KR100952076B1 - Method and apparatus for processing aggregation in sensor networks - Google Patents

Abstract

PURPOSE: A method and an apparatus for processing aggregation in sensor networks are provided to evaluate a normalized distance ratio and the normalized distribution ratio of each aggregation of each sensor node. CONSTITUTION: A distribution ratio calculating unit(206) calculates a normalized distribution ratio by using the sensing data of each sensor node saved in a database. A sample group extracting unit(208) extracts a fixed number of sensor nodes to be used as a sample group for a normalized distance ratio and the normalized distribution ratio of each sensor node. A totalizing operating unit(210) receives the sensing data from the sample group extracted from the sample group extracting unit. The totalizing operating unit calculates a total result value.

Description

Method and apparatus for processing aggregation in sensor networks

Embodiments of the present invention relate to data sensing techniques and aggregate computation processing techniques in a sensor network environment.

The sensor network is a network configured to collect the information collected from various sensors wirelessly.It is an intelligent transportation system that monitors or controls the external environment, automatically controls the production process, remotely detects the patient's condition, and controls the environment inside the intelligent building. It is widely used for such purposes.

In order to know the result of the calculation operation such as the average value, maximum value, minimum value, median value, etc. of the sensing data collected by each sensor node on the sensor network, the base station requests the sensor data from all the sensor nodes. Propagation and the sensing data received from all sensor nodes should be analyzed to derive the result. In the sensor network, a query refers to a data transmission command that requires each sensor node to transmit data for a specific condition in order to derive a result requested by an end user.

In this way, all sensor nodes in the sensor network consume a considerable amount of energy in order to sense data and transmit it to the base station. Research is needed to obtain.

Embodiments of the present invention analyze the data sensed on the sensor network composed of sensor nodes and evaluate the normalized average distance ratio and normalized distribution ratio for each aggregate operation for each sensor node, while minimizing energy consumption of the sensor network. It is intended to provide an energy efficient sensing data aggregation method that can obtain an aggregate operation value with small error.

Aggregation arithmetic processing unit of the sensor network for solving the above problems, the database that stores the sensing data of each sensor node; An average distance ratio calculator configured to calculate a normalized average distance ratio of the query data of each sensor node by using the sensing data of each sensor node stored in the database; A distribution ratio calculator for calculating a normalized distribution ratio of sensing data of each sensor node stored in the database; A sample group extracting unit extracting a predetermined number of sensor nodes into a sample group by using the normalized average distance ratio and the normalized distribution ratio of each sensor node; And an aggregation operation unit receiving sensing data from the sample group extracted by the sample group extracting unit and calculating a result value for the query data.

On the other hand, in the aggregation calculation processing apparatus of the sensor network for solving the above problems, in the aggregation calculation processing apparatus, the aggregation calculation of each sensor node using the sensing data of each sensor node accumulated for a certain period in the past. Calculating a normalized average distance ratio to; Calculating, by the counting arithmetic processing unit, a normalized distribution ratio of sensing data of each sensor node accumulated in the past period; Extracting, by the aggregate arithmetic processing unit, a predetermined number of sensor nodes into a sample group by using the normalized average distance ratio and the normalized distribution ratio of each sensor node; And estimating a result of the aggregation operation using the extracted sample population in the aggregation operation processing apparatus.

Embodiments of the present invention minimize the energy consumption of the sensor network by propagating an aggregate query to only the sensor nodes with a small normalized average distance ratio and a small normalized distribution ratio for a specific aggregate operation and estimating the result value. Even if only part of is used, the result of the aggregation operation with small error can be estimated.

In addition, it can be usefully used to cope with failure by replacing the sensor with a similar normalized average distance ratio when a specific sensor node occurs.

Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. However, this is only an example and the present invention is not limited thereto.

