Information processing device, stability calculation method, and program

Abstract

This information processing device comprises an acquisition unit that acquires quality information relating to the quality of communication, and a calculation unit that calculates the stability of the quality on the basis of variations in the quality information.

Description

Information processing device, stability calculation method, and program

[0001] The present invention relates to a technique for performing control based on communication quality.

[0002] A conventional technique for estimating communication quality involves creating a history of quality information based on the device's actual usage of a network (NW), storing the history in a database (DB), and estimating quality by referring to the quality history in the DB in response to a request from an external function.

[0003] Furthermore, Non-Patent Document 1 discloses a technique for reducing the amount of quality information managed in a DB by converting quality information into feature quantities.

[0004] By using the quality information stored in the DB, it is possible to predict future communication quality in the controlled object and perform proactive control on the controlled object.

[0005] Non-patent document 1. Ono et al., Study on quality information management method for network quality estimation, 2024 Institute of Electronics, Information and Communication Engineers General Conference, B-6-16

[0006] When the reliability of quality prediction is low, implementing control based on the prediction result may result in further deterioration of communication quality or wasteful use of large resources. Therefore, it is necessary to estimate the reliability of quality prediction, but conventional techniques have not been able to estimate the reliability of quality prediction.

[0007] The present invention has been made in view of the above points, and has an object to provide a technique that makes it possible to estimate the reliability of prediction of communication quality.

[0008] According to the disclosed technique, an information processing device is provided that includes: an acquisition unit that acquires quality information related to communication quality; and a calculation unit that calculates stability of the quality based on a variation in the quality information.

[0009] The disclosed technology provides a technology that makes it possible to estimate the reliability of communication quality predictions.

[0010] 1 is a diagram for explaining an application area of ​​proactive control. FIG. 1 is a diagram showing an example of the configuration of a communication system in an embodiment of the present invention. FIG. 2 is a diagram showing an image of stability calculation. FIG. 3 is a diagram showing experimental results regarding improvement in quality estimation accuracy by utilizing stability. FIG. 4 is a diagram for explaining the principle of Accuracy / Recall. FIG. 5 is a diagram showing an image of optimization of an application area based on NW quality behavior and stability. FIG. 6 is a diagram showing an example of quality estimation results and cooperative control instructions. FIG. 7 is a diagram showing an example of control performed depending on the quality and stability. FIG. 8 is a diagram showing an example of an area that can be handled by proactive control. FIG. 9 is a diagram showing a processing sequence of a communication system. FIG. 10 is a diagram for explaining an example of a control policy for a mobile device based on stability S. FIG. 11 is a diagram showing an example of the configuration of an information processing device 400. FIG. 12 is a diagram showing an example of the configuration of an information processing device 500. FIG. 13 is a diagram showing an example of the hardware configuration of an apparatus.

[0011] 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.

[0012] In the following, first, the problems and application areas of proactive control will be described, and then the system configuration and operation according to this embodiment will be described.

[0013] (Problem) As described above, by using the information on communication quality (which may also be called network quality) stored in the DB, it is possible to predict the future communication quality of a device to be controlled (e.g., a connected car). Furthermore, it is possible to proactively control the device based on the prediction.

[0014] However, network quality predictions based on quality information stored in a database are not always accurate, and predictions may be wrong. If the network quality prediction is wrong, incorrect proactive control may be implemented, which may lead to a deterioration in the quality of the entire system.

[0015] Therefore, it is necessary to stabilize or improve the accuracy of proactive control in order to maintain and improve the quality of the entire system. In other words, there is a problem in that it is necessary to clarify the applicability of proactive control by clarifying the reliability of quality prediction.

[0016] (Causes of Error in Future Prediction and Application Area of ​​Proactive Control) Here, we will explain the causes of error in prediction of future network quality (future prediction) and application area of ​​proactive control. Note that the term "quality estimation" used in the following explanation includes quality prediction (estimating (predicting) future quality).

[0017] Reasons why future predictions may be inaccurate include, for example, that "events that have not occurred in the past cannot be predicted" and "in a 'low-level environment' where no regularity or stationarity is observed, it is difficult to improve the accuracy of predictions based on statistical information."

