JP2019215807A - Data controller, data control method, and program - Google Patents

Abstract

To provide a data controller with which it is possible to stabilize the accuracy of executing an application in even the edge environment of a limited calculation resource.SOLUTION: Provided is a data controller comprising: a data control unit for receiving sensor data in which application at a delivery destination is designated, controlling the sensor data in accordance with the result of comparison of a future load that can occur when the sensor data is processed by an application at the delivery destination with a load threshold, and transmitting the controlled sensor data to the application at the delivery destination; and a load prediction unit for predicting, on the basis of the load information of a server in which the application at the delivery destination is installed and the load characteristics of the application, a future load that can occur when the sensor data controlled by the data control unit is processed by the application at the delivery destination.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a data control device, a data control method, and a program for controlling a load resulting from an operation of an application.

  An IoT (Internet of Things) system collects sensor data acquired by a sensor such as a camera, and executes processing using the collected sensor data. The sensor sends sensor data to the application via the IoT gateway. The IoT gateway delivers data collected from the sensors to a desired application implemented on any computer. The application processes the delivered sensor data.

  Generally, an application configuring an IoT system is operated on a computer in a cloud environment or an edge environment in consideration of cost and processing delay.

  When operating an application in a cloud environment, sensor data is sent to an application on the cloud environment via a public network or a dedicated network. In general, the cloud environment has the advantage that the computing resources can be used at low cost and the computing resources can be increased or decreased according to the application load. On the other hand, the cloud environment requires the transmission of sensor data through a public network or a dedicated network, and thus has a disadvantage that communication cost and communication time are required.

  In the edge environment, a computer is located near the sensor. The sensor sends the sensor data to a nearby computer application. Generally, in the edge environment, since the computation resources are arranged near the sensor, there is a disadvantage that the computation resources are expensive. Also, in the edge environment, it is difficult to allocate a large amount of computational resources due to restrictions on the location, and the computational resources are limited. Therefore, there is a disadvantage that the computational resources cannot be increased or decreased according to the application load. is there. On the other hand, in the edge environment, there is an advantage that communication is unnecessary, so that communication costs can be suppressed and communication time can be reduced.

  Whether to operate the application in the cloud environment or the edge environment is determined by the requirements of the system.

  On the other hand, in the IoT system, the server load resulting from the processing of the application may fluctuate dynamically. For example, for an application that performs data processing triggered by data input from a sensor, if there is data input from a sensor whose data amount fluctuates dynamically, the load of the application depends on the data input amount. And fluctuate dynamically.

  Further, the application load may fluctuate depending on not the amount of data input from the sensor but the quality of the data. For example, in an application that analyzes camera video data and analyzes a person in the camera, the application load varies depending on the number of people in the camera. When operating an application whose load dynamically fluctuates in an edge environment where computational resources are limited, there is a possibility that computational resources will be exhausted. When computing resources are exhausted, there is a possibility that system operation becomes unstable or processing is delayed.

  Patent Literature 1 discloses a data distribution system that allows a plurality of client devices connected to a server device via a network to upload and download data.

  Patent Literature 2 discloses a server that provides an M2M (Machine to Machine) service.

  Patent Literature 3 discloses a data warehouse system in a distributed network computing environment.

  Patent Document 4 discloses a distributed data storage server system in which a plurality of servers connected via a network provide services to clients.

JP 2014-041487 A JP 2013-200645 A JP 2009-146425 A JP 2003-248669 A

  According to the system of Patent Document 1, by adjusting the processing on the client device side according to the load status of the server device, the processing efficiency of the entire system can be improved. The system of Patent Literature 1 requests a client device to change a setting in accordance with the load status of a server device predicted based on the number of client devices that are responsible for adding and acquiring images to a stream. Therefore, the system of Patent Document 1 has a problem in that it takes a long time to add an image to a server device from a client device that plays a role of adding an image to a stream.

  According to the server of Patent Literature 2, even when the server is under a heavy load, it is possible to reduce the load on the server without delaying the activation of the service. However, the server disclosed in Patent Literature 2 has a problem in that data processing is delayed because the server obtains the priority of the data collection method after the server itself becomes in a high load state.

  According to the system of Patent Literature 3, data can be shared between clients by grouping the clients, so that the storage device capacity, update processing cost, and network load on the client side can be reduced. However, the system disclosed in Patent Literature 3 has no problem unless it is in an environment where clients are connected to each other, and therefore cannot be used for collecting data from independent devices such as sensors. .

  According to the system disclosed in Patent Literature 4, since data transfer between servers when a server is changed can be suppressed, the load on the server, the storage device, and the network can be suppressed when the connection destination server of the client is changed. However, the system of Patent Document 4 selects the second server with the lowest load from the server load information when it is determined that the load on the first server is higher than the load threshold. There was a problem that it was necessary to share.

