阅读说明：本技术 基于方向角分解优化和冗余节点去除的传感器网络覆盖增强方法 (Sensor network coverage enhancement method based on direction angle decomposition optimization and redundant node removal ) 是由 张磊 许淼 文方青 王可 黑创 胡林 于 2019-09-29 设计创作，主要内容包括：本发明公开一种基于方向角分解优化和冗余节点去除的传感器网络覆盖增强方法,属于传感器监测领域。本发明首先建立一种三维锥体有向感知模型；然后将传感器节点的感知范围分解为俯仰角和偏向角,提出一种基于差分进化算法的偏向角优化方法；最后针对传感器网络中存在大量冗余节点的问题,提出一种冗余节点去除方法。该方法能够有效改善传感器网络的覆盖率和节点数量。(The invention discloses a sensor network coverage enhancement method based on direction angle decomposition optimization and redundant node removal, and belongs to the field of sensor monitoring. Firstly, establishing a three-dimensional cone directional perception model; then decomposing the sensing range of the sensor node into a pitch angle and a deflection angle, and providing a deflection angle optimization method based on a differential evolution algorithm; and finally, aiming at the problem that a large number of redundant nodes exist in the sensor network, a redundant node removing method is provided. The method can effectively improve the coverage rate and the number of nodes of the sensor network.)

1. The sensor network coverage enhancement method based on the direction angle decomposition optimization and the redundant node removal is characterized by comprising the following steps of:

step 1, establishing a three-dimensional cone directional perception model;

step 2, calculating the volume of the coverage area of the perception model;

step 3, calculating an optimal coverage volume S and an optimal pitch angle lambda corresponding to S;

step 4, constructing the space coverage rate of the whole sensor network;

step 5, optimizing a deflection angle by using a differential evolution algorithm;

and 6, removing redundant nodes in the sensor network.

2. The method for enhancing the coverage of the sensor network based on the directional angle decomposition optimization and the redundant node removal according to claim 1, wherein the specific process of the step 1 is as follows:

step 1.1, rotating by taking a sensor node P (x, y, z) as a circle center, a radius of R and a central angle of 2 β to form a spatial sensing area PO₁O₂O₃O₄And is represented by a quintuple array (P, B, α, R);

step 1.2, setting P ' as the projection of the sensing node P on an X-Y plane, setting B (lambda, phi) as the main sensing direction of P, wherein the included angle lambda (∠ P ' PB) between P ' and a Z axis is called a pitch angle, the projection of P ' on the X-Y plane is called a deflection angle phi (∠ BP ' X), and 2 α,2 β and R respectively represent the horizontal sensing angle, the vertical sensing angle and the sensing radius of the sensor node.

3. The method for enhancing the coverage of the sensor network based on the directional angle decomposition optimization and the redundant node removal according to claim 1, wherein the specific process of the step 2 is as follows:

step 2.1, set the Cone PO₁O₂O₃O₄Each vertex coordinate is respectivelyComprises the following steps:

P＝(x,y,z)

O₁＝(x+|P'O₁|cos(α+φ),y+|P'O₁|sin(α+φ),0)

O₂＝(x+|P'O₂|cos(α-φ),y-|P'O₂|sin(α-φ),0)

O₃＝(x+|P'O₃|cos(α-φ),y-|P'O₃|sin(α-φ),0)

O₄＝(x+|P'O₄|cos(α+φ),y+|P'O₄|sin(α+φ),0)

wherein, | P' O₁|＝|P'O₂|＝ztan(λ-β)/|cosα|，|P'O₃|＝|P'O₄|＝ztan(λ+β)/|cosα|；

Step 2.2, calculate Cone PO₁O₂O₃O₄The volume S of (A) is:

4. the method for enhancing sensor network coverage based on directional angular resolution optimization and redundant node removal according to claim 1, wherein the step 4 specifically comprises:

step 4.1, discrete gridding of a monitoring area;

step 4.2, calculating the discrete grid number N of the monitoring area_totalThe number of the covered grids of the sensor network is

Wherein N is_totalIn order to be able to count the number of sensors,the number of grids covered by each sensor, wherein N is a positive integer;

step 4.3, obtain the space coverage of the sensor network as η ═ S_total/N_total。

5. The method for enhancing sensor network coverage based on directional angular resolution optimization and redundant node removal according to claim 1, wherein the step 5 specifically comprises:

step 5.1, setting initial parameters including population scale N, sensor number m and maximum evolution times G_maxScaling factor F, crossover factor CR;

step 5.2, generating an initial population P₀Each of the individuals

Step 5.3, mutation operation:

V_i(t+1)＝X_r1(t)+F×(X_r2(t)-X_r3(t))

wherein t is the evolution iteration number, and F is within 0,1]Called scaling factor, r₁,r₂,r₃E {1,2, …, N }, N being the population size, and r₁,r₂,r₃Different from the current target index i;

step 5.4, cross operation:

wherein t is the evolution iteration number, CR belongs to [0,1 ]]Called the crossover factor, rand (j) is [0,1 ]]A random number of (c); j-k guarantees test individual U_iAt least one-dimensional component of (t +1) is derived from the variant individual V_i(t + 1);

and 5.5, selecting operation:

where t is the number of evolutionary iterations, f (U)_i(t +1)) and f (X)_i(t)) isA response evaluation function;

step 5.6, judging whether the maximum evolution times G is met_max(ii) a If yes, ending, and outputting an optimal solution, namely a group of optimal deflection angles

6. The method for enhancing sensor network coverage based on directional angular resolution optimization and redundant node removal according to claim 1, wherein the step 6 specifically comprises:

step 6.1, calculating the dormancy decision parameter SDP of all nodes in the sensor network;

step 6.2, identifying redundant nodes according to the dormancy decision parameter SDP, and sequencing CCPs of all redundant nodes in ascending order;

6.3, removing the redundant node with the minimum CCP;

and 6.4, updating the working nodes in the sensor network, and turning to the step 6.1 until the number of the nodes in the sensor network is reduced to a set numerical value.

Technical Field

The invention relates to the field of sensor monitoring, in particular to a sensor network coverage enhancement method based on direction angle decomposition optimization and redundant node removal.

Background

With the progress of the world urbanization, the sensor network is widely applied to many fields such as intelligent transportation, intelligent home and the like. In fact, sensor networks have become an indispensable backbone for various practical applications. Generally, monitoring a specific area using a sensor network requires a large number of low-energy nodes, often involving two key issues, namely coverage performance and energy consumption.

The coverage enhancement problem is a fundamental problem in wireless sensor networks. When constructing a sensor network in an area of interest, a large number of sensors need to be deployed. Each sensor is capable of monitoring a target within its sensing region. Therefore, coverage enhancement is primarily to determine whether objects located at different locations can be covered by the sensor. The coverage performance is often referred to as the percentage of the total monitored area that the sensor successfully covers the target area. Better coverage performance generally means higher quality of service. These challenges are primarily related to the massive parametric optimization of large-scale sensing models. Most existing coverage enhancement methods aim to improve coverage without considering more targets such as cost or energy consumption, and are not suitable for the view of multi-objective optimization.

Lifecycle is another important issue in sensor networks. The deployment of sensors in inappropriate locations or patterns, as well as the difficulty of replacing batteries, further highlights the energy consumption problem of the sensor nodes. Therefore, the strategy for optimizing energy consumption is very critical, especially considering that the sensor network cannot work normally after a small number of nodes exhaust energy. The challenge is primarily related to determining whether certain portions of the region of interest covered by the sensors are also covered by other sensors, and determining the sequence of sensor activation or deactivation.

The invention aims to improve the coverage rate and the number of nodes of a sensor network simultaneously, and realizes the optimization of service quality and the minimization of energy consumption.

Disclosure of Invention

Aiming at the problems in the prior art, the two aspects of the invention provide a sensor network coverage enhancement method based on direction angle decomposition optimization and redundant node removal.

In order to achieve the purpose, the invention adopts the following technical scheme:

the sensor network coverage enhancement method based on the direction angle decomposition optimization and the redundant node removal is characterized by comprising the following steps of:

step 1, establishing a three-dimensional cone directional perception model;

step 2, calculating the volume of the coverage area of the perception model;

step 3, calculating an optimal coverage volume S and an optimal pitch angle lambda corresponding to S;

step 4, constructing the space coverage rate of the whole sensor network;

step 5, optimizing a deflection angle by using a differential evolution algorithm;

and 6, removing redundant nodes in the sensor network.

In the above aspect, the step 1 specifically includes:

step 1.1, rotating by taking a sensor node P (x, y, z) as a circle center, a radius of R and a central angle of 2 β to form a spatial sensing area PO₁O₂O₃O₄And is represented by a quintuple array (P, B, α, R);

step 1.2, setting P ' as the projection of the sensing node P on an X-Y plane, setting B (lambda, phi) as the main sensing direction of P, wherein the included angle lambda (∠ P ' PB) between P ' and a Z axis is called a pitch angle, the projection of P ' on the X-Y plane is called a deflection angle phi (∠ BP ' X), and 2 α,2 β and R respectively represent the horizontal sensing angle, the vertical sensing angle and the sensing radius of the sensor node.

