CN110234146B - Distributed self-adaptive clustering method suitable for self-organizing network - Google Patents
Distributed self-adaptive clustering method suitable for self-organizing network Download PDFInfo
- Publication number
- CN110234146B CN110234146B CN201910442322.4A CN201910442322A CN110234146B CN 110234146 B CN110234146 B CN 110234146B CN 201910442322 A CN201910442322 A CN 201910442322A CN 110234146 B CN110234146 B CN 110234146B
- Authority
- CN
- China
- Prior art keywords
- node
- clustering
- cluster head
- cluster
- self
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W40/00—Communication routing or communication path finding
- H04W40/24—Connectivity information management, e.g. connectivity discovery or connectivity update
- H04W40/32—Connectivity information management, e.g. connectivity discovery or connectivity update for defining a routing cluster membership
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W40/00—Communication routing or communication path finding
- H04W40/34—Modification of an existing route
- H04W40/36—Modification of an existing route due to handover
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Abstract
The invention discloses a distributed self-adaptive clustering method suitable for a self-organizing network, and aims to provide a distributed self-adaptive clustering method capable of reducing network control overhead and improving network spectrum utilization rate and network energy efficiency. The invention is realized by the following technical scheme: in the self-organizing network, each network node constructs three functional modules of environment perception, self-learning and self-decision which periodically carry out control information interaction with the neighbor node according to a self-organizing network protocol, each network node receives the clustering control information of the neighbor node through a wireless receiving unit by utilizing the capability of a wireless transmission module to interact the control information with the neighbor node in the control period of the node, and reports the clustering control information of the neighbor node acquired by the wireless receiving unit to an environment perception functional module; and circularly executing a cluster head selection process, a cluster head replacement process, a cluster head switching process, a clustering fission process and a clustering merging process, and optimizing cluster head selection in real time.
Description
Technical Field
The invention relates to a distributed self-adaptive clustering method suitable for a self-organizing network.
Background
An ad hoc network (ad hoc) is a centerless multi-hop network formed spontaneously by a number of mobile nodes via a distributed network protocol. The self-organizing network can efficiently process the problems of network topology change, transmission link failure and the like, and has strong flexibility and survivability. As a distributed network, a mobile ad hoc network is an autonomous, multi-hop network, and the entire network has no fixed infrastructure and can provide intercommunication between terminals without utilizing or inconveniently utilizing existing network infrastructure (e.g., base stations, APs). Due to the limited transmission power and wireless coverage of the terminals, two terminals at a longer distance must perform packet forwarding by means of other intermediate nodes if communication is to be performed, so that a wireless multi-hop network is formed between the nodes. It is noted that, unlike multi-hop in general networks, multi-hop routing in wireless ad hoc networks is performed by common nodes in cooperation, not by dedicated routing devices. Unlike cellular networks with preset base stations, ad hoc networks offer all network nodes the ability to periodically interact with neighboring nodes for control information, allowing network nodes to build wireless communication systems anywhere and anytime without any preset infrastructure. In the self-organizing network, a network node communicates by means of a wireless transceiver and interacts control information and data service with a neighbor node. Since the network bandwidth provided by wireless communication is much smaller than that of wired communication, the network capacity provided by the ad hoc network is also much smaller than that of the wired network. In addition, ad hoc networks often operate in harsh field environments, and network nodes may only rely on limited energy sources such as batteries to supply power. Therefore, how to effectively utilize limited bandwidth resources and limited energy resources to improve the network spectrum utilization rate and the network energy efficiency of the ad hoc network is a concern of the clustering method.
In the wireless self-organizing network environment, wireless links among nodes and a network topology structure formed by the wireless links show the characteristic of dynamic change along with the factors of the position distribution and movement of the nodes, the change of channels and the like. The dynamic change of the topological structure makes the acquisition, management and maintenance of the link state information between the nodes difficult. Due to mutual interference of a hidden terminal, an exposed terminal, an intrusive terminal and the like between adjacent nodes, the state of a wireless link is difficult to determine, and link parameters such as bandwidth, time delay jitter and the like are difficult to acquire, update and maintain in time. Wireless ad hoc networks have two different hierarchical structures: planar structures and layered structures. In the planar structure, all nodes are equal in status, and therefore, the structure is also called a peer-to-peer structure. Each node in the planar structure needs to know the route to all other nodes. Maintaining this dynamic route requires a large amount of control information due to the mobility of the nodes. The larger the network size, the greater the overhead of route maintenance. The network scalability of the planar structure is poor. The hierarchy requires a corresponding clustering algorithm and cluster maintenance mechanism. In a self-organizing network (ad hoc) with a hierarchical structure, a cluster head is responsible for communication among member nodes in a cluster and communication among the member nodes of the cluster and other cluster member nodes, so how to dynamically select the most reasonable node to serve as the cluster head becomes a key problem of a clustering algorithm. In addition, because the nodes are added or removed at any time and any place, the network link is easy to be disconnected at any time, and the communication service quality is influenced. In the hierarchical structure, the cluster head nodes not only maintain the routing information reaching other cluster heads, but also maintain the communication with the members of the cluster and the members in the cluster. Therefore, the task of cluster head nodes is rather heavy, and due to the limited resources of wireless communication, a good clustering structure and a routing algorithm are needed to maintain the service quality of hierarchical ad hoc network routing and communication. The current typical clustering algorithm comprises a minimum ID clustering algorithm, a highest node degree clustering algorithm, a lowest mobility clustering algorithm, a clustering algorithm based on node mobility prediction and the like. And the minimum ID clustering algorithm selects the cluster head according to the node ID, and the algorithm is simple and easy to implement. When the node mobility is strong, the cluster head updating frequency is high, and the cluster maintenance cost is high. And selecting the nodes with the node degree larger than all uncovered adjacent nodes as cluster heads by the highest node degree clustering algorithm, and selecting the nodes with smaller IDs as the cluster heads when the node degrees are the same. The algorithm does not take into account load balancing, etc. The minimum mobility clustering algorithm elects a cluster head according to the weight of the mobility of the node. The cluster head calculation cost of the algorithm is high, and the problems of load balance and node energy loss are not considered. The clustering algorithm based on node mobility prediction is used for predicting the motion trend of a mobile node by learning the movement historical behavior of the mobile node so as to select a cluster head. The algorithm is not suitable for ad hoc networks that change paths often without habitual mobility.
