KR101044807B1 - Query processing method and system for sensor networks - Google Patents

Abstract

The present invention relates to a query processing method and system in a sensor network environment.

According to the present invention, when a user or an application requests a new query from the sensor network, it is possible to derive the result of the newly requested query by using the results of the existing queries executed in the sensor network. Can be reduced. In addition, when executing a merged query, instead of executing the queries individually, the queries are merged and executed as a single query, thereby reducing energy consumption in the sensor network than executing several queries at once.

Sensor Networks, Query Processing, Energy, Candidate Queries, Properties, Projection

Description

Query processing method and system for sensor networks

The present invention relates to a query processing method and system in a sensor network environment.

The present invention is derived from the research conducted as part of the IT growth engine technology development project of the Ministry of Knowledge Economy and the Ministry of Information and Communication Research and Development. [Task management number: 2006-S-022-03, Task name: USN middleware platform technology development].

The sensor network refers to a network composed of sensor nodes capable of sensing various phenomena in the periphery, and generally communicates with nodes using wireless communication. In a sensor network environment, because sensor nodes are powered by batteries, there is a need for a technology that can maximize the life of the sensor network by efficiently using limited energy.

The user or application requests a query from the sensor network to obtain data from the sensor network. In general, a query includes conditions for acquiring data, data or attributes to be acquired, and how often the query should be executed, and sensor nodes of a sensor network execute a query given to the user or an application. The requested data is sensed, processed and transmitted.

In order to execute a query, sensor nodes require a lot of energy, such as receiving and sensing a query from a user or an application program, processing the sensed value, and transmitting the result again. Therefore, before requesting a query from the sensor network, a technique is needed to optimize the queries to reduce the energy consumption of the sensor network.

Several techniques have been proposed to optimize and execute queries in the current sensor network environment. These conventional techniques attempt to create a virtual query by merging a newly requested query with a query previously executed in the sensor network when a user or an application requests a query from the sensor network. Conventional techniques have the advantage of not having to execute all queries requested by a user or an application program, but have the following disadvantages.

First, most of the queries executed in the sensor network describe the period in which the queries are executed, but conventional techniques are not efficient for processing the execution cycle. For example, if a user or application wants to get a sensing result every 10 seconds from the sensor network, the query should describe a 10 second execution cycle. Conventional techniques execute a merged query every cycle corresponding to the greatest common divisor of the execution cycle of the queries to be merged when merging queries having different execution cycles. As a result, there is a problem that the energy of the sensor network can be consumed unnecessarily because the query can be executed frequently.

In addition, queries requested by a user or an application have different conditions for acquiring different data. When query merging, the conditions of each query must also be merged. However, since the merged query generally has a broader query condition than the existing query, there is a problem in that the sensor network must transmit a result that is not desired by the existing query. This causes the sensor node to consume unnecessary communication energy, so there is a limit to the efficient use of the energy of the sensor node.

Accordingly, the present invention relates to a query processing method and system, and more particularly, to provide an energy efficient query processing method and system in a sensor network environment. In particular, the present invention provides a query processing method and system that efficiently uses energy by deriving a result of a newly requested query by a user or an application by utilizing the result of a query that has been executed in a sensor network environment.

A system for processing a query in a sensor network environment, which is one feature for achieving the technical problem of the present invention,

A query receiving unit for receiving and transmitting a newly requested query; Candidate queries are extracted to execute the newly requested query delivered from the query receiver, and whether the result of the requested query can be derived from the extracted candidate query or whether the requested query is merged with the query being executed. A query optimizer for determining; Information about the queries currently being executed, mapping information between queries projecting an arbitrary attribute, mapping information between parent queries producing a result of a specific attribute, and inducible query information that can derive the result of the newly requested query. A query storage unit for storing; A query result deriving unit for deriving a query result by combining query results stored in the query storage unit at every desired execution period of the newly requested query; And a query execution unit that executes queries determined to be impossible to derive from queries belonging to the candidate set in the query optimizer.

Method for processing a query in a sensor network environment that is another feature for achieving the technical problem of the present invention,

Receiving a request for processing a new query; Constructing a candidate query set from a plurality of queries previously stored for the new query; Checking whether a result of the new query can be generated from the configured candidate query set; And if it is determined that a result of the new query can be generated, deriving a result of the new query from the candidate query set.

According to the present invention, when a user or an application requests a new query to the sensor network, the energy consumed in the sensor network can be derived by using the results of the existing queries that are being executed in the sensor network. Can be reduced.

