KR20090021906A - Wireless sensor network and cluster optimising method therefor - Google Patents

Abstract

A radio sensor network and a cluster optimizing method minimizing the number of clusters using a plurality of sensors nodes performing the breakup and annexation of the cluster are provided to extend the existence time of sensor network by minimizing the number of head nodes. A head node transmits a cluster merge message that the head node sets up cluster to the other head node(S510). A head node receiving the cluster merge message transmits a notice message to the member node(S520). The member node irradiates the head node is positioned within its own communications region. The investigation result is transmitted to the head node(S530). The head node receiving the cluster merge invariable message transmits the head node abort invariable message to member nodes. The member node in which the new head node is affiliates the member node positioned within its own communications area or the sensor node giving up the head node in its own cluster.

Description

Wireless sensor network and cluster optimising method therefor}

The present invention relates to a wireless sensor network and a cluster optimization method, and more particularly, to a wireless sensor network and cluster optimization capable of minimizing the number of clusters by using a plurality of sensor nodes capable of dissolving and merging clusters on their own. It is about a method.

The wireless network is a technology that integrates various sensor devices for situational awareness in a widely installed wired / wireless network infrastructure based on the recent research of ubiquitous computing, and interlocks the detected environmental data with an application service server. The difference from standard sensors such as temperature sensors is the interconnectivity in intelligent clusters and the ability to collect and process data collectively. Using this technology, it is attempting to apply to disaster control, environmental monitoring, intelligent logistics management, real-time security, and mobile healthcare.

In wireless sensor networks, a large number of nodes are distributed for surveillance missions in a wide range of areas, and the data sensed through the sensors is transformed into higher-level events through data processing inside the network and then delivered from remote managers. Data is then transmitted over low-cost, low-power single-hop or multihop wireless networks. The characteristics of the wireless sensor network are as follows.

First, nodes in wireless sensor networks are low-power, low-cost systems that operate on batteries. However, due to the limitations of current battery technology, nodes are severely constrained in their energy use and will not be recharged or replaced if the battery is depleted and the node stops operating. In other words, the nodes of the wireless sensor network have the characteristics of disposable consumables. Therefore, the wireless sensor network should be able to maximize the lifetime of the network by using power efficiently. When analyzing the power consumption of each node by function, the power consumption used in wireless communication is larger than that in sensing or data processing. Therefore, how to operate the wireless communication protocol to reduce power consumption is an important problem to be solved.

Second, the wireless sensor network consists of a large number of communication nodes. Because of the nature of the monitoring itself, more situational information can be grasped if more observation points are collected from more observation points. This improves the accuracy of information by distributing the task from a single sensor to more nodes. This requires three requirements. First, in order to construct a network with a large number of nodes, a low-cost platform should be developed at the implementation stage. Second, a single hop in which the network type does not need an infrastructure in order for these large network nodes to exchange information. -hop or multihop wireless network. A network without infrastructure does not have a fixed network access point such as a base station of a mobile communication or an access point of a wireless LAN. Third, a data merging method that can extract only meaningful results by processing redundant data between nodes in a network and a network structure for performing the same must be presented.

Third, nodes in wireless sensor networks are randomly deployed by plane or other vehicle rather than distributed by humans. This is because the first application of wireless sensor networks will be in areas that are difficult to reach, such as disaster areas or combat areas. This feature has the technical requirement that nodes be able to self-configure the network without human intervention or interference.

Fourth, as mentioned above, the area where the wireless sensor network is deployed is very poor and variable in physical environment. This assumes that node failure, destruction, malfunction, etc. can occur frequently. Thus, in the face of rapidly changing surroundings, it must be equipped with disability tolerance to continue a given task. To meet these requirements, the wireless sensor network increases the density of the nodes so that the monitoring area is overlapped so that a certain level of information can be collected even if some nodes fail.

However, since the energy of the sensor nodes is generally limited and cannot be charged, a method of fairly distributing energy consumption of the sensor nodes is required in order to maximize the survival time of the sensor network. In the process of operating each sensor node, the highest power consumption is used for wireless communication.

Therefore, in order to extend the survival time of the entire sensor network, sensor nodes included in a certain range are formed into one cluster, and one sensor node among the sensor nodes included in the cluster is selected as the head node, so that the head node is the remaining sensor in the cluster. It uses a method of collecting information from nodes to communicate with the server.

In this case, since the sensor node selected as the head node communicates wirelessly with the server, energy consumption is large, and thus, the gap is gradually increased in the sensor network due to the end of life from the node selected as the head node.

To address this problem, the paper "AnApplication-Specific Protocol Architecture for Wireless Microsensor Networks" (Wendi B. Heinselman), published in IEEE Trans.On Wireless Communications Vol. 1, No. 4, dynamically plays the role of node heads between nodes. A method of cycling has been proposed.

