KR100627328B1 - Energy Efficient Data Aggregation Method in Wireless Sensor Networks - Google Patents

Abstract

The present invention relates to a system that can increase the network survival time by reducing the amount of data transmission in the sensor network, interworking with the sensor network that collects data from at least one sensor node, and transmits the monitoring task through a wired or wireless communication network including the Internet. And a remote management server that integrates and manages the collected monitoring data and reports it to an administrator, receives a monitoring task from the remote management server, manages the sensor node, and collects data that has been merged from the sensor node and manages the remote management. It includes a sink node that forwards to the server. The sensor node obtains an average value and a standard deviation of data detected by the one or more sensor nodes, and compares the data with previous data to determine whether to transmit the data to the sink node. . According to the present invention, it is possible to reduce the amount of packet transmission of the entire network by transmitting the detected data only at the time of the data change, thereby making it possible to efficiently use energy in a sensor network with severe energy constraint.

Sensor network, data merging, routing protocol

Description

Energy Efficient Data Aggregation Method in Wireless Sensor Networks

1 is a schematic diagram illustrating a sensor network system.

2 is a block diagram of a sensor network system according to an embodiment of the present invention.

3 is a flowchart showing a data merging method performed in a node of a sensor network.

4A and 4B are conceptual views illustrating energy efficiency in a sensor network according to an embodiment of the present invention.

<Description of the symbols for the main parts of the drawings>

100: sensor network node 110: processor device

120: wireless communication device 130: sensor device

140: sink node 150: remote management server

The present invention relates to a data acquisition method of a sensor network for recognizing the situation of the real world, specifically, when transmitting sensor data collected from a node of a sensor network consisting of a sensor, a processor, and a wireless communication device, a sensor of a previous time The present invention relates to a system and a method for reducing the traffic of a network by transmitting sensor data only when there is a difference over a predetermined threshold through comparison with data.

The sensor network, which has been used primarily for military operations, has been miniaturized and improved in performance due to the improvement of semiconductor technology, the memory capacity has been increased, and the cost has been realized. It is just before commercialization. Sensor network technology applications that are currently proposed are diverse, such as unmanned security, environmental monitoring, remote meter reading, and facility monitoring to monitor conditions such as temperature or pollution level in a certain area or water area. Applications that operate are also proposed and implemented.

As shown in FIG. 1, the basic structure of the sensor network is a structure in which a plurality of network nodes having independent sensing and computing capabilities are interconnected by a communication network, and power of each node is supplied through a local battery located for each node. .

However, since the battery supplying power to each node is a relatively small capacity battery considering the mobility of each node, there is an extremely limited disadvantage in using energy. To overcome this problem, studies on power consumption reduction are conducted in all areas of the network. Has been. The main flow of research direction is to maximize the network survival time by reducing the number of radio communication or the amount of communication between each node, but until now, no specific method has been proposed to achieve energy consumption efficiently. Although research on the point has been published, it has not been able to reduce the actual data transmission amount.

An object of the present invention for solving the above-described problems of the prior art is to measure the amount of change in the data sensed by the sensor and the amount of change in the distribution in the node group over time, so that the two changes are set by the user It is to provide a sensor network system and method for transmitting the actual measured data to the base station node only when the threshold is exceeded.

Another object of the present invention is to apply a threshold that varies according to the position of a node in the sensor network, and to increase the width that becomes a reference for the change by applying a higher threshold as the position of the node to be transmitted from the base station node increases. To provide a sensor network system and method for transmitting data by applying a threshold.

In order to achieve the above object, the sensor network system according to the present invention is a sensor network system for collecting data from at least one sensor node connected to a communication network, and transmits a monitoring task through a wired or wireless communication network including the Internet; The remote management server that integrates and manages the collected monitoring data and reports to the administrator, and receives the monitoring task from the remote management server, manages the sensor node, and collects data merged from the sensor node to manage the remote management. Including a sink node for transmitting to the server, the sensor node obtains the average value and the standard deviation of the sensed data in order to perform the merge process for the data, and compares the average value and the standard deviation before the reference time to the sink To determine whether to send to a node The features.

