Dataset Creation Device, Dataset Creation Method, and Program

Abstract

A dataset generation device (1) according to the present invention is equipped with: a first collection function unit (11) for receiving data which pertains to communication from a circuit or from a communication device over a circuit and is regularly or irregularly transmitted from one or more first data sources (21); a second collection function unit (12) which accesses one or more second data sources (22), and receives data which pertains to communication from the circuit or from the communication device over the circuit at a pre-set arbitrary period from the one or more second data sources (22); a data recording unit (13) for recording the data received from the first collection function unit (11) and from the second collection function unit (12); and a calculation function unit (14) for generating one data set by extracting a plurality of features from the data recorded in the data recording unit (13), arranging the plurality of features in a time series in prescribed time intervals, and integrating the plurality of features which have been arranged in a time series.

Description

Technical Field

[0001] The present disclosure relates to a dataset creation device, a dataset creation method, and a program used for calculating a communication bandwidth required when a plurality of lines are accommodated in one communication link.

Background Art

[0002] Conventionally, in calculating the communication bandwidth, the communication bandwidth has been calculated using time-series information of feature quantities such as traffic volume and line contract bandwidth as input. However, when performing bandwidth estimation using machine learning, it is desirable to create a dataset including a plurality of feature quantities, not limited to the traffic volume or the line contract bandwidth. For example, Non-Patent Document 1 describes a method for calculating a communication bandwidth for satisfying a given communication quality in a communication network in which a plurality of lines are accommodated in one communication link. In the method described in Non-Patent Document 1, as shown in FIG. 14, the communication bandwidth is estimated using time-series information (information of t1, ···, t2) of the traffic volume and the line contract bandwidth as input. Also, generally when using machine learning, it is known that the prediction accuracy improves as the number of input feature quantities increases. When performing bandwidth estimation using machine learning, it is desirable to create a dataset including a plurality of feature quantities, not limited to the traffic volume or the line contract bandwidth. For this reason, the dataset including a plurality of feature quantities used for machine learning needs to be data in which given feature quantities are continuously recorded in time series at given time intervals.

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

[0004] However, there are problems when creating a dataset including a plurality of feature amounts. The first problem is related to data collection. Since the data extracted as feature amounts differ in generation conditions, storage conditions of data sources, etc., a data collection method according to the type of data is required. For example, there is data that needs to be collected because it is generated regularly, data that needs to be collected regularly because it does not hold information before update, data that needs to be collected at an appropriate timing because it is generated irregularly, and the like.

[0005] The second problem is related to time-series conversion of feature amounts. Since the collection intervals of the data from which feature amounts are extracted are different, an excess or deficiency of feature amounts occurs when extracting feature amounts from the collected data. For this reason, it is necessary to perform complementation / decimation on the extracted feature amounts to obtain feature amounts time-series converted at a predetermined time interval. For example, when creating a dataset at 10-minute intervals from feature amount 1 extracted at 5-minute intervals and feature amount 2 extracted every hour, it is necessary to obtain feature amounts time-series converted at 10-minute intervals by an appropriate method such as complementation / decimation for feature amount 1 and feature amount 2. Therefore, there is a demand for a technique for creating a dataset that can be used for calculating the communication bandwidth required when a plurality of lines are accommodated in one communication link, which can address the problems described above.

[0006] In view of such circumstances, an object of the present invention is to provide a dataset creation device, a dataset creation method, and a program that can create a dataset that can be used for calculating the communication bandwidth required when a plurality of lines are accommodated in one communication link. Means for Solving the Problems

[0007] To solve the above problems, a dataset creation device according to the present disclosure is a dataset creation device that creates a dataset used for calculating a communication bandwidth required when a plurality of lines are accommodated in one communication link, including: a first collection function unit that receives data related to communication via the line by the line or a communication device, which is periodically or aperiodically distributed from one or more first data sources; a second collection function unit that accesses one or more second data sources and receives data related to communication via the line by the line or a communication device from the one or more second data sources at a preset arbitrary period; a data recording unit that records the data received by the first collection function unit and the second collection function unit respectively; and a computer function unit that extracts a plurality of feature amounts from the data recorded in the data recording unit, time-series each of the plurality of feature amounts at a predetermined time interval, and combines the time-series plurality of feature amounts to create one dataset.

[0008] To solve the above problems, a dataset creation method according to the present disclosure is a dataset creation method that creates a dataset used for calculating a communication bandwidth required when a plurality of lines are accommodated in one communication link, including: a step of receiving, by a dataset creation device, data related to communication via the line by the line or a communication device, which is periodically or aperiodically distributed from one or more first data sources; a step of accessing one or more second data sources and receiving data related to communication via the line by the line or a communication device from the one or more second data sources at a preset arbitrary period; a step of recording the received data; a step of extracting a plurality of feature amounts from the recorded data; a step of time-series each of the plurality of feature amounts at a predetermined time interval; and a step of combining the time-series plurality of feature amounts to create one dataset.

[0009] To solve the above problems, a program according to the present disclosure causes a computer to function as the above dataset creation device.

