CN101729331B - Clustering method and device, routing method and device of cluster head and base station - Google Patents

Clustering method and device, routing method and device of cluster head and base station Download PDF

Info

Publication number
CN101729331B
CN101729331B CN2008101728400A CN200810172840A CN101729331B CN 101729331 B CN101729331 B CN 101729331B CN 2008101728400 A CN2008101728400 A CN 2008101728400A CN 200810172840 A CN200810172840 A CN 200810172840A CN 101729331 B CN101729331 B CN 101729331B
Authority
CN
China
Prior art keywords
node
described
bunch
head
base station
Prior art date
Application number
CN2008101728400A
Other languages
Chinese (zh)
Other versions
CN101729331A (en
Inventor
闫斌
周小佳
谭勇
耿军涛
Original Assignee
华为技术有限公司
电子科技大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司, 电子科技大学 filed Critical 华为技术有限公司
Priority to CN2008101728400A priority Critical patent/CN101729331B/en
Publication of CN101729331A publication Critical patent/CN101729331A/en
Application granted granted Critical
Publication of CN101729331B publication Critical patent/CN101729331B/en

Links

Classifications

    • YGENERAL 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
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02DCLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THIR OWN ENERGY USE
    • Y02D70/00Techniques for reducing energy consumption in wireless communication networks
    • Y02D70/30Power-based selection of communication route or path

Abstract

The embodiment of the invention discloses clustering method and device, routing method and device of a cluster head and a base station, belonging to the technical field of wireless sensor networks. The clustering method comprises the steps of: carrying out logic region division on a target region and determining a region ID of the target region; selecting one agent node in each region, wherein a node in an agent node notification region is used as an agent; sending node state information by the node in an agent node receiving region; and selecting one node in the region by the agent node according to the information as the cluster head, wherein the region becomes one cluster. The routing method comprises the steps of: selecting one base station in which a link has lower load when transmitting data from two base stations; determining the gradient of the cluster head according to the region ID in the cluster to transmit the data to the selected base station along the diminishing direction of gradient; selecting N cluster heads with gradient less than that of a source cluster head as a next hop cluster head to transmit the data along the direction of the selected base station by the source cluster head; and selecting N cluster heads with gradient less than that of a middle hop cluster head as a next ho cluster head to transmit the data along the direction of the selected base by the middle hop cluster head.

Description

Method for routing and the device of clustering method and device, bunch head and base station

Technical field

The present invention relates to the wireless sensor network technology field, particularly method for routing and the device of a kind of clustering method and device, bunch head and base station.

Background technology

Wireless sensor network (WSN, Wireless Sensor Network) be a kind of multi-hop mobility peer-to-peer network that combines sensor technology, embedded computing technique, distributed information processing and the communication technology, generally by one or more base stations (Sink node) and be deployed in the monitored area in a large number, the wireless network node of being furnished with various kinds of sensors constitutes.Each node cost is low, and power consumption is little, has certain computing ability, communication capacity.WSN adopts the mode dynamic group net of radio communication, various environment in monitoring in real time collaboratively, perception and the collection network distributed areas or the information of monitoring target, and these information are handled, obtain detailed and information accurately, and be ultimately delivered to the user who needs these information.Though individual node image data and inaccuracy, also unreliable, but great deal of nodes cooperates mutually and forms the network configuration of high unity, accuracy and the reliability of operation of data acquisition have been improved, can be deployed in the enemy-occupied area, disaster district, the inaccessible special area of manpower such as nuclear reactor etc. carry out data acquisition, transmission etc., have the incomparable characteristic of other networks, be widely used in national defence, environmental monitoring, Smart Home etc.

But wireless sensor network is different from the conventional wireless MANET, its node adopts not replaceable (can not fill), the powered battery of finite energy usually, computing capability, disposal ability, storage capacity, communication capacity are all limited, and interstitial content often very huge (thousands of even up to ten thousand), node distributes more intensive, because environmental impact and depleted of energy, node is easier to break down, therefore, traditional wireless network protocol can't directly be applied in the wireless sensor network.

In wireless sensor network technology, network topology control can generate the good network topological structure automatically, for improve the network energy service efficiency, prolong network lifecycle, improve router efficiency, to reduce a lot of aspects such as network delay and target localization all significant.Main topological control technology has levels at present, and type is topological to be controlled.The control of level type topology is a plurality of bunches with node division, give different functional roles to node, form the backbone network handling and transmit data by leader cluster node, other non-backbone nodes can the temporary close communication module, enters resting state to save energy.Bunch be the set with certain related network node, each bunch is made up of member (Cluster Member) in a bunch of head (CH, Cluster Head) and a plurality of bunches.Member node is only sent data to a bunch hair, and bunch head is collected data and the service data blending algorithm of member node, forms new packet, by certain routing policy data is sent to aggregation node (Sink node) then.Bunch head by respective algorithms bunch in elect, be in charge of or control whole bunch operational mode, coordinate bunch member's work.The sub-clustering topological structure is convenient to management, is conducive to the application of distributed algorithm, can make fast reaction to system change, has extensibility preferably, is fit to large scale network.Most typical bunch of head select and cluster-dividing method in, whether each node determines elected bunch head according to random number is autonomous, shortcoming is that every bunch head of taking turns generation not have definite quantity and position, can not guarantee that a bunch head evenly distributes.For solving this deficiency of this method, prior art improve one's methods that each node need report to the base station to self geographical position and current energy in one, so just can guarantee that a bunch head evenly distributes.But it is frequent to do node communication like this, and energy consumption is bigger.In addition, dump energy is the important state information of node, also is bunch important parameter of head election contest.Therefore, improving one's methods of prior art two selected bunch head based on residue energy of node, and realized the even distribution of node.But state information such as each residue energy of node in the cluster process need collecting zone of this method, the energy consumption spending is bigger, and the selection of bunch head only considered residue energy of node information, and the average communication distance is big and energy consumption is higher in causing easily bunch.

After prior art is analyzed, the inventor finds: though dump energy status in a bunch process of selection is important, but under complex environment, seem also insufficient with dump energy for bunch most important condition of head election contest, other factors are very important equally such as the geographical location information of node.When the maximum node of dump energy concentrates on some zones, perhaps relatively isolated certain zone that is distributed in, perhaps the node of a plurality of high dump energies is apart nearer, and then bunch member is far away to the distance of bunch head, what form bunch is not optimum, may cause more energy consumption on the contrary.

In addition, the inventor also finds: existing cluster algorithm can not carry out effective monitoring to bunch head, can not ensure the robustness of cluster management, if bunch head is because of certain reason sudden failure then can cause in various degree influence to network.

In order to guarantee the route robustness of WSN, prior art proposes to set up the direction staggered-mesh that the cost gradient from the source node to the destination node is successively decreased, and each is jumped with a plurality of node for data forwarding, has guaranteed the reliability of data.But the inventor finds: the energy consumption cost of its maintain communications gradient is higher.In routing procedure, realize route according to global information, node needs continuous exchange message to upgrade routing table, however the node energy of WSN is limited, so this mode is not suitable for WSN.

Summary of the invention

For in order efficiently to utilize limited Internet resources, compress unnecessary spending and energy-conservation as far as possible, the embodiment of the invention provides method for routing and the device of a kind of clustering method and device, bunch head and base station.Described technical scheme is as follows:

A kind of clustering method may further comprise the steps:

According to the communication coverage of the different discrete power grades of two base station communication energy in the target area logic region being carried out in described target area divides and determines its regional ID, described two base stations are positioned at the two ends of described target area homonymy, and described regional ID is logic region ID;

Select an agent node in each described zone, described agent node notification area interior nodes oneself becomes the agency, and described zone is logic region;

The agent node in described zone receives the described regional interior nodes state information that described regional interior nodes sends;

The agent node in described zone is selected a node according to described state information in described zone be a bunch head, and described zone becomes one bunch.

The burden of bunch head has been shared in the introducing of agency mechanism in the embodiment of the invention, has prolonged the duty cycle of bunch head, and regional interior nodes management and data processing efficiency are improved.

A kind of method for routing according in the network of described clustering method cluster bunch of head and base station may further comprise the steps:

A lower base station of link load when from two base stations, selecting data are transmitted;

According to bunch in described regional ID determine the gradient of described bunch of head, make direction that data successively decrease along gradient to the selection of base stations transmission;

Source bunch head is selected N along the selection of base stations direction 1An individual gradient ratio source bunch little bunch head is jumped a bunch Data transmission as next;

Middle hop bunch head is selected N along the selection of base stations direction 2The described middle hop of an individual gradient ratio bunch little bunch head is jumped a bunch Data transmission as next;

Wherein, source bunch head refers to the starting point that described data are transmitted, N 1〉=N 2〉=2.

In the embodiment of the invention, because Data transmission is each regional leader cluster node, reduce the influence that bottom layer node moves the network topology change that causes, reduced the route energy consumption.In addition, make the data transmission directly towards the base station with regional ID control route direction, avoided appearance and the meaningless transmission of data on link of route ring, improved router efficiency, shortened network delay.

