Communication environment evaluation device, communication environment evaluation system, and communication environment evaluation method

Abstract

This communication environment evaluation device (10) is characterized by comprising: an information acquisition unit that acquires reception data of communication packets of a plurality of applications transmitted between a vehicle-mounted wireless communication device (3) installed on a vehicle (2) and a terrestrial wireless communication device (8) installed on the ground; an application identification unit that identifies the applications; and an evaluation unit that analyzes the reception data of the identified applications and evaluates a change in the communication environment for each application.

Description

Communication environment evaluation device, communication environment evaluation system, and communication environment evaluation method

[0001] The present disclosure relates to a communication environment evaluation device, a communication environment evaluation system, and a communication environment evaluation method for evaluating the communication environment of wireless communication between on-board and ground stations.

[0002] 3GPP (registered trademark), a standardization organization for mobile communication systems, is studying fifth-generation (hereinafter sometimes referred to as "5G") wireless access systems. Currently, mobile communication systems used in railway operations are often built as private networks designed, constructed, and managed by railway operators, but it is expected that 5G systems will also be used for wireless communication between on-board and ground-based systems in the railway industry.

[0003] In Japan, telecommunications carriers have been gradually providing 5G commercial services since around 2020. Local 5G is also being institutionalized as a mechanism for local businesses, local governments, and other organizations to build spot 5G networks within their own buildings or premises. Compared to conventional communication systems, 5G systems offer features such as high speed, large capacity, low latency, high reliability, and multiple simultaneous connections. Taking advantage of these features, an increase in new applications is expected in the railway industry, such as autonomous driving, train control, and on-board video surveillance.

[0004] In a wireless communication system, since communication quality changes over time, it is important to understand the current communication quality for stable operation of the system. For example, Patent Document 1 proposes a method for evaluating the communication environment for each communication path in a system in which multiple types of communication paths are available.

[0005] International Publication No. 2022 / 219730

[0006] However, in the above-mentioned conventional technology, even when data of multiple applications are transmitted over a single communication path, the communication environment is evaluated for each communication path, which makes it impossible to evaluate the communication environment for each application. Because the underlying design requirements and required performance requirements differ for each application, it is important to evaluate the communication environment for each application and manage the communication quality.

[0007] The present disclosure has been made in view of the above, and aims to provide a communication environment assessment device capable of assessing a communication environment for each application.

[0008] In order to solve the above-mentioned problems and achieve the objectives, the communication environment evaluation device disclosed herein is characterized by comprising an information acquisition unit that acquires received data of communication packets of multiple applications transmitted between an on-board wireless communication device installed in a vehicle and a ground wireless communication device installed on the ground, an application identification unit that identifies the application of the communication packets, and an evaluation unit that analyzes the received data of the identified application and evaluates changes in the communication environment for each application.

[0009] The communication environment assessment device according to the present disclosure has an effect of being able to assess the communication environment for each application.

[0010] 2 is a diagram showing an example of the configuration of a communication environment evaluation system according to a first embodiment; FIG. 1 is a diagram showing an example of the functional configuration of the communication environment evaluation device shown in FIG. 1; FIG. 2 is a diagram showing an example of received data acquired by the information acquisition unit shown in FIG. 2; FIG. 2 is a diagram showing an example of data classification by the application identification unit shown in FIG. 2; FIG. 13 is a diagram showing an example of received data acquired by the communication environment evaluation device shown in FIG. 12. FIG. 14 is a flowchart for explaining the learning process in the second embodiment. FIG. 15 is an explanatory diagram of the calculation process of the reception interval of communication packets in the third embodiment. FIG. 16 is a flowchart for explaining the learning process in the third embodiment. FIG. 17 is an explanatory diagram of the learning process in the third embodiment. FIG. 18 is a diagram showing an example of the configuration of the communication environment evaluation system according to the fourth embodiment. FIG. 19 is a diagram showing an example of the configuration of the communication environment evaluation system according to a modified example of the fourth embodiment. FIG. 20 is a diagram showing an example of the configuration of the communication environment evaluation system according to the fifth embodiment. FIG. 21 is a diagram showing an example of the configuration of the communication environment evaluation system according to a modified example of the fifth embodiment. FIG. 22 is a diagram showing an example of the hardware configuration of the communication environment evaluation devices according to the first to fifth embodiments and the data aggregating device according to the second embodiment.

[0011] A communication environment evaluation device, a communication environment evaluation system, and a communication environment evaluation method according to embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0012] First Embodiment. Fig. 1 is a diagram showing an example of the configuration of a communication environment evaluation system 1 according to a first embodiment. The communication environment evaluation system 1 includes an on-board wireless communication device 3 and an on-board terminal 4 mounted on a train car 2, and a base station 5, a core network device 6, a connection router 7, a ground wireless communication device 8, and a communication environment evaluation device 10 installed on the ground. A plurality of applications represented by app #1 to app #n share a wireless section for ground-to-vehicle communication. Data is transmitted at different cycles for each application.

[0013] For example, a communication packet transmitted by the on-board wireless communication device 3 is transmitted to the trackside wireless communication device 8 via the on-board terminal 4, base station 5, core network device 6, and connection router 7. The communication environment evaluation device 10 acquires received data, which is data of the communication packets received by the trackside wireless communication device 8 from the on-board wireless communication device 3 as described above, and evaluates the communication environment for each application. The communication environment evaluation device 10 may be located on the cloud.

