CN103619016A - Self-adaptive grid safe routing method in wireless sensor network - Google Patents

Abstract

The invention relates to a self-adaptive grid safe routing method in a wireless sensor network, and belongs to the technical field of wireless sensors. The method is characterized in that grids are divided to perform redundancy control by using a base station so that the right that low-energy nodes are selected in a cluster head can be kicked, a cluster head election mechanism is improved, after a cluster head election is completed, each cluster head node randomly distributes a single chain secret key from a secret key pool composed of hash chains, other nodes use a one-way hash function and a pseudorandom function for generating communication secret keys, a secondary cluster head election mechanism is added, and after being attacked, the network is recovered by using a grid self-rule and the serial number of identification of the cluster head to be elected. The safe routing algorithm can effectively reduce the death rate of the nodes, enhance anti-attack capacity, improve the data fusion degree and prolong the lifetime of the network.

Description

Wireless sensor network adaptive mesh safety routing method

Technical field

The present invention relates to a kind of wireless sensor network adaptive mesh safety routing method, belong to wireless sensor network technology field.

Background technology

Wireless sensor network is widely used in the severe situation of some environment such as detection, medical treatment, space exploration now, and along with the development of computer technology and information technology, its importance becomes increasingly conspicuous.The applied environment of wireless sensor network determines that network itself is easy to be subject to the attack of malicious node, so the safety problem of wireless sensor network is the problem of overriding concern.The key management that the safe and reliable secure communication of take is target is the emphasis of WSN security study, is also the basis of safe Routing Protocol.

LEACH(Low Energy Adaptive Clustering Hierarchy) Routing Protocol is a kind of low power consumption adaptive Routing Protocol based on cluster, is that the people such as Heinzelman and Chandrakasan proposed in 2000.The basic thought of LEACH is the concept of introducing " wheel ", by selecting randomly cluster head, a utilization bunch head converge data then with base station communication, so just reduced the energy loss of node and base station communication.LEACH agreement is assigned to each node in network by the energy loss of overall network.

Bunch each of LEACH agreement is taken turns circulation can be divided into two stages, establishment stage and data transfer phase.Bunch establishment stage, network will be chosen sensor node randomly as cluster head, the principle of choosing is: the number between one 0 to 1 of the random generation of each sensor node, if the random number generating is less than threshold value T (n), this node is elected as cluster head, and threshold value T (n) calculates by following formula

$T\left(n\right)=\left\{\begin{array}{cc}\frac{p}{1-p\left[r\mathrm{mod}(1/p)\right]}& n\∈G\\ 0& n\∉G\end{array}\right.---\left(1\right)$

Wherein P is the percentage that node becomes leader cluster node, and r is election wheel number, and rmod (1/p) represents that this takes turns the node number of being elected to bunch head in circulation, and G is the node set of elected bunch head not in nearest takes turns circulation.After bunch head is selected, leader cluster node releases news in network in the mode of broadcast, thus other nodes in network according to the signal strength signal intensity receiving determine under it bunch and inform the foundation that bunch head completes bunch.Stabilization sub stage is data transfer phase namely, and ordinary node is sent to leader cluster node by the data that collect.Bunch data that head converges it carry out being transmitted to base station after fusion treatment.After stabilization sub stage continues for some time, network reenters next round, and constantly circulation thus, until the life cycle of whole network finishes.

Distributed, self adaptation cluster mechanism that LEACH agreement adopts, periodically elect bunch head mechanism, the life span that has extended greatly radio sensing network.But LEACH agreement still exists a lot of shortcomings to be badly in need of solving.The election of cluster head of LEACH agreement itself is random and each node probability of being elected as bunch head is etc. large, so just likely occurs that a bunch skewness or low-yield node be elected as the situation of bunch head.And a lot of agreements that the later stage grows up according to LEACH are not all considered safety problem.

