KR101875341B1 - Message relay system and method based on machine learning - Google Patents

Abstract

The present invention discloses a technology about a machine learning based message relay system, and a method thereof. According to an embodiment of the present invention, the machine learning based message relay system receives a message including sensing data of a sensor from a local device that senses local situations through the sensor, performs validation procedures on the message, and relays the message that has gone through the validation procedures to a relevant destination. In addition, the machine learning based message relay system comprises: a message learning unit which stores both message identification information identifying a message of a local device and a message reliability value corresponding to the message identification information, in a learning result database, by learning preceding messages received and processed in the system; a header parsing unit which acquires header information of the received messages by parsing the header of the messages received from the local device; a validation range determining unit which compares the message reliability value corresponding to the message identification information included in the header information with a preset reference value, among message reliability values stored in the learning result database, and determines to omit a part or the entire process of the validation process when the relevant message reliability value is above the reference value; and a message transmission processing unit which processes the received message and transmits the same to a relevant destination when omitting the part or the entire process of the validation process depending on the above destination. The present invention is able to improve performance and efficiency of message processing.

Description

[0001] The present invention relates to a message relay system and method based on machine learning,

The present invention relates to a system and method for relaying messages based on machine learning, and more particularly, to a system and method for relaying messages to a destination by receiving a message including sensing data of the sensor from a field device, And more particularly, to a machine-based message relay system and method.

In general, a smart city or a ubiquitous city can detect various events occurring in various sites in a city through sensor network infrastructures composed of various sensors, and can detect events such as weather, traffic, accidents, Refers to a city that efficiently manages resource situations such as water, electricity, and gas, and provides appropriate response services. In such a smart city or the like, a message relay system or communication middleware that receives a message including sensing data of the sensor from a field device that senses a situation of a site through a sensor and relays the message to a destination such as a control system or a service providing system is applied do.

However, as disclosed in Korean Patent Laid-Open Nos. 10-2011-0117431 and 10-2011-0002362, existing technologies relay and process sensed data messages received from various field devices through the same procedure , There is a problem that message processing time is unnecessarily delayed and message processing performance and efficiency are deteriorated. Also, there is a problem that conventional technologies can not adaptively adapt to changes in reliability such as operational errors or failures of field devices that sense the field situation through sensors.

SUMMARY OF THE INVENTION The present invention is directed to a machine learning based message relay system and method for reducing message processing time and improving message processing performance and efficiency while ensuring reliability of sensed data messages relayed in smart city or ubiquitous city .

A machine learning based message relay system according to an embodiment of the present invention receives a message including sensing data of the sensor from a field device that senses a field situation through a sensor and performs a validation procedure And relaying a message passed through the validation procedure to a corresponding destination, the system comprising: message identification information for identifying a message of the field device; A message learning unit for storing a message reliability value corresponding to the identification information in a learning result database; A header parser for parsing a header of a message received from the field device to obtain header information of the received message; Comparing the message reliability value corresponding to the message identification information included in the header information and the preset reference value among the message reliability values stored in the learning result database, and if the corresponding message reliability value is equal to or greater than the reference value, A verification scope determining unit that determines to omit some procedures or the entire procedure; And a message transmission processing unit for processing the received message and transmitting the processed message to the corresponding destination in the absence of a part of the procedures or the entire procedures of the validity verification procedure according to the determination.

In an exemplary embodiment, the message learning unit may include: a learning data collection unit for collecting message identification information of the received message and metadata necessary for determining reliability of the received message from the header information acquired by the header parsing unit; A message reliability determination unit for determining whether the collected message identification information and meta data match a preset message trust condition and determining a message reliability value corresponding to the collected message identification information according to the determination result; And a learning result database for storing the determined message reliability value corresponding to the collected message identification information.

In one embodiment, the learning data collection unit may include, as message identification information of the received message, a message identification including a field device ID for identifying the field device, and a message ID for identifying a message type of the received message Information can be collected.

In one embodiment, the learning data collection unit may collect meta data including message generation time data and message size data as metadata required for determining the reliability of the received message.

In one embodiment, the message reliability determination unit determines whether or not the message trust condition has the same message identification information and exists or not in the received preceding message, has the same message identification information, relays success of the immediately preceding message received, The message reliability value corresponding to the collected message identification information can be determined in consideration of the periodicity compliance and the message size compliance.

In one embodiment, the message reliability determination unit may increase the existing message reliability value corresponding to the collected message identification information when the collected message identification information and meta data meet the message trust condition, If the collected message identification information and meta data do not match the message trust condition, the existing message reliability value corresponding to the collected message identification information may be reduced to a certain extent.

In one embodiment, the learning result database may receive the message transmission result of the message transmission processing unit and store the transmission result of the last received message for each message identification information.

In one embodiment, when the message reliability value corresponding to the message identification information included in the header information is equal to or greater than the first reference value but less than the second reference value, the verification range determination unit determines the validity of the received message It may decide to omit the header verification procedure and decide to omit the entire procedure of the validity verification procedure if the message reliability value is equal to or greater than the second reference value.

In one embodiment, the system may perform the validation procedure on the received message, but omit the header verification procedure of the received message during the validity verification procedure in accordance with the determination of the verification range determination unit, And a message verification unit that omits the entire procedure of the verification procedure.