[0014] Fig. 2 is a diagram showing an example of the functional configuration of the communication environment assessment device 10 shown in Fig. 1. The communication environment assessment device 10 has an information acquisition unit 11, an application identification unit 12, a feature calculation unit 13, a feature DB (DataBase) 14, an evaluation unit 15, a storage unit 16, and a display unit 17.

[0015] The information acquisition unit 11 extracts received data, which is information on communication packets received by the ground wireless communication device 8. The information acquisition unit 11 may acquire, as the received data, not only data included in the received communication packets but also data related to the received communication packets, such as data such as a reception time generated upon reception of the communication packets by the ground wireless communication device 8, and data acquired from inside or outside the communication environment evaluation system 1 based on the data included in the received communication packets.

[0016] FIG. 3 is a diagram illustrating an example of received data acquired by the information acquisition unit 11 illustrated in FIG. 2 . The received data illustrated in FIG. 3 includes an application type, which is information for identifying the type of application; a time zone, a reception time, and an acquisition date, which indicate the time when the communication packet was received; and vehicle information, which includes a vehicle position, vehicle conditions, and operating conditions. The application type is, for example, data included in the received communication packet, such as a protocol, a port number, and a data length. The time zone represents the reception time as a time range, and is information assigned by the information acquisition unit 11 based on, for example, information about the reception time assigned by the ground wireless communication device 8. In the example illustrated in FIG. 3 , the vehicle position includes a section and location information. The vehicle position may be included in the received communication packet, or the vehicle 2 may be identified from information included in the communication packet and the location information of the vehicle 2 may be acquired from a train traffic management system (not illustrated) or the like. The section represents the vehicle position as a location range, and is information assigned by the information acquisition unit 11 based on, for example, the acquired vehicle position.

[0017] The vehicle conditions are, for example, information such as the vehicle type and the vehicle antenna installation position. The antenna installation position may be, for example, information indicating whether the antenna is installed indoors or outdoors. The operation conditions are, for example, information such as the line section and direction of travel, which are vehicle operation information. The direction of travel may be, for example, information indicating whether the direction is uphill or downhill. The information acquisition unit 11 outputs the acquired received data to the application identification unit 12.

[0018] Returning to the description of FIG. 2 , the application identification unit 12 classifies communication packets by application based on the received data acquired by the information acquisition unit 11. The application identification unit 12 identifies applications based on the application type information acquired by the information acquisition unit 11 and classifies the received data by application. Examples of application identification methods include methods based on protocols, port numbers, data lengths, and the like. For example, if the application type information acquired by the information acquisition unit 11 corresponds one-to-one with the application, the application identification unit 12 can identify applications for each application type value. Alternatively, the application identification unit 12 may identify applications by a combination of multiple types of information. Furthermore, the application identification unit 12 can classify the received data based on the received data acquired by the information acquisition unit 11. For example, the application identification unit 12 can classify the received data by time zone, section, vehicle conditions, and operating conditions. The application identification unit 12 classifies the received data for each combination of time zone, section, vehicle conditions, and operating conditions. The application identification unit 12 associates the received data with the identified application and classification information and outputs the received data to the feature calculation unit 13.

[0019] Fig. 4 is a diagram showing an example of data classification by the application identification unit 12 shown in Fig. 2. In the example shown in Fig. 4, the application identification unit 12 classifies the received data based on the application type, the time period indicating the time when the communication packet was received, the reception time and acquisition date, the section and location information indicating the vehicle location, the vehicle conditions, and the operating conditions.

[0020] The feature calculation unit 13 calculates a feature indicating the characteristics of the communication packet for each classification by the application identification unit 12. Specifically, the feature calculation unit 13 calculates the reception interval from the difference in reception time of the communication packets, and can calculate the probability density function (PDF) of the reception interval as the feature. In addition to the probability density function of the reception interval, the feature may be a packet loss rate calculated from sequence numbers per unit time, or a receiving-side throughput calculated from the data amount of the received communication packets.

[0021] FIG. 5 is an explanatory diagram of an example of the operation of the feature calculation unit 13 shown in FIG. 2. As shown in FIG. 5, the feature calculation unit 13 calculates the reception interval at the ground wireless communication device 8 of communication packets transmitted from the on-board wireless communication device 3 to the ground wireless communication device 8 via a wireless section. In FIG. 5, the packet reception interval of application #k is calculated. In the middle diagram of FIG. 5, the horizontal axis represents the packet number, and the vertical axis represents the reception interval of the communication packet having that packet number. Here, the packet number is incremented each time a communication packet is received, so the communication packet with the largest packet number is the most recently received communication packet. Based on the calculated reception interval, the feature calculation unit 13 calculates a probability density function of the reception interval for each specific interval. The interval used to calculate the PDF is referred to as the PDF calculation interval, as shown in FIG. 5.

[0022] Although there is some fluctuation in the reception intervals of communication packets, there is a tendency for the distribution to concentrate on specific values ​​depending on the transmission cycle of the application. Fluctuations in the reception intervals of communication packets become greater when the radio wave environment or traffic environment deteriorates and packet loss occurs in the wireless section. By using the PDF of the reception interval as a feature and analyzing changes in the fluctuations in the reception interval for each application, the communication environment can be evaluated.

[0023] However, if the communication packets are received on different days or if the data is received by a subsequent vehicle 2 in the bus schedule, the reception interval will be long. Also, in applications where traffic volume is high and data is divided into multiple communication packets for transmission, and many packets are transmitted at once, the reception interval may become short regardless of the application's transmission cycle. For this reason, it is desirable to exclude data with reception intervals that are not subject to the above-mentioned analysis from the analysis.