In key managing project Shang， academia, generally believe that asymmetrical key managing project is because the intrinsic communication energy bottleneck of node is not suitable for.Eschenauer in 2002 and Gligor propose encryption key distribution Managed Solution first, and Anjum F has proposed symmetric key Managed Solution on this basis, and the key managing project based on symmetric cryptography has become the emphasis of current research.The core of symmetric key Managed Solution is the distribution of key; distribute certain key information in advance before network design each node, latter two of deployment needs the node of secure communication to protect the traffic that will produce future with a shared pairwise key of key information establishment separately.

Existing wireless sensor network symmetric key Managed Solution can be divided into two kinds: stochastic pattern and definite type.To be exactly network obtain part set of cipher key and become key ring disposing early stage node stochastic pattern from pool of keys with random fashion, and after disposing, adjacent node is shared public keys with certain probability.Determine that type is exactly for each node and other arbitrary node are set up independently pairwise key.

Above-mentioned two schemes is all static key distribution scheme, and the shortcoming of existence is: the random topology environment that 1. can not adapt to WSN completely; 2. the following several attacks that can not avoid agreement to face completely: false routing iinformation, Sybil attack, Hello Flood attack, collusion, selective forwarding, Sinkhole attack; 3. key distribution scheme not and Routing Protocol perfectly combine.

Summary of the invention

The object of this invention is to provide a kind of wireless sensor network adaptive mesh safety routing method, can effectively overcome the shortcoming that prior art exists.

Wireless sensor network adaptive mesh safety routing method is achieved in that to it is characterized in that operating procedure as follows:

A. netinit

A-1. node is before disposing, and base station server generates pool of keys, and pool of keys is comprised of L bar one-way hash chain, and every hash chain comprises M chain key, and pool of keys size is L * M.

A-2. Redundant Control

Before each takes turns circulation, first utilize base station that monitored area is subdivided into N equal-sized grid, in each grid, choose the mean value of all node energies as fiducial value E _aver, reject energy lower than the node of fiducial value, rejecting node will be in resting state as to be selected bunch of head, then using all the other nodes as leader cluster node to be selected and give ID numbering.Have ID label node we be referred to as preferential node.

B. election of cluster head

B-1. utilize the election of cluster head algorithm of LEACH to carry out after election of cluster head finishes, each to be selected bunch of head is chosen at random q chain key from pool of keys, when preserving the commitment value of hash chain and chain sign with function F (x _n| c _i) be encrypted.Cluster process, leader cluster node CH _iat q message { ID of the whole network broadcast _ch, S, G, F (x _n| C _i), message comprises node identification, chain key seed, growth factor, message utilizes F (x _n| c _i) encrypt.Non-leader cluster node receives after message, utilizes RSSI to calculate the distance d of bunch head and generate formula according to the key in steps A and carries out the authentication of chain key, if chain key authentication passes through, generates a derivative key K _{i, j, w}=F (K _i,j|| ID _chi).In bunch the chain key of all the other leader cluster nodes to be selected will be retained by node stand-by, but node itself remains and can use derivative key.Non-leader cluster node, after generating derivative key, send and confirms to join message from the shortest bunch hair of its distance of trend, and message comprises apart from d, node ID, derivative key K _{i.j, w}.

B-2. in a second cluster election process of data transfer phase, preferential selection has to be selected bunch of head of ID numbering.Take turns after bunch capitiform becomes when one, bunch head re-starts numbering to having to be selected bunch of head of preferential ID numbering in its place bunch according to following formula:

T(n)=0.6E(n)+0.4H(n) （2）

E (n) is residue energy of node, and H (n) is the distance of nodal distance base station.After prevariety intra-cluster head lost efficacy, according to arbitrary bunch of head of ID number order election.After the to be selected bunch of head that has ID numbering in arbitrary bunch was all elected to bunch head, network restarts new one and takes turns election.