Ad hoc networks allow each network node to communicate with any node within the communication coverage of the node. As shown in fig. 5, each node has a unique node ID, and ID numbers of 32 nodes are 1, 2, …,32, respectively. The communication coverage area of the node 12 includes 8 one-hop neighbor nodes, which form 8 communication links. Although the flexible networking feature provides convenience for data transmission, the complex physical topology of the network makes the overhead of control signaling for maintaining network topology information large, so that the control signaling may occupy too much wireless channel bandwidth resources. In addition, under the flat physical topology, each network node needs node information of the whole network, and each network node is required to broadcast control signaling frequently, so that the energy consumption of the network node is increased.
The main technical scheme in the prior art is as follows:
the highest node degree clustering method comprises the following steps: the method carries out cluster head election according to the neighbor list, the node with the highest number of neighbor nodes in the adjacent nodes becomes a cluster head, and one-hop neighbor nodes of the cluster head automatically become cluster members of the cluster head. The method is simple in rule and easy to operate, but the cluster scale is difficult to control, and network congestion and network resource utilization rate are easily reduced due to the fact that the number of cluster members is unbalanced.
The clustering method based on the node positions comprises the following steps: the method requires that a node estimates the geographical position of the node in the self-organizing network through positioning equipment such as a GPS (global positioning system) or algorithms such as network ranging, and the like, and utilizes neighbor list information to interactively obtain the approximate topological structure of the network, so that a clustering structure is constructed by utilizing a relatively balanced clustering algorithm. The method can effectively maintain the balance of the number of cluster members, but needs additional positioning equipment or a ranging algorithm, and brings higher complexity and control overhead.
The existing technical scheme is limited to a single clustering rule, and although the control overhead can be reduced to a certain extent, the performance requirements of self-organizing network clustering on multiple dimensions such as optimized cluster head selection, balanced clustering scale, stable network communication and the like cannot be effectively considered.
In an ad hoc network, dynamically changing network topology nodes have arbitrary mobility and can join and leave at any time. Meanwhile, the radio propagation conditions change rapidly, such as the change of the transmission power, the influence of factors such as terrain and weather, and the network topology formed by the mobile terminals through the radio channel may also change at any time. The wireless channel can provide a much lower network bandwidth than the wired channel. The multi-hop routing wireless node has limited transmission power and limited signal propagation range. The wireless node needs an intermediate node to assist in storing and forwarding so as to realize data communication between the source node and the destination node. The clustering method can effectively reduce the number of transmission links of the self-organizing network by constructing a clustering structure of a logical topology, greatly reduce network control overhead, reduce the frequency of broadcasting control signaling by cluster member nodes, save energy consumption of a large number of network nodes, prolong the survival time of the self-organizing network, and more efficiently support a routing protocol and a resource management protocol. However, the clustering structure may affect the stability and connectivity of network links, and easily cause network data flow congestion. Therefore, how to optimize cluster head selection, balance clustering scale and guarantee network connectivity in the clustering method of the self-organizing network is the problem to be solved by the invention.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides the distributed self-adaptive clustering method which can reduce network control overhead, improve the network spectrum utilization rate and the network energy efficiency and ensure network connectivity and is suitable for the self-organizing network.
The above object of the present invention can be achieved by the following measures, a distributed adaptive clustering method suitable for an ad hoc network, having the following technical features:
in the self-organizing network, each network node constructs three functional modules of environment perception, self learning and self decision which periodically carry out control information interaction with adjacent nodes according to a self-organizing network protocol, and executes a cluster head selection process, a cluster head replacement process, a cluster head switching process, a clustering fission process and a clustering merging process; each network node receives the clustering control information of the neighbor node through the wireless receiving unit by utilizing the capability of the wireless transmission module to interact the control information with the neighbor node in the control period of the node, and reports the clustering control information of the neighbor node acquired by the wireless receiving unit to the environment perception function module; the environment perception function module counts clustering control information of neighbor nodes in each control period, and reports neighbor list information, clustering state information, connected cluster head set information and cluster head weight information in a two-hop range of the node to the self-learning function module; the self-learning function module utilizes the clustering control information of the neighbor nodes obtained by the environment perception function module to refine the clustering control information of all the neighbor nodes into clustering state change information of a local network, refines the clustering state change information of the local network according to predefined clustering state transfer conditions and the definition of two cluster head sets, and concentrates the fussy node clustering information into clear network clustering information to be transmitted to the self-decision function module; the self-decision function module utilizes the clustering state change information of the local network extracted by the self-learning function to self-adaptively adjust the clustering state of the local node according to the current clustering role of the local node and the corresponding clustering state transfer condition, and circularly executes a cluster head selection process, a cluster head replacement process, a cluster head switching process, a clustering fission process and a clustering merging process, so that the network node can self-adaptively process the dynamic change of network topology and the change of the network clustering state according to the neighbor list information, and the cluster head selection is optimized in real time.