In the above aspect, the step 2 specifically includes:

step 2.1, set the Cone PO₁O₂O₃O₄The vertex coordinates are respectively:

P＝(x,y,z)

O₁＝(x+|P'O₁|cos(α+φ),y+|P'O₁|sin(α+φ),0)

O₂＝(x+|P'O₂|cos(α-φ),y-|P'O₂|sin(α-φ),0)

O₃＝(x+|P'O₃|cos(α-φ),y-|P'O₃|sin(α-φ),0)

O₄＝(x+|P'O₄|cos(α+φ),y+|P'O₄|sin(α+φ),0)

wherein, | P' O₁|＝|P'O₂|＝ztan(λ-β)/|cosα|，|P'O₃|＝|P'O₄|＝ztan(λ+β)/|cosα|；

Step 2.2, calculate Cone PO₁O₂O₃O₄The volume S of (A) is:

in the above aspect, the step 4 specifically includes:

step 4.1, discrete gridding of a monitoring area;

step 4.2, calculating the discrete grid number N of the monitoring area_totalThe number of the covered grids of the sensor network is

Wherein N is_totalIn order to be able to count the number of sensors,

the number of grids covered by each sensor, wherein N is a positive integer;

step 4.3, obtain the space coverage of the sensor network as η ═ S_total/N_total。

In the above aspect, the step 5 specifically includes:

step 5.1, setting initial parameters including population scale N, sensor number m and maximum evolution times G_maxScaling factor F, crossover factor CR;

step 5.2, generating an initial population P₀Each of the individuals

j＝1,2,…,m，0≤φⁱ≤2π；

Step 5.3, mutation operation:

V_i(t+1)＝X_r1(t)+F×(X_r2(t)-X_r3(t))

wherein t is the evolution iteration number, and F is within 0,1]Called scaling factor, r₁,r₂,r₃E {1,2, …, N }, N being the population size, and r₁,r₂,r₃Different from the current target index i;

step 5.4, cross operation:

wherein t is the evolution iteration number, CR belongs to [0,1 ]]Called the crossover factor, rand (j) is [0,1 ]]A random number of (c); j-k guarantees test individual U_iAt least one-dimensional component of (t +1) is derived from the variant individual V_i(t + 1);

and 5.5, selecting operation:

where t is the number of evolutionary iterations, f (U)_i(t +1)) and f (X)_i(t)) is a fitness evaluation function;

step 5.6, judging whether the maximum evolution times G is met_max(ii) a If yes, ending, and outputting an optimal solution, namely a group of optimal deflection angles

Otherwise go to step 5.3.

In the above aspect, the step 6 specifically includes:

step 6.1, calculating the dormancy decision parameter SDP of all nodes in the sensor network;

step 6.2, identifying redundant nodes according to the dormancy decision parameter SDP, and sequencing CCPs of all redundant nodes in ascending order;

6.3, removing the redundant node with the minimum CCP;

and 6.4, updating the working nodes in the sensor network, and turning to the step 6.1 until the number of the nodes in the sensor network is reduced to a set numerical value.

The invention discloses a sensor network coverage enhancement method based on direction angle decomposition optimization and redundant node removal. Firstly, establishing a three-dimensional cone directional perception model conforming to a real physical environment; secondly, decomposing a coverage enhancement process into pitch angle optimization and deflection angle optimization, and providing a deflection angle optimization method based on a differential evolution algorithm to improve the spatial coverage rate of the sensor network; and finally, a redundant node removing method is provided, the number of working nodes is reduced, and the optimization of service quality and the minimization of energy consumption are realized.

Drawings

FIG. 1 is a schematic flow chart of a method for enhancing sensor network coverage based on direction angle decomposition optimization and redundant node removal in an embodiment of the present invention;

fig. 2 is a schematic diagram of a three-dimensional cone directional perception model established in the sensor network coverage enhancement method based on direction and angle decomposition optimization and redundant node removal in the embodiment of the present invention.

Detailed Description

The invention provides a sensor network coverage enhancement method based on direction angle decomposition optimization and redundant node removal, which is characterized in that a three-dimensional cone directional perception model is established; providing a deviation angle optimization method based on a differential evolution algorithm; a redundant node removal method is provided.

The following will clearly and completely describe and demonstrate the method scheme in the embodiment of the present invention with reference to the accompanying drawings 1-2, which will help to understand the present invention, but not limit the content of the present invention.