After in bunch, new bunch capitiform becomes, leader cluster node CH _iat q message { ID of the whole network broadcast _ch, S, G, F (x _n| C _i), all the other nodes restart chain key authentication and generate derivative key.The leader cluster node to be selected of not being elected as bunch head still retains chain key, is elected as the node of bunch head and chain key can be abandoned.

C. election of cluster head after out non-leader cluster node according to the broadcasting packet receiving, select to add by a nearest bunch head for authentication, and derive with leader cluster node between the key of communicating by letter.

D. set up communication path: between bunch head, carry out multi-hop or single-hop transfer of data, the communication key between bunch head is its chain key phase XOR value.

E. set up dynamic key management mechanism: Yi Cuwei unit, according to bunch interior nodes number, set up suitable EBS structure, and by single-hop security process, EBS structure is kept in base station, bunch interior nodes by bunch in key generate node and know the managing keys number in EBS structure to base station login, generate managing keys and again to bunch in destroy after broadcast.

F. data transfer phase: non-leader cluster node image data in bunch, send to a bunch head, leader cluster node carries out giving next bunch head after fusion treatment to data, until send to base station.In any bunch, all preferential leader cluster nodes to be selected were all elected to after bunch head, re-started B and circulated to F.

A kind of wireless sensor network adaptive mesh safety routing method of the present invention, comprises that malicious node catches the recovering step after normal node:

(1) a bunch interior nodes is hunted down

1. a bunch interior nodes is hunted down and does not form collusion:

In bunch, broadcast be hunted down in the upgrading bunch key of node of m packet.Utilize key to generate node according to k _s2generate new managing keys K ' with one-way function _s2, and bunch in broadcast, therefore the node that is hunted down does not obtain m packet cannot decipher the managing keys that obtains oneself, so is ejected out network.The one-way function is here a random function f, meets for any one x that belongs to the domain of definition, can obtain f(x)=y; For the x that does not belong to the domain of definition, on calculating, can not obtain y=f(x).

2. a bunch interior nodes is hunted down and forms collusion:

Key generates node and between bunch interior nodes of different distance and leader cluster node, sets up unique collusion and recover key, and with sending to preferential leader cluster node to be selected after session key, preferential leader cluster node to be selected is forwarded to base station.

The present invention adopts other preferential node at large in the cluster network grid in early stage to recover captive node.

(2) leader cluster node is hunted down

After leader cluster node is hunted down, base station utilizes a to be selected bunch of ID of storage re-start in single bunch election of cluster head and set up new key distribution scheme, and idiographic flow as shown in Figure 1.

The present invention has improved traditional LEACH Routing Protocol and has allocated the scheme of private cipher key in advance for each node, innovation add Redundant Control, adaptive mesh secondary election of cluster head and preferential leader cluster node to be selected random from pool of keys, choose q chain key.

Node in each bunch, according to the message of the leader cluster node broadcast in one's respective area, utilizes hash function checking legitimacy, adds nearby bunch, by random function, derives communication key.

EBS key management system, in conjunction with practical problem, be take bunch as the suitable EBS parameter of contractor selection, make to meet

wherein n is interstitial content, and k is the managing keys number of each node of distribution, and k+m is managing keys sum, and this process is also to adopt the form that sends message, adjustment key code system that like this can be spontaneous in the situation that network is under attack.

Introduce adaptive mesh, according to energy and geographical position, preferentially indicate leader cluster node to be selected, utilize secondary election of cluster head to extend network lifetime and the recovery leader cluster node that is hunted down.

The present invention adopts above technical scheme compared with prior art, has following technique effect:

1. utilize pool of keys, for each leader cluster node distributes some keys, other node utilizes hash function authentication legitimacy and derives communication key with random function, reduces traffic load, strengthens the correlation of key management.

2. adopt Redundant Control mechanism, dispose base station in earlier stage network integral body is carried out to network grid division, select preferential leader cluster node to be selected, reduce the energy consumption of cluster process, extend network lifetime.