  An object of the present invention is to solve the above-described problems and to provide a data control device capable of stabilizing the execution accuracy of an application even in an edge environment with limited computational resources.

  The data control device according to one embodiment of the present invention is configured such that the destination application receives the specified sensor data, and a comparison result between a future load and a load threshold that can occur when the destination application processes the sensor data is used. A data control unit that controls the sensor data in accordance with the data, and transmits the controlled sensor data to a destination application; and a data control unit based on load information of a server on which the destination application is installed and load characteristics of the application. And a load estimating unit for estimating a future load that may occur when processing the sensor data controlled by the application in the destination application.

  A data processing system according to one embodiment of the present invention processes a server, at least one application installed on the server, at least one sensor for acquiring sensor data to be processed by the application, and sensor data acquired by the sensor. A data delivery device that specifies the application of the delivery destination and delivers the sensor data with the designated delivery destination to the data control device, and receives the sensor data with the designated delivery destination from the data delivery device and installs the application of the delivery destination. Based on the server load information provided and the load characteristics of the destination application, predict the future load that may occur when the sensor data is processed by the destination application, and according to the comparison result between the future load and the load threshold. To control the sensor data, and And a data controller that sends the data to the destination application.

  In the data control method according to one aspect of the present invention, the destination application receives the designated sensor data, and converts the sensor data based on the load information of the server on which the destination application is installed and the load characteristics of the application. Predict the future load that may occur when processing in the destination application, and control the sensor data according to the result of comparing the future load that may occur when processing the sensor data with the destination application with the load threshold. , And sends the controlled sensor data to the destination application.

  According to an embodiment of the present invention, there is provided a program for distributing sensor data based on a process of receiving sensor data specified by a destination application and load information of a server on which the destination application is installed and load characteristics of the application. Controls sensor data according to the result of predicting the future load that may occur when processing with the destination application and comparing the load threshold with the future load that may occur when processing the sensor data with the destination application And transmitting the controlled sensor data to the destination application.

  According to the present invention, it is possible to provide a data control device capable of stabilizing the execution accuracy of an application even in an edge environment with limited computational resources.

FIG. 1 is a conceptual diagram illustrating an example of a configuration of a data processing system according to a first embodiment of the present invention. FIG. 2 is a block diagram illustrating an example of a configuration of a data control device included in the data processing system according to the first embodiment of the present invention. 6 is a table illustrating another example of control method information stored in a control method storage unit of the data control device provided in the data processing system according to the first embodiment of the present invention. 6 is a table illustrating an example of server load information stored in a server load storage unit of the data control device provided in the data processing system according to the first embodiment of the present invention. 6 is a table illustrating an example of application load characteristic information stored in an application load characteristic storage unit of the data control device provided in the data processing system according to the first embodiment of the present invention. 5 is a table illustrating another example of application load characteristic information stored in an application load characteristic storage unit of the data control device provided in the data processing system according to the first embodiment of the present invention. 5 is a flowchart for explaining an operation of the data processing system according to the first embodiment of the present invention. It is a conceptual diagram showing an example of the composition of the data processing system concerning a 2nd embodiment of the present invention. FIG. 2 is a block diagram illustrating an example of a hardware configuration of a data control device included in a data processing system according to each embodiment of the present invention.

  An embodiment for carrying out the present invention will be described below with reference to the drawings. However, the embodiments described below have technically preferable limitations for carrying out the present invention, but do not limit the scope of the invention. In all the drawings used in the description of the following embodiments, the same parts are denoted by the same reference numerals unless otherwise specified. In the following embodiments, repetitive description of similar configurations and operations may be omitted.

(1st Embodiment)
First, a data processing system according to a first embodiment of the present invention will be described with reference to the drawings.

(Constitution)
FIG. 1 is a conceptual diagram illustrating an example of a configuration of a data processing system 1 according to the present embodiment. As shown in FIG. 1, the data processing system 1 includes a sensor group 11, a data delivery device 12, a data control device 13, an application group 14, and a server 15.

  The sensor group 11 includes at least one sensor 111. FIG. 1 shows an example in which a sensor group 11 is configured by m sensors 111-1 to 111-m (m is a natural number). The application group 14 includes at least one application 141. FIG. 1 illustrates an example in which an application group 14 is configured by n applications 141-1 to 141-n (n is a natural number).

  Each of the sensors 111 constituting the sensor group 11 converts spatial information and time information to be detected into an electric signal (hereinafter, referred to as sensor data SD), and transmits the converted sensor data SD to the data distribution device 12. . The sensor 111 transmits, to the data delivery device 12, transmission data TD in which an ID indicating a transmission destination application 141 (hereinafter, referred to as a destination application ID) and sensor data SD are set. The destination application ID may be configured to be set with the sensor data SD by the data distribution device 12 or the data control device 13.