Compared with the prior art, the distributed self-adaptive clustering method applicable to the self-organizing network has the following beneficial effects.
The invention adopts three functional modules of environment perception, self learning and self decision to construct a clustering structure in a self-adaptive manner. Firstly, environment perception is a process that a network node monitors and stores clustering control information of a neighbor node in real time by utilizing the capability of exchanging the control information with the neighbor node; secondly, self-learning can utilize clustering control information of neighbor nodes obtained by an environment sensing function, and according to eight clustering state transfer conditions and the definition of two cluster head sets, clustering state change information of a local network is extracted, and fussy node clustering information is concentrated into clear network clustering information; and finally, self-decision can utilize the clustering state change information of the local network extracted by the self-learning function to self-adaptively adjust the clustering state of the node according to the clustering role of the node and the clustering state transfer condition. Therefore, each network node can dynamically maintain the clustering structure of the self-organizing network through a self-adaptive clustering method, multi-dimensional optimization targets of self-organizing network clustering in cluster head selection optimization, cluster scale equalization, network communication guarantee and the like are achieved, network control overhead is reduced, and network spectrum utilization rate and network energy efficiency are improved. Secondly, the invention adopts a cluster head selection process, a cluster head replacement process and a cluster head switching process; in the cluster head selection process, the node to be clustered can selectively add a cluster head with a higher cluster head weight value or promote the node as a cluster head node according to the candidate cluster head set; in the cluster head replacement process, when a network node finds that a neighbor node of the node and a set formed by the node are a true superset of a set formed by a certain neighbor cluster head and a neighbor node thereof according to neighbor list information in a two-hop range, the clustering role of the node can be adaptively promoted to be a cluster head node; in the cluster head switching process, when a cluster head node finds that all neighbor cluster heads of the node are communicated according to neighbor list information in a two-hop range and the union of all neighbor cluster heads and neighbor nodes thereof is a superset of a set formed by the node and the neighbor nodes of the node, the cluster role of the node can be set as a cluster member node in a self-adaptive manner and the cluster member of the node is informed to switch the cluster head; therefore, the invention enables the network node to adaptively process the dynamic change of the network topology according to the neighbor list information through the cluster head selection process, the cluster head replacement process and the cluster head switching process, thereby achieving the purpose of optimizing the cluster head selection in real time. Compared with the prior art, the method has the advantages that a better clustering structure can be obtained by optimizing a cluster head selection mechanism, so that the purposes of improving the network spectrum utilization rate and the network energy efficiency are achieved;
the invention designs the clustering fission flow by adopting the clustering scale with balanced node number, so that the network clustering can dynamically balance the clustering scale according to the maximum cluster member capacity limit pre-configured by the self-organizing network and the cluster members of each cluster head, when the cluster head node receives a new clustering application and the number of the current cluster members plus the number of the new cluster members reaches the preset maximum cluster member capacity limit, the cluster head node selects the node with the maximum neighbor node number from the cluster members of the cluster as a new cluster head, and informs the new cluster head and the cluster members of the cluster to carry out clustering fission operation through clustering control information, thereby achieving the effect that the clustering scale is self-adaptive to the network node density, and effectively reducing the probability of network data stream congestion. Therefore, compared with the prior art, the method can achieve the purpose of balancing the clustering scale without additional positioning equipment or a ranging algorithm, and prevent the network data flow from congestion while reducing the number of network transmission links.
The invention designs a clustering merging process by adopting a clustering structure for guaranteeing connectivity, network clustering can dynamically guarantee the connectivity of the clustering structure of the self-organizing network according to a connected cluster head set, when the minimum node ID in the connected cluster head set of a certain neighbor node is not equal to the minimum node ID in the connected cluster head set of the node, the current clustering structure of the self-organizing network destroys the connectivity of each network node in a physical topological structure, the node sets a clustering role and a clustering state as a cluster head node and a clustering merging state respectively, and indicates the neighbor node to promote the clustering role to the cluster head node through clustering control information, and restores the connectivity between network clusters, so that the distributed self-adaptive clustering method can guarantee the connectivity of the network clustering structure in real time, thereby achieving the purpose of keeping network data multi-hop relay transmission while reducing network control overhead. Therefore, compared with the prior art, the method and the device can adaptively promote the nodes maintaining the network connectivity to the cluster head nodes, thereby effectively ensuring the connectivity of the self-organizing network and achieving the purpose of excavating the multi-hop relay transmission gain.
The invention is applied to the fields of aviation, explosion prevention, disaster relief, environment, medical treatment, health care, home furnishing, industry, commerce and the like.
Drawings
FIG. 1 is a functional block diagram of the distributed adaptive clustering of the present invention.
Fig. 2 is a schematic diagram of a clustering structure of an ad hoc network.
FIG. 3 is a process flow diagram of the context awareness and self-learning function module of FIG. 1.
Fig. 4 is a process flow diagram of the self-decision function module of fig. 1.
Fig. 5 is a schematic diagram of the physical topology of an ad hoc network.