3. adopt a second cluster election mechanism, utilize the to be selected bunch of head that has ID numbering electing in grid early stage to carry out secondary election of cluster head, optimize election mechanism, reduce election energy loss.

4. connection setup multi-hop, the single-hop mechanism that mutually combines bunch.

5. dynamic management key code system, network rear spontaneous adjustment managing keys system under attack, guarantees the accuracy of transfer of data to improve internet security.

6. adaptive mesh mechanism, is hunted down in leader cluster node recovery process, utilizes the leader cluster node preferential to be selected in grid, has avoided large-scale bunch of nose heave choosing, has saved energy, has extended network lifetime.

Accompanying drawing explanation

Fig. 1 is bunch head recovery flow chart that is hunted down

Fig. 2 is method for routing flow chart of the present invention

Fig. 3 is the network topology structure figure that removes redundant node

Fig. 4 is the dead node temporal evolution of network curve chart

Fig. 5 is network energy consumptions profile

Fig. 6 is that network base station receives data and curves figure

In figure: 1 '---captive bunch of head; 2 '---preferential to be selected bunch of head; 3 '---former captive bunch of head; 4 '---bunch head that secondary election of cluster head generates; 5 '---eliminate the malicious node of network; 6 '---new key more, reject the network after malicious cluster heads node; 7 '---ordinary node

Concrete execution mode

Introduce in detail the safe routing method of wireless sensor network adaptive mesh of the present invention below.

In the present invention of side signal transmission face, adopt free space model and multipath decline model, when the distance of sending node and receiving node is less than certain value, adopt free space model, otherwise adopt multipath fading model, according to the energy loss ratio of mentioning in document, node sends the big or small energy consuming for the information of K-bit and is:

$E_{\mathrm{send}}=\left\{\begin{array}{c}k*E_{\mathrm{elec}}+k*{\mathrm{\&xi;}}_{\mathrm{fs}}*d^2(d<d_0)\\ k*E_{\mathrm{elec}}+k*{\mathrm{\&xi;}}_{\mathrm{mp}}*d^4(d>d_0)\end{array}\right.---\left(3\right)$

The energy that the information of sensor node reception K-bit consumes is:

E _receive=K*E _elec (4)

D in formula 3 ₀by following formula, determined:

${\mathrm{<figure-callout\; id="0"\; label="d"\; filenames="HDA0000419670070000021.png,HDA0000419670070000022.png"\; state="\{\{state\}\}">d</figure-callout>}}_0=\sqrt{\frac{{\mathrm{\&epsiv;}}_{\mathrm{fs}}}{{\mathrm{\&epsiv;}}_{\mathrm{mp}}}}---\left(5\right)$

ε wherein _fsand ε _mprespectively that amplifier is less than d in transmission range ₀be more than or equal to d ₀time multiplication factor, E _sendbe the energy that reception and transtation mission circuit consume, d is transmission range.From formula 3, can find out that suitable transmission range plays very large help for the communication energy consumption of saving node.

As shown in Figure 1, LEACH-SEED agreement performing step of the present invention is as follows:

A. netinit in earlier stage

A-1.100 random being distributed in the region that will monitor of sensor node, static node, node primary power is 1.5J, aggregation node is at monitored area middle position.Each node has the fixing ID identifier of oneself, and node set is also N={m ₁..., m _j..., m ₁₀₀.Node is before disposing, and base station server generates pool of keys, and pool of keys is comprised of L bar hash chain, and every hash chain comprises M chain key and growth factor G is separate, and node shares a seed seed.J chain key generating mode of this hash chain is as follows:

K _i,j=H ^j(seed,G _i) （6）

H wherein ^j(seed, G _i)=H (H ^j-1(seed, G _j-1), G _i) and the value of j between 1 to M.

