Database management system and communication system

Abstract

To provide a database management system and a communication system using the same capable of stably and efficiently performing a search even when search target data does not exist in a narrow area.SOLUTION: In a hierarchical database 1a of a database management system 1, at least one starting point intermediate database includes a local reference data table and a local master table. The local reference data table includes: a search path leading to registration data stored in a lower layer side intermediate database directly connected to an intermediate database including this table, a directly lower layer database being a terminal end database, an intermediate database indirectly connected through the directly lower layer database, or the terminal end database; and a reference value indicating an attribute of a group obtained by grouping the registration data by related classification information. The local master table includes a tree structure of databases with the starting point intermediate database as a starting point and classification information for grouping.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a database management system and a communication system.

Background Art

[0002] The fifth-generation mobile communication system (5G), which has recently been launched globally, has enhanced functions such as even lower latency and a large number of simultaneous connections compared to the fourth-generation mobile communication system (4G). Using the functions of this 5G, the operation of a database for efficiently collecting and distributing a vast amount of diverse sensor information obtained from factories and automobiles is being considered.

[0003] As such a database, a database used for Mobility as a Service (MaaS) can be cited. MaaS is a technology that is expected to have an impact on the national life, economy, industry, etc. of our country in the future. For its business expansion, time-series actual data having a large dynamic range in terms of time and space obtained by automobiles is required. If a vast amount of data targeted at such mobile bodies (such as automobiles) and distributed across the network is accumulated in a non-hierarchical database only by a conventional cloud-based method, a huge amount of computing resources and storage are required for construction, update, and testing. Therefore, it is considered that the energy consumption is large, the data arrival time to the terminal is long, and it is difficult to provide an efficient service.

[0004] For example, as a method of efficiently searching for database resources distributed across a network, and efficiently searching for highly valuable data without imposing a large load on the network or database, a search method using an answer aggregation device has been proposed (see Japanese Patent Application Laid-Open No. 2000-235583). In the above search method, first, a search process is performed on a database located in a narrow area, and an answer counter counts the answer data accumulated in the answer aggregation device, that is, the number of valid pieces of information. Then, when it is determined that a predetermined number of answer data cannot be obtained from this narrow area, a wider range of search is performed from the search device via a communication device.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] In the above conventional search method using an answer aggregation device, when the search target data does not exist in a narrow area, a wider area search is performed, which requires a search via a network. As the number of searches via the network increases, the time required for the search increases dramatically. Therefore, it cannot be said that the above conventional search method can stably and efficiently perform searches.

[0007] The present invention has been made based on the above circumstances, and an object thereof is to provide a database management system capable of stably and efficiently searching for registered data and a communication system using this database management system.

Means for Solving the Problems

[0008] A database management system according to one aspect of the present invention includes a hierarchical database having 11 main databases and a plurality of terminal databases indirectly connected to the main databases via one or more intermediate databases, and a search unit that searches for matching data that matches a search condition from registered data registered in the hierarchical database. The hierarchical database has at least one starting intermediate database as the intermediate database, at least the terminal databases store a plurality of registered data, the starting intermediate database has a local reference data table and a local master table, the local reference data table is a directly lower-layer database that is a lower-layer intermediate database or a terminal database directly connected to the intermediate database having this local reference data table, or a search path to the registered data stored in an intermediate database or a terminal database indirectly connected via this directly lower-layer database, and a reference value indicating an attribute of a group obtained by grouping the registered data by related classification information. The local master table has a tree structure of databases starting from the starting intermediate database and classification information for the grouping.

[0009] A communication system according to another aspect of the present invention includes one or more local edge networks, a cloud data center, and the database management system of the present invention. The local edge network includes one or more mobile base stations connected by a common wireless interface. The cloud data center is configured to be communicable with at least one of the mobile base stations. The starting intermediate database is arranged in the mobile base station.

Advantages of the Invention

[0010] The database management system of the present invention and the communication system using this database management system can stably and efficiently search for registered data.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Mode for Carrying Out the Invention

[0012] [Explanation of Embodiment of the Present Invention] First, the embodiments of the present invention will be listed and explained.

[0013] A database management system according to an aspect of the present invention includes a hierarchical database having one main database and a plurality of end databases indirectly connected to the main database via one or more intermediate databases, and a search means for searching for matching data that matches a search condition from registered data registered in the hierarchical database. The hierarchical database has at least one starting intermediate database as the intermediate database, at least the end databases store a plurality of registered data, the starting intermediate database has a local reference data table and a local master table, the local reference data table is a directly lower-layer database which is an intermediate database or an end database on the lower layer side directly connected to an intermediate database having this local reference data table, or the registered data stored in an intermediate database or an end database indirectly connected via this directly lower-layer database, and a reference value indicating an attribute of a group obtained by grouping the registered data by related classification information. The local master table has a tree structure of databases starting from the starting intermediate database and classification information for the grouping.

[0014] In the database management system, the local master table has a tree structure of databases starting from the starting intermediate database. Therefore, since only the databases included in the tree structure can be targeted for search, the databases to be searched are restricted, and it is possible to prevent the search target from covering the entire hierarchical database. Also, in the database management system, since the end databases store a plurality of registered data, it is possible to select an end database including registered data that can be a search target and freely set the tree structure, so that a tree structure considering the search hit rate and network load can be selected. Therefore, the database management system can stably and efficiently search for registered data.

[0015] The hierarchical database has an intermediate database on the lower layer that is directly or indirectly connected to the starting intermediate database, and the lower-layer intermediate database has a local reference data table and a local master table. When the intermediate databases are included in a tree structure starting from the starting intermediate database in this way, by providing the intermediate databases with local reference data tables and local master tables as well, registered data can be retrieved stably and efficiently.

[0016] A communication system according to another aspect of the present invention includes one or more local edge networks, a cloud data center, and the database management system of the present invention. The local edge network includes one or more mobile base stations connected by a common wireless interface, the cloud data center is configured to be communicable with at least one of the mobile base stations, and the starting intermediate database is disposed in the mobile base station.

[0017] The communication system includes the database management system of the present invention, and the starting intermediate database is disposed in the mobile base station. Since the mobile base station governs a local edge network within a certain geographical range, by disposing the starting intermediate database in the mobile base station, the intermediate databases and the terminal databases directly or indirectly connected to the starting intermediate database are also those disposed within the local edge network governed by the mobile base station. Therefore, a search targeting the starting intermediate database is closed within the local edge network. Accordingly, the communication system can suppress the use of wide-area network communication by using the database management system of the present invention and disposing the starting intermediate database in the mobile base station, so that registered data can be retrieved stably and efficiently.