A kind of cluster device comprises:

Module is divided in the zone, being used for according to the communication coverage of the different discrete power grades of two the base station communication energy in target area logic region being carried out in described target area divides and determines its regional ID, described two base stations are positioned at the two ends of described target area homonymy, and described regional ID is logic region ID;

Agent node is selected module, is used for selecting an agent node in each described zone, and described agent node notification area interior nodes oneself becomes the agency, and described zone is logic region;

Receiver module, the agent node that is used for described zone receives the state information of the described regional interior nodes of described regional interior nodes transmission;

Bunch head is selected module, and the agent node that is used for described zone is selected a node according to described state information in described zone be a bunch head, and described zone becomes one bunch.

Embodiment of the invention agent node selects the introducing of agency mechanism in the module to share the burden of bunch head, has prolonged the duty cycle of bunch head, and regional interior nodes management and data processing efficiency are improved.

A kind of route device by bunch head and base station in the network of described cluster device cluster comprises:

Direction of transfer is selected module, and a lower base station of link load is the data direction of transfer when being used for selecting to make data to transmit from two base stations;

The gradient determination module, be used for according to bunch in described regional ID determine the gradient of described bunch of head, make direction that data successively decrease along gradient to the selection of base stations transmission;

First bunch of head selected module, is used for source bunch head along the selection of base stations direction, selects a N gradient ratio source bunch little bunch head to jump a bunch Data transmission as next;

Second bunch of head selected module, is used for middle hop bunch head along the selection of base stations direction, selects N the described middle hop of a gradient ratio bunch little bunch head to jump a bunch Data transmission as next;

Wherein, source bunch head refers to the starting point that described data are transmitted, N 〉=2.

In the embodiment of the invention, because Data transmission is each regional leader cluster node, reduce the influence that bottom layer node moves the network topology change that causes, reduced the route energy consumption.In addition, make the data transmission directly towards the base station with regional ID control route direction, avoided appearance and the meaningless transmission of data on link of route ring, improved router efficiency, shortened network delay.

Description of drawings

Fig. 1 is the clustering method flow chart that the embodiment of the invention one provides;

Fig. 2 is free-space communication model schematic diagram;

Fig. 3 is the clustering method flow chart that the embodiment of the invention two provides;

Fig. 4 is the monitored area schematic diagram of the embodiment of the invention;

Fig. 5 is the initial message data format schematic diagram of the embodiment of the invention;

Fig. 6 is the relative coordinate schematic diagram of the logic region interior nodes of the embodiment of the invention;

Fig. 7 is that the leader cluster node of the embodiment of the invention carries out data in the packet schematic diagram that forms after the fusion treatment;

Fig. 8 is bunch head that provides of the embodiment of the invention three and the method for routing flow chart of base station;

Fig. 9 is the information conveyer belt schematic diagram of the embodiment of the invention;

Figure 10 is the schematic diagram of the information conveyer belt of the embodiment of the invention " bounce-back " 1 cell when extending to outside the target area;

Figure 11 is the border schematic diagram of the target area of the embodiment of the invention;

Figure 12 is the cluster schematic representation of apparatus of the embodiment of the invention;

Figure 13 is bunch head that provides of the embodiment of the invention five and the route device schematic diagram of base station.

Embodiment

For making the purpose, technical solutions and advantages of the present invention clearer, embodiment of the present invention is described further in detail below in conjunction with accompanying drawing.

Embodiment one

The embodiment of the invention provides a kind of clustering method, as shown in Figure 1, may further comprise the steps:

110: the target area is carried out the logic region division and determined its regional ID, be provided with two base stations in the described target area, described two base stations are positioned at the two ends of target area homonymy;

120: select an agent node in each described zone, described agent node notification area interior nodes oneself becomes the agency;

130: the agent node in described zone receives the described regional interior nodes state information that described regional interior nodes sends;

140: the agent node in described zone is selected a node according to described state information in described zone be a bunch head, and described zone becomes one bunch.

The burden of bunch head has been shared in the introducing of agency mechanism in the embodiment of the invention, has prolonged the duty cycle of bunch head, and regional interior nodes management and data processing efficiency are improved.

Embodiment two

The embodiment of the invention is applied to following network model:

1. node is distributed in plane domain, does not have the communication barrier between the node.

2. node location is fixed, and each node is furnished with power control unit, can dynamic adjustments node transmitting power, and node primary power random distribution.

3. communicating by letter between member node and a bunch head, bunch head and bunch head satisfied free-space communication model E=ε Ampd n, as shown in Figure 2.Wherein: E is energy consumption; D is signal transmission distance; ε AmpBe transmission amplifying circuit energy consumption index, unit is pJ/bit/m 2N is the path attenuation index relevant with environment, and its value is different and difference to some extent sees the following form according to network environment:

In the embodiment of the invention, communication distance is nearer, gets n=2.

4. all nodes have identical function, and have the data fusion ability, and node can detect the signal strength values that receives, and sensitivity is enough high.

5. (Base Station BS), lay respectively at both sides, target monitoring zone boundary, and the distance between two BS is known, as shown in Figure 2 to set up two base stations.

6. the base station communication energy is not limited, and its communication range can cover whole monitored area, and base station communication power τ level is discrete adjustable.

As shown in Figure 3, the embodiment of the invention has proposed a kind of clustering method, and this method comprises:

S310: the target area is carried out the logic region division and determined its regional ID.

The communication energy of base station can be divided into a lot of discrete power grades, and the more high communication coverage of power is more big, otherwise more little, the target monitoring zone can be divided into a lot of relatively independent logic regions accordingly.Node can be similar to definite own distance to two base stations according to the signal strength signal intensity of the different base station that receives, thereby determines the logic region under oneself, and each logic region has oneself independently ID, i.e. regional ID.

If the target monitoring zone is the square of length of side M, two high-performance base stations (BS1, BS2) are deployed in the two ends of monitored area homonymy respectively, shown in Fig. 4 (a), the not limited and big I discrete adjustment of power grade of its energy, maximum power level is made as τ, the more high communication coverage of power grade is more big, otherwise more little.Two base stations (BS1, BS2) are respectively with from small to large discrete power grade 1~τ broadcasting initial message " initial ", the initial message data format comprises shown in Fig. 5 (a): base station numbering (base station IDs), energy grade (Level), transmitting power (Power) etc.The coverage of base station adjacent power grade forms an annular section, and the annular section of two each self-formings of base station intersects mutually, and the target area is divided into a series of virtual grids unit, and this grid cell is exactly the logic region of dividing in the present embodiment.

Each node in the logic region has a logic ID, and the logic ID of all nodes in the same logic region is identical, and the logic ID of node is exactly the regional ID of its logic region of living in.Each node also has a physical I D, in order to distinguish other node.The obtain manner of node logic ID (the regional ID of logic region) is as follows:

After node in the target area is received the initial message of base station broadcast first, detect and record the signal strength values that receives; Resolve the base station and the power grade number information that comprise in the initial message simultaneously, and determine oneself to be in respectively in the ring which two adjacent power grade of base station 1 and base station 2 form according to power grade numbering.The logic ID of node is defined as the combination of two base station adjacent power Grade numbers, and based on two Grade numbers of BS1 preceding, based on two bit numbers of BS2 after, for example, certain node receives that from BS1 with power grade 5 (its adjacent power grade is 4) with from the initial message of BS2 with power grade 8 (its adjacent power grade is 7) broadcasting, then the logic ID of this node is 4578 for the first time respectively.Next continue with high power level broadcasting initial message more two base stations, logic ID is updated when avoiding this node to receive the initial message of two base station broadcasts again, the present embodiment regulation: the node of having set logic ID no longer receives all the other initial messages of base station broadcast, be that node is only handled the initial message of receiving for the first time, unless the instruction that system reinitializes is sent in the base station.The logic region division methods is specific as follows:

On the other hand, the node that all logic ID are identical constitutes a virtual grid unit, the ID of this grid cell is exactly the logic ID of node, is that 4578 grid cell represent the ring 45 and the intersection region of BS2 with adjacent power grade 7, formed ring 78 formation of 8 broadcasting that base station BS 1 forms with adjacent power class 4,5 broadcasting as logic ID among Fig. 4 (b).Under the hypothesis based on the free space model, regulate base station power, make power between the adjacent power grade with τ 2Doubly increase progressively, can obtain equally distributed virtual grid unit.This virtual grid unit is exactly above-mentioned logic region, and its regional ID (regionID) is exactly the logic ID of this logic region interior nodes, wherein based on the ID of BS1 preceding, BS2 after, as Fig. 5 (b).

S320: in each logic region, select an agent node.

A selected agent node (Agent) is for the node administration in the logic region and data are handled separately in each logic region.Agent node is the manager in the logic region, is responsible for selection, the one's respective area interior nodes management of information of bunch head.Itself does not participate in the collection of Monitoring Data, the independent management of being responsible for the one's respective area interior nodes specially.The desirable mode of selecting is:

When selecting Agent for the first time, adopt random fashion to generate.Be chosen as agency's node to other node broadcasts " hello " message, notifying oneself becomes Agent, has only regional ID just to receive this message with the identical node of Agent zone ID.After receiving message, the packet of the signal strength signal intensity that node returns the signal strength signal intensity that comprises following information: physical I D (being the unique identify label of each node), dump energy, receive BS1, receive BS2 is given Agent, announces oneself to become the member node of Agent.Agent safeguards an area data table, deposits the physical I D, dump energy of each node in the region within the jurisdiction, the information of receiving such as signal strength signal intensity, and average residual energy in the zoning.