[0024] Returning to the explanation of FIG. 2 , the feature DB 14 is a database that accumulates feature quantities calculated by the feature quantity calculation unit 13. The feature quantity DB 14 stores feature quantities for each combination of application type, time period when a communication packet is received, section, vehicle conditions, and operating conditions. The feature quantity DB 14 also accumulates trained models. The feature quantity DB 14 accumulates feature quantities and trained models for each application type and for each classification that combines time period, section, vehicle conditions, and operating conditions. Note that the feature quantity DB 14 only stores data when communication is being performed normally, and does not store data when an abnormality occurs in communication.

[0025] 6 is a diagram showing an example of data stored in the feature DB 14 shown in FIG. 2. FIG. 6 shows an example in which a probability density function (PDF) of reception intervals is stored as a feature. A feature is stored for each application represented by app #1 to app #n. The feature and the trained model are associated with each classification by the application identification unit 12.

[0026] The trained model is generated by having a learning device learn data stored in the feature DB 14. In this case, the learning device may be provided in the feature DB 14, or the feature DB 14 and the learning device may be provided separately.

[0027] 7 is a flowchart illustrating the learning process for generating a trained model. The information acquisition unit 11 acquires received data of a communication packet received by the ground wireless communication device 8 (step S101).

[0028] The application identification unit 12 classifies the received data of communication packets for each application (step S102).The application identification unit 12 also classifies the received data based on the contents of the received data, such as time zone, section, vehicle conditions, and operating conditions (step S103).

[0029] The feature calculation unit 13 calculates feature quantities of the communication packets for each classification by the application identification unit 12 (step S104). The feature calculation unit 13 stores data of the calculated feature quantities in the feature DB 14 (step S105). The feature DB 14 causes a learning device to learn the stored data and generate a trained model (step S106). The feature DB 14 stores the generated trained model in association with the feature quantity data. The trained model is used to obtain evaluation results from the feature quantities. The learning device may be provided in the feature DB 14, or the feature DB 14 and the learning device may be provided separately.

[0030] Furthermore, the information about the features and the trained model stored in the feature DB 14 may be stored in advance in the feature DB 14. In this case, the learning flow can be omitted.

[0031] FIG. 8 is an explanatory diagram of the generation of a trained model. The feature DB 14 uses data stored in the feature DB 14 to generate a trained model for outputting an evaluation result of a communication environment from the features. As shown in FIG. 8 , the data stored in the feature DB 14 is features stored for each application and for each classification based on the value of the received data. In FIG. 8 , packet reception intervals are stored as features for each combination of time period, section, vehicle conditions, and operating conditions. In each of the multiple graphs in the upper diagram of FIG. 8 , the horizontal axis represents the reception interval of communication packets, the vertical axis of the bar graph represents the number of data, and the vertical axis of the line graph represents PDF.

[0032] The accumulated data as described above is input to a learning device, and a trained model is generated for each application and for each classification based on the values ​​of the received data. The generated trained model is stored for each application and for each of the above classifications.

[0033] Here, a case where kNN (k-Nearest Neighbor), a clustering technique, is applied will be described. kNN is a technique for analyzing data to be evaluated using training data. Therefore, unlike other machine learning algorithms, there is no process such as parameter adjustment in the learning phase. When kNN is used, the process of generating the trained model described above refers to the process of classifying and storing the training data. Specifically, the learner acquires feature amounts when the communication environment is normal as training data, and classifies and stores the data by application type, time period when the data was acquired, section, vehicle conditions, and operating conditions.

[0034] Returning to the description of Fig. 2, the evaluation unit 15 has a function of analyzing received data of communication packets and evaluating changes in the communication environment for each application. Specifically, the evaluation unit 15 can evaluate the communication environment using the features calculated by the feature calculation unit 13 and the trained model stored in the feature DB 14.

[0035] 9 is an explanatory diagram of the operation of the evaluation unit 15 shown in FIG. 2. The evaluation unit 15 queries the feature DB 14 for a trained model under the same conditions, using as input conditions the type of application of the received data to be evaluated, the time period in which the data was acquired, the section, the vehicle conditions, and the operating conditions. The feature DB 14 returns a trained model under the same conditions as the input conditions as a response to the evaluation unit 15. The evaluation unit 15 inputs features into a trained model that matches the input conditions, and obtains an evaluation result that is output.

[0036] The evaluation result may be, for example, a risk level. In kNN, k neighboring data for the data to be evaluated is extracted from the training data, and the distance between the data to be evaluated and the neighboring data is calculated. In other words, when the evaluation unit 15 inputs the feature quantities of the evaluation target into the trained model, a distance is output as the evaluation result. The trained model is trained only on data obtained when the communication environment is normal, and this distance is the difference between the feature quantities of the training data, which are the feature quantities in normal conditions, and the feature quantities of the data to be evaluated. Therefore, it can be used as a risk level indicating the difference from normal conditions. Therefore, the higher the risk level, the greater the change in the data to be evaluated from the normal communication environment.

[0037] In this embodiment, the evaluation unit 15 has been described as outputting the evaluation result of the communication environment using a trained model trained by a learning device contained in the feature DB 14, but it may also be configured to acquire a trained model from outside and output the evaluation result based on this trained model.

[0038] In this way, the evaluation unit 15 can output the evaluation result obtained based on the feature amount of the communication packet to the display unit 16.

[0039] In this embodiment, the case where kNN is applied to the learning algorithm has been described, but the present invention is not limited to this. Reinforcement learning, supervised learning, unsupervised learning, semi-supervised learning, etc. can also be applied to the learning algorithm.

[0040] Furthermore, the learning algorithm used by the learning device may be deep learning, which learns to extract the feature values ​​themselves, or any other known method.