In describing the present invention, when it is determined that the detailed description of the known technology related to the present invention may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted. In addition, terms to be described below are terms defined in consideration of functions in the present invention, which may vary according to the intention or custom of a user or an operator. Therefore, the definition should be made based on the contents throughout the specification.

The technical spirit of the present invention is determined by the claims, and the following embodiments are merely means for efficiently explaining the technical spirit of the present invention to those skilled in the art.

1 is a flowchart illustrating an aggregation operation processing method 100 according to an embodiment of the present invention. In the following description, an aggregation operation refers to an operation for obtaining an average value, a maximum value, a minimum value, and a median value for data sensed by each sensor node.

First, by using the sensing data collected from each sensor node constituting the sensor network for a certain period of time in the past, each sensor node for each sensor node from the deviation between the accumulated sensing data of each sensor node and the result value for each calculation operation calculated therefrom Compute 102 the normalized average distance ratio for the aggregation operation.

Next, the normalized distribution ratio of the sensing data of each sensor node is calculated (104). The normalized distribution ratio is proportional to the variance or standard deviation of the data sensed for a certain time in each sensor node, and the smaller the normalized distribution ratio value, the more constant the data value measured by the corresponding sensor node. This is a measure for determining how consistently the data is sensed for each sensor node.

Thereafter, a predetermined number of sensor nodes are selected as a sample group for an aggregation operation to be queried using the calculated normalized average distance ratio and normalized distribution ratio of each sensor node (106). The method for selecting a sample group among the sensor nodes is to calculate the priority of each sensor node by weighting the normalized average distance ratio and the normalized distribution ratio, and extracting k sensor nodes in the order of high priority. This can be used.

Finally, the sensing data is received from the selected sample population to estimate a result of the aggregate operation query data (108). In other words, an aggregate calculation value for the entire sensor network is estimated using the data sensed from the sample population.

Hereinafter, each step shown in FIG. 1 will be described in detail.

Normalized mean Ratio of distance  Calculation

In the exemplary embodiment of the present invention, the normalized average distance ratio means "a ratio of normalized average distance to data sensed by each sensor node for a predetermined period of time and an aggregate calculation result calculated therefrom." That is, the closer the value of the data sensed by each sensor node is to the actual calculated result of the calculation, the smaller the normalized average distance ratio of the sensor node is. The aggregation operation may be, for example, an operation for obtaining a maximum value, a minimum value, an average value, a median value, or the like for the sensing data value.

The data used in the normalized average distance ratio calculation is sensing data collected by each sensor node constituting the sensor network in the past, which is stored in a database. That is, sensing data of each sensor node stored in the database is used to calculate a normalized average distance ratio.

The normalized average distance ratio can be calculated as follows.

를 모든 센서 노드들의 j 번째 센싱 데이터의 집계 연산 결과 값이라 하면, x_kj의 상기 집계 연산 결과 값에 대한 정규화된 거리비(e(x_kj))는 다음의 수학식 1과 같다.First, it is assumed that m sensor nodes exist from node 1 to m in a sensor network, and n sensing data sensed from each sensor node are stored in the database. x _kj is the jth (1≤j≤n) sensing data of node k (1≤k≤m),

When the counted result value as the j-th sensing data from all sensor nodes, normalized length ratios (e (x _kj)) on the aggregation result of the x _kj is: Equation (1).

(

≠ 0)는 각 노드들의 j 번째 센싱 데이터와 모든 센서 노드들의 j 번째 센싱 데이터의 집계 연산 결과 값과의 차이 중 가장 큰 값을 의미한다. 만일

의 값이 0인 경우에는 e(x_kj)의 값을 0으로 한다. 상기 정규화된 거리비는 0에서 1 사이의 값을 가지며, 정규화된 거리비가 0에 가까울수록 집계 연산의 결과 값에 가까운 것을 나타낸다.At this time,

(

≠ 0) means the largest difference between the j-th sensing data of each node and the j-th sensing data of all sensor nodes. if

When the value of 0 is 0, the value of e (x _kj ) is set to 0. The normalized distance ratio has a value between 0 and 1, and the closer the normalized distance ratio is to 0, the closer to the result of the aggregation operation.