[0018] As the maximum value of the distance between the above-mentioned starting intermediate database and the terminal database directly or indirectly connected to this starting intermediate database, it is preferably 5 km or more and 20 km or less. Generally, the average daily movement distance of a person is about 8 km to 12 km within the living area, and in many cases, information within this living area is required. By setting the above maximum value within the above range, one starting intermediate database can cover this living area, so the use of wide-area network communication can be further suppressed, and the search efficiency can be improved.

[0019] The above local edge network includes one or more moving bodies, the above moving bodies have sensors and the above terminal database, the above database management system has a registration means for registering the sensor information acquired by the above sensors as the above registration data in the above terminal database possessed by the above moving bodies, the registration data registered by the above registration means has actual data corresponding to the above sensor information and a plurality of labels that can be used as the above classification information, and when the above registration means registers the above registration data in the above terminal database, it is preferable to assign the values of the above plurality of labels predetermined according to the above actual data to the above registration data. By assigning a plurality of labels that can be used as the above classification information to the above registration data in this way, the search and management of the registration data can be facilitated.

[0020] The above database management system is provided with a prediction means for predicting the energy consumption per unit time and the value index of the registration data during the operation of the communication system, and it is preferable that the above registration means changes the value of the above label of the above registration data based on the prediction of the above prediction means. By changing the value of the above label of the registration data based on the prediction of the above prediction means in this way, it is possible to suppress the local concentration of energy consumption and the depletion of the hardware resources of the database.

[0021] The classification information includes items such as the retention time after registration in the terminal database, the collection area of the sensor information, and the urgency of the actual data corresponding to the sensor information. The plurality of labels include slicing labels that include slicing for end-to-end selection and connection of network functions and database functions corresponding to the application using the search means. When the registration means registers the registration data in the terminal database, the value of the slicing label may be determined according to the items. By determining the value of the slicing label in this way, it is possible to easily provide more optimal slicing for each application.

[0022] [Details of Embodiments of the Present Invention] Hereinafter, a database system according to an embodiment of the present invention will be described with appropriate reference to the drawings.

[0023] The database management system 1 shown in FIG. 1 includes a hierarchical database 1a, a search means 1b, a prediction means 1c, and a registration means 1d.

[0024] The hierarchical database 1a has one main database 10 and a plurality of terminal databases 30 indirectly connected via the main database 10 and one or more intermediate databases 20. The hierarchical database 1a also has at least one starting intermediate database 21 as the intermediate database 20. Note that there may be a terminal database 30 directly connected to the main database 10.

[0025] Further, at least the terminal database 30 stores a plurality of registration data. In this embodiment, for simplicity of explanation, hereinafter, it is assumed that all the registration data are stored in the terminal database 30, but it does not prevent a part of the registration data from being stored in the main database 10 or the intermediate database 20. The present invention functions in the same manner even if some of the registration data are stored in the main database 10 or the intermediate database 20.

[0026] The database management system 1 is used in the communication system 100 shown in FIG. 2. The communication system 100 includes one or more local edge networks 100a, a fog network 100b, a cloud network 100c, and the database management system 1. In FIG. 2, only the main database 10, the intermediate database 20, the origin intermediate database 21, and the terminal database 30 of the database management system 1 are illustrated.

[0027] The local edge network 100a includes one or more mobile base stations 110 connected by a common wireless interface. The fog network 100b and the cloud network 100c are configured to be communicable with at least one mobile base station 110 via a gateway, and are also configured to be communicable with each other via a gateway between the fog network 100b and the cloud network 100c.

[0028] The communication system 100 can constitute, for example, a Mobility as a Service (MaaS) in-vehicle data comprehensive utilization platform (hereinafter, also referred to as the "MaaS platform"). In the MaaS platform, for example, physical data such as outside air temperature and driver body temperature, abstract data such as road congestion and driver health status, in-vehicle data, image data such as scenery data, and abstract data such as road congestion and driver health status are acquired from vehicles for the purpose of taxis, delivery services, rental cars, carsharing, mobile sales, etc. The acquired data is stored in an edge database (terminal database 30), and after being processed, data for referring to those data is stored in the database (intermediate database 20) of the fog network 100b and the database (main database 10) of the cloud network 100c. Service providers and users use the data accumulated in each database as one registered data, for example, for in-vehicle services such as monitoring inside and outside the vehicle and monitoring the driver's physical condition, and for driver analysis (driving, biometrics, behavior), passenger analysis (attributes, behavior, emotions), road and surrounding situation analysis, vehicle movement line analysis, etc.

[0029] It has become unrealistic to collect all the vast in-vehicle sensor information, including images and videos distributed in time and space such as those of a MaaS platform, in the cloud network 100c. In the communication system 100, the sensor information collected by each vehicle is accumulated and utilized as registered data in the local edge network 100a. Then, each registered data accumulates necessary information in the fog network 100b or the cloud network 100c based on its distribution requirements. The database management system 1 centrally manages the whole including the cloud network 100c, enables comprehensive search for all vehicles, and utilizes in-vehicle hardware resources in a MaaS platform to which a large number of vehicles are connected, reducing the utilization of the cloud network 100c and data communication costs. Hereinafter, the communication system 100 will be described in detail.

[0030] 〔Local Edge Network〕 The local edge network 100a includes one or more mobile bodies 120 in addition to the mobile base station 110. Also, the local edge network 100a can include a roadside unit 130.

[0031] The mobile base station 110, the plurality of mobile bodies 120, and the roadside unit 130 are connected by the above-described common wireless interface. As the wireless interface, 5G, local 5G, 4G, DSRC (Dedicated Short Range Communications), LPWA (Low Power Wide Area), WiFi, IEEE802.11 series, an interface for ITS (Intelligent Transport System) communication, etc. can be used.

[0032] <Mobile Base Station> The mobile base station 110 is a base station that generalizes and controls the geographical range in which wireless communication of the local edge network 100a can be performed. Specifically, the mobile base station 110 performs, for example, data management processing and distribution control processing of a plurality of mobile bodies 120.