According to another aspect of the present invention, a method for energy efficient data merging of a sensor network system including a remote management server, a sink node, and a sensor node may include: distributing a monitoring task to the sensor node; Forming a group and designating a head node for each group, acquiring sensor data periodically and / or continuously according to each monitoring node, and detecting the sensor node in the group by the head node for each group. And calculating a mean value and standard deviation of one data, and determining whether to transmit data to the sink node by comparing the mean value, standard deviation, and change amount of a previous reference time. Thresholds, which are the criteria for determining whether to transmit, are preset or dynamically reset according to the monitoring purpose and environment, and the distance factor may be reflected to consider the data transmission path in the network.

As described above, the sensor network is a structure in which a plurality of independent nodes composed of a sensor, a communication device, and a processor are interconnected through a communication network. The general task of passing from a sensor node to a sensor node is to monitor a particular area. Network nodes in this region form a group to collect data, merge the collected data, and send it to the sink node. Since sensor nodes in the sensor network are assigned tasks in a certain area, it is highly likely that multiple nodes in the same monitoring area detect the same event in duplicate.

Accordingly, the present invention provides a sensor network data communication protocol for analyzing a change amount of data generated in the sensor network and merging and transmitting data based on the change amount, and specifically, a standard deviation of data values detected by each sensor node. Obtaining. When the difference between this value and the data sensed in the previous time exceeds a predetermined threshold, the sensor data is transmitted to merge the transmission data.

The merging process of the present invention is performed step by step using the following equations (1) to (3).

Equation 1 represents an average of data detected at time t from n nodes from group nodes 1 to n of sensor network nodes performing the same task in the same region. In this case, t represents a time variable and x _i (t) is a sensor value at time t of node i .

From the average obtained in Equation 1, Equation 2 below is used to obtain a standard deviation of n sensor node data performing the same task.

Equation 2 shows the standard deviation of data detected at time t from nodes 1 to n of group nodes of the sensor network node performing the same task.

The merging method proposed in the present invention calculates the average and standard deviation of data collected by the nodes in the group, and compares it with the value obtained in the previous time, and transmits the data when a certain threshold is exceeded, that is, when the condition of Equation 3 is satisfied. It is.

Equation 3 shows a method of determining whether data is transmitted using the values obtained in Equations 1 and 2, a threshold set by the user, and a distance from the sink node. The threshold field value is a threshold value that is set in advance in order to determine the change of the sensor value over time, and the distance factor is a factor value for changing the distance between the sink node and the sensor node when performing data merging. . The distance factor has a value between 0 and 1, and the further away from the sink, the closer to 1 is assigned.

The reason for considering the distance factor is that the farther the distance between the sensor node and the sink node is, the higher the energy consumption of the entire network is. Therefore, the distance between the sensor node and the sink node reflects the distance factor. The greater the distance, the more dense merging is done to reduce power consumption across the network.

Hereinafter, the configuration of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram illustrating an example of a sensor network structure in which at least one sensor node 100 is connected to a remote management server 150 through a sink node 140 according to the present invention. The sensor nodes 100 are distributed in their respective monitoring areas and receive instructions from the sink node 140 to transmit the monitoring data to the sink nodes. The sink node 140 directly communicates with the remote management server 150 through a wired or wireless network including the Internet, and transmits commands of the remote management server to each sensor node and transmits monitoring data received from the sensor node to the remote management server. .

2 is a block diagram illustrating an internal structure of a sensor node and a relationship between a sink node and a remote management server according to an embodiment of the present invention. 2 illustrates a processor device 110, a wireless communication device 120, and a sensor device 130 included in each sensor node 100, a sink node 140 that manages a plurality of sensor nodes 100, and remote management. Server 150 is shown.