Using Equation 1, the normalized average distance ratio E (n _k ) of the k-th sensor node is calculated as in Equation 2 below.

The above equation represents the normalized average distance ratio by n sensing data of the k-th sensor node (1 ≦ k ≦ m). The normalized average distance ratio has a value between 0 and 1, and the closer it is to 0, the mean that the sensing data value of the corresponding sensor is closer to the result of the aggregation operation.

Normalized distribution ratio calculation

In an embodiment of the present invention, the normalized distribution ratio of each sensor node is a measure for determining how consistently the data is sensed for each sensor node. In other words, if the sensor nodes with the normalized distribution ratio of the sensing values are extracted as the sample population, the estimation result closer to the actual value may be obtained.

In the normalized distribution ratio calculation, a database storing sensing data collected from each sensor node constituting the sensor network is used. That is, n sensing data stored in a database is used for each of the m sensor nodes.

The normalized distribution ratio is calculated using one of the following equations (3) or (4).

Equation 3 is for calculating a normalized distribution ratio by using a standard deviation of an aggregate operation of sensing data.

In the above equation, D (n _k ) is the normalized distribution ratio of the k-th sensor node, and max (standard deviation of x _i ) means the largest value among the standard deviations of each of the sensor nodes.

Equation 4 is for calculating the normalized distribution ratio using the variance of the aggregate operation of the sensing data.

In the above equation, D (n _k ) is the normalized distribution ratio of the k-th sensor node, max (variance of x _i ) means the largest value of the distribution of each of the sensor nodes.

Sampling

In this step, the normalized average distance ratio and normalized distribution ratio of each sensor node are calculated using the above-described method, and the sample population is extracted using the calculated normalized average distance ratio and normalized distribution ratio.

The sampling of the sample population is given weights to the normalized average distance ratio and the normalized distribution ratio, and the priority of each sensor node is calculated therefrom, and the predetermined number of sample groups are extracted in the order of high priority. It can be configured to. In this case, the weight is expressed as ω and has a value between 0 and 1. For example, if the weight of the normalized average distance ratio is assumed to be ω, the weight of the normalized distribution ratio may be expressed as (1-ω).

The priority is calculated by the following equation (5).

Where R (n _k ) is the priority of the k-th sensor node, E (n _k ) is the normalized average distance ratio of the k-th sensor node, and D (n _k ) is the normalized distribution ratio of the k-th sensor node, ω Is the weight. The priority according to Equation 5 has a value between 0 and 1, and the closer to 0, the higher the priority. That is, prior to selecting the sensor nodes for the aggregation operation, the sensors close to the priority is selected first.

According to the priority calculated by Equation 5, a predetermined number or a predetermined ratio of sensor nodes are selected in consideration of the remaining power (energy) amount of the sensor nodes, and for the aggregation operation on the selected sensor nodes. Sensing can be performed. Since the selected sensor nodes are selected by calculating a distance ratio and a distribution ratio for each historical sensing data, a more accurate aggregation operation is performed with fewer sensor nodes than a method of arbitrarily selecting sensor nodes and performing an aggregation operation. can do.

On the other hand, the size of the sample group can be calculated using statistical theory in order to estimate the result of the more accurate aggregation operation, for example, it can be calculated by the following equation (6) to determine the size of the sample group in the normal distribution .

Where Z is the standard normal distribution probability variable, σ is the standard deviation of the population, α is the significance level, and d is the margin of error.