In one embodiment, the message verification unit includes: a header verification unit for comparing the header information of the received message with predetermined header verification reference information to perform validity verification; A body decryption unit for decrypting the body when the body of the received message is encrypted; A body parser for parsing a body or a decoded body of the received message to obtain body information of the received message; And a body verification unit for comparing the body information with predetermined body verification reference information to perform validity verification.

In one embodiment, the message transmission processing unit may be configured such that, when the entire procedure of the validity verification procedure is omitted, or when the received message passes the procedure other than the procedure in which the omission is determined in the validity verification procedure, A routing path setting unit for setting a routing path of the received message to a destination for collecting sensing data; And a message reconfiguration unit for reconfiguring the format of the received message when the format of the received message is different from the format used at the destination.

A machine learning based message relaying method according to an embodiment of the present invention includes receiving a message including sensing data of a sensor from a field device for sensing a field situation through a sensor, and a message relaying method for relaying a message that has passed the validation procedure to a corresponding destination, the method comprising: learning the message received and processed by the system, And storing a message reliability value corresponding to the message identification information in a learning result database; A header parsing step of parsing a header of a message received from the field device to obtain header information of the received message; The system compares a message reliability value corresponding to the message identification information included in the header information among message reliability values stored in the learning result database with a predetermined reference value, and if the corresponding message reliability value is equal to or greater than the reference value, A step of determining the scope of the verification to decide to omit some or all procedures of the verification procedure; And a message transmission processing step in which the system processes the received message and transmits the processed message to a corresponding destination in a state in which some procedures or entire procedures of the validity verification procedure are omitted in accordance with the determination.

In one embodiment, the message learning step may include: learning data that collects, from the header information obtained in the header parsing step, the message identification information of the received message and the metadata necessary for determining the reliability of the received message, Collecting step; Determining a message reliability value corresponding to the collected message identification information according to a result of the determination, determining whether the collected message identification information and the meta data match the predetermined message trust condition, and determining a message reliability value corresponding to the collected message identification information; And a learning result storing step of storing the determined message reliability value in the learning result database in response to the collected message identification information by the system.

In one embodiment, the learning data collection step includes a step of acquiring, as the message identification information of the received message, the system, a field device ID for identifying the field device, and a message ID for identifying a message type of the received message And collecting the data.

In one embodiment, the learning data collection step may include collecting message generation time data and message size data as metadata required for the system to determine the reliability of the received message.

In one embodiment, the message trustworthiness determination step may include: determining, by the system, that the message has the same message identification information as the message trust condition, the existence of the received preceding message, the same message identification information, Determining a message reliability value corresponding to the collected message identification information in consideration of whether or not to comply with the message generation period, and whether or not to comply with the message size.

In one embodiment, the message reliability determination step may include determining whether the collected message identification information and meta data match the message trust condition, and if the system satisfies the message trust condition, And if the collected message identification information and the meta data do not meet the message trust condition, the system may reduce the existing message reliability value corresponding to the collected message identification information by a certain amount have.

In one embodiment, the verification range determination step may include: when the message reliability value corresponding to the message identification information included in the header information is equal to or greater than a first reference value but less than a second reference value, Deciding to omit the header verification procedure of the received message and determining that the system omits the entire procedure of the validity verification procedure if the corresponding message reliability value is equal to or greater than the second reference value.

In one embodiment, the method further comprises: before the message transmission processing step after the verification range determination step, the system performs the validity verification procedure on the received message, wherein, according to the determination in the verification range determination step And a message verification step of omitting the header verification procedure of the received message during the validity verification procedure or omitting the entire procedure of the validity verification procedure.

In one embodiment, the validity verification procedure performed in the message verification step may include: a header verification procedure for comparing validity of the received message with header verification information of the received message to verify validity; A body decryption procedure in which the system decrypts the body when the body of the received message is encrypted; A body parsing procedure in which the system parses the body or decoded body of the received message to obtain body information of the received message; And a body verification procedure in which the system compares the body information with predetermined body verification reference information to perform validity verification.

In one embodiment, the message transmission processing step may include: when the entire procedure of the validity verification procedure is omitted, or when the received message passes the procedure other than the procedure in which the omission is determined in the validity verification procedure, A routing path setting step of setting a routing path of the received message to a destination for collecting sensing data of the sensor; And a message reconfiguration step in which the system reconfigures the format of the received message if the format of the received message is different from the format used at the destination.

The embodiments according to the present invention can be implemented using a computer program recorded on a recording medium as a computer program for executing the above-described operations or methods through a computer system.

According to the present invention, the message relay system learns the processed messages received and processed in the system, determines the message reliability value for each message received at each field device, and determines the validity of the message according to the reliability value of the currently received message By reducing or omitting the verification procedure, it is possible to reduce the message processing time of the message relay system and improve the message processing performance and efficiency while ensuring the accuracy and reliability of the sensed data message relayed through the message relay system in the smart city or ubiquitous city have.

That is, when the reference performance value of the message relay system or the communication middleware is 1200 TPS, the existing technology uses about 1,089 TPS among 1200 TPS, and the present invention applying the machine learning based message learning algorithm has a Using only 392 TPS shows a usage rate of about 33%. According to the present invention, it is possible to improve the processing performance by 2.7 times or more as compared with the existing technology.

In addition, according to the present invention, the message relay system continuously learns the messages received and processed by the system and appropriately changes the determined determined message reliability values, It can adaptively cope with the change of the situation, and can easily manage the on-site apparatus.