[0041] Furthermore, the clustering method is not limited to the above kNN, and known clustering methods such as non-hierarchical clustering using the K-means method or hierarchical clustering such as the shortest distance method may also be used.

[0042] The functions of the learning device and the evaluation unit 15 may be, for example, connected to the communication environment evaluation device 10 via a network and may be separate from the communication environment evaluation device 10. The functions of the learning device and the evaluation unit 15 may reside on a cloud server. The learning data may be feature quantities acquired by another communication environment evaluation device 10 as long as the conditions, such as the application used, are the same.

[0043] 10 is a flowchart for explaining the evaluation process for evaluating the communication environment. The information acquisition unit 11 acquires reception data of a communication packet received by the ground wireless communication device 8 (step S201).

[0044] The application identification unit 12 classifies the received data of communication packets for each application (step S202). The application identification unit 12 also classifies the received data based on the contents of the received data, such as time zone, section, vehicle conditions, and operating conditions (step S203).

[0045] The feature amount calculation unit 13 calculates the feature amount of the communication packet for each classification by the application identification unit 12 (step S204).

[0046] Next, the evaluation unit 15 queries the feature DB 14 to acquire a trained model under the same conditions as the received data to be evaluated (step S205).

[0047] The evaluation unit 15 evaluates the communication environment using the acquired trained model (step S206). The display unit 17 displays the evaluation result on the display screen (step S207).

[0048] Returning to the explanation of Fig. 2, the storage unit 16 stores the evaluation results of the evaluation unit 15. The display unit 17 displays the evaluation results stored in the storage unit 16 on a display screen. The storage unit 16 stores the generated evaluation results every time the evaluation results are generated. Furthermore, the display unit 17 updates the display content every time the evaluation results are updated.

[0049] 11 is a diagram showing an example of a display screen output by the display unit 17 shown in FIG. 2. The display unit 17 displays the risk level, which is the evaluation result, for each application and for each combination of vehicle conditions and driving conditions. k , operating conditions O k The horizontal axis of FIG. 11 represents the interval, and the vertical axis represents the time period. n The risk level for each of the time periods T1 to T n Each of the above can be expressed by the color or pattern used to fill each square. In the example of Fig. 11, the pattern used to fill each square is changed depending on the risk level.

[0050] As described above, the communication environment evaluation device 10 according to the first embodiment is characterized by including an information acquisition unit 11 that acquires received data of communication packets of a plurality of applications transmitted between the on-board wireless communication device 3 installed in the vehicle 2 and the ground wireless communication device 8 installed on the ground, an application identification unit 12 that identifies the application of the communication packets, and an evaluation unit 15 that analyzes the received data of the identified application and evaluates changes in the communication environment for each application. With this configuration, it is possible to evaluate changes in the communication environment for each type of application from the received data of the received communication packets.

[0051] The importance of changes in the communication environment and responses to changes in the communication environment vary depending on the application. For example, even if the communication environment of an application for displaying advertisements to passengers in vehicle 2 deteriorates, the advertisements will simply not be displayed for a certain period of time, and so there will be little impact on the operation of vehicle 2. However, if the communication environment of an application for autonomous driving of vehicle 2 deteriorates, this could in some cases escalate into a situation that puts human life at risk. Therefore, for example, when the communication environment begins to deteriorate, responses can be taken, such as stopping vehicle 2 in a safe place or stopping communication of less important applications other than autonomous driving, before a communication failure occurs.

[0052] The communication environment assessment device 10 further includes a feature amount calculation unit 13 that calculates feature amounts indicating characteristics of communication packets from received data, and the assessment unit 15 can also analyze the communication packets based on the feature amounts calculated by the feature amount calculation unit 13. The feature amount may be any value that corresponds to the communication quality, and may be, for example, the reception interval of communication packets, the packet loss rate, or the receiving-side throughput.

[0053] The communication environment assessment device 10 may further include a feature DB 14 that accumulates the feature calculated by the feature calculation unit 13 for each classification based on the received data. The "classification" is determined, for example, by a combination of application type, time period, section, vehicle conditions, and operating conditions. In the above embodiment, the classification is determined using all of the application type, time period, section, vehicle conditions, and operating conditions. However, the classification can be determined based on at least one of the application type, time period, section, vehicle conditions, and operating conditions. The feature DB 14 may also store, for each of the above classifications, trained models for outputting evaluation results of the communication environment from the feature. This allows the assessment unit 15 to acquire, from the feature DB 14, trained models corresponding to the above classifications of the received data of the communication packets acquired by the information acquisition unit 11, and evaluate the communication environment using the feature of the communication packets and the acquired trained models.

[0054] The communication environment evaluation device 10 may further include a storage unit 16 that accumulates the evaluation results of the evaluation unit 15, and a display unit 17 that displays the evaluation results on a display screen for each classification based on the received data, based on the evaluation results accumulated in the storage unit 16. As shown in Fig. 11 , for example, the display unit 17 can display the evaluation results for each time period and for each section in which the vehicle 2 is located, regarding the application for which the evaluation results are to be displayed, the vehicle conditions of the vehicle 2 in which the on-board wireless communication device 3 that transmits communication packets is installed, and the operating conditions of the vehicle 2.

[0055] In the first embodiment, the communication environment assessment device 10 is installed on the ground. The information acquisition unit 11 acquires reception data of communication packets transmitted by the on-board wireless communication device 3 and received by the ground wireless communication device 8. Since the ground wireless communication device 8 receives communication packets from the on-board wireless communication devices 3 of multiple vehicles 2, the communication environment assessment device 10 acquires reception data of communication packets between the multiple on-board wireless communication devices 3 and the ground wireless communication device 8.