[0033] The mobile base station 110 includes a main base station 111 and a local base station 112, and can communicate with other mobile base stations 110 of the other local edge network 100a, or with both or one of the fog network 100b and the cloud network 100c, and constitutes the backbone network of the communication.

[0034] (Main base station) The main base station 111 oversees and controls the entire geographical area. The main base station 111 has a communication function and a processor. The origin intermediate database 21 is arranged in the main base station 111 which is the mobile base station 110. Also, the main base station 111 may have sensors. By having sensors in this way, the information obtained at the main base station 111 can be used as registered data.

[0035] (Local base station) The local base station 112 belongs to the lower layer of the main base station 111 and subdivides and oversees the geographical area overseen by the main base station 111. Therefore, usually, a plurality of local base stations 112 are provided. Conversely, when there is no need to subdivide the above geographical area, the local base station 112 can be omitted, and the mobile base station 110 is composed of only the main base station 111 and directly oversees the entire above geographical area.

[0036] The local base station 112 has a communication function and a processor. An intermediate database 20 is arranged in the local base station 112. Also, the local base station 112 may have sensors. By having sensors in this way, the information obtained at the local base station 112 can be used as registered data. The local base station 112 can be, for example, a local 5G base station.

[0037] <Mobile body> The plurality of mobile bodies 120 are devices that move within the data sharing area (geographical area where wireless communication can be performed) managed by the local edge network 100a.

[0038] The plurality of moving bodies 120 may include a vehicle 121, a drone 122, and a mobile device 123. The plurality of moving bodies 120 may also include other mobile robots, bicycles, etc. Among these, it is preferable that the moving body 120 includes the vehicle 121. The database management system 1 can be particularly preferably used in a system including the vehicle 121, for example, a V2X communication system used in autonomous driving or the MaaS business. "V2X" is an abbreviation for Vehicle to X, and is a technology that collectively refers to the connection and mutual cooperation between the vehicle 121 and something else (other vehicles 121, pedestrians, infrastructure, network, etc.). This V2X is indispensable for, for example, the autonomous driving of the vehicle 121. In the above V2X communication system, it is necessary to handle time-series data having a large dynamic range in terms of time and space acquired by the vehicle 121. The database management system 1 can handle such time-series data particularly efficiently. Hereinafter, the description will be made on the premise that the moving body 120 includes the vehicle 121, but it does not mean that the database management system 1 regards the vehicle 121 as an essential component as the moving body 120.

[0039] The plurality of moving bodies 120 have a communication function and a processor. Further, the moving body 120 has a sensor 120a and a local database 30. Note that a "sensor" refers to an element or device that detects a phenomenon in the real world and replaces the detected phenomenon with a signal that can be processed by a computer. In FIG. 2, a configuration in which the drone 122 has a sensor 120a is illustrated, but although not illustrated for other moving bodies 120, they similarly have a sensor 120a.

[0040] Each moving body 120 processes the information collected by the sensor 120a by the above-mentioned processor and stores it in the local database 30. Examples of the above information include any natural phenomena such as weather and temperature, phenomena caused by artificial objects such as traffic jams, human vital information such as heart rate, respiratory rate, and blood pressure, approach of a person to a vehicle, information on fallen objects on the road, time, moving speed, position, and any other things that can be detected by a sensor.

[0041] <Roadside device> The roadside device 130 is a device installed on the road for controlling and monitoring the traffic of the vehicle 121. The roadside device 130 includes traffic lights, street lights, etc., and is connected to the local base station 112 through the wireless interface.

[0042] The roadside device 130 has a communication function. Also, the roadside device 130 preferably has a database (intermediate database 20 or terminal database 30), a processor, and a sensor. By having these configurations, the information acquired by the roadside device 130 can be used as registered data.

[0043] Also, the roadside device 130 may have a function as a mobile base station 110. That is, the roadside device 130 can function as a moving body 120 that does not move, a mobile base station 110, or both.

[0044] 〔Fog network〕 The fog network 100b has a router 140 and performs data transfer between the local edge network 100a and the cloud network 100c. An intermediate database 20 is arranged in the fog network 100b.

[0045] 〔Cloud network〕 The cloud network 100c has a router 140 and a server 141. A main database 10 is arranged in the cloud network 100c. The cloud network 100c is configured to be communicable with at least one mobile base station 110 directly or via the fog network 100b.

[0046] The cloud network 100c collects information from the mobile base stations 110 of each local edge network 100a via the router 140. The server 141 manages the collected information, and the main database 10 stores the above information under the control of the server 141. Note that the information stored in the main database 10 is the necessary information among the information collected and processed by the mobile base station 110. Not all data is necessarily stored.

[0047] When the communication system 100 is used in a MaaS platform, a large number of vehicles 121 are not stand-alone single functions, but are interconnected, for example, by the massive terminal simultaneous connection function of a 5G network, to achieve the effect of clustering, so that the vehicles 121 can be arranged according to the demand for the mobility of information and energy. As a result, it becomes possible to apply to a wide range of fields, such as power supply (safety, security, energy) from the vehicle 121 during disasters, virtual meeting places (life, living) using large screens and high-definition images between the vehicle 121 and between the vehicle 121 and the community center, virtual exhibition halls (tourism), etc.

[0048] 〔Database Management System〕 According to the Ministry of Internal Affairs and Communications, in the case of 2005, the average daily travel distance per person (living area) was 12 km in urban areas and 8 km in rural areas. Also, the traffic flow density is 10 vehicles / km on average nationwide 2 and 128 vehicles / km in Tokyo 2It is so. Although the database management system 1 can handle data on a national scale, for example, users often need information within their living area. In the database management system 1, by targeting only the intermediate database 20 and the terminal database 30 included in the tree structure of the database starting from the starting intermediate database 21, it aims to efficiently handle the data within this living area. That is, when the starting intermediate database 21 is arranged at the main base station 111 and a tree structure of the database including the terminal database 30 in which the registered data within the above living area is stored is constructed, the management and search of the registered data can be generally performed through communication within the local edge network 100a, so that large-capacity and high-speed database processing can be enabled. The database management system 1 can be said to be a regionally focused database that aims to localize data production and consumption.