If not selecting Agent for the first time, then be responsible for triggering the Agent election process by a bunch head.Bunch head is by the current dump energy of acting on behalf of of heartbeat message monitoring, if dump energy is lower than certain percentage (embodiment of the invention is made as 50%) of average residual energy in the zone, then begin new agency's election process: bunch head selects the maximum node of the interior dump energy in zone as the agency, in the zone, broadcast the ID of this agent node then, node sends information such as self rest energy, physical I D to the new post agency after receiving this information, the agency puts the information of receiving in order, forms and the maintenance area tables of data.

S330: agent node is determined the relative coordinate of each node in the one's respective area, and it is joined the area data table.

Establish respectively the lower boundary based on BS1 and BS2 in each zone is made as virtual Y-axis and X-axis, as Fig. 6.As previously mentioned, after each regional node is received the initial message of base station broadcast first, can detect and record the signal strength values that receives.The Agent node is the border with the node that record receives BS1 signal strength signal intensity maximum, and its coordinate is decided to be 0, represents that this node is positioned at the one's respective area lower boundary nearest apart from BS1, and this border is called the reception signal " 0 isogam " based on BS1, and is considered as the Y reference axis.The Agent node receives the node of BS1 signal strength signal intensity minimum with record, namely in the one's respective area apart from BS1 node farthest, the boundary line at place is defined as " 1 isogam " based on BS1, its coordinate is decided to be 1.To the relative value of all the other nodes according to the reception signal strength signal intensity size distribution 0~1 of record, represent that it is apart from the relative position of up-and-down boundary.In like manner, obtain " 0 isogam " and " 1 isogam " with respect to BS2, and node is based on the relative value on this border, and then obtains node based on the relative X coordinate figure of this virtual coordinates system.By above method, each node in the zone obtains one group based on the relative coordinate value of BS1 and BS2 up-and-down boundary isogam, and based on the isogam relative coordinate value of BS1 preceding, as Fig. 6, node a (0.7,08), expression node a is positioned at 0.7 unit of 0 isogam apart from BS1,0.8 unit place of 0 isogam of BS2, Agent joins the coordinate information of all nodes in the area data table.

S340: the agent node of each logic region is selected a leader cluster node in the one's respective area.

After the relative coordinate of each node, Agent namely begins to select bunch head in selected Agent and the definite one's respective area.For the network of a clustering architecture, more little then its global communication energy consumption of bunch head and bunch interior nodes communication distance is more low, and namely bunch head is to member node square distance and more little, and then this node more has the good geographical position that becomes bunch head.Except node geographical position of living in, it is more important that dump energy seems, has only the high node of dump energy could effectively bear high-intensity data processing task.Agent determines leader cluster node by node competitiveness expression formula:

p i = ω E ri - E ‾ E ‾ + ( 1 - ω ) ( 1 - 1 n - 1 Σ j = 1 n - 1 δ 2 ij ) ω ∈ [0,1], i ∈ [1, n], and j ≠ i (1)

Wherein ω is weights, and expression energy information and range information are to the influence degree of node competition bunch head;

E RiDump energy for node;

The expression bunch in the average residual energy,

δ IjRepresent the distance that i node arrives j node in the zone:

δ ij = ( x i - x j ) 2 + ( y i - y j ) 2 i,j∈[1,n],j≠i

Agent is according to the competition intensity p of each node iValue is selected wherein the maximum conduct bunch head.

S350: agent node compute cluster head node is to the distance of node farthest and send it to leader cluster node.

After leader cluster node is determined, Agent query node coordinate, the compute cluster head node arrives the distance R of node farthest CH_node, and with R CH_nodeSend to selected leader cluster node.

S360: leader cluster node is according to the distance of receiving of node farthest that arrives, and its transmitting power of dynamic adjustments is with all nodes in the overlay area.

Leader cluster node is according to the R that receives CH_nodeValue, its transmitting power of dynamic adjustments is with all nodes in the overlay area.Bunch head is to the one's respective area broadcast then, announce oneself to become a bunch head, receive that the one's respective area node of this message is according to the power of communications of the signal strength signal intensity setting that receives oneself and bunch head, and send (join-request) message that joins request to a bunch hair, announce to add this bunch, become the member node of this bunch.

Since then, enter between member node and bunch head and stablize the communication stage.

In the communication of stablizing the communication stage, comprise between member node and the leader cluster node two aspects of the information interaction between bunch head and the neighbours bunch head.

In the communication process between member node and bunch head, member node mainly comprises to the data message that bunch hair send: regional ID, the physical I D etc. of the information that transducer is gathered, residue energy of node, node.For avoiding data collision, bunch head is based on time division multiplexing (TDMA, Time-Division Multiple-Access) the low energy consumption medium access control (MAC of system, MediaAccess Control) agreement distributes the communication time slot for local agent, each member node, only in the time slot of oneself, node is just opened wireless sending module and is communicated by letter with leader cluster node, and this module is in closed condition always in all the other communication time slots, thereby saves energy.

In addition, because broadcast mode is adopted in the node communication, therefore node is when communicating by letter with bunch head of oneself, all the other neighbours' bunch node also may receive this information, for disturbing between avoiding bunch, each bunch uses regional ID in communication process be unique communications identification sign indicating number, and node filters out non-bunch information of receiving according to this identification code.

Leader cluster node is received from behind the monitoring information of member node data being carried out fusion treatment, is formed packet, and calculate this cluster knot point average residual energy.Packet is mainly jumped bunch head region ID, original data region ID, data acquisition time and Monitoring Data etc. by next and is partly constituted, as Fig. 7.Wherein, original data region ID is used for the locator data source.The base station is regularly with maximum power airtime synchronization message, bunch head receive and forwarding time information, node is specified data acquisition time and send to leader cluster node accordingly.

Identical with bunch regional ID, bunch in when communicating by letter, bunch head receives only the identical packet of this ID according to regional ID.But bunch between communication in (between bunch communication in routing procedure, take place, route is the data transmission procedure between each bunch head), send bunch head and bunch head region code that receives data and inequality of data, therefore bunch head may will abandon these route datas, and this should not take place, and should allow bunch head to receive neighbours' bunch head and forward data message.

In order to allow bunch head to receive the packet that neighbours' bunch head forwards, the embodiment of the invention is by arranging a sign position flag in packet, when member's node when sending packet to a bunch hair, this sign position is " 0 ", and bunch head send this sign position of packet of data to be " 1 " to bunch hair, accordingly, receiving node can judge this packet be bunch in or bunch between data, thereby determine whether to receive.When leader cluster node is received the packet that other leader cluster node sends, just received; When leader cluster node is received the packet that rank and file's node sends, just further judge its regional ID, if one's respective area ID is received, otherwise abandon.

In the communication process between bunch head and neighbours' bunch head, leader cluster node communication coverage should reach neighbours' bunch head at least, and the distribution of neighbours zone intra-cluster head both may be in zone central authorities, also may be positioned at edge, zone.Establish the square that each logic region is approximately length of side r in the embodiment of the invention, consider the worst situation, bunch head in the adjacent area all is positioned at the edge, distal side in zone separately, as Fig. 7, and the smallest coverage R of an adjacent area bunch communication then CH-CHFor:

2 r 2 + r 2 ≤ R CH - CH Perhaps R CH - CH ≥ 2 2 r

It is R with the covering radius that leader cluster node is regulated own transmitting power CH-CHCommunications zone.

In order to ensure the robustness of cluster management, the hop-information of making up one's mind mutually between Agent and bunch head is to confirm state each other.Bunch head is circulated a notice of information such as each node current remaining of this bunch, a bunch self rest energy after in the end a communication time slot finishes immediately to Agent.For avoiding leader cluster node frequently to substitute, ensure the stability of a bunch identity, a regulation bunch dump energy is lower than certain percentage of bunch average residual energy, perhaps Agent detects bunch head and can't communicate by letter with this bunch member, neighbours' bunch head again, illustrating then that this node is no longer competent serves as a bunch head, so Agent initiates bunch head election contest process of a new round.On the other hand, Agent gives the dump energy of itself to bunch hair in each oneself communication time slot, equally, for guaranteeing the stability of agent node identity, if regulation agent node dump energy is lower than bunch certain percentage of average residual energy, illustrate that then this node no longer is fit to fill the post of the agency, act on behalf of the election contest process so a bunch hair plays a new round.In addition, if bunch head or agency because of irresistible factor sudden failure, heartbeat timer is not received the other side's heartbeat message, then the either party also can initiate the other side's new round election contest process.Like this, by the heartbeat inspecting mechanism between Agent and bunch head, either party can monitor the other side's state, and can rationally make a strategic decision according to the state information that obtains, and has strengthened the reliability of cluster management.

The embodiment of the invention is divided into the logic region of a plurality of concurrent workings with big zone, shortens node single-hop communication distance, reduces the communication energy consumption, has shortened the cluster time simultaneously, has effectively improved the rapidity of network response.Each logic region is selected an agency, be in charge of the cluster process, the burden of bunch head has been shared in the introducing of agency mechanism, has prolonged the duty cycle of bunch head, regional interior nodes management and data processing efficiency are improved, have strengthened the reliability of cluster management simultaneously.Select bunch head by residue energy of node, node geographical location information (node coordinate), formed bunch of overall average communication distance is littler, bunch is optimized thereby make, global energy consumption consumption in also effectively having reduced bunch.Each regional independent cluster has been realized the even distribution of bunch head simultaneously, has avoided node constantly to transmit the node global information to telepoint base station.