Detailed Description
See fig. 1. According to the invention, in the self-organizing network, each network node constructs three functional modules of environment perception, self-learning and self-decision which periodically carry out control information interaction with adjacent nodes according to a self-organizing network protocol, and executes a cluster head selection process, a cluster head replacement process, a cluster head switching process, a clustering fission process and a clustering merging process; each network node receives the clustering control information of the neighbor node through the wireless receiving unit by utilizing the capability of the wireless transmission module to interact the control information with the neighbor node in the control period of the node, and reports the clustering control information of the neighbor node acquired by the wireless receiving unit to the environment perception function module; the environment perception function module counts clustering control information of neighbor nodes in each control period, and reports neighbor list information, clustering state information, connected cluster head set information and cluster head weight information in a two-hop range of the node to the self-learning function module; the self-learning function module utilizes the clustering control information of the neighbor nodes obtained by the environment perception function module to refine the clustering control information of all the neighbor nodes into clustering state change information of a local network, refines the clustering state change information of the local network according to predefined clustering state transfer conditions and the definition of two cluster head sets, and concentrates the fussy node clustering information into clear network clustering information to be transmitted to the self-decision function module; the self-decision function module utilizes the clustering state change information of the local network extracted by the self-learning function to self-adaptively adjust the clustering state of the local node according to the current clustering role of the local node and the corresponding clustering state transfer condition, and circularly executes a cluster head selection process, a cluster head replacement process, a cluster head switching process, a clustering fission process and a clustering merging process, so that the network node can self-adaptively process the dynamic change of network topology and the change of the network clustering state according to the neighbor list information, and the cluster head selection is optimized in real time.
In the cluster head selection process, the node to be clustered can selectively add a cluster head with a higher cluster head weight value or promote the node as a cluster head node according to the candidate cluster head set; in the cluster head replacement process, when a network node finds that a neighbor node of the node and a set formed by the node are a true superset of a set formed by a certain neighbor cluster head and a neighbor node thereof according to neighbor list information in a two-hop range, the clustering role of the node can be adaptively promoted to be a cluster head node; in the cluster head switching process, when the cluster head node finds that all the neighbor cluster heads of the node are communicated and the union of all the neighbor cluster heads and the neighbor nodes thereof is a superset of the set formed by the node and the neighbor nodes of the node according to the neighbor list information in the two-hop range, the clustering role of the node can be set as a cluster member node in a self-adaptive manner and the cluster member of the node is informed to switch the cluster head.
In the clustering fission process, when a cluster head node receives a new clustering application and the sum of the number of current cluster members and the number of new cluster members reaches the preset maximum cluster member capacity limit, the cluster head node selects a node with the largest number of neighbor nodes from the cluster members of the cluster as a new cluster head, and informs the new cluster head and the cluster members of the cluster to perform clustering fission operation through clustering control information; in the clustering merging process, when the minimum node ID in a connected cluster head set of a certain neighbor node is not equal to the minimum node ID in the connected cluster head set of the node, the node sets a clustering role and a clustering state as a cluster head node and a clustering merging state respectively, and indicates the neighbor node to promote the clustering role to be the cluster head node through clustering control information; the self-decision function module constructs the clustering control information of the node and inputs the clustering control information into the wireless transmitting unit of the wireless transmission module for broadcast transmission.
In an alternative embodiment, the wireless transmission module receives the clustering control information of the neighbor node through the wireless receiving unit. Each network node reports the clustering control information of the neighbor node acquired by the wireless receiving unit to the environment perception function module in the control period of the node. The clustering control information comprises cluster head weight values of the neighbor nodes, clustering states, target node IDs, a connected cluster head set and a neighbor list, wherein the cluster head weight values represent the degree that the neighbor nodes are suitable for serving as cluster head nodes, and the nodes with higher cluster head weight values are more suitable for serving as cluster head nodes in the self-organizing network; the clustering state comprises seven states of applying for clustering, cluster members, cluster heads, cluster head replacement, cluster head switching, cluster fission and cluster combination, and the clustering state and the target node ID indicate the clustering role of the neighbor node, the clustering operation being executed by the neighbor node and the target node matched with the neighbor node to complete the clustering operation; the connected cluster head set comprises all cluster head nodes which can be connected with the neighbor nodes and can be used for judging the connectivity of the clustering structure in the self-organizing network; the neighbor list contains all the neighbor nodes of the neighbor node, indicating the local topology of the neighbor node in the ad hoc network.
The environment perception function module classifies and stores the acquired clustering control information of all neighbor nodes according to four categories of a neighbor list, a clustering state, a connected cluster head set and a cluster head weight, wherein the neighbor list information indicates a physical topological structure in a two-hop range of the node, which is beneficial for the node to carry out cluster head optimization selection and judge the connection relation between the node and a specified cluster head; the clustering state statistics is beneficial to the analysis of the clustering state change of the local network by the node on the clustering roles of all neighbor nodes of the node and the clustering operation being executed; the connected cluster head set summarizes a union set of cluster heads which can be connected by all neighbor nodes of the node and is beneficial to the node to analyze the connectivity of the clustering structure of the local network; the cluster head weight values rank all the neighbor nodes of the node according to the degree suitable for serving as cluster heads, and optimization and selection of the cluster heads are facilitated. The environment perception function module counts clustering control information of neighbor nodes and reports neighbor list information, clustering state information, connected cluster head set information and cluster head weight information in a two-hop range of the node to the self-learning function module. The self-learning function module is used for refining the clustering control information of all the neighbor nodes into clustering state change information of the local network. The self-learning function module analyzes the clustering state change information of the local network in the two-hop range of the node according to predefined cluster head replacement conditions, successful cluster entering conditions, cluster head switching conditions, clustering fission conditions, neighbor merging states, whether the node is fissile, whether the cluster head is switched, clustering merging conditions, eight state transfer conditions of a connected cluster head set and a candidate cluster head set and two cluster head sets. And the self-learning function module transmits the eight state transition conditions and the analysis results of the two cluster head sets to the self-decision function module. And the self-decision function module extracts a corresponding clustering state transfer condition according to the current clustering role of the node and adaptively selects an optimal clustering state. The self-decision function module constructs the clustering control information of the node and inputs the clustering control information into the wireless transmitting unit of the wireless transmission module for broadcast transmission, so that the purpose that the clustering structure can be self-adaptive to the network clustering state change is achieved.