[0049] For example, a radius of 12 km and a traffic flow density of 128 vehicles / km 2 is assumed, and if the number of registered data per vehicle 121 is 1000 and 10% of them are active vehicles 121, the number of registered data is estimated to be approximately 6 million (=π×12 2 ×128×1000×0.1). The data scale handled by the database starting from the starting intermediate database 21 of the database management system 1 is preferably in the scale of this class, that is, the total number of registered data is 1 million or more and 10 million or less.

[0050] <Search means> The search means 1b searches for matching data that meets the search conditions from the registered data registered in the hierarchical database 1a.

[0051] The search means 1b identifies the target starting intermediate database 21 according to the search conditions, and searches for the registered data registered in the intermediate database 20 and the terminal database 30 included in the tree structure of the database starting from this starting intermediate database 21. That is, the search means 1b does not search the intermediate database 20 and the terminal database 30 not included in the tree structure of the starting intermediate database 21. Note that the above tree structure is specified by the local master table 20b of the starting intermediate database 21 described later.

[0052] <Prediction means> The prediction means 1c predicts the energy consumption per unit time of the registered data during the operation of the communication system 100 and the value index of the registered data.

[0053] The energy consumption per unit time of the registered data refers to the amount of energy consumed by the communication system 100 when accessing the target registered data and managing the above registered data (for example, data update, etc.). Also, the value index of the registered data is the relative value of a plurality of registered data, and can be, for example, the number of accesses. It is advisable to add a time weight to the number of accesses. That is, it is advisable to weight so that the relative value increases as the number of recent accesses increases.

[0054] The value of the registered data tends to decrease as time elapses after registration. For example, since road debris or accidents that occur suddenly are resolved in a relatively short time, their value often decreases significantly with respect to time. Data about the weather also changes every moment, so its value tends to decrease with time. On the other hand, for example, if it is newly opened road information, its value is maintained for a relatively long time. Thus, the energy consumption per unit time of the registered data and the value index of the registered data can change depending on the content of the registered data or the changes in the current situation. In the prediction means 1c, the energy consumption per unit time of this registered data and the value index of the registered data are predicted according to the content of the registered data.

[0055] As the prediction means 1c, a known AI module can be used. That is, a learning model is constructed from the past energy consumption per unit time for the registered data and the value index of the registered data, and the future energy consumption per unit time and the value index of the registered data can be predicted according to this learning model.

[0056] When the prediction means 1c predicts the existence of, for example, the terminal database 30 with a large energy consumption per unit time, the increase in the energy consumption of a specific terminal database 30 can be suppressed by methods such as dispersing and arranging the registered data of this terminal database 30 to other terminal databases 30. Also, when it is expected that the registered data of the database will become too large, the parameters of the registered data may be adjusted by the registration means 1d described later, and the amount of active registered data may be limited according to the value index of the registered data.

[0057] <Registration means> The registration means 1d registers the sensor information acquired by the sensor 120a of the mobile body 120 as registered data in the terminal database 30 possessed by the mobile body 120.

[0058] The above registration data has actual data corresponding to the above sensor information and a plurality of labels that can be used as classification information described later. Specifically, as the above plurality of labels, the above registration data includes a first label (first layer) including service provider information that provides a service using the above actual data, a second label (second layer) including the collection area of the above sensor information and the holding time and update time after registration in the terminal database 30, a third label (third layer) including the deliverable area and deliverable users of the above registration data, a fourth label (fourth layer) including the communication quality when the above registration data is delivered, and a fifth label (fifth layer) including slicing that end-to-end selects and connects network functions and database functions corresponding to an application using the search means 1b. When the registration means 1d registers the above registration data in the terminal database 30, it assigns the values of the first to fifth labels determined in advance according to the actual data (that is, the type of actual data, the type of the acquired sensor 120a, the numerical value of the actual data, etc.) to the above registration data. By assigning the values of the first to fifth labels to the above registration data in this way, the search and management of the registration data can be facilitated.

[0059] The above first layer corresponds to the usage requirements of the registration data. Service providers include MaaS operators, transportation operators, logistics service providers, etc. Only the service providers registered here and the users who receive services from this service provider can use this registration data.

[0060] The second layer above is a collection and accumulation requirement. The collection area of sensor information is one of the important pieces of information in the database management system 1 aiming to achieve local production and local consumption of registered data. Based on this collection area, it is possible to determine which tree structure of the intermediate database 21 of the starting point should belong to. Also, as described above, the value of registered data tends to decline over time, and the rate of decline varies depending on the type of registered data. Therefore, it is effective to predetermine the retention period according to the registered data and prevent the depletion of the hardware resources of the database by deleting the data after the retention period has elapsed. The retention period of the second layer above can be used for this purpose, for example. Among the registered data, there is also data whose freshness is maintained by updating the content, such as weather data. For such registered data, the update time may be used instead of the retention period. That is, the registered data is refreshed every time the update time elapses.

[0061] The third layer above is a distribution requirement. By restricting the distribution area of the registered data, the wide-area network traffic with the fog network 100b and the cloud network 100c can be reduced. Also, by restricting the distributable users of the registered data, it is possible to prevent communication from being locally concentrated during the search for the registered data.

[0062] The fourth layer above is a quality requirement. The communication quality includes the radio wave reception level, transmission delay time, packet transmission error rate, transfer speed, and the like. By grasping the communication quality when communicating the registered data, it becomes easier to estimate the load on the communication network during the search for the registered data, and it is also possible to take measures such as moving the registered data to another terminal database 30 so as to improve this communication quality.

[0063] The fifth layer above is a slicing requirement. By adding this requirement, it is possible to facilitate the setting of slicing according to the end-to-end application.

[0064] When the registration means 1d registers the registration data in the terminal database 30, it is preferable to assign the values of the first to fifth labels, and to update the values of the first to fifth labels of the registration data based on the prediction of the prediction means 1c. For example, when it is predicted that the hardware resources of the database will be exhausted, for the registration data with a low value index of the registration data predicted by the prediction means 1c, by shortening the retention time of the second layer, the deletion of the registration data can be promoted. This update may be performed on the registration data that has already been registered, or the registration data to be registered may be registered with the updated value, or both may be performed. By updating the values of the first to fifth labels of the registration data based on the prediction of the prediction means 1c in this way, it is possible to suppress the local concentration of the energy consumption and the exhaustion of the hardware resources of the database.