  For example, the sensor 111 can be realized by a device that detects light, such as a light receiving element such as a photodiode or an image sensor, a device that detects weight, pressure, strain, and the like, and a device that detects speed, acceleration, rotation speed, and the like. . The sensor 111 can be realized by a device that detects ultrasonic waves, sound waves, and vibrations, a device that detects temperature, humidity, and heat, and a device that detects a magnetic field, a current, a voltage, and the like. The sensor 111 can be realized by a device that detects the presence or absence, position, and shape of an object, and a device that detects smell, taste, face, fingerprint, and the like. It should be noted that the implementation example of the sensor 111 is merely an example, and does not limit the present embodiment.

  The data delivery device 12 receives the transmission data TD from the sensor group 11. The data delivery device 12 sends delivery data DD in which the address of the application 141 specified by the destination application ID included in the transmission data and the sensor data SD are set to the data control device 13.

  The data control device 13 acquires load information of the server 15. Further, the data control device 13 holds the load characteristic information of the application group 14. The load characteristic information of the application 141 is a correlation between the load of the application 141 and at least one of the data amount and the content.

  Further, the data control device 13 receives the delivery data DD from the data delivery device 12. The data control device 13 determines a control method for the sensor data SD included in the delivery data DD, and generates controlled sensor data CSD by applying the determined control method to the sensor data SD.

  Specifically, the data control device 13 uses the load information of the server 15, the load characteristic information of the application 141, and the sensor data SD to load the future load FL that can occur when the application 141 is executed on the server 15. Predict. The data control device 13 compares the predicted future load FL with a load threshold TL held as a control parameter by the own device, and determines a control method in which the future load FL falls below the load threshold TL. The data control device 13 generates the controlled sensor data CSD by applying the control method to the sensor data SD.

  The data control device 13 uses the load information of the server 15, the load correlation of the application 141, and the controlled sensor data CSD to predict a future load FL ′ that may occur when the application 141 is executed on the server 15. I do. The data control device 13 compares the predicted future load FL 'with the load threshold TL held as a control parameter by the own device, and determines a control method in which the future load FL' is lower than the load threshold TL. The data control device 13 updates the controlled sensor data CSD by applying the control method to the controlled sensor data CSD. The data control device 13 repeats the update of the controlled sensor data CSD until the future load FL 'falls below the load threshold TL.

  Then, the data control device 13 transmits the sensor data SD or the controlled sensor data CSD to the application 141 set at the destination address. In the following, the uncontrolled sensor data SD may also be referred to as controlled sensor data CSD (controlled sensor data) if the load is verified by the data control device 13.

  The application 141 set to the destination address receives the sensor data SD or the controlled sensor data CSD from the data control device 13. The application 141 executes processing using the sensor data SD or the controlled sensor data CSD received from the data control device 13.

  Each application 141 constituting the application group 14 is software programmed according to the purpose. For example, examples of the application 141 include a face image analysis application and a river flood detection application. However, any software can be applied to the application 141. In FIG. 1, the application group 14 and the server 15 are illustrated as having different configurations. However, each application 141 configuring the application group 14 is installed on the server 15.

  The server 15 is a server on which an application 141 constituting the application group 14 is installed. For example, the server 15 can be installed at an arbitrary location such as a data center, a server room, an office, or a private home. The server 15 may be constructed as a virtual server. Although FIG. 1 shows an example in which one server 15 is provided, a plurality of servers 15 may be provided.

  The above is the description of the configuration of the data processing system 1. Next, a detailed configuration of the data control device 13 included in the data processing system 1 will be described with reference to the drawings.

[Data controller]
FIG. 2 is a block diagram illustrating an example of a detailed configuration of the data control device 13. 2, the data control device 13 includes a data control unit 131, a control method storage unit 132, a server load storage unit 133, a server load collection unit 134, a future load prediction unit 135, and an application load characteristic storage unit 136. . The data control unit 131 and the control method storage unit 132 constitute a data control unit. The server load storage unit 133 and the server load collection unit 134 constitute a server load collection unit. The future load prediction unit 135 and the application load characteristic storage unit 136 constitute a load prediction unit.

  The data control unit 131 receives the delivery data DD from the data delivery device 12. The data control unit 131 transmits the sensor data SD included in the delivery data DD and a prediction instruction of a future load FL which is a load that can be generated when the application 141 set to the destination address processes the sensor data SD in the future. This is transmitted to the load prediction unit 135. The data control unit 131 receives the future load FL predicted according to the prediction instruction from the future load prediction unit 135.

