CN105898784A - Fault-tolerant reparation method and system of wireless sensor network - Google Patents

Abstract

The invention discloses a fault-tolerant reparation method and system of a wireless sensor network. The method comprises following steps of obtaining wireless sensor nodes, and coordinates and communication radiuses of the nodes; determining an external convex hull of the nodes; calculating a center node coordinate and side lengths of the external convex hull; connecting the inflection points and the center node of the external convex hull; determining the nodes and quantity of the nodes needing to be deployed on the connection lines according to the communication radiuses; deploying the nodes at upper parts of the sides of an internal convex hull, wherein three-connection is carried out on the inflection points of the internal convex hull; obtaining non- adjacent sides on the external convex hull; deploying the nodes on the non- adjacent sides, thus forming two-connection. Through adoption of the mode, the connection reparation problem of the wireless sensor network can be solved from different angles; the fault tolerance of the network is realized; the number of the deployed nodes is reduced; the larger coverage area and better topology quality can be realized; the follow-up attack and damage can be resisted better; and the practical application demand of the wireless sensor network can be satisfied better.

Description

The fault-tolerant restorative procedure of wireless sensor network and system

Technical field

The present invention relates to wireless sensor technology field, especially relate to the fault-tolerant restorative procedure of a kind of wireless sensor network And system.

Background technology

Wireless sensor network is due to the bad environments disposed and the battery electric quantity that carries of node is limited etc. that reason is easy Break down.If network topology has fault tolerance, then, when the fault of network generation small-scale, network can continue to normally Run, it is not necessary to artificial reparation in time, it is not necessary to buy new equipment, the most just saved manpower and materials.So, build one and have The network topology of fault tolerance is the most important for the wireless sensor network being susceptible to fault.But with without fault tolerance Network topology compare, the network topology structure with fault tolerance is increasingly complex, and the interstitial content of required consuming is more.Cause This, the main target of existing a lot of repair mechanisms is how to minimize the interstitial content of deployment.But, these algorithms are building When there is the network topology of fault tolerance, it is still desirable to expend more node, and other performances of network are (such as the network coverage Area and network topology quality etc.) improved the most well.

Some algorithm is although it is contemplated that other performances of network, but the fault-tolerant energy of the network topology constructed by these algorithms Power remains unchanged the highest.As a example by 2C-SpiderWeb algorithm, should be by one network topology being similar to " polygonal star " structure of structure To realize the fault freedom of network.Although this algorithm is repairing the topological mass problem that considers while network connects, but exists Problem below: as a example by " five-pointed star " topological structure, in the position the closer to 5 summits of five-pointed star, the distance between node The nearest, the limit of node is the most, and therefore the average node degree of node is the highest.But when there is fault on a small scale, such as, near top Break down in the position of point, then whole " five-pointed star " is divided into 2 or the sub-block of more than 2 by being easy to.Therefore, use Although the network topology that 2C-SpiderWeb algorithm is constructed has higher average node degree, but in such network topology Under structure, higher average node degree can not represent preferable fault freedom.Additionally, due to the closer to the nodes of locations on summit Distance is the nearest, causes there is more covering overlapping area between node so that the total area coverage of network after reparation is not Greatly.In a word, 2C-SpiderWeb algorithm is although it is contemplated that network topology quality problems, but network topology that this algorithm is formed Network after structure can not tolerate small-scale fault next time, and this algorithm reparation well has bigger covering Overlapping area.

Summary of the invention

The technical problem to be solved is: provide a kind of fault-tolerant recovery scenario of wireless sensor network, with not Solve wireless sensor network with angle and connect reparation problem, and achieve the fault-tolerance of network respectively.

In order to solve above-mentioned technical problem, the technical solution used in the present invention is: provide a kind of wireless sensor network Fault-tolerant restorative procedure, including:

Obtain wireless sensor node and coordinate, communication radius；

Determine the outer convex closure of node, and calculate the Centroid coordinate of described outer convex closure and each length of side；

Respectively outer convex closure flex point is connected with Centroid, according to communication radius, determine need on line dispose node and Its quantity；

Convex bound edge is disposed node so that interior convex closure flex point is three connections；

Obtain evagination and wrap non-conterminous limit, and dispose node thereon, form two connections.

For solving the problems referred to above, the present invention also provides for the fault-tolerant restorative procedure of a kind of wireless sensor network, including:

Obtain wireless sensor node and coordinate, communication radius；

Determine the outer convex closure of node, and calculate the Centroid coordinate of described outer convex closure and each length of side；

Respectively outer convex closure flex point is connected with Centroid, according to communication radius, determine need on line dispose node and Its quantity；

Convex bound edge is disposed node so that interior convex closure flex point is three connections；

Node is disposed outside so that outer convex closure flex point is three connections on convex closure limit.

