KR101039250B1 - Apparatus and method of distributed processing of RFID middleware - Google Patents

Abstract

The present invention relates to a large-capacity distributed processing apparatus and method of RFID middleware, and displays and manages contents corresponding to events submitted through the manager and the user device unit, which are interface means for receiving a specific event from the outside, and the manager and the user device unit. A load information processor for selecting an optimal distributed middleware server according to a scheduling of the web application processor and the web application processor, and a server process manager that manages each distributed middleware sub in the cluster in association with the manager and the user device; In addition, it is possible to continuously cope with system scalability, infrastructure expansion, and system safety maintenance, thereby having a significant economic effect.

RFID, large capacity

Description

Apparatus and method of distributed processing of RFID middleware}

The present invention relates to a large capacity distributed processing apparatus and method for RFID middleware, and to a large capacity distributed processing apparatus and method for RFID middleware using a separate apparatus for managing servers by connecting RFID servers in parallel to process large data. will be.

RFID (Radio Frequency Identification) technology stores the product identification information in a very small RFID tag and transmits the goods and the surrounding environment information to the network through an antenna and a reader at a radio frequency and processes them. Technology. In addition, RFID technology is a promising technology that will lead the future IT (Information Technology) market from the conventional people-oriented informatization to goods-oriented. In addition, RFID technology is a next-generation core technology that can be applied to various fields such as information communication, logistics, supply chain, transportation, environment, and security as well as generality that can be easily applied to various fields by changing frequency bands. .

RFID technology is regarded as a key technology for realizing ubiquitous computing. To make it easy for application service developers to build various ubiquitous application services using RFID technology, middleware that acts as a bridge between RFID tagged objects and application services is required. RFID middleware manages various sensors to collect data from the sensors using various protocols, and extracts meaningful or easy-to-use information from the collected and processed data and delivers them to application services. Is basically done.

As the ubiquitous computing environment is gradually expanded and the range of utilization of RFID and USN technology is widened, research and development are proceeding on how to process a large amount of data while minimizing bottlenecks and delay time in RFID systems. Large RFID systems, such as those used in large warehouses, regularly monitor a large number of tags by multiple readers, and wirelessly form an ad hoc network to deliver information to the RFID database.

In such a system, there are load balancing between RFID readers as well as collision between tags and tags. When a large number of readers perform a recognition task in a large RFID system, reader recognition ranges may overlap in some areas. Recent studies have shown that load balancing between readers in this situation can reduce tag recognition latency and reduce reader energy consumption.

In addition, research and development are being conducted on load balancing methods between RFID middlewares to solve the problem of load concentration among RFID middlewares, which plays a key role in RFID middleware systems, as well as load balancing between readers in large RFID systems. In middleware RFID systems, RFID middleware plays a key role in enabling data exchange between reader groups and higher application layers. The load problem between RFID middlewares is to distribute the load evenly by appropriately distributing the resources according to the state of the resource and the load state of each node in order to efficiently use distributed resources and obtain high performance. If the load concentration on the RFID middleware persists, processing efficiency will be lowered and business delays and interruptions due to processing delays will occur. In order to prevent the load from being concentrated on a specific node, it is necessary to introduce a load balancing technique.

Therefore, the present invention is to solve the above problems, a large-capacity distributed processing apparatus of RFID middleware using a separate device for defining RFID as a cluster by connecting the RFID server in parallel, and managing the RFID middleware servers in the cluster and To provide a way.

Another object of the present invention is to provide an apparatus and method for large-capacity distributed processing of RFID middleware, which enables load information, process, and resource management in RFID middleware.

In order to achieve the above object, a large-capacity distributed processing apparatus for RFID middleware according to the present invention displays an administrator and a user device unit, which is an interface means for receiving a specific event from the outside, and contents corresponding to an event submitted through the manager and the user device unit. A load information processing unit for selecting an optimal distributed middleware server according to a scheduling of the web application processing unit and the web application processing unit, and a server process management unit managing each distributed middleware sub in the cluster in cooperation with the manager and the user device unit. Include.

The load information processor is configured to receive and store current load information from the distributed middleware server. A load information manager and a load information manager are used to analyze and manage a request specification transmitted from the web application processor. And a dispatcher control unit configured to change the algorithm according to the information, and determine a schedule for transmitting the request specification to the distributed middleware server.