  The data control unit 131 uses the received future load FL and the sensor data SD to generate controlled sensor data CSD according to the load. That is, the data control unit 131 compares the future load FL based on the sensor data SD with the load threshold TL, and generates controlled sensor data CSD according to the comparison result.

  When the future load FL is lower than the load threshold TL, the data control unit 131 determines that the load generated when the application 141 processes the sensor data SD does not cause a server resource tightness. In this case, the data control unit 131 transmits the sensor data SD to the destination application 141.

  On the other hand, when the future load FL exceeds the load threshold TL, the data control unit 131 determines that a server resource tightness occurs due to a load generated when the application 141 processes the sensor data SD. In this case, the data control unit 131 generates the controlled sensor data CSD by referring to the control method storage unit 132 and applying the control method stored in the control method storage unit 132 to the sensor data SD. Specifically, the data control unit 131 acquires a control method for the application 141 using the sensor data SD with reference to the control method storage unit 132, applies the acquired control method to the sensor data SD, Obtain sensor data CSD. Note that whether to control the sensor data SD when the future load FL is equal to the load threshold TL may be determined in advance.

  The data control unit 131 transmits to the future load prediction unit 135, together with the controlled sensor data CSD, a prediction instruction of a future load FL which is a load that may occur when the controlled sensor data CSD is processed by the application 141. The data control unit 131 receives the future load FL predicted according to the prediction instruction from the future load prediction unit 135.

  The data control unit 131 transmits or updates the controlled sensor data CSD according to the load, using the received future load FL and the controlled sensor data CSD. Specifically, the data control unit 131 transmits or updates the controlled sensor data CSD according to the comparison result between the future load FL based on the controlled sensor data CSD and the load threshold TL.

  When the future load FL is lower than the load threshold TL, the data control unit 131 determines that a server resource tightness does not occur due to a load generated when the application 141 processes the controlled sensor data CSD. . In this case, the data control unit 131 transmits the controlled sensor data CSD to the destination application 141.

  On the other hand, when the future load FL exceeds the load threshold TL, the data control unit 131 determines that a server resource tightness occurs due to a load generated when the application 141 processes the controlled sensor data CSD. In this case, the data control unit 131 updates the controlled sensor data CSD by referring to the control method storage unit 132 and applying the control method stored in the control method storage unit 132 to the controlled sensor data CSD. Specifically, the data control unit 131 acquires a control method for the application 141 using the controlled sensor data CSD with reference to the control method storage unit 132. The data control unit 131 updates the controlled sensor data CSD by applying the obtained control method to the controlled sensor data CSD. The updated controlled sensor data CSD is also referred to as updated controlled sensor data RCSD. Note that whether to update the controlled sensor data CSD when the future load FL is equal to the load threshold TL may be determined in advance.

  The data control unit 131 transmits, to the future load prediction unit 135, an instruction to predict the future load FL, which is a load that may occur when the application 141 processes the updated controlled sensor data RCSD, together with the updated controlled sensor data RCSD. . The data control unit 131 receives the future load FL predicted according to the prediction instruction from the future load prediction unit 135. The data control unit 131 repeats updating the updated controlled sensor data RCSD until the received future load FL falls below the load threshold TL.

  The control method storage unit 132 stores a sensor data control method for each application 141. The sensor data control method can be configured by performing a characteristic check for each application 141 in advance.

  FIG. 3 is an example of the control method stored in the control method storage unit 132 (control method 120). The control method 120 includes a control method for the face image analysis application (application A) and the river flood detection application (application B). In the case of the face image analysis application (application A), the data control unit 131 acquires a control method with reference to the control method storage unit 132 and performs control to reduce the number of image bits of the sensor data SD or the controlled sensor data CSD by half. I do. In the case of a river flood detection application (application B), the data control unit 131 acquires a control method with reference to the control method storage unit 132 and performs control to reduce the number of data of the sensor data SD or the controlled sensor data CSD by half. .

  The server load storage unit 133 stores load information of the server 15 (hereinafter, referred to as server load information).

  FIG. 4 is an example (server load information 130) of server load information stored in the server load storage unit 133. The server load information 130 includes information for associating a server ID (Identification) for identifying the server 15 with a server load indicating a load state of the server 15.

  The server load collecting unit 134 periodically acquires load information of the server 15 from the server 15. The server load collecting unit 134 stores the obtained load information in the server load storage unit 133.

  The future load prediction unit 135 receives a prediction instruction of the future load FL from the data control unit 131 together with the sensor data SD. Upon receiving the prediction instruction, the future load prediction unit 135 acquires the current load CL by referring to the server load storage unit 133. The future load prediction unit 135 combines the current load CL, which is the current load of the server 15, and the additional load NL, which is a load that can be generated on the server 15 by the received sensor data SD, according to the received prediction instruction. Predict the future load FL. The future load prediction unit 135 transmits the future load FL estimated according to the prediction instruction to the data control unit 131.