For solving the problems referred to above, the present invention also provides for the fault-tolerant repair system of a kind of wireless sensor network, including:

Acquisition module, is used for obtaining wireless sensor node and coordinate, communication radius；

Outer convex closure module, for determining the outer convex closure of node, and calculates the Centroid coordinate of described outer convex closure and each limit Long；

Outer convex closure deployment module, for outer convex closure flex point being connected with Centroid respectively, according to communication radius, determines even Need on line to dispose node and quantity thereof；

Interior convex closure deployment module, for disposing node so that interior convex closure flex point is three connections on convex bound edge；

Outer convex closure deployment module is additionally operable to obtain evagination and wraps non-conterminous limit, and disposes node thereon, forms two Connect.

The beneficial effects of the present invention is: be different from prior art, the present invention is after determining the outer convex closure of node, in inside Dispose node, and on convex bound edge and evagination bound edge, dispose node so that inside/outside convex closure flex point is three connections or two even Connect.By the way, the present invention solves wireless sensor network with different angles and connects reparation problem, and achieves respectively The fault-tolerance of network, had both reduced the interstitial content of deployment, had bigger area coverage and more preferable topological mass, more resistant against Follow-up attack and destruction, more meet the practical application request of wireless sensor network.

Accompanying drawing explanation

Fig. 1 is the schematic flow sheet of the fault-tolerant restorative procedure embodiment one of wireless sensor network of the present invention；

Fig. 2 is the schematic flow sheet of the fault-tolerant restorative procedure embodiment two of wireless sensor network of the present invention；

Fig. 3 is to dispose node in the specific embodiment of the invention to calculate triangle schematic diagram；

In the case of Fig. 4 is same communication radius, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton The via node number contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm；

In the case of Fig. 5 is same sub-block number, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton The via node number contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm；

In the case of Fig. 6 is same communication radius, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton Total area coverage contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm；

In the case of Fig. 7 is same sub-block number, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton Total area coverage contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm；

In the case of Fig. 8 is same communication radius, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton Each node average area coverage contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm；

In the case of Fig. 9 is same sub-block number, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton Each node average area coverage contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm；

In the case of Figure 10 is same communication radius, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton The average node degree contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm；

In the case of Figure 11 is same sub-block number, the F2CR algorithm of the present invention and P3CR algorithm and existing Hamilton The average node degree contrast schematic diagram of Path algorithm and 2C-SpiderWeb algorithm.

Detailed description of the invention

By describing the technology contents of the present invention in detail, being realized purpose and effect, below in conjunction with embodiment and coordinate attached Figure is explained.

The design of most critical of the present invention is: connect restorative procedure F2CR and portion by wireless sensor network whole two Divide three connection restorative procedure P3CR, solve wireless sensor network with different angles and connect reparation problem, and achieve respectively The fault-tolerance of network.

Refer to Fig. 1, the embodiment of the present invention one provides the fault-tolerant restorative procedure of a kind of wireless sensor network, including:

Obtain wireless sensor node and coordinate, communication radius；

Determine the outer convex closure of node, and calculate the Centroid coordinate of described outer convex closure and each length of side；

Respectively outer convex closure flex point is connected with Centroid, according to communication radius, determine need on line dispose node and Its quantity；

Convex bound edge is disposed node so that interior convex closure flex point is three connections；

Obtain evagination and wrap non-conterminous limit, and dispose node thereon, form two connections.

Being different from prior art, the embodiment of the present invention one is after determining the outer convex closure of node, at On-premise node, and Node is disposed so that interior convex closure flex point is three connections or the connection of outer convex closure flex point two on convex bound edge and evagination bound edge.Logical Crossing aforesaid way, the embodiment of the present invention one solves wireless sensor network and connects reparation problem, by opening up of structure fan-shaped Flutter structure to reduce the interstitial content of deployment, thus reduce node deployment cost, there is bigger area coverage and preferably open up Flutter quality, more resistant against follow-up attack and destruction, more meet the practical application request of wireless sensor network.