The sensor network node 100 is distributed and distributed in an area to be monitored, and collects data in various ways and transmits the data to the sink node 140 through the wireless network. After the arrangement of the sensor node 100 is confirmed and the sensor network is established, the sink node 140 periodically checks the size of the entire network to classify the geographic area-specific steps. In addition, the sink node 140 may cause the sensor node to periodically report sensor data of a specific region or to report when an event is detected. For example, when the sensor network node 100 is equipped with a temperature sensor, the sink node 140 causes the one or more sensor nodes 100 located in the area to report every 10 seconds. can do. As another example, the sensor node 100 may cause the motion sensor and the camera sensor to report when a person is detected in the area A.

The sensor network node 100 generally includes a processor device 110, a wireless communication device 120, and a sensor device 130. The sensor device may be equipped with various sensors in duplicate, and converts the analog signal detected by the analog-to-digital converter into a digital signal. The sensor data collected from the sensor device 130 is not transmitted as it is, but is first processed by the processor device 110. The fact that the sensor node 100 is equipped with the processor device 110 means that the sensor network is not a simple set of sensors that transmit the collected data to a user, but is a situation in which the sensed data is intelligently processed and transmitted inside the network. Means network.

The sensor node according to the present invention performs a merge operation by exchanging data collected by forming a group of nodes 100 performing the same task. The data processed by the processor device 110 in one sensor node 100 is transmitted to the adjacent sensor node through the wireless communication device 120. The wireless communication device 120 of each sensor network node simultaneously performs a routing role of directly transmitting the data packet generated by the sensor network node to the sink node or transmitting the data packet received from the neighbor node back to the sink node.

The reason why each node 100 performs the routing function is because the energy consumption rule in wireless communication consumes 2 to 4 times the amount of energy to be transmitted, so that the sensor node 100 is connected to the sink node 140. If not very close to the network path, each sensor node 100 transmits along the network path via the adjacent node 100 or through multi-hop rather than directly transmitting the sensor data to the sink node 140 This can significantly reduce the energy consumption of the entire network.

The sink node 140 manages sensor nodes in the jurisdiction. That is, it transmits the task received from the manager to each node and reports the sensor data collected from the network to the manager. In addition, a plurality of sensor nodes are grouped and head sensor node nodes are assigned to the group. Group definitions and head sensor node nodes can be dynamically determined according to changes in sensor data in the surveillance area. In addition, since the sink node 140 does not have the capability of being directly connected to the Internet, the sensor network node 100 developed so far also performs a network gateway function for connecting the sensor network and the Internet.

The remote management server 150 receives a task directly from an administrator, allows each node 100 of the sensor network to perform necessary tasks through the sink node 140, and integrates and manages sensor data collected through the above-described process. And store and display. The remote management server 150 may be directly connected to the sink node 140 or may be connected to the sink node 140 by being connected to the Internet. The fact that the remote management server 150 and the sink node 140 can access the mutual Internet is a major point that can give considerable flexibility in the configuration of the sensor network. That is, the administrator may receive a command and report on the temperature monitoring near the South Pole from the sensor network node installed in the Antarctic region through the remote management server 150 installed in Korea.

Hereinafter, a process of merging sensor data according to the present invention will be described in detail with reference to FIG. 3.

One or more sink nodes 140 connected to the remote monitoring server 150 and a plurality of sensor nodes 100 receiving commands from each sink node 140 are appropriately arranged according to the monitoring purpose, and the remote monitoring server 150 is provided. When a specific monitoring task is given to the sink node 140 from the sink node 140, the sink node 140 receives the received monitoring task (S10). In the next step (S20), the plurality of sensor nodes 100 in the jurisdiction of the sink node 140 itself is classified and grouped by region or detected target criteria, and at least one head sensor node is designated for each group. Alternatively, a plurality of sensor nodes 100 that perform the same task without the sink node designating the group and the head node may independently identify each other to form a group and designate any one sensor node in the group as the head node. have. In this case, a criterion for designating a head node in a group may be a random method or a method in which a node having a smallest average distance from other nodes in the group becomes a head node. It is not limited.

