CN102238561B - Based on the node deployment method of the layering cooperation overlay model of Energy Efficient - Google Patents

Abstract

The invention discloses the node deployment method of a kind of layering based on Energy Efficient cooperation overlay model, belong to wireless sensor network technology field.Concrete grammar is: (1) sets up the overlay model of target area, by carrying out layering to reach the object of saving energy consumption to target area; (2) solve heuristic greedy method, make when with this model of ant colony optimization for solving, convergence rate is accelerated greatly; (3) solution node quantitatively lower limit, make with during this model of ant colony optimization for solving iterations reduce; (4) the position deployment strategy of node is chosen; (5) optimal solution of ant colony optimization for solving layering cooperation overlay model is utilized.Beneficial effect is the node deployment that the present invention can meet the region with specific region shape.

Description

Based on the node deployment method of the layering cooperation overlay model of Energy Efficient

Technical field

The present invention relates to the node deployment method of a kind of layering based on Energy Efficient cooperation overlay model; Belong to wireless sensor network technology field.

Background technology

Wireless sensor network is the computer system that a kind of intelligent radio sensing node is formed, and its appearance extends the natural ability of mankind's mobile awareness, changes the mode of people's managing family, factory and environment.Wireless sensor network is disposed flexibly with it, and feature with low cost is widely used in fields such as military surveillance, traffic monitoring, intelligent industrial controls.

The feature of wireless sensing node comprises: resource-constrained, and network coverage is large, and node is intensive, dynamic topology.Generally speaking, for a certain measurement task, the layout quantity of sensing node all will exceed its requirement.Do the problem that can make up not high the caused certainty of measurement of sensing node positioning precision on the one hand and reduce like this; Also System Error-tolerance Property can be improved further on the other hand.But wireless sensing node is all subject to hard constraints in size, weight, energy and bandwidth, these will affect the availability of Internet resources.Therefore, how to ensure when consuming least energy that network coverage is the focus of research at present simultaneously.

The network coverage is that wireless sensor network is measured the one of its target monitoring regionally detecting, is a major issue in wireless sensor network research contents.Overlay model certainty can be divided into cover according to configuration mode and randomness covers.The research contents that certainty covers is that the state of target area is relatively fixing, according to pre-configured node location determination network topology structure or the node density setting subregion as required.The research contents that random node covers is in node random distribution and the situation of node location the unknown completes covering to target area.The main standard evaluating overlay model quality has network to the level of coverage of target area, energy efficiency, the accuracy of algorithm and the robustness of network topology.In wireless sensor network, because node energy is limited, actual application environment is complicated and in most cases do not allow to change node battery, and the useful life and the network lifetime that therefore extend node to greatest extent become research contents important in wireless sensor network.

Conventional traditional overlay model has Boolean Model, general overlay model, cooperation overlay model and Probability Coverage Model etc.Boolean Model is comparatively simple, but the feature that the detectivity can not expressing transducer increases with distance and reduces; General overlay model then only considers the covering power of single-sensor; Although cooperation overlay model and Probability Coverage Model take into account multiple transducer to the contribution covered, do not study from the angle of node energy consumption.