According to the clustering method described in this paper, the role of the cluster head is increased by increasing the probability that nodes that do not act as head nodes become the next head node based on whether the sensor nodes play the role of head nodes for a certain period of time. Circulate dynamically

However, according to this method, a sensor node with less energy than a neighboring sensor node may perform the function of a head node, which causes a rapid energy exhaustion and eventually destroys the node function of the corresponding sensor node. There was a problem.

In addition, Korean Patent Laid-Open Publication No. 2006-0055635 uses a method of measuring the remaining energy when setting the head node to evenly distribute the energy consumption so that the node having a high residual amount is set as the head node. However, the above method has a problem in that the survival time of the sensor network is shortened because the excessive number of clusters is set as the head node because the number of clusters is not optimized.

The present invention devised to solve the problems of the prior art as described above is a wireless sensor network and cluster optimization method that can minimize the number of clusters by using a plurality of sensor nodes that can perform the dissolution and merge of the cluster itself It is an object of the present invention to provide.

The object of the present invention is a wireless sensor including a plurality of sensor nodes capable of wireless communication with the surrounding sensor node and server and capable of performing dissolution and merging of a cluster, and a server for receiving and processing data from the sensor node. Achieved by the network.

Another object of the present invention is to provide a first step of transmitting a head node abandonment notification message to all member nodes in a cluster when at least one head node of another cluster is located in a communication area of the head node of a cluster; A second step of all the member nodes which have received the message check whether a head node of another cluster is located in their communication area and transmit a response message to the head node; A third step of receiving a result of the survey of all member nodes, when the response of all member nodes is possible according to the response message, transmitting the head node abandonment confirmation message to all member nodes; A fourth step of releasing the cluster from the head node and subscribing to the head node of another cluster in its communication area as a member node and maintaining the head node state as it is when at least one of the response messages is not available. It is achieved by a cluster optimization method comprising a.

In addition, another object of the present invention is the first step of transmitting a cluster merge message to the head node when there is more than one head node in its communication area in the sensor node that is a member of any cluster; Each head node receiving the cluster merge message transmits a head node abandon notice message to all member nodes belonging to its cluster; A third step of all the member nodes which have received the message check whether a head node of another cluster is located in their communication area and transmit a response message to the head node; Receiving a result of the survey of all member nodes, the head node transmitting a discardable message when the response of all member nodes is possible and a non-stopping message when at least one is impossible, to the sensor node according to the response message; A fifth step in which the sensor node receiving the message maintains a current member node when at least one message is a head node abandonment message; A sixth step of transmitting a merge determination message to each head node when the messages are all abandonable to the head node, and becoming a head node; The head node receiving the merge confirmation message sends a head node abandonment confirmation message to a member node in the cluster, and a seventh step of giving up the head node, a node giving up the head node, and a member node included in the cluster of the head node. A cluster optimization method comprising an eighth step of subscribing as a member node to a head node located within its own communication area, respectively.

Accordingly, the wireless sensor network and the cluster optimization method of the present invention use the plurality of sensor nodes capable of dissolving and merging the clusters themselves to dissolve redundant clusters or to merge two or more clusters into one to consume energy. By minimizing the number of head nodes, it is possible to extend the survival time of the entire sensor network.

The terms or words used in this specification and claims are not to be construed as being limited to their ordinary or dictionary meanings, and the inventors may appropriately define the concept of terms in order to best describe their invention. It should be interpreted as meaning and concept corresponding to the technical idea of the present invention based on the principle that the present invention.

Therefore, the embodiments described in the specification and the drawings shown in the drawings are only the most preferred embodiment of the present invention and do not represent all of the technical idea of the present invention, various modifications that can be replaced at the time of the present application It should be understood that there may be equivalents and variations.

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

1 is a structural diagram showing the basic structure of a wireless sensor network according to the present invention. Referring to FIG. 1, a wireless sensor network is capable of wirelessly communicating with neighboring sensor nodes and servers, a plurality of sensor nodes capable of dissolving and merging clusters, and a server for receiving and processing data from the sensor nodes ( 100).

The server 100 may wirelessly communicate with the head node 130 and collect a data transmitted from the head node 130.

The sensor node can communicate with the server 100 and subscribe to the head node 130 and the head node 130, which can transmit data received from the member node 120 to the server, and collect the data collected by the head node (head node). It consists of a member node 120 that can be transmitted to 130. The head node 130 and the member node 120 joined as a member of the head node 130 is referred to as a cluster (110).