After the group and the head sensor nodes in the group are designated, the sensor network node 100 in the group collects sensor data periodically or in an event driven manner using its sensor device 130 (S30). Transmit to group head via wireless communication device 120. In step S40 and step S50, the group head sensor node collects data collected from sensor nodes in the group and calculates a data mean value and a standard deviation using Equations 1 and 2 above. Next, in order to determine whether to transmit the data, the difference between the average value, the standard deviation and the currently calculated average value and the standard deviation stored in the previous cycle time is calculated using Equation 3 and compared with the preset threshold (S60). ). If the difference with the value of the previous cycle time exceeds a predetermined threshold, the data is transmitted (S65), otherwise, the monitoring operation and the calculation operation are continuously performed until the monitoring task is terminated (S70).

The threshold is pre-set to be appropriate for determining whether previously transmitted sensor data and currently collected sensor data have made a significant change. For example, in order to detect cold temperature, the temperature related threshold difference may be larger than that of the warm zone. In addition, the sensing data transmission threshold from the motion sensor installed in the museum at night will be set to a minimum so that any motion can be reported. Alternatively, the sink node 140 or the remote monitoring server 150 may be dynamically changed by determining an appropriateness of the threshold set according to the monitoring data value and / or the data transmission amount. For example, if the data transfer occurs frequently according to the preset threshold, and the determination after the start of the monitoring determines that the data transmission occurs even though it is not a meaningful change for monitoring purposes, the reset operation for increasing the threshold is performed in the sink node or the remote management server. It is performed at the head node by the command of.

In the method shown in FIG. 4, the reason for comparing the change between the sensor data average value and the standard deviation calculated in the previous cycle time is whether or not the change in the monitoring area occurs over time and the difference in data between the nodes in the monitoring area. To check for two things. If the value of the standard deviation increases, this is the basis for judging that different situations are developing within the same monitoring area, so the group head sensor reports on its own or to the sink node to modify the strategy of the sensing task so that each sensor node is responsible for each sensor node. Dynamically change tasks such as changing regions or regrouping.

On the other hand, in the comparison between the current data and the previous data, the amount and the number of data transmissions to the sink node may differ depending on when the previous data is set as the data. For example, the amount of transfer and the number of times when the previous data is calculated as the initial sensor data and compared with the current data while maintaining it and the previous data are calculated as the sensor data sensed immediately before and the previous data are not considered are different. will be.

As a result of sampling the data generated from the sensor network in various monitoring environments, the inventors found that the change of the situation desired by the administrator is extremely rare because the width of the data change is very small. However, even in this case, when comparing the initial data and the current data at the start of monitoring, the amount of change over time may accumulate, and after a certain time, the difference between the current data and the comparative data may exceed the threshold, and thus the group The head node will send sensor data to sink node 140. On the other hand, if the comparison data is used as the sensor data collected immediately before, since the amount of change over time does not accumulate, if the sensor data changes linearly and slowly, no data transmission will occur within the sensor network. . Therefore, the optimum value of the amount of data transmission and the number of times of transmission in the sensor network can be found by appropriately changing and adjusting the data to be compared with the currently collected data at what point in time according to the requirements of the monitoring environment and the monitoring manager.

On the other hand, as described above, each sensor node 100 may perform a routing function and the transmission path from the group head node 100 to the sink node 140 in the same group or in consideration of the communication efficiency of the entire sensor network. Most are via the sensor node 100 belonging to another group. In this case, in a plurality of head nodes belonging to the jurisdiction of the same sink node 140, the network paths from each head node to the sink node 140 are different.

As described above, when the network path is long, it will pass through more sensor nodes 100 than when the network path is short, and each pass-through sensor must perform routing functions and be involved in data transmission, thus consuming power throughout the network. Is bigger. Therefore, when the network path is long, it is necessary to reduce the network-wide power consumption by reducing the number of data transmissions than is shorter. Therefore, if the distance between the sink node 140 and the sensor node, exactly the head node in the group is more than a predetermined range to reflect the distance factor value to reduce the amount of communication of the entire network as possible. In other words, as the distance from the head node and the sink node 140 increases, the amount of transmission to the entire network increases. In this case, the amount of transmission is reduced by performing more data merging.

To this end, a distance factor is included in Equation 3, which determines the transmission time, and the transmission time is adjusted according to the distance between the head node and the sink node, that is, the network path. Intake is as described above. Of course, depending on the monitoring environment, it may not be necessary to change the transmission amount according to the distance. In this case, the monitoring factor of Equation 3 may be set to 1.

The reason for the above-mentioned data merging is generally possible because sensors mounted in the entire sensor network or nodes in the same group are of the same kind, and nodes in a short distance simultaneously detect similar data.

4A and 4B are diagrams schematically illustrating a data transmission amount in a case where data merging is not performed in the sensor network according to the present invention.

First, referring to FIG. 4A, when the data flow is not performed, the source nodes 1,2 and 3 (100) for monitoring the same region are individually passed through the collected data (100). Transfer to sink node 140 through. In this case, a total of 11 sensor data transmissions occur from sources 1, 2, and 3 to the pass through node a and the pass through node b through the sink node. Here, source nodes 1, 2, 3 and via nodes a, b are all sensor nodes 100 of the sensor network.

On the other hand, the example shown in Figure 4a is assuming that all data paths from the source 1, 2, 3 to the sink node is the optimal path, for example, if some or all of the source nodes take the bypass path network The total amount of transmission and the number of transmissions will increase.

On the other hand, as shown in FIG. 4B, when data is merged, the data transfer amount between the pass-through node a, the pass-through node b, and the sink node 140 is reduced to 1/3, so that only a total of 5 times of data are used throughout the network. The transmission was made. This shows that the energy consumption of the entire network is reduced to less than half. In particular, the more the pass-through node between the head node and the sink node, that is, the larger the sensor network system, the higher the power consumption reduction effect of the present invention.

The present invention has been described with reference to some preferred embodiments as described above, but this is for the purpose of illustration and the scope of the present invention is not limited to the above-described embodiments and the present invention falls within the scope without departing from the spirit of the present invention. Various changes and modifications may be made by those skilled in the art. Therefore, of course, the scope of the present invention should be defined by the claims described below.

According to the present invention, it is possible to substantially and efficiently reduce the power consumption due to data transmission, which is one of the main technical challenges of the sensor network that can be applied to various applications. As the range of the sensor network is expanded, the power consumption of each node due to the data transmission increases rapidly, and there have been considerable difficulties in practical use. However, by adopting the present invention, thousands of sensor nodes capable of wireless communication in a specific monitoring area can be obtained. From hundreds to hundreds of thousands, you can get visibility into the surveillance area while minimizing power usage.

That is, the data is transmitted only when the change is made by comparing the amount of change of data detected by the energy efficient data merging method of the present invention with the previous time zone, thereby reducing the energy consumption of the entire network and detecting when the event is not performed. The same effect is collected.

In addition, the energy-efficient data merging method of the present invention can create a new ubiquitous computing space in which physical space and electronic space are fused to collect data from various sensors to accurately recognize and respond to real-world situations. . In addition, by applying the variable threshold reflecting the distance factor, it is possible to consider the energy consumption in the network according to the paths of the sink node and the sensor node, thereby minimizing the power consumption due to the expansion of the network, and thus a significant contribution to the commercialization of the sensor network. It is expected.