  The application load characteristic storage unit 136 stores a load characteristic (hereinafter, referred to as an application load characteristic) for each application 141 constituting the application group 14. The application load characteristics can be configured by performing a load test for each application 141 in advance. For example, the application load characteristic storage unit 136 is configured for each application group 14. Note that the application load characteristic storage unit 136 may be configured for each application 141 included in the application group 14.

  FIG. 5 is a configuration example of an application load characteristic 161 stored in the application load characteristic storage unit 136 when the application 141 is a face image analysis application (application A). When the application 141 is a face image analysis, the sensor data SD is image data, and the application load is determined according to the number of persons appearing in the image and the image size. Therefore, the application load characteristics storage unit 136 stores the application load characteristics 161 in which the loads generated per person in the image are summarized for each image size.

  In the case of the example of FIG. 5, the future load prediction unit 135 acquires the number of persons in the image by searching in the image. Further, the future load prediction unit 135 refers to the application load characteristic storage unit 136, and acquires the load generated per person in the image according to the image size. The future load prediction unit 135 calculates the additional load NL using the generated load per person in the image according to the image size acquired from the application load characteristic storage unit 136 and the number of people in the image acquired by the search in the image. Predict. For example, when the image size is 768 and the number of people in the image is 5, the additional load NL is 1.0 (= 0.2 × 5) because the generated load per person in the image is 0.2. Is calculated.

  FIG. 6 is a configuration example of the application load characteristic 162 stored in the application load characteristic storage unit 136 when the application 141 is a river flood detection application (application B). The river inundation detection application detects inundation using the water level meter data of the river as the sensor data SD. River flood detection applications have a low need to process data when river water levels are low and the risk of flooding is low. Therefore, in order to avoid an unnecessary increase in the load on the server 15, the number of times that the sensor 111 transmits data may be reduced. In this case, the application 141 generates a load proportional to the number of bits of the sensor data SD. Therefore, the load generated per bit of the sensor data SD is stored in the application load characteristic storage unit 136.

  In the case of the example of FIG. 6, the future load prediction unit 135 measures the data size of the sensor data SD. Further, the future load prediction unit 135 refers to the application load characteristic storage unit 136 and acquires the generated load per bit number of the sensor data SD. The future load prediction unit 135 predicts an additional load NL using the generated load per bit number of the sensor data SD acquired from the application load characteristic storage unit 136 and the measured data size. For example, when the data size is 100 bits, the additional load NL is calculated as 0.1 (= 0.001 × 100) because the generated load per bit of the sensor data SD is 0.001.

  The above is the description of the configuration of the data processing system 1. Next, the operation of the data processing system 1 will be described.

(motion)
FIG. 7 is a flowchart for explaining the operation of the data control device 13. Note that the processing according to the flowchart of FIG. 7 will be described with the data control device 13 as the subject of operation.

  7, first, the data control device 13 receives the delivery data DD from the data delivery device 12 (step S11).

  Next, the data control device 13 predicts a future load FL when the sensor data SD included in the delivery data DD is applied to the application 141 set at the destination address (step S12).

  Here, the data control device 13 compares the predicted future load FL with the load threshold TL (step S13).

  When the future load FL exceeds the load threshold TL (Yes in Step S13), the data control device 13 controls the sensor data SD and generates the controlled sensor data CSD (Step S14). On the other hand, when the future load FL is lower than the load threshold TL (No in Step S13), the process proceeds to Step S18.

  Next, the data control device 13 predicts a future load FL 'when the controlled sensor data CSD is applied to the application 141 set at the destination address (step S15).

  Here, the data control device 13 compares the predicted future load FL 'with the load threshold TL (step S16).

  When the future load FL 'exceeds the load threshold TL (Yes in step S16), the data control device 13 controls the controlled sensor data CSD to generate the controlled sensor data CSD (step S17). After step S17, the process returns to step S15. On the other hand, when the future load FL 'is lower than the load threshold TL (No in Step S16), the process proceeds to Step S18.

  Then, the data control device 13 transmits the sensor data SD or the controlled sensor data CSD to the application 141 set to the destination address (Step S18).

  The above is the description of the operation of the data control device 13.

  As described above, the data processing system according to the present embodiment includes the server, at least one application installed on the server, and at least one sensor that acquires sensor data processed by the application. Further, the data processing system of the present embodiment includes a data distribution device and a data control device. The data delivery device specifies a delivery destination application that processes the sensor data acquired by the sensor, and delivers the sensor data with the designated delivery destination to the data control device. The data control device receives, from the data delivery device, the sensor data whose delivery destination is specified. The data control device predicts a future load that may occur when the sensor data is processed by the destination application based on the load information of the server on which the destination application is installed and the load characteristics of the destination application. The data control device controls the sensor data according to a result of the comparison between the future load and the load threshold, and transmits the controlled sensor data to the application at the delivery destination.

