CN106454858B - A Method to Solve the Hot Zone Problem Existing in Multi-Hop Sensor Networks - Google Patents

Abstract

A kind of method solving the problems, such as the hot-zone present in multi-hop sensor network disclosed by the invention, comprising the following steps: S1, the topology policy model unevenly arranged using node, number of nodes are distributed in the form of arithmetic progression；S2, control area data production rate: the acquisition by each round to residue energy of node just calculates the mathematic expectaion and standard deviation of each region energy consumption by SINK, then judges whether to trigger control command；S3, the coverage met the requirements by the way that the distribution shape of node is arranged.Method of the invention, it is possible to realize its energy-consuming balance, efficient studies also in addition are carried out to the distribution shape of node, improve the utilization rate of node.

Description

A Method to Solve the Hot Zone Problem Existing in Multi-Hop Sensor Networks

technical field

The invention relates to the field of wireless sensors, in particular to a method for solving the problem of hot spots in multi-hop sensor networks.

Background technique

In wireless sensor networks, the power of nodes is usually limited and difficult to replenish, and the communication distance is not enough. In a huge node network, a multi-hop strategy is usually used to send information to the sink node, but the disadvantage of this strategy is Nodes closer to the sink node not only have to send their own data, but also serve as information routing transit nodes for nodes far away from the sink node, which undertake a larger amount of information sending tasks, and the nodes near the sink node consume more energy than other nodes. Faster, if things go on like this, once the nodes near the sink node fail earlier than other nodes, there will be a large area of the network that cannot be monitored, and the information of other nodes will be difficult to send to the sink node. This unbalanced network energy consumption is called the hot zone problem.

As shown in Figure 1, the following is the analysis and modeling of the hot area problem: Let the node density in the area be p, each ring represents the next hop area of the outer ring, and there are M rings in total, and the data generation speed of each node The same is b, i is the ring number, then the load of ring i can be understood as the ratio of the data flow to the number of nodes in ring i, then

The amount of data traffic forwarding to be undertaken by ring i is

ring _fu ＝p*((2*M*r) ² -π*(i*r) ² )*b

The total number of nodes in ring i is

ring _s = p*(π*((i+1)*r) ² -π*(i*r) ² )

The load on ring i is

It can be seen that the smaller i is, the larger the Load is, which can indicate that the closer to the sink node in the multi-hop wireless sensor network, the greater the load of the node.

The hot zone problem is essentially a problem of uneven energy distribution. In theory, the following strategy can be made: let the width of the ring be different, but the node density be uniform. Let the width of each ring decrease as the distance from the base station increases, and the nodes in each ring can be used as the next hop node of the nodes in the outer ring, so there are more nodes in the ring to undertake routing transfer , to achieve the purpose of greater energy near the sink node. But in actual operation, the node density of this strategy is uniform, and it is not realistic to increase the energy near the converging node by drawing a large circular area in the drawing.

It is very necessary to alleviate the hot spot problem through topology design in a multi-hop wireless sensor network, and there have been many studies on this aspect, such as a strategy for the uneven initial energy distribution of nodes proposed in literature [1] and literature [1]. [2] proposes a functional relational formula that uses distance to determine the number of node distributions, but both have the disadvantage of being difficult to operate in reality.

The literature [1] is: W.R.Heinzelman, A.Chandrakasan, H.Balakrishnan. Energy-efficient communication protocol for wireless micro-sensor networks[C].InProc.of the 33rd Annual Hawaii International Conference on System Sciences,January 2000.

The literature [2] is: Lian J, Naik K, Agnew G.Date Capacity Improvement of Wireless Sensor Networks Using Non-Uniform Sensor Distribution[J].inter-nationalJournal of Distributed Sentor Networks,2006,2(2):121 -1.

The literature is [3]: Lu Kezhong, Liu Yingling. A node layout scheme for linear wireless sensor networks [J]. Computer Applications, 2007,27(7):1566-1568

Contents of the invention

The purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art, and provide a data collection method based on the adaptive collection rate of the node linear topology.

