KR100994732B1 - Method for clustering a ad-hoc network - Google Patents

Abstract

The present invention relates to a cluster forming method of an ad hoc network, wherein each node transmits and receives a message including information on the corresponding node, calculates mobility, and compares mobility with neighboring nodes to determine whether the node is a cluster head or a cluster member. By recognizing, the resources of the network can be used efficiently. By selecting a node with less movement as the cluster head, it is possible to stably form and manage a cluster and to maintain a stable overall network topology. In addition, the mobile ad hoc network can prevent limited network resource waste and increase efficiency. In addition, by reducing the frequency of cluster head re-election and cluster reconfiguration due to the movement of the cluster head, it is possible to maintain a stable network topology and improve network performance.

Wireless network, mobile, ad hoc

Description

Method for clustering a ad-hoc network

The present invention relates to a method for forming a cluster in an ad hoc network, and more particularly, to a method for forming a cluster in consideration of mobility of nodes in a mobile ad hoc network.

With the development of electronic and communication technologies, users can send and receive data anytime, anywhere through the network. As described above, a network for transmitting and receiving data includes a wired network that directly connects electronic devices using a wire to form a network, and a wireless network that forms a network without using a wire.

The wireless network includes all types of computer networks formed wirelessly. A wireless network can communicate without using wires between network nodes. The wireless network may be divided into a wireless personal area network (WPAN), a wireless local area network (WLAN), a wireless metropolitan area network (WMAN), a wireless wide area network (WWAN), and the like, according to a range in which the network is formed.

The wireless network may be connected to a fixed wired network, or may be a network consisting only of a mobile host without being connected to a fixed wired network. A network composed only of mobile hosts, not connected to the wired network, is called a mobile ad-hoc network (or a mobile ad-hoc network). Each mobile host (or mobile node) in a mobile ad hoc network functions as one router as well as a host and may have multiple paths to other nodes. You can also set the path dynamically.

The present invention provides a method for forming a cluster in an ad hoc network that can form a cluster in consideration of the mobility of the nodes in the process of forming a cluster of mobile nodes in the ad hoc network.

In the cluster forming method according to the present invention, each node transmits and receives a message including information on the corresponding node. When the first node receives a message of the corresponding node from a neighboring node, the first node is included in the message. Updating neighboring node information of the first node using information of the corresponding node; and calculating mobility of the first node using the updated neighboring node information.

In another aspect, the method for forming a cluster according to the present invention includes the steps of: each node transmitting and receiving a message including information of a corresponding node with a neighboring node, and each node storing the information of the neighboring node included in the received message. The node compares its mobility with the mobility of the neighboring node included in the message, and each node compares the mobility of the neighboring node with the cluster member if its mobility is higher than that of the neighboring node. And recognizing and operating itself as a cluster head if its mobility is lower than that of a neighboring node.

According to the method for forming a cluster in the ad hoc network of the present invention, it is possible to efficiently manage mobile nodes and to efficiently use resources of the network. By selecting a node with less movement as the cluster head, it is possible to stably form and manage a cluster and to maintain a stable overall network topology. In addition, the mobile ad hoc network can prevent limited network resource waste and increase efficiency.

In addition, by reducing the frequency of cluster head re-election and cluster reconfiguration due to the movement of the cluster head, it is possible to maintain a stable network topology and improve network performance.

The objects, features and advantages of the present invention will become apparent from the following detailed description of embodiments taken in conjunction with the accompanying drawings. In addition, the terms used in the present invention was selected as a general term widely used as possible now, but in certain cases, the term is arbitrarily selected by the applicant, in which case the meaning is described in detail in the corresponding description of the invention, It is to be clear that the present invention is to be understood as the meaning of terms rather than names.

The operation of the method for forming a cluster in the ad hoc network according to the present invention configured as described above will be described in detail with reference to the accompanying drawings.

1 is a diagram schematically illustrating a mobile ad hoc network. A mobile ad hoc network (MANET) refers to a network consisting only of mobile hosts, not connected to a wired network.