[0056] The received data includes information contained in the communication packet, information generated in association with the transmission of the communication packet, information relating to the device from which the communication packet is transmitted, and information relating to the device that is the source or destination of the communication packet. For example, the information acquisition unit 11 can acquire received data including at least one of information indicating the type of application, information indicating the time at which the communication packet is received, and vehicle information indicating the characteristics of the vehicle 2 in which the on-board wireless communication device 3 that transmits the communication packet is installed. The application identification unit 12 classifies the communication packet based on the received data. This makes it possible to identify the time and section in which a change in the communication environment is occurring.

[0057] The vehicle information is information including at least one of the position information of the vehicle 2, the vehicle conditions of the vehicle 2, and the operating conditions of the vehicle 2. The vehicle conditions are, for example, information such as the vehicle type and the antenna installation position of the vehicle 2. The antenna installation position may be, for example, information indicating whether the antenna is installed indoors or outdoors. The operating conditions are, for example, information such as the line section and direction of travel, which are operating information of the vehicle 2. The direction of travel may be, for example, information indicating whether it is going up or down.

[0058] Moreover, according to the first embodiment, it is possible to provide a communication environment evaluation system 1. This communication environment evaluation system 1 is a communication environment evaluation system 1 that evaluates the communication environment of communication packets of a plurality of applications transmitted between an on-board wireless communication device 3 installed in a vehicle 2 and a ground wireless communication device 8 installed on the ground, and is characterized by including a communication environment evaluation device 10 that has an information acquisition unit 11 that acquires received data of communication packets of a plurality of applications transmitted between the on-board wireless communication device 3 and the ground wireless communication device 8, an application identification unit 12 that identifies the application of the communication packets, and an evaluation unit 15 that analyzes the received data of the identified application and evaluates changes in the communication environment for each application.

[0059] Furthermore, according to the first embodiment, a communication environment evaluation method can also be provided. This communication environment evaluation method is executed by the communication environment evaluation device 10, and is characterized by including the steps of acquiring received data of communication packets of a plurality of applications transmitted between an on-board wireless communication device 3 installed on a vehicle and a ground wireless communication device 8 installed on the ground, identifying the applications of the communication packets, and analyzing the received data of the identified applications to evaluate the communication environment for each application. The communication environment evaluation method may also be executed by the communication environment evaluation system 1.

[0060] Second Embodiment Fig. 12 is a diagram showing an example of the configuration of a communication environment evaluation system 1A according to a second embodiment. In addition to the components of the communication environment evaluation system 1, the communication environment evaluation system 1A further includes a data aggregation device 9. Furthermore, while the communication environment evaluation system 1 includes a communication environment evaluation device 10 connected to a ground wireless communication device 8, the communication environment evaluation system 1A includes a communication environment evaluation device 10A connected to an on-board wireless communication device 3. The following mainly describes the differences from the communication environment evaluation system 1.

[0061] The communication environment evaluation system 1A analyzes communication packets received on the vehicle 2 side. Note that while only one vehicle 2 is shown in Fig. 12, in reality, the communication environment evaluation system 1A has an on-board wireless communication device 3 and a communication environment evaluation device 10A installed in each of a plurality of vehicles 2, and a data aggregation device 9 installed on the ground side aggregates the evaluation results and feature information from the communication environment evaluation devices 10A installed in each of the plurality of vehicles 2.

[0062] 13 is an explanatory diagram of the functions of the data aggregating device 9 of FIG. 12. The data aggregating device 9 has an evaluation result aggregating DB 91 and a feature aggregating DB 92. The data aggregating device 9 aggregates the evaluation results of the communication environment evaluation device 10A and feature information from each of the multiple vehicles 2-1 to 2-n. The data aggregating device 9 stores the aggregated evaluation results in the evaluation result aggregating DB 91, and stores the aggregated features and trained models in the feature aggregating DB 92.

[0063] The functional configuration of the communication environment evaluation device 10A is the same as that of the communication environment evaluation device 10 according to the first embodiment. However, since the data acquired by the information acquisition unit 11 is only data of communication packets received by the vehicle 2 in which the communication environment evaluation device 10A is installed, the vehicle conditions and operating conditions are fixed.

[0064] Fig. 14 is a diagram showing an example of received data acquired by the communication environment assessment device 10A shown in Fig. 12. The received data shown in Fig. 14 includes an application type, a time period indicating the time when the communication packet was received, the reception time and acquisition date, section and location information indicating the location of the vehicle 2, vehicle conditions, and operating conditions. Compared to the received data in the first embodiment shown in Fig. 3, this differs in that the vehicle conditions and operating conditions are fixed.

[0065] The operation of the communication environment assessment device 10A is similar to that of the communication environment assessment device 10 according to the first embodiment, but a learning device may be provided in the feature aggregation DB 92 to generate a trained model based on data aggregated for each vehicle condition and operating condition.

[0066] 15 is a flowchart for explaining the learning process in the second embodiment. The processes of steps S101 to S104 are the same as those in the first embodiment shown in FIG. 7. However, the "received communication packet" is a "communication packet received by the on-board wireless communication device 3." The processes of steps S101 to S104 are executed in each of the multiple vehicles 2.

[0067] Once the features are calculated for each vehicle 2, the data for each vehicle 2 is aggregated and stored in the feature aggregation DB 92 of the data aggregating device 9 (step S107). The data accumulated in the feature aggregation DB 92 is used as training data for training a learner to generate a trained model (step S108). The data aggregating device 9 then stores the generated trained model in each vehicle 2 (step S109).