As in figure 2 it is shown, the embodiment of the present invention two provides the fault-tolerant restorative procedure of a kind of wireless sensor network, including:

Obtain wireless sensor node and coordinate, communication radius；

Determine the outer convex closure of node, and calculate the Centroid coordinate of described outer convex closure and each length of side；

Respectively outer convex closure flex point is connected with Centroid, according to communication radius, determine need on line dispose node and Its quantity；

Convex bound edge is disposed node so that interior convex closure flex point is three connections；

Node is disposed outside so that outer convex closure flex point is three connections on convex closure limit.

Being different from prior art, the embodiment of the present invention two is after determining the outer convex closure of node, at On-premise node, and Node is disposed so that interior convex closure flex point is three connections or the connection of outer convex closure flex point two on convex bound edge and evagination bound edge.Logical Crossing aforesaid way, the embodiment of the present invention two solves wireless sensor network and connects reparation problem, by constructing the net of Crossed Circle Network topology improves the fault-tolerance of network.Compared with embodiment one, the network topology constructed by embodiment two has preferably appearance Wrong ability.It is applicable to fault-tolerance and the top-priority situation of area coverage of network, there is bigger area coverage with more preferable Topological mass, more resistant against follow-up attack and destruction, more meets the practical application request of wireless sensor network.

Understand two kinds of methods of the present invention for convenience, further illustrate below in conjunction with accompanying drawing.

When wireless sensor network is divided into multiple sub-block not communicated relatively far apart and mutually, need to take effectively Network is repaired by node deployment algorithm in time.Owing to network is susceptible to the characteristic of fault, at the node that can be used for deployment In the case of number more abundance, while repairing network connection with minimum cost, should make every effort to so that the network after repairing There is fault tolerance.But this is two conflicting requirements, one is that the interstitial content that requirement is disposed minimizes, and two is requirement Network after reparation has fault-tolerance.The network with fault tolerance generally requires the node that deployment is more, disposes less joint The network of point does not the most have fault-tolerance.It is thus desirable to the requirement to these two aspects is weighed, effective scheme is proposed so that While disposing relatively little of node, it is achieved network has the target of fault tolerance.

1. related definition

Definition 1, Minimum Convex Closure: node set S of random distribution in a given plane, find out comprise minimum a little Convex polygon, then this polygon is referred to as Minimum Convex Closure.The present invention needs to build two convex closures, for convenience of description, here by minimum Convex closure is referred to as outer convex closure.Outer convex closure is made up of each isolated sub-block in network.

Definition 2, interior convex closure: so that the network after Xiu Fuing has fault tolerance, the present invention is inside Minimum Convex Closure Build a convex closure again.This convex closure within Minimum Convex Closure is referred to as interior convex closure by the present invention.

Definition 3, convex closure flex point: the point of composition convex closure is referred to as convex closure flex point.The flex point of outer convex closure is each sub-block in network, Interior convex closure flex point is the node that the present invention needs to dispose.

2. step describes

2.1 whole two connect reparation algorithm

Whole two targets connecting reparation algorithm (F2CR) are to use minimum number destination node to build one to have fault-tolerant merit The network of energy so that all nodes after reparation have two connections.F2CR step is as follows:

Step 1. initializes.Assume in plane, there be n node (sub-block), the coordinate of all nodes and node to be disposed Communication radius R be known.

Step 2. finds outer convex closure.Outer convex closure is found out in all nodes.Assume that the outer convex closure found comprises m node, The coordinate of each node is known.Calculate Centroid coordinate and each length of side of convex closure of convex closure.By outer convex closure flex point and convex closure Center line one by one, calculates the interstitial content that every line segment needs to dispose.

Step 3. finds interior convex closure.From the beginning of outer convex closure flex point, dispose corresponding interstitial content towards convex closure center.Every Last node (i.e. near the node at convex closure center) of line segment is the flex point of interior convex closure.Edge part along interior convex closure Administration's node.Now, all flex points of interior convex closure are three connections.

The outer convex closure of step 4. forms two connections.Outside the limit set of convex closure is foundThe non-conterminous limit of bar, along These edge administration node.If the flex point number of outer convex closure is odd number, then a flex point will be had to be formed without two connections, now from interior Find on convex closure one and this flex point not conllinear a bit, dispose node along the line between 2 o'clock.So far, evagination wraps institute Some nodes are respectively formed two connections.

Perform above step, in the network topology formed, all of node all at least two connection.F2CR flows Journey figure is as shown in Figure 1.

F2CR algorithm specifically comprises the following steps that

S1. to point set S in find minimum convex closure.