In addition, when executing a merged query, instead of executing the queries individually, the queries are merged and executed as a single query, thereby reducing energy consumption in the sensor network than executing several queries at once.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when a part is said to "include" a certain component, it means that it can further include other components, without excluding other components unless specifically stated otherwise. In addition, the terms “… unit”, “… unit”, “module”, etc. described in the specification mean a unit that processes at least one function or operation, which may be implemented by hardware or software or a combination of hardware and software. have.

Hereinafter, an energy efficient query processing system and a query processing method according to an embodiment of the present invention will be described with reference to the drawings. First, a query processing system will be described with reference to FIG. 1.

1 is an overall structure diagram of a query processing system in a sensor network environment according to an embodiment of the present invention.

As shown in FIG. 1, a query processing system according to an exemplary embodiment of the present invention includes a query receiver 100, a query optimizer 200, a query execution unit 300, a query storage unit, and a query result derivation unit 400. Include. The query storage unit may include a query storage unit 540 for each projection attribute, a query storage unit 530 that is being executed, a derived query storage unit 520, and a parent query storage unit 510.

The query receiving unit 100 receives a query requested by a user or an application program.

The query optimizer 200 optimizes the query received by the query receiver 100 to be executed efficiently, and includes a candidate query set constructor 210, a derivability checker 220, and a query merger 230. .

First, the candidate query set constructor 210 extracts queries having the possibility of producing a result of a newly requested query among the queries executed in the sensor network. Then construct a candidate query set and print it out. To this end, information about the query currently being executed in the sensor network and query information for each projection attribute are received from the query storage unit 540 and the query storage unit 530 for each of the execution attributes to be described below.

The derivability checker 220 checks whether queries belonging to a given candidate query set can actually derive the result of the newly requested query. If it is determined that the newly requested query can be derived from queries belonging to the candidate set, the newly requested query is registered in the derived query storage unit 520. On the other hand, the sets of candidate stores determined to derive the newly requested query result are registered in the parent query store 510. In this case, a method of determining whether a newly requested query can be derived from queries belonging to the candidate set will be described later.

In addition, if it is determined that the result of the newly requested query cannot be derived from the results of the queries belonging to the candidate set, it is advantageous to merge the newly requested query with the existing running query or to execute only the newly requested query. Determine if it is. In this case, whether the benefit is judged is predicted by the number of query results generated when the newly requested query is executed alone in the sensor network, and the number of query results generated when merging with the existing running query. Determine by checking whether the query result is generated.

Benefit determination is a known matter, detailed description thereof will be omitted in the embodiment of the present invention. If it is determined that merging with an existing running query is a benefit, then the newly requested query is sent to the query merger 230. On the other hand, if it is determined that it is beneficial to execute only the newly requested query, the newly requested query is sent to the query executing unit 300 so that the sensor network can be requested and executed.

The query merger 230 merges the newly requested query received by the inducibility checker 220 with the currently executing query stored in the existing executing query storage unit 530 to be executed.

Next, the query execution unit 300 requests and executes the sensor network from the queries that are determined to be impossible to derive from the queries belonging to the candidate set in the inducibility checker 220. The query is registered in the query storage unit 530 which is being executed, and is registered in the query storage unit 540 for each projection attribute by using the attribute information projected by the query.

The query result deriving unit 400 combines the results of the queries registered in the parent query storage unit 510 for each execution cycle of the query to derive the result of the query registered in the deriving query storage unit 520. Induce query results.

The query storage unit 530 for each projection property stores mapping information between a specific property and queries for projecting the property. For example, assuming that there are Q1 and Q2 as a query for projecting an attribute called temperature, mapping information in the form of (temperature, Q1, Q1) is stored in the query storage unit 530 for each projection attribute.

The parent query store 510 stores mapping information between a specific attribute of the inducible query and the parent queries that produce the result of this attribute. For example, if query Q3 is an inducible query, and queries Q4 and Q5 produce the result of the attribute of temperature of Q3, mapping information of the form ((Q3, temperature), Q4, A5) is stored.

The executing query storage unit 530 and the derived query storage unit 520 respectively store information of queries currently executing in the sensor network and information of queries capable of deriving a result from the executing query.

When a user or an application program requests a new query on such a system, a process of processing the query will be described with reference to FIG. 2.

2 is a flowchart illustrating a process of processing a query in a query processing system according to an exemplary embodiment of the present invention.