The purpose of the present invention is achieved through the following technical solutions:

A method for solving the hot zone problem existing in a multi-hop sensor network, comprising the following steps:

S1. A topological strategy model with uneven distribution of nodes is adopted, and the number of nodes is distributed in the form of an arithmetic sequence;

S2. Data generation rate in the control area: Through the collection of the remaining energy of the nodes in each round, calculate the mathematical expectation and standard deviation of energy consumption in each area through SINK, and then judge whether to trigger the control command;

S3. Satisfy the required coverage by setting the distribution shape of the nodes.

In step S1, the number of nodes is distributed in the form of an arithmetic sequence, specifically: the density of the nodes in the ring decreases sequentially with the increase of the distance from the base station, and the sequential decrease satisfies the arithmetic sequence. In this way, the purpose of greater energy near the sink node can be achieved, and more energy can be used to undertake the task of energy transfer.

The step S1 is specifically:

S101. Divide the network: Divide the linear network into n regions, mark each region as a _i , set the length of each region as L, the area of each region as S, and the node density in region a _i as ρ _i ; where 1≤i≤n;

S102. Calculate the total energy consumption of the area a _i :

Record the power e of each node in the area, and assume that the data generation rate in the area is the same, and each round is M; it can be seen from the energy consumption formula of the sensor sending module that the data sending distance has a great impact on energy consumption, and the formula The values of 2 and 4 for the middle distance are also closely related to the actual environment. If the outdoor transmission conditions are good, it can usually be taken as 2. In places with poor transmission environments such as indoors or tunnels, it can only be taken as 4. However, the original research is on the linear network topology applicable to bridges, and bridges belong to the environment with good outdoor transmission conditions, so this technical solution takes 2.

Since the data generation rate of each area is the same and has nothing to do with the number of nodes, the amount of data received and sent by each area in each round is n*M;

From a macro point of view, the area a _n needs to forward all the data collected in the area and generated in the peripheral area. Since the topology of the cluster is adopted in the reception, the amount of data received can also be regarded as receiving the data generated in the area and the peripheral area. data, the data transmission distance is assumed to be the center distance L between regions; then for region a _i :

The data consumption energy E _Rx received in each round is

E _Rx =E _clec *n*M;

Where E _clec is the energy consumed by receiving a unit message;

The energy E _Tx consumed by sending data in each round is

E _Tx =n*M*(E _elec +E _amp *L ² );

Among them, E _elec is the loss energy of the transmitted unit message; E _amp is the amplification power energy;

The total energy consumption E of area a _i in each round is

E＝E _Rx +E _Tx ＝n*M*(E _clec +E _elec +E _amp *L ² );

S103, calculate the ratio K of the total energy of the area and the total energy consumption of the area:

To make the overall energy consumption of the network uniform, it can be understood that the energy consumption of each node and each area is balanced, and to make the life of the area and each node close, the energy consumption is balanced, that is, to make the nodes in each area The ratio of the total energy to the energy consumption of each round is consistent, so that the energy consumption of each region can be balanced from the overall network, so that the energy consumption of each node can also be balanced in general;

The total energy of each area a _i is

ρ _i *S*e

Record the ratio of the total energy of area a _i to the total energy consumption of each round of area a _i as K:

S104, use ρ _i in K to show the total number of nodes and find the expression of ρ _i with respect to the total number of network nodes g, the area is S; then

Then the total number of network nodes is g as

but

Simplification:

substitute

Since M, E _clec , E _elec , E _amp , and L are all fixed values, the simplification is

Then the number of nodes a _s in area a _i is:

It can be seen from this formula that the closer to the converging node, the more nodes in area a _i , and the relationship between the number of nodes between areas is an arithmetic sequence relationship. The existing technology also proposes some topology designs, but they are not perfect. For example, a strategy with different initial energy of nodes is proposed in [2]. It is difficult to operate, and this approach is not realistic. Literature [3] also proposed a node deployment function formula that can make the network energy distribution uneven, but in fact, if it is arranged according to a formula of node density about the distance from the sink node, it is quite difficult to operate. In order to balance energy consumption And it has a certain operability in reality. The present invention divides the network into n equally divided areas, and the node density of the area decreases as the distance from the converging node increases. Strong operability in reality.