Embodiment three

Present embodiment embodiment two set up bunch on the basis of network communication unit, data communication between carrying out bunch, in order to improve reliability of data transmission, referring to Fig. 8, the embodiment of the invention provides the method for routing of a kind of bunch of head and base station, the not direct and far-end BS communication of bunch head in the method, its data are to send to BS in the multi-hop mode by neighbours' bunch head.In the present embodiment, node refers to leader cluster node.This method comprises:

810: the communications direction that a lower base station of link load is source node when selecting data are transmitted from two base stations.

Two base station x and y respectively with from the discrete power grade of 1~τ from small to large successively to target area broadcasting initial message, the node of target area is determined oneself regional ID, for example regional ID (x of node i according to the initial message that receives i, x i+ 1, y i, y i+ 1).Wherein, preceding two be described logic region corresponding to two of the first base station x adjacent power grades, back two is that described logic region is corresponding to two of the second base station y adjacent power grades; From the base station more away from, the numerical value of power grade is more big; Wherein, i is natural number.

If the next-hop node of node communication is always certain neighbor node of its nearly base station side.For node i (x i, x i+ 1, y i, y i+ 1), if x i<y i, then under equal communication condition, identical as link error rate, node failure rate etc., data are that the link load of aggregation node is low than the link load that is aggregation node with BS1 with BS2, i.e. φ BS1<φ BS2If x i>y i, φ then BS1>φ BS2, namely be that the link load of aggregation node is low with BS2; If x i=y i, then above-mentioned two kinds of situation link loads are suitable, φ BS1BS2The reasons are as follows:

Under free space model hypothesis, source node to the link load φ that aggregation node sends packet is: φ=E * Num * hop, and wherein E is node single communication energy consumption, and Num is single-hop communication (re-transmission) number of times, and hop is that source node is to total jumping figure of aggregation node.Because the next-hop node of node communication is always certain neighbor node of its nearly base station side, therefore be respectively the 2nd (x of node region ID based on the total jumping figure of the route of BS1, BS2 i+ 1) and the 4th (y i+ 1), works as x i<y i, namely as (x i+ 1)<(y i+ 1) time, source node is to the total jumping figure hop of route of base station BS1<hop BS2Under equal communication condition (link error rate, node failure rate etc. are identical), single communication energy consumption E, single-hop communication number of retransmissions Num are equal, therefore, and φ BS1<φ BS2In like manner, work as x i>y iThe time, φ BS1>φ BS2And work as x i=y iThe time, be that the aggregation node link load is suitable with any base station.

For the communication link load that makes source node is minimum, for source node i (x i, x i+ 1, y i, y i+ 1), if x i<y i, then select the direction transmission data to BS1; If x i>y i, then select the direction transmission data to BS2; And if x i=y i, then selecting BS1 or BS2 at random is data transfer direction.

820: determine the gradient of node according to bunch inner region ID, make direction that data successively decrease along gradient to the selection of base stations transmission.

Carry out logic region in the target area and divide, set up the virtual grid unit after, there is a leader cluster node each unit, it has 4 unique regional ID.To any source node i, its regional ID is (x i, x i+ 1, y i, y i+ 1), the communication gradient of this node is the front two of regional ID of node or back two.Work as x i<y iThe time, the communication gradient of this node is the front two (x of regional ID i, x i+ 1); Work as x i>y iThe time, the communication gradient of this node is back two (y of regional ID i, y i+ 1); And work as x i=y iThe time, getting node region ID front two or back two with equiprobability is gradient.The direction transmission that data are always successively decreased along gradient, and increase in gradient, all do not have the data transmission on the identical direction, therefore avoided the appearance of route ring, guarantee that data transmit towards certain base station.

830: source node is along the selection of base stations direction, selects the little node of N gradient ratio source node as the next-hop node Data transmission.Wherein, N 〉=2.

After determining the gradient of leader cluster node, along the principle that the selection of base stations direction is successively decreased according to gradient, select the next-hop node of transfer of data.For guaranteeing the robustness of transfer of data, not with single path or non-intersect multithreading Data transmission, but according to the communication covered radius of node, set up " information conveyer belt " from the source node to the base station, the data that source node sends are passed through N 1(N 1〉=2) an individual gradient ratio source bunch little bunch head is transmitted as next-hop node and is entered the information conveyer belt, and the width of conveyer belt is for participating in each hop node quantity that data are transmitted.In embodiments of the present invention, N gets 3, and bandwidth fixed is 3, and the node that is in the information band edge has two next-hop nodes, and the node in the middle of the information band has three next-hop nodes, as Fig. 9.The method for selecting of the next-hop node of source node is:

If the ID of source node is S (x s, x s+ 1, y s, y s+ 1), supposes x s<y s, then according to step 810, should select BS1 as the purpose base station, gradient is the front two of node ID.The direction transmission that data are always successively decreased along gradient, and increase in gradient, all do not have the data transmission on the identical direction.Therefore, the next-hop node collection ψ of S s3 neighbor node: ψ for the nearly BS1 side of node S S={ n S, L(x S-1, x S, y S+ 1, y S+ 2), n S, M(x S-1, x S, y S, y S+ 1), n S, R(x S-1, x S, y S-1, y S), they are the starting points of " information conveyer belt ".In like manner, if x s>y sThe time, the next-hop node collection ψ of S sBe { n S, L(x S-1, x S, y S-1, y S), n S, M(x S, x S+ 1, y S-1, y S), n S, R(x S+ 1, x S+ 2, y S-1, y S).

840: the middle hop node is along the selection of base stations direction, selects the little node of N gradient ratio middle hop node as the next-hop node Data transmission.

The information conveyer belt is made up of node, and each jumping has the present count destination node.Wherein, the source node index starting point of reportedly passing.Each node after having determined next-hop node, when transmitting data in bag ID number of the binding next-hop node automatically, have only the node that is complementary for ID number could receive this packet, make packet towards the specific direction transmission.Preset number is the width of information conveyer belt, and preset number is 3 in the present embodiment.In the information conveyer belt, except starting point, the method for selecting of the next-hop node of remaining middle hop node is:

If the purpose base station is BS1, for guaranteeing data along the gradient direction transmission of successively decreasing, present node ID front two 1 front two as its next-hop node ID that successively decreases respectively, and behind the next-hop node ID two with the position of node in band different and difference.Because the node communication covering radius is fixed, the leader cluster node in a zone can only be communicated by letter with the leader cluster node in its neighbours zone, and the width of information conveyer belt is defined as 3 in the present embodiment, therefore, be positioned at the middle node of information conveyer belt 3 next-hop nodes are arranged, and the node that is in band edge can only be communicated by letter with two next-hop nodes, referring to Fig. 9 and Figure 10.

If 3 nodes that i jumps are respectively: ψ i={ n I, L(x i, x i+ 1, y i+ 1, y i+ 2), n I, M(x i, x i+ 1, y i, y i+ 1), n I, R(x i, x i+ 1, y i-1, y i).Then the i+1 hop node collection of " information conveyer belt " is ψ I+1={ n I+1, L(x i-1, x i, y i+ 1, y i+ 2), n I+1, M(x i-1, x i, y i, y i+ 1), n I+1, R(x i-1, x i, y i-1, y i) (2) left margin node n wherein I, LThe i+1 hop node be respectively: n I+1, L, n I+1, MBand the right is along node n I, RTwo i+1 hop nodes be: n I+1, M, n I+1, RAnd intermediate node n I, M3 next-hop nodes for being respectively: n I+1, L, n I+1, M, n I+1, RAs seen, not all 3 next-hop nodes can both be received the information of broadcasting of all 3 nodes of previous dive.After having selected next-hop node, " information conveyer belt " extends one towards the base station direction and jumps, if when extending to Grad for the node of " 01 ", illustrates that its next jumping is the purpose base station.

In whole data transfer, " information conveyer belt " need not be set up in advance, but when data transmitted, according to the node local message, autonomous foundation had the feature of distributed control.The information band difference that different source nodes is set up, in case and the information end of transmission, information band namely is dropped.Buffering area of node maintenance is deposited the packet of having transmitted, if node is received the identical data packet from different upstream nodes, then deletes from formation.For reducing conflict, wait for a random time before the node for data forwarding, avoid sending data simultaneously and causing conflict with neighbor node.

As Figure 10, the information conveyer belt may extend to outside the target area.For avoiding this situation to take place, " bounce-back " 1 cell when the conveyer belt middle hop is run into grid cell and is positioned at the border, target area makes behind the ID of next-hop node collection two to successively decrease 1 respectively, next-hop node collection ψ that namely should jumping I+1To cell of right translation, become ψ I+1={ n I+1, L(x i-1, x i, y i, y i+ 1), n I+1, M(x i-1, x i, y i-1, y i), n I+1, R(x i-1, x i, y i-2, y i-1) } (3)

Corresponding left margin node n I, LThe i+1 hop node be: n I+1, L, intermediate node n I, MNext-hop node become 2: n I+1, L, n I+1, M, and band the right is along node n I, R3 i+1 hop node: n are arranged I+1, L, n I+1, M, n I+1, RAfter this next-hop node collection reverts to the set of node of determining in the formula (1) again.