Advantages of the Invention

[0010] According to the present disclosure, a data set that can be used to calculate the communication bandwidth required when a plurality of lines are accommodated in one communication link can be created.

Brief Description of the Drawings

[0011]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Mode for Carrying Out the Invention

[0012] Hereinafter, embodiments according to the present disclosure will be described in detail with reference to the drawings.

[0013] (First Embodiment) FIG. 1 is a block diagram showing a configuration example of a dataset creation device 1 according to the first embodiment. The dataset creation device 1 according to the first embodiment will be described below. As shown in FIG. 1, the dataset creation device 1 includes a first collection function unit 11, a second collection function unit 12, a data recording unit 13, a calculation function unit 14, and a data bus 15. The dataset creation device 1 creates a dataset used for calculating the communication band required when a plurality of lines are accommodated in one communication link.

[0014] The first collection function unit 11, the second collection function unit 12, and the calculation function unit 14 constitute a control unit 10 (controller 10). The control unit 10 (controller 10) may be constituted by dedicated hardware such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array), may be constituted by a processor, or may be constituted including both.

[0015] As shown in FIG. 1, the dataset creation device 1 is connected to one or more first data sources 21 and one or more second data sources 22. The first data source 21 is a data source that performs push-type information distribution. Push-type information distribution refers to a method of information distribution in which the sender actively transmits information without a request from the recipient side. The first data source 21 is, for example, a line information database in which information (line information) regarding subscriber lines is recorded. The second data sources 22a, 22b, and 22c are data sources that perform pull-type information distribution. Pull-type information distribution refers to a method of information distribution in which the sender passively transmits information based on a request from the recipient side. The second data source 22a is, for example, a facility database in which information (connection information) regarding the connection relationship between interfaces of communication devices that communicate via a line is registered. The second data source 22b is, for example, a link information database in which information (link information) regarding the communication link in which the line is accommodated is recorded. The second data source 22c is, for example, a traffic information collection database in which information (traffic information) regarding the traffic transmitted and received at each interface of the communication device is recorded.

[0016] The databases described in this embodiment include files stored in a format such as csv and are not limited to databases created simply using a database language such as SQL. Also, the communication links described in this embodiment include physical links, Link Aggregation in which a plurality of physical links are bundled and treated as one logical link, or virtual links such as VLAN (Virtual Local Area Network), and are not limited to a specific technology. Further, the interfaces described in this embodiment include physical, Link Aggregation, or virtual interfaces such as VLAN and are not limited to a specific technology.

[0017] Also, for the sake of convenience of explanation, in the present embodiment, as shown in FIG. 1, the first data source 21 and the second data sources 22a, 22b, 22c are each described as a single database, but each database may be composed of two or more units. Further, although it is described that information (line information) of all subscriber lines is acquired from the first data source 21, the line information may be obtained from a plurality of data sources. Also, the second data source 22c is described as including traffic data, but in addition to the traffic data, it may include time-series data such as delay information, the number of MAC address learnings, and the number of IP addresses.

[0018] The first data source 21 is connected to the first collection function unit 11, and the second data sources 22a, 22b, 22c are connected to the second collection function unit 12. Also, the first collection function unit 11, the second collection function unit 12, the data recording unit 13, and the calculation function unit 14 are connected via a data bus 15. In the present embodiment, an example in which the first data source 21 is connected to the first collection function unit is described, but the first data source 21 may be connected to the second collection function unit 12, and the second data sources 22a, 22b, 22c may be connected to the first collection function unit 11.

[0019] The first collection functional unit 11 receives data regarding communication via a line by a line or a communication device, which is periodically or non-periodically distributed from one or a plurality of first data sources 21. Hereinafter, the data received by the first collection functional unit 11 may be referred to as push-type data. In the case of push-type information distribution, the first data source 21 actively transmits data without requesting from the first collection functional unit 11. For example, as shown in FIG. 1, it is assumed that the first data source 21 transmits data of line information to the first collection functional unit 11 on a daily basis. Note that the period in which the first data source 21 transmits data is not limited to being periodic. The first data source 21 shall transmit data non-periodically according to its specifications, or each time an update occurs. When the first collection functional unit 11 receives data from the first data source 21, the received data is transmitted to the data recording unit 13. In FIG. 1, the first collection functional unit 11 is connected to one first data source 21, but it may be connected to a plurality of first data sources 21. Note that the format of data exchanged between the first data source 21 and the first collection functional unit 11 is not limited to a specific format such as csv format, json format, Mib format, etc.

[0020] The second collection functional unit 12 accesses one or more second data sources 22 and receives data related to communication via a line or a line by a communication device from the one or more second data sources 22 at a preset arbitrary period. Hereinafter, the data received by the second collection functional unit 12 may be referred to as Pull-type data. In the case of Pull-type information distribution, the second data source 22 passively transmits data based on a request from the second collection functional unit 12. When the second collection functional unit 12 successfully receives data, the received data is transmitted to the data recording unit 13. When the second collection functional unit 12 fails to receive data, it waits for a preset specified time or attempts to receive data again at a preset arbitrary period. In this embodiment, it is assumed that data on connection information is acquired from the second data source 22a on a daily basis, data on link information is acquired from the second data source 22b on an hourly basis, and data on traffic information is acquired from the second data source 22c on a minute-by-minute basis. Note that the format of data exchanged between the second data source 22 and the second collection functional unit 12 is not limited to a specific format such as csv format, json format, or Mib format.