  The data processing system of the present embodiment repeatedly updates the controlled sensor data CSD until the future load FL falls below the load threshold TL. As a result, a load that can be generated when the sensor data SD or the controlled sensor data CSD is processed by the application set at the destination address can be reduced, so that the server resources can be prevented from being tight. In other words, according to the data processing system of the present embodiment, it is possible to avoid a halt or delay of the application and to prevent a significant reduction in accuracy, so that the execution accuracy of the application can be stabilized even in an edge environment with limited computational resources. It becomes possible to make it.

  The data control device of the present embodiment determines a control method for the sensor data SD received from the data distribution device, and generates controlled sensor data CSD in which the determined control method is applied to the sensor data SD.

  The data control device of the present embodiment acquires load information from a server and holds load characteristic information of an application. The data control device predicts a future load that will occur in the future from the sensor data SD using the load information of the server and the load correlation of the application. Further, the data control device compares the predicted future load with a load threshold which is a control parameter held by the device itself, and determines a control method in which the future load falls below the load threshold. As a result, according to the data control device of the present embodiment, it is possible to stabilize the execution accuracy of the application while stabilizing the system.

  In other words, according to the present embodiment, it is possible to avoid a system instability due to a shortage of server resources and to avoid a decrease in application execution accuracy due to application stoppage and delay. Therefore, according to the present embodiment, it is possible to increase the execution accuracy of the application.

(Second Embodiment)
Next, a data control device according to a second embodiment of the present invention will be described with reference to the drawings. The data control device of the present embodiment is a higher-level concept of the data control device of the data processing system according to the first embodiment.

  FIG. 8 is a block diagram showing a configuration of the data control device 20 of the present embodiment. As shown in FIG. 8, the data control device 20 includes a data control unit 210 and a load prediction unit 250. In the data control unit 210, the data control unit 210 is connected to a data delivery device 22 that delivers sensor data obtained by a sensor (not shown) and an application 240 designated as a delivery destination of the sensor data.

  The data control unit 210 receives from the data delivery device 22 the sensor data designated by the delivery destination application 240. The data control unit 210 controls the sensor data according to a comparison result between a future load and a load threshold which may occur when the sensor data is processed by the application 240 of the delivery destination. The data control unit 210 transmits the controlled sensor data to the application 240 of the delivery destination.

  The load estimating unit 250 is based on load information of a server (not shown) on which the application 240 of the delivery destination is installed and load characteristics of the application 240, and a future that may occur when the application 240 of the delivery destination processes the sensor data. Predict the load. Further, the load prediction unit 250 predicts a future load that may be generated when the sensor data controlled by the data control unit 210 is processed by the destination application 240.

  As described above, the data control device according to the present embodiment includes the data control unit and the load prediction unit. The data control unit receives the sensor data in which the application of the delivery destination is specified. The data control unit controls the sensor data according to the result of comparing the load threshold with a future load that may occur when the sensor data is processed by the destination application, and transmits the controlled sensor data to the destination application I do. The load estimating unit is configured to calculate a future load that may occur when processing the sensor data controlled by the data control unit in the destination application based on the load information of the server on which the application of the destination is installed and the load characteristics of the application. Predict.

  According to the data control device of the present embodiment, the execution accuracy of an application can be stabilized even in an edge environment with limited computational resources.

  For example, the data control device repeats the control of the sensor data until the future load falls below the load threshold.

  For example, the data control device includes a server load collecting unit that collects, from the server, load information of the server on which the application is installed.

  For example, the data control unit includes a control method storage unit that stores a control method for each application. The data control unit acquires the control method of the application at the delivery destination with reference to the control method storage unit, and controls the sensor data based on the acquired control method.

  For example, the server load collection unit includes a server load storage unit that stores load information of a server on which an application is installed. Further, the load prediction unit includes an application load characteristic storage unit that stores a load characteristic for each application. The load prediction unit refers to the server load storage unit to obtain the load information of the server on which the destination application is installed, and obtains the load characteristics of the destination application by referring to the application load characteristic storage unit. The load prediction unit predicts a future load that may occur when the sensor data controlled by the data control unit is processed by the destination application based on the obtained server load information and the application load characteristics.

  For example, the load prediction unit predicts a future load based on a correlation between at least one of the amount and content of the sensor data and the load of the application.