[0068] In addition, the feature information and the trained model stored in the feature DB 14 of each vehicle 2 may be stored in a database in advance.

[0069] As described above, the communication environment evaluation device 10A according to the second embodiment is installed in a vehicle 2, whereas the communication environment evaluation device 10 according to the first embodiment is installed on the ground. In this case, the information acquisition unit 11 acquires reception data of communication packets transmitted by the ground wireless communication device 8 and received by the on-board wireless communication device 3. The communication environment evaluation device 10 installed on the ground as in the first embodiment acquires reception data of communication packets from multiple vehicles 2, but as described above, the communication environment evaluation device 10A installed in a vehicle 2 acquires reception data of communication packets received by a single vehicle 2. Therefore, the communication environment evaluation system 1A according to the second embodiment further includes a data aggregation device 9 installed on the ground that aggregates evaluation results transmitted from the multiple communication environment evaluation devices 10A. This makes it possible to check the communication environment of each vehicle 2 operating along the managed line. Furthermore, by aggregating information on the feature quantities of each vehicle 2, it becomes possible to share data with other vehicles 2 having the same vehicle conditions and operating conditions.

[0070] Third Embodiment A communication environment evaluation system 1B (not shown) according to a third embodiment has a communication environment evaluation device 10B instead of the communication environment evaluation device 10 of the communication environment evaluation system 1 according to the first embodiment. Other configurations of the communication environment evaluation system 1B are the same as those of the first embodiment.

[0071] The communication environment assessment device 10B has the same functional configuration as the communication environment assessment device 10. However, since some operations differ from those of the communication environment assessment device 10, only the different parts will be described below using the reference numerals of the communication environment assessment device 10. The communication environment assessment device 10B performs learning data labeling in advance for data acquired for each application type, time period, section, vehicle conditions, and operating conditions, and analyzes the data. The differences from the first embodiment are the processing in the feature calculation unit 13 and the operation related to the generation of a trained model in the feature DB 14. In the third embodiment, the parts that differ from the first embodiment will be mainly described, and a description of the parts that are the same as those in the first embodiment will be omitted.

[0072] 16 is an explanatory diagram of the calculation process of the reception interval of communication packets in embodiment 3. After the application identification unit 12 classifies the reception data by application type, time period, section, vehicle conditions, and operating conditions, the feature calculation unit 13 calculates the reception interval of communication packets. In embodiment 1, the calculation of the reception interval was performed for each data classified by time period, section, vehicle conditions, and operating conditions for each application, and by dividing the data into PDF calculation sections. However, in embodiment 3, the reception interval is calculated collectively without providing a PDF calculation section. In addition, the feature calculation unit 13 calculates a probability density function from the reception interval.

[0073] 17 is a flowchart illustrating the learning process in the third embodiment. The processes in steps S101 to S105 are the same as those in the first embodiment. However, as described above, in calculating the features in step S104, the reception interval and PDF are calculated without providing a PDF calculation interval. The feature DB 14 groups the classified data by clustering (step S110). The feature DB 14 also trains a learner on the accumulated data to generate a trained model for each group (step S111).

[0074] FIG. 18 is an explanatory diagram of the learning process in the third embodiment. As shown in FIG. 16 , the PDF of the reception interval is subdivided for each application based on time period, section, vehicle conditions, and operating conditions. In this case, the number of labels in the learning data increases by the number of combinations of classification conditions. Therefore, the feature DB 14 uses clustering to group the data subdivided based on time period, section, vehicle conditions, and operating conditions into data with high similarity. The clustering method is not particularly limited. For example, the feature DB 14 may use a hierarchical clustering method such as the group average method, Ward's method, or shortest distance method, or a non-hierarchical clustering method such as the k-means method. This has the effect of reducing the number of labels in the learning data and further reducing the amount of training data. The number of groups classified by clustering can be set arbitrarily. The feature DB 14 generates a trained model from a learner for each group obtained by clustering.

[0075] As described above, in the communication environment assessment device 10B according to the third embodiment, the feature DB 14 can group the features calculated by the feature calculation unit 13 for each of the above classifications based on the values ​​of the received data by clustering, into groups of data with high similarity, and store trained models for outputting evaluation results of the communication environment from the features generated for each of the classified groups in association with the groups. This achieves the effect of the first embodiment, in addition to the effect of reducing the number of labels in the training data because the training data is grouped by clustering before generating the trained models. Therefore, training data can be collected efficiently.

[0076] 19 is a diagram showing an example of the configuration of a communication environment evaluation system 1C according to a fourth embodiment. The communication environment evaluation system 1C includes a communication environment evaluation device 10C, which evaluates the communication environment for each application group, each group including a plurality of applications, instead of the communication environment evaluation device 10 of the communication environment evaluation system 1.

[0077] The functional configuration of the communication environment evaluation device 10C is the same as the functional configuration of the communication environment evaluation device 10. However, the communication environment evaluation device 10C evaluates only one representative application from a group of applications and uses this as the evaluation result for applications included in the same application group. Alternatively, the communication environment evaluation device 10C may acquire and evaluate data on multiple applications included in the same application group, regarding them as being of the same type. As other operations are the same as those in embodiment 1, detailed description thereof will be omitted.