See fig. 2. In the ad hoc network, the network nodes 10, 12, 14, 15, 18, 19, 21, 23 are defined as cluster head nodes by a distributed adaptive clustering method, and other network nodes adaptively join neighboring clusters; the cluster member-cluster head link is a single-hop link in a cluster and maintains data transmission in the cluster, and the inter-cluster head link is a multi-hop link between clusters and maintains data transmission between the clusters. Each network node periodically performs control information interaction with neighboring nodes according to a self-organizing network protocol, and becomes a cluster head node or a cluster member node in a self-adaptive manner, and the cluster head node maintains and manages routing information in and among clusters, so that a clustering structure of the self-organizing network is realized.
The clustering control information comprises a source node ID, a cluster head weight, a clustering state, a destination node ID, a connected cluster head set and a neighbor list in a name domain, wherein the content of the source node ID is the ID of a sending node; the contents of the cluster head weight are the normalized cluster head weight of the sending node.
See table 1. An example table of clustering control information is given in the table.
TABLE 1 example table of clustering control information
TABLE 1 example table of clustering control information
The clustering state comprises 7 states with serial numbers of 0, 1,. And 6; the number 0 clustering state is an application for entering a cluster, and the ID of a target node is an appointed cluster head ID; the No. 1 clustering state is a cluster member, and the ID of a target node is the ID of a cluster head of the cluster; the number 2 clustering state is a cluster head, and the target node ID is a whole network broadcast ID; the No. 3 clustering state is cluster head replacement, and the ID of the target node is the ID of the replaced cluster head; the No. 4 clustering state is cluster head switching, and the ID of a target node is an intra-cluster broadcast ID; the No. 5 clustering state is clustering fission, and the ID of the target node is the ID of a new cluster head; the No. 6 clustering state is clustering combination, and the destination node ID is the cluster member ID of the specified adjacent cluster. The content of the connected cluster head set comprises the number (Nc) of connected cluster heads and the IDs of the Nc connected cluster heads; the contents of the neighbor list include the number of neighbor nodes (Nb) and the IDs of the Nb neighbor nodes.
See fig. 3. In the processing flow of the environment sensing and self-learning function module, the environment sensing function module 400 divides the clustering control information content of the neighbor nodes reported by the wireless receiving unit into four categories, namely a neighbor list, a clustering state, a connected cluster head set and a cluster head weight, and then inputs the information into the self-learning function module 401 and transfers to execute the self-learning function; the self-learning function module 401 extracts neighbor list information, clustering state information, connected cluster head information and cluster head weight information, and analyzes clustering state change information of a local network in a two-hop range of a local node according to predefined cluster head replacement conditions, successful cluster entering conditions, cluster head switching conditions, clustering fission conditions, neighbor merging states, whether the local cluster is cracked, whether the cluster head is switched, clustering merging conditions, eight clustering state transition conditions including a connected cluster head set and a candidate cluster head set, and two cluster head sets; the analysis steps of the eight clustering state transition conditions and the two cluster head sets are as follows:
and step 410, the self-learning function module updates the connected cluster head set of the node according to the clustering state and the connected cluster head set. If the node is a cluster head, the node takes a union set of connected cluster heads of all cluster head nodes received in the current control period; otherwise, the cluster member replaces the connected cluster head set of the cluster head received in the current control period with the connected cluster head set of the node;
See fig. 4. In the processing flow of the self-decision function module, the self-decision function module takes the clustering role of the node as a starting point to perform a clustering state transfer flow and output the clustering state of the node.
500, starting clustering state transfer by the self-decision function module according to the clustering role of the node; if the node finishes initialization loading but never sends clustering control information, the clustering role of the node is a node to be clustered, and the step 501 is carried out to start a self-decision function; if the node has sent the clustering control information and the clustering state of the latest clustering control information is an application cluster, the clustering role of the node is the application cluster node, and the step 502 is switched to start the self-decision function; if the node has sent the clustering control information and the clustering state of the latest clustering control information is a cluster member, the clustering role of the node is a cluster member node, and the step 503 is switched to start a self-decision function; if the three conditions are not met, the clustering role of the node is a cluster head node, and the step 504 is switched to start a self-decision function;
otherwise, go to step 506 to determine whether the candidate cluster head set is non-empty;
509, removing the currently selected cluster head from the candidate cluster head set by the self-decision function module, updating the candidate cluster head set, and turning to 506 to judge whether the candidate cluster head set is non-empty;
523, the self-decision function module triggers the cluster fission, plans to notify a new cluster head and cluster members of the cluster to perform cluster fission, sets a destination node ID of the cluster control information of the node as the new cluster head ID, and goes to 532 to update the clustering state;
See fig. 5. In the self-organizing network, each node has a unique node ID, and the ID numbers of 32 nodes are 1, 2, … and 32 respectively. Each network node and the neighbor nodes in the communication coverage range form a communication link. The communication coverage area of the node 12 includes 8 one-hop neighbor nodes, which form 8 communication links.