[0021] Note that the data related to the line includes, for example, data on information about the line (line information) accommodated in the link for which the calculation of the required communication bandwidth is targeted, and data on information about the communication link in which the line is accommodated (link information). Further, the data related to communication via the line by the communication device includes, for example, data on information about the connection relationship between the interfaces of the communication devices that communicate via the line (connection information), and data on information about the traffic transmitted and received at each interface of the communication device (traffic information). The communication bandwidth required for the link greatly depends on the line accommodated in the link and the communication by the communication device via that line. The data related to the line or the communication by the communication device via the line acquired by the first collection functional unit 11 and the second collection functional unit 12 described above is merely an example, and the first collection functional unit 11 and the second collection functional unit 12 acquire arbitrary data related to the line and the communication by the communication device via that line necessary for the calculation of the communication bandwidth required for the link.

[0022] The data recording unit 13 records the data received by the first collection functional unit 11 and the second collection functional unit 12, respectively. That is, the data recording unit 13 records the Push-type data (data on line information) and the Pull-type data (data on connection information, data on link information, data on traffic information), respectively. An example of the recording by the data recording unit 13 is shown in FIGS. 2 to 5.

[0023] FIG. 2 is a table showing an example of the line information recorded in the data recording unit 13. In FIG. 2, as the line information, the lines accommodated in the target link A (in this embodiment, line 1 and line 2 are accommodated), the opening date for each line, the contract bandwidth, redundancy, priority control, delay upper limit, and the number of connections to the ground are recorded, but the line information is not limited to these. For example, the line information may be information such as the number of via links, the number of via communication devices, the used bandwidth, delay, the number of MAC learning, the number of IP addresses, the number of VLANs, the number of evis, the number of VRFs, the number of tunnels, the number of TCP / UDP sessions, and the number of flows.

[0024] FIG. 3 is a table showing an example of the connection information recorded in the data recording unit 13. In FIG. 3, as the connection information, the creation date for each communication link, the fact that interface 1 and interface 2 are connected to link A, and the fact that interface 1 of link A is interface Y / Z / 0 of communication device XXX and interface 2 is interface Y / Z / 2 of communication device XXZ are recorded. Note that the name of the communication device may be information such as the host name or IP address of the communication device.

[0025] FIG. 4 is a table showing an example of the link information recorded in the data recording unit. In FIG. 4, it is recorded that line 1 and line 2 are accommodated in link A and line 2 is accommodated in link B at 00:00 on September 10, 2021.

[0026] FIG. 5 is a table showing an example of the traffic information recorded in the data recording unit. In FIG. 5, the traffic volume transmitted and received on interface Y / Z / 0 of node XXX and interface Y / Z / 2 of node XXZ registered in link A is recorded at 1-minute intervals from 00:00 on October 11, 2021.

[0027] Also, in FIGS. 2 to 5, although each piece of information is recorded in a separate table format, as long as each piece of information is saved in a form that can be searched and referenced, it does not have to be saved as an independent table.

[0028] The computing function unit 14 extracts a plurality of feature amounts from the data recorded in the data recording unit 13 and time-series each of the plurality of feature amounts at a predetermined time interval. The computing function unit 14 combines the time-series plurality of feature amounts to create one data set. The feature amount refers to a numerical value that quantitatively represents the features and characteristics of the data or object to be analyzed.

[0029] The computing function unit 14 creates a dataset consisting of any plurality of feature amounts at any timing. The any timing is, for example, the timing at which calculations are performed by a machine learning system outside the present invention or the like. In such a case, it is assumed that the start time, end period, data output cycle, etc. of the dataset are specified from an external system. Alternatively, the dataset may be automatically created at a preset time interval.

[0030] FIG. 6 is a diagram for explaining the operation of the computing function unit 14. Hereinafter, in the present embodiment, the operation of the computing function unit 14 will be described with reference to FIGS. 2 to 6. In the present embodiment, a dataset is created with the contract bandwidth, delay upper limit, logarithm to the ground, and traffic volume of each line in link A in a 10-minute cycle from 00:00 on October 11, 2021 to 23:59 on November 10, 2021 as a plurality of feature amounts. In FIG. 6, the predetermined time interval is 10 minutes.