[0018] The application area of ​​proactive control based on future predictions is where it is possible to guarantee the accuracy of quality estimation (limiting the uncertainty of future predictions to a certain level). In areas where uncertainty in future predictions remains, a "reactive response + efficient advance measures" is required.

[0019] As an example of factors that affect network quality, Figure 1(a) shows a plane divided into four quadrants with "changes in the radio wave propagation environment" and "changes in traffic congestion status" on the vertical and horizontal axes.

[0020] In Fig. 1(a), for example, the first quadrant is an area where reliable quality estimation is difficult because both "changes in the radio wave propagation environment" and "changes in the traffic congestion situation" are large. On the other hand, the third quadrant is an area where reliable quality estimation is high because both "changes in the radio wave propagation environment" and "changes in the traffic congestion situation" are small.

[0021] Figure 1(b) shows the weight of proactive control and reactive control for each quadrant. As shown in Figure 1(b), the first quadrant is an area that is mainly handled by reactive control, and the third quadrant is an area that is mainly handled by proactive control.

[0022] (Outline of the embodiment) In this embodiment, a quality stability estimation unit 110 (described later) acquires quality information continuously collected by passive measurement, calculates (estimates) the stability of quality based on the variability of the quality information, and clarifies the reliability of future prediction (= applicability of proactive control). In other words, the stability is an index showing the reliability of future prediction.

[0023] By taking into account the stability of quality in addition to conventional quality estimates to determine the content of cooperative control, including whether or not to implement proactive control, proactive control is applied only in areas where future predictions are valid, making it possible to avoid reduced resource utilization efficiency and unstable operation due to excessive control.

[0024] (System Configuration Example) Fig. 2 shows an example of the configuration of a communication system according to this embodiment. As shown in Fig. 2, the communication system according to this embodiment includes a quality stability estimation unit 110, a storage unit 120, a quality estimation unit 130, a cooperation control unit 200, and a quality information distribution unit 210. Here, the configuration including the quality stability estimation unit 110, the storage unit 120, and the quality estimation unit 130 is referred to as an information processing device 100. The storage unit 120 may also be referred to as a DB (database).

[0025] Furthermore, each of the quality stability estimation unit 110, the storage unit 120, the quality estimation unit 130, the collaboration control unit 200, and the quality information distribution unit 210 may be called an information processing device. The information processing device is, for example, a computer. The computer may be a physical machine or a computing resource on the cloud.

[0026] 2 also shows a control target 300 that is an object of control by the cooperative control unit 200. The control target 300 is, for example, a device (e.g., a connected car), a network, an edge cloud server, or the like.

[0027] The storage unit 210 stores quality information that is continuously acquired by passive measurement from the control target 300 and the like via the quality information distribution unit 210. The quality information is quality information related to communication of the control target 300. For example, if the device is a connected car, quality information (throughput, packet loss rate, delay, jitter, etc.) in information transmission between the connected car and a remote monitoring and control system and the like is continuously stored in the storage unit 120.

[0028] The quality stability estimation unit 110 refers to the quality information stored in the storage unit 120, calculates (estimates) the stability of the quality information, and stores the calculated stability in the storage unit 120. The quality stability estimation unit 110 also updates the stability, for example, periodically. The method of calculating the stability will be described in detail later.

[0029] The quality estimation unit 130 refers to the quality information and stability stored in the storage unit 120, performs quality estimation based on the quality information and stability, and notifies the cooperative control unit 200 of the quality estimation result. The quality estimation result includes a quality estimation value and stability. The quality estimation method is not limited to a specific method. For example, a percentile value is calculated from N pieces of quality information (measured values), and this value is used as the estimated value of the NW quality. Alternatively, as described in Non-Patent Document 1, a percentile value may be calculated from one quality index (such as an average value or a standard deviation) per region, and this may be used as the estimated value of the NW quality. Each of the above percentile values ​​may be determined based on stability. Alternatively, a final estimated value may be obtained by multiplying the above estimated value by a coefficient based on stability.