When if source node is positioned at the border of target area, source node S (x s, x s+ 1, y s, y s+ 1) next-hop node present embodiment is defined as 1:

Work as x s<y sThe time, be L (x s-1, x s, y s-1, y s).

Work as x s>y sThe time, be R (x s-1, x s, y s-1, y s).

This wherein, arbitrary node (comprising source node and middle hop node) judges that according to all detection packet of receiving the method on the border that oneself whether is positioned at the target area is:

If this node fails to receive that regional ID is identical with the regional ID front two of self, then two increase progressively 1 detection packet respectively, judge that then this node is positioned at the border in left side, target area,

If this node fail to receive behind regional ID and self the regional ID two identical, and front two increases progressively 1 detection packet respectively, judges that then this node is positioned at the target area right side boundary,

If this node is not received identical two 1 the detection packet of successively decreasing respectively then with self regional ID front two of regional ID, perhaps do not receive behind regional ID and self the regional ID successively decrease respectively 1 detection packet of two identical and front twos, judge that then this node is positioned at lower boundary.

As Figure 11, do not contain the detection packet from 3478 in the detection packet that node 3467 is received, therefore 3467 are positioned at left border; 6734 to can not receive virtual ID be 7834 information, because it is in right margin; In like manner, 2323 fail to receive 2312,1223 information, and then node is positioned at regional lower boundary.But if communication quality is relatively poor, node may can not receive the information in certain neighbours zone and cause erroneous judgement, and can adopt some other corresponding strategies for improving detection accuracy this moment.Such as: node repeatedly sends detection information, perhaps bunch in each member node send a probe message, and the probe message of receiving is sent to the node node, judged according to said method by the node node, thereby improve detection accuracy.

Under the node motion situation, network topology constantly changes, and in the prior art, need repeatedly rebulid node to the communication gradient of base station, and energy consumption is bigger.And in embodiments of the present invention, because each hop node is each regional leader cluster node, how pipe node does not move, as long as whenever there is a leader cluster node in each zone, then data communication will be undertaken by set mode, reduced the influence that bottom layer node moves the network topology change that causes, not only improved the robustness that data are transmitted, the more important thing is that system's route energy consumption can be along with bottom layer node moves and extra increasing.In addition, make the data transmission directly towards the base station with regional ID control route direction, the direction transmission that data are always successively decreased along gradient, and increase in gradient, all do not have the data transmission on the identical direction, therefore appearance and the meaningless transmission of data on link of route ring have been avoided, improve router efficiency, shortened network delay.And bunch adaptively selected next paths, the route flexibility improves, and leader cluster node needn't maintaining routing list, has also saved memory space.

Embodiment four

The embodiment of the invention provides a kind of cluster device, as shown in figure 12, comprising:

Module 1201 is divided in the zone, is used for the target area is carried out the logic region division and determined its regional ID, is provided with two base stations in the target area, and two base stations are positioned at the two ends of target area homonymy;

Agent node is selected module 1202, is used for selecting an agent node in each above-mentioned zone, and agent node notification area interior nodes oneself becomes the agency;

Receiver module 1203, the agent node that is used for above-mentioned zone receives the regional interior nodes state information that the one's respective area interior nodes sends;

Bunch head is selected module 1204, and the agent node that is used for above-mentioned zone is selected a node according to above-mentioned state information in above-mentioned zone be a bunch head, and above-mentioned zone becomes one bunch.

Wherein, the target area is carried out logic region is divided and determine that the method for its regional ID is specially:

Initial message is broadcasted to described target area from small to large successively with the discrete power grade from 1~τ respectively in two base stations;

The node of target area is determined the logic ID of oneself according to the initial message that receives;

The identical node of all logic ID constitutes a logic region, with the logic ID of this logic region interior nodes as its regional ID;

Wherein, initial message comprises base station numbering and emitting power grade, τ>1.

Regional ID shape such as the ID (x of i logic region i, x i+ 1, y i, y i+ 1).Wherein, preceding two be above-mentioned logic region corresponding to two of the first base station x adjacent power grades, back two is that above-mentioned logic region is corresponding to two of the second base station y adjacent power grades; The numerical value of power grade is more big, and the power that the base station sends information is more big, and the overlay area is more wide; Wherein, i is natural number.

Further, agent node selects module 1202 to comprise:

Selected cell 12021 is used for selecting for the first time the agency for the first time, adopts random fashion to generate;

The n time selected cell 12022 is used for the n time and selects to act on behalf of, and when old agency's dump energy is lower than predetermined threshold value, selects dump energy is maximum in the described zone node as the agency by described regional intra-cluster head; Wherein, n is>1 natural number.

Predetermined threshold value is half of the interior average residual energy of logic region in the present embodiment.

Above-mentioned zone interior nodes state information comprises:

The regional ID of node, the physical I D of node, dump energy and receive the signal strength signal intensity of described two base station signals respectively.

Further, said apparatus also comprises:

Node relative coordinate acquisition module 1205, the agent node that is used for above-mentioned zone is determined the relative coordinate (X of each node in the one's respective area according to the regional interior nodes state information of receiver module 1203 receptions i, Y i).

Node relative coordinate acquisition module 1205 comprises:

Set up coordinate system unit 12051, the node position that is used for to receive from the signal strength signal intensity maximum of base station x is the y axle, be that the x axle is set up coordinate system with the node position that receives from the signal strength signal intensity maximum of base station y, described y axle, x axle are respectively as 0 isogam based on base station x and base station y; Be 1 isogam based on base station x to receive node position from the signal strength signal intensity minimum of base station x, is 1 isogam based on base station y with reception from the node position of the signal strength signal intensity minimum of base station y;

Determine relative coordinate unit 12052, be used for the node between described 0 isogam and 1 isogam is distributed 0 to 1 relative value X according to the signal strength signal intensity size that receives from base station x i, distribute 0 to 1 relative value Y according to the signal strength signal intensity size that receives from base station y i

Further, bunch head selects module 1204 to comprise:

Node competitiveness computing unit 12041 is used for the agent node in described zone according to the relative coordinate computing node competitiveness index p of described residue energy of node information and node i, p i = ω E ri - E ‾ E ‾ + ( 1 - ω ) ( 1 - 1 n - 1 Σ j = 1 n - 1 δ 2 ij ) Wherein ω is weights, ω ∈ [0,1], and the influence degree that expression energy information and range information are competed bunch head to node,

E RiBe the dump energy of node,

Represent average residual energy in the described zone,

N is the number of described regional interior nodes,

δ IjRepresent i node to the distance of interior j the node in zone, I, j ∈ [1, n], and j ≠ i;

A bunch selected cell 12042, the agent node that is used for described zone is selected described p iThe node of the maximum correspondence is bunch head in this zone.

When the agent node of above-mentioned zone is selected in the one's respective area according to above-mentioned state information after a node is bunch head, the relative coordinate of this agent node inquiry one's respective area interior nodes, the compute cluster head arrives the distance of the node farthest of one's respective area, and this distance is sent to a bunch head; Bunch head according to its transmitting power of this distance adjustment to cover all nodes in the one's respective area; Bunch head announces oneself to become a bunch head to the one's respective area broadcast; The one's respective area node of receiving broadcast send adding message according to the signal strength signal intensity setting that receives is own with the power of communications of bunch head and to a bunch hair, announces to add described bunch.

The embodiment of the invention is divided into the logic region of a plurality of concurrent workings with big zone, shortens node single-hop communication distance, reduces the communication energy consumption, has shortened the cluster time simultaneously, has effectively improved the rapidity of network response.Each logic region is selected an agency, be in charge of the cluster process, the burden of bunch head has been shared in the introducing of agency mechanism, has prolonged the duty cycle of bunch head, regional interior nodes management and data processing efficiency are improved, have strengthened the reliability of cluster management simultaneously.Select bunch head by residue energy of node, node geographical location information (node coordinate), formed bunch of overall average communication distance is littler, bunch is optimized thereby make, global energy consumption consumption in also effectively having reduced bunch.Each regional independent cluster has been realized the even distribution of bunch head simultaneously, has avoided node constantly to transmit the node global information to telepoint base station.

Embodiment five

The embodiment of the invention comes on the network foundation of cluster at the cluster device that is provided by embodiment four, and the route device of a kind of bunch of head and base station is provided, and referring to Figure 13, comprising:

Direction of transfer is selected module 1301, and a lower base station of link load is the data direction of transfer when being used for selecting to make data to transmit from two base stations;

Gradient determination module 1302, be used for according to bunch in described regional ID determine the gradient of described bunch of head, make direction that data successively decrease along gradient to the selection of base stations transmission;

First bunch of head selected module 1303, is used for source bunch head along the selection of base stations direction, selects a N gradient ratio source bunch little bunch head to jump a bunch Data transmission as next;

Second bunch of head selected module 1304, is used for middle hop bunch head along the selection of base stations direction, selects N the described middle hop of a gradient ratio bunch little bunch head to jump a bunch Data transmission as next;

Wherein, source bunch head refers to the starting point that described data are transmitted, N 〉=2.