[0031] As described above, the computing function unit 14 starts calculations at an arbitrary timing. The processing of the computing function unit 14 is roughly classified into three types of processing: extraction processing, time series processing, and combination processing. In the extraction processing, the computing function unit 14 refers to the link information (Figure 4) recorded in the data recording unit 13 and extracts the lines (lines 1 and 2 in the example shown in Figure 4) accommodated in link A, which is the target of communication bandwidth estimation. Also, the computing function unit 14 refers to the connection information (Figure 3) recorded in the data recording unit 13 and extracts the interfaces (interface Y / Z / 0 of node XXX and interface Y / Z / 2 of node XXZ in the example shown in Figure 3) included in link A. In the time series processing, the computing function unit 14 refers to the line information (Figure 2) recorded in the data recording unit 13 and extracts, for example, the contracted bandwidth, delay limit, and number of ground connections, which are recorded corresponding to each of lines 1 and 2, as feature quantities. Also, the computing function unit 14 refers to the traffic information (Figure 5) recorded in the data recording unit 13 and refers to the traffic volume transmitted and received on interfaces 1 and 2 as a feature quantity. The computing function unit 14 time-seriesizes the extracted multiple feature quantities at time intervals of 10 minutes each. In the combination processing, the computing function unit 14 combines the contracted bandwidth, delay limit, and number of ground connections included in lines 1 and 2 time-seriesized in the time series processing with the traffic volume transmitted and received on interfaces 1 and 2 to create a single data set including multiple feature quantities. Details of the time series processing and combination processing will be described below.

[0032] First, the details of the time-series processing of the line information executed by the computing function unit 14 will be described. The number of lines accommodated in Link A may vary over time. In this embodiment, as described above, the link information is information with a one-hour period. Therefore, the feature amounts extracted from the link information also have a one-hour period. On the other hand, as described above, the dataset to be created has a 10-minute period. Thus, when the time period of the link information (feature amount) is larger than the predetermined time interval for creating the dataset, the computing function unit 14 complements the values that become the feature amounts at each predetermined time interval to time-series the feature amounts at the predetermined time interval. To match the predetermined time interval (10 minutes), for example, the computing function unit 14 complements the link information (Figure 4) at 00:00 on October 11, 2021, with the link information (accommodating Line 1 and Line 2) at 00:00, 10 minutes, 20 minutes, 30 minutes, 40 minutes, and 50 minutes on October 11, 2021. Next, the line information on October 11, 2021 is extracted from the line information (Figure 2) recorded in the data recording unit 13 as the feature amounts at 00:00, 10 minutes, 20 minutes, 30 minutes, 40 minutes, and 50 minutes on October 1, 2021. Since the accommodated lines at each time of Link A are known, the contract information (contracted bandwidth, bandwidth upper limit, and number of connections to the ground) of the line is extracted. In this embodiment, at 00:00 on October 11, 2021, Line 1 and Line 2 are accommodated in Link A. Therefore, based on the line information of Line 1 and Line 2 at 00:00 on October 11, 2021, the contracted bandwidth, bandwidth upper limit, and number of connections to the ground of Line 1 and Line 2 at 00:00, 10 minutes, 20 minutes, 30 minutes, 40 minutes, and 50 minutes on October 1, 2021 are complemented. In this embodiment, the feature amounts of the line information with numerical values recorded in advance are specified. However, for example, feature amounts recorded as character data, such as redundancy and priority control, are converted into numerical values using methods such as categorical variables.

[0033] Next, the details of the time-series processing of traffic information executed by the computing function unit 14 will be described. Since the interfaces constituting Link A are identified from the connection information (Figure 3) recorded in the data recording unit 13, the time-series data of the traffic information (Figure 5) recorded in the data recording unit 13 is referred to. In this embodiment, while it is necessary to create a data set with a 10-minute cycle, the traffic information is recorded at a 1-minute cycle. Thus, when the time cycle of the feature amount is smaller than a predetermined time interval, the computing function unit 14 time-seriesizes the feature amount at the predetermined time interval by determining the representative value of the feature amount from 00:00 to 00:09 on October 11, 2021. Here, the representative value refers to the first value, intermediate value, approximate value, maximum value, minimum value, average value, etc. within the time cycle. For example, from the traffic information (Figure 5) recorded in the data recording unit 13, 35.609, which is the maximum value of the feature amount from 00:00 to 00:09 on October 11, 2021, is taken as the feature amount of the traffic at 00:00 on October 11, 2021. Here, the maximum value is selected as the representative value, but the first value, intermediate value, approximate value, minimum value, average value, etc. may also be used as the feature amount of the traffic at that time. Different from this embodiment, when the time cycle of the feature amount of the traffic is larger than the predetermined time interval for creating the data set, for example, when it is necessary to create a data set with a 10-minute cycle (00:00, 10 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes on October 11, 2021), while the traffic information is recorded at a 1-hour cycle (··· 23:00 on October 10, 00:00, 01:00 on October 11 ···), the computing function unit 14 selects the values of the traffic feature amounts at 00:00, 10 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes on October 11, 2021, and uses methods such as selecting the first value of the traffic feature amount (for example, the value at 00:00 on October 11, 2021), using the intermediate value of the preceding and following values (for example, the intermediate value of the values at 23:00 on October 10 and 00:00 on October 11), and approximating a plurality of preceding and following values (for example, the values at 23:00 on October 10, 00:00, 01:00 on October 11) to perform time-seriesization.