A-2. before each takes turns circulation, first utilizing base station that monitored area is subdivided into N(size can determine according to concrete network, the words N that node is intensive can be a little bigger, here we to choose N be 16) individual equal-sized grid, in each grid, choose the mean value of all node energies as fiducial value E _aver, reject energy lower than the node of fiducial value, rejecting node will be in resting state as to be selected bunch of head, then using all the other nodes as leader cluster node to be selected and be endowed ID numbering.Have ID label node we be referred to as preferential node, ID numbering data send to base station and preserve after first round network is set up.

B. cluster and election of cluster head thereof

B-1. in preferential range of nodes, utilize LEACH election of cluster head algorithm to carry out after election of cluster head finishes, for each leader cluster node to be selected is chosen at random q chain key from pool of keys, use function F (x when preserving the commitment value of hash chain and chain sign _n| c _i) be encrypted.Leader cluster node CH _iat q message { ID of the whole network broadcast _ch, S, G, F (x _n| C _i), message comprises node identification, chain key seed, growth factor, message utilizes F (x _n| c _i) encrypt.Non-leader cluster node receives after message, utilizes RSSI to calculate bunch head and carries out the authentication of chain key apart from the distance d of base station and according to formula (6), if chain key authentication passes through, generates a derivative key K _{i, j, w}=F (K _i,j|| ID _chi).In bunch the chain key of all the other leader cluster nodes to be selected will be retained by node stand-by, but node itself remains and can use derivative key.Non-leader cluster node, after generating derivative key, send and confirms to join message from the shortest bunch hair of its distance of trend, and message comprises apart from d, node ID, derivative key K _{i.j, w}, to this cluster, complete.

B-2. one, take turns after network sets up, bunch head renumbers according to following formula the preferential node that has ID numbering in place bunch.

T(n)=0.6E(n)+0.4H(n) （7）

E (n) is residue energy of node, and H (n) is the distance of nodal distance base station.After prevariety intra-cluster head lost efficacy, according to arbitrary bunch of head of ID number order election.After the to be selected bunch of head that has ID numbering in arbitrary bunch was all elected to bunch head, network restarts new one and takes turns election.After in bunch, new bunch capitiform becomes, leader cluster node CH _iat q message { ID of the whole network broadcast _ch, S, G, F (x _n| C _i), all the other nodes restart chain key authentication and generate derivative key.The leader cluster node to be selected of not being elected as bunch head still retains chain key, is elected as the node of bunch head and chain key can be abandoned.

C. election of cluster head after out non-leader cluster node according to the broadcasting packet receiving, select to add in its bunch by a nearest bunch head for authentication, and derive with leader cluster node between the key of communicating by letter.

D. set up communication path

Route establishment between carrying out bunch, selects multi-hop or single-hop path, and the communication key between leader cluster node is its chain key phase XOR value.

E. set up dynamic key management

In fact one-way hash chain is exactly a hashed value sequence { x ₁..., x _j..., x _n, this sequence meets

hash function H meets lower Column Properties: (1) given x _j-1be easy to calculate x with G _j; (2) not given x _j-1with any one in G, be difficult to calculate x _j.G is called growth factor, last element x of hash chain _nbe called as commitment value, clearly every hash chain has unique x _n, other element is referred to as chain key.

EBS is a kind of communication key management method based on combinatorial principle being put forward by people such as Eltoweissy.Known when EBS matrix according to pertinent literature

time, matrix has

individual column vector, in matrix, n compound mode just can form EBS (n, k, m) arbitrarily, forms a key distribution scheme.By being broadcast to many m packet, just can cancel and upgrade whole keys that arbitrary node has, thereby this node is evicted out of to network.