In FIG. 1, six mobile nodes 100, 110, 120, 130, 140, and 150 form an ad hoc network. In the network, a computer that supports wireless communication or various mobile communication terminals may be used as a node. Each node has a range of transmission ranges, and may route data through other nodes within the transmission range and transmit the data to the destination node. For example, nodes can route data to their destination nodes using routing protocols such as Zone Routing Protocol (ZRP), Ad-Hoc Demand Distance Vertor Routing (AODV), Temporally Ordered Routing Algorithm (TORA), or Dynamic Source Routing (DSR). Can transmit

In the mobile ad hoc network environment as described above, frequent topology changes may occur due to the movement of nodes, and frequent head selection and topology recognition rate may also decrease in cluster formation. Therefore, hereinafter, a method of measuring a node's mobility and forming a cluster based thereon will be described in order to efficiently form a cluster and stabilize a network topology.

The mobility of the node may use the movement speed of the node measured through the Global Positioning System (GPS). However, in the present invention, the mobility of the node is considered in consideration of a connection state with neighboring nodes, not simply using the movement speed of the node. Measure and use it for cluster formation. In the example of the present invention, each node transmits a Hello message including its various information to neighboring nodes, and calculates mobility using the received Hello message. Each node selects a cluster head using the calculated mobility information and forms a cluster.

2 is a diagram schematically illustrating a format of a Hello message according to an embodiment of the present invention. The message refers to a message including various kinds of information of the node transmitting the message. The message includes an MH_ID field 200, a CH_ID field 210, a K field 220, and an HP field 230, and each field consists of 8 bits. The name of the field is one example, and the content of the invention is not limited to the name.

The MH_ID field 200 includes identification information (ID) of a node transmitting the Hello message. Each node contains its own identification information (ID) that distinguishes it from other nodes. The CH_ID field 210 includes head identification information (ID) of a cluster to which the node transmitting the Hello message belongs. That is, it includes identification information of the head node of the cluster to which it belongs. The K field 220 includes mobility information of a node transmitting the Hello message, and the HP field 230 includes cycle information of the message.

Each node of the mobile ad hoc network broadcasts a Hello message as shown in FIG. 2 to neighboring nodes within a transmission range. Each node may broadcast its Hello message at predetermined intervals and periodically transmit its information to neighboring nodes. When receiving the above Hello message from the neighbor node, each node is a table (for example, neighbor node table (NNT)) to identify the neighbor node's identification information, mobility of the neighbor node, cluster head information of the neighbor node, and the like. Save it. Each node may calculate its mobility using the information that is periodically updated.

The mobility refers to a numerical value of the change of neighboring nodes. That is, it is determined that the more the number of changes of the neighboring nodes, the higher the mobility, and the less the number of changes of the neighboring nodes, the lower the mobility. The information includes the number of nodes that are determined to be their own neighbors as a new node moves into the transmission range of an arbitrary node, and the number of nodes that were their own neighbors but lost their link state by moving out of the transmission range. It can be calculated using.

The mobility of each node may be calculated using Equation 1 or Equation 2 below.

는 현재 시간 t에서 임의의 노드 i의 이동성을 나타낸다.

는 노드 i의 전송 범위 내로 이동하여 새로 가입된 노드의 수,

는 노드 i의 전송 범위 밖으로 이동하여 탈퇴하는 노드의 수를 나타낸다.

는 현재 시간 t의 바로 전 주기인 t-1 시간에서 노드 i의 주변 노드의 총 수를 나타낸다.

는 이전 시간 t-1 의 주변 노드의 수로써 현재 시간 t의 주변 노드의 수와 비교하여 클러스터 내 노드의 변화를 인지하는데 사용된다. 상기 수학식 1에서 각 노드의 이동성은 해당 노드의 전송 범위에 가입하거나 탈퇴함으로써 변동되는 노드의 총 수를 이전 시간의 주변 노드의 수로 나눈 값으로 나타낼 수 있다.

Denotes the mobility of any node i at the current time t.

Is within the transmission range of node i and the number of newly joined nodes,

Denotes the number of nodes moving out of the transmission range of node i and leaving.

Denotes the total number of neighboring nodes of node i at time t-1, just before the current time t.