The step S2 is specifically:

It is set that the nodes in an area are assembled into a cluster, and the members do not change after n rounds, that is, nodes in other areas cannot become members of this cluster;

Since the cluster head can collect the energy information of each member, together with the sensing data, the cluster head calculates the energy distribution, so in order to balance the communication energy consumption between the cluster heads and the cluster head and the node members in the area The energy consumption of the communication between the nodes, this design adopts the mechanism of electing the cluster head according to the remaining energy of the nodes in the area;

From the first round, the cluster head of each area is artificially set first. During the period when the node members start to collect data and communicate with the cluster head, they send their remaining energy remaining information to the cluster head node. At the moment before the rotation, According to the energy status of the members, the cluster head sends the designated information to the member node with the largest remaining energy. If there is more than one node with the most remaining energy, the cluster head is randomly selected from the members with the largest remaining energy to complete the rotation.

The step S2 is specifically:

Each round of the cluster head can grasp the remaining energy of its own members. After the cluster head sends the information to the sink node, the sink node obtains the remaining energy of the nodes in each area as a whole, and then calculates the remaining energy e _j of each node and the The average remaining energy E _j , the average value of remaining energy E _ave , where 1≤j≤g, then

After obtaining E _ave , the sink node can know the deviation of the remaining energy of each region from E _ave by calculating the standard deviation X and variance x of the average remaining energy E _j and E _ave in each region;

Standard DeviationX:

Variance x:

By presetting the threshold K, when K is at a specific level, the sink node broadcasts and controls to reduce the data occurrence rate of the area where the energy of the remaining nodes in the area is lower than the average E _ave and deviates the most.

Since the energy consumption of nodes is mainly related to the communication distance, the uncertainty of the routing path under the node arrangement strategy and the cluster head election mechanism can make the network energy consumption have a certain balance, but this random balance is not enough Well, especially in areas far away from the sink node, the number of nodes distributed is far from the number of nodes in the area closer to the sink node, so the possibility of routing paths formed with the next-hop cluster head is relatively far away from the sink node The difference in the number of possible routing paths in the near area is also very large. Since the energy adjustment of the nodes in the area is adjusted by electing the node with the largest remaining energy as the cluster head, this difference in the number of routing paths makes the area far away from the sink node The energy of the node is relatively less likely to obtain a routing path with less energy consumption. In the case of multiple work rotations, the area far away from the sink node is relatively more likely to become the most energy-consuming area. fast area. In order to alleviate the problem that energy consumption in areas far away from the sink node is faster than other areas, this technical solution proposes a mechanism for the sink node to broadcast and control the data generation rate of the area with excessive energy consumption, which can further balance the energy consumption of the network.

Step S3, the distribution shape of the nodes, when the nodes are set on both sides of the bridge, the distribution shape of the nodes includes: paired isosceles triangle, paired rectangle.

The specific setting process of the isosceles triangle is as follows:

The induction radius of nodes O, A, and B is r, the width of the bridge model is h, J is one of the nodes, the base OA of the isosceles triangle is equal to d, assuming that r>h, the triangle OAB must be seamlessly covered, and the coverage degree is at least 2, the length of the line segment needs to satisfy OJ≥OA; in the deployment of an isosceles triangle with a coverage degree of at least 2, it must meet:

If it is necessary to meet the seamless coverage with a coverage of at least 3, the triangle OAB must all be within the monitoring range of the circles O, A, and B, and it must meet:

r≥d.

For the pair of rectangles, the specific setting process is:

The induction radius of nodes A, B, C, and D is r, the width of the bridge model is h, the base of the isosceles triangle is AB=d, assuming that r>h, to achieve the seamless coverage of the rectangular ABCD coverage of at least 2, E, F , G are one of the nodes, and the area EFG only covers the area covered by circle A to be covered by circle B or D, then AG+FB≥AB is required, that is

Similarly, to satisfy the seamless coverage of the rectangle ABCD coverage of at least 3, at least the vertical distance between the intersection point E of circle A and circle C to AC is greater than AB, that is

This formula is also a satisfying condition for the seamless coverage of the rectangular ABCD coverage of at least 4.