The above description is only exemplary of the present invention and should not be taken as limiting the invention, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. A distributed self-adaptive clustering method suitable for a self-organizing network has the following technical characteristics: in the self-organizing network, each network node constructs three functional modules of environment perception, self learning and self decision which periodically carry out control information interaction with adjacent nodes according to a self-organizing network protocol, and executes a cluster head selection process, a cluster head replacement process, a cluster head switching process, a clustering fission process and a clustering merging process; each network node receives clustering control information of a neighbor node through a wireless receiving unit by using the capability of a wireless transmission module for exchanging the control information with the neighbor node in the control period of the node, wherein the clustering control information comprises a source node ID, a clustering state, a destination node ID, a connected cluster head set and a neighbor list, and reports the clustering control information of the neighbor node acquired by the wireless receiving unit to an environment sensing function module; the environment perception function module classifies and stores clustering control information of all neighbor nodes according to four categories of a neighbor list, a clustering state, a connected cluster head set and cluster head weight, the clustering control information of the neighbor nodes is counted in each control period, a physical topological structure in a two-hop range of the node is constructed from the neighbor list information, clustering roles and executing clustering operations of all the neighbor nodes are counted from the clustering state information, a union set of cluster heads which can be communicated by all the neighbor nodes is summarized from the connected cluster head set information, all the neighbor nodes of the node are sorted according to the cluster head weight information and the degree suitable for serving as cluster heads, and the neighbor list information, the clustering state information, the connected cluster head set information and the cluster head weight information in the two-hop range of the node are reported to the self-learning function module; the self-learning function module extracts the clustering control information of all the neighbor nodes into clustering state change information of a local network comprising judgment results of eight state transfer conditions and update results of two cluster head sets according to predefined cluster head replacement conditions, successful clustering conditions, cluster head switching conditions, clustering fission conditions, neighbor merging states, whether the cluster head is split or not, and clustering merging conditions, and connected cluster head set update rules and candidate cluster head set update rules of the node, and transmits the clustering state change information of the local network to the self-decision function module; the self-decision function module circularly executes a cluster head selection process, a cluster head replacement process, a cluster head switching process, a cluster fission process and a cluster merging process according to the current cluster role of the node and the corresponding cluster state transfer condition by utilizing the cluster state change information of the local network extracted by the self-learning function, adjusts the cluster state of the node and updates the cluster head selection.
2. The distributed adaptive clustering method for the self-organizing network according to claim 1, wherein: in the cluster head selection process, the node to be clustered can selectively add a cluster head with a higher cluster head weight value or promote the node as a cluster head node according to the candidate cluster head set; in the cluster head replacement process, when a network node finds that a neighbor node of the node and a set formed by the node are a true superset of a set formed by a certain neighbor cluster head and a neighbor node thereof according to neighbor list information in a two-hop range, the clustering role of the node can be adaptively promoted to be a cluster head node; in the cluster head switching process, when the cluster head node discovers that all neighbor cluster heads of the node are communicated and the union of all neighbor cluster heads and neighbor nodes thereof is a superset of a set formed by the node and neighbor nodes of the node according to neighbor list information in a two-hop range, the clustering role of the node can be set as a cluster member node in a self-adaptive manner and the cluster member of the node is informed to switch the cluster head.
3. The distributed adaptive clustering method for an ad-hoc network according to claim 1 wherein: the clustering control information comprises cluster head weight values of neighbor nodes, clustering states and target node IDs, a connected cluster head set and a neighbor list, wherein the cluster head weight values represent the degree that the neighbor nodes are suitable for serving as cluster head nodes, and the nodes with higher cluster head weight values are more suitable for serving as cluster head nodes in the self-organizing network; the clustering state comprises a cluster application state, a cluster member state, a cluster head replacement state, a cluster head switching state, a cluster fission state and a clustering merging state, and the clustering state and the target node ID indicate that the clustering role of the neighbor node and the clustering operation which is being executed by the neighbor node and the target node which is matched with the neighbor node to complete the clustering operation; the connected cluster head set comprises all cluster head nodes which can be connected by the neighbor nodes, so as to judge the connectivity of the cluster structure in the self-organizing network; the neighbor list contains all neighbor nodes of the neighbor nodes, and indicates the local topology of the neighbor nodes in the self-organizing network.
