modification unit 15 determines that it is necessary to modify the threshold for a risk determination result that does not match the actual application quality among the risk determination results for each predicted value (for each prediction unit 12), and determines that it is not necessary to modify the threshold for other risk determination results. A risk determination result that does not match the actual application quality is a determination result of "no degradation" if the actual application quality has deteriorated, or a determination result of "degradation" if the actual application quality has not deteriorated. Note that the determination of whether the application quality has deteriorated may be performed based on a threshold set for each application for this determination.

[0042] When it is determined that a change in threshold is necessary for any of the risk assessment results, the threshold change unit 15 changes the threshold (S110). Specifically, the threshold change unit 15 corrects upward the threshold associated with the risk assessment result of "no degradation" when the actual application quality has deteriorated. Alternatively, the threshold change unit 15 corrects downward the threshold associated with the risk assessment result of "degraded" when the actual application quality has not deteriorated.

[0043] For example, suppose that the predicted value of the received power output by the prediction unit 12-1 is −82 dBm, and the threshold corresponding to this predicted value is −83 dBm. In this case, if the risk assessment result based on this predicted value and this threshold is “no risk,” and the actual application quality is deteriorating, the threshold change unit 15 changes the threshold to a value greater than this predicted value (for example, −81 dBm) so that the predicted value is determined to be “risky.” By doing so, −82 dBm becomes less than the threshold, and it becomes possible to determine that “risk exists.”

[0044] If the risk assessment result based on the predicted value and the threshold is "risky" and the actual application quality has not deteriorated, the threshold change unit 15 lowers the threshold to the predicted value so that the predicted value is judged to be "no risk" (for example, the predicted value is set as the threshold).

[0045] The method for changing the threshold is not limited to the above. The threshold may be changed by a predetermined change amount, or the change amount may be determined according to the degree of deviation between the threshold and the predicted value.

[0046] Next, the threshold modification unit 15 stores the modified threshold in the threshold DB 16 as the threshold for the predicted situation category to which the target terminal situation belongs (S111). At this time, if there is no record corresponding to the predicted situation category in the threshold DB 16 (i.e., if the threshold before the modification was a threshold for each prediction unit 12), the threshold modification unit 15 adds a record corresponding to the predicted situation category to the threshold DB 16 and records the modified threshold in the record.

[0047] 7 is a diagram showing an example of the configuration of the threshold DB 16 after adding a new predicted situation category. In FIG. 7, a record in the second row has been newly added. The record in the second row contains values ​​for all items that make up the predicted situation category. Therefore, the thresholds in the record in the second row correspond to the terminal type, location, time information, application ID, and purpose related to the target terminal situation in addition to the specific prediction unit 12.

[0048] Thereafter, when a future terminal situation that belongs to the predicted situation category in the second row occurs, in step S104, it is determined whether or not there is a risk of deterioration in application quality based on the threshold value in the second row.

[0049] Note that when step S109 and subsequent steps are performed in a batch manner, such as at night, and there are multiple future device situations that match the target device situation, it may be determined whether or not to change the threshold based on a majority vote regarding the consistency between the risk assessment result and the actual application quality. Specifically, the threshold change unit 15 may not change the threshold if there are more future device situations in which the risk assessment result matches the actual application quality, and may change the threshold if there are more future device situations in which the risk assessment result does not match the actual application quality.

[0050] As described above, according to the present embodiment, a threshold for determining whether or not there is a risk of degradation in application quality based on a predicted value of wireless quality can be set according to the status of the terminal 20 (location, time, application used, and purpose), etc. Therefore, flexibility in determining the threshold for the predicted value of wireless communication quality can be improved. As a result, the likelihood of accurate determination can be increased compared to, for example, a case where a unique threshold is set for each wireless quality prediction method. By improving the accuracy of determining the risk of degradation in application quality, stable use of wireless communication and applications in the terminal 20 can be expected.

[0051] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various modifications and changes are possible within the scope of the gist of the present invention described in the claims.

[0052] REFERENCE SIGNS LIST 10 Wireless usage risk assessment device 11 Information receiving unit 12 Prediction unit 13 Risk assessment unit 14 Quality acquisition unit 15 Threshold change unit 16 Threshold DB 17 Performance DB 20 Terminal 21 Information acquisition unit 22 Information transmission unit 23 Control determination unit 24 Control unit 25 Application unit 100 Drive device 101 Recording medium 102 Auxiliary storage device 103 Memory device 104 Processor 105 Interface device B Bus

Claims

1. A wireless usage risk judgment device comprising: a risk judgment unit configured to judge whether or not there is a risk of deterioration in quality related to the use of wireless communication by comparing a predicted value of the quality of the wireless communication with a threshold; and a threshold change unit configured to change the threshold based on the actual quality related to the use of the wireless communication and the judgment result of the presence or absence of the risk.

2. The wireless usage risk assessment device according to claim 1, characterized in that the threshold change unit is configured to change the threshold when the actual quality and the assessment result do not match.

3. The wireless usage risk judgment device as described in claim 1 or 2, characterized in that the threshold change unit is configured to change the threshold so that, when the changed threshold is compared with the predicted value, the judgment result of whether or not the risk is present is reversed.

4. A wireless usage risk assessment method, characterized in that a computer executes the following steps: a risk assessment procedure for determining whether or not there is a risk of deterioration in quality related to the use of wireless communication by comparing a predicted value of the quality of the wireless communication with a threshold; and a threshold change procedure for changing the threshold based on the actual quality related to the use of wireless communication and the result of the risk assessment.

5. A program that causes a computer to execute a risk determination procedure for determining whether or not there is a risk of deterioration in quality related to the use of wireless communication by comparing a predicted value of the quality of the wireless communication with a threshold, and a threshold change procedure for changing the threshold based on the actual quality related to the use of wireless communication and the result of the determination of the presence or absence of the risk.

Images