Compared with the prior art, the present invention has the following advantages and beneficial effects:

The present invention proposes a linear topology strategy suitable for real bridges. This strategy adopts the strategy of uneven distribution of nodes to avoid the problem of uneven network energy consumption caused by hot zone problems, and calculates the distribution of nodes in a specific area through mathematical calculations. Quantity, the strategy selects the cluster head based on the remaining energy of the node in the clustering mechanism, and adopts a method of sacrificing the amount of data collection to further balance the energy consumption of the network. Under these strategies, the energy balance is confirmed in the simulation In addition, the efficiency research on the distribution shape of the nodes is also carried out, which improves the utilization rate of the nodes.

Description of drawings

Figure 1 is a schematic diagram of modeling the hot zone problem.

FIG. 2 is a flow chart of a method for solving the hot spot problem existing in a multi-hop sensor network according to the present invention.

FIG. 3 is a flow chart of cluster head election in the method shown in FIG. 2 .

Fig. 4 is a simulation diagram of unevenly distributed node residual energy.

Fig. 5 is a simulation diagram of uniformly distributed residual energy of nodes.

Figure 6 is a node location diagram.

FIG. 7 is a simulation diagram of the remaining energy of a node without a data generation rate adjustment mechanism.

FIG. 8 is a simulation diagram of the variance trend under the data generation rate adjustment mechanism.

FIG. 9 is a simulation diagram of the remaining energy of a node under the data generation rate adjustment mechanism.

Fig. 10 is a graph comparing the efficiency of a rectangle with a coverage of 2 and an isosceles triangle.

Figure 11 is a comparison diagram of the efficiency of a rectangle with a coverage of 3 and an isosceles triangle.

Detailed ways

The present invention will be further described in detail below in conjunction with the embodiments and the accompanying drawings, but the embodiments of the present invention are not limited thereto.

As shown in Figure 2, a method for solving the hot zone problem existing in a multi-hop sensor network includes the following steps:

S1. A topological strategy model with uneven distribution of nodes is adopted, and the number of nodes is distributed in the form of an arithmetic sequence;

S2. Data generation rate in the control area: Through the collection of the remaining energy of the nodes in each round, calculate the mathematical expectation and standard deviation of energy consumption in each area through SINK, and then judge whether to trigger the control command;

S3. Satisfy the required coverage by setting the distribution shape of the nodes.

In step S1, the number of nodes is distributed in the form of an arithmetic sequence, specifically: the density of the nodes in the ring decreases sequentially with the increase of the distance from the base station, and the sequential decrease satisfies the arithmetic sequence.

The step S1 is specifically:

S101. Divide the network: Divide the linear network into n regions, mark each region as a _i , set the length of each region as L, the area of each region as S, and the node density in region a _i as ρ _i ; where 1≤i≤n;

S102. Calculate the total energy consumption of the area a _i :

Record the power e of each node in the area, and assume that the data generation rate in the area is the same, and each round is M; then the amount of data received and sent by each area in each round is n*M;

The data transmission distance is assumed to be the center distance L between regions; then for region a _i :

The data consumption energy E _Rx received in each round is

E _Rx =E _clec *n*M;

Where E _clec is the energy consumed by receiving a unit message;

The energy E _Tx consumed by sending data in each round is

E _Tx =n*M*(E _elec +E _amp *L ² );

Among them, E _elec is the loss energy of the transmitted unit message; E _amp is the amplification power energy;

The total energy consumption E of area a _i in each round is

E＝E _Rx +E _Tx ＝n*M*(E _clec +E _elec +E _amp *L ² );

S103, calculate the ratio K of the total energy of the area and the total energy consumption of the area:

The total energy of each area a _i is

ρ _i *S*e

Record the ratio of the total energy of area a _i to the total energy consumption of each round of area a _i as K:

S104, use ρ _i in K to show the total number of nodes and find the expression of ρ _i with respect to the total number of network nodes g, the area is S; then

Then the total number of network nodes is g as

but

Simplification:

substitute

Since M, E _clec , E _elec , E _amp , and L are all fixed values, the simplification is

Then the number of nodes a _s in area a _i is:

It can be seen from this formula that the closer to the converging node, the more nodes in area a _i , and the relationship between the number of nodes between areas is an arithmetic sequence relationship.