[0034] In the combination process, the computing function unit 14 creates a single dataset containing multiple feature quantities by combining the feature quantities (contracted bandwidth, delay upper limit, number of connections to the ground) included in Line 1 and Line 2, for which time series processing has been performed, and the feature quantities of the traffic volume transmitted and received over Interface 1 and Interface 2. FIG. 7 is a table showing an example of a single dataset created by combining multiple time-series feature quantities. In this embodiment, to create a dataset with a 10-minute cycle, the date and time in the first row of the table in FIG. 7 are divided into 10-minute cycles from 00:00 on October 11, 2021 to 23:59 on November 10, 2021. In the traffic in the second row, the maximum value of the traffic volume for each 10-minute period is recorded. In Line 3 for Line 1 and Line 4 for Line 2, the contracted bandwidth, delay upper limit, and number of connections to the ground for each 10-minute period are recorded.

[0035] The data bus 15 interconnects the first collection function unit 11, the second collection function unit 12, the data recording unit 13, and the computing function unit 14, which are provided in the dataset creation device 1.

[0036] FIG. 8 is a flowchart showing an example of a dataset creation method executed by the dataset creation device 1 according to the first embodiment.

[0037] In step S101, the first collection function unit 11 receives data periodically or aperiodically distributed from one or more first data sources 21.

[0038] In step S102, the second collection function unit 12 accesses one or more second data sources 22 and receives data at a preset arbitrary cycle. As shown in FIG. 8, the process of step S101 and the process of step S102 may be performed in parallel, or either one of the processes may be performed first and the other process may be performed later.

[0039] In step S103, the data recording unit 13 records the data received by the first collection function unit 11 and the second collection function unit 12, respectively.

[0040] In step S104, the computing function unit 14 extracts a plurality of feature amounts from the data recorded in the data recording unit 13.

[0041] In step S105, the computing function unit 14 time - seriesizes each of the extracted plurality of feature amounts at a predetermined time interval.

[0042] In step S106, the computing function unit 14 combines the time - seriesized plurality of feature amounts to create one data set.

[0043] According to the data set creation device 1 according to the present embodiment, it is possible to create a data set that can be used for calculating the communication bandwidth required when a plurality of lines are accommodated in one communication link, and it is possible to realize bandwidth estimation with high prediction accuracy.

[0044] (Second Embodiment) FIG. 9 is a block diagram showing a configuration example of a data set creation device 2 according to the second embodiment. As shown in FIG. 9, the data set creation device 2 includes a first collection function unit 11, a second collection function unit 12, a data recording unit 13, a computing function unit 14, a data bus 15, and a recording determination unit 16. The data set creation device 2 creates a data set used for calculating the communication bandwidth required when a plurality of lines are accommodated in one communication link. The data set creation device 2 according to the present embodiment is different from the data set creation device 1 according to the first embodiment in that it further includes a recording determination unit 16. For the same configuration as that of the first embodiment, the same reference numerals as those in the first embodiment are given and the description is appropriately omitted.

[0045] The first collection function unit 11, the second collection function unit 12, the computing function unit 14, and the recording determination unit 16 constitute a control unit 10a (controller 10a). The control unit 10a (controller 10a) may be configured by dedicated hardware such as an ASIC or an FPGA, may be configured by a processor, or may be configured to include both.

[0046] The recording determination unit 16 compares the data received by the first collection function unit 11 or the second collection function unit 12 with the data recorded in the data recording unit 13. When a difference is found in the content of the data, it is determined that the data has been updated, and the received data is transmitted to the data recording unit 13 for recording.

[0047] The operation of the data set creation device 2 according to this embodiment will be described with reference to FIG. 9. When the first collection function unit 11 or the second collection function unit 12 acquires data, it transmits the data to the recording determination unit 16. When the recording determination unit 16 receives data from the first collection function unit 11 or the second collection function unit 12, it compares the received data with the latest data recorded in the data recording unit 13. When the recording determination unit 16 finds a difference in the content of the data, it determines that the data has been updated and transmits the data received by the first collection function unit 11 or the second collection function unit 12 to the data recording unit 13. When the recording determination unit 16 finds no difference in the content of the data, it does not transmit the data to the data recording unit 13 and waits for the reception of the next data from the first collection function unit 11 or the second collection function unit 12.

[0048] FIG. 10 is a flowchart showing an example of the data set creation method executed by the data set creation device 2 according to the second embodiment.

[0049] In step S201, the first collection function unit 11 receives data periodically or irregularly distributed from one or more first data sources 21.

[0050] In step S202, the second collection function unit 12 accesses one or more second data sources 22 and receives data at a preset arbitrary period. As shown in FIG. 10, the processes of step S201 and step S202 may be performed in parallel, or either one of the processes may be performed first and the other process may be performed later.

[0051] In step S203, the recording determination unit 16 compares the data received by the first collection function unit 11 with the data recorded in the data recording unit 13. If a difference is recognized in the content of the data, the recording determination unit 16 determines that there has been an update of the data, and transmits the data received by the first collection function unit 11 to the data recording unit 13 for recording. If no difference is recognized in the content of the data, the recording determination unit 16 determines that there has been no update of the data, does not transmit the data to the data recording unit 13, and waits for the reception of the next data from the first collection function unit 11.