Assume that obtaining outer convex closure flex point collection isCorresponding coordinate is

{(x₁,y₁),(x₂,y₂),…,(x_m,y_m)}.Being in the point set within outer convex closure is { s_m+1,s_m+2,…,s_n}。

1) centre coordinate (x of outer convex closure is calculated₀,y₀)

$x_0=\frac{1}{m}\underset{i=1}{\overset{m}{\Σ}}{x}_i;$

$y_0=\frac{1}{m}\underset{i=1}{\overset{m}{\Σ}}{y}_i.$

2) length of side of outer convex closure is calculated.

Following steps are performed for all of i=1 to m

J=i+1；

If i=m, perform following steps

J=1；

After having performed previous step, perform following steps

${\mathrm{side}}_i=\sqrt{{(x_i-x_j)}^2+{(y_i-y_j)}^2};$

Then the collection of evagination bound edge length is combined into { side₁,side₂,…,side_m}。

S2. by outer convex closure flex point and the convex closure line of centres, the length of every line is calculated:

1) following steps are performed for all of i=1 to m

The collection then obtaining these lines is combined into { l₁,l₂,…,l_m}.2) adjacent two are calculated The angle of bar line:

Following steps are performed for all of i=1 to m

J=i+1；

If i=m, perform following steps

J=1；

After having performed previous step, perform following steps

${\mathrm{\θ}}_i=arccos\left(\frac{l_i^2+l_{i+1}^2-{\mathrm{side}}_i^2}{2\×l_i\×l_{i+1}}\right);$

Calculate the node number that every line segment needs to dispose:

$x_i^{\′}=x_0\±\frac{R\×\left|x_0-x_i\right|}{2\×l_i\×\sin \frac{{\mathrm{\θ}}_i}{2}};$

$y_i^{\′}=y_0\±\frac{R\×\left|y_0-y_i\right|}{2\×l_i\×\sin \frac{{\mathrm{\θ}}_i}{2}};$

From convex closure flex point s_iStart, dispose node towards O direction, convex closure center, dispose a node every distance R, obtain n_iThese points are added in set P, and assume n-th by ' deployed position that individual node is corresponding_i' individual node is p_i, its coordinate is (x_i″,y_i″).The flex point collection of interior convex closure is combined into { p₁,p₂,…,p_m}。

3) node is disposed at the edge along interior convex closure.

Following steps are performed for all of i=1 to m

If i=m, perform following steps

J=1；

At p_iAnd p_i+1Between dispose node, by these point add set P in.

S4.1) if m%2=0, following steps are performed

Set { side from evagination bound edge length₁,side₂,…,side_mSelect in }The shortest non-conterminous of bar summation Limit, disposes node on limit, is added in set P by these points.

Otherwise perform following steps

A summit u ∈ S will be had on convex closure not form two connections.Now find distance u nearest also on convex bound edge And with the some v of u not conllinear, between u and v, dispose node equably, and these points added in set P.Set P is to dispose Set of relay nodes, during initialization, P is empty set.When disposing new node in a network, just the node of deployment is joined P In.Last P is exactly the set of the via node of all deployment in network.

Give an example: if any five summits A, B, C, D, E, it is assumed that A with B couples together, C with D couples together, then A, B, C, D define 2 connections, but E is but formed without 2 connections, then u refers to evagination and wraps the point being formed without two connections, i.e. In set S, certain does not forms the point of two connections, here this point is set to u.As an illustrative example, u refers to summit E.

S4.2) point being isolated in plane sets up two connections

For all of k=m+1 to n execution following steps:

For s_k∈ S finds nearest two some u ' and v ', u ', v ' ∈ P, respectively at s_kWith u ' and s_kAnd between v ' uniformly Node is disposed on ground, is added in set P by these points.

2.2 parts three connect repairs algorithm

F2CR algorithm can set up the network topology with fault-tolerance in the case of using less interstitial content.Therefore, F2CR algorithm is applicable to the less situation of interstitial content that can use.When available interstitial content is more sufficient, can be to F2CR Algorithm improves so that the network topology after reparation has higher fault-tolerance.Proposed here the reparation algorithm of improvement P3CR.P3CR algorithm is similar with F2CR algorithm, but the network after using P3CR algorithm reparation has the structure that part three connects. Part three connection means that the most all of sub-block the most at least has three connections, and the node disposed the most at least has two Connect.P3CR flow chart is as shown in Figure 2.

Specifically comprising the following steps that of P3CR algorithm

It is identical that S1 with S2 whole with wireless sensor network two connects restorative procedure.

S3 is that the point that evagination is wrapped sets up 3 connections.

1) following steps are performed for all of i=1 to m

At s_iAnd s_i+1Between dispose node.