It will be apparent to those skilled in the art that various embodiments of the present invention can be made without departing from the spirit and scope of the present invention.

1 is a diagram illustrating an example of a smart city network environment to which the present invention is applied.
2 is a block diagram illustrating a machine learning based message relay system according to an embodiment of the present invention.
3 is a flowchart illustrating a machine learning based message relaying method according to an embodiment of the present invention.
FIG. 4 is a flowchart illustrating a message reliability learning process applied to a message relaying method based on a machine learning according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings in order to clarify solutions for technical problems of the present invention. In the following description of the present invention, however, the description of related arts will be omitted if the gist of the present invention becomes obscure. In addition, the terms described below are defined in consideration of the functions of the present invention, and may be changed depending on the intention or custom of the designer, the manufacturer, and the like. Therefore, the definition should be based on the contents throughout this specification.

FIG. 1 shows an example of a smart city network environment to which the present invention is applied.

As shown in FIG. 1, the smart city network includes a plurality of field devices 10 installed at various sites of the smart city, a message relay system 10 for receiving a message from each of the field devices 10 and relaying the message to a corresponding destination 20 and a plurality of destination systems 30 for receiving messages of the field devices 10 via the message relay system 20.

Each of the field devices 10 includes at least one sensor 12 to periodically sense the field conditions at the site through the sensor 12 and to detect the sensing data of the sensor 12 according to a predetermined transmission protocol And transmits the generated message to the message relay system 20. The field device 10 may be configured as a variety of devices for detecting weather conditions such as weather, traffic, accidents, disasters, etc., or water resources such as water, electricity, and gas. In this case, the sensors provided in the field device 10 may be various types and shapes such as a temperature sensor, a humidity sensor, a pressure sensor, an infrared sensor, a magnetic field sensor, or a sensor for sensing a resource usage amount of water, electricity, Sensors. For example, the field apparatus 10 for monitoring the occurrence of forest fires in a forest area site senses a situation of a site through a plurality of temperature sensors or thermal sensors, periodically generates messages including sensing data of the respective sensors, To the message relay system 20. The field device 10 may be configured as a single physical device, or may be configured in the form of a local sensor network in which a plurality of sensors are distributed and operated in a wide area.

On the other hand, each message transmitted from the field devices has a format including a header and a body. The header of the message may include header information such as a field device ID that identifies the field device, a message ID that identifies the message type, a message generation time, a message size, a message generation period, a source IP address, a destination IP address, . In addition, the body of the message may include body information such as a sensor ID for identifying a corresponding sensor of the corresponding field device, sensing data of the corresponding sensor, and the like.

The destination system 30 corresponds to a server system or a computer system. The destination system 30 periodically receives messages of the field device 10 through the message relay system 20 and collects the sensing data included in each message And perform actions or services corresponding to the analysis results. The destination system 30 may be composed of various systems for controlling a meteorological situation such as weather, traffic, accidents, disasters, etc., or resource situations such as water, electricity, and gas, and providing corresponding services. For example, the destination system 30 may comprise a forest fire central control system. In this case, the destination system 30 periodically receives messages of the field device 10 that detects the occurrence of forest fires through a plurality of sensors at the forest area site through the message relay system 20, Sensing data can be collected and analyzed. If it is determined that the forest fire has occurred, the message relay system 20 provides the related information to the computer system of the fire fighting headquarters responsible for the forest fire occurrence area and transmits the information about the occurrence of the forest fires to the nearby residents' Emergency disaster letters can be sent out.

The machine learning based message relay system 20 according to the present invention receives a message including the sensing data of the sensor 12 from the field device 10 that senses the situation of the field through the sensor 12, And may relay the messages that have passed the validity verification procedure to the destination system 30 corresponding to the destination.

FIG. 2 is a block diagram of a machine learning-based message relay system 20 according to an embodiment of the present invention.

2, the machine learning-based message relay system 20 according to an embodiment of the present invention may include a communication unit 22, a communication middleware unit 24, a storage unit 26, and the like .

The communication unit 22 includes wired or wireless communication means and can perform data communication with the field device 10 and the destination system 30 or the like via a wired or wireless communication network.

The storage unit 26 includes a storage medium such as a memory, a buffer, and the like, and temporarily stores messages received from the field devices through the communication unit 22. In addition, the storage unit 26 stores message reliability condition information for determining reliability of received messages, verification criteria information used in a validity verification procedure of each message, routing path information corresponding to each destination system 30, And so on.

The communication middleware unit 24 may perform a validation procedure for messages received from the respective field devices 10 and may relay the message that has passed the validity verification procedure to the corresponding destination. In particular, the communication middleware unit 24 learns the prior messages received and processed by the message relay system 20 based on machine learning, and calculates a message reliability value for each message repeatedly received in each of the field devices 10 And the validity verification procedure of the corresponding message can be reduced or omitted according to the message reliability value of the currently received message. The communication middleware unit 24 includes a message receiving unit 100, a header parsing unit 110, a message learning unit 120, a verification range determination unit 130, a message transmission processing unit 140, (150), and the like.