When the sample population is extracted as described above, a desired aggregation operation query is sent to the sensor nodes selected as the sample population, the data is sensed and collected from them, and the result value of the query is estimated from the collected sensing data. For example, if the aggregation operation query is to obtain an average value of the sensing values of all the sensor nodes, the average value of the values sensed from the sample population may be estimated as the average value of all the sensor nodes. Similarly, if the aggregation operation query is to obtain the maximum or minimum value of the sensing values of all the sensor nodes, the maximum or minimum value among the values sensed from the sample population may be estimated as the maximum or minimum value of the all sensor nodes.

Hereinafter, the above-described aggregation calculation processing method and the result value estimating method using the selected sensor node will be described as an example. In the following example, the aggregation operation is to find the average value for the temperature.

The following is an example of collecting current national temperature data from the Korea Meteorological Agency website on July 20, 2009 and using it as sensing data. It is assumed that 83 locations across the country where temperature is provided by the Korea Meteorological Administration as sensor nodes, and 3 hour intervals are used as sensing values for each sensor node. The following example is for the sake of clarity, and the present invention can be applied in an actual sensor network environment, and can be used for general aggregation operations.

First, temperature data is collected from 83 sensor nodes from node 1 to node 83 at three-hour intervals and stored in a database. Calculate the ratio. Each item of Table 1 is a normalized distance ratio of sensing values of the corresponding number of sensing, and the rightmost column is a normalized average distance ratio of 10 temperatures. Table 1 is arranged in ascending order of the normalized average distance ratio.

Sensor node 1 time Episode 2 3rd time 4 times 5 times 6 times 7 times 8 times 9 times 10 times Normalized Average Distance Ratio Node78 0.04 0.03 0 0.01 0.15 0.03 0.09 0.15 0.27 0.3 0.107 Node 81 0.26 0.26 0.15 0.07 0.02 0.09 0.04 0.02 0.06 0.11 0.108 Node 34 0.16 0.16 0.04 0.07 0.05 0.2 0.02 0.22 0.27 0.05 0.124 Node 33 0.12 0.16 0.02 0.12 0.3 0.16 0.02 0.04 0.05 0.27 0.126 Node 36 0.23 0.32 0.12 0.03 0.06 0.02 0.12 0.02 0.05 0.36 0.133 Node 39 0.24 0.1 0.09 0.14 0.11 0.37 0.25 0.02 0.06 0 0.138 Node 13 0.31 0.42 0.19 0.15 0.04 0.05 0.03 0.13 0 0.07 0.139 Node 29 0.05 0.01 0.05 0.12 0.25 0.18 0.1 0.3 0.21 0.16 0.143 Node 24 0.03 0.17 0.15 0.14 0.22 0.02 0.11 0.06 0.35 0.34 0.159 Node 40 0.05 0.1 0.06 0.28 0.36 0.26 0.04 0.11 0.13 0.22 0.161 Node 11 0.14 0.11 0.08 0.23 0.21 0.05 0.21 0.22 0.24 0.16 0.165 Node 25 0.14 0.11 0.07 0.23 0.31 0.14 0.16 0.24 0.16 0.11 0.167 Node 32 0.07 0.1 0.07 0.19 0.31 0.43 0.31 0.18 0.11 0 0.177 Node 31 0.03 0.17 0.26 0.14 0.08 0.06 0.13 0.35 0.33 0.23 0.178 Node 64 0.4 0.42 0.04 0.18 0.21 0.2 0.13 0.11 0.05 0.06 0.18 Node 42 0.3 0.25 0.2 0.03 0.26 0.19 0 0.28 0.21 0.11 0.183 Node 44 0.31 0.32 0.17 0.03 0.46 0.12 0.03 0.11 0.16 0.12 0.183 Node 14 0.27 0.4 0.29 0.01 0.12 0.04 0.18 0.28 0.16 0.11 0.186 Node 41 0.33 0.32 0.18 0.14 0.01 0.17 0.18 0.15 0.14 0.25 0.187 Node 27 0.07 0.08 0.14 0.3 0.43 0.34 0.29 0.15 0.03 0.05 0.188 Node 15 0.15 0.15 0.1 0.19 0.17 0.11 0.26 0.31 0.29 0.19 0.192 Node57 0.54 0.5 0.22 0 0.01 0.29 0.38 0.48 0.25 0.14 0.281 Node 52 0.56 0.52 0.29 0.09 0.08 0.04 0.33 0.45 0.32 0.2 0.288 Node 23 0.16 0.3 0.43 0.37 0.47 0.58 0.33 0.15 0.11 0.01 0.291 Node 49 0.31 0.26 0.12 0.03 0.12 0.39 0.34 0.5 0.48 0.37 0.292 …