2) point being isolated in plane sets up 3 connections

Following steps are performed for all of k=m+1 to n

For s_k∈ S finds nearest three some u ', v ', w ' (u ', v ', w ' are not arranged on the same straight line be) at 3

Respectively at s_kWith u ', s_kWith v ' and s_kAnd dispose node between w ' equably.

It is the improvement connecting two and repairing algorithms that algorithm is repaired in three connections.Wherein two connect the net repaired after algorithm is repaired Network there may be bigger covering overlapping area, but the network after algorithm reparation is repaired in three connections would not exist bigger Covering overlapping area.But compared to and connect reparation algorithm, three connect repair algorithms need to expend more node.So Algorithm is repaired in two connections and three connections are repaired algorithms and can be emphasized particularly on different fields in actual applications.

3. Algorithm Analysis

The coordinate of known outer each flex point of convex closure and convex closure centre coordinate are respectively { (x₁,y₁),(x₂,y₂),…,(x_n, y_n) and (x₀,y₀), the communication radius of via node is R, and assumes to dispose a node every distance R.

Theorem 1: when the coordinate of interior convex closure flex point isThe joint now disposed Count out minimum.

Prove: as it is shown on figure 3, set in Δ ABC A, B, C tri-point coordinates be respectively (x₁,y₁)、(x₂,y₂)、(x₀,y₀).Here A point and B point represent the different flex points of outer convex closure respectively, and C point represents the central point of convex closure.The algorithm of the present invention is from convex closure flex point (A point and B point) is disposed to the direction of central point (C point).Because the value of AB, AC, BC and R is fixed, the joint disposed to be made Count out minimum, then require that the total length of AD and BE is the shortest, i.e. the total length of CD and CE is the longest.Therefore, the asking of required proof Topic is converted into: ask so that the D point when total length of CD and CE is the longest and the coordinate figure of E point.

If ∠ is ACB=θ, the length of DE, CD, CE is respectively R, a, b, and ∠ CDE=α, ∠ CED=β, a, b, α, β are the unknown.

It is respectively by the known length obtaining AB, AC and BC

${\mathrm{side}}_1=\sqrt{{(x_1-x_2)}^2+{(y_1-y_2)}^2}---(3-1)$

$l_1=\sqrt{{(x_1-x_0)}^2+{(y_1-y_0)}^2}---(3-2)$

$l_2=\sqrt{{(x_2-x_0)}^2+{(y_2-y_0)}^2}---(3-3)$

The value being understood θ by (3-1), (3-2), (3-3) formula is

$\mathrm{\θ}=arccos\left(\frac{l_1^2+l_2^2-{\mathrm{side}}_1^2}{2\×l_1\×l_2}\right)---(3-4)$

From trigonometric function theorem

$\frac{R}{\sin \mathrm{\θ}}=\frac{a}{sin\mathrm{\α}}=\frac{b}{sin\mathrm{\β}}=\frac{a+b}{sin\mathrm{\α}+sin\mathrm{\β}}---(3-5)$

So

$a+b=\frac{sin\mathrm{\α}+sin\mathrm{\β}}{sin\mathrm{\θ}}R$

$\begin{array}{c}=\frac{2\sin \frac{\mathrm{\α}+\mathrm{\β}}{2}\cos \frac{\mathrm{\α}-\mathrm{\β}}{2}}{\sin \mathrm{\θ}}R\\ =\frac{2\sin \frac{\mathrm{\π}-\mathrm{\θ}}{2}\cos \frac{\mathrm{\α}-\mathrm{\β}}{2}}{\sin \mathrm{\θ}}R\\ =\frac{2\cos \frac{\mathrm{\θ}}{2}\cos \frac{\mathrm{\α}-\mathrm{\β}}{2}}{\sin \mathrm{\θ}}R\\ =\frac{\cos \frac{\mathrm{\α}-\mathrm{\β}}{2}}{\sin \frac{\mathrm{\θ}}{2}}R\end{array}---(3-6)$

Because θ and R be it is known that so, as α=β,Take maximum, then the value of a+b is maximum, now a= b.That is, as a=b, the total length that a+b obtains maximum, i.e. CD and CE is the longest.