The step S2 is specifically:

After the cluster head sends the information to the sink node, the sink node obtains the remaining energy of the nodes in each region as a whole, and then calculates the remaining energy e _j of each node and the average remaining energy E _j of each region, and the average value of remaining energy E _ave , where 1≤j ≤g, then

After obtaining E _ave , the sink node can know the deviation of the remaining energy of each region from E _ave by calculating the standard deviation X and variance x of the average remaining energy E _j and E _ave in each region;

Standard DeviationX:

Variance x:

By presetting the threshold K, when K is at a specific level, the sink node broadcasts and controls to reduce the data occurrence rate of the area where the energy of the remaining nodes in the area is lower than the average E _ave and deviates the most.

As shown in Figure 3, the step S2 is specifically:

It is set that the nodes in an area are assembled into a cluster, and the members do not change after n rounds, that is, nodes in other areas cannot become members of this cluster;

Since the cluster head can collect the energy information of each member, along with the sensing data, the cluster head calculates the energy distribution;

From the first round, the cluster head of each area is set. During the period when the node members start to collect data and communicate with the cluster head, they send their remaining energy and remaining information to the cluster head node. At the moment before the rotation, the cluster head according to For the energy status of the members, the specified information is given to the member node with the largest remaining energy. If there is more than one node with the largest remaining energy, the cluster head is randomly selected from these members with the largest remaining energy to complete the rotation.

Step S3, the distribution shape of the nodes, when the nodes are set on both sides of the bridge, the distribution shape of the nodes includes: paired isosceles triangle, paired rectangle.

The specific setting process of the isosceles triangle is as follows:

The induction radius of nodes O, A, and B is r, the width of the bridge model is h, J is one of the nodes, the base OA of the isosceles triangle is equal to d, assuming that r>h, the triangle OAB must be seamlessly covered, and the coverage degree is at least 2, the length of the line segment needs to satisfy OJ≥OA; in the deployment of an isosceles triangle with a coverage degree of at least 2, it must meet:

If it is necessary to meet the seamless coverage with a coverage of at least 3, the triangle OAB must all be within the monitoring range of the circles O, A, and B, and it must meet:

r≥d.

For the pair of rectangles, the specific setting process is:

The induction radius of nodes A, B, C, and D is r, the width of the bridge model is h, the base of the isosceles triangle is AB=d, assuming that r>h, to achieve the seamless coverage of the rectangular ABCD coverage of at least 2, E, F , G are one of the nodes, and the area EFG only covers the area covered by circle A to be covered by circle B or D, then AG+FB≥AB is required, that is

Similarly, to satisfy the seamless coverage of the rectangle ABCD coverage of at least 3, at least the vertical distance between the intersection point E of circle A and circle C to AC is greater than AB, that is

This formula is also a satisfying condition for the seamless coverage of the rectangular ABCD coverage of at least 4.

In order to verify whether the above node distribution strategy and control data generation rate control mechanism can really achieve energy balance and node deployment shape efficiency, this section uses Matlab for programming simulation verification and analyzes the results.

1. Comparison of energy consumption between uneven deployment and uniform deployment of nodes

parameter settings

Table 1 parameter settings

Simulation results are shown in Figures 4 and 5.

The experimental results show that, under the setting of the above parameters, in the strategy of uneven distribution of nodes, the first node dies in the 549th round, and the remaining total energy of the node is 0.91, accounting for 3.3% of the initial total energy, while the node is uniform Distribution The first node died in the 121st round, and the remaining total energy of the node was 13.50, accounting for 49.09% of the initial total energy. From this result, it can be seen that the strategy of uneven distribution of nodes can indeed improve the uniformity of network energy consumption and improve the life of the network.

2. The effect of data generation rate adjustment mechanism

Table 2 parameter setting

The experimental results are shown in Figures 6, 7, 8, and 9.

Analysis of results:

Due to the random distribution of nodes, the two mechanisms are used to take the average of 3 experiments, and the average total remaining energy with the data generation rate adjustment mechanism is 5.35, and the number of death rounds of the first node is 1049, while that without the data generation rate adjustment mechanism The average total remaining energy is 9.38, and the average number of death rounds for the first node is 933. The experimental results show that the data generation rate adjustment mechanism can indeed further extend the life of the network to a certain extent.