Table 2 below calculates the normalized distribution ratio for the average for each sensor node, and is calculated using Equation 4. That is, the normalized distribution ratio is calculated using the variance. Table 2 is arranged in ascending order of the normalized distribution ratio.

Sensor node Normalized distribution ratio Node 81 0.117 Node78 0.117 Node 34 0.129 Node 33 0.141 Node 36 0.142 Node 13 0.147 Node 39 0.148 Node 29 0.156 Node 24 0.179 Node 11 0.181 Node 40 0.182 Node 25 0.184 Node 64 0.189 Node 32 0.19 Node 42 0.192 Node 14 0.193 Node 31 0.193 Node 44 0.198 Node 41 0.199 Node 15 0.207 Node 27 0.207 Node 38 0.208 Node 50 0.372 Node 5 0.378 … …

Then, using the normalized average distance ratio and the normalized distribution ratio calculated in Table 1 and Table 2, the sample population is extracted by Equation 5 as follows.

When the weight is 0.5 in Equation 5, the priority of each sensor node is calculated as shown in Table 3 below. Table 3 is arranged in ascending order of priority.

Sensor node Normalized Average Distance Ratio Normalized distribution ratio Priority Node78 0.107 0.117 0.112 Node 81 0.108 0.117 0.113 Node 34 0.124 0.129 0.127 Node 33 0.126 0.141 0.134 Node 36 0.133 0.142 0.138 Node 39 0.138 0.148 0.143 Node 13 0.139 0.147 0.143 Node 29 0.143 0.156 0.150 Node 24 0.159 0.179 0.169 Node 40 0.161 0.182 0.172 Node 11 0.165 0.181 0.174 Node 25 0.167 0.184 0.176 Node 32 0.177 0.190 0.184 Node 64 0.180 0.189 0.185 Node 31 0.178 0.193 0.186 Node 42 0.183 0.192 0.188 Node 14 0.186 0.193 0.190 Node 44 0.183 0.198 0.191 Node 41 0.187 0.199 0.193 Node 27 0.188 0.207 0.198 Node 15 0.192 0.207 0.200 ... ...

In Table 3, the sample population from which a certain number (10 in this embodiment) is extracted in the order of high priority is as follows.

Sample population Nodes 78 Nodes 81 Nodes 34 Nodes 33 Nodes 36 Nodes 39 Nodes 13 Nodes 29 Nodes 24 Nodes 40

After sensing the extracted sensor nodes of the sample group, an aggregation operation may be performed using the collected sensing data, and the result of the aggregation operation for all the sensor nodes may be estimated therefrom. For example, if the aggregation operation is to obtain the maximum value, the maximum value of the sensing values of the sensor nodes of the sample group may be selected. If the minimum value is to be obtained, the minimum value of the sensing value may be selected. . In addition, if the average value is obtained, the average of the sensing values of the sensor nodes of the sample group may be calculated and estimated as the result of the aggregation operation.

On the other hand, embodiments of the present invention may include a computer readable recording medium including a program for performing the methods described herein on a computer. The computer-readable recording medium may include program instructions, local data files, local data structures, etc. alone or in combination. The media may be those specially designed and constructed for the purposes of the present invention, or they may be of the kind well-known and available to those having skill in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical recording media such as CD-ROMs, DVDs, magnetic-optical media such as floppy disks, and ROM, RAM, flash memory, and the like. Hardware devices specifically configured to store and execute program instructions are included. Examples of program instructions may include high-level language code that can be executed by a computer using an interpreter as well as machine code such as produced by a compiler.