When ensureing that sensing network has certain covering power to target area, the Main Means reducing node energy consumption has the optimized network topology when interstitial content is determined, dormancy dispatching is carried out when node redundancy is disposed, optimize interstitial content and node location, optimize the position etc. of sink node.Covering research at present for Energy Efficient in radio sensing network focuses mostly in the first two field.The people such as K.Shashi Prab propose a kind of distributed hexagon covering algorithm, the network coverage of this algorithm by selecting optimized neighbor node to realize Energy Efficient.The people such as Paul Sounak propose the balanced dormancy dispatching strategy of dynamic power under isomorphism multi-node network environment, by three factors such as residue energy of node, this algorithm determines that the dormancy of node reaches the object of saving energy, but this algorithm does not support multihop network.The people such as Costa propose the multi-path selecting solution algorithm of topology ambiguity node refusal, and this algorithm, by investigating node and covering power and carrying out balance node energy consumption by multi-path selecting solution, can extend the life span of network by this selection algorithm.The people such as Martins then propose and solve when having a node failure, and the Dynamic-Recovery of region overlay, proposes the dynamic coverage algorithm MGoDA of Energy Efficient.Minimum covering set approximate data based on greedy method can go out Minimum covering set in linear Time Calculation, namely with minimum coverage target area, thus allows more node enter dormancy dispatching, reaches the object of saving node energy consumption.Above algorithm is all analyze and optimized network node energy consumption from the first two angle.In multihop network, most node needs to carry out routing forwarding by via node could be sent to sink node by data, and via node also needs to forward the Monitoring Data of other node except sending self Monitoring Data, node energy loss is very fast.Therefore, the loss rate being in the node energy of diverse location is in a network not quite similar.If via node death is too fast, the monitor data of ordinary node cannot be forwarded by via node, and network just cannot provide normal service, and the energy of ordinary node fails to be fully used.And in many applications, the node that many positions are important, because dispose difficulty comparatively greatly, first should ensure that the energy of these nodes is fully used, as the front node in the data monitoring of battlefield, mine bottom node in coal information monitoring etc.Therefore only consider coverage rate and ignore because node location is different and cause energy loss model that the node of some position can be made dead faster, thus reducing the life span of whole network.Therefore need at diverse location when node deployment, zones of different lays the node of different densities, to reach balancing energy, extends the object of network lifetime.

The people such as calendar year 2001 Seapahn establish cooperation overlay model (CCM), reflect the coverage condition of network at this point by defining certain any covering index.Covering index definition is:

$I(p,K)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}S(p,s_k)---\left(1\right)$

Wherein

The people such as 2007 Nian Yang point out on the basis of cooperation overlay model, are ensureing under the condition that region is completely covered, and optimum Node distribution is: any 3 form an equilateral triangle, and the leg-of-mutton length of side is (R is the maximum communication radius of node), all nodes become cellular coverage goal region.As shown in Figure 1, region S is become cellular covering by 14 nodes.

Cooperation overlay model only investigates node to the coverage of target area from the angle of overlay area, and does not add the analysis of the energy loss to node.In multihop network, also need to forward the data from the node away from sink node from the node that sink node is relatively near except will sending self monitor data, when sending data volume and being identical, energy loss is relatively very fast.In single-hop networks, the energy sending data degradation from the node away from sink node is relatively many.Therefore, the covering of network only should not consider area coverage, also will consider diverse location, the energy loss of the node of different levels.

Summary of the invention

The object of the present invention is to provide the node deployment method of a kind of layering based on Energy Efficient cooperation overlay model, realizing the minimum power consumption making node when meeting coverage requirement.

The present invention takes following technical scheme to realize: a kind of node deployment method of the cooperation of the layering based on Energy Efficient overlay model, comprises the following steps:

1) set up the overlay model of target area, this model is by carrying out layering to reach the object of saving energy consumption to target area.

2) solve heuristic greedy method, make when with this model of ant colony optimization for solving, convergence rate is accelerated greatly.

3) solution node quantitatively lower limit, makes when with this model of ant colony optimization for solving, and iterations reduces.

4) the position deployment strategy of node is chosen.

5) optimal solution of ant colony optimization for solving layering cooperation overlay model is utilized.

Described step 1 specifically comprises the following steps:

11) draw out the drawing of target area, and calculate the area of target area.

12) require and the size of target area according to coverage, choose distinguishing hierarchy granularity.

13) the layering cooperation overlay model of target area is set up.

Described step 3 is specially: the first target function of Modling model, and this target function is that the life span of the minimum layer of the life span after making score layer is maximum.Its secondary constraints setting up this model: one, every node layer quantity sum equals total number of nodes.Two, the coverage rate of every node layer is greater than the coverage rate of requirement.