The server process manager may include a middleware internal process processor that manages a process and a thread setting according to a request specification received from the dispatcher controller, and a user command processor that interprets a predefined command to be issued to the distributed middleware server. And a middleware server resource management unit for updating the load information inside the middleware server changed by the processing unit and the middleware internal process processing unit and managing resources of the middleware server.

The event is characterized in that it includes a data collection cycle, purification conditions, load information of the distributed middleware server.

The web application processor may control a scheduling and include a specification backup manager to store all processed results in a database.

The analysis of the load information management unit may include contents of load information to which contents and weights of data processed in the middleware are applied.

The weight processor may set weights using a collection time, a collection period, a response time, the number of readers, and the number of processes.

The one cluster is characterized in that it comprises a plurality of middleware servers.

The setting of the threads is characterized in that it includes starting or terminating or stopping or restarting the thread.

Setting of the process and the thread is characterized in that it is performed automatically.

The setting of the process and the thread may be performed according to an instruction from the outside.

The mass distribution processing method of the RFID middleware of the present invention for achieving the above object comprises the steps of submitting a web application request specification, updating the load information of the servers in the cluster, analyzing the request specification and weight information Determining whether there is no weight information and determining whether there is no weight information, selecting an optimal request specification based on the message, and transmitting it through load information of a distributed middleware server and transmitting the result In the step of delivering the information to the requesting user, analyzing and managing load information, and determining whether there is a separate command for managing the distributed middleware server, and when there is no separate command as a result of the determination The method may include transmitting a result value according to the request to the outside.

And when the weighting information is present as a result of the determination, changing the algorithm according to the weight and recalculating the load information according to the changed algorithm.

If a separate command exists as a result of the determination, analyzing the command, executing the command of setting the log information and the process through the analyzed command, and updating and updating the current status information of the log and the process. Transmitting the received information.

The setting of the process is characterized in that it includes starting or ending or stopping or restarting the process.

The large-capacity distributed processing apparatus and method of the RFID middleware of the present invention can continuously cope with system scalability, infrastructure expansion, and system safety maintenance, and thus, economically significant effects occur.

Hereinafter, with reference to the accompanying drawings will be described in detail a large-capacity distributed processing apparatus and method of the RFID middleware of the present invention.

1 is a view showing the configuration of a large-capacity distributed processing apparatus of the RFID middleware of the present invention. As shown, the large-capacity distributed processing apparatus of the RFID middleware of the present invention includes a manager and user device unit 100, a web application processing unit 110, a load information processing unit 120, a distributed middleware server 130, and a server. It includes a process manager 140.

The manager and user device unit 100 is an interface means for the administrator and the user to make a specific request. The administrator and the user monitor the matters related to data collection such as data collection cycle and purification conditions in the middleware through the manager and the user device unit 100, load information of distributed middleware servers, and processes and threads of individual middleware servers. Allows you to enter various arguments in the form of a predefined specification for management.

The web application processing unit 110 displays and manages content corresponding to a predefined specification submitted through the administrator and the user device unit 100. The web application processing unit 110 mainly includes a specification backup manager which controls scheduling in submission and output of request and response specifications, and stores all processed results in a database.

The load information processor 120 analyzes the request specification transmitted according to the scheduling of the web application processor 110 and selects an optimal middleware server through load balancing of the connected distributed middleware server.

FIG. 2 is a detailed view of the load information processor of FIG. 1. As shown, the load information processor includes a distributed middleware server load manager 200, a load information manager 210, a weight processor 220, and a dispatcher controller 230.

The distributed middleware server load manager 200 receives load information of the current server from the distributed middleware server 130 connected to the load information manager 120 to generate and store a table. This changes the algorithm according to the request specification analysis result transmitted from the web application processing unit 110, so that load information of the servers is dynamically updated. However, there are various methods for changing the load information, but not limited to the above methods.

The load information manager 210 analyzes and manages a request specification transmitted from the web application processor 110. The analysis results can be summarized into two broad categories. The first is about the data that needs to be processed inside the middleware. This is the same content as the ECSpec (Event Cycle Specification) referred to in the existing ALE-based middleware. The second part is about weighted load information. It is the content of load information for recalculating and updating the load information table of the server generated by the distributed middleware server load management unit 200 based on the criteria presented in the request specification. For example, the distributed middleware server load management unit 200 thinks that the load information of the middleware server has been calculated according to the waiting process. The middleware server A has a collection time of 10 minutes and one process having 10 readers, Assuming that B middleware server is running 2 processes each with 5 readers for 3 minutes of collection time, load information is changed according to collection time, number of readers to collect, number of processes, etc. Change. It is about what weight to select the load information.