As a=b, from trigonometric function theorem

$a=b=\frac{\frac{R}{2}}{sin\frac{\mathrm{\θ}}{2}}=\frac{R}{2\sin \frac{\mathrm{\θ}}{2}}---(3-7)$

Assume that D point coordinates is (x₁′,y₁'), then have

$\sqrt{{(x_1^{\′}-x_0)}^2+{(y_1^{\′}-y_0)}^2}=\frac{R}{2\sin \frac{\mathrm{\θ}}{2}}---(3-8)$

Because D point is on line segment AC, so having

$\frac{y_0-y_1}{x_0-x_1}=\frac{y_0-y_1^{\′}}{x_0-x_1^{\′}}---(3-9)$

Can be obtained by formula (4-8) and (4-9)

$x_1^{\′}=x_0\±\frac{R\×\left|x_0-x_1\right|}{2\×l_1\×sin\frac{\mathrm{\θ}}{2}}---(3-10)$

$y_1^{\′}=y_0\±\frac{R\×\left|y_0-y_1\right|}{2\×l_1\×sin\frac{\mathrm{\θ}}{2}}---(3-11)$

If x₁<x₀, then

$x_1^{\&prime;}=x_0-\frac{R\&times;\left|x_0-x_1\right|}{2\&times;l_1\&times;sin\frac{\mathrm{\&theta;}}{2}}---(3-12)$

Otherwise

$x_1^{\&prime;}=x_0+\frac{R\&times;\left|x_0-x_1\right|}{2\&times;l_1\&times;sin\frac{\mathrm{\&theta;}}{2}}---(3-13)$

y₁' situation and x₁' similar, E point coordinates ask method similar with D point.It is to say, when D point and the coordinate minute of E point It is notTime, CD's and CE Total length is the longest, and the total length of AD and BE is the shortest, and the interstitial content of deployment is minimum.

In sum, when the flex point coordinate of interior convex closure it isTime, can make The interstitial content disposed is minimum.

4. simulation analysis

The respectively F2CR algorithm of the present invention and P3CR algorithm and Hamilton Path algorithm and 2C-SpiderWeb algorithm Total area coverage, the average area coverage of each node and average node degree these four after required interstitial content, reparation These four algorithms are contrasted by aspect, the F2CR algorithm of the checking present invention and the feasibility of P3CR algorithm and superiority.? In emulation, sub-block is randomly dispersed on the two dimensional surface of fixed size.Additionally, Fig. 4, Fig. 6, Fig. 8, Figure 10 interior joint communication distance For fixed value 50m, in Fig. 5, Fig. 7, Fig. 9, Figure 11, piecemeal number is fixed as 8.

4.1 interstitial contents disposed

Figure 4, it is seen that when the communication radius of node is fixed value, along with the increase of piecemeal number, four kinds of calculations Interstitial content used by method increases the most therewith.This is because when piecemeal number is the most, and the path summation between piecemeal is the biggest, Needing the interstitial content disposed the most, therefore total interstitial content used is the biggest.It can also be seen that not managed network from figure Being divided into how many sub-blocks, the interstitial content used by F2CR algorithm of the present invention is all less than used by 2C-SpiderWeb algorithm Interstitial content, more than the interstitial content used by Hamilton Path algorithm, this is to be determined by the topological structure that each algorithm is different Fixed.The P3CR algorithm of the present invention is owing to reaching part tee joint, so the interstitial content used by this algorithm is higher than other three kinds Algorithm.

From fig. 5, it can be seen that in the case of fixed block number, along with the increase of communication radius, used by four kinds of algorithms Interstitial content all reduce.This is because when piecemeal position determines, the distance between piecemeal is fixed, the node number of deployment is just Depend on the communication distance of node.When the communication radius of node is the biggest, between piecemeal, the required interstitial content disposed is the fewest, used Total interstitial content the fewest.From this figure it can be seen that along with the increase of node radius, used by the F2CR algorithm of the present invention Node number will be closer to the node number used by 2C-SpiderWeb algorithm.This is because when node radius is the biggest, 2C- The network topology that SpiderWeb algorithm is formed is more closely similar to the network topology structure that the F2CR algorithm of the present invention is formed, institute Will be closer to the interstitial content used by both algorithms.