The message receiving unit 100 may periodically and repeatedly receive a message including sensing data from the field device 10 by communicating with the field device 10 through the communication unit 22. [ When the field device 10 includes a plurality of sensors, the message receiving unit 100 may simultaneously receive or sequentially receive a plurality of messages corresponding to the plurality of sensors. According to an embodiment, the message receiving unit 100 may receive a single message including a plurality of sensing data corresponding to the plurality of sensors.

The header parsing unit 110 parses the header of the message received from the field device 10 to obtain header information of the received message and stores the header information in the storage unit 26 have.

The message learning unit 120 learns the preceding messages received and processed by the system 20, and generates message reliability information for identifying a message of the on-site device 10 and a message reliability value Can be stored in the learning result database. For this, the message learning unit 120 may include a learning data collection unit 122, a message reliability determination unit 124, and a learning result database 126.

The learning data collection unit 122 of the message learning unit 120 extracts the message identification information of the received message from the header information acquired by the header parsing unit 110 and the meta information necessary for determining the reliability of the received message Data can be collected. In this case, the learning data collection unit 122 may include a field device ID for identifying the field device 10 as message identification information of the received message, and a message ID for identifying a message type of the received message Message identification information can be collected. The learning data collection unit 122 collects metadata including message generation time data, message size data, message generation cycle data, and the like of the received message as metadata required for determining the reliability of the received message .

The message reliability determination unit 124 of the message learning unit 120 determines whether the collected message identification information and the meta data satisfy a predetermined message trust condition, A corresponding message reliability value may be determined and stored in the learning result database 126. [ In this case, the message reliability determination unit 124 has the same message identification information as the message trust condition, and has the same message identification information as the received message and has the same message identification information. The message reliability determination unit 124 determines whether the preceding message received successfully, The message reliability value corresponding to the collected message identification information can be determined in consideration of the periodicity compliance and the message size compliance.

The message reliability determination unit 124 may increase the existing message reliability value corresponding to the collected message identification information when the collected message identification information and meta data meet the message trust condition, If the collected message identification information and meta data do not match the message trust condition, the existing message reliability value corresponding to the collected message identification information may be reduced to a certain extent. For example, if the received message has the same message identification information as the received message, the received pre-message exists, the pre-message having the same message identification information as the received message is successfully relayed, The message reliability determining unit 124 may increase the existing message reliability value corresponding to the message identification information of the received message by a certain amount.

On the other hand, if the received message is the first received message from the corresponding field device 10, the message reliability determination unit 124 ends the reliability determination procedure and the validity verification procedure by the message verification unit 150 proceeds . In addition, although the received message has the same message identification information as the received message and the received pre-message exists, the pre-message having the same message identification information as the received message is not relayed due to failure of the validity verification Or if the received message does not comply with a predetermined message generation period, message size, or the like, the message reliability determination unit 124 determines an existing message reliability value corresponding to the message identification information of the received message as a predetermined .

The learning result database 126 of the message learning unit 120 may store the determined message reliability value corresponding to the collected message identification information. In this manner, the learning result database 126 may store the message reliability value for each message identification information of messages received from the field devices. In this case, the learning result database 126 may receive the message transmission result of the message transmission processing unit 140 and store the transmission result of the last received message for each message identification information.

The verification range determination unit 130 may determine a message reliability value corresponding to the message identification information of the received message included in the header information among the message reliability values stored in the learning result database 126, , And may decide to omit some procedures or entire procedures of the validity verification procedure if the corresponding message reliability value is greater than or equal to the reference value. In this case, if the message reliability value corresponding to the message identification information of the received message belongs to the lower class as being less than the first reference value, the verification range determination unit 130 may directly check the validity of the received message Can be determined. On the other hand, when the corresponding message reliability value is equal to or higher than the first reference value but falls below the second reference value that is larger than the first reference value, the verification range determination unit 130 determines that the received message has a validity It may decide to skip the header verification procedure of the received message during the procedure. Also, when the corresponding message reliability value belongs to the upper class as the second reference value or more, the verification range determination unit 130 may decide to omit the entire procedure of the validity verification procedure for the received message.

The message transmission processing unit 140 may process the received message and transmit the processed message to the corresponding destination in a state in which some procedures or entire procedures of the validity verification procedure are omitted according to the determination of the verification range determination unit 130. [ For example, when the verification range determination unit 130 determines to omit the entire validity verification procedure for the received message, the message transmission processing unit 140 processes the received message without the validity verification procedure To the destination. In addition, when the verification range determination unit 130 determines to omit the header verification procedure of the received message, the message transmission processing unit 140 determines that the received message has been subjected to the body verification related procedure of the received message The message can be processed and transmitted to the destination. For this, the message transmission processing unit 140 may include a routing path setting unit 142 and a message reconstruction unit 144.

The routing path setting unit 142 of the message transmission processing unit 140 determines whether the entire procedure of the validity verification procedure is omitted or when the received message passes the procedures other than the procedure in which the omission is determined in the validity verification procedure The routing path of the received message can be set to a destination for collecting the sensing data of the sensor. In addition, when a new destination system for collecting sensing data is further registered by receiving the message of the field device 10 in addition to the existing destination system, the routing path setting unit 142 sets the routing path setting unit 142, The system can set the routing path of the received message. In this case, the storage unit 26 of the message relay system 20 may store the routing path information corresponding to the message identification information of the messages relayed through the message relay system 20 in advance.