2 is a diagram illustrating an aggregation operation processing apparatus 200 according to an embodiment of the present invention.

As shown, the aggregation calculation processing apparatus 200 according to an embodiment of the present invention, the database 202, the average distance ratio calculation unit 204, the distribution ratio calculation unit 206, the sample group extraction unit 208 ) And an aggregate calculation unit 210.

The database 202 receives and stores sensing data from each sensor node (not shown) connected to the aggregation operation processing apparatus 200.

The average distance ratio calculation unit 204 uses the deviation of the sensing data of each sensor node accumulated in the past time period stored in the database 202 and the aggregate calculation result calculated therefrom to perform each of the sensor node calculations. Calculate the normalized mean distance ratio for. In this case, the result of the aggregation operation is an aggregation operation value such as an average value, a maximum value, a minimum value, and a median value of the sensing data of each sensor node. The normalized average distance ratio calculation may be performed by Equations 1 and 2 described above.

The distribution ratio calculator 206 calculates a normalized distribution ratio using a standard deviation or variance of sensing data of each sensor node accumulated in the past period of time stored in the database 202. The normalized distribution ratio may be calculated by Equation 3 or 4 described above.

The sampling group extracting unit 208 calculates an aggregate using the normalized average distance ratio of each sensor node calculated by the average distance ratio calculating unit 204 and the normalized distribution ratio calculated by the distribution ratio calculating unit 206. Extract a predetermined number of sensor nodes for a sample population.

Extraction of the sample group may be configured to calculate the priority of each sensor node according to Equation 5, and extract a predetermined number of sensor nodes into the sample group in order of high priority.

Finally, the aggregate calculating unit 210 receives sensing data from the sample group extracted by the sample group extracting unit 208 and estimates a result value of the query data.

Although the present invention has been described in detail with reference to exemplary embodiments above, those skilled in the art to which the present invention pertains can make various modifications to the above-described embodiments without departing from the scope of the present invention. Will understand.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the claims below and equivalents thereof.

1 is a flowchart illustrating an aggregation operation processing method 100 according to an embodiment of the present invention.

2 is a diagram illustrating an aggregation operation processing apparatus 200 according to an embodiment of the present invention.

<Description of Major Symbols in Drawing>

200: aggregate operation processing unit 202: database

204: average distance ratio calculation unit 206: distribution ratio calculation unit

208: sample group extraction unit

Claims (19)

In the aggregation operation processing device, calculating the aggregation calculation result value by using the sensing data of each sensor node accumulated for a certain period, and calculating the normalized average distance ratio of the aggregation calculation result value for each sensor node; Calculating, by the aggregation calculation processing device, a normalized distribution ratio of sensing data of each sensor node accumulated over a period of time; And Extracting, by the aggregate arithmetic processing unit, a predetermined number of sensor nodes into a sample group by using the normalized average distance ratio and the normalized distribution ratio of each sensor node; Aggregate operation processing method comprising a. The method of claim 1, And said aggregation operation is an operation relating to any one of an average value, a maximum value, a minimum value, and a median value of said sensing data. The method of claim 1, The normalized average distance ratio with respect to the aggregate calculation result value of each sensor node is represented by the following equation.