Described step 5 specifically comprises the following steps:

51) parameter used in initialization algorithm, comprises the number N of total node, the primary power of node, the maximum communication radius of node; In ant group algorithm, the number of ant, the pheromones amount τ on the ij of path _ijbe initially 0, the relative importance α of pheromones, the relative importance β of heuristic greedy method, heuristic greedy method η _ijin constant M=1, W=1, ant ring constant Q=10;

52) start outermost layer iteration, for every ant, present node is added taboo list;

53) heuristic greedy method is calculated;

54) calculate the transition probability on every paths by pheromones and heuristic greedy method, and choose next node;

55) node location deployment strategy is chosen;

56) multihop routing algorithm is performed, computing network life span;

57) Pheromone update is carried out.

Described step 55 is specially: the area according to overlay area obtains each orthohexagonal size divided by node number, thus obtains the length of side and center.Each sensor node is just placed on orthohexagonal center, and this length of side is less than or equal to the orthohexagonal length of side of general overlay model.

Layering based on Energy Efficient cooperation overlay model provided by the invention, the distribution of the node location that it takes into full account is for the impact of network lifetime, achieve when meeting coverage requirement by carrying out hierarchy optimization to target area, the lifetime maximization of network.Solve this model by ant group algorithm and can reach Fast Convergent, avoid the situation of global optimum.The present invention can meet the node deployment in the region with specific region shape.

The present invention is further illustrated below in conjunction with drawings and Examples.

Accompanying drawing explanation

Fig. 1 is the schematic diagram of the layering cooperation overlay model that the present invention adopts.

Fig. 2 is the conversion method of the ant group algorithm that the present invention adopts.

Embodiment

As shown in Figure 1, the present invention in the specific implementation, first will set up the overlay model of target area.Layering cooperation overlay model Energy Efficient Hierarchical Collaboration CoverageModel (EEHCCM) of Energy Efficient, solves the maximum of network lifetime exactly under certain constraints.The target function of this model is:

$T_{\mathrm{Sur}}=\max \left(\min T_{{\mathrm{Sur}}_i}\right),i=\mathrm{1,2}...l---\left(3\right)$

Constraints is:

$\underset{i=1}{\overset{j}{\mathrm{\Σ}}}{b}_i=N---\left(4\right)$

N＞N′ (5)

S _monitor≥S _area(6)

The implication of formula 4 is that the number sum of every node layer equals total number of network nodes, and formula 5 and formula 6 are that the deployment of requirement node covers whole target monitoring region, wherein T _surrepresent network lifetime, represent the network lifetime of i-th layer, l represents the total number of plies divided in network, b _irepresent the quantity of the sensor node of i-th layer, N represents the quantity of total node in network, the minimum nodes number of N ' representative cooperation overlay model needed for entirely covering under a condition, S _monitorthe area of node coverage areas, S _areait is the area of target area; The problem that EEHCC solves is in a certain monitored area, when total nodal point number amount is known, realizes node energy equilibrium by the node density optimizing diverse location, extends network lifetime.

After model has been set up, ant colony optimization for solving is carried out to model, before solving, set the selection rule of parameter and the heuristic greedy method used in this algorithm.The selection rule of the heuristic greedy method of the ant colony optimization algorithm of EEHCCM is:

If i > is j

If i < is j

Wherein: η _ijk () represents the heuristic greedy method from layer i selection kth paths to layer j, the optimum number of nodes of presentation layer i is estimated, M and W is any normal number.

Solving model employs number of nodes constraint, if l _ilayer guarded region area is wherein d is node maximum communication distance, then, under will ensureing the prerequisite of all standing, the lower limit of every node layer quantity is:

If the set having determined the layer of number of nodes is P, do not determine that the set of the layer of number of nodes is Z, l _ithe upper limit of node layer quantity is:

Wherein N is total number of nodes, n _ibe the i-th node layer quantity, S _iit is the area of i-th layer of guarded region.