The message reconstruction unit 144 of the message transmission processing unit 140 may reconstruct the format of the received message when the format of the received message is different from the format used in the destination. For example, when the message formats used in the plurality of destination systems are different from each other, the message reconstruction unit 144 may reconfigure the format of the received message in a format suitable for the plurality of destination systems, and transmit the formatted message to the destination system .

Meanwhile, the message verification unit 150 performs a verification procedure on the received message, and performs a header verification procedure of the received message in the validity verification procedure according to the determination of the verification range determination unit 130 Or omit the entire procedure of the validation procedure. The message verification unit 150 may include a header verification unit 152, a body decoding unit 154, a body parsing unit 156, and a body verification unit 158.

The header verification unit 152 of the message verification unit 150 may compare the header information of the received message with predetermined header verification reference information to perform validity verification. However, if the verification range determination unit 130 determines to skip the entire header verification procedure or the entire verification procedure in the validity verification procedure for the received message, the header verification unit 152 may determine Do not perform header verification.

The body decryption unit 154, the body parsing unit 156 and the body verification unit 158 of the message verification unit 150 perform a procedure for body verification of the received message. That is, when the body of the received message is encrypted, the body decryption unit 154 can decrypt the body. The body parsing unit 156 may obtain the body information of the received message by parsing the body or the decoded body of the received message. The body verification unit 158 may verify the validity by comparing the body information with predetermined body verification reference information. The body decryption unit 154, the body parsing unit 156, and the body verification unit 158 may determine whether or not the entire validity verification procedure for the received message is to be omitted, Do not perform the procedure.

In one embodiment, the communication middleware portion 24 of the above-described message relay system 20 may be implemented as a microprocessor that executes the communication middleware to perform the above-described operations or procedures.

FIG. 3 is a flowchart illustrating a machine learning based message relaying method according to an exemplary embodiment of the present invention. 3, detailed configurations of the message relay system 20 shown in FIG. 2 and their operations will be described in detail.

As shown in FIG. 3, the message relay system 20 basically performs a validation procedure for messages received from each of the field devices 10, and transmits a message that has passed the validity verification procedure To the destination. In particular, the message relay system 20 learns the preceding messages received and processed by the system 20 based on machine learning, and determines a message reliability value for each message repeatedly received at each of the field devices 10 , The validity verification procedure of the corresponding message can be reduced or omitted according to the message reliability value of the currently received message.

More specifically, the message receiving unit 100 of the message relay system 20 communicates with the field device 10 through the communication unit 22 of the system 20, A message including data may be periodically repeatedly received (S300).

The header parsing unit 110 of the system 20 then parses the header of the message received from the field device 10 to obtain the header information of the received message, And stored in the storage unit 26 (S310).

The message learning unit 120 of the system 20 learns the preceding messages received and processed by the system 20 and generates message identification information for identifying the message of the field device 10, The corresponding message reliability value may be stored in the learning result database (S320). Hereinafter, the message reliability learning process performed by the message learning unit 120 of the system 20 will be described in more detail.

FIG. 4 is a flowchart illustrating a message learning process applied to a message relaying method based on a machine learning according to an embodiment of the present invention.

4, the learning data collection unit 122 of the message learning unit 120 extracts the message identification information of the received message from the header information acquired by the header parsing unit 110, The metadata necessary for determining the reliability of the received message can be collected (S400). In this case, the learning data collection unit 122 may include a field device ID for identifying the field device 10 as message identification information of the received message, and a message ID for identifying a message type of the received message Message identification information can be collected. The learning data collection unit 122 collects metadata including message generation time data, message size data, message generation cycle data, and the like of the received message as metadata required for determining the reliability of the received message .

Next, the message reliability determination unit 124 of the message learning unit 120 determines whether the collected message identification information and the meta data satisfy a predetermined message trust condition (S410 to S440) The message reliability value corresponding to the collected message identification information is determined (S450 and S460), and the determined message reliability value is stored in the learning result database 126 of the message learning unit 120 together with the collected message identification information (S470).

In this case, the message reliability determination unit 124 determines whether or not the message reliability condition has the same message identification information, the existence (S410) of the received preceding message, the same message identification information, The message reliability value corresponding to the collected message identification information can be determined in consideration of the message generation period (S420), the message generation period compliance (S430), and the message size compliance status (S440). In addition, if the collected message identification information and meta data satisfy both the message trust conditions, the message reliability determining unit 124 increases the existing message reliability value corresponding to the collected message identification information by a certain amount (S450). For example, if the received message has the same message identification information as the received message, the received pre-message exists, the pre-message having the same message identification information as the received message is successfully relayed, The message reliability determination unit 124 may increase the existing message reliability value corresponding to the message identification information of the received message by one.

On the other hand, if the collected message identification information and meta data do not match the message trust condition, the message reliability determination unit 124 reduces the existing message reliability value corresponding to the collected message identification information by a certain amount (S460). For example, if the received message has the same message identification information as the received message and the received pre-message does not exist, that is, if the received message is the first message received from the corresponding on-site device 10, the message reliability determination unit 124 The reliability verification procedure may be terminated and the validity verification procedure by the message verification unit 150 may be performed. In addition, although the received message has the same message identification information as the received message and the received pre-message exists, the pre-message having the same message identification information as the received message is not relayed due to failure of the validity verification Or if the received message does not comply with a predetermined message generation period, a message size, or the like, the message reliability determination unit 124 determines an existing message reliability value corresponding to the message identification information of the received message as 288 .