[0030] Furthermore, the quality estimation unit 130 may perform quality estimation using only quality information whose stability is equal to or greater than a threshold value.

[0031] The cooperative control unit 200 issues control instructions to the control target 300 based on the quality estimation result received from the quality estimator 130. The control instructions include, for example, changing the connection network, changing the monitoring system, changing the information transmission method, changing the route, changing the speed, and issuing other alert notifications.

[0032] Fig. 3 shows an image of stability calculation by the quality stability estimation unit 110. In the example of Fig. 3, as NW quality information continuously acquired by passive measurement, for example, "throughput, packet loss rate, communication delay, jitter" at each time in a certain location (or area) is stored in the storage unit 120.

[0033] In the example of FIG. 3, the quality stability estimation unit 110 estimates the stability of each of "throughput, packet loss rate, communication delay, and jitter" as S Ｔ , S Ｌ , S Ｄ , S Ｊ The quality stability estimating unit 110 may calculate (update) the stability, for example, every time the number of quality information items increases by a predetermined number.

[0034] Examples (1) to (4) of the estimation formula (calculation formula) of the stability S are shown below. In all of the following examples, the calculated value is based on the "variation of quality information."

[0035] (1) S = f １ (CV) f １ (CV)=1 / CV CV is the coefficient of variation, and is expressed as "coefficient of variation CV=standard deviation / average value." In the example of Fig. 3, both the standard deviation and the average value are values ​​for n pieces of quality information.

[0036] (2) S = f ２ (N, CV) f ２ (N, CV) = ln N / CV That is, f ２ (N, CV) is the value obtained by dividing "ln N" (the natural logarithm of N) by CV, where N is the number of quality information items (number of samples).

[0037] (3) S = f ３ (N, V) f ３ (N, V) = ln N / V, that is, f ３ (N, V) is the value obtained by dividing "ln N" (the natural logarithm of N) by V. V is the variance.

[0038] (4) S = f ４ (x) f ４ (x) = mean value / x% confidence interval In other words, f ４(x) is the value obtained by dividing the average value of the target quality information by the x% confidence interval for the quality information.

[0039] (Improving Quality Estimation Accuracy by Utilizing Stability) As described above, the quality estimation unit 130 can perform quality estimation using only quality information whose stability is equal to or greater than a threshold. More specifically, the quality estimation unit 130 can improve quality estimation accuracy by utilizing the stability estimated by the quality stability estimation unit 110. An experiment demonstrating this was conducted and is described below.

[0040] Here, quality information was stored in the storage unit 120 by implementing packet transfer (iperf, UDP) simulating real-time video transmission using commercial mobile networks (Company A and Company B), and the stability S of quality (e.g., throughput, communication delay) and quality estimation values ​​were calculated based on the stored information to evaluate the quality estimation accuracy (Accuracy) and degradation detection accuracy (Recall). The experimental results are shown in Figure 4. S ｔｈ is a predetermined threshold value for the stability S.

[0041] As shown in FIG. ｔｈ When quality estimation is attempted only when S is satisfied, accuracy improves compared to when quality estimation is attempted without considering S. In other words, in an environment where regularity or stationarity is observed (high reproducibility), quality estimation (for example, quality estimation based on statistical information) is effective, and it is thought that by applying quality estimation only to effective regions, it is possible to avoid destabilization of proactive control based on erroneous future predictions.

[0042] Note that S≧S ｔｈ When quality estimation is attempted only when S is satisfied, Recall decreases compared to when quality estimation is attempted without considering S. This is because deterioration detection due to excessive quality deterioration alerts is not included.

[0043] (Relationship between the Principle of Accuracy / Recall and Clarification of Applicable Area) With regard to the above-mentioned Accuracy / Recall, the relationship between the principle of Accuracy / Recall and clarification of applicable area will be described with reference to Fig. 5. In Fig. 5, for example, "correct answers in areas with high reproducibility" means the number of "correct answers in areas with high reproducibility". The same applies to other cases.