Because solving of EEHCC model is NP difficulty, therefore there is not the derivation algorithm of polynomial time.Needing when applying ant colony optimization for solving EEHCCM to change this problem, making input parameter type meet the form of ant group algorithm.If the set L of distinguishing hierarchy is regarded as the set C in the city in TSP problem, and the nodes set N of every layer regards the distance set D in TSP problem between city as, then this problem is converted into the class TSP problem of a multipath.Shown in Fig. 2, the circle in figure represents certain one deck in Region dividing, and corresponding to the city in TSP problem, the camber line between circle represents the quantity n of every node layer _l, corresponding to distance d intercity in TSP problem _{ci, cj}, network lifetime T _kcorresponding to path L _k, selecting kth paths to transfer to a layer j definition of probability from layer i is a _ijk (), the pheromone concentration on this paths is τ _ij(t, k), k represents the number of nodes in layer i, and the heuristic greedy method between layer i and layer j on kth paths is η _ij(k).When algorithm performs, ant carries out random walk between each layer, determines transfer by transition probability, and the pheromones more on new route, finally reach convergence.Therefore EEHCCM is converted into the number of nodes compound mode that in the set of paths of each circle in traversing graph, search makes network lifetime the longest.

The ant group derivation algorithm model of EEHCCM can be expressed as:

$a_{\mathrm{ij}}\left(k\right)=\frac{{\left[{\mathrm{\&tau;}}_{\mathrm{ij}}\left(k\right)\right]}^{\mathrm{\&alpha;}}{\left[{\mathrm{\&eta;}}_{\mathrm{ij}}\left(k\right)\right]}^{\mathrm{\&beta;}}}{\mathrm{\&Sigma;}{\left[{\mathrm{\&tau;}}_{\mathrm{ij}}\left(c\right)\right]}^{\mathrm{\&alpha;}}{\left[{\mathrm{\&tau;}}_{\mathrm{ij}}\left(c\right)\right]}^{\mathrm{\&beta;}}},\&ForAll;i\&NotEqual;j,i,j=\mathrm{1,2}...l,c=\mathrm{1,2}...V---\left(11\right)$

τ _ij(t，k)＝ρτ _ij(t-1，k)+Δτ _ij(k) (12)

$\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}\left(k\right)=\underset{a=1}{\overset{m}{\mathrm{\&Sigma;}}}\mathrm{\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}^a\left(k\right)---\left(13\right)$

Wherein a _ij(k) probability for selecting kth paths to transfer to layer j from layer i, τ _ijk () is for transferring to the pheromone concentration the kth paths of layer j, η from layer i _ijk () is the heuristic greedy method transferred to from layer i the kth paths of layer j, τ _ijc () is for transferring to the pheromone concentration the c paths of layer j, η from layer i _ijc () is the heuristic greedy method transferred to from layer i the c paths of layer j, τ _ij(t, k) transfers to the pheromone concentration the kth paths of layer j from layer i, τ in t _ij(t-1, k) transfers to the pheromone concentration the kth paths of layer j from layer i, Δ τ in the t-1 moment _ijthe t-1 moment to the change from pheromone concentration the path that layer i transfers to layer j of t, Δ τ _ij(k) be transfer to layer j the t-1 moment to t from layer i kth paths the change of pheromone concentration, be the t-1 moment to t a ant through transferring to the kth paths of layer j and the variable quantity of pheromone concentration that brings from layer i, m represents the quantity of the ant used in algorithm, and V representative transfers to the feasible path sum of layer j, T from layer i _knetwork lifetime during kth paths is selected in representative, the relative importance of α representative information element, and β represents the relative importance of heuristic greedy method, and ρ is pheromones evaporation coefficient, and Q is ant ring constant, and it is any one normal number;