The learning result database 126 of the message learning unit 120 may store the determined message reliability value corresponding to the collected message identification information (S470). In this manner, the learning result database 126 may store the message reliability value for each message identification information of messages received from the field devices. In this case, the learning result database 126 may receive the message transmission result of the message transmission processing unit 140 and may further store the transmission result of the last received message for each message identification information.

The message learning unit 120 of the system 20 may repeat the above-described procedures when a subsequent message having the same message identification information is received (S480).

3, when the message reliability value for each message identification information is stored in the learning result database 126 as described above, the verification range determination unit 130 of the system 20 determines whether or not the learning result database 126 , A validity range of the received message may be determined by comparing a message reliability value corresponding to the message identification information of the received message included in the header information with a preset reference value have. That is, if the message reliability value is equal to or greater than the reference value, the verification range determination unit 130 may determine to omit some procedures or entire procedures of the validity verification procedure (S330, S332, and S334).

In this case, the message reliability value may be classified into a lower class, a middle class, and a higher class based on a first reference value and a second reference value that is larger than the first reference value. For example, if the specific field device 10 is configured to generate a message at five minute intervals and send it to the message relay system 20, the message relay system 20 may send a message of the field device 10 for 28 days When all the messages received for one day are successfully relayed, the message reliability value corresponding to the message identification information of the corresponding messages is 288. In this way, the message reliability value 2880, which is calculated when all the messages received for 10 days are successfully relayed, is set to the first reference value, and all the messages received for 30 days are successfully relayed The calculated message reliability value 8640 may be set to the second reference value.

If the message reliability value corresponding to the message identification information of the received message belongs to the lower class as being less than the first reference value of 2880, the verification range determination unit 130 determines the validity of the received message It can be decided to keep it. On the other hand, when the corresponding message reliability value belongs to the middle level as the first reference value of 2880 or more but less than the second reference value of 8640, the verification range determination unit 130 determines that the reception Lt; RTI ID = 0.0 > header verification < / RTI > Also, when the corresponding message reliability value belongs to the upper class as 8640 or more, which is the second reference value, the verification range determination unit 130 may decide to omit the entire procedure of the validity verification procedure for the received message.

Meanwhile, the message verification unit 150 of the system 20 performs a validity verification procedure (S340 to S346) on the received message, and the validity verification procedure The header verification procedure (S340) of the received message may be omitted or the entire procedure of the validity verification procedure (S340 to S346) may be omitted.

That is, if the message reliability value corresponding to the message identification information of the received message belongs to the lower class as less than the first reference value of 2880 (S332), the header verification unit 152 of the message verification unit 150 The header information obtained by the header parser 110 may be compared with predetermined header verification reference information to perform validity verification (S340). However, if the message reliability value belongs to a middle grade that is equal to or greater than the first reference value of 2880 or less than the second reference value of 8640 or belongs to an upper grade that is equal to or greater than the second reference value of 8640 (S334) The control unit 152 may not perform header verification on the received message.

In addition, the message reliability value corresponding to the message identification information of the received message belongs to the lower rank as less than the first reference value of 2880 (S332), or the message reliability value corresponding to the second reference value of less than 8640 The body decrypting unit 154, the body parsing unit 156 and the body verification unit 158 of the message verifying unit 150 may perform a process of verifying the body of the received message (S334) Can be performed. That is, when the body of the received message is encrypted, the body decryption unit 154 may decrypt the body (S342). The body parsing unit 156 may obtain the body information of the received message by parsing the body or the decoded body of the received message (S344). The body verification unit 158 may compare the body information with predetermined body verification reference information to perform validity verification (S346). If the message reliability value belongs to the upper class as 8640 or more as the second reference value in step S334, the body decryption unit 154, the body parsing unit 156, and the body verification unit 158 compare the received message The procedure of the body verification of FIG.

Then, the message transmission processing unit 140 of the system 20 processes the received message in a state where a part of the procedures or entire procedures of the validity verification procedure are omitted according to the determination of the verification range determination unit 130 To the corresponding destination (S350 to S370).

For example, when the message reliability value of the received message belongs to the upper class as 8640 or more, which is the second reference value, and the verification range determination unit 130 determines to omit the entire validity verification procedure for the received message, The message transmission processing unit 140 may process the received message without transmitting the validity verification procedure and transmit the processed message to the corresponding destination.

On the other hand, if the message reliability value of the received message belongs to the middle grade or lower grade as less than the second reference value of 8640, the verification range determination unit 130 may perform the header verification procedure of the received message during the validity verification procedure The message transmission processing unit 140 determines whether the received message is to be passed through the body verification related procedure or the entire validity verification procedure (S348), the received message can be processed and transmitted to the corresponding destination.

In the message transmission process, the routing path setting unit 142 of the message transmission processing unit 140 determines whether the entire procedure of the validity verification procedure is omitted, or when the received message is determined to be omitted in the validity verification procedure If a procedure other than the procedure is passed, the routing path of the received message can be set to a destination for collecting the sensing data of the sensor (S350). In this case, when a new destination system for collecting the sensing data is additionally registered in addition to the existing destination system, the routing path setting unit 142 sets the routing path setting unit And may set the routing path of the received message to the destination system, respectively. In this case, the storage unit 26 of the message relay system 20 may store the routing path information corresponding to the message identification information of the messages relayed through the message relay system 20 in advance.