[0078] As described above, in the communication environment evaluation system 1C according to the fourth embodiment, multiple applications are classified into application groups, each group consisting of multiple applications with similar communication conditions, and the evaluation unit 15 can evaluate the communication environment for each application group. For example, the evaluation unit 15 can evaluate only one representative application from the application group and use it as an evaluation result for the application group, or it can handle data by treating multiple applications included in the same application group as the same type. In either case, it is no longer necessary to analyze each application individually, and highly similar applications can be evaluated as an application group. Therefore, in addition to the effects of the first embodiment, the present invention has the effect of improving processing efficiency in data acquisition and evaluation.

[0079] Fig. 20 is a diagram showing an example of the configuration of a communication environment evaluation system 1D according to a modified example of the fourth embodiment. While the communication environment evaluation system 1C shown in Fig. 19 has the communication environment evaluation device 10C installed on the ground, Fig. 20 differs in that a communication environment evaluation device 10D is installed on a vehicle 2 and a data aggregation device 9 is installed on the ground. Similar to the communication environment evaluation device 10C, the communication environment evaluation device 10D evaluates the communication environment for each application group, which is a group of multiple applications. Similarly to the communication environment evaluation device 10A of the second embodiment, the communication environment evaluation device 10D is installed on a vehicle 2 and analyzes communication packets received by the on-board wireless communication device 3.

[0080] Fifth Embodiment Fig. 21 is a diagram showing a configuration example of a communication environment evaluation system 1E according to a fifth embodiment. In the communication environment evaluation system 1 according to the first embodiment, the on-board wireless communication device 3 and the trackside wireless communication device 8 are connected via one type of wireless communication network, whereas in the communication environment evaluation system 1E, the on-board wireless communication device 3 and the trackside wireless communication device 8 are connected via multiple types of wireless communication networks.

[0081] The communication environment evaluation system 1E includes an on-board wireless communication device 3 mounted on a vehicle 2, on-board terminals 4-1 and 4-2 corresponding to each of the multiple wireless communication networks, base stations 5-1 and 5-2 corresponding to each of the multiple wireless communication networks, core network devices 6-1 and 6-2 corresponding to each of the multiple wireless communication networks, a connection router 7, a ground wireless communication device 8, and a communication environment evaluation device 10.

[0082] In the example shown in Figure 21, the on-board wireless communication device 3 and the trackside wireless communication device 8 are connected via a route via a 5G network and a route via an LTE (Long Term Evolution) network. Note that the 5G network and the LTE network may be public networks or private networks. In other words, examples of the wireless communication network to be used include a public 5G network, a public LTE network, a private 5G network, a private LTE network, and a private network network designed and managed by the railway operator itself.

[0083] In the communication environment evaluation system 1E, each of a plurality of applications uses a different wireless communication network according to the requirements of the application. The distribution of reception intervals of communication packets for each application acquired by the communication environment evaluation device 10 tends to change depending on the transmission cycle of the application and the scheduling of the wireless communication network used.

[0084] The communication environment assessment device 10 collects learning data for each application in advance, and therefore can perform assessments according to the requirements of the application and the characteristics of the wireless communication network. However, it is not assumed that the wireless communication network to be used will be switched midway depending on the communication environment, priority, etc.

[0085] As described above, in the communication environment evaluation system 1E according to the fifth embodiment, the on-board wireless communication device 3 and the trackside wireless communication device 8 are connected via a plurality of types of wireless communication networks. Each application uses a wireless communication network having characteristics appropriate for the application's requirements. In this way, in addition to the effect of the first embodiment, the communication environment evaluation system 1E according to the fifth embodiment can achieve the effect of being able to evaluate the communication environment for each application even when each application is connected via a different wireless communication network.

[0086] FIG. 22 is a diagram illustrating an example of the configuration of a communication environment evaluation system 1F according to a modified example of the fifth embodiment. While the communication environment evaluation system 1E illustrated in FIG. 21 has a communication environment evaluation device 10 installed on the ground, FIG. 22 differs in that the communication environment evaluation device 10 is installed on a vehicle 2 and a data aggregation device 9 is installed on the ground. In the communication environment evaluation system 1F, the on-board wireless communication device 3 and the ground wireless communication device 8 are also connected via multiple types of wireless communication networks. Each application uses a wireless communication network with appropriate characteristics according to the requirements of the application. Similarly to the communication environment evaluation device 10A of the second embodiment, the communication environment evaluation device 10 in the communication environment evaluation system 1F is also installed on a vehicle 2 and analyzes communication packets received by the on-board wireless communication device 3.

[0087] 23 is a diagram illustrating an example of the hardware configuration of the communication environment assessment devices 10, 10A, 10B, 10C, and 10D according to the first to fifth embodiments and the data aggregating device 9 according to the second embodiment.

[0088] 23 includes a memory 101, a processor 102, a storage 103, a communication I / F (Interface) 104, and an input / output I / F 105. The memory 101, the processor 102, the storage 103, the communication I / F 104, and the input / output I / F 105 are connected by a data transmission path for transmitting and receiving data to and from each other.

[0089] The memory 101 is, for example, a random access memory (RAM) or a read-only memory (ROM). The processor 102 is, for example, a central processing unit (CPU) or a graphics processing unit (GPU). The storage 103 is, for example, a storage device such as a hard disk drive (HDD), a solid state drive (SSD), or a memory card. The storage 103 may also be a memory such as a RAM or a ROM.

[0090] The storage 103 stores programs for realizing the functions of the components included in the communication environment assessment devices 10, 10A, 10B, 10C, and 10D or the data aggregating device 9. The processor 102 executes each of these programs to realize the functions of the components included in the communication environment assessment devices 10, 10A, 10B, and 10C or the data aggregating device 9.

[0091] The communication I / F 104 transmits and receives data to and from an external device. For example, the communication I / F 104 communicates with the external device via a wired communication path.