Algorithm needs the parameter used in initialization algorithm before performing, comprise the number N of total node, the primary power of node, the maximum communication radius of node. in ant group algorithm, and the number of ant, the pheromones amount τ on the ij of path _ijbe initially 0, the relative importance α of pheromones, the relative importance β of heuristic greedy method, heuristic greedy method η _ijin constant M=1, W=1, ant ring constant Q=10

Utilize the algorithm steps of the layering coordination model of ant colony optimization for solving Energy Efficient as follows:

Algorithm, after several times iteration, can converge on optimal solution, and this optimal solution specifies the optimal number ratio of every node layer, after determining the number of nodes of every layer, needs position or the distribution of specified node.The mode that the present invention takes obtains each orthohexagonal size according to the area of overlay area divided by node number, thus obtain the length of side and center.Each sensor node is just placed on orthohexagonal center, and this length of side is less than or equal to the orthohexagonal length of side of general overlay model.

Claims (5)

1., based on a node deployment method for the layering cooperation overlay model of Energy Efficient, it is characterized in that: concrete steps are as follows:

1) set up the overlay model of target area, this model reaches the object of saving energy consumption by carrying out layering to target area; The described overlay model setting up target area adopts layering cooperation overlay model, is specially: the target function of this model is:

$T_{\mathrm{Sur}}\text{=max}\left(\min T_{{\mathrm{Sur}}_i}\right),i=\mathrm{1,2}...l$

Constraints is:

$\underset{i=1}{\overset{l}{\mathrm{\&Sigma;}}}{b}_i=N$

N＞N′

S _monitor≥S _area

Wherein T _surrepresent network lifetime, represent the network lifetime of i-th layer, l represents the total number of plies divided in network, b _irepresent the quantity of the sensor node of i-th layer, N represents the quantity of total node in network, the minimum nodes number of N ' representative cooperation overlay model needed under all standing condition, S _monitorthe area of node coverage areas, S _areait is the area of target area; The problem that model solves is in a certain monitored area, when total number of nodes is known, realizes node energy equilibrium by the node density optimizing diverse location, extends network lifetime;

2) solve heuristic greedy method, make when with this model of ant colony optimization for solving, convergence rate is accelerated greatly;

The heuristic greedy method computational methods used in derivation algorithm are:

If i > is j

If i < is j

Its η _ijk () represents the heuristic greedy method from layer i selection kth paths to layer j, the optimum number of nodes of presentation layer i is estimated, M and W is any normal number;

Described solving model adopts ant group algorithm, is specially:

$a_{\mathrm{ij}}\left(k\right)=\frac{{\left[{\mathrm{\&tau;}}_{\mathrm{ij}}\left(k\right)\right]}^{\mathrm{\&alpha;}}{\left[{\mathrm{\&eta;}}_{\mathrm{ij}}\left(k\right)\right]}^{\mathrm{\&beta;}}}{\mathrm{\&Sigma;}{\left[{\mathrm{\&tau;}}_{\mathrm{ij}}\left(c\right)\right]}^{\mathrm{\&alpha;}}{\left[{\mathrm{\&tau;}}_{\mathrm{ij}}\left(c\right)\right]}^{\mathrm{\&beta;}}}\&ForAll;i\&NotEqual;j,i,j=\mathrm{1,2}...l,c=\mathrm{1,2}...V$

τ _ij(t，k)＝ρτ _ij(t-1，k)+Δτ _ij(k)

${\mathrm{\&Delta;\&tau;}}_{\mathrm{ij}}\left(k\right)=\underset{a=1}{\overset{m}{\mathrm{\&Sigma;}}}{\mathrm{\&Delta;\&tau;}}_{\mathrm{ij}}^a\left(k\right)$