4. The distributed adaptive clustering method for the self-organizing network according to claim 1, wherein: the self-learning function module predefines eight state transfer conditions including a cluster head replacement condition, a successful cluster entering condition, a cluster head switching condition, a cluster fission condition, a neighbor merging state, whether the cluster is fission or not, whether the cluster head is switched or not, a cluster merging condition, a connected cluster head set and a candidate cluster head set, and two cluster head sets to analyze the cluster state change information of the local network in the two-hop range of the node, wherein the cluster head replacement condition is that a neighbor node of the node and a set formed by the node are true supersets of a set formed by a certain neighbor cluster head and a neighbor node thereof; the condition of successful clustering is that the clustering state of the cluster head appointed by the node is the cluster head, and the neighbor list information of the cluster head comprises the ID of the node; the cluster head switching condition is that all the neighbor cluster heads of the node are communicated, and the union of all the neighbor cluster heads and the neighbor nodes thereof is a superset of the set formed by the node and the neighbor nodes of the node; the cluster fission condition is that the node is a cluster head and receives a new cluster entering application, and the sum of the number of the current cluster members and the number of the new cluster entering members reaches the preset maximum cluster member capacity limit; the neighbor merging state is a clustering state of cluster heads with neighbor clusters in the current control period and is clustering merging; the cluster fission condition is that the node receives the clustering control information of the cluster head and the clustering state of the cluster head is cluster fission; the cluster head switching condition of the cluster is that the node receives the clustering control information of the cluster head and the clustering state of the cluster head is cluster head switching; the clustering and merging condition is that the minimum node ID in the connected cluster head set of the adjacent node in the current control period is not equal to the minimum node ID in the connected cluster head set of the node; the updating rule of the connected cluster head set of the node is that if the node is a cluster head, the node takes a union set of the connected cluster head sets of all cluster head nodes received in the current control period; otherwise, the cluster member replaces the connected cluster head set of the cluster head received in the current control period with the connected cluster head set of the node; the updating rule of the candidate cluster head set of the node is to fill all neighbor node IDs with cluster heads in the clustering state into the candidate cluster head set and record the weight of the cluster heads.
5. The distributed adaptive clustering method for the self-organizing network according to claim 1, wherein: the self-learning function module transmits the total eight state transition conditions and the analysis results of the two cluster head sets to the self-decision function module; the self-decision function module extracts corresponding clustering state transfer conditions according to the current clustering role of the node, adaptively selects the optimal clustering state, constructs clustering control information of the node and inputs the clustering control information into a wireless transmitting unit of the wireless transmission module for broadcast transmission, and therefore the purpose that a clustering structure can be adaptive to network clustering state changes is achieved.
6. The distributed adaptive clustering method for an ad-hoc network according to claim 1 wherein: the clustering state comprises 7 states with serial numbers of 0, 1,. And 6; the number 0 clustering state is a cluster application state, and the destination node ID is a designated cluster head ID; the No. 1 clustering state is a cluster member, and the ID of a target node is the ID of a cluster head of the cluster; the number 2 clustering state is a cluster head, and the target node ID is a whole network broadcast ID; the No. 3 clustering state is cluster head replacement, and the ID of the target node is the ID of the replaced cluster head; the No. 4 clustering state is cluster head switching, and the ID of a target node is an intra-cluster broadcast ID; the No. 5 clustering state is clustering fission, and the ID of the target node is the ID of a new cluster head; the No. 6 clustering state is clustering combination, and the destination node ID is the cluster member ID of the specified adjacent cluster.
7. The distributed adaptive clustering method for the self-organizing network according to claim 1, wherein: in the processing flow of the environment perception and self-learning function module, the environment perception function module divides the clustering control information content of the neighbor nodes reported by the wireless receiving unit into four categories of a neighbor list, a clustering state, a communicated cluster head set and a cluster head weight, and then inputs the information into the self-learning function module and transfers the information to the self-learning function; the self-learning function module extracts neighbor list information, clustering state information, connected cluster head information and cluster head weight information, and analyzes clustering state change information of a local network in a two-hop range of the node according to predefined cluster head replacement conditions, successful cluster entering conditions, cluster head switching conditions, clustering fission conditions, neighbor merging states, whether the cluster is cracked, whether the cluster head is switched, clustering merging conditions, connected cluster head sets, candidate cluster head set clustering state transfer conditions and two cluster head sets.