[0044] As shown in Figure 5, when S is not taken into consideration, Accuracy is calculated as follows: Accuracy = ("correct answer in high reproducibility area" + "correct answer in low reproducibility area") ÷ ("estimation attempt in high reproducibility area" + "estimation attempt in low reproducibility area").

[0045] S≧S ｔｈ If quality estimation is attempted only when the accuracy is high, Accuracy = "correct answers in areas with high reproducibility" / "estimation attempts in areas with high reproducibility." As shown in Figure 4, avoiding estimation attempts outside the applicable area has the effect of increasing Accuracy.

[0046] On the other hand, with regard to Recall, if S is not taken into consideration, Recall = ("detection of deterioration in areas with high reproducibility" + "detection of deterioration in areas with low reproducibility") ÷ ("quality deterioration in areas with high reproducibility" + "quality deterioration in areas with low reproducibility").

[0047] S≧S ｔｈ If quality estimation is attempted only when the above condition is satisfied, Recall = ("detection of deterioration in areas with high reproducibility") ÷ ("quality deterioration in areas with high reproducibility" + "quality deterioration in areas with low reproducibility").

[0048] That is, S≧S ｔｈIf quality estimation is attempted only when the condition is met, irregular / unsteady degradation that would have been detected by excessive quality degradation alerts cannot be detected. On the other hand, regular / steady quality degradation can be detected. Therefore, there is no problem with the degradation detection function in the application area of ​​proactive control. Rather, it is considered that the benefit of being able to avoid excessive quality degradation alerts is greater in cooperative control that also uses reactive control. (Image of Optimizing the Application Area Based on Network Quality Behavior and Stability) The application decision of a control method by the cooperative control unit 200 will be described later. The concept behind this application decision will be explained with reference to FIG. 6. FIG. 6 shows an image of optimizing the application area based on network quality behavior and stability. The horizontal axis in FIG. 6 represents quality, with quality improving toward the right and worsening toward the left. The vertical axis represents the frequency of quality information. Each curve in the figure represents an image of the frequency distribution for quality.

[0049] The distribution region indicated by A (the region of poor quality) has a narrow distribution width (high stability) and is a region of quality degradation with high reproducibility that is the target of detection. The cooperative control unit 200, having obtained the quality estimation result for this region, instructs the controlled object 300 to avoid using the network, for example, by proactive control.

[0050] The distribution region shown in B (region of better quality) has a narrow distribution width (high stability) and is a high-quality region with high reproducibility where network resources are desirable to utilize. The cooperative control unit 200, having obtained the quality estimation result in this region, instructs the controlled object 300 to use the network, for example, by proactive control.

[0051] In the distribution region shown in C, the frequency of poor quality is high, but the distribution width is wide and the stability is low. This distribution region is a quality degradation region that can be detected with quality estimation that is on the safe side, and there is a trade-off with a decrease in accuracy.

[0052] Moreover, in the distribution region shown in D, the frequency of high quality is high, but the distribution width is wide and stability is low. This distribution region is a region where quality estimation leaning towards the safe side (quality estimation leaning towards the bad side) has a large disadvantage. In other words, excessive quality degradation alerts increase the possibility that network resources that would otherwise be available will not be available. Regions C and D are suitable for reactive-mainly cooperative control.

[0053] By selectively using proactive control and reactive control as described above, cooperative control based on quality estimation with an optimized application range becomes possible, thereby making it possible to avoid the following two problems: Avoidance Point 1: System instability caused by excessive control based on quality estimation that is too safe (underestimation of quality); Avoidance Point 2: Occurrence of negative events such as video interruption caused by erroneous control due to overestimation of quality. (Examples of Quality Estimation Results and Cooperative Control Instructions) Examples of quality estimation results and cooperative control instructions in this embodiment will be described with reference to FIG. 7. As shown in FIG. 7, the quality estimation unit 130 transmits a quality estimation value (prediction value of future quality) and quality stability S to the cooperative control unit 200 as the quality estimation result. The stability S is a value obtained by the quality stability estimation unit 110 and corresponds to the reliability of the future prediction.