[0065] <Hierarchical database> In the hierarchical database 1a, the lower limit of the maximum value of the distance between the starting intermediate database 21 and the terminal database 30 directly or indirectly connected to the starting intermediate database 21 is preferably 5 km, more preferably 10 km. On the other hand, the upper limit of the maximum value of the distance is preferably 20 km, more preferably 15 km. If the maximum value of the distance is less than the lower limit, one starting intermediate database 21 may not be able to cover the living area, the usage amount of the wide area network communication may increase, and the search efficiency may decrease. Conversely, if the maximum value of the distance exceeds the upper limit, the registration data outside the living area, that is, the registration data that is not required, may also be the search target, so the search efficiency may decrease.

[0066] The starting intermediate database 21, the lower-layer intermediate database 20 and the terminal database 30 directly or indirectly connected to the starting intermediate database 21 are preferably included in the same or adjacent local edge network 100a. By configuring in this way, the management and search of the registration data can be performed by short-distance communication centered within the local edge network 100a, so that a large-capacity and high-speed database process can be enabled.

[0067] <Data Configuration of Database System> In the hierarchical database 1a, the starting intermediate database 21 has the local reference data table 20a and the local master table 20b shown in FIG. 3. Further, the terminal database 30 has a terminal actual data table 30a and a terminal master table 30b. Note that in FIG. 3, for simplicity, an example is shown in which the starting intermediate database 21 is directly connected to the main database 10, and one terminal database 30 is directly connected to the starting intermediate database 21. However, an intermediate database 20 may be arranged between the main database 10 and the starting intermediate database 21 or between the starting intermediate database 21. Also, the number of intermediate databases 20 or terminal databases 30 connected to the main database 10 and the starting intermediate database 21 may be plural. In this case, the lower-layer intermediate database 20 directly or indirectly connected to the starting intermediate database 21 also has a local reference data table 20a and a local master table 20b.

[0068] The local reference data table 20a and the local master table 20b will be described with reference to FIGS. 4 and 5. The local reference data table 20a and the local master table 20b are tables starting from the starting intermediate database 21. FIGS. 4 and 5 show the lower-layer intermediate database 20 and the terminal database 30 directly or indirectly connected to the starting intermediate database 21 downstream of the starting intermediate database 21. The local reference data table 20a and the local master table 20b are deployed in the databases shown in FIGS. 4 and 5.

[0069] In the database management system 1, one starting intermediate database 21 does not prevent having a plurality of pairs of local reference data tables 20a and local master tables 20b with different classification information to be described later. In this case, it is not necessary for the intermediate databases 20 and the terminal databases 30 belonging to the downstream side of the starting intermediate database 21 to match for different classification information. That is, if the classification information is different, the lower-layer intermediate databases 20 and the terminal databases 30 directly or indirectly connected to the same starting intermediate database 21 can be different.

[0070] Also, it does not prevent the existence of a plurality of starting intermediate databases 21. In this case, for each starting intermediate database 21, a unique local reference data table 20a, a local master table 20b, and a terminal master table 30b are generated.

[0071] Furthermore, it does not prevent one intermediate database 20 and one terminal database 30 from being directly or indirectly connected downstream of a plurality of starting intermediate databases 21. Whether each intermediate database 20 and the terminal database 30 belong to which starting intermediate database 21 can be grasped, for example, by which local master table 20b or terminal master table 30b corresponding to which starting intermediate database 21 is being deployed.

[0072] In FIGS. 4 and 5, the numbers shown in ( ) for each database indicate the database ID of each database. For example, the database ID of the starting intermediate database 21 is 2100. This database ID is a unique numerical value, and there is no other database having the same numerical value as a certain database. Therefore, a database can be specified by indicating the database ID.

[0073] (Local Reference Data Table) The local reference data table 20a has a search path leading to registered data stored in the immediate lower-level database, which is the intermediate database 20 or the terminal database 30 on the lower layer directly connected to the intermediate database 20 having this local reference data table 20a, or the intermediate database 20 or the terminal database 30 indirectly connected via this immediate lower-level database, and a reference value indicating the attribute of a group obtained by grouping the above registered data according to classification information related to the registered data.

[0074] The above path information indicates the path up to the terminal database 30 where the actual data (registered data) corresponding to the reference value exists. For example, the first row "2400 3300" of the local reference data table 20a in the starting intermediate database 21 indicates that there is a path from the starting intermediate database 21 having this local reference data table 20a, via the intermediate database 20 with database ID 2400, to the terminal database 30 with database ID 3300. Also, as reference values for this path information, it is shown that there are two data with reference value = 1 and reference value = 3. The above path information may be the entire path from the starting intermediate database 21 to the terminal database 30, but as shown in FIG. 4, it is preferable to use only the path on the downstream side of the intermediate database 20 having the local reference data table 20a as information. By using only the path on the downstream side as information in this way, the capacity of the local reference data table 20a can be reduced.

[0075] The above reference values indicate classifications based on actual data and are numerical values determined based on the master table described later. In the example of FIG. 4, where the actual data values range from 0 to 100, they are classified into 0 to 30 (reference value 1), 31 to 60 (reference value 2), and 61 to 100 (reference value 3) (see the master table in FIG. 5). That is, the actual data is grouped by relevant classification information (reference value 1, reference value 2, or reference value 3). When classified in this way, while 7 bits are required for the actual data to represent 0 to 100 in binary, it can be stored in 2 bits (which can represent 1 to 3), so the actual data can be compressed to 2 bits / 7 bits = 29%. It is considered that this compression rate improves as the actual data becomes more complex. Particularly in the case of image data, etc., the capacity of the actual data itself is large, and when the registration means 1d assigns the values of the first to fifth labels to the above registration data as described above, the capacity of these assigned data is also added. Therefore, in the data actually handled, by using reference values in this way, the data can be greatly compressed, enabling high-speed search and reduction of the data storage capacity.

[0076] The local reference data table 20a appears to be a single database by itself, but in reality, only the terminal database 30 holds the actual data. The local reference data table 20a is a mechanism that makes the distributed hierarchical databases 1a cooperate to appear as one integrated single table. That is, the registration data, which is the actual data, is stored only in the terminal actual data table 30a, and the local reference data table 20a accumulates data including path information and reference values, which are classification information of the data necessary for searching, as the minimum necessary data for searching and acquiring the terminal actual data table 30a. Thereby, the three layers of the starting intermediate database 21, the intermediate database 20 on its downstream side, and the terminal database 30 are coordinated, and when searching for registration data, the necessary data can be provided to the user as one virtual database. In other words, by enabling the information of the registration data distributed and stored in the hierarchical database 1a to be handled as one database collectively, high-speed search and reduction of the storage capacity are made possible.