If the format of the received message is different from the format used in the destination, the message reconstruction unit 144 of the message transmission processing unit 140 may reconstruct the format of the received message and transmit the formatted message to the corresponding destination (S360, S370). For example, when the message formats used in the plurality of destination systems are different from each other, the message reconstruction unit 144 may reconfigure the format of the received message in a format suitable for the plurality of destination systems, and transmit the formatted message to the destination system .

Next, the message transmission processing unit 140 may store the transmission result of the received message in the learning data collection unit 122 in response to the message identification information of the received message (S380).

Meanwhile, the embodiments according to the present invention can be implemented by a computer system and a computer program for driving such a computer system. When embodiments of the present invention are implemented as a computer program, the elements of the present invention are program segments that execute corresponding operations or tasks through the computer system. Such computer programs and program segments may be stored in various computer-readable media. A computer-readable recording medium includes all kinds of media for recording data that can be read by a computer system. For example, the computer-readable recording medium may include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like. In addition, the recording medium may be distributed to computer systems connected by various networks, and may store and execute program codes in a distributed manner.

As described above, according to the present invention, the message relay system learns the preceding messages received and processed by the system, determines the message reliability value for each message received at each field device, and determines the reliability value of the currently received message Accordingly, it is possible to reduce the message processing time of the message relaying system and to improve the message processing performance and reliability of the message relaying system by ensuring the accuracy and reliability of the sensed data message relayed through the message relay system in the smart city or ubiquitous city, The efficiency can be improved. That is, when the reference performance value of the message relay system or the communication middleware is 1200 TPS, the existing technology uses about 1,089 TPS among 1200 TPS, and the present invention applying the machine learning based message learning algorithm has a Using only 392 TPS shows a usage rate of about 33%. According to the present invention, it is possible to improve the processing performance by 2.7 times or more as compared with the existing technology. In addition, according to the present invention, the message relay system continuously learns the messages received and processed by the system and appropriately changes the determined determined message reliability values, It can adaptively cope with the change of the situation, and can easily manage the on-site apparatus. Furthermore, it should be understood that the embodiments according to the present invention can solve various technical problems other than those mentioned in the specification in the related technical field as well as the related art.

The present invention has been described with reference to specific embodiments. It will be apparent, however, to one skilled in the art that various modifications may be practiced within the technical scope of the invention. Therefore, the above-described embodiments should be considered from an illustrative point of view, not from a restrictive viewpoint. That is, the scope of the true technical idea of the present invention is shown in the claims, and all differences within the scope of equivalents thereof should be construed as being included in the present invention.

20: Message relay system 22:
24: communication middleware section 26: storage section
100: message receiving unit 110: header parsing unit
120: message learning unit 122: learning data collection unit
124: message reliability determination unit 126: learning result database
130: verification range determination unit 140: message transmission processing unit
142: routing path setting unit 144: message reconstruction unit
150: message verification unit 152: header verification unit
154: Body decryption unit 156: Body parsing unit
158: body verification unit

Claims (22)