The weight processor 220 changes the algorithm according to the weight information analyzed by the load information manager 210 and recalculates the load information based on the changed algorithm. The algorithm is selected based on various weights such as collection time, collection period, response time, number of readers, and number of processes. Based on the selected algorithm, the load information of the distributed middleware server 130 managed by the distributed middleware server load management unit 200 is renewed and stored in a table.

The dispatcher controller 230 determines a schedule for transmitting the request specification to the distributed middleware server. The dispatcher controller 210 creates a message containing load information of the re-updated distributed middleware server including the request specification and determines an optimal distributed middleware server based on the written message.

The distributed middleware server 130 is an ALE based RFID middleware. Various types of code processing are possible, including EPC codes. The distributed middleware server 130 forms a cluster of individual RFID middleware servers. The load information processor 120 manages load information of the cluster unit.

The server process manager 140 manages each middleware server 130 in the cluster in cooperation with the manager and the user device.

3 is a diagram illustrating in detail a server processor manager of FIG. 1. As illustrated, the server process manager includes a middleware internal process processor 300, a user command processor 310, and a middleware server resource manager 320.

The middleware internal process processor 300 manages a process and a thread setting according to a request specification received from the dispatcher controller 210. The setting of the thread means the start / end / stop / restart of the thread. This can be done according to a user's separate command or can be done automatically. In ALE-based RFID middleware, clients receive ECReports as a result of various processing methods such as Define, which registers ECSpec specifying processing conditions in middleware, Subscribe, Pool, and Immediately request processing. For example, the Subscribe command that wants to request a result every time the processing specified in the ECSpec registered in the middleware needs to be unsubscribed must release a separate release command to stop the process.

The user command processor 310 interprets a predefined command for mass processing including a command that can be given to middleware in the ALE specification. When an administrator and a user issue a command for starting, terminating, stopping, or restarting a process and a thread or an application of an individual distributed middleware server in a cluster, the user command processor 310 interprets and executes the command. Update load information and transmit the result.

The middleware server resource manager 320 renews the load information inside the middleware server changed by the user command processor 310 and the middleware internal process processor 300 and manages resources of the middleware server.

4 is a flowchart illustrating a large-capacity distributed processing method of the RFID middleware of the present invention. As shown, the mass distribution processing method of the RFID middleware of the present invention submits a web application request specification. (S100) The manager and the user device unit 100 submit the web application request specification to the web application processing unit 110. do. The request specification submitted above is transmitted to the load information processing unit 120.

The load information of the servers in the cluster is updated (S105). The load information of the servers in the classer is updated through the distributed middleware server load management unit 200 of the load information processing unit 120.

The request specification is analyzed through the load information management unit 210 of the load information processing unit 120 (S110).

On the basis of the analyzed result, it is determined whether there is additional weight information as a result of the load information analysis (S115).

If there is weight information after determining whether the separate weight information exists, the algorithm according to the weight is changed (S120). The algorithm according to the weight is changed through the weight processor 220 of the load information processor 120.

The load information is recalculated using the changed algorithm. (S125)

The load information of the distributed middleware server is updated through the recalculated load information (S130).

A dispatcher message is created to execute a command in the dispatcher (S135). Based on the updated distributed middleware server information, the dispatcher controller 230 creates a dispatcher message so that the dispatcher can execute the command.

The most appropriate request specification is selected based on the contents of the written message, and the effective transmission is performed through load information of the distributed middleware server.

The transmission result information is transmitted to the requesting user (S145).

The middleware server receiving the transferred load information analyzes and manages the corresponding log information (S150).

It is determined whether there is a separate command for managing an individual distributed middleware server. (S155) The server process manager 140 separates the individual distributed middleware server from an external manager and user device unit 100 except for the received load information. It is determined whether there is a separate command for management of the 130.

If there is a user command after the determination, the user command is analyzed (S160). The user command processor 310 of the server process manager 140 analyzes the user command.

Through the analyzed command, the log information in the middleware server and a command such as start / end / stop / restart of the process are executed (S165).