[0054] The cooperative control unit 200 controls the control target 300 based on the quality estimation result received from the quality estimation unit 130. Fig. 7 shows a device 301, a network 302, and an edge cloud 302 as examples of the control target 300. A more specific example of the operation will be described below.

[0055] The cooperative control unit 200 may, for example, set a threshold value (S ｔｈ ) and determines whether or not to implement proactive control based on a comparison between the threshold value and the stability S.

[0056] The quality estimator 130 transmits to the cooperative controller 200 the stability S (for example, the aforementioned “ln N / CV”) for each of the throughput, packet loss rate, communication delay, and jitter.

[0057] The cooperative control unit 200 determines whether S≧S for each of the throughput, packet loss rate, communication delay, and jitter. ｔｈ If so, the corresponding quality estimate is used.

[0058] The cooperative control unit 200 determines whether S<S for each of the throughput, packet loss rate, communication delay, and jitter. ｔｈ If the quality estimation value is not applicable, it is treated as an area where estimation is not possible. Examples of control targets or change targets (control / change targets) for each control target are as follows:

[0059] The control / change targets for the device 301 (e.g., connected car) include, for example, the travel speed, the travel route, the format of transmission information such as video, and the bit rate of video.

[0060] The control / change targets for the NW 302 include, for example, line switching, bonding, information transmission flow allocation, packet allocation, QoS class / priority, and the like.

[0061] The objects to be controlled / changed for the edge cloud 303 include, for example, a remote monitoring system, computing resources, applications, and operational methods including device migration plans.

[0062] FIG. 8 shows an example of control performed depending on the quality and stability. As shown in FIG. 8, when the quality estimation value is a value indicating good quality and the stability of the quality is high (i.e., S≧S ｔｈ In this case, the cooperative control unit 200 uses the quality estimation value for proactive control. The determination of "good quality" or "bad quality" can be made by, for example, comparing with a threshold value.

[0063] If the quality estimation value indicates poor quality and the stability of the quality is high (i.e., S≧S ｔｈ In this case, the cooperative control unit 200 avoids using the network having the quality estimation value by, for example, proactive control.

[0064] If the quality estimation value indicates good quality and the stability of the quality is low (i.e., S<S ｔｈ In this case, the cooperative control unit 200 does not use the quality estimation value for proactive control and avoids proactive control altogether. However, even in this case, the quality estimation value may be used for proactive control. In addition, proactive measures are taken to prepare for reactive responses.

[0065] If the quality estimation value indicates poor quality and the stability of the quality is low (i.e., S<S ｔｈ In this case, the cooperative control unit 200 does not use the quality estimation value for proactive control, and avoids proactive control altogether. Even in this case, proactive control may be used to avoid, for example, the use of a network with poor quality as much as possible. If the quality estimation value must be used, advance measures are taken to prepare for a reactive response.

[0066] (Clarification of Areas Addressable by Proactive Control) As an example, clarification of areas addressable by proactive control in the case where the control target is a connected car will be described with reference to Fig. 9. Note that the divisions shown in Fig. 9 are just an example.

[0067] 9 shows an operational area of ​​a connected car, which is made up of nine areas. For ease of explanation, each of the nine areas is numbered.

[0068] Areas 1, 2, 3, 4, and 5 have good estimation quality and high stability. Areas 1, 2, 3, 4, and 5 are areas that are primarily addressed by proactive control and are the basic ODD (Operational Design Domain) of connected cars. Note that in areas 1, 2, 3, 4, and 5, reactive control is used in a complementary manner to reduce the workload of the trouble-shooting system.

[0069] Region 6 is a region with good estimation quality and low stability, regions 7 and 8 are regions with poor estimation quality and low stability, and region 9 is a region with poor estimation quality and high stability.

[0070] Areas 6, 7, 8, and 9 are areas that are primarily handled by reactive control and are optional ODDs. Areas 6, 7, 8, and 9 also use proactive control as an auxiliary measure to strengthen the troubleshooting system.

[0071] Area 9 is an area to be avoided by proactive control and is basically outside the ODD. If quality information from this area must be used, the troubleshooting system should be strengthened.