4.2 total area coverages

It can be seen from figures 6 and 7 that along with the increase of piecemeal number/node communication radius, the node of four kinds of algorithms covers The lid gross area is consequently increased.The area coverage of the F2CR algorithm of the present invention is more than 2C-SpiderWeb algorithm and Hamilton The area coverage of Path algorithm, and the area coverage of the P3CR algorithm of the present invention is much larger than the area coverage of other three kinds of algorithms. Although 2C-SpiderWeb algorithm deploys the more node of F2CR algorithm than the present invention, but whether sub-block number or joint How the communication radius of point changes, and the area coverage of the F2CR algorithm of the present invention is always greater than the covering of 2C-SpiderWeb algorithm Area.Compared with 2C-SpiderWeb algorithm, the F2CR algorithm of the present invention has less covering overlapping area, therefore, not pipe How the communication radius of block number or node changes, and the F2CR algorithm of the present invention all has bigger than 2C-SpiderWeb algorithm Area coverage.Similarly, the P3CR algorithm of the present invention of the present invention has the covering weight less than the F2CR algorithm of the present invention How folded area, therefore, change regardless of the communication radius of sub-block number or node, and the P3CR algorithm of the present invention all has than it The area coverage that his algorithm is bigger.

4.3 average area coverages

As can be seen from Figure 8, the F2CR algorithm of the present invention and the average area coverage of P3CR algorithm are essentially identical, but little The average area coverage of each node obtained by Hamilton Path algorithm, more than averagely covering of 2C-SpiderWeb algorithm Capping amasss.Reason is that 2C-SpiderWeb algorithm exists bigger overlapping area.Substantial amounts of overlapping area makes 2C- The actual area coverage of the network topology that SpiderWeb algorithm is formed diminishes, and the average area coverage of the most each node also becomes Little.And compared with other three kinds of algorithms, it is overlapping to there is the minimum network coverage in the network topology that Hamilton Path algorithm is formed Area, the average area coverage shared by the most each node is maximum.

From fig. 9, it can be seen that along with the increase of communication radius, the area coverage shared by per node on average is also with increasing Greatly.Because when sub-block locations one timing, the number disposed between sub-block depends on the communication radius of node.Communication radius is the biggest, The area coverage of node is the biggest, and in fixed area, the required node number disposed is the fewest, therefore, and all nodes average Area coverage is the biggest.

4.4 average node degree

It can be seen from fig. 10 that the average node degree of F2CR and the P3CR algorithm of the present invention is respectively less than 2C-SpiderWeb The average node degree of algorithm.Because the covering lap that the existence of 2C-SpiderWeb algorithm is bigger, and the node of lap There is bigger node degree, so the average node degree of this algorithm is more than other three kinds of algorithms.Additionally, Hamilton Path calculates The topology that method is formed can regard a ring as, and the degree of the node on ring is essentially 2, but there is phase mutual respect between part of nodes Folded situation, so, final average node degree is less times greater than 2.The average node degree of two kinds of algorithms proposed by the invention is situated between Between 2C-SpiderWeb algorithm and Hamilton Path algorithm.Compared with 2C-SpiderWeb algorithm, the two of present invention proposition Plant algorithm and be respectively provided with less overlapping area coverage, and compared with Hamilton Path algorithm, two kinds of algorithms of the present invention have Bigger covering overlapping area, therefore the average node degree of both algorithms is between 2C-SpiderWeb algorithm and Hamilton Between Path algorithm.Additionally, it can also be seen that the average node degree of the P3CR algorithm of the present invention is slightly lower than this from Figure 10 Bright F2CR algorithm, this is because compared with the F2CR algorithm of the present invention, the network topology that P3CR algorithm is formed has less Covering overlapping area, therefore the average node degree of the P3CR algorithm of the present invention can be slightly lower than F2CR algorithm.

The situation of Figure 11 is similar with Figure 10, but it can be seen from fig. 11 that along with the increase of communication radius, node average Node degree is also with increase.This is because when the position of node deployment determines, the communication radius of node is the biggest, node and node Between covering overlapping area the biggest, therefore, the average node degree of node is the biggest.

Accordingly, the embodiment of the present invention three also provides for the fault-tolerant repair system of a kind of wireless sensor network, including:

Acquisition module, is used for obtaining wireless sensor node and coordinate, communication radius；

Outer convex closure module, for determining the outer convex closure of node, and calculates the Centroid coordinate of described outer convex closure and each limit Long；

Outer convex closure deployment module, for outer convex closure flex point being connected with Centroid respectively, according to communication radius, determines even Need on line to dispose node and quantity thereof；

Interior convex closure deployment module, for disposing node so that interior convex closure flex point is three connections on convex bound edge；

Outer convex closure deployment module is additionally operable to obtain evagination and wraps non-conterminous limit, and disposes node thereon, forms two Connect.

On wireless sensor node, wherein choose several nodes, be linked in sequence, form convex-edge shape, and remainder can be comprised All nodes；Wherein this convex-edge shape is outer convex closure, the flex point that several nodes are outer convex closure chosen；

The node near Centroid will be deployed on line as interior convex closure flex point, and convex closure in the formation that is linked in sequence Limit.