As shown in FIG. 2, the query receiving unit 100 receives a newly requested query from a user or an application program (S100). If the new query is accepted in step S100, the query optimizer 200 checks whether it is possible to derive the result of the newly requested query using the queries executed in the sensor network (S110). The query optimizer 200 determines whether it is possible to derive the result of the newly requested query from the results of the queries executed in the sensor network based on the check of the step S110 (S120).

Herein, a process of constructing a candidate query set having a possibility of deriving a result of a newly requested query for checking in step S110 will be described with reference to FIG. 3.

3 is a flowchart illustrating a process of constructing a candidate query set according to an embodiment of the present invention.

As shown in FIG. 3, the candidate query set constructor 210 first extracts queries that can send a query result at each execution cycle of a newly requested query. To this end, the execution cycle of each query registered in the executing query storage unit 530 and the execution cycle of the newly requested query are examined (S300). The candidate query set generator 210 determines whether the execution cycle of each query registered in the query storage unit 530 being executed is a divisor of the execution cycle of the newly requested query, based on the check in operation S300 (S310). .

As a result of the determination of S310, if the execution cycle of the specific query registered in the executing query storage unit 530 is a divisor of the newly requested execution cycle, the candidate query set constructor 210 determines the condition of the query and the newly requested query. The condition of whether there is an overlap even in part even between the conditions of the condition is checked (S320). For example, if the condition of the newly requested query is "Temperature> 10", and the specific existing query condition is "Temperature <20", the conditions of these two queries are determined to have overlapping portions (11 ≤ temperature ≤ 19). do. Accordingly, the candidate query set generator 210 determines whether the condition of the specific query that is being executed overlaps with the condition of the newly requested query based on the check of 303 (S330).

As a result of the determination in step S330, the query determined to have a part overlapping with the condition of the newly requested query is registered in the candidate query set which is a set of queries having the possibility of deriving the result of the newly requested query (S340). . Here, the candidate query set is stored in the candidate query set constructor 210. On the other hand, the determination result in step S310 is a query having an execution cycle that is not an abbreviation of the execution cycle of the newly requested query, or the query having no part overlapping the condition of the newly requested query as a result of determination in step S330 is a candidate query storage unit ( Not registered at 210).

Next, if the determination result of step S120 of FIG. 2 indicates that the result of the newly requested query can be derived from the results of the queries already executed in the sensor network, the query result deriving unit 400 returns the result of the newly requested query in the sensor network. Derived from the results of running queries (S130). However, if it is determined that derivation is not possible from the results of existing queries, the query optimizer 200 calculates the benefit of merging the newly requested query with the query running in the sensor network (S140). Through the calculation at step S140, the query optimizer 200 may calculate whether it is beneficial to merge and execute the newly requested query with the query being executed in the sensor network or whether it is advantageous to execute only the newly requested query.

The query optimizer 200 determines whether the query merging is profit based on the calculation result at step S140 (S150), and if it is determined that merging with a query that is already being executed is a benefit, the query optimizer 200 may merge the existing query. The query is stopped in the sensor network (S160). At the same time, the information of the suspended query is deleted from the list of the query storage unit 530, the query storage unit 540 for each projection attribute, and the parent query storage unit 510 that are being executed.

The query optimizer 200 merges the newly requested query with the existing query whose execution is stopped, and then registers the queries to derive the results from the merged query in the derived query storage unit 520 and the parent query storage unit 510. (S170). Then, the merged query is transferred to the query executing unit 300 so that the merged query can be executed in the sensor network. Then, the query execution unit 300 registers the information of the merged query in the query storage unit 530 and the projection storage unit 540 for each projection property while executing the merged query (S180).

On the other hand, if the determination result of step S150 determines that there is no benefit in merging the newly requested query with the query running in the sensor network, the query optimizer 200 transmits the newly requested query to the query execution unit 300. Even if the newly requested query is not directly projected, since the specific attribute value can be found through the condition, the query executing unit 300 can find the value from the condition, although the newly requested query does not project to handle such a case. Check for the existence of an attribute. If there is an attribute, the attribute and the newly requested query are registered in the query storage unit 540 for each projection attribute (S190).

For example, assuming that the condition of the newly requested query is "temperature = 10" and do not project the temperature attribute, the result of this query shows that the temperature is 10. Therefore, the query executing unit 300 transmits the newly requested query to be newly registered as a query projecting the temperature attribute to the projection storage unit 540 for each projection attribute.