A message including sensing data of the sensor is received from a field device for sensing a field situation through a sensor to perform a validation procedure for the message and a message passed through the validation procedure is relayed to a corresponding destination A message relay system comprising:
A message learning unit that learns the preceding messages received and processed by the system and stores message identification information for identifying a message of the field device and message reliability values corresponding to the message identification information in a learning result database;
A header parser for parsing a header of a message received from the field device to obtain header information of the received message;
Comparing the message reliability value corresponding to the message identification information included in the header information and the preset reference value among the message reliability values stored in the learning result database, and if the corresponding message reliability value is equal to or greater than the reference value, A verification scope determining unit that determines to omit some procedures or the entire procedure; And
And a message transmission processing unit for processing the received message and transmitting the processed message to the corresponding destination in the absence of some procedures or entire procedures of the validity verification procedure according to the determination. The method according to claim 1,
The message learning unit,
A learning data collection unit for collecting message identification information of the received message and metadata necessary for determining reliability of the received message from the header information acquired by the header parsing unit;
A message reliability determination unit for determining whether the collected message identification information and meta data match a preset message trust condition and determining a message reliability value corresponding to the collected message identification information according to the determination result; And
And a learning result database for storing the determined message reliability value corresponding to the collected message identification information. 3. The method of claim 2,
Wherein the learning data collection unit collects, as message identification information of the received message, message identification information including a field device ID for identifying the field device and a message ID for identifying a message type of the received message Based message relay system. 3. The method of claim 2,
Wherein the learning data collection unit collects metadata including message generation time data and message size data as metadata required for determining the reliability of the received message. 3. The method of claim 2,
Wherein the message reliability determining unit determines whether or not the preceding message having the same message identification information and having the same message identification information and having the same message identification information and succeeding in the relaying of the immediately preceding message, Wherein the message reliability value corresponding to the collected message identification information is determined in consideration of compliance. 3. The method of claim 2,
The message reliability determination unit may increase the existing message reliability value corresponding to the collected message identification information to a certain extent if the collected message identification information and meta data match the message trust condition, And if the meta data does not match the message trust condition, reduces the existing message reliability value corresponding to the collected message identification information by a certain amount. 3. The method of claim 2,
Wherein the learning result database stores the transmission result of the last received message for each message identification information in response to the message transmission result of the message transmission processing unit. The method according to claim 1,
Wherein the validation range determining unit omits the header verification procedure of the received message during the validity verification procedure when the message reliability value corresponding to the message identification information included in the header information is equal to or greater than the first reference value but less than the second reference value And determines to omit the entire procedure of the validity verification procedure if the message reliability value is equal to or greater than the second reference value. 9. The method of claim 8,
Wherein the system is configured to perform the validity verification procedure on the received message and to omit the header verification procedure of the received message in the validity verification procedure according to the determination of the verification scope determination unit or to perform the entire procedure of the validity verification procedure And a message verifying unit for deleting the message from the received message. 10. The method of claim 9,
Wherein the message verification unit comprises:
A header verification unit for verifying validity by comparing header information of the received message with predetermined header verification reference information;
A body decryption unit for decrypting the body when the body of the received message is encrypted;
A body parser for parsing a body or a decoded body of the received message to obtain body information of the received message; And
And a body verification unit for comparing the body information with predetermined body verification reference information to perform validity checking. The method according to claim 1,
The message transmission processing unit,
If the entire procedure of the validity verification procedure is omitted or if the received message passes the procedure other than the procedure in which the omission is determined during the validity verification procedure, A routing path setting unit for setting a routing path; And
And a message reconstruction unit for reconstructing a format of the received message when the format of the received message is different from a format used at a corresponding destination. The message relay system receives a message including sensing data of the sensor from a field device that senses a field situation through a sensor, performs a validation procedure for the message, and transmits a message that passes the validity verification procedure And relaying the message to the destination,
A message learning step of storing the message identification information for identifying the message of the field device and the message reliability value corresponding to the message identification information in the learning result database by learning the preceding messages received and processed by the system;
A header parsing step of parsing a header of a message received from the field device to obtain header information of the received message;
The system compares a message reliability value corresponding to the message identification information included in the header information among message reliability values stored in the learning result database with a predetermined reference value, and if the corresponding message reliability value is equal to or greater than the reference value, A step of determining the scope of the verification to decide to omit some or all procedures of the verification procedure; And
And a message transmission processing step of processing the received message and transmitting the processed message to a corresponding destination, with the system skipping some of the procedures or entire procedures of the validity verification procedure according to the determination. 13. The method of claim 12,
The message learning step comprises:
The system collects message identification information of the received message and metadata necessary for determining reliability of the received message from the header information obtained in the header parsing step;
Determining a message reliability value corresponding to the collected message identification information according to a result of the determination, determining whether the collected message identification information and the meta data match the predetermined message trust condition, and determining a message reliability value corresponding to the collected message identification information; And
And storing the determined message reliability value in the learning result database in response to the collected message identification information by the system. 14. The method of claim 13,
Wherein said collecting of learning data comprises collecting, as message identification information of said received message, said system, a field device ID identifying said field device and a message ID identifying a message type of said received message Wherein the machine learning based message relaying method comprises: 14. The method of claim 13,
Wherein the learning data collection step includes collecting message generation time data and message size data as metadata necessary for the system to determine the reliability of the received message. 14. The method of claim 13,
Wherein the message reliability determination step determines that the system has the same message identification information as the message trust condition, and that the system has the same message identification information, the presence or absence of the received preceding message, the same message identification information, the success or failure of the preceding message received immediately before, And determining a message reliability value corresponding to the collected message identification information, in consideration of whether or not the received message identification information matches the message size. 14. The method of claim 13,
Wherein the message reliability determination step determines that the system increases the existing message reliability value corresponding to the collected message identification information to a certain extent when the collected message identification information and meta data match the message trust condition, And if the message identification information and the meta data do not match the message trust condition, the system reduces the existing message reliability value corresponding to the collected message identification information by a certain amount Message relaying method. 13. The method of claim 12,
Wherein the verification range determination step includes a step of determining whether or not the message reliability value corresponding to the message identification information included in the header information is equal to or greater than a first reference value and less than a second reference value, And determining that the system omits the entire procedure of the validity verification procedure if the message reliability value is equal to or greater than the second reference value. 19. The method of claim 18,
The method may further comprise: before the message transmission processing step after the verification range determination step, the system performs the validity verification procedure on the received message, wherein, in accordance with the determination in the verification range determination step, Further comprising a message verification step of omitting the header verification procedure of the received message or omitting the entire procedure of the validity verification procedure. 20. The method of claim 19,
Wherein the validation procedure performed in the message validation step comprises:
A header verification step of the system verifying validity by comparing header information of the received message with predetermined header verification reference information;
A body decryption procedure in which the system decrypts the body when the body of the received message is encrypted;
A body parsing procedure in which the system parses the body or decoded body of the received message to obtain body information of the received message; And
And a body verification procedure in which the system compares the body information with predetermined body verification reference information to perform validity verification. 13. The method of claim 12,
Wherein the message transmission processing step comprises:
When the entire procedure of the validity verification procedure is omitted or when the received message passes the procedure other than the procedure in which the omission is determined in the validity verification procedure, A routing path setting step of setting a routing path of the received message; And
Wherein the system comprises a message reconstruction step of reconstructing the format of the received message if the format of the received message is different from the format used at the destination. 22. A computer program for executing a method according to any one of claims 12 to 21 via a computer system, the computer program being recorded on a recording medium.

Images