3. Efficiency comparison of node layout shapes

Table 3 parameter setting

bridge width 8 meters Node Induction Radius 10 m

The experimental results are shown in Figures 10 and 11. The experimental results show that the rectangular distribution is more efficient than the isosceles triangular distribution when the coverage is at least 2, and the isosceles triangular distribution is more efficient than the rectangular distribution when the coverage is at least 3.

According to the above analysis, if the bridge of 1000m*8m is divided into 10 areas, and the length of each area is 100m, then the first area away from SINK can use 5 nodes to achieve a network with a coverage of 1. When the first Arrange 5 nodes in the area. According to the previous work, it means that 10 nodes are placed in the second area, and 15 nodes are placed in the third area. By analogy, starting from the third area, the coverage is 2, the highest Efficient rectangular arrangement. Starting from the fifth area, it is arranged in an isosceles triangle with a coverage of 3 and the highest efficiency. Starting from the seventh area, the coverage is 4 and the most efficient rectangle is arranged. So as to achieve the purpose of not wasting node functions and improving monitoring accuracy.

The above-mentioned embodiment is a preferred embodiment of the present invention, but the embodiment of the present invention is not limited by the above-mentioned embodiment, and any other changes, modifications, substitutions, combinations, Simplifications should be equivalent replacement methods, and all are included in the protection scope of the present invention.

Claims (4)

1. a kind of method for solving the problems, such as the hot-zone present in multi-hop sensor network, it is characterised in that: the following steps are included:

S1, the topology policy model unevenly arranged using node, number of nodes are distributed in the form of arithmetic progression；

S2, control area data production rate: the acquisition by each round to residue energy of node just calculates each area by SINK The mathematic expectaion and standard deviation of domain energy consumption, then judge whether to trigger control command；

S3, the coverage met the requirements by the way that the distribution shape of node is arranged；

The distribution shape of the node, when node is arranged at bridge two sides, the distribution shape of node includes: to isoceles triangle Shape, to rectangle；

It is described to isosceles triangle, specific setting up procedure are as follows:

The induction radius of node O, A, B are r, and bridge model width is h, and J is one of node, bottom OA of isosceles triangle etc. In d, it is assumed that r > h will meet triangle OAB by seamless coverage, and coverage is at least 2, then need line segment length meet OJ >= OA；It is at least 2 isosceles triangle deployment in coverage, it is necessary to meet:

If desired meet the seamless coverage that coverage is at least 3, then triangle OAB must all circle O, A, B monitoring range in, Then it must satisfy:

r≥d。

2. solving the problems, such as the method for the hot-zone present in multi-hop sensor network according to claim 1, it is characterised in that: In step S1, the number of nodes is distributed in the form of arithmetic progression, specifically: the density of annulus interior nodes with base station The increase of distance and successively reduce, successively reduce meet arithmetic progression.

3. solving the problems, such as the method for the hot-zone present in multi-hop sensor network according to claim 1, which is characterized in that The step S2, specifically:

The node set in a region is set into cluster, and member does not change by n wheel, i.e. the node in other regions can not Member as this cluster；

Since cluster head can collect the energy information of each member, it is subsidiary sensing data together, cluster head calculates Energy distribution Situation；

The cluster head in each region is set since the first round, during node member starts to acquire data and communicate with cluster head, The energy residual information of self residual is sent to leader cluster node together, and in the eve of rotation, cluster head is according to the energy feelings of member Condition gives dump energy maximum member node specify information, if the most node of dump energy has more than one, surplus Cluster head is selected in these maximum members of complementary energy at random, completes rotation.

4. solving the problems, such as the method for the hot-zone present in multi-hop sensor network according to claim 1, it is characterised in that: It is described to rectangle, specific setting up procedure are as follows:

The induction radius of node A, B, C, D are r, and bridge model width is h, the bottom AB=d of isosceles triangle, it is assumed that r > h will reach To the seamless coverage of rectangle ABCD coverage at least 2, E, F, G are one of node, and region EFG is only by circle A covering Region will by circle B or D cover, then need AG+FB >=AB, i.e.,

Equally to meet the seamless coverage that rectangle ABCD coverage is at least 3, then at least need the intersection point E of round A and circle C to AC's Vertical range is greater than AB, i.e.,

The formula be also simultaneously rectangle ABCD coverage be at least 4 seamless coverage meet condition.