Thereafter, the query executing unit 300 executes the newly requested query in the sensor network (S200). The query executing unit 300 registers the information of the newly requested query in the query storage unit 530 and the projection storage unit 540 for each projection attribute while executing the newly requested query.

Next, when the inducibility checker 220 receives the candidate query set from the candidate query set constructor 210 having the possibility of producing the result of the newly requested query through the procedure of FIG. 3 described above, the candidate query set. A process of checking whether the queries belonging to can actually derive the result of the newly requested query will be described with reference to FIG. 4.

4 is a flowchart illustrating a process of checking whether a result of a newly requested query can be derived according to an embodiment of the present invention.

As shown in FIG. 4, the inducibility checker 220 selects queries that project the attribute from the candidate query set for the attributes (denoted as 'attribute a' in FIG. 4) projected by the newly requested query. do. Then, after obtaining the union of the conditions of the selected queries, it is checked whether the obtained union includes the conditions of the newly requested query as a subset (S400).

Then, the inducibility checker 220 determines whether the queries belonging to the candidate query set include the conditions of the newly requested query, based on the check in step S400 (S410). At this time, if there is at least one attribute that does not satisfy the condition among the attributes projected by the newly requested query, it is determined in step S410 that the condition is not satisfied.

If it is determined in step S410 that the queries belonging to the candidate query set include the condition of the newly requested query, the queries of the candidate query set projecting the corresponding attribute for each attribute projected by the newly requested query are stored in the parent query. It registers to the unit 510 (S430). In addition, the derivability checker 220 projects the corresponding property among the queries belonging to the candidate query set for each property (marked as 'attribute b' in FIG. 4) included in the condition, although the newly requested query is not projected. After selecting the queries to perform, the union of the conditions of the selected queries is obtained.

The derivability checker 220 checks whether the union thus obtained includes a condition of a newly requested query as a subset (S440). Based on the check of 405, the inducibility checker 220 determines whether the queries belonging to the candidate query set include the condition of the newly requested query (S450). If the newly requested query is not projected but there is one attribute in the condition that does not satisfy the condition, 406 determines that the condition is not satisfied.

If it is determined in step S450 that the condition of the newly requested query is satisfied, the parent query storage unit queries the candidate query set for projecting the corresponding attribute for each attribute not newly projected but included in the condition. It registers to the 510 (S460). Thereafter, the derivability checker 220 determines that the result of the newly requested query can be derived using the results of the queries belonging to the query storage unit 530 being executed (S470). The newly requested query is registered in the derived query storage unit 520 (S480_).

As a result of the determination in steps S410 and S450, if it is determined that the queries belonging to the candidate query set cannot satisfy the conditions of the newly requested query, the inducibility checker 220 determines that the result of the newly requested query cannot be derived. (S420). However, if it is determined in step 406 that the condition is not satisfied, since there may be information registered in the parent query storage unit 510 through step S440, the corresponding information is deleted from the parent query storage unit 510.

Next, as mentioned in step S130 of FIG. 2, when the queries executed in the sensor network send the results of the query, the query result deriving unit 400 derives the results of the queries belonging to the deriving query storage unit 520. The process of making will be described with reference to FIG. 5.

5 is a flowchart illustrating a process of deriving the results of queries according to an embodiment of the present invention.

As illustrated in FIG. 5, the query result deriving unit 400 receives a query result for each execution cycle described in each query for the query executed in the sensor network (S500). When the query result deriving unit 400 receives the query result, the query result deriving unit 400 selects queries which should receive the result in the current execution cycle among the queries belonging to the deriving query storing unit 520 through the information in the parent query storing unit 510. (S510).

When the queries to receive the results in the current execution cycle are selected, the condition of the query to derive the results is applied to the query results in order to filter the query results sent by the queries running in the sensor network (S520). Thereafter, to filter the attributes projected by the query to derive the result, the filtered query result is projected in step S520 (S530). In order to aggregate the query results received from the various queries into one, the query results which have been filtered and projected are joined to derive the query results (S540).

In order to actually explain a query processing method according to an embodiment of the present invention using an SQL type query, five queries Q1, Q2, Q3, Q4 and Q5 are currently running in the sensor network, and the respective queries are Assume that it is equal to Equation 1.