After path is set up, bunch head is selected suitable EBS parameter K and M according to a bunch interior nodes number n, and EBS structure is kept in base station.Bunch interior nodes is sent logon message to a bunch hair, and logon message comprises node ID, and apart from d, message utilizes the communication key K of bunch interior nodes and bunch head _{i, j, w}encrypt, then leader cluster node carries out secure registration to base station, and process is as follows: bunch head sends an application message to base station, the communication key of bunch head and base station is K _p,j, K _p,jalso belong in q chain key, base station generates registered seed S after receiving message ₁, base station communication key K _p,jafter encryption, send to a bunch head, bunch head bunch in broadcast seed, all the other nodes utilize one-way Hash function to generate login key K _s1, node is with login key encrypting registration message and send request message, leader cluster node CH _iaccording to registered seed S ₁with one-way Hash function authentication node legitimacy, and bunch in record node ID and apart from d, last bunch of head K _p,jencrypting registration message is registered to base station.

After EBS structure sets, utilize key to generate node, mainly generate managing keys and collusion recovery key.Here we choose at random one and generate node as key in preferential leader cluster node, and the communication process that leader cluster node and key generate between node is as follows: leader cluster node CH _ito base station, send message, base station receives message and generates seed S ₂and send message to a bunch head, and bunch head generates node at transmission message to key, and key generates node and generates encryption key distribution to bunch head.Bunch interior nodes obtains the managing keys of oneself according to the cura generalis key of EBS matrix and bunch head broadcast.After the secondary election of cluster head that the present invention takes turns at each finishes, leader cluster node CH _iat q message { ID of the whole network broadcast _ch, S, G, F (x _n| C _i), all the other nodes restart chain key authentication and generate derivative key.

F. data transfer phase: non-leader cluster node image data in bunch, send to a bunch head, leader cluster node carries out giving next bunch head after fusion treatment to data, until send to base station.In any bunch, all preferential leader cluster nodes to be selected were all elected to after bunch head, re-started B to the circulation of F.

Restoration Mechanism after network is under attack

For node cheating, attack, the hash chain that this method is used can be isolated this attack, because assailant cannot puppet produce seed and growth factor, therefore cannot obtain the authentication of normal node, thereby cannot carry out node cheating attack.

But bunch interior nodes is hunted down and does not form in the situation of collusion in attacking for Equations of The Second Kind, by bunch in broadcast the key of the node that is hunted down in m packet renewal bunch.Utilize key to generate node according to k _s2new managing keys K ' with one-way function generation _s2, and bunch in broadcast, therefore the node that is hunted down does not obtain m packet cannot decipher the managing keys that obtains oneself, so is ejected out network, other key is similar.

In attacking for Equations of The Second Kind, bunch interior nodes is hunted down and forms in the situation of collusion, has introduced a collusion and has recovered key K _r.Here can adopt node at large in the cluster network grid in early stage to recover captive node.After leader cluster node is hunted down, base station utilizes a to be selected bunch of ID of storage to carry out again election of cluster head and set up new key distribution scheme.Idiographic flow as shown in Figure 2.

Simulated environment and parameter configuration

Network model is herein as follows: what sensor node was random is distributed in the region that will monitor, static node, and node primary power is 1.5J.Base station location is fixed, and between node, communication can pass through multi-hop or single-hop mode, and traffic model is single order radio communication model, and it is 5000 to take turns that number is taken turns in emulation.Simulated environment is Matlab R2011a, and according to general optimum bunch ratio, this test election of cluster head probability is 5%.Physical simulation parameter is in Table 1.

Table 1 analog parameter table

Performance evaluation criterion and experimental performance analysis

Emulation experiment is just for network situation under attack, malicious node is caught normal node, obtains network management scheme, and so we just analyze network recovery ability, and contrast the data volume that analysis network life, network energy consumption, base station receive with original LEACH agreement.

Fig. 3 is the network topology structure figure that removes redundant node.In figure, circle represents preferential bunch node to be selected, and cross represents redundant node, and middle position asterisk represents base station.In this MATLAB experiment simulation, node energy is 1.5J for the first time, in first round of voting, does not have redundant node.Fig. 3 the 500th takes turns and removes redundant node design sketch.Can find out, LEACH-SEED removes redundant node, makes preferentially to choose high-energy node in network later stage election of cluster head process, and the election of cluster head in simplified network circulation is machine-processed, conserve energy.