Is the number of neighboring nodes at the previous time t-1 and is used to recognize the change of nodes in the cluster compared to the number of neighboring nodes at the current time t. In Equation 1, the mobility of each node may be expressed as a value obtained by dividing the total number of nodes changed by joining or leaving the transmission range of the corresponding node by the number of neighboring nodes of the previous time.

The mobility may be calculated by the same method as in Equation 1 above, or the mobility may be calculated by the same method as in Equation 2 below. In Equation 2, the average mobility of the node is calculated by weighting the average average mobility of the past and the current mobility of each node.

는 현재 시간 t에서 노드 i의 가중치 평균 이동성을,

는 이전 시간 t-1에서 노드 i의 가중치 평균 이동성을,

는 현재 시간 t에서 노드 i의 이동성을 나타낸다. α는 임의의 가중치 변수를 나타낸다. 상기 α는 0 < α < 1의 값을 가지며, 구현 예에 따라 상기 가중치 변수 값은 달라질 수있다.

Is the weighted average mobility of node i at current time t,

Is the weighted average mobility of node i at previous time t-1,

Denotes the mobility of node i at the current time t. α represents an arbitrary weight variable. Α has a value of 0 <α <1, and the weight variable value may vary according to embodiments.

Each node may be transmitted including the mobility value calculated by Equation 1 in the K field 220 of the Hello message as shown in FIG. 2, or may include the average mobility value calculated by Equation 2 above. . Alternatively, when the first Hello message is transmitted, the mobility message including the mobility value calculated by Equation 1 may be transmitted, and after that, the average mobility value calculated by Equation 2 may be included. Which value is not used may vary depending on implementation.

Each node compares its mobility with the mobility of its neighbors, which is included in the Hello message from its neighbors, and determines whether it will be a cluster head or a cluster member. Each node compares its mobility with that of its neighbors, and if its mobility is greater than its mobility, it recognizes itself as a cluster member and operates. If its mobility is less than that of neighboring nodes, it recognizes itself as a cluster head and operates.

For example, assume that neighboring nodes of the second, third, and fourth nodes exist within the transmission range of the first node. The first, second, third, and fourth nodes transmit and receive a Hello message to each other to recognize each other as neighboring nodes, and compare mobility. If the mobility of the first node is less than the mobility of the second, third, and fourth nodes, which are neighboring nodes, the first node recognizes and operates as a cluster head. However, if the mobility of the first node is greater than at least one of the second, third, and fourth nodes that are neighboring nodes, the first node recognizes and operates as a cluster member.

3 is a diagram schematically illustrating a movement of nodes and a transmission range of each node according to an embodiment of the present invention. In FIG. 3, ten nodes form an ad hoc network. In the following description, a process of forming a cluster will be described in detail on the assumption that nodes 3 and 5 move.

Each node 1 to 10 of FIG. 3 transmits and receives a Hello message and transmits its information to the neighboring nodes, and receives the information of the corresponding node from the neighboring nodes. Each node uses the received information to create a peripheral node table (NNT). After sending and receiving a Hello message at time t-1, each node sends and receives a Hello message and updates the neighbor node table at time t after one transmission period has elapsed. 3 shows the positions of each node before the third and fifth nodes move at time t-1, and the positions of the respective nodes after the third and fifth nodes move at time t.

Each node calculates mobility of the node using the information included in the Hello message transmitted and received at each time. The mobility calculation is as described above.

)가 된다. 3번 노드의 경우, t-1 시간에 Hello 메시지를 주고받는 노드는 4번, 6번, 7번, 및 9번 노드이고, t 시간에 Hello 메시지를 주고받는 노드는 4번, 및 6번 노드가 된다. 따라서, 수학식 1에 의하면 t시간에 3번 노드의 이동 성은 0.5(

)가 된다. 각 노드에서의 이동성을 산출하면 도 4와 같은 값을 얻을 수 있다.4 is a diagram illustrating mobility of each node according to an embodiment of the present invention. MH_ID represents identification information of each node, and K represents mobility of each node. 4 is a table illustrating a result calculated using the mobility calculation equation of Equation 1. FIG. For example, in the case of node 1, the nodes that send and receive Hello messages at time t-1 are nodes 2, 4, 5, 7, and 8, and the nodes that send and receive Hello messages at time t are Nodes 2, 4, 7, and 8. Therefore, according to Equation 1, the mobility of node 1 at time t is 0.2 (

) In the case of node 3, nodes 4 and 6, 7, and 9 send and receive Hello messages at time t-1, and nodes 4 and 6 send and receive Hello messages at time t. Becomes Therefore, according to equation 1, the mobility of node 3 at time t is 0.5 (

) By calculating the mobility at each node, the same value as in FIG. 4 can be obtained.