Q1: SELECT x FROM sensors WHERE x <10 SAMPLE PERIOD 8s

Q2: SELECT y FROM sensors WHERE 0 <x <5 SAMPLE PERIOD 4s

Q3: SELECT x, y FROM sensors WHERE y <= 15 SAMPLE PERIOD 8s

Q4: SELECT x, y FROM sensors WHERE 5 <= x and y <= 5 SAMPLE PERIOD 2s

Q5: SELECT x FROM sensors WHERE 7 <x SAMPLE PERIOD 3s

Such queries are stored in the executing query storage unit 530. In the query storage unit 540 for each projection attribute, projection attributes of these queries and mapping information of queries for projecting each attribute are stored. That is, the queries projecting the property x are Q1, Q3, Q4, Q5, and the queries projecting the property y are Q2, Q3, Q4, so (x, Q1, Q3, Q4, Q5), (y, Q2, Q3, Q4) is stored in the same form.

In this state, it is assumed that the user requests the query Qnew as shown in Equation 2 below.

Qnew: SELECT x, y FROM sensors WHERE 0 <x <10 and 0 <y <10 SAMPLE PERIOD 8s

When the query receiver 100 receives a new query as shown in Equation 2, the query optimizer 200 constructs a candidate query set through the candidate query set constructor 210. When constructing the candidate query set, the process described with reference to FIG. 3 is performed. Since the execution cycle 3s of Q5 is not a divisor of the Qnew execution cycle 8s, it is excluded from the candidate query set. Therefore, candidate query sets constructed through the procedure of FIG. 3 are Q1, Q2, Q3, and Q4.

The derivability checker 220 receives the candidate query set configured in the candidate query set constructor 210 and checks whether Q1, Q2, Q3, and Q4 can actually produce the result of Qnew. This test follows the procedure of FIG. 4 described above.

First, because the attributes projected by the new query Qnew are x and y, check whether the union of the conditions of the candidate queries projecting these attributes, as shown in 401 of FIG. 4 for these attributes, includes the conditions of the newly requested query. do. First, since the queries projecting the attribute x in the candidate query set are Q1, Q3, and Q4, it is checked whether the union of the conditions of these queries includes the condition of Qnew.

If we combine the conditions of Q1, Q3, and Q4, we include the conditions of Qnew as a subset, so the check for attribute x passes. The queries projecting the property y are Q2, Q3, and Q4. When the conditions of these queries are combined, the condition of Qnew is included as a subset, so the check on the property y is also passed. As a result, the union of the conditions of candidate queries that project these attributes for each attribute x, y includes the condition of Qnew.

Q1, Q3, and Q4 are registered in the parent query storage unit 510 as queries for generating a value for the property x of Qnew, and Q2, Q3, and Q4 are queries for generating a value for the property x of Qnew. Also, since Qnew has no attributes that appear only in conditions, the part that checks that the union of the conditions of candidate queries that project this attribute for each attribute included in Qnew does not project but contains the conditions of the newly requested query. Are satisfied. Finally, Qnew is determined to be an inducible query, and registers this query in the derived query storage 520.

The process of deriving the result of Qnew stored in the derivation query storage unit 520 follows the procedure of FIG. 5 described above. When the query result deriving unit 400 receives the query result from the sensor network, the query result deriving unit 400 selects a query to derive the result in the current execution cycle. In this case, it is assumed that Qnew has been selected.

Parent queries to retrieve Qnew's results are retrieved from the parent query storage unit 510. Thereafter, the query result deriving unit 400 applies Qnew's condition to the results of the parent queries for each property of Qnew among the received query results and filters the result. When the filtering result is generated, the result is projected for each attribute of Qnew. In the embodiment of the present invention, the result of projecting the attributes x and y is generated. Finally, the result of Qnew is generated by joining the result of projecting on property x with the result of projecting on y.

The embodiments of the present invention described above are not implemented only through the apparatus and the method, but may be implemented through a program for realizing a function corresponding to the configuration of the embodiment of the present invention or a recording medium on which the program is recorded. Implementation may be easily implemented by those skilled in the art from the description of the above-described embodiments.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements of those skilled in the art using the basic concepts of the present invention defined in the following claims are also provided. It belongs to the scope of rights.

1 is an overall structure diagram of a query processing system in a sensor network environment according to an embodiment of the present invention.

2 is a flowchart illustrating a process of processing a query in a query processing system according to an exemplary embodiment of the present invention.

3 is a flowchart illustrating a process of constructing a candidate query set according to an embodiment of the present invention.

4 is a flowchart illustrating a process of checking whether a result of a newly requested query can be derived according to an embodiment of the present invention.

5 is a flowchart illustrating a process of deriving the results of queries according to an embodiment of the present invention.