Regional ID shape such as the ID (x of i described logic region i, x i+ 1, y i, y i+ 1), wherein, preceding two be described logic region corresponding to two of the first base station x adjacent power grades, back two is that described logic region is corresponding to two of the second base station y adjacent power grades; Wherein, i is natural number.

Gradient determination module 1302 is determined the gradient of each bunch head according to the regional ID of each bunch head:

A bunch i (x i, x i+ 1, y i, y i+ 1), if its x i<y i, then its gradient is (x i, x i+ 1);

If its x i>y i, then its gradient is (y i, y i+ 1);

If x i=y i, then its gradient is (x i, x i+ 1) or (y i, y i+ 1).

Direction of transfer selection module 1301 is come the specified data direction of transfer according to the regional ID of source bunch head:

A source bunch S (x s, x s+ 1, y s, y s+ 1), if its x s<y s, then under equal communication condition, select the x base station to carry out the data transmission;

If its x s>y s, then under equal communication condition, select the y base station to carry out the data transmission;

If x s=y s, then selecting arbitraryly among both to carry out the data transmission under the equal communication condition.

Further, said apparatus also comprises:

Detection packet broadcast module 1305, all bunches head broadcasting that is used for the target area comprises the detection packet of self zone ID;

Border judge module 1306 is used for arbitrary bunch of head and judges according to all detection packet of receiving whether oneself is positioned at the border of target area:

If this bunch head fails to receive that regional ID is identical with the regional ID front two of self, then two increase progressively 1 detection packet respectively, judge that then this bunch head is positioned at the border in left side, target area,

If this bunch head fail to receive behind regional ID and self the regional ID two identical, and front two increases progressively 1 detection packet respectively, judges that then this bunch head is positioned at the target area right side boundary,

If this bunch head is not received identical two 1 the detection packet of successively decreasing respectively then with self regional ID front two of regional ID, perhaps do not receive behind regional ID and self the regional ID successively decrease respectively 1 detection packet of two identical and front twos, judge that then this bunch head is positioned at lower boundary.

Correspondingly, first bunch of head selected module 1303, specifically is used for source bunch head along the selection of base stations direction, selects a N gradient ratio source bunch little bunch head to jump a bunch Data transmission as next according to the judged result of border judge module 1306;

Second bunch of head selected module 1304, specifically is used for middle hop bunch head along the selection of base stations direction, selects N the described middle hop of a gradient ratio bunch little bunch head to jump a bunch Data transmission as next according to the judged result of border judge module 1306.

In the present embodiment, N gets 3.

First bunch of head selected module 1303 pairs of sources bunch S (x s, x s+ 1, y s, y s+ 1), works as x s<y sThe time, next of a source bunch S jumped a bunch head and is respectively: L (x S-1, x S, y S+ 1, y S+ 2), M (x S-1, x S, y S, y S+ 1), R (x S-1, x S, y S-1, y S); Work as x s>y sThe time, next of a source bunch S jumped a bunch head and is respectively: L (x S-1, x S, y S-1, y S), M (x S, x S+ 1, y S-1, y S), R (x S+ 1, x S+ 2, y S-1, y S).

When if above-mentioned source bunch head is positioned at the border of target area, it is 1 that next of a source bunch S jumped bunch head:

Work as x s<y sThe time, be L (x s-1, x s, y s-1, y s).

Work as x s>y sThe time, be R (x s-1, x s, y s-1, y s).

Second bunch of head selected module 1304, and when the selected first base station x was the data direction of transfer, the front two of the regional ID of the N of middle hop bunch head collection next jumping bunch head collection subtracted 1 respectively, and back two remain unchanged; When the selected second base station y was the data direction of transfer, back two of the regional ID of the N of described middle hop bunch head collection next jumping bunch head collection subtracted 1 respectively, and front two remains unchanged.

If in the above-mentioned middle hop bunch head during the arbitrary border that is positioned at the target area, its next jump bunch head all four of regional ID subtract 1 respectively.

Under the node motion situation, network topology constantly changes, and in the prior art, need repeatedly rebulid node to the communication gradient of base station, and energy consumption is bigger.And in embodiments of the present invention, because each hop node is each regional leader cluster node, how pipe node does not move, as long as whenever there is a leader cluster node in each zone, then data communication will be undertaken by set mode, reduced the influence that bottom layer node moves the network topology change that causes, not only improved the robustness that data are transmitted, the more important thing is that system's route energy consumption can be along with bottom layer node moves and extra increasing.In addition, make the data transmission directly towards the base station with regional ID control route direction, the direction transmission that data are always successively decreased along gradient, and increase in gradient, all do not have the data transmission on the identical direction, therefore appearance and the meaningless transmission of data on link of route ring have been avoided, improve router efficiency, shortened network delay.And bunch adaptively selected next paths, the route flexibility improves, and leader cluster node needn't maintaining routing list, has also saved memory space.

The embodiment of the invention can utilize software to realize, corresponding software program can be stored in the storage medium that can read, for example, and in the hard disk of router, buffer memory or the CD.The above embodiment of the invention is the example explanation with two base stations, but the number of base stations of the embodiment of the invention is not limited to two, also can be more than two.

The above only is preferred embodiment of the present invention, and is in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, is equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (26)