[0072] (Processing Sequence) With reference to FIG. 10, a processing sequence of the communication system according to the present embodiment shown in FIG. 2 will be described.

[0073] In S1, quality information measured in the control target 300 is transmitted to the quality information distribution unit 210. In S2, the quality information distribution unit 210 transmits the quality information to the storage unit 120. The quality information is stored in the storage unit 120.

[0074] In S3, the quality stability estimation unit 110 requests quality information from the storage unit 120 and acquires the quality information in S4. In S5, the quality stability estimation unit 110 calculates the quality stability based on the quality information acquired in S4. In S6, the quality stability estimation unit 110 stores the calculated stability in the storage unit 120. If a stability already exists, it is updated. The processes of S1 to S6 are performed continuously.

[0075] In S7, the quality estimation unit 130 requests the quality information and the stability corresponding to the quality information from the storage unit 120. In S8, the quality estimation unit 130 acquires the quality information and the stability corresponding to the quality information from the storage unit 120.

[0076] In S9, the quality estimation unit 130 calculates a future quality estimate based on the quality information acquired in S8. The method for calculating the quality estimate may be changed depending on the stability. For example, quality estimation may be performed using only quality information whose stability is higher than a threshold.

[0077] In S10, the quality estimator 130 transmits a quality estimation result including a quality estimation value and stability to the cooperative control unit 200. In S11, the cooperative control unit 200 determines a control method (e.g., proactive control or reactive control) based on the quality estimation value and stability. In S12, the cooperative control unit 200 issues a control instruction to the control target based on the control method determined in S11.

[0078] (Example of control policy for moving device based on stability S) As a specific example related to the determination of the control method by the cooperative control unit 200, an example of a control policy when the controlled object 300 is a moving device (here, a connected car) will be described. Here, the following control policy 1 and control policy 2 are assumed as the control policies.

[0079] Control policy 1: Control policy 1 is a policy that minimizes reactive responses to the occurrence of irregular events. Therefore, it is a policy that prioritizes the selection of routes with a certain degree of stability S, including detour routes.

[0080] Control policy 2: Control policy 2 is a policy for cases where a reactive response system can be ensured, and is a policy that prioritizes the selection of the shortest route (including the case where S is small) that allows the use of a high-quality network.

[0081] An example will be described with reference to Figure 11. Assume that a connected car moves along a vertical and horizontal route connecting each point on the (x, y) coordinate plane shown in Figure 11. Also, assume that on each route connecting two coordinate points in Figure 11, one or more networks to which the connected car can connect are provided, and that at least one network with good quality exists. For example, assume a situation in which there are mobile networks from company A, company B, and company C, and on one route, the quality of company A's mobile network is good.

[0082] As shown in FIG. 11, the path (0,0)-(0,1), the path (1,0)-(1,1), and the path (2,1)-(2,2) are all S<S ｔｈ The route is as follows: S ｔｈis the threshold value already explained, and is a value indicating the stability of the network required to ensure the reliability of quality estimation.

[0083] The route "(0,0)-(1,0)-(2,0)-(2,1)-(1,1)-(1,2)-(2,2)" shown by A in FIG. 11 is an example of route selection based on control policy 1. S≧S ｔｈ The route is composed of only routes where

[0084] The route indicated by B, "(0,0)-(1,0)-(2,0)-(2,1)-(2,2)", is an example of route selection based on control policy 2. The shorter route is given priority, and S<S ｔｈ On this route, remote monitoring and control will be used to respond to irregularities, including reducing travel speed.

[0085] (Other Examples) Quality information such as throughput, packet loss rate, communication delay, and jitter, and their respective stability levels, may be managed for each area defined in space-time. This makes it possible to clarify the "coverage range of various use cases, such as connected cars," and the "application area of ​​various prediction / estimation utilization methods, such as proactive control based on quality estimation." Note that the area defined in space-time is, for example, a plane area having a space axis and a time axis, as disclosed in Non-Patent Document 1.