[0077] The local reference data table 20a has actual data registered in the terminal database 30 by the registration means 1d and is updated when the registered data is generated. Hereinafter, the procedure will be described with reference to FIG. 4.

[0078] For example, when new actual data "23" and "62" occur in the terminal database (3300) of FIG. 4, these actual data are recorded in the terminal database 30 together with reference values. The reference values can be determined by the terminal actual master table 30b. That is, since the actual data "23" is in the range "0-30", the reference value is "1", and since the actual data "62" is in the range "61-100", the reference value is "3".

[0079] Since this terminal database (3300) is connected to the lower layer of the intermediate database (2400), the information updated in the terminal database (3300) is transmitted to this intermediate database (2400) as table information 30c as shown in FIG. 4.

[0080] The table information 30c includes information excluding the actual data of the updated terminal actual data table 30a, that is, the database ID (3300) of the terminal database 30 storing the actual data and the reference value.

[0081] In the intermediate database (2400) that receives the table information 30c, the local reference data table 20a is updated based on this information. Further, the intermediate database (2400) adds its own database ID "2400" to the path information to create table information 20c and transmits it to the starting intermediate database 21 which is the immediate upper layer database.

[0082] Furthermore, the start intermediate database 21 receives the table information 20c from the intermediate database (2400) and updates the local reference data table 20a of the start intermediate database 21. Since this local reference data table 20a starts from the start intermediate database 21, all local reference data tables 20a related to the actual data registration newly in the end database (3300) can be updated accordingly.

[0083] (Local master table) The local master table 20b has the database tree structure starting from the start intermediate database 21 and the classification information for the above grouping. The local master table 20b is a table that guarantees the same data structure and data content as the entire database starting from the start intermediate database 21.

[0084] As shown in FIG. 5, the search path stored in the local master table 20b is the same information as the search path of the corresponding local reference data table 20a.

[0085] Also, as shown in FIG. 5, the above classification information has record classification and reference values. The above reference values represent the reference values used in the local reference data table 20a. The record classification indicates the range of actual data values classified into the above reference values. In the start intermediate database 21 and the intermediate database 20, by referring to the local master table 20b, it is possible to determine to which range the actual data belongs from the reference values.

[0086] By providing the above-described local master table 20b in this way, it becomes possible to guarantee the identity of the data structure and data content between the local reference data table 20a included in the intermediate database 20 and the actual database composed of the registered data.

[0087] Regarding the terminal master table 30b in the terminal database 30, since there is no search path below it, the search path is an empty set (data not required). Therefore, the terminal master table 30b has only the above classification information. In the terminal database 30, by referring to the terminal master table 30b, it is possible to determine to which reference value the actual data should be classified.

[0088] Regarding the update of the local master table 20b and the terminal master table 30b, the update information is aggregated in the starting intermediate database 21. For example, when a new terminal database 30 is added along with the addition of registration data, it is transmitted to the starting intermediate database 21 that a new terminal database 30 has been added simultaneously with the update of the local reference data table 20a.

[0089] The starting intermediate database 21 distributes the updated local master table 20b to the intermediate database 20 or the terminal database 30 in the immediate lower layer. In the example of FIG. 5, the starting intermediate database 21 distributes to three databases: the intermediate database (2400), the intermediate database (2410), and the terminal database (3340). The corresponding database can be determined by referring to the path information of the local master table 20b in the starting intermediate database 21. That is, the database having the database ID described at the head of the path described in the path information is the intermediate database 20 or the terminal database 30 in the immediate lower layer, and it is advisable to distribute to these.

[0090] At this time, the route information to be distributed to each database is only those including the corresponding intermediate database 20 or the terminal database 30 from the perspective of data volume reduction, and the database ID at the beginning (of the corresponding intermediate database 20 or terminal database 30) may be omitted. For example, as the route information of the local master table 20b distributed to the intermediate database (2400), among the route information of the local master table 20b of the starting intermediate database 21, "2400 3300" and "2400 3310" which include the database ID 2400 of this intermediate database 20 are the target route information, and "3300" and "3310" obtained by deleting the database ID "2400" of this intermediate database 20 from this become the route information. According to this method, when the transfer destination is the terminal database 30, the route information will always be an empty set. That is, route information is unnecessary in the terminal database 30.

[0091] In addition, when a new local database (a database composed of the starting intermediate database 21, the intermediate database 20 belonging to the downstream side of the starting intermediate database 21, and the terminal database 30) is provided, first this local master table 20b is distributed, and then the registration data is registered in the terminal database 30.

[0092] (Main reference data table and main master table) As shown in FIG. 3, it is preferable that the main database 10 has a main reference data table 10a and a main master table 10b. By providing the main reference data table 10a and the main master table 10b in the main database 10 in this way, it is possible to manage the registration data of the entire hierarchical database 1a beyond the starting intermediate database 21.

[0093] As registration data managed across the entire hierarchical database 1a, for example, it may be the registration data itself managed by a plurality of origin intermediate databases 21 with the same classification information but different jurisdictions, or for example, aggregated and statistical data obtained by aggregating registration data in units of short term (1 day), medium term (1 month), and long term (1 year). The aggregated and statistical data can be accumulated as time series data in a mobile body 120 or the like equipped with a terminal database 30, aggregated, and then the reference value can be distributed to the main database 10 via an intermediate database 20 belonging to the fog network 100b.

[0094] Based on the reference data aggregated in the main database 10, for example, a server 141 performs analysis, learning, anomaly detection, etc. on the usage status and quality of the terminal database 30, and performs overall optimization of the hierarchical database 1a such as the arrangement of each database, slicing setting, data consistency during communication switching, and acceptance of new databases. The results can be shared, for example, by distributing the main master table 10b to each database.

[0095] <Functions of the communication system> FIG. 6 shows the communication system 100 as a functional block diagram. The communication system 100 is a system that transmits and receives information between the local edge network 100a and the cloud network 100c within the mobile communication network 101 and within the local edge network 100a, and is controlled by the mobile core unit 200. Note that the mobile core unit 200 shown in FIG. 6 is a mobile core unit of the fifth generation mobile communication system (5G). The mobile communication network 101 includes a fog network 100b as a relay network to the cloud network 100c.