Based on the executed command, the log in the middleware and the current state information of the process are updated (S170).

The updated information is transmitted. (S175) The middleware server resource management unit 320 of the server process management unit 140 updates the middleware server resource management information based on the updated contents, and transmits the information to the load information processing unit 120. Send to distributed middleware server load management unit.

The distributed middleware server 130 transmits a result value according to the request to the manager and the user (S180).

On the other hand, while shown and described with respect to specific preferred embodiments of the present invention, various modifications and changes of the present invention without departing from the spirit or field of the invention provided by the claims below It will be readily apparent to one of ordinary skill in the art that changes may be made.

1 is a view showing the configuration of a large-capacity distributed processing apparatus of the RFID middleware of the present invention.

FIG. 2 is a detailed view of the load information processor of FIG. 1. FIG.

3 is a diagram illustrating in detail a server processor manager of FIG. 1;

Figure 4 is a flow chart showing a large capacity distributed processing method of the RFID middleware of the present invention.

Claims (15)

delete An administrator and a user device unit which is an interface means for receiving a specific event from the outside; A web application processing unit for displaying and managing contents corresponding to the event submitted through the manager and the user device unit; A middleware server load manager for receiving and storing current load information from a distributed middleware server, a load information manager for analyzing and managing a request specification transmitted from the web application processor, and an algorithm according to weight information analyzed by the load information manager Selects an optimal distributed middleware server according to the scheduling of the web application processor, including a weight processor and a dispatcher controller configured to determine a schedule and transmit a request specification to the distributed middleware server according to the changed algorithm. A load information processor; And And a server process manager configured to manage each distributed middleware sub in the cluster in association with the manager and the user device. The method of claim 2, wherein the server process management unit; A middleware internal process processor configured to manage settings of processes and threads according to the request specification received from the dispatcher controller; A user command processor for interpreting a predefined command to be issued to the distributed middleware server; And a middleware server resource management unit for updating the load information in the middleware server changed by the user command processor and the middleware internal process processor and managing resources of the middleware server. The method of claim 3, The event is a large-capacity distributed processing device of the RFID middleware, characterized in that it comprises a data collection cycle, purification conditions, load information of the distributed middleware server. The method of claim 3, The web application processing unit is a large-capacity distributed processing device of the RFID middleware, characterized in that for controlling the scheduling and a specification backup management unit for storing all the processed results in a database. The method of claim 3, wherein the analysis of the load information management unit Large-capacity distributed processing apparatus for RFID middleware, comprising the content of data processed in the middleware and the content of load information to which weights are applied. The method of claim 3, wherein the weight processing unit A large-capacity distributed processing device for RFID middleware, characterized in that a weight is set using a collection time, a collection period, a response time, a number of readers, and a number of processes. The method of claim 3, The cluster is a large-capacity distributed processing device of the RFID middleware, characterized in that it comprises a plurality of middleware servers. The method of claim 3, Setting of the thread is a mass distribution processing device of the RFID middleware, characterized in that it comprises the start or end or stop or restart of the thread. The method of claim 3, Setting of the process and the thread is a large-capacity distributed processing device of the RFID middleware, characterized in that performed automatically. The method of claim 3, Setting of the process and the thread is a large-capacity distributed processing apparatus of the RFID middleware, characterized in that performed according to the command from the outside. Submitting a web application request statement; Updating load information of servers in the cluster; Analyzing the request specification; Determining the presence or absence of weight information; Creating a dispatcher message when there is no weight information as a result of the determination; Selecting an optimal request specification based on the message and transmitting the optimal request specification through load information of a distributed middleware server; Delivering the transmission result information to a requesting user; Analyzing and managing load information; Determining whether a separate command for managing the distributed middleware server exists; And And if there is no separate command as a result of the determination, transmitting a result value according to the request to the outside. The method of claim 12, In the case where the weight information exists as a result of the determination, Changing the algorithm according to the weight; And Recalculating the load information according to the changed algorithm; Mass distribution processing method of RFID middleware. The method of claim 13, If a separate command exists as a result of the determination, Analyzing the instructions; Executing a command of setting log information and a process through the analyzed command; Updating the log and current status information of the process; And Transmitting the updated information; Mass distribution processing method of RFID middleware comprising a. 15. The method of claim 14, Setting the process comprises starting or terminating or stopping or restarting the process.

Images