Described outer convex closure deployment module is additionally operable to dispose on convex closure limit outside node so that outer convex closure flex point is three even Connect.

The foregoing is only embodiments of the invention, not thereby limit the scope of the claims of the present invention, every utilize this The equivalents that bright description and accompanying drawing content are made, or directly or indirectly it is used in relevant technical field, the most in like manner include In the scope of patent protection of the present invention.

Claims (10)

1. the fault-tolerant restorative procedure of a wireless sensor network, it is characterised in that including:

Obtain wireless sensor node and coordinate, communication radius；

Determine the outer convex closure of node, and calculate the Centroid coordinate of described outer convex closure and each length of side；

Respectively outer convex closure flex point is connected with Centroid, according to communication radius, determines and need on line to dispose node and number thereof Amount；

Convex bound edge is disposed node so that interior convex closure flex point is three connections；

Obtain evagination and wrap non-conterminous limit, and dispose node thereon, form two connections.

The most fault-tolerant restorative procedure of wireless sensor network, it is characterised in that at wireless sensor node Choose several nodes on Dian, be linked in sequence, form convex-edge shape, and remaining all nodes can be comprised；Wherein this convex-edge shape is outward Convex closure, the flex point that several nodes are outer convex closure chosen.

The most fault-tolerant restorative procedure of wireless sensor network, it is characterised in that will be deployed on line Near the node of Centroid as interior convex closure flex point, and be linked in sequence formed in the limit of convex closure.

The most fault-tolerant restorative procedure of wireless sensor network, it is characterised in that according to outer convex closure flex point Coordinate, calculate Centroid coordinate and each limit length of side of outer convex closure；

According to outer convex closure flex point and the coordinate of Centroid, calculate the length of outer convex closure flex point and each line of Centroid, and count Calculate the angle of adjacent connection lines；

In conjunction with communication radius, calculate the quantity needing to dispose node on each line；

From outer convex closure flex point, convex closure Centroid direction outwardly, each line is disposed node.

The most fault-tolerant restorative procedure of wireless sensor network, it is characterised in that obtain evagination and wrap not Adjacent edge, and dispose node thereon, form the steps of two connections particularly as follows:

Whether the quantity judging evagination bound edge is even number；

The most then obtain the non-conterminous limit of half that evagination is wrapped so that its total length is minimum, and dispose on these non-conterminous limits Node；

Otherwise, on convex closure, obtain the flex point not forming two connections the most outside；

And wrap the acquisition limit without this flex point at convex, and on limit, obtain the auxiliary magnet minimum with this flex point distance；

This flex point line with auxiliary magnet is disposed node.

The most fault-tolerant restorative procedure of wireless sensor network, it is characterised in that outside in convex closure, remove In other nodes of interior convex closure flex point, obtain three points of closest two point or not conllinear, and at its interconnecting line Upper deployment node.

7. the fault-tolerant restorative procedure of a wireless sensor network, it is characterised in that including:

Obtain wireless sensor node and coordinate, communication radius；

Determine the outer convex closure of node, and calculate the Centroid coordinate of described outer convex closure and each length of side；

Respectively outer convex closure flex point is connected with Centroid, according to communication radius, determines and need on line to dispose node and number thereof Amount；

Convex bound edge is disposed node so that interior convex closure flex point is three connections；

Node is disposed outside so that outer convex closure flex point is three connections on convex closure limit.

The most fault-tolerant restorative procedure of wireless sensor network, it is characterised in that outside in convex closure, remove In other nodes of interior convex closure flex point, obtain three points of closest two point or not conllinear, and at its interconnecting line Upper deployment node.

9. the fault-tolerant repair system of a wireless sensor network, it is characterised in that including:

Acquisition module, is used for obtaining wireless sensor node and coordinate, communication radius；

Outer convex closure module, for determining the outer convex closure of node, and calculates the Centroid coordinate of described outer convex closure and each length of side；

Outer convex closure deployment module, for outer convex closure flex point being connected with Centroid respectively, according to communication radius, determines on line Need to dispose node and quantity thereof；

Interior convex closure deployment module, for disposing node so that interior convex closure flex point is three connections on convex bound edge；

Outer convex closure deployment module is additionally operable to obtain evagination and wraps non-conterminous limit, and disposes node thereon, forms two connections.

The most fault-tolerant repair system of wireless sensor network, it is characterised in that outer convex closure disposes mould Block is additionally operable to dispose on convex closure limit outside node so that outer convex closure flex point is three connections.