  For example, when the destination application receives the specified sensor data, the data control unit sends a future load prediction instruction and sensor data that may occur when the destination application processes the sensor data to the load prediction unit. Send. The data control unit receives a future load predicted according to the prediction instruction from the load prediction unit, and compares the received future load with a load threshold. When the future load exceeds the load threshold, the data control unit controls the sensor data. Then, the data control unit transmits to the load prediction unit an instruction to predict a future load that may occur when the controlled sensor data is processed by the application of the delivery destination and the controlled load threshold. The data control unit receives a future load predicted according to the prediction instruction from the load prediction unit, and compares the received future load with a load threshold. The data control unit controls the controlled sensor data when the future load exceeds the load threshold. The data control unit transmits the load instruction with a predicted load instruction and a controlled load threshold that may occur when the controlled sensor data is processed by the application of the delivery destination. Repeat until below.

  For example, when the load prediction unit receives the prediction instruction and the sensor data from the data control unit, the load prediction unit predicts an additional load that may be generated when the destination application processes the sensor data based on the load characteristics of the application. The load prediction unit predicts a future load by combining the current load and the additional load of the server based on the load information of the server on which the application is installed.

(hardware)
Here, a hardware configuration for executing processing of the data processing system according to each embodiment of the present invention will be described using the information processing device 90 in FIG. 9 as an example. For example, the information processing device 90 is realized by a computer or a server device. Note that the information processing apparatus 90 in FIG. 9 is a configuration example for executing the processing of the data processing system of each embodiment, and does not limit the scope of the present invention.

  As shown in FIG. 9, the information processing device 90 includes a processor 91, a main storage device 92, an auxiliary storage device 93, an input / output interface 95, and a communication interface 96. In FIG. 9, the interface is abbreviated as I / F (Interface). The processor 91, the main storage device 92, the auxiliary storage device 93, the input / output interface 95, and the communication interface 96 are connected to each other via a bus 99 so as to be able to perform data communication. The processor 91, the main storage device 92, the auxiliary storage device 93, and the input / output interface 95 are connected to a network such as the Internet or an intranet via a communication interface 96.

  The processor 91 expands the program stored in the auxiliary storage device 93 or the like into the main storage device 92 and executes the expanded program. In the present embodiment, a configuration using a software program installed in the information processing device 90 may be used. The processor 91 executes processing by the data processing system according to the present embodiment.

  The main storage device 92 has an area where programs are loaded. The main storage device 92 may be a volatile memory such as a DRAM (Dynamic Random Access Memory). Further, a nonvolatile memory such as an MRAM (Magnetoresistive Random Access Memory) may be configured and added as the main storage device 92.

  The auxiliary storage device 93 stores various data. The auxiliary storage device 93 is configured by a local disk such as a hard disk or a flash memory. Note that it is also possible to store various data in the main storage device 92 and omit the auxiliary storage device 93.

  The input / output interface 95 is an interface for connecting the information processing device 90 and peripheral devices. The communication interface 96 is an interface for connecting to an external system or device via a network such as the Internet or an intranet based on standards and specifications. The input / output interface 95 and the communication interface 96 may be shared as an interface for connecting to an external device.

  The information processing device 90 may be configured to connect input devices such as a keyboard, a mouse, and a touch panel as needed. These input devices are used for inputting information and settings. When a touch panel is used as an input device, the display screen of the display device may be configured to also serve as an interface of the input device. Data communication between the processor 91 and the input device may be mediated by the input / output interface 95.

  Further, the information processing apparatus 90 may be provided with a display device for displaying information. When a display device is provided, it is preferable that the information processing device 90 includes a display control device (not shown) for controlling display on the display device. The display device may be connected to the information processing device 90 via the input / output interface 95.

  Further, the information processing device 90 may be provided with a disk drive as necessary. The disk drive is connected to the bus 99. The disk drive mediates between a processor 91 and a recording medium (program recording medium) not shown, such as reading of a data program from the recording medium and writing of a processing result of the information processing device 90 to the recording medium. The recording medium can be realized by an optical recording medium such as a CD (Compact Disc) and a DVD (Digital Versatile Disc). The recording medium may be realized by a semiconductor recording medium such as a USB (Universal Serial Bus) memory or an SD (Secure Digital) card, a magnetic recording medium such as a flexible disk, or another recording medium.

  The above is an example of the hardware configuration for enabling the data processing system according to each embodiment of the present invention. Note that the hardware configuration in FIG. 9 is an example of a hardware configuration for executing arithmetic processing of the data processing system according to each embodiment, and does not limit the scope of the present invention. Further, a program that causes a computer to execute a process related to the data processing system according to each embodiment is also included in the scope of the present invention. Furthermore, a program recording medium on which a program according to each embodiment is recorded is also included in the scope of the present invention.

  The components of the data processing system of each embodiment can be arbitrarily combined. Also, the components of the data processing system of each embodiment may be realized by software or may be realized by a circuit.

  As described above, the present invention has been described with reference to the exemplary embodiments. However, the present invention is not limited to the above exemplary embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

Reference Signs List 1 data processing system 11 sensor group 12 data delivery device 13 data control device 14 application group 15 server 20 data control device 22 data delivery device 111 sensor 131 data control unit 132 control method storage unit 133 server load storage unit 134 server load collection unit 135 Future load prediction unit 136 Application load characteristic storage unit 141 Application 210 Data control unit 240 Application 250 Load prediction unit

Claims (10)

The destination application receives the specified sensor data, and controls the sensor data in accordance with a comparison result between a future load and a load threshold that may occur when processing the sensor data in the destination application, Data control means for transmitting the controlled sensor data to the destination application,
The future that may occur when the sensor application controlled by the data control unit is processed by the application of the destination based on the load information of the server on which the application of the destination is installed and the load characteristics of the application. A data control device comprising: a load prediction unit that predicts a load. The data control means includes:
The data control device according to claim 1, wherein control of the sensor data is repeated until the future load falls below the load threshold.   The data control device according to claim 1, further comprising a server load collecting unit configured to collect, from the server, load information of the server on which the application is installed. The data control means includes:
Including a control method storage means for storing a control method for each application,
The server load collecting means,
Server load storage means for storing load information of the server on which the application is installed,
The load prediction means includes:
Application load characteristics storage means for storing the load characteristics for each application,
The data control means includes:
Acquiring the control method of the application of the delivery destination with reference to the control method storage means, controlling the sensor data based on the acquired control method,
The load prediction means includes:
Acquiring the load information of the server on which the application of the delivery destination is installed by referring to the server load storage unit, and acquiring the load characteristics of the application of the delivery destination by referring to the application load characteristic storage unit, A method for predicting the future load that may occur when the sensor data controlled by the data control unit is processed by the application at a delivery destination based on the acquired load information of the server and load characteristics of the application. 4. The data control device according to 3. The load prediction means includes:
The data control device according to claim 1, wherein the future load is predicted based on a correlation between at least one of an amount and a content of the sensor data and a load of the application. The data control means includes:
When the destination application receives the specified sensor data,
Sending the future load prediction instruction and the sensor data that may occur when the sensor data is processed by the application at the delivery destination to the load prediction unit,
Receiving the future load predicted according to the prediction instruction from the load prediction means,
Comparing the received future load with the load threshold,
When the future load exceeds the load threshold,
Controlling the sensor data,
Transmitting the predicted instruction and the controlled load threshold of the future load that may occur when the controlled sensor data is processed by the application of the delivery destination to the load predicting unit,
Receiving the future load predicted according to the prediction instruction from the load prediction means,
Comparing the received future load with the load threshold,
When the future load exceeds the load threshold,
Controlling the controlled sensor data,
The future load is configured to transmit, to the load prediction unit, the prediction instruction of the future load and the controlled load threshold that may occur when the controlled sensor data is processed by the application at the delivery destination. The data control device according to any one of claims 1 to 5, wherein the data control device repeats the process until the load value falls below a load threshold value. The load prediction means includes:
When receiving the prediction instruction and the sensor data from the data control means,
Based on the load characteristics of the application, predict an additional load that may occur when the sensor data is processed by the application at the delivery destination, and calculate a current load of the server based on load information of the server on which the application is installed. 7. The data control device according to claim 6, wherein the future load is predicted by combining the load with the additional load. Server and
At least one application installed on the server;
At least one sensor for acquiring sensor data to be processed by the application,
A data delivery device that designates the application of a delivery destination that processes the sensor data acquired by the sensor, and delivers the sensor data whose delivery destination is designated;
Receiving the sensor data whose delivery destination is specified from the data delivery device, and delivering the sensor data based on load information of the server on which the application of the delivery destination is installed and load characteristics of the application of the delivery destination. Predicting a future load that may occur when processing by the application, controlling the sensor data according to a comparison result between the future load and a load threshold, and controlling the sensor data under the control of the application as a delivery destination. A data processing system comprising: The destination application receives the specified sensor data,
Based on the load information of the server where the application of the delivery destination is installed and the load characteristics of the application, predict a future load that may occur when the sensor data is processed by the application of the delivery destination,
Controlling the sensor data according to a comparison result between the future load and a load threshold that may occur when processing the sensor data in the application at the destination,
A data control method for transmitting the controlled sensor data to the destination application. A process in which the destination application receives the specified sensor data,
Based on the load information of the server where the application of the delivery destination is installed and the load characteristics of the application, a process of predicting a future load that may occur when the sensor data is processed by the application of the delivery destination,
A process of controlling the sensor data according to a comparison result between the future load and a load threshold that may occur when processing the sensor data in the application at the delivery destination;
Transmitting a controlled sensor data to the application at a delivery destination.

Images