[0096] As a function of controlling the hierarchical database 1a (main database 10, intermediate database 20, starting intermediate database 21, and terminal database 30), the database management system 1 includes a database operation management unit 40 and an interface unit 50 between databases. Each element constituting the database management system 1 is distributed and arranged within the communication system 100. For example, in the communication system 100 shown in FIG. 3, the terminal database 30 is arranged in the user device 150 within the local edge network 100a, and the intermediate database 20 is arranged in the mobile base station 110 within the local edge network 100a and within the fog network 100b. The main database 10 is arranged in the cloud 160, which is computer resources within the cloud network 100c.

[0097] The database operation management unit 40 is a function of managing the registered data registered in the hierarchical database 1a, and includes the functions of a search means 1b, a prediction means 1c, and a registration means 1d. The database operation management unit 40 can manage the registered data and perform the hierarchical configuration of the database via the interface unit 50 between databases. Also, a part of the functions of the database operation management unit 40 is incorporated within the mobile core unit 200. Note that these arrangements are merely examples and do not mean that the arrangement of the database management system 1 is limited to the configuration shown in FIG. 6.

[0098] In the mobile communication network 101, the user device 150 is located at the communication terminal and includes the mobile body 120 and the roadside unit 130. The U-Plane 201 is a function of setting the routing route between the mobile base station 110 and the cloud 160 and is also a part of the functions of the mobile core unit 200.

[0099] The mobile core unit 200 controls the establishment of communication, etc. The mobile core unit 200 may operate and manage the mobile communication network 101 using an API on the Web.

[0100] When the mobile core unit 200 is implemented with a 5G core, the mobile core unit 200 includes a UDM 202 (Unified Data Management) which is a subscriber information database and manages authentication information and location information, an AUSF 203 (AUthentication Server Function) specialized in authentication processing functions, a PCF 156 (Policy Control Function) for policy control, a PCF 204 (Policy Control Function) for policy control including QoS (Quality of Service) and charging control, a NEF 205 (Network Exposure Funciton) that enables control of network functions from external applications, an NSSF 206 (Network Slice Selection Functions) that selects network slicing functions, an AMF 207 (Access and Mobility management Function) that performs access and mobility control between the user equipment 150 and the mobile base station 110, an SMF 208 (Session Management Function) that controls the U-Plane 201 and performs session management from the mobile base station 110 to the cloud 160 via the fog network 100b, an NRF 209 (Network Repository Function) that performs service registration, authorization, etc.

[0101] Also, the mobile core unit 200 includes a DB management 210 and an energy resource management 211 as applications (APIs) for operating and managing the database management system 1, and manages processing resources and energy resources during collection, distribution, and accumulation of registration data.

[0102] It is preferable that the first to fifth labels assigned by the registration means 1d of the database management system 1 to the registered data be operated in cooperation with the functions of the above-mentioned 5G core. Specifically, the usage requirements of the first label may be such that the use of data is managed in cooperation with the PCF 204 and the NEF 205 based on information such as the service provider ID and the user ID. The collection and accumulation requirements of the second label may be such that the use of data is managed in cooperation with the AMF 207 and the SMF 208 based on the collection area of sensor information and the holding time and update time after registration in the terminal database 30. The distribution requirements of the third label may be such that the use of data is managed in cooperation with the AMF 207 and the SMF 208 based on the information on the deliverable area of the registered data and the deliverable users. The quality requirements of the fourth label may be such that the use of data is managed in cooperation with the PCF 204 based on the communication quality when the registered data is distributed. The slicing requirements of the fifth label may be such that end-to-end slicing corresponding to the end-to-end application is performed in cooperation with the NRF 209 and the NSSF 206.

[0103] <Slicing Selection Function> In the database management system 1, the fifth label of the registered data includes slicing that end-to-end selects and connects network functions and database functions corresponding to the application using the search means 1b. By this slicing function, it is possible to realize stable and efficient search of the registered data by selecting the optimal hierarchical database 1a configuration according to the application for requirements regarding delay, quality, and massive connection. Fig. 7 shows a configuration diagram of the slice selection function.

[0104] Examples of such applications include, as shown in FIG. 7, an ultra-low latency application 171, a large-capacity transmission application 172, a regional cooperation application 173, and the like. Taking the application for the vehicle 121 as an example, the ultra-low latency application 171 includes driver health support, high-security environment support, high-data reliability environment support, ultra-low latency vehicle-to-vehicle communication, and the like. Examples of the large-capacity transmission application 172 include video streaming support, continuous service, data path optimization, battery consumption minimization, and the like. Examples of the regional cooperation application 173 include mobility support, high-security environment support, CO 2 emission optimization, and the like.

[0105] The RAN-based slicing 174 is selected according to the access setting 175 (type of user equipment 150, quality of service, mobility, data traffic, and slice type). Also, the CN-based slicing 176 is selected by the slice matching 177 and performs slicing for the relay line.

[0106] According to the slice type of the application set by the user, the database operation management unit 40 selects and connects network functions and database functions end-to-end, and provides slicing for each application optimized for the user. At that time, as an interface, it enables dynamic changes to the database configuration that allows slicing requirements for each end-to-end service and optimal placement of the hierarchical database 1a.

[0107] When using the slicing selection function, the classification information of the database management system 1 includes the retention time included in the second label of the registered data, the collection area of the sensor information included in the second label, and the urgency item of the actual data corresponding to the sensor information. When the registration means 1d registers the registered data in the terminal database 30, it is preferable to determine the value of the fifth label (slicing label) according to the above items. By determining the value of the fifth label in this way, it is possible to easily provide more optimal slicing for each application.

[0108] <Advantages> The database management system 1 has a tree structure of databases with the local master table 20b as the starting point of the intermediate database 21. Therefore, only the intermediate database 20 and the terminal database 30 included in the tree structure can be searched, so the databases to be searched are limited, and it is possible to prevent the search target from covering the entire hierarchical database 1a. Further, in the database management system 1, since the terminal database 30 stores a plurality of registered data, the terminal database 30 including the registered data that can be the search target can be selected, and the tree structure can be freely set. Therefore, a tree structure considering the search hit rate and the network load can be selected. Therefore, the database management system 1 can stably and efficiently search for registered data.