1. a clustering method is characterized in that, may further comprise the steps:
According to the communication coverage of the different discrete power grades of two base station communication energy in the target area logic region being carried out in described target area divides and determines its regional ID, described two base stations are positioned at the two ends of described target area homonymy, and described regional ID is logic region ID;
Select an agent node in each described zone, described agent node notification area interior nodes oneself becomes the agency, and described zone is logic region;
The agent node in described zone receives the described regional interior nodes state information that described regional interior nodes sends;
The agent node in described zone is selected a node according to described state information in described zone be a bunch head, and described zone becomes one bunch.
2. clustering method according to claim 1 is characterized in that, describedly the target area is carried out logic region is divided and determine that the step of its regional ID specifically comprises:
Be provided with two base stations in the described target area, initial message is broadcasted to described target area from small to large successively with the discrete power grade from 1~τ respectively in two described base stations;
The node of described target area is determined the logic ID of oneself according to the initial message that receives;
The identical node of all logic ID constitutes a logic region, with the logic ID of this logic region interior nodes as its regional ID;
Wherein, described initial message comprises base station numbering and emitting power grade, τ>1.
3. clustering method according to claim 2 is characterized in that, regional ID shape such as the ID (x of i described logic region i, x i+ 1, y i, y i+ 1), wherein, preceding two be described logic region corresponding to two of the first base station x adjacent power grades, back two is that described logic region is corresponding to two of the second base station y adjacent power grades; The numerical value of power grade is more big, and the power that the base station sends information is more big, and the overlay area is more wide;
Wherein, i is natural number.
4. clustering method according to claim 1 is characterized in that, the described step of an agent node of selecting in each described zone is specially:
If select the agency for the first time, adopt random fashion to generate;
If select the agency for the n time, when old agency's dump energy is lower than predetermined threshold value, select dump energy is maximum in the described zone node as the agency by described regional intra-cluster head;
Wherein, n is the natural number greater than 1.
5. clustering method according to claim 4 is characterized in that, described predetermined threshold value is half of the interior average residual energy of described logic region.
6. clustering method according to claim 1 is characterized in that, is provided with two base stations in the described target area, and described regional interior nodes state information comprises:
The regional ID of node, the physical I D of node, dump energy and receive the signal strength signal intensity of described two base station signals respectively.
7. clustering method according to claim 6 is characterized in that, the agent node in described zone selects a node also to comprise for before the step of bunch head in described zone according to described state information:
The agent node in described zone is determined the relative coordinate (X of each node in the one's respective area according to described regional interior nodes state information i, Y i), be specially:
Being the y axle with the border, node place that receives from the signal strength signal intensity maximum of base station x, is that the x axle is set up coordinate system with the border, node place that receives from the signal strength signal intensity maximum of base station y, and described y axle, x axle are respectively as 0 isogam based on base station x and base station y; Be 1 isogam based on base station x to receive node position from the signal strength signal intensity minimum of base station x, is 1 isogam based on base station y with reception from the node position of the signal strength signal intensity minimum of base station y;
Node between described 0 isogam and 1 isogam is distributed 0 to 1 relative value X according to the signal strength signal intensity size that receives from base station x i, distribute 0 to 1 relative value Y according to the signal strength signal intensity size that receives from base station y i
8. clustering method according to claim 7 is characterized in that, the agent node in described zone selects a node specifically to comprise for the step of bunch head in described zone according to described state information:
The agent node in described zone is according to the relative coordinate computing node competitiveness index p of described residue energy of node information and node i, p i = ω E ri - E ‾ E ‾ + ( 1 - ω ) ( 1 - 1 n - 1 Σ j = 1 n - 1 δ 2 ij )
Wherein ω is weights, ω ∈ [0,1], and the influence degree that expression energy information and range information are competed bunch head to node,
E RiBe the dump energy of node,
Represent average residual energy in the described zone,
N is the number of described regional interior nodes,
δ IjRepresent i node to the distance of interior j the node in zone, I, j ∈ [1, n], and j ≠ i;
The agent node in described zone is selected described p iThe node of the maximum correspondence is bunch head in this zone.
9. clustering method according to claim 7 is characterized in that, the agent node in described zone is selected also to comprise after the step of a node for bunch head in described zone according to described state information:
Agent node is inquired about the described relative coordinate of described regional interior nodes in the described zone, and the compute cluster head arrives the distance of the node farthest in described zone, and described distance is sent to described bunch of head;
Described bunch of head according to its transmitting power of described distance adjustment to cover all nodes in the described zone;
Described bunch of head announces oneself to become a bunch head to described regional broadcast message;
The described Area Node of receiving described broadcast send adding message according to the signal strength signal intensity setting that receives is own with the power of communications of described bunch of head and to described bunch of hair, announces to add described bunch.
10. clustering method according to claim 1 is characterized in that, the agent node in described zone is selected also to comprise after the step of a node for bunch head in described zone according to described state information:
Described bunch of head distributes the communication time slot for local agent and each member node;
Described local agent only communicates by letter with this bunch bunch head in the communication time slot of oneself with each member node.
11. clustering method according to claim 1 is characterized in that,
Described bunch is used regional ID as unique communication identifier sign indicating number in communication process, simultaneously binding data type identification in packet;
Described regional interior nodes filters out non-the cluster data bag of receiving according to this communication identifier sign indicating number;
Described bunch of head receives the packet that neighbours' bunch hair send according to this data type sign, receives the packet that this cluster knot point sends according to this data type sign and this communication identifier sign indicating number.
12. the method for routing according in the network of the described clustering method cluster of claim 1 bunch of head and base station is characterized in that, may further comprise the steps:
A lower base station of link load is the data direction of transfer when selecting data are transmitted from two base stations;
According to bunch in described regional ID determine the gradient of described bunch of head, make direction that data successively decrease along gradient to the selection of base stations transmission;
Source bunch head selects a N gradient ratio source bunch little bunch head to jump a bunch Data transmission as next along the selection of base stations direction;
Middle hop bunch head selects N the described middle hop of a gradient ratio bunch little bunch head to jump a bunch Data transmission as next along the selection of base stations direction;
Wherein, source bunch head refers to the starting point that described data are transmitted, N 〉=2.
13. the method for routing of bunch head and base station is characterized in that in the network according to claim 12, regional ID shape such as the ID (x of i described logic region i, x i+ 1, y i, y i+ 1), wherein, preceding two be described logic region corresponding to two of the first base station x adjacent power grades, back two is that described logic region is corresponding to two of the second base station y adjacent power grades; Wherein, i is natural number; This logic region ID is the logic ID of all nodes in the unified logic region, and namely the logic ID of node is i (x among the regional i i, x i+ 1, y i, y i+ 1).
14. the method for routing of bunch head and base station is characterized in that in the network according to claim 13, according to bunch in described regional ID determine that the method for the gradient of described bunch of head is specially:
A bunch i (x i, x i+ 1, y i, y i+ 1), if its x i<y i, then its gradient is (x i, x i+ 1);
If its x i>y i, then its gradient is (y i, y i+ 1);
If x i=y i, then its gradient is (x i, x i+ 1) or (y i, y i+ 1).
15. the method for routing of bunch head and base station is characterized in that in the network according to claim 13, the described step that a lower base station of link load is the data direction of transfer when selecting data are transmitted from two base stations is specially:
The logic ID of source bunch head is S (x s, x s+ 1, y s, y s+ 1), if its x s<y s, then under equal communication condition, select the x base station to carry out the data transmission;
If its x s>y s, then under equal communication condition, select the y base station to carry out the data transmission;
If x s=y s, then selecting arbitraryly among both to carry out the data transmission under the equal communication condition.
16. the method for routing of bunch head and base station is characterized in that in the network according to claim 12, also comprises step:
All of target area bunch head broadcasting comprises the detection packet of self zone ID;
Arbitrary bunch of head judges according to all detection packet of receiving whether oneself is positioned at the border of target area;
If this bunch head fails to receive that regional ID is identical with the regional ID front two of self, then two increase progressively 1 detection packet respectively, judge that then this bunch head is positioned at the left border of target area,
If this bunch head fail to receive behind regional ID and self the regional ID two identical, and front two increases progressively 1 detection packet respectively, judges that then this bunch head is positioned at the right side boundary of target area,
If this bunch head is not received identical two 1 the detection packet of successively decreasing respectively then with self regional ID front two of regional ID, perhaps do not receive behind regional ID and self the regional ID successively decrease respectively 1 detection packet of two identical and front twos, judge that then this bunch head is positioned at the lower boundary of target area.
17. the method for routing of bunch head and base station is characterized in that in the network according to claim 16,
Described source bunch head selects a N gradient ratio source bunch little bunch head to be specially 3 as N in the step of next jumping bunch Data transmission along the selection of base stations direction;
Work as x s<y sThe time, a source bunch S (x s, x s+ 1, y s, y s+ 1) next jumped a bunch head and is respectively: a bunch L (x S-1, x S, y S+ 1, y S+ 2), a bunch M (x S-1, x S, y S, y S+ 1), a bunch R (x S-1, x S, y S-1, y S),
Work as x s>y sThe time, a source bunch S (x s, x s+ 1, y s, y s+ 1) next jumped a bunch head and is respectively: a bunch L (x S-1, x S, y S-1, y S), a bunch M (x S, x S+ 1, y S-1, y S), a bunch R (x S+ 1, x S+ 2, y S-1, y S).
18. the method for routing of bunch head and base station is characterized in that in the network according to claim 16,
When the selected first base station x was the data direction of transfer, the front two of the regional ID of the N of described middle hop bunch head next jumping bunch head subtracted 1 respectively, and back two remain unchanged;
When the selected second base station y was the data direction of transfer, back two of the regional ID of the N of described middle hop bunch head next jumping bunch head subtracted 1 respectively, and front two remains unchanged.
19. the method for routing of bunch head and base station is characterized in that in the network according to claim 17, if when described source bunch head is positioned at the border of target area,
A source bunch S (x s, x s+ 1, y s, y s+ 1) it is 1 that next jumped bunch head:
Work as x s<y sThe time, be L (x s-1, x s, y s-1, y s);
Work as x s>y sThe time, be R (x s-1, x s, y s-1, y s).
20. the method for routing of bunch head and base station is characterized in that in the network according to claim 18, if in the described middle hop bunch head during the arbitrary border that is positioned at the target area, all four of regional ID that next of described middle hop bunch head jumped bunch head subtract 1 respectively.
21. a cluster device is characterized in that, comprising:
Module is divided in the zone, being used for according to the communication coverage of the different discrete power grades of two the base station communication energy in target area logic region being carried out in described target area divides and determines its regional ID, described two base stations are positioned at the two ends of described target area homonymy, and described regional ID is logic region ID;
Agent node is selected module, is used for selecting an agent node in each described zone, and described agent node notification area interior nodes oneself becomes the agency, and described zone is logic region;
Receiver module, the agent node that is used for described zone receives the state information of the described regional interior nodes of described regional interior nodes transmission;
Bunch head is selected module, and the agent node that is used for described zone is selected a node according to described state information in described zone be a bunch head, and described zone becomes one bunch.
22. cluster device according to claim 21 is characterized in that, agent node selects module to comprise:
Selected cell is used for selecting for the first time the agency for the first time, adopts random fashion to generate;
The n time selected cell is used for the n time and selects to act on behalf of, and when old agency's dump energy is lower than predetermined threshold value, selects dump energy is maximum in the described zone node as the agency by described regional intra-cluster head;
Wherein, n is the natural number greater than 1.
23. cluster device according to claim 21 is characterized in that, described device also comprises:
Node relative coordinate acquisition module, the agent node that is used for described zone is determined the relative coordinate (X of each node in the one's respective area according to the regional interior nodes state information of described receiver module reception i, Y i);
Wherein, described node relative coordinate acquisition module comprises:
Set up the coordinate system unit, the border, node place that is used for to receive from the signal strength signal intensity maximum of base station x is the y axle, be that the x axle is set up coordinate system with the border, node place that receives from the signal strength signal intensity maximum of base station y, described y axle, x axle are respectively as 0 isogam based on base station x and base station y; Be 1 isogam based on base station x to receive node position from the signal strength signal intensity minimum of base station x, is 1 isogam based on base station y with reception from the node position of the signal strength signal intensity minimum of base station y;
Determine the relative coordinate unit, be used for the node between described 0 isogam and 1 isogam is distributed 0 to 1 relative value X according to the signal strength signal intensity size that receives from base station x i, distribute 0 to 1 relative value Y according to the signal strength signal intensity size that receives from base station y i
24. cluster device according to claim 23 is characterized in that, described bunch of head selects module to comprise:
Node competitiveness computing unit is used for the agent node in described zone according to the relative coordinate computing node competitiveness index p of described residue energy of node information and node i, p i = ω E ri - E ‾ E ‾ + ( 1 - ω ) ( 1 - 1 n - 1 Σ j = 1 n - 1 δ 2 ij )
Wherein ω is weights, ω ∈ [0,1], and the influence degree that expression energy information and range information are competed bunch head to node,
E RiBe the dump energy of node,
Represent average residual energy in the described zone,
N is the number of described regional interior nodes,
δ IjRepresent i node to the distance of interior j the node in zone, I, j ∈ [1, n], and j ≠ i;
A bunch selected cell, the agent node that is used for described zone is selected described p iThe node of the maximum correspondence is bunch head in this zone.
25. the route device by bunch head and base station in the network of the described cluster device of claim 21 cluster is characterized in that, comprising:
Direction of transfer is selected module, and a lower base station of link load is the data direction of transfer when being used for selecting to make data to transmit from two base stations;
The gradient determination module, be used for according to bunch in described regional ID determine the gradient of described bunch of head, make direction that data successively decrease along gradient to the selection of base stations transmission;
First bunch of head selected module, is used for source bunch head along the selection of base stations direction, selects a N gradient ratio source bunch little bunch head to jump a bunch Data transmission as next;
Second bunch of head selected module, is used for middle hop bunch head along the selection of base stations direction, selects N the described middle hop of a gradient ratio bunch little bunch head to jump a bunch Data transmission as next;
Wherein, source bunch head refers to the starting point that described data are transmitted, N 〉=2.
26. the route device of bunch head and base station is characterized in that in the network of cluster device cluster according to claim 25, also comprises:
The detection packet broadcast module, all bunches head broadcasting that is used for the target area comprises the detection packet of self zone ID;
The border judge module is used for arbitrary bunch of head and judges according to all detection packet of receiving whether oneself is positioned at the border of target area;
Described border judge module, concrete being used for if this bunch head fails to receive that regional ID is identical with the regional ID front two of self, then two increase progressively 1 detection packet respectively, judge that then this bunch head is positioned at the left border of target area,
If this bunch head fail to receive behind regional ID and self the regional ID two identical, and front two increases progressively 1 detection packet respectively, judges that then this bunch head is positioned at the right side boundary of target area,
If this bunch head is not received identical two 1 the detection packet of successively decreasing respectively then with self regional ID front two of regional ID, perhaps do not receive behind regional ID and self the regional ID successively decrease respectively 1 detection packet of two identical and front twos, judge that then this bunch head is positioned at the lower boundary of target area;
Correspondingly, described first bunch of head selected module, specifically is used for source bunch head along the selection of base stations direction, selects a N gradient ratio source bunch little bunch head to jump a bunch Data transmission as next according to the judged result of border judge module;
Second bunch of head selected module, specifically is used for middle hop bunch head along the selection of base stations direction, selects N the described middle hop of a gradient ratio bunch little bunch head to jump a bunch Data transmission as next according to the judged result of border judge module.
CN2008101728400A 2008-10-28 2008-10-28 Clustering method and device, routing method and device of cluster head and base station CN101729331B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2008101728400A CN101729331B (en) 2008-10-28 2008-10-28 Clustering method and device, routing method and device of cluster head and base station