Fig. 1 is a schematic flow chart of a method for enhancing sensor network coverage based on direction angle decomposition optimization and redundant node removal in an embodiment of the present invention:

step 1, rotating by taking a sensor node P (x, y, z) as a circle center, a radius of R and a central angle of 2 β to form a spatial sensing area PO₁O₂O₃O₄And a quintuple array (P, B, α, R) is used for representing, and a three-dimensional cone directional sensing model is established, as shown in FIG. 2, wherein P ' is the projection of the sensing node P on an X-Y plane, B (lambda, phi) is the main sensing direction of P, an included angle lambda (∠ P ' PB) between the sensing node P and a Z axis is called a pitch angle, the projection on the X-Y plane is called a deflection angle phi (∠ BP ' X), and 2 α,2 β and R are respectively the horizontal sensing angle, the vertical sensing angle and the sensing radius of the sensor node.

Step 2: calculating the stereo area covered by the perception model, namely cone PO₁O₂O₃O₄The volume of (a). Wherein the vertex coordinates are respectively:

P＝(x,y,z)

O₁＝(x+|P'O₁|cos(α+φ),y+|P'O₁|sin(α+φ),0)

O₂＝(x+|P'O₂|cos(α-φ),y-|P'O₂|sin(α-φ),0)

O₃＝(x+|P'O₃|cos(α-φ),y-|P'O₃|sin(α-φ),0)

O₄＝(x+|P'O₄|cos(α+φ),y+|P'O₄|sin(α+φ),0)

wherein, | P' O₁|＝|P'O₂|＝ztan(λ-β)/|cosα|，|P'O₃|＝|P'O₄|＝ztan(λ+β)/|cosα|；

Cone PO₁O₂O₃O₄The volume S of (A) is:

and step 3: because only the pitch angle lambda is a variable in the volume formula S, the optimal coverage volume S and the corresponding optimal pitch angle lambda can be further calculated.

And 4, step 4: the accuracy of the discrete gridding of the monitoring area can be determined according to the actual requirement,such as 1 m.times.1 m; calculating the discrete grid number N of the monitoring area_totalThe number of the covered grids of the sensor network is

i is 1, …, N, N is the number of sensors,

the number of grids covered for each sensor the spatial coverage of the sensor network is obtained as η ═ S_total/N_total。

And 5: and the pitch angle of the sensor nodes is optimized, the perception overlapping area and the perception blind area between the sensor nodes are reduced, so that the space coverage rate is improved, and the optimization is carried out by utilizing a differential evolution algorithm.

Step 5.1: setting initial parameters including population scale N, sensor number m and maximum evolution times G_maxScaling factor F, crossover factor CR.

Step 5.2: generating an initial population P₀Each of the individuals

i＝1,2,…,NP，j＝1,2,…,m，0≤φⁱ≤2π。

Step 5.3: and (5) performing mutation operation.

V_i(t+1)＝X_r1(t)+F×(X_r2(t)-X_r3(t))

In the formula, t is the evolution iteration number, and F is belonged to [0,1 ]]Called scaling factor, r₁,r₂,r₃E {1,2, …, N }, N being the population size, and r₁,r₂,r₃Unlike the current target index i.

Step 5.4: and (4) performing a crossover operation.

In the formula, t is the evolution iteration number, CR is belonged to [0,1 ]]Called the crossover factor, rand (j) is [0,1 ]]The random number of (2). j-k guarantees test individual U_iIn (t +1)At least one-dimensional component of which is formed by a variant individual V_i(t + 1).

Step 5.5: and (6) selecting operation.

Where t is the number of evolutionary iterations, f (U)_i(t +1)) and f (X)_i(t)) is a fitness evaluation function.

Step 5.6: judging whether the maximum evolution times G is met_max. If yes, ending, and outputting an optimal solution, namely a group of optimal deflection angles

Otherwise go to step 5.3.

Step 6: redundant nodes in the sensor network are removed.

Step 6.1: a Sleep Decision Parameter (SDP) is calculated for all nodes in the sensor network, which is defined as follows.

Sleep Decision Parameter (SDP): setting the total number of grids covered by the sensor network to be N_tIf one of the sensor nodes P is turned off, the number of grids covered by the sensor network is N_P. Let SDP equal to 1-N_P/N_tWhen the SDP is smaller than δ (a small real number), the node P enters a sleep state, which is called a sleep decision parameter.

Step 6.2: redundant nodes are identified according to the SDP, and CCPs of all the redundant nodes are sorted in ascending order.

Step 6.3: the redundant node with the smallest CCP is removed.

Step 6.4: and updating the working nodes in the sensor network, and turning to the step 6.1 until the number of the nodes in the sensor network is reduced to a certain number.

The parts not described in the specification are prior art or common general knowledge. The embodiment is only used for illustrating the invention, and is not used for limiting the scope of the invention, and those skilled in the art should consider that the three-dimensional cone directional perception model, the bias angle optimization method based on the differential evolution algorithm, the redundant node removal method, and the like in the invention fall within the protection scope of the claims of the invention.