Each node may store its mobility value, and transmit the mobility value to the neighboring node by including the mobility value in the next Hello message. Alternatively, the average mobility value (Equation 2) obtained by weighting the average mobility value at time t-1 and the mobility value at time t such as 4 may be included in the Hello message of the next period and transmitted to the neighboring node. Which value to use may vary depending on implementation. In each node, the mobility value may be used to compare the mobility of neighboring nodes with their own mobility and form a cluster.

FIG. 5 is a diagram schematically illustrating a process of forming a cluster according to an embodiment of the present invention, and FIG. 6 is a diagram illustrating a node selected as a cluster head according to time as an embodiment according to the present invention.

5A illustrates an example of forming a cluster using mobility of each node at time t, and (b) illustrates an example of forming a cluster using mobility of each node at time t + 1. to be. 6 shows the cluster head CH_ID of each node at time t and the cluster head CH_ID of each node at time t + 1. For convenience of explanation, it is assumed that each node does not move at time t + 1. When each node does not move at time t + 1, the mobility of each node becomes 0 according to Equation 1, and it is assumed that after t + 1 time, the average mobility of Equation 2 is transmitted and received. That is, when calculating the mobility of each node by transmitting and receiving a Hello message at first, the mobility is calculated using Equation 1, and then, using the Equation 2, the weighted average of the past mobility and the current mobility is averaged and averaged. Assume the case of calculating mobility.

According to the mobility of FIG. 4, the cluster head can be selected at time t as shown in FIG. 6. For example, for node 1, the neighboring nodes in the transmission range at time t are numbered 2, 4, 7, and 8. The mobility of node 1 is 0.2, node 2 is 0.25, node 4 is 0, node 7 is 0.4, node 8 is 0. Node 1 is considered to be a cluster member because its mobility is greater than that of nodes 4 and 8. Node 1 elects node 4 as the cluster head. Node 8 of node 1's peripheral node also has the same mobility as node 4 with 0 mobility, but node 1 selects node 4 with low node ID as the cluster head. Alternatively, a node having a high node ID may be selected as a cluster head according to an embodiment.

At time t, the neighboring nodes of node 3 are nodes 4 and 6. Although node 4 and node 6 have the same mobility as 0, node 3 selects node 4 with a low node ID as the cluster head and recognizes itself as a cluster member.

Each node may elect node 4 and node 6 as the cluster head through the above process. Node 1, node 3, node 7, and node 8 form a cluster with node 4 as a cluster head, and node 6 is elected as a cluster head without cluster members. Nodes 9 and 6 elect node 10 as the cluster head, and node 10 elects node 2 as the cluster head. Node 2 may elect node 1 as a cluster head, but node 1 recognizes node 4 as a cluster head and itself as a cluster member, so no more cluster heads are selected. That is, at time t, only nodes 4 and 6 are recognized as cluster heads. A cluster head node must recognize itself as a cluster head, and if it recognizes itself as a member node of another cluster, it cannot be a cluster head.

In the case of t + 1 time that a Hello message is transmitted and received between nodes after one cycle has elapsed, assuming that there is no change in the position of each node, the current mobility of each node at time t + 1 is zero. In the case of initially calculating the mobility of each node, it is assumed that the mobility is calculated using Equation 1, and then the average mobility is calculated by weighting the average mobility of the past and the current mobility using Equation 2 and calculating the average mobility. According to the average mobility calculation method, the difference in mobility between nodes at time t + 1 is not significantly different from the difference in mobility between nodes at time t.

Thus, nodes 4 and 6 continue to be elected cluster heads to form clusters. In the case of node 2, nodes 1 and 7 of the surrounding nodes 1, 7, and 10 are recognized as cluster members, so they are recognized as cluster heads compared to node 10. Node 10 selects node 2 as the cluster head and forms a cluster. In the case of node 9, node 10 of node 5 and node 10 recognize node 2 as cluster head and recognize itself as cluster member, so it recognizes itself as cluster head compared to node 5. do. Node 5 selects node 9 as the cluster head and forms a cluster.

In this manner, each node can compare mobility with neighboring nodes to elect a cluster head and form a cluster.

7 is a flowchart schematically illustrating a cluster formation process as an embodiment according to the present invention.

Each node forming the network in the mobile ad hoc network transmits a Hello message including various information of the corresponding node to the neighboring nodes (S700). Each node may broadcast a Hello message having a format as shown in FIG. 2 and transmit the same to a plurality of unspecified neighbor nodes.

Each node stores the information of the neighboring nodes included in the received Hello message in a table and calculates its mobility (S710). As described above, each node may calculate its current mobility using Equation 1 or Equation 2. Currently, its mobility is included in the next Hello message and sent to neighboring nodes in the next cycle.

Each node compares its mobility with that of the neighboring nodes included in the received Hello message (S720). Each node compares its mobility with that of its neighbors (S730), and if its mobility is higher than its neighbors, the node recognizes it as a cluster member and operates (S740). By comparing the mobility of the neighboring nodes and their mobility in each node (S730), if their mobility is lower than that of the neighboring nodes, the node is recognized as a cluster member and operates (S750).

Each node transmits and receives a Hello message at a set cycle, and steps S700 to S750 are repeatedly performed at every cycle of transmitting and receiving a Hello message.

In this manner, a node with less movement can be selected as a cluster head to stably form and manage a cluster, and can stably maintain the overall network topology.

The present invention is not limited to the above-described embodiments, and as can be seen in the appended claims, modifications can be made by those skilled in the art to which the invention pertains, and such modifications are within the scope of the present invention.

1 schematically illustrates a mobile ad hoc network

2 is a diagram schematically illustrating a format of a Hello message according to an embodiment of the present invention.

3 is a diagram schematically illustrating a movement of nodes and a transmission range of each node according to an embodiment of the present invention.

4 is a diagram illustrating mobility values of each node according to an embodiment of the present invention.

5 is a diagram schematically illustrating a process of forming a cluster according to an embodiment of the present invention.

6 is a diagram illustrating a node selected as a cluster head according to an embodiment according to the present invention.

7 is a flowchart schematically illustrating a cluster formation process as an embodiment according to the present invention.

* Explanation of symbols for the main parts of the drawings

100: first node 110: second node

120: third node 130: fourth node

140: fifth node 150: sixth node

Claims (13)

delete delete delete delete delete delete delete In an ad hoc network, Each node transmitting and receiving a message with a neighboring node, the message including information of a node transmitting the message, each node storing information of the neighboring node included in the received message; Each node comparing its mobility with that of a neighboring node included in the message; And Each node recognizes itself as a cluster member when its mobility is higher than that of neighboring nodes, and when it is lower than its mobility, it recognizes itself as a cluster head. A cluster forming method comprising the step of operating. The method of claim 8, Wherein each node transmits / receives the message to a neighboring node at every set period. The method of claim 8, The message is A cluster forming method comprising at least one of identification information of a node transmitting the message, head node identification information of a cluster including a node transmitting the message, mobility information of a node transmitting the message, and message transmission period information. . The method of claim 8, The mobility of the node, The sum of the number of nodes that are determined to be neighboring nodes of the node as a new node moves into the transmission range of the node, and the number of nodes that were neighbors of the node but are disconnected from the node by moving out of the transmission range. Cluster formation method proportional to. The method of claim 11, The mobility of the node, And a value calculated by dividing the total number of nodes changed by entering or exiting a transmission range of the node by the number of nodes around the previous period of the node. The method of claim 11, The mobility of the node, And a weighted average of the mobility of the current period of the node and the weighted average mobility of the previous period.

Images