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2008101728400A CN101729331B (en) 2008-10-28 2008-10-28 Clustering method and device, routing method and device of cluster head and base station

Publications (2)

Publication Number Publication Date
CN101729331A CN101729331A (en) 2010-06-09
CN101729331B true CN101729331B (en) 2013-08-28

Family

ID=42449601

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2008101728400A CN101729331B (en) 2008-10-28 2008-10-28 Clustering method and device, routing method and device of cluster head and base station

Country Status (1)

Country Link
CN (1) CN101729331B (en)

Families Citing this family (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102340855B (en) * 2011-11-15 2013-10-30 江南大学 Low energy-consumption data acquisition method based on Zigbee wireless sensor network
CN102448139B (en) * 2011-12-28 2014-08-06 南昌大学 Hierarchical routing method for wireless sensor network
CN103249054B (en) * 2012-02-09 2016-01-06 中国联合网络通信集团有限公司 The transmission method of business datum, Apparatus and system
CN102624748B (en) * 2012-04-16 2014-09-03 暨南大学 Peer-to-peer (P2P) network access control method
CN102685688B (en) * 2012-06-12 2014-06-11 西华大学 Wireless sensor network clustering method based on first-price-sealed bid auction
CN103052130B (en) * 2012-11-20 2015-08-05 南京邮电大学 A kind of wireless multimedia sensor network data fusion method based on rough set
CN104159258B (en) * 2013-05-15 2018-03-02 电信科学技术研究院 A kind of state of communication nodes method of adjustment, apparatus and system
CN103281741B (en) * 2013-05-16 2016-01-20 东南大学 A kind of wireless sensor network clustering routing divided based on hexagonal area
CN104254081B (en) * 2013-06-29 2018-05-25 华为技术有限公司 A kind of method and terminal device of centralization networking
CN103491612B (en) * 2013-09-16 2016-03-02 电子科技大学 Mobile Sink node location information acquisition method in a kind of sensor network
CN104080144B (en) * 2014-07-18 2017-08-29 重庆邮电大学 A kind of Energy Efficient Uneven Cluster data forwarding method based on gradient
CN104469854B (en) * 2014-11-27 2017-09-19 重庆邮电大学 A kind of wireless industrial WIA PA network clustering methods of adaptive-flow
WO2017068597A1 (en) * 2015-10-23 2017-04-27 Telefonaktiebolaget Lm Ericsson (Publ) Establishment of operational status of a machine-to-machine device
CN105682212A (en) * 2016-04-01 2016-06-15 中国联合网络通信集团有限公司 Ad hoc network information transfer method and ad hoc network
CN106487586A (en) * 2016-10-13 2017-03-08 北京东土科技股份有限公司 A kind of self-organized network topology method for building up and device
CN106658381B (en) * 2016-12-31 2019-08-30 中国铁道科学研究院电子计算技术研究所 A kind of landslide method for early warning based on wireless sensor network
CN107197514A (en) * 2017-04-27 2017-09-22 上海大学 The cognizable global clock synchronous method of wireless sensor network and the application in multihop network
CN107071849A (en) * 2017-06-20 2017-08-18 海信集团有限公司 A kind of method and apparatus for electing leader cluster node
CN107340017A (en) * 2017-06-28 2017-11-10 深圳源广安智能科技有限公司 Aerial power transmission line on-line monitoring system
CN108471603A (en) * 2018-03-29 2018-08-31 深圳汇创联合自动化控制有限公司 A kind of highway overload remediation system based on block chain technology
CN108900587A (en) * 2018-05-22 2018-11-27 四川斐讯信息技术有限公司 A kind of intelligent terminal acts on behalf of decision system
CN108881436A (en) * 2018-05-22 2018-11-23 四川斐讯信息技术有限公司 A kind of decision system of proxy terminal

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101282290A (en) * 2008-03-05 2008-10-08 中科院嘉兴中心微系统所分中心 Method for balancing and clustering load of wireless sensing network based on gradient

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101282290A (en) * 2008-03-05 2008-10-08 中科院嘉兴中心微系统所分中心 Method for balancing and clustering load of wireless sensing network based on gradient

Also Published As

Publication number Publication date
CN101729331A (en) 2010-06-09

Similar Documents

Publication Publication Date Title
Kim et al. Challenging the IPv6 routing protocol for low-power and lossy networks (RPL): A survey
Zungeru et al. Classical and swarm intelligence based routing protocols for wireless sensor networks: A survey and comparison
Ho et al. A ladder diffusion algorithm using ant colony optimization for wireless sensor networks
Ganesh et al. Efficient and secure routing protocol for wireless sensor networks through SNR based dynamic clustering mechanisms
Junhai et al. A survey of multicast routing protocols for mobile ad-hoc networks
Agarwal et al. Survey of clustering algorithms for MANET
Zeng et al. Directional routing and scheduling for green vehicular delay tolerant networks
Al-Turjman Cognitive routing protocol for disaster-inspired internet of things
Jurdak Wireless ad hoc and sensor networks: A cross-layer design perspective
Akyildiz et al. A cross-layer protocol for wireless sensor networks
Ben-Othman et al. Energy efficient and QoS based routing protocol for wireless sensor networks
Ergen et al. PEDAMACS: Power efficient and delay aware medium access protocol for sensor networks
Rathi et al. A review on routing protocols for application in wireless sensor networks
Stojmenovic et al. Broadcasting and activity scheduling in ad hoc networks
Hong et al. Load balanced, energy-aware communications for mars sensor networks
Tamboli et al. Coverage-aware connectivity restoration in mobile sensor networks
US9386504B2 (en) Energy-efficient network protocol and node device for sensor networks
CN100446496C (en) Method for establishing radio sensor network rout ebased on route neighbour list
Garcia et al. Saving energy and improving communications using cooperative group-based wireless sensor networks
CN102149160B (en) Energy perception routing algorithm used for wireless sensing network
Sheikhpour et al. Comparison of energy efficient clustering protocols in heterogeneous wireless sensor networks
CN101360051B (en) Energy efficient wireless sensor network routing method
Li et al. A cone-based distributed topology-control algorithm for wireless multi-hop networks
Pazzi et al. Mobile data collector strategy for delay-sensitive applications over wireless sensor networks
CN102395173B (en) Clustering method for efficiently utilizing energy in perception layer of environment monitoring internet of things

Legal Events

Date Code Title Description
PB01 Publication
C06 Publication
SE01 Entry into force of request for substantive examination
C10 Entry into substantive examination
GR01 Patent grant
C14 Grant of patent or utility model