The head node 130 and the member node 120 have the same structure, and their functions are not determined from the beginning of network creation to the end of network generation. Accordingly, the member node 120 may also be the head node 130 according to a situation, and the head node 130 may also be in a state transition to the member node 120. It has a random timer inside, and when the cluster is formed at the beginning of sensor network creation or when the cluster is reconfigured by the surrounding environment such as adding or deleting sensor nodes, the cluster optimization function is not performed at the same time, but with a certain time difference. Only one or fewer sensor nodes should perform cluster configuration or reconfiguration.

The sensor node configured as the head node 130 configures the cluster 110 by joining the member node 120 in its wireless communication area as its member. When one or more other head node 130 is located in its wireless communication area, the head node 130 may be abandoned and joined as a member node of another head node 130 (120). It transmits the data collected by the member nodes 120 in the) to the server 100. In addition, the head node 130 may be given up according to the remaining energy of the head node 130, and the neighboring head node 130 may be joined as the member node 120.

In addition, the member node 120 transmits the collected data to the head node 130 according to the function of the sensor, and other head node 130 is located in the wireless communication area in addition to the head node 130 subscribed to. When present, the user may directly become the head node 130 to join another head node 130 to his or her member node 120.

The sensor node may transmit and receive state information with surrounding sensor nodes, and store state information received from surrounding sensor nodes, thereby reducing the number of unnecessary communication between sensor nodes.

2 is a conceptual diagram showing the cluster dissolution method of the present invention. Referring to FIG. 2, when the head node 220 is activated by the timeout of the random timer, it is searched whether one or more head nodes 240 are located in its cluster 210. When the cluster 210 overlaps, the number of head nodes 220 and 230 is unnecessarily increased, and the survival time of the entire wireless sensor network is short because energy consumption is large due to the characteristics of the head nodes 220 and 230 that communicate with the server. You lose.

Accordingly, the head node 220 in which the cluster 210 is duplicated has the member nodes 240 joined to all the other head nodes 240 surrounding the head node 220, so that the head node 220 also gives up the head node state and the other head nodes. Join as a member node of 240. The cluster 210 established by the head node 220 is dissipated as the head node state is abandoned, and the number of clusters in the entire sensor network is reduced.

3 is a flowchart showing the cluster dissolution method of the present invention. Referring to FIG. 3, the head node is activated when the random timer included therein times out and searches for the presence of the head node in its communication area. If there is more than one head node, the head node abandon notice message is transmitted to the member nodes subscribed as their members (S310).

When the member node receives the head node abandonment notification message, the member node searches whether another head node is located in its communication area, and transmits the result to its head node (S320). The resulting message sends a discardable message if there is a head node to which it subscribes, even if the head node to which it is a member gives up the head node status. If not within the wireless communication area, a message for not giving up is transmitted.

The head node grasps the received message (S330) and if all the messages are abandonable messages, transmits a head node abandonment confirmation message that gives up the head node status to the member node so that the member node joins the neighboring head nodes as members. , He becomes a member node and joins another head node as a member (S340). On the other hand, if there is more than one message that cannot be abandoned among the received messages, the head node state is maintained (S350) to continue to perform the function of the head node.

In addition, the head node may perform the above process according to the amount of energy of itself, except when another head node is located in its area, and reaches a state where almost no energy remains, so that the head node no longer serves as a member node as well as a member node. When it is difficult to perform a function, it can send a head node abandon message. In this case, a member node without a head node that can be newly joined may become a head node by itself.

4 is a conceptual diagram illustrating a cluster merging method of the present invention. Referring to FIG. 4, when member node B 430 of head node A 420 is activated by a timeout of a random timer, it searches for two head nodes A 420 and C 440 within a communication area. do. When the clusters 410 and 450 overlap, the number of head nodes 420 and 440 is unnecessarily increased and the energy consumption of the head nodes 420 and 440 communicating with the server is high, so the survival time of the entire wireless sensor network is increased. Will be shorter.

Member node B (430) in a position where two clusters (410, 450) overlap, causes two head nodes A (420) and C (440) to give up the head node state, and 430 itself direct head node. Then, to join the sensor node in its own communication area as its member node to create a new cluster (460).

As member node B 430 becomes a new head node, a new cluster 460 is created, but two existing clusters 410 and 450 are reduced, thereby reducing the number of clusters of the entire sensor network.

5 is a flowchart illustrating a cluster merging method of the present invention. Referring to FIG. 5, the member node is activated when the random timer included therein times out and checks whether there is a head node in its communication area. If there is more than one head node, it transmits a cluster merge message to the corresponding head node to establish a cluster as the head node (S510). The head node receiving the cluster merge message transmits a head node abandon notice message to its member node (S520).