Fig. 4 is the dead node temporal evolution of network curve chart.In this figure, transverse axis represents to take turns number, and vertical pivot represents node mortality, and what dotted line represented is the dead variation diagram of network node that uses LEACH, and what solid line represented is the dead variation diagram of network node that uses LEACH-SEED.LEACH agreement is dead first node when 750 take turns, dead 100 nodes when 5000 take turns, radio sensing network paralysis.Improved LEACH-SEED agreement is dead first node when 1100 take turns, dead 90 nodes when 5000 take turns, and the dead interstitial content overall trend of wireless sensor network is mild.

Fig. 5 is network energy consumptions profile.In this figure, transverse axis represents it is wheel number, and vertical pivot represents it is the energy of network consumption.Can it is evident that, in same time, use the network power consumption of LEACH-SEED agreement few.This is because LEACH-SEED has increased control information and security mechanism improves the survival rate of node greatly, has saved network energy consumption, makes every network energy consumption of taking turns more balanced.

Fig. 6 is that network base station receives data and curves figure.Can find out, within the same time, use the data volume of the network reception that improves the safe Routing Protocol of LEACH large.Reason is that LEACH agreement is used bunch head directly and base station communicates, and after a bunch head is hunted down, produces the misleading in path, and packet cannot be delivered in time.Agreement after improvement adds key management system, and after a bunch head is hunted down, preferential leader cluster node to be selected adds automatically, and distributing key system again, guarantees the timely transmission of data.

The present invention has proposed a kind of wireless sensor network adaptive mesh safety routing method on the basis of original LEACH agreement, with base station grid division form, carrying out network area division carries out Redundant Control and selects preferential leader cluster node to be selected with this, preferential select leader cluster node to be selected for bunch head the chain key that distributes it to be formed by hash chain, increase a second cluster election mechanism of utilizing preferential to be selected bunch of head, in bunch, ordinary node forms new communication key according to one-way Hash function.Adopt dynamic key management mechanism and utilize the adaptive mesh mechanism node that is hunted down to recover.This new LEACH-SEED Routing Protocol can effectively be resisted the attack of malicious node, and the fast quick-recovery node that is hunted down, extends network lifetime, makes dead node time equalization, improves capacity usage ratio.

Claims (1)

1. a wireless sensor network adaptive mesh safety routing method, is characterized in that operating procedure is as follows:

A. netinit:

A-1. node is before disposing, and base station server generates pool of keys, and pool of keys is comprised of L bar one-way hash chain, and every hash chain comprises M chain key, and pool of keys size is L * M;

A-2. Redundant Control:

Before each takes turns circulation, first utilize base station that monitored area is subdivided into N equal-sized grid, in each grid, choose the mean value of all node energies as fiducial value E _aver, reject energy lower than the node of fiducial value, rejecting node will be in resting state as to be selected bunch of head, then using all the other nodes as leader cluster node to be selected and give ID numbering, have ID label node we be referred to as preferential node;

B. election of cluster head:

B-1. utilize LEACH algorithm to carry out after election of cluster head finishes, each to be selected bunch of head is chosen at random q chain key from pool of keys, when preserving the commitment value of hash chain and chain sign with function F (x _n| c _i) be encrypted cluster process, leader cluster node CH _iat q message { ID of the whole network broadcast _ch, S, G, F (x _n| C _i), message comprises node identification, chain key seed, growth factor, message utilizes F (x _n| c _i) encrypt, non-leader cluster node receives after message, utilizes RSSI to calculate the distance d of bunch head and generate formula according to the key in steps A and carries out the authentication of chain key, if chain key authentication passes through, generates a derivative key K _{i, j, w}=F (K _i,j|| ID _chi), in bunch the chain key of all the other leader cluster nodes to be selected will be retained by node stand-by, but node itself remains and can use derivative key, non-leader cluster node is after generating derivative key, from the shortest bunch hair of its distance of trend, send and confirm to join message, message comprises apart from d, node ID, derivative key K _{i.j, w};