[0052] In step S204, the recording determination unit 16 compares the data received by the second collection function unit 12 with the data recorded in the data recording unit 13. If a difference is recognized in the content of the data, the recording determination unit 16 determines that there has been an update of the data, and transmits the data received by the second collection function unit 12 to the data recording unit 13 for recording. If no difference is recognized in the content of the data, the recording determination unit 16 determines that there has been no update of the data, does not transmit the data to the data recording unit 13, and waits for the reception of the next data from the second collection function unit 12.

[0053] In step S205, the data recording unit 13 records the data transmitted from the recording determination unit 16.

[0054] In step S206, the calculation function unit 14 extracts a plurality of feature amounts from the data recorded in the data recording unit 13.

[0055] In step S207, the calculation function unit 14 time - seriesizes each of the plurality of extracted feature amounts at a predetermined time interval.

[0056] In step S208, the calculation function unit 14 combines the time - seriesized plurality of feature amounts to create one data set.

[0057] According to the dataset creation device 2 according to this embodiment, by determining whether data can be recorded, it is possible to reduce the amount of data recorded by the data recording unit 13 of the dataset creation device 2.

[0058] (Third Embodiment) FIG. 11 is a block diagram showing a configuration example of a dataset creation device 3 according to the third embodiment. As shown in FIG. 11, the dataset creation device 3 includes a first collection function unit 11, a second collection function unit 12, a data recording unit 13, a calculation function unit 14, a data bus 15, a recording determination unit 16, and a third collection function unit 17. The dataset creation device 3 creates a dataset used for calculating the communication bandwidth required when a plurality of lines are accommodated in one communication link. The dataset creation device 3 according to this embodiment is different from the dataset creation device 2 according to the second embodiment in that it further includes a third collection function unit 17. For the same configurations as those in the first embodiment or the second embodiment, the same reference numerals as those in the first embodiment or the second embodiment are given, and the description is omitted as appropriate.

[0059] The control unit 10b (controller 10b) is constituted by the first collection function unit 11, the second collection function unit 12, the calculation function unit 14, the recording determination unit 16, and the third collection function unit 17. The control unit 10b (controller 10b) may be constituted by dedicated hardware such as an ASIC or an FPGA, may be constituted by a processor, or may be constituted by including both.

[0060] The third collection functional unit 17 is communicable with one or more third data sources 23 and a management system 24 that manages the states of the one or more third data sources 23. When a state change of the third data source 23 is notified from the third data source 23 or the management system 24, the third collection functional unit 17 accesses the third data source 23 to receive data. The third data source 23 is, for example, a communication device, and the management system 24 is an element management system (EMS) that performs inventory management of the communication device. Inventory management refers to the centralized management of information such as the hardware configuration and software configuration of communication devices connected on a network. An element management system (EMS) is a system for directly managing each communication device (element) that constitutes a network. In this embodiment, it is described assuming that there is one third data source 23 and one management system 24, respectively, but two or more third data sources 23 and management systems 24 may be connected to a communication network.

[0061] As shown in FIG. 11, the third data source 23, the management system 24, and the third collection function unit 17 operate in cooperation in the following order of (1) to (3). (1) The third data source 23 notifies the management system 24 of events such as setting changes and state changes during a failure. (2) Upon receiving the notification from the third data source 23, the management system 24 notifies the third collection function unit 17 of the event. The third data source 23 may directly notify the third collection function unit 17 of events such as state changes. (3) When the third collection function unit 17 is notified of a state change of the third data source 23 from the third data source 23 or the management system 24, the third collection function unit 17 accesses the third data source 23 and receives data. In this way, the third data source 23 is a data source that performs both Push-type information distribution and Pull-type information distribution. That is, the third data source 23 actively sends notifications of state changes without being requested by the third collection function unit 17, and passively sends data based on an access from the third collection function unit 17. When the third data source 23 is a communication device, the third collection function unit 17 connects to the third data source 23 using telnet, ssh, etc., issues a command to obtain the state, and records the response. Further, information such as a change in the line accommodated in the communication link and a change in the connection relationship of the interface is extracted from the response.

[0062] FIG. 12 is a flowchart showing an example of a dataset creation method executed by the dataset creation device 3 according to the third embodiment.

[0063] In step S301, the first collection function unit 11 receives data periodically or irregularly distributed from one or more first data sources 21.

[0064] In step S302, the second collection function unit 12 accesses one or more second data sources 22 and receives data at a preset arbitrary period.

[0065] In step S303, the third collection functional unit 17 receives a notification of a state change of the third data source 23 from the third data source 23 or the management system 24.

[0066] In step S304, when the third collection functional unit 17 is notified of a state change of the third data source 23, it accesses the third data source 23 and receives data. The processing of step S301, the processing of step S302, and the processing from step S303 to S304 may be performed in parallel as shown in FIG. 12, or either one of the processing may be performed first and the other processing may be performed later.