(In this case, E (n _k) is a k-th sensor node (1 ≤ k ≤ the average distance ratio normalization m), e (x _kj) is x _kj of the aggregation result of the normalized distance for value ratio, x _kj Is j-th sensing data of k-th sensor node, n is the number of sensing data) Calculated by the method. The method of claim 3, The normalized distance ratio for the aggregation operation is represented by the following equation,

(At this time,

Is the average of the j th sensing data of each node,

Is the largest difference between the j th sensing data of each node and the result of the aggregation operation of the j th sensing data of each node) Calculated by the method. The method of claim 4, wherein

If the value is zero, the e (xkj) value is set to zero. The method of claim 1, The normalized distribution ratio of each sensor node is represented by the following equation

Where D (n _k ) is the normalized distribution ratio of the kth sensor node, and max (standard deviation of x _i ) is the largest of the standard deviations of each of the sensor nodes. Calculated by the method. The method of claim 1, The normalized distribution ratio of each sensor node is represented by the following equation

Where D (n _k ) is the normalized distribution ratio of the kth sensor node, and max (variance of x _i ) is the largest of each variance of the sensor nodes. Calculated by the method. The method of claim 1, And the sampling group extracting step is configured to calculate a priority of each sensor node and extract a predetermined number of sensor nodes into a sample group in order of high priority. The method of claim 8, The priority is the following equation

Where R (n _k ) is the priority of the kth sensor node, E (n _k ) is the normalized average distance ratio of the kth sensor node, D (n _k ) is the normalized distribution ratio of the kth sensor node, ω is the weight 0≤ω≤1) Calculated by the method. A computer-readable recording medium having recorded thereon a program for performing the method according to any one of claims 1 to 9 on a computer. A database in which sensing data of each sensor node is stored; An average distance ratio calculator configured to calculate a normalized average distance ratio with respect to an aggregate calculation result value of each sensor node using sensing data of each sensor node stored in the database; A distribution ratio calculator for calculating a normalized distribution ratio of sensing data of each sensor node stored in the database; A sample group extracting unit extracting a predetermined number of sensor nodes into a sample group by using the normalized average distance ratio and the normalized distribution ratio of each sensor node; And An aggregation operation unit which receives sensing data from the sample group extracted by the sample group extraction unit and calculates a result value for the aggregation operation; Aggregate operation processing apparatus comprising a. The method of claim 11, And the aggregation operation is an operation relating to any one of an average value, a maximum value, a minimum value, and a median value of the sensing data. The method of claim 11, The normalized average distance ratio for the aggregate operation of each sensor node is represented by the following equation.

(In this case, E (n _k) is a k-th sensor node (1 ≤ k ≤ the average distance ratio normalization m), e (x _kj) is x _kj of the aggregation result of the normalized distance for value ratio, x _kj Is j-th sensing data of k-th sensor node, n is the number of sensing data) Calculated by the aggregate operation processing device. The method of claim 13, The normalized distance ratio for the aggregation operation is represented by the following equation,

(At this time,

Is the average of the j th sensing data of each node,

Is the largest difference between the j th sensing data of each node and the result of the aggregation operation of the j th sensing data of each node) Calculated by the aggregate operation processing device. The method of claim 14,

And when the value is 0, the e (xkj) value is set to 0. The method of claim 11, The normalized distribution ratio of each sensor node is represented by the following equation

Where D (n _k ) is the normalized distribution ratio of the kth sensor node, and max (standard deviation of x _i ) is the largest of the standard deviations of each of the sensor nodes. Calculated by the aggregate operation processing device. The method of claim 11, The normalized distribution ratio of each sensor node is represented by the following equation

Where D (n _k ) is the normalized distribution ratio of the kth sensor node, and max (variance of x _i ) is the largest of each variance of the sensor nodes. Calculated by the aggregate operation processing device. The method of claim 11, And the sampling group extracting unit calculates the priority of each sensor node and extracts a predetermined number of sensor nodes into a sample group in order of high priority. The method of claim 18, The priority is the following equation

Where R (n _k ) is the priority of the kth sensor node, E (n _k ) is the normalized average distance ratio of the kth sensor node, D (n _k ) is the normalized distribution ratio of the kth sensor node, ω is the weight 0≤ω≤1) Calculated by the aggregate operation processing device.

Images