Wherein a _ij(k) probability for selecting kth paths to transfer to layer j from layer i, τ _ijk () is for transferring to the pheromone concentration the kth paths of layer j, η from layer i _ijk () is the heuristic greedy method transferred to from layer i the kth paths of layer j, τ _ijc () is for transferring to the pheromone concentration the c paths of layer j, η from layer i _ijc () is the heuristic greedy method transferred to from layer i the c paths of layer j, τ _ij(t, k) transfers to the pheromone concentration the kth paths of layer j from layer i, τ in t _ij(t-1, k) transfers to the pheromone concentration the kth paths of layer j from layer i, Δ τ in the t-1 moment _ijthe t-1 moment to the change from pheromone concentration the path that layer i transfers to layer j of t, Δ τ _ij(k) be transfer to layer j the t-1 moment to t from layer i kth paths the change of pheromone concentration, be the t-1 moment to t a ant through transferring to the kth paths of layer j and the variable quantity of pheromone concentration that brings from layer i, m represents the quantity of the ant used in algorithm, and V representative transfers to the feasible path sum of layer j, T from layer i _knetwork lifetime during kth paths is selected in representative, the relative importance of α representative information element, and β represents the relative importance of heuristic greedy method, and ρ is pheromones evaporation coefficient, and Q is ant ring constant, and it is any one normal number;

3) solution node quantitatively lower limit, makes when with this model of ant colony optimization for solving, and iterations reduces;

Solving model employs number of nodes constraint, if l _ilayer guarded region area is wherein d is node maximum communication distance, then, under will ensureing the prerequisite of all standing, the lower limit of every node layer quantity is:

If the set having determined the layer of number of nodes is P, do not determine that the set of the layer of number of nodes is Z, l _ithe upper limit of node layer quantity is:

Wherein N is total number of nodes, n _ibe the i-th node layer quantity, S _iit is the area of i-th layer of guarded region;

4) the position deployment strategy of node is chosen;

5) optimal solution of ant colony optimization for solving layering cooperation overlay model is utilized.

2. the node deployment method of the layering based on Energy Efficient according to claim 1 cooperation overlay model, is characterized in that, described step 1) concrete grammar of setting up the overlay model of target area is:

11) draw out the drawing of target area, and calculate the area of target area;

12) require and the size of target area according to coverage, choose distinguishing hierarchy granularity;

13) the layering cooperation overlay model of target area is set up.

3. the node deployment method of the cooperation of the layering based on Energy Efficient overlay model according to claim 1, it is characterized in that, described step 3) concrete grammar of solution node quantitatively lower limit is: the first target function of Modling model, this target function is that the life span of the minimum layer of the life span after making score layer is maximum; Next sets up the constraints of this model: one, every node layer quantity sum equals total number of nodes; Two, the coverage rate of every node layer is greater than the coverage rate of requirement.

4. the node deployment method of the layering based on Energy Efficient according to claim 1 cooperation overlay model, is characterized in that, described step 5) the optimal solution concrete grammar that solves layering cooperation overlay model is:

51) parameter used in initialization algorithm, comprises the number N of total node, the primary power of node, the maximum communication radius of node; In ant group algorithm, the number of ant, the pheromones amount τ on the ij of path _ijbe initially 0, the relative importance α of pheromones, the relative importance β of heuristic greedy method, heuristic greedy method η _ijin constant M=1, W=1, ant ring constant Q=10;

52) start outermost layer iteration, for every ant, present node is added taboo list;

53) heuristic greedy method is calculated;

54) calculate the transition probability on every paths by pheromones and heuristic greedy method, and choose next node;

55) node location deployment strategy is chosen;

56) multihop routing algorithm is performed, computing network life span;

57) Pheromone update is carried out.

5. the node deployment method of the cooperation of the layering based on Energy Efficient overlay model according to claim 4, it is characterized in that, described step 55) to choose node location deployment strategy and be specially: the area according to overlay area obtains each orthohexagonal size divided by node number, thus obtains the length of side and center; Each sensor node is just placed on orthohexagonal center, and this length of side is less than or equal to the orthohexagonal length of side of general overlay model.