[0067] In step S305, the recording determination unit 16 compares the data received by the first collection functional unit 11 with the data recorded in the data recording unit 13. If a difference is recognized in the content of the data, the recording determination unit 16 determines that there has been an update of the data, and transmits the data received by the first collection functional unit 11 to the data recording unit 13 for recording. If no difference is recognized in the content of the data, the recording determination unit 16 determines that there has been no update of the data, does not transmit the data to the data recording unit 13, and waits for the reception of the next data from the first collection functional unit 11.

[0068] In step S306, the recording determination unit 16 compares the data received by the second collection functional unit 12 with the data recorded in the data recording unit 13. If a difference is recognized in the content of the data, the recording determination unit 16 determines that there has been an update of the data, and transmits the data received by the second collection functional unit 12 to the data recording unit 13 for recording. If no difference is recognized in the content of the data, the recording determination unit 16 determines that there has been no update of the data, does not transmit the data to the data recording unit 13, and waits for the reception of the next data from the second collection functional unit 12.

[0069] In step S307, the recording determination unit 16 compares the data received by the third collection function unit 17 with the data recorded in the data recording unit 13. If a difference is recognized in the content of the data, the recording determination unit 16 determines that there has been an update of the data, and transmits the data received by the third collection function unit 17 to the data recording unit 13 for recording. If no difference is recognized in the content of the data, the recording determination unit 16 determines that there has been no update of the data, does not transmit the data to the data recording unit 13, and waits for the reception of the next data from the third collection function unit 17.

[0070] In step S308, the data recording unit 13 records the data transmitted from the recording determination unit 16.

[0071] In step S309, the calculation function unit 14 extracts a plurality of feature amounts from the data recorded in the data recording unit 13.

[0072] In step S310, the calculation function unit 14 time - seriesizes each of the extracted plurality of feature amounts at a predetermined time interval.

[0073] In step S311, the calculation function unit 14 combines the time - seriesized plurality of feature amounts to create one data set.

[0074] According to the data set creation device 3 according to the present embodiment, since it is possible to recognize the occurrence of feature amounts based on the notification from the third data source 23 or the management system 24, more efficient data collection can be performed by acquiring data only when the data is generated.

[0075] In order to operate the above-described dataset creation devices 1, 2, and 3, it is also possible to use a computer capable of executing program instructions. FIG. 13 is a block diagram showing a schematic configuration of a computer 100 that functions as a dataset creation device. Here, the computer 100 that functions as the dataset creation devices 1, 2, and 3 may be a general-purpose computer, a dedicated computer, a workstation, a PC (Personal Computer), a tablet computer, or the like. The program instructions may be program codes, code segments, etc. for executing necessary tasks.

[0076] As shown in FIG. 13, the computer 100 includes a processor 110, a ROM (Read Only Memory) 120, a RAM (Random Access Memory) 130, and a storage 140 as storage units, an input unit 150, an output unit 160, and a communication interface (I / F) 170. Each component is communicably connected to each other via a data bus 180.

[0077] The ROM 120 stores various programs and various data. The RAM 130 temporarily stores a program or data as a work area. The storage 140 is composed of an HDD (Hard Disk Drive) or an SSD (Solid State Drive) and stores various programs and various data including an operating system. In the present disclosure, the program according to the present disclosure is stored in the ROM 120 or the storage 140.

[0078] The processor 110 is specifically a CPU (Central Processing Unit), MPU (Micro Processing Unit), GPU (Graphics Processing Unit), DSP (Digital Signal Processor), SoC (System on a Chip), etc., and may be composed of a plurality of processors of the same type or different types. The processor 110 reads a program from the ROM 120 or the storage 140, and executes the program using the RAM 130 as a working area, thereby controlling each of the above components and performing various arithmetic processes. Note that at least a part of these processing contents may be realized by hardware.

[0079] The program may be recorded on a recording medium readable by the dataset creation devices 1, 2, and 3. By using such a recording medium, it is possible to install it on the dataset creation devices 1, 2, and 3. Here, the recording medium on which the program is recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, and may be, for example, a CD-ROM, a DVD-ROM, a USB (Universal Serial Bus) memory, etc. Also, this program may be in a form downloaded from an external device via a network.

[0080] Regarding the above embodiments, the following additional remarks are further disclosed.