Claims (20)

A sensor network system for collecting data from at least one sensor node connected to a communication network, A remote management server that transmits surveillance tasks through the wired / wireless communication network including the Internet, integrates and manages the collected surveillance data and reports to the administrator; Receiving a monitoring task from the remote management server and managing the sensor node, including a sink node collecting data passed through a merge process from the sensor node and delivering the data to the remote management server, In order to perform the merging process on the data, the sensor node obtains the average value and the standard deviation of the detected data, and compares the average value and the standard deviation before the reference time to determine whether to transmit the data to the sink node. Sensor network system. The method of claim 1, wherein the sensor node is And a sensor device for collecting data, a wireless communication device for performing data transmission and routing functions, and a processor device for performing data processing including data merging. 3. The device of claim 2, wherein the wireless communication device of the sensor node is A sensor network system for automatically configuring a wireless network, performing transmission and reception of data, performing a routing function of retransmitting a received data packet toward a sink node, and performing group communication with other sensor nodes. The processor device of claim 2, wherein the processor device of the sensor node comprises: And a basic computing function for calculating sensor data obtained from the sensor device, and performing the merging process on data transmitted from a plurality of sensor nodes. delete 6. The processor device of claim 5, wherein the processor device of the sensor node is The sensor network system to determine whether to transmit the sensor data further considering the distance factor reflecting the network path between itself and the sink node. 6. The processor device of claim 5, wherein the processor device of the sensor node is And notifying the sink node when the standard deviation exceeds a predetermined range. The method of claim 1, wherein the remote management server A sensor network system in which an administrator has an analytical tool for distributing surveillance missions from a remote site to the surveillance area via the Internet and determining a situation based on collected data. The method of claim 1, wherein the sink node is And a plurality of the sensor nodes, forming a group including at least one sensor node and designating one head sensor for each group. The method of claim 9, wherein the sink node is Sensor network system for receiving data from the head sensor and delivering it to the remote management server. The method of claim 10, wherein the sink node is The sensor network system that receives data from the head sensor and forwards the merged sensor data to the remote management server, and the other data is analyzed and processed by the head sensor. The method of claim 11, wherein the sink node is And the redirection of the group and / or the redirection of the head sensor for each group is performed dynamically when the other data is a standard deviation error range notification from the head node. In the energy efficient data merging method of a sensor network system including a remote management server, a sink node, a sensor node, Distributing a monitoring task to the sensor node; Forming a group with sensor nodes performing the same task and designating a head node for each group; Acquiring sensor data periodically and / or continuously by each sensor node according to a monitoring task; Obtaining, by the head node of each group, an average value and standard deviation of data detected by sensor nodes in the group; Determining whether data is transmitted to the sink node by comparing an average value, a standard deviation, and a change amount of a previous reference time Energy efficient data merging method of the sensor network comprising a. 14. The method of claim 13, wherein deploying the surveillance mission Receiving surveillance task data including a surveillance region, a sensor type, a sensor data reporting period, and an end point of a task from an administrator and distributing the remote management server to sensor nodes in the surveillance region through the Internet and a user network; Energy efficient data merging method The method of claim 13, wherein forming the group of sensor nodes comprises: If you are located within the surveillance zone of the received mission, you record the received mission on internal storage, form a group with other nodes in the surveillance zone, and determine any sensor in the group as the head node. Energy efficient data merging method. The method of claim 13, wherein forming the group of sensor nodes comprises: And the sink node forms a group in consideration of the installation position and type of at least one sensor node under its control, and designates a head node for each group. The method of claim 13, wherein obtaining the sensor data comprises Collecting sensor data at a duty cycle from a sensor device of the sensor node and converting the sensor data into a digital signal through an analog-to-digital converter for reading by a processor device of the sensor node. How to merge. The method of claim 13, wherein collecting data in the group and obtaining the mean and standard deviation Transmitting data collected by the at least one sensor node belonging to the group to the head node; The head node collecting all of the data to obtain an average value and a standard deviation Energy efficiency data merging method of a sensor network comprising a. The method of claim 13, wherein determining whether to transmit sensor data to the sink node Comparing the difference between the mean obtained from the current time and the standard deviation from the previous reference time, and comparing the stored value with a threshold determined by the administrator; If the difference is more than the threshold value, it is determined that the change of the situation of the surveillance area has occurred, and the data transmission to the sink node is performed. Energy efficiency data merging method of a sensor network comprising a. 20. The apparatus of claim 19, wherein the threshold is And a distance factor considering a network path between the head node and the sink node is determined.

Images