[0092] The input / output I / F 105 is connected to an input device 106 and a display device 107. The input device 106 is a device that accepts operational inputs from an operator, such as a keyboard, a mouse, a touch sensor, etc. The display device 107 is a device that displays the drawing data processed by the processor 102 on a monitor.

[0093] The configurations shown in the above embodiments are merely examples, and may be combined with other known technologies, or some or all of the embodiments may be combined with each other, or some of the configurations may be omitted or modified within the scope of the invention.

[0094] 1, 1A, 1B, 1C, 1D, 1E, 1F Communication environment evaluation system, 2, 2-1 to 2-n vehicles, 3 On-board wireless communication device, 4 On-board terminal, 5 Base station, 6 Core network device, 7 Connection router, 8 Ground wireless communication device, 9 Data aggregation device, 10, 10A, 10B, 10C, 10D Communication environment evaluation device, 11 Information acquisition unit, 12 Application identification unit, 13 Feature calculation unit, 14 Feature DB, 15 Evaluation unit, 16 Storage unit, 17 Display unit, 101 Memory, 102 Processor, 103 Storage, 104 Communication I / F, 105 Input / output I / F, 106 Input device, 107 Display device.

Claims

1. A communication environment evaluation device comprising: an information acquisition unit that acquires received data of communication packets of multiple applications transmitted between an on-board wireless communication device installed in a vehicle and a ground wireless communication device installed on the ground; an application identification unit that identifies the applications of the communication packets; and an evaluation unit that analyzes the received data of the identified applications and evaluates changes in the communication environment for each application.

2. The communication environment evaluation device according to claim 1, further comprising a feature calculation unit that calculates feature quantities indicating characteristics of the communication packets from the received data, wherein the evaluation unit analyzes the communication packets based on the feature quantities calculated by the feature calculation unit.

3. The communication environment assessment device according to claim 2, further comprising a feature database that stores the feature calculated by the feature calculation unit for each classification based on the received data.

4. The communication environment assessment device according to claim 3, characterized in that the feature database stores, for each category, a trained model for outputting an assessment result of the communication environment from the feature.

5. The communication environment evaluation device described in claim 4, characterized in that the evaluation unit obtains the trained model corresponding to the classification of the received data of the communication packet from the feature database, and evaluates the communication environment using the features of the communication packet and the obtained trained model.

6. The communication environment evaluation device according to claim 1, further comprising: a memory unit that stores the evaluation results of the evaluation unit; and a display unit that displays the evaluation results on a display screen for each classification based on the received data, based on the evaluation results stored in the memory unit.

7. The communication environment evaluation device according to claim 1, wherein the communication environment evaluation device is installed on the ground, and the information acquisition unit acquires the received data of the communication packet transmitted by the on-board wireless communication device and received by the ground wireless communication device.

8. The communication environment evaluation device according to claim 1, which is installed in the vehicle, characterized in that the information acquisition unit acquires the received data of the communication packet transmitted by the ground wireless communication device and received by the on-board wireless communication device.

9. A communication environment evaluation device as described in any one of claims 1 to 8, characterized in that the information acquisition unit acquires the received data including at least one of information indicating the type of application, information indicating the reception time of the communication packet, and vehicle information indicating the characteristics of the vehicle in which the on-board wireless communication device that transmits the communication packet is installed, and the application identification unit classifies the communication packet based on the received data.

10. The communication environment evaluation device according to claim 9, wherein the vehicle information includes at least one of the location information of the vehicle, the vehicle conditions of the vehicle, and the operating conditions of the vehicle.

11. The communication environment evaluation device described in claim 6, characterized in that the display unit displays the evaluation results for the application for which the evaluation results are to be displayed, the vehicle conditions of the vehicle in which the on-board wireless communication device that transmits the communication packets is installed, and the operating conditions of the vehicle, for each time period and for each section in which the vehicle is located.

12. The communication environment assessment device described in claim 2, further comprising a feature database that groups the features calculated by the feature calculation unit for each classification based on the values ​​of the received data by clustering into data with high similarity, and stores trained models that correspond to the groups and are used to output assessment results of the communication environment from the features generated for each classified group.

13. The communication environment evaluation device according to claim 1, characterized in that the plurality of applications are classified into application groups that group together a plurality of applications with similar communication conditions, and the evaluation unit evaluates the communication environment for each of the application groups.

14. A communication environment evaluation system for evaluating the communication environment of communication packets of multiple applications transmitted between an on-board wireless communication device installed in a vehicle and a ground wireless communication device installed on the ground, comprising a communication environment evaluation device having: an information acquisition unit that acquires received data of communication packets of multiple applications transmitted between the on-board wireless communication device and the ground wireless communication device; an application identification unit that identifies the application of the communication packets; and an evaluation unit that analyzes the received data of the identified application and evaluates changes in the communication environment for each application.

15. The communication environment evaluation system described in claim 14, further comprising: a data aggregation device that is installed on the ground and aggregates the evaluation results transmitted from the plurality of communication environment evaluation devices, and the communication environment evaluation device is mounted on each of the plurality of vehicles.

16. The communication environment evaluation system according to claim 14, characterized in that the on-board wireless communication device and the ground wireless communication device are connected by a plurality of types of wireless communication networks.

17. A communication environment evaluation method comprising the steps of: acquiring received data of communication packets of a plurality of applications transmitted between an on-board wireless communication device installed on a vehicle and a ground wireless communication device installed on the ground; identifying the applications of the communication packets; and analyzing the received data of the identified applications and evaluating the communication environment for each application.

Images