[0081] (Additional Clause 1) A dataset creation device for creating a dataset used for calculating the communication bandwidth required when a plurality of lines are accommodated in one communication link, a memory, and a controller connected to the memory, wherein the controller receives data related to communication via the line by the line or a communication device, which is periodically or irregularly distributed from one or more first data sources, and records the data in the memory, Access one or more second data sources, and receive, from the one or more second data sources at a preset arbitrary period, data related to communication via the line or by the communication device over the line, and record the data in the memory. Extract a plurality of feature quantities from the data recorded in the memory. Time-series each of the plurality of feature quantities at a predetermined time interval. Combine the time-series plurality of feature quantities to create one data set. Data set creation device. (Additional item 2) When the time period of the feature quantity is smaller than the predetermined time interval, the controller time-series the feature quantity by determining a representative value of the feature quantity within the time period. When the time period of the feature quantity is larger than the predetermined time interval, the controller time-series the feature quantity by complementing the value that becomes the feature quantity for each predetermined time interval. The data set creation device according to claim 1. (Additional item 3) The controller compares the received data with the data recorded in the memory, and when a difference is recognized in the content of the data, records the received data in the memory. The data set creation device according to claim 1 or 2. (Additional item 4) The controller is communicable with one or more third data sources and a management system that manages the states of the one or more third data sources. When the controller is notified of a state change of the third data source from the third data source or the management system, the controller accesses the third data source and receives data. The data set creation device according to any one of claims 1 to 3. (Additional item 5) A data set creation method for creating a data set used for calculating a communication bandwidth required when a plurality of lines are accommodated in one communication link, By a data set creation device Receiving data related to communication via the line by the line or communication device, which is periodically or aperiodically distributed from one or more first data sources; Accessing one or more second data sources and receiving data related to communication via the line by the line or communication device from the one or more second data sources at a preset arbitrary period; Recording the received data; Extracting a plurality of feature quantities from the recorded data; Time - serializing each of the plurality of feature quantities at a predetermined time interval; Combining the time - serialized plurality of feature quantities to create one data set; A data set creation method including the above. (Additional item 6) A non - transitory storage medium storing a computer - executable program, which stores a program for causing the computer to function as the data set creation device according to any one of claims 1 to 4.

[0082] Although the above - described embodiments have been described as representative examples, it is obvious to those skilled in the art that many changes and substitutions can be made within the spirit and scope of the present disclosure. Therefore, the present invention should not be construed as being limited by the above - described embodiments, and various modifications or changes are possible without departing from the scope of the claims. For example, it is possible to combine a plurality of constituent blocks described in the configuration diagrams of the embodiments into one, or to divide one constituent block.

Explanation of reference numerals

[0083] 1, 2, 3 Data set creation device 10, 10a, 10b Control unit (controller) 11 First collection functional unit 12 Second collection functional unit 13 Data recording unit (memory) 14 Computing functional unit 15,180 Data Bus 16 Recording Determination Unit 17 Third Collection Function Unit 21 First Data Source 22, 22a, 22b, 22c Second Data Source 23 Third Data Source 24 Management System 100 Computer 110 Processor 120 ROM 130 RAM 140 Storage 150 Input Unit 160 Output Unit 170 Communication Interface (I / F)

Claims

1. A data set creation device for creating a data set used for calculating a communication bandwidth required when a plurality of lines are accommodated in one communication link, a first collection function unit that receives data related to communication via the line by the line or a communication device, which is distributed regularly, irregularly, or each time an update occurs, from a data source that performs push-type information distribution installed outside the own device; a second collection function unit that accesses a data source that performs pull-type information distribution installed outside the own device and receives data related to communication via the line by the line or a communication device at an arbitrary period set in advance from the data source that performs the pull-type information distribution; a data recording unit that records the data received by the first collection function unit and the second collection function unit respectively; a calculation function unit that extracts a plurality of feature amounts from the data recorded in the data recording unit, time-series each of the plurality of feature amounts at a predetermined time interval, and combines the time-series plurality of feature amounts to create one data set; A data set creation device comprising the above.

2. When the time period of the feature amount is smaller than the predetermined time interval, the calculation function unit determines a representative value of the feature amount within the time period to time-series the feature amount at the predetermined time interval. When the time period of the feature amount is larger than the predetermined time interval, the value that becomes the feature amount is complemented for each predetermined time interval to time-series the feature amount at the predetermined time interval. The data set creation device according to Claim 1.

3. The data set creation device according to Claim 1 or 2, further comprising a recording determination unit that compares the data received by the first collection function unit or the second collection function unit with the data recorded in the data recording unit, and when a difference is recognized in the content of the data, causes the received data to be recorded in the data recording unit.

4. The data set creation device further comprises a third collection function unit that can communicate with one or more third data sources and a management system that manages the states of the one or more third data sources, when the third collection function unit is notified of a state change of the third data source from the third data source or the management system, it accesses the third data source and receives data, When the third data source is a communication device, the third collection function unit issues a command for acquiring the state of the third data source, and extracts information on a change in a line accommodated in a communication link or a change in a connection relationship of an interface from a response of the third data source. The data set creation device according to any one of claims 1 to 3.

5. A data set creation method for creating a data set used for calculating a communication bandwidth required when a plurality of lines are accommodated in one communication link, by a data set creation device, receiving data regarding communication via the line by the line or a communication device, which is distributed periodically, irregularly, or each time an update occurs, from a data source that performs push-type information distribution installed outside the own device; accessing a data source that performs pull-type information distribution installed outside the own device, and receiving data regarding communication via the line by the line or a communication device from the data source that performs pull-type information distribution at an arbitrarily set period in advance; recording the received data; extracting a plurality of feature amounts from the recorded data; time-seriesizing each of the plurality of feature amounts at a predetermined time interval; combining the time-seriesized plurality of feature amounts to create one data set; A data set creation method including the above.

6. A program for causing a computer to function as the data set creation device according to any one of claims 1 to 4.

Images