B-2. in a second cluster election process of data transfer phase, preferential to be selected bunch of head selecting to have ID numbering, takes turns after bunch capitiform becomes when one, and bunch head re-starts numbering to having to be selected bunch of head of preferential ID numbering in its place bunch according to following formula

T(n)=0.6E(n)+0.4H(n) （2）

E (n) is residue energy of node, H (n) is the distance of nodal distance base station, after prevariety intra-cluster head lost efficacy, according to the lower cluster head of ID number order election, after the to be selected bunch of head that has ID numbering in any bunch was all elected to bunch head, network restarts new one and takes turns election, after in bunch, new bunch capitiform becomes, and leader cluster node CH _iat q message { ID of the whole network broadcast _ch, S, G, F (x _n| C _i), all the other nodes restart chain key authentication and generate derivative key, and the leader cluster node to be selected of not being elected as bunch head still retains chain key, is elected as the node of bunch head and chain key can be abandoned;

C. election of cluster head after out non-leader cluster node according to the broadcasting packet receiving, select to add by a nearest bunch head for authentication, and derive with leader cluster node between the key of communicating by letter;

D. set up communication path: between bunch head, carry out multi-hop or single-hop transfer of data, the communication key between bunch head is its chain key phase XOR value;

E. set up dynamic key management mechanism: Yi Cuwei unit, according to bunch interior nodes number, set up suitable EBS structure, and by single-hop security process, EBS structure is kept in base station, bunch interior nodes by bunch in key generate node and know the managing keys number in EBS structure to base station login, generate managing keys and again to bunch in destroy after broadcast;

F. data transfer phase: non-leader cluster node image data in bunch, send to a bunch head, leader cluster node carries out giving next bunch head after fusion treatment to data, until send to base station, in any bunch, all preferential leader cluster nodes to be selected were all elected to after bunch head, re-started B and circulated to F;

In aforesaid operations step, malicious node is caught the recovering step after normal node and is:

(1) a bunch interior nodes is hunted down

A. a bunch interior nodes is hunted down and does not form collusion:

In bunch, broadcast be hunted down in the upgrading bunch key of node of m packet, utilize key to generate node according to k _s2generate new managing keys K ' with one-way function _s2, and bunch in broadcast, therefore the node that is hunted down does not obtain m packet cannot decipher the managing keys that obtains oneself, therefore be ejected out network, the one-way function is here a random function f, meets for any one x that belongs to the domain of definition, can obtain f(x)=y; For the x that does not belong to the domain of definition, on calculating, can not obtain y=f(x);

B. a bunch interior nodes is hunted down and forms collusion:

Key generates node and between bunch interior nodes of different distance and leader cluster node, sets up unique collusion and recover key, and with sending to preferential leader cluster node to be selected after session key, preferential leader cluster node to be selected is forwarded to base station;

The present invention adopts other preferential node at large in the cluster network grid in early stage to recover captive node;

(2) leader cluster node is hunted down

After leader cluster node is hunted down, base station utilizes a to be selected bunch of ID of storage re-start in single bunch election of cluster head and set up new key distribution scheme:

EBS key management system, in conjunction with practical problem, be take bunch as the suitable EBS parameter of contractor selection, make to meet

wherein n is interstitial content, and k is the managing keys number of each node of distribution, and k+m is managing keys sum, and this process is also to adopt the form that sends message, adjustment key code system that like this can be spontaneous in the situation that network is under attack;

Introduce adaptive mesh, according to energy and geographical position, preferentially indicate leader cluster node to be selected, utilize secondary election of cluster head to extend network lifetime and the recovery leader cluster node that is hunted down.

Images