8. The distributed adaptive clustering method for the self-organizing network according to claim 1, wherein: the self-decision function module starts clustering state transfer according to the clustering role of the node; if the node finishes initialization loading but never sends clustering control information, the clustering role of the node is a node to be clustered, and a self-decision function is started; if the node has sent the clustering control information and the clustering state of the latest clustering control information is an application cluster, the clustering role of the node is the application cluster node, and a self-decision function is started; if the node has sent the clustering control information and the clustering state of the latest clustering control information is a cluster member, the clustering role of the node is a cluster member node, and a self-decision function is started; if the three conditions are not met, the clustering role of the node is a cluster head node, and a self-decision function is started.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910442322.4A CN110234146B (en) | 2019-05-25 | 2019-05-25 | Distributed self-adaptive clustering method suitable for self-organizing network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910442322.4A CN110234146B (en) | 2019-05-25 | 2019-05-25 | Distributed self-adaptive clustering method suitable for self-organizing network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110234146A CN110234146A (en) | 2019-09-13 |
| CN110234146B true CN110234146B (en) | 2022-12-13 |
Family
ID=67861130
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910442322.4A Active CN110234146B (en) | 2019-05-25 | 2019-05-25 | Distributed self-adaptive clustering method suitable for self-organizing network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110234146B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110995333B (en) * | 2019-11-29 | 2020-12-01 | 北京航空航天大学 | Cluster QoS route design method |
| CN110943920B (en) * | 2019-12-12 | 2020-10-27 | 电子科技大学 | Distributed cluster head election method based on wireless self-organizing network |
| CN111212455B (en) * | 2020-01-16 | 2022-02-22 | 中国人民解放军陆军工程大学 | Dynamic clustering system and method for wireless self-organizing network |
| CN111356198B (en) * | 2020-02-10 | 2022-02-08 | 西安电子科技大学 | Clustering cross-layer communication processing method and system based on geographic position and Q learning |
| CN111555798B (en) * | 2020-05-09 | 2022-02-11 | 中国航空无线电电子研究所 | Multi-platform aviation network clustering method |
| CN112822653B (en) * | 2020-12-30 | 2023-03-31 | 国网甘肃省电力公司信息通信公司 | Clustering routing method in wireless sensor network |
| CN112822752B (en) * | 2021-01-07 | 2022-10-25 | 西安电子科技大学 | Route establishing method and system for unmanned aerial vehicle self-organizing network |
| CN114356240B (en) * | 2021-12-30 | 2023-09-26 | 重庆市紫建电子股份有限公司 | Method for rapidly and efficiently checking capacity flow |
| CN114585047B (en) * | 2022-02-22 | 2024-02-23 | 西安电子科技大学 | Wireless self-organizing network self-adaptive clustering method based on clustering management information |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105744566A (en) * | 2016-01-28 | 2016-07-06 | 北京邮电大学 | Self-adaptive on-demand weighted clustering method based on perceptron |
| CN108712758A (en) * | 2018-04-27 | 2018-10-26 | 中山大学 | It is a kind of to be applied to dynamic stability cluster-dividing method safe in vehicular ad hoc network network |
| CN109041127A (en) * | 2018-09-26 | 2018-12-18 | 北航（四川）西部国际创新港科技有限公司 | The adaptive stabilizing cluster-dividing method and system of load balancing suitable for highly dynamic radio network |
-
2019
- 2019-05-25 CN CN201910442322.4A patent/CN110234146B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105744566A (en) * | 2016-01-28 | 2016-07-06 | 北京邮电大学 | Self-adaptive on-demand weighted clustering method based on perceptron |
| CN108712758A (en) * | 2018-04-27 | 2018-10-26 | 中山大学 | It is a kind of to be applied to dynamic stability cluster-dividing method safe in vehicular ad hoc network network |
| CN109041127A (en) * | 2018-09-26 | 2018-12-18 | 北航（四川）西部国际创新港科技有限公司 | The adaptive stabilizing cluster-dividing method and system of load balancing suitable for highly dynamic radio network |
Non-Patent Citations (4)
| Title |
|---|
| "Hybrid_Self-Organized_Clustering_Scheme_for_Drone_Based_Cognitive_Internet_of_Things";FAROOQ AFTAB et al;《IEEE Access》;20190509;全文 * |
| 基于分簇结构的移动自组织网络接入控制协议关键技术研究;刘强等;《北京交通大学学报》;20180415(第02期);全文 * |
| 基于加权分簇的移动无线传感器网络拓扑维护算法;姚玉坤等;《微电子学与计算机》;20180805(第08期);全文 * |
| 基于稳定分簇的移动自组织网络路由协议;黄金科等;《北京航空航天大学学报》;20161130(第11期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110234146A (en) | 2019-09-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110234146B (en) | Distributed self-adaptive clustering method suitable for self-organizing network | |
| CN108600942A (en) | A kind of method for routing of unmanned plane ad hoc network | |
| CN101808382B (en) | Data communication method of wireless sensor network | |
| CN111263419B (en) | Unmanned aerial vehicle-based dynamic routing method for stereo heterogeneous network in emergency scene | |
| CN101883048A (en) | Routing method of multi-dimensional network | |
| Bouzid et al. | Efficient routing protocol for wireless sensor network based on reinforcement learning | |
| US9980317B2 (en) | Protocol stack for a multi-waveform AD hoc network | |
| CN111526557B (en) | Wireless ad hoc network routing information acquisition method | |
| CN116545923A (en) | Distributed routing protocol method suitable for large-scale unmanned aerial vehicle cluster network | |
| Osman et al. | A survey of clustering algorithms for cognitive radio ad hoc networks | |
| ECE | A survey on energy efficient routing in wireless sensor networks | |
| CN105007570B (en) | A kind of mobile ad-hoc network jamming control method based on OLSR agreement | |
| CN102083164A (en) | Method for opportunistic routing of wireless sensor network (WSN) based on energy perception | |
| Abdollahi et al. | Dynamic routing protocol selection in multi-hop device-to-device wireless networks | |
| Faheem et al. | Energy based efficiency evaluation of cluster-based routing protocols for wireless sensor networks (WSNs) | |
| CN110121185A (en) | A kind of power distribution communication network route optimization method | |
| Savyanavar et al. | Survey of clustering schemes in mobile ad hoc networks | |
| Abolhasan et al. | Joint Mobile Node Participation and Multihop Routing for Emerging Open Radio-Based Intelligent Transportation System | |
| Tan et al. | A distributed and dynamic data gathering protocol for sensor networks | |
| Wang et al. | IMCR: Influence maximisation-based cluster routing algorithm for SDVN | |
| CN108040346B (en) | Automatic relay method for high-dynamic ground-air network | |
| CN112423356A (en) | Unmanned equipment cluster AODV routing method based on energy balance | |
| Kuppusamy et al. | A survey on end-to-end performance metrics of QoS-aware multicast routing protocols for mobile ad hoc networks | |
| Saqib et al. | Distributed SDN based network state aware architecture for flying ad-hoc network | |
| Hajela et al. | LCSD: Leveling clustering and sectoring with dissemination nodes to perform energy efficient routing in mobile cognitive wireless sensor networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |