KR20210083117A - Edge computing processing method based on data type and sensitivity - Google Patents

Abstract

The present invention relates to a method for processing edge computing based on a data type and sensitivity, capable of assigning a priority based on attributes of data input from a sensor and a controller and of transmitting the data to a cloud server based on priority. According to the present invention, the method for processing edge computing based on the data type and the sensitivity includes: a first step of collecting sensor data and control data generated from a plurality of Internet of Thing (IoT) devices and a plurality of controllers; a second step of verifying the validity of the sensor data and the control data collected in the first step or of determining whether previous sensor data and previous control data are changed; a third step of assigning a weight based on urgency and sensitivity of each effective data and of classifying data; and a fourth step of classifying the data classified in the third step into priorities in at most (NxM) stages, of producing at most (N×M) number of queues with respect to each priority, and of storing the data into the queue based on the priorities. According to the present invention, in the method for processing edge computing based on data type and sensitivity, the conventional IoT or automatic control service operation is maintained while a data traffic is minimized, and cloud service costs are minimized.

Description

EDGE COMPUTING PROCESSING METHOD BASED ON DATA TYPE AND SENSITIVITY

The present invention relates to a data type and sensitivity-based edge computing processing method, and more specifically, to give priority according to the properties of data input from a sensor device and a control device, and to transmit data to a cloud server according to the priority It relates to a data type and sensitivity-based edge computing processing method.

In the rapidly developing Internet of Thing (IoT) market, the current cloud computing model has limitations in satisfying the explosively increasing traffic volume of end devices and the increasing demands of users. Cloud computing has many problems such as concentration of load, limitation of real-time connection, security problems, and the need for complex DB design. An edge computing model has been proposed to overcome these problems. The edge computing model performs calculations at a physically nearby gateway, PC, phone, set top box, etc., and responds quickly in real time, provides a convenient interface, distributes the load on the cloud server, and improves security do it

The IoT server or automatic control server installed in the cloud processes data collected from numerous sensor devices and control devices installed in the field in real time, and not only communication traffic but also computing resources for data processing are burdened. Have.

Therefore, efforts are being made to reduce the data transmitted from the field to the cloud using the edge computing concept, but most are implemented as a gateway, so data generated from sensor devices and control devices are delivered to the cloud, resulting in data traffic and service load. had an increasing problem.

Korean Patent Publication No. 10-2018-0047070 (published on May 10, 2018)

An object of the present invention is to provide a data type and sensitivity-based edge computing processing method that can reduce cloud server usage fees and service load generated in the server by reducing data traffic delivered to the cloud.

According to the data type and sensitivity-based edge computing processing method according to the present invention, there is an effect of minimizing the cloud service cost while minimizing data traffic while maintaining the operation of the conventional IoT or automatic control service. In addition, there is an effect of reducing the overload of the cloud server by reducing the collection of unnecessary data transmitted to the cloud server or the data that is frequently collected unnecessarily by using edge computing to focus on the unique functions that the server needs to perform.

According to the present invention, since transmission priority is determined according to urgency and sensitivity according to data type, it is possible to solve the problem of delay in transmission of important data due to transmission of insignificant data.

Since the present invention has a data queue, data is continuously updated in the queue according to urgency and sensitivity even when there is a problem in data transmission due to a transmission line problem or a system down of the cloud server in the process of being transmitted to the cloud server. is performed, there is an advantage that the latest data can be transmitted according to the priority when the transmission line is restored or the cloud server is operating normally.

In addition, in the present invention, since the policy delivered to the edge computer can be changed in real time, the IoT server or automatic control server can change the urgency and sensitivity policy according to the time zone and the day of the week to change the system operation in real time, not a single policy. This has the effect of improving the existing fixed system operation method into a flexible system.

1 is a block diagram of an edge computing system according to an embodiment of the present invention.
2 is a block diagram of an edge computer according to the present invention.
3 is a flowchart showing a data processing flow performed in an edge computer.

The terms used in the present invention are only used to describe specific embodiments, and are not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly dictates otherwise. In this specification, terms such as "comprises" or "have" are intended to designate that the features, numbers, steps, operations, components, parts, or combinations thereof described in the specification exist, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

In addition, in this specification, "on or on top of" means to be located above or below the target part, and does not necessarily mean to be located above the direction of gravity. Also, when a part of a region, plate, etc. is said to be “on or on” another part, it is not only when another part is in contact with or spaced “on or on” another part, but also when another part is in the middle. Including cases where there is

In addition, in this specification, when a component is referred to as "connected" or "connected" with another component, the component may be directly connected or directly connected to the other component, but in particular It should be understood that, unless there is a description to the contrary, it may be connected or connected through another element in the middle.

Also, in this specification, terms such as first, second, etc. may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another.

In the present invention, the edge computer classifies each value collected from the sensor device and the control device into previously set importance levels, sets the urgency of data transmission to N levels for each classified importance type, and provides sensitivity to the collected current values. Through verification, M level is determined for the sensitivity of data to the system, and (N x M) transmission queues are created according to the combination of these N level urgency and M level sensitivity. The created queue determines the priority of data delivered to the cloud server. In other words, the maximum value is given to the type of high urgency and the minimum weight is given to the type with low urgency, respectively, and the sensitivity of the M level is weighted in the same way as the urgency, so that the final priority is the data. It is judged by the product of the urgency weight and the sensitivity weight for each type, and the larger the value, the higher the transmission priority, so that data transmission has priority.

Hereinafter, preferred embodiments, advantages and features of the present invention will be described in detail with reference to the accompanying drawings.

Warden. Please check the underlined parts below, and if necessary, make corrections and supplementary explanations.

1 is a block diagram of an edge computing system according to an embodiment of the present invention. It consists of a plurality of sensor devices 10 , a plurality of control devices 20 , one or a plurality of edge computers 30 and a cloud server 50 .

The sensor device 10 is a device that includes a communication module and senses and transmits physical properties such as ambient temperature and pressure. The control device 20 is a device that includes a communication module and controls the operation of the IoT device. An example of the control device 20 may be a servo motor. The sensor device 10 and the control device 20 transmit sensing data and control data to the edge computer 30 through a local wired or wireless network.

The edge computer 30 does not send the data generated from the plurality of sensor devices 10 and the plurality of control devices 20 to a centralized data center such as the cloud, but processes the data in real time at the site or near field where the data is generated. A computer used in a computing method that supports acceleration. The edge computer 30 may be implemented in various forms because it only needs to have data processing capability. For example, in addition to embedded systems, smartphones or personal computers can also be used as edge computers. The edge computer 30 efficiently processes large-capacity data that can be processed in the vicinity of the source (source), thereby significantly shortening the data processing time of the cloud server and reducing the Internet bandwidth usage used for communication.

The edge computer 30 according to the present invention receives data collected from a sensor device and a control device in the field, classifies the urgency and sensitivity of the data, sets a priority according to the classified urgency and sensitivity, and according to the priority It is a computer device that creates a data transmission queue and transmits it to the cloud computer 50 through the Internet network with a predetermined period to the cloud server according to the priority.

The cloud server 50 is a server computer that processes data input from a plurality of edge computers 30, and controls data input from the IoT server 51 and the control device that processes and controls sensor data input from the IoT device. An automatic control server 55 for processing and controlling is provided.

Hereinafter, data processing performed by the edge computer 30 according to the present invention will be described. 2 is a block diagram of an edge computer according to the present invention, and FIG. 3 is a flowchart showing a data processing flow performed in the edge computer. The edge computer 30 includes a data receiving unit, a data filtering unit, an urgency classifying unit, a sensitivity classifying unit, a priority classifying unit, a queue generating unit, a transmission period determining unit, and a data transmitting unit.

1. Data collection stage

The data collection step is performed by the data receiving unit, and the sensor data measured from the sensor device and the control device or the controlled control data (hereinafter, 'sensor data' and 'control data' collectively referred to as 'data') are transferred to the local network. It is the step of collecting through As a communication network type used for data collection, both wired and wireless communication may be used as shown in FIG. 1 .

The collected data may include both analog data in which values change gradually and continuously, and digital data in which values change discretely, such as 1 and 0 (On and Off). The data collected by the edge computer is basically acquired through communication using a set protocol, and it also functions to connect the sensor directly to the edge computer if necessary.

2. Data Filtering Steps

The data filtering step is performed by the data filtering unit, and the validation step of checking whether the collected sensor data and control data is within the range of values that should be in the sensor or control device property, and the corresponding sensor data and control after validation is completed This is the step of verifying the necessity of data transmission by checking whether the data has changed between the previous value and the current value. That is, the same value as the previous value is not transmitted by default, and if transmission is not performed due to a problem in the transmission path or the cloud server system is down, all values are transmitted.

The policy for data validation for each data collected from all sensor devices and control devices is preset in the storage device inside the edge computing, and the validity of the data is determined by comparing the collected values with the set input value range, and the amount of change It performs a function of passing the data filtering step only when the size of .

The policy for validation is set in the IoT server or automatic control server, and it is downloaded to the edge computer and used. If another sensor device is added or the verification method is changed, reset the policy for verification in the IoT server or automatic control server and download it to the edge computer to enable data filtering according to the new policy.

3. Urgency classification stage

The urgency classification step is performed by the urgency classification unit, and is a step of classifying the data urgency classification into N levels based on the importance of the validated data.

Data with high urgency is assigned a high weight, and data with low urgency is assigned a low weight, and all collected data is processed to have only one level among N levels.

The urgency quantifies the large and low impact on the operation of the IoT server or automatic control server. Data used for the purpose of monitoring simply corresponds to the low priority, and the system operator must immediately detect a change in the value. In the case of data, it is included in data with high urgency.

4. Sensitivity Classification Step

The sensitivity classification step is performed in the sensitivity classification unit, and the sensitivity of the sensor data or control data on the operation of the system is evaluated by comparing the value of the data with the value of the validated data and the range of the value to be monitored in the system. It is a step of classifying into dog levels.

By assigning a high weight to data with high sensitivity, and assigning a low weight to data with low sensitivity, data classified by urgency is classified to have one stage among M levels for each level.

Even for data with the same urgency, data that is not in the high sensitivity range may have a lower priority transmitted to the cloud server than the data in the relatively high sensitivity range. You can determine the priority of sending to the cloud server.

As illustrated in FIG. 3 , urgency classification is first performed on the filtered data and then sensitivity classification is processed later. However, urgency classification and sensitivity classification can be processed regardless of the execution order. For example, it goes without saying that sensitivity classification may be performed first and then urgency classification may be performed later.

5. Priority classification stage

The priority level classification step is performed in the priority classification unit, and the priority for data transmission considering urgency and sensitivity is determined according to weights assigned to N levels divided by urgency and M levels classified by sensitivity. is a step to

Since the same priority can be generated according to the weight design policy of urgency and sensitivity, the priority can be calculated as a maximum (N urgency levels x M sensitivity levels).

The higher the priority, the higher the priority when contention occurs in data transmission, and the number of data transmitted to the cloud per hour increases by designing the data transmission cycle to have a shorter transmission period compared to the low-priority data.

6. Queue creation and data loading phase

The queue creation and data loading steps are performed by the queue generator, and when the priority is determined, a separate queue is created for each priority in order to transmit data, and loading is performed on the queue that matches the priority calculated for each data collected in the created queue. step is made.

When a maximum of (N x M) queues are created, the transmission period determining unit is designed to transmit immediately rather than being loaded into the queue as the priority is higher, and as the priority decreases, after all data with higher priority is transmitted. can be designed to do so.

In addition, in order to reduce data traffic transmitted to the cloud server, data with low priority can be collected and transmitted at once rather than individually transmitted.

For this purpose, each priority has a different transmission period, and sensor data and control data that arrive within a predetermined transmission period are continuously accumulated in the queue until the transmission time comes, and when the transmission period reaches the transmission time, the data stored in the transmission queue until now are stored in the transmission queue. All data can be bundled into one transmission unit and transmitted to the cloud.

7. Data transfer steps

The data transmission step is performed by the data transmission unit, and the data loaded in the queue is sequentially transmitted to the cloud by priority. The edge computer 30 transmits to the cloud while repeatedly repeating the transmission task indefinitely according to the priority and the transmission period.

Although preferred embodiments of the present invention have been described above using specific terms, such terms are only for clearly describing the present invention, and the embodiments and described terms of the present invention depart from the spirit and scope of the following claims. It is self-evident that various changes and changes can be made without being performed. Such modified embodiments should not be individually understood from the spirit and scope of the present invention, but should be considered to fall within the scope of the claims of the present invention.

10: sensor device
20: control device
30: edge computer
50: cloud server
51: IoT server
55: automatic control server

Claims (5)

Edge computing processing based on data type and sensitivity that receives sensor data and control data generated from multiple IoT devices and multiple control devices, sets priorities, and processes data from edge computers that are transmitted to the cloud server according to the priority As a method,
A first step of collecting sensor data and control data generated from a plurality of IoT devices and a plurality of control devices;
a second step of verifying the validity of the sensor data and control data collected in the first step, or verifying the valid data by checking whether there is a change with the previous sensor data and control data;
A third step of classifying the data according to the importance of each valid data, and
A fourth step of classifying the data classified in the third step with the priority of the maximum (N×M) step, creating a maximum of (N×M) queues for each priority, and storing the queues according to the priority
- The highest priority data is stored in the maximum (N×M)th queue, and the lowest priority data is stored in the first queue -
Data type and sensitivity-based edge computing processing method comprising a.
According to claim 1,
The third step is
a step 3-1 of classifying each valid data into data having 1 to N urgency by applying a first weight according to the degree of urgency;
- The first weight is composed of a total of N differential weights, and the higher the urgency, the higher the weight is given to valid data -
Step 3-2 of classifying each data classified according to urgency by step 3-1 into data having 1 to M sensitivities by applying a second weight according to sensitivity
- The second weight consists of all M differential weights, and is performed so that the more sensitive the valid data, the higher the weight is given -
Data type and sensitivity-based edge computing processing method, characterized in that configured to include.
According to claim 1,
The third step is
Step 3-3 of classifying each valid data into data having 1 to M sensitivities by applying a second weight according to the sensitivity;
- The second weight consists of all M differential weights, and is performed so that the more sensitive the valid data, the higher the weight is given -
Step 3-4 of classifying each data classified according to sensitivity in step 3-1 into data having 1 to N urgency by applying a first weight according to urgency
- The first weight is composed of a total of N differential weights, and the higher the urgency, the higher the weight is given to valid data -
Data type and sensitivity-based edge computing processing method, characterized in that configured to include.
4. according to any one of claims 2 or 3,
As a step performed after the fourth step,
Data type and sensitivity-based edge computing processing method, characterized in that the data stored in the high-priority queue is transmitted to the cloud server at a shorter period than the data stored in the low-priority queue.
5. The method of claim 4,
Data type and sensitivity-based edge computing processing method, characterized in that the data stored in the low priority queue in the fifth step is collected and transmitted to the cloud server.

Images