[0086] Furthermore, the technology according to the present embodiment may be applied to a cooperative control method that combines proactive control and reactive control. For example, proactive control is switched on and off in real time based on the stability S. More specifically, a mechanism may be provided in which proactive control is automatically turned on when S reaches a certain level or more.

[0087] Furthermore, the stability and quality estimation methods may be automatically improved by a feedback loop that minimizes the difference between the estimation results and the actual results. Furthermore, in cooperative control, the quality estimation results, the applied control method, and the actual results may be compared and analyzed, and the quality estimation method and the control method may be automatically improved by a feedback loop that optimizes them.

[0088] Furthermore, the technology according to this embodiment may be utilized for base station control of a mobile network (such as beamforming control for improving the quality itself or the reliability of future predictions) or for base station design.

[0089] The quality estimation unit 130 may also change the quality estimation method depending on the quality stability, or may change (optimize) the calculated quality estimation value itself by multiplying the calculated quality estimation value by a stability function as a coefficient.

[0090] Furthermore, in calculating the stability by the quality stability estimation unit 110, not only statistical indices such as standard deviation and variance may be incorporated into the stability estimation formula, but also MOS values, or other quality evaluation indices may be incorporated as functions.

[0091] Furthermore, the concept of stability may be applied not only to network quality (communication quality) but also to image / video quality indicators such as SSIM (Structural SIMilarity) and VMAF (Video Multi-Method Assessment Fusion) and other application quality.

[0092] (Configuration example of information processing device) As described above, a configuration including the quality stability estimation unit 110, the storage unit 120, and the quality estimation unit 130 is called the information processing device 100. Furthermore, each of the quality stability estimation unit 110, the storage unit 120, the quality estimation unit 130, the cooperation control unit 200, and the quality information distribution unit 210 may be called an information processing device. An information processing device is, for example, a computer. The computer may be a physical machine or a computing resource on the cloud.

[0093] 12 shows an example configuration of an information processing device 400 having the functions of the quality stability estimation unit 110 and the quality estimation unit 130. As shown in Fig. 12, the information processing device 400 includes an acquisition unit 410 that acquires quality information related to the quality of communication, a calculation unit 420 that calculates the stability of the quality based on the variability of the quality information, and a prediction unit 430 that predicts the quality of the communication based on the quality information and the stability. Note that the information processing device 400 may not include the prediction unit 430.

[0094] Fig. 13 shows an example of the configuration of an information processing device 500 having the functions of the cooperative control unit 200. As shown in Fig. 13, the information processing device 500 includes an acquisition unit 510 that acquires the stability of quality calculated based on the variability of quality information related to the quality of communication of the control target, and a control unit 520 that determines a control method for the control target based on the stability.

[0095] (Hardware Configuration Example) Any of the information processing devices described in this embodiment can be realized, for example, by causing a computer to execute a program. This computer may be a physical computer or a virtual machine on a cloud.

[0096] That is, the device can be realized by executing a program corresponding to the processing performed by the device using hardware resources such as a CPU and memory built into a computer. The program can be recorded on a computer-readable recording medium (such as a portable memory) and stored or distributed. The program can also be provided via a network such as the Internet or email.

[0097] Fig. 14 is a diagram showing an example of the hardware configuration of the computer. The computer in Fig. 14 includes 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, and the like, all of which are interconnected via a bus B. The computer may further include a GPU.

[0098] The program that realizes the processing on the computer is provided by a recording medium 1001, such as a CD-ROM or a memory card. When the recording medium 1001 storing 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, but may be downloaded from another computer via a network. The auxiliary storage device 1002 stores the installed program as well as necessary files, data, etc.

[0099] The memory device 1003 reads and stores a program from the auxiliary storage device 1002 when an instruction to start the program is received. The CPU 1004 realizes functions related to the device in accordance with 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) or the like according to the program. The input device 1007 is composed of a keyboard, mouse, buttons, a touch panel, etc., and is used to input various operation instructions. The output device 1008 outputs the results of calculations.

[0100] (Summary, Effects, etc. of the Embodiment) Key points of the technology according to the embodiment include, for example, the following (1) to (4).