Claims (10)

In a system for processing a query in a sensor network environment, A query receiving unit for receiving and transmitting a newly requested query; Candidate queries are extracted to execute the newly requested query delivered from the query receiver, and whether the result of the requested query can be derived from the extracted candidate query or whether the requested query is merged with the query being executed. A query optimizer for determining; Information about the queries currently being executed, mapping information between queries projecting an arbitrary attribute, mapping information between parent queries producing a result of a specific attribute, and inducible query information that can derive the result of the newly requested query. A query storage unit for storing; A query result deriving unit for deriving a query result by combining query results stored in the query storage unit at every desired execution period of the newly requested query; And A query execution unit that executes queries determined to be impossible to derive from queries belonging to the candidate set by the query optimizer Query processing system comprising a. The method of claim 1, The query storage unit, A running query storage unit for storing information of queries currently being executed in the sensor network; A projection storage unit for each projection attribute that stores mapping information between arbitrary attributes and queries for projecting the arbitrary attributes; A parent query storage unit for storing mapping information between a specific attribute of an inducible query and a parent query producing a result of the specific attribute; And A derivation query storage unit for storing derivable query information capable of deriving a result of the requested query Query processing system comprising a. The method of claim 2, The query optimizer, A candidate query set constructer for extracting candidate queries from among queries executed in a sensor network that can produce a result of the requested query; Examines whether candidate queries extracted by the candidate query set constructor can derive the result of the requested query, whether the requested query can be derived from the candidate queries, and the requested query and the query being executed Induction probability checker for determining whether or not to merge; And A query merger for merging the requested query with a running query Query processing system comprising a. In the method for processing a query in a sensor network environment, Receiving a request for processing a new query; Constructing a candidate query set from a plurality of previously stored queries based on an execution cycle for the new query; Checking whether a result of the new query can be generated based on an execution cycle and query conditions of the queries belonging to the configured candidate query set; And If it is determined that a result of the new query can be generated, deriving a result of the new query from a candidate query set based on a result of the query generated at each execution cycle; Query processing method comprising a. 5. The method of claim 4, Comprising the candidate query set, Checking the execution cycle of the new query; Constructing a candidate query set of queries having a period corresponding to a divisor of a plurality of queries; Determining whether a condition for the configured candidate query set is partially similar to a condition of the new query; And If it is determined to be similar, registering the selected candidate query in the candidate query storage. Query processing method comprising a. 5. The method of claim 4, The checking step, Selecting a query projecting the first attribute from the candidate query set with respect to the first attribute projected by the new query to obtain a union of conditions of the selected queries; Checking whether the obtained union includes a condition of the new query as a subset; If the first attribute is not included in the union, determining the new query as a non-derivable query Query processing method comprising a. The method of claim 6, If the first attribute is included in the union, registering the queries of the candidate query set projecting the first attribute into a parent query set; Obtaining a union by selecting queries that project the second attribute from among queries belonging to a candidate query set, with respect to a second attribute not projected by the new query; Determining if the union includes the condition of the new query as a subset; If the union includes a condition of the new query, registering queries of the candidate query set projecting the second attribute with the parent query set; And Determining that the new query can be derived using the results of queries belonging to the currently executing query set, and storing the new query in the derived query storage unit. Query processing method comprising a. 5. The method of claim 4, The deriving step, Receiving a result of a query that occurs at each execution cycle of a currently executed query; If the new query belongs to a query whose results are to be derived in the execution cycle, retrieving parent queries in a parent query storage unit for deriving a result of the new query; Filtering and applying a condition of the new query to a result of the retrieved parent query among results of the received query; Projecting each property of the new query to a parent query belonging to a filtering result; Joining the projected result to derive the new query Query processing method comprising a. 5. The method of claim 4, If it is determined that the result of the new query cannot be generated, Calculating a benefit of merging the new query with the currently executing query; If it is determined that merging the new query with the currently running query is a benefit, stopping execution of the query to be merged; Deleting the query that has been suspended from the query storage unit, the query storage unit for each projection attribute, and the parent query storage unit; Merging the new query with the query that has stopped execution and registering the queries to derive the results from the merged query in a derived query storage; And Executing the merged query Query processing method comprising a. 10. The method of claim 9, If it is determined that it is beneficial to not merge the new query with the currently running query, checking that the new query is not projected but has an attribute from which a value can be derived; If there is an attribute, registering the attribute and a new query in a query storage for each projection attribute; And Executing the new query Query processing method comprising a.

Images