The member node that receives the head node abandon notice message checks whether the head node is located in its communication area and transmits the result to the head node (S530). The head node sends a discardable message to the member node that sent the initial cluster merge message if all member nodes are abandonable based on the result received from the member node, and a non-surrenderable message if one or more non-abandonable messages are received. Send it.

The member node that transmits the first cluster merge message transmits the cluster merge confirmation message to the head node when all the head nodes can be given up according to the result received from the head node (S540) to give up the head node state.

Receiving the cluster merge confirmation message, the head node transmits a head node abandonment confirmation message to member nodes subscribed as its members (S550). A member node that has become a new head node joins a member node or a sensor node that has abandoned the head node within its communication area to its cluster, thereby creating a new cluster (S570), and the existing cluster is naturally dissolved, so that the entire sensor network Will reduce the number of clusters.

However, when one or more non-cancellation messages are received from the head node, the member node that transmitted the first cluster merge message maintains the existing member node state (S560), and operates as a member node of the cluster as before.

Although the present invention has been shown and described with reference to the preferred embodiments as described above, it is not limited to the above embodiments and those skilled in the art without departing from the spirit of the present invention. Various changes and modifications will be possible.

1 is a structural diagram showing a basic structure of a wireless sensor network according to the present invention;

2 is a conceptual diagram showing the cluster dissolution method of the present invention;

3 is a flowchart showing a cluster dissolution method according to the present invention;

4 is a conceptual diagram illustrating a cluster merging method of the present invention;

5 is a flowchart illustrating a cluster merging method of the present invention.

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

            100: server

            110, 210, 410, 450, 460: cluster

            120, 240, 430: member nodes

            130, 220, 420: head node

Claims (12)

A plurality of sensor nodes capable of wireless communication with neighboring sensor nodes and servers and capable of dissolving and merging clusters, and Wireless sensor network comprising a server for receiving and processing data from the sensor node. The method of claim 1, The sensor node incorporates a random timer to perform cluster optimization so that the sensor node is activated as soon as the random timer times out. The method of claim 1, The sensor node may transmit its state information to all sensor nodes located within its communication area, and receive and store state information of neighboring sensor nodes. The method of claim 1, The sensor node sets an arbitrary sensor node as a head node, joins a sensor node around the head node as a member node of the head node, creates a cluster, and then collects and transmits data transmitted from the member node to a server. Wireless sensor network. The method of claim 4, wherein And the head node can set or give up the head node state according to the state of the sensor nodes in the communication area. The method of claim 4, wherein The member node transmits data collected by the member node to the head node and, upon receiving a head node abandonment notification message from the head node, searches for the presence of another head node in the communication area and transmits a result to the head node. network. The method of claim 4, wherein The member node may be a head node by itself, when there are two or more head nodes in a communication area, and merge existing clusters. A first step of transmitting a head node abandon notice message to all member nodes when there is at least one head node of another cluster in a communication area in a head node of one cluster; A second step of all the member nodes which have received the message, searching whether another head node is located in a communication area and transmitting a response message to the head node; The head node receiving the survey result of all the member nodes sends a head node abandonment confirmation message to all member nodes when the response of all member nodes is abandonable according to the response message; Dissolving the cluster at the head node and joining the other head node in the communication area as a member node; and A fifth step of maintaining a state of the head node when at least one of the second messages of the second message cannot be abandoned; Cluster optimization method comprising a. A first step of transmitting a cluster merge message to the head node when there are two or more head nodes in a communication area in a sensor node that is a member of a cluster; Each head node receiving the cluster merge message transmits a head node abandon notice message to all member nodes belonging to its cluster; A third step of all the member nodes which have received the message check whether a head node of another cluster is located in their communication area and transmit a response message to the head node; The head node receiving the survey result of all the member nodes sends a discardable message when the response of all member nodes is abandonable, and a non-cancelable message when at least one is impossible, to the sensor node according to the response message. ; A fifth step in which the sensor node receiving the message maintains a current member node when at least one message is a head node abandonment message; A sixth step of transmitting a merge determination message to each head node when the messages are all abandonable to the head node, and becoming a head node; Receiving a merge confirmation message, the head node sends a head node abandonment confirmation message to a member node in a cluster and abandons the head node; An eighth step in which the node giving up the head node and the member node included in the cluster of the head node respectively join the head node located in its communication area as a member node; Cluster optimization method comprising a. The method according to claim 8 or 9, And all the sensor nodes include a random timer, and are activated as soon as the random timer times out such that the head node performs the process of claim 8 and the member node performs the process of claim 9. The method according to claim 8 or 9, And all the sensor nodes are configured as head nodes when the sensor network is initially created or added to an existing sensor network. The method of claim 9, In the fifth step, if there is no head node to which the sensor node is joined, the cluster optimization method joins a head node located in a wireless communication area as a member node.

Images