[0101] (1) The reliability of future predictions is estimated (quantified as to the certainty of quality estimation) by utilizing the constancy (stability) of quality that can be grasped through the continuous accumulation of quality information.

[0102] In relation to the above points, the following findings were gained from the demonstration of remote monitoring of connected cars using a commercial mobile network.

[0103] The more quality information samples that can be used for quality estimation, the higher (more stable) the estimation accuracy tends to be.

[0104] - When the quality estimation target does not show regularity / steadiness or the degree of regularity is low (numerical variations are observed), it is difficult to improve the estimation accuracy.

[0105] (2) It clarifies the scope of application of future predictions that include uncertain variables using numerical standards.

[0106] (3) The stability of quality can be defined for each category of quality information management (area defined on the time-space axis, type of quality information, etc.).

[0107] (4) Based on the quality information continuously obtained through passive measurement, the quality stability as well as the quality information can be updated at any time.

[0108] The technology according to this embodiment makes it possible to avoid proactive control based on uncertain future predictions, thereby making it possible to avoid reduced resource utilization efficiency and unstable operation due to excessive control.

[0109] Furthermore, since the application area of ​​proactive control can be clarified, cooperative control including reactive control and operational response can be stabilized and made more efficient.

[0110] The following additional notes are provided regarding the above-described embodiments.

[0111] <Additional Notes> (Additional Item 1) An information processing device comprising: an acquisition unit that acquires quality information related to communication quality; and a calculation unit that calculates stability of the quality based on variations in the quality information. (Additional Item 2) The information processing device according to Additional Item 1, further comprising: a prediction unit that predicts the quality of the communication based on the quality information and the stability. (Additional Item 3) The information processing device according to Additional Item 1, wherein the calculation unit calculates the stability using the number of samples of the quality information. (Additional Item 4) An information processing device comprising: an acquisition unit that acquires stability of the quality calculated based on variations in quality information related to the quality of communication of a control object; and a control unit that determines a control method for the control object based on the stability. (Additional Item 5) The information processing device according to Additional Item 4, wherein the control unit determines whether to implement proactive control as the control method based on the stability. (Additional Item 6) The information processing device according to Additional Item 4, wherein the control unit determines the control method by comparing the stability with a threshold for the stability. (Supplementary Item 7) A stability calculation method executed by an information processing device, comprising: an acquisition step of acquiring quality information relating to communication quality; and a calculation step of calculating the stability of the quality based on a variation in the quality information. (Supplementary Item 8) A non-transitory storage medium storing a program for causing a computer to function as each unit in the information processing device according to any one of Supplementary Items 1 to 6.

[0112] 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.

[0113] REFERENCE SIGNS LIST 100 Information processing device 110 Quality stability estimation unit 120 Storage unit 130 Quality estimation unit 200 Collaboration control unit 210 Quality information distribution unit 400 Information processing device 410 Acquisition unit 420 Calculation unit 430 Prediction unit 500 Information processing device 510 Acquisition unit 520 Control 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: an acquisition unit that acquires quality information related to communication quality; and a calculation unit that calculates the stability of the quality based on the variation in the quality information.

2. The information processing device according to claim 1, further comprising a prediction unit that predicts the quality of the communication based on the quality information and the stability.

3. The information processing device according to claim 1, wherein the calculation unit calculates the stability using the number of samples of the quality information.

4. An information processing device comprising: an acquisition unit that acquires the stability of quality calculated based on the variability of quality information regarding the quality of communication of a control object; and a control unit that determines a control method for the control object based on the stability.

5. The information processing device according to claim 4, wherein the control unit determines whether or not to implement proactive control as the control method based on the stability.

6. The information processing device according to claim 4, wherein the control unit determines the control method by comparing the stability with a threshold value for the stability.

7. A stability calculation method executed by an information processing device, comprising: an acquisition step of acquiring quality information relating to communication quality; and a calculation step of calculating the stability of the quality based on the variation in the quality information.

8. A program for causing a computer to function as each unit in the information processing device according to any one of claims 1 to 6.

Images