[0109] The communication system 100 includes the database management system 1 of the present invention, and the origin intermediate database 21 is arranged in the mobile base station 110. Since the mobile base station 110 governs the local edge network 100a within a certain geographical range, by arranging the origin intermediate database 21 in the mobile base station 110, the intermediate database 20 and the terminal database 30 directly or indirectly connected to the origin intermediate database 21 are also those arranged within the local edge network 100a governed by the mobile base station 110. Therefore, the search targeting the origin intermediate database 21 is closed within the local edge network 100a. Accordingly, the communication system 100 can suppress the use of wide area network communication by using the database management system 1 of the present invention and arranging the origin intermediate database 21 in the mobile base station 110, so that the registered data can be stably and efficiently searched.

[0110] [Other Embodiments] The above embodiments do not limit the configuration of the present invention. Therefore, based on the description in this specification and common technical knowledge, it is possible to omit, replace, or add the constituent elements of each part of the above embodiments, and all of them should be interpreted as belonging to the scope of the present invention.

[0111] In the above embodiment, the case where the communication system includes a fog network has been described. However, the fog network is not an essential component. The communication system can also be composed only of a cloud network and a local edge network.

[0112] In the above embodiment, a communication system including vehicles on the premise of a MaaS platform has been described. However, the platform is not limited to MaaS, and other platforms may also be used. Further, the communication system is not limited to a network including vehicles, and can also be used for other networks. The communication system can be used, for example, in enterprise / factory networks, agricultural / fishery networks, etc. where services are provided using local 5G.

[0113] In the above embodiment, the case where the database management system includes a prediction means has been described. However, the prediction means is not an essential component, and a database management system without a prediction means is also within the scope of the present invention. In this case, the registration means does not necessarily have to have a function of changing the values of the first to fifth labels of the registration data. Alternatively, the registration means may change the values of the first to fifth labels of the registration data according to other criteria instead of the prediction by the prediction means.

[0114] In the above embodiment, the case where the registration means assigns the values of the first to fifth labels to the registration data has been described. However, the assignment of the values of the first to fifth labels is not an essential component. The registration means does not necessarily have to assign some or all of these labels to the registration data.

[0115] In the above embodiment, the case where an intermediate database is connected to the lower layer side of the starting intermediate database has been described. However, a configuration in which only a terminal database is connected to the lower layer side of the starting intermediate database is also within the scope of the present invention.

Industrial Applicability

[0116] As described above, the database management system of the present invention and the communication system using this database management system can stably and efficiently search for registration data.

Explanation of Signs

[0117] 1 Database management system 1a Hierarchical database 1b Search means 1c Prediction means 1d Registration means 10 Main database 10a Main reference data table 10b Main master table 20 Intermediate database 20a Local reference data table 20b Local Master Table 20c Table Information 21 Origin Intermediate Database 30 Terminal Database 30a Terminal Actual Data Table 30b Terminal Master Table 30c Table Information 40 Database Operation and Management Department 50 Database Interface Department 100 Communication System 100a Local Edge Network 100b Fog Network 100c Cloud Network 101 Mobile Communication Network 110 Mobile Base Station 111 Main Base Station 112 Local Base Station 120 Mobile Unit 120a Sensor 121 Vehicle 122 Drone 123 Portable Device 130 Roadside Unit 140 Router 141 Server 150 User Equipment 160 Cloud 171 Ultra-Low Latency Application 172 High-Capacity Transmission Application 173 Regional Collaboration Application 174 RAN-Based Slicing 175 Access Setting 176 CN-Based Slicing 177 Slice Matching 200 Mobile Core 201 U-Plane 202 UDM 203 AUSF 204 PCF 205 NEF 206 NSSF 207 AMF 208 SMF 209 NRF 210 DB Management 211 Energy Resource Management

Claims

1. A database management system comprising: a hierarchical database having one main database, and a plurality of end databases indirectly connected to the main database via one or a plurality of intermediate databases; and a search means for searching for matching data matching a search condition from registered data registered in the hierarchical database, the hierarchical database has at least one origin intermediate database as the intermediate database, At least the terminal database stores a plurality of registration data; the origin intermediate database having a local reference data table and a local master table; The local reference data table has a search path leading to the registration data stored in a direct-lower layer database which is a lower-layer intermediate database or a terminal database directly connected to the intermediate database having the local reference data table, or in an intermediate database or a terminal database indirectly connected via the direct-lower layer database, and a reference value indicating an attribute of a group obtained by grouping the registration data according to related classification information, A database management system in which the local master table has a tree structure of databases starting from the starting intermediate database and classification information for the grouping.

2. the hierarchical database has a lower-level intermediate database directly or indirectly connected to the origin intermediate database, 2. The database management system of claim 1, wherein said lower level intermediate database comprises a local reference data table and a local master table.

3. one or more local edge networks; Cloud network, The database management system according to claim 1 or 2. Equipped with the local edge network includes one or more mobile base stations connected by a common wireless interface; the cloud network is configured to be in communication with at least one of the mobile base stations; A communications system in which the origin intermediate database is located at the mobile base station.

4. 4. The communication system according to claim 3, wherein a maximum distance between said origin intermediate database and any of the end databases directly or indirectly connected to said origin intermediate database is between 5 km and 20 km.

5. the local edge network includes one or more mobile entities; the mobile object has a sensor and the terminal database; the database management system includes a registration means for registering sensor information acquired by the sensor as the registration data in the terminal database of the mobile object, The registration data registered by the registration means is Actual data corresponding to the sensor information; and There are multiple labels available for the above classification information. having 5. A communication system according to claim 3, wherein said registration means assigns to said registration data predetermined label values ​​in accordance with said actual data when said registration data is registered in said terminal database.

6. The database management system includes a prediction means for predicting an energy consumption per unit time of registered data during operation of the communication system and a value index of the registered data, 6. The communication system according to claim 5, wherein said registration means changes the value of said label of said registration data based on the prediction of said prediction means.

7. the classification information includes items of retention time after registration in the terminal database, collection area of ​​the sensor information, and urgency of actual data corresponding to the sensor information, The plurality of labels includes a slicing label including slicing for selecting and connecting a network function and a database function end-to-end in response to an application using the search means; 7. The communication system according to claim 5, wherein said registration means determines a value of said slicing label in accordance with said item when said registration data is registered in said terminal database.

Images