CN106211187A - A kind of water sound sensor network dynamic gateway node deployment method based on prediction - Google Patents

Abstract

The network node that the invention belongs to water sound sensor network disposes field, it is characterized in that based on prediction, dynamic gateway deployment method, it is possible to ensures that in specifying duration gateway coverage area maximizes, extends to the bit error rate when reducing Information Communication.In the gateway deployment method of the present invention, first based on prediction, dynamic gateway deployment method is proposed, and it is carried out mathematical modeling, then introduce prediction algorithm and linear optimization method solves the gateway deployment method optimization problem that the present invention is put forward, and then the maximization of gateway coverage area in realization appointment duration, extend to the bit error rate when reducing Information Communication.

Description

A kind of water sound sensor network dynamic gateway node deployment method based on prediction

Technical field

The network node that the invention belongs to water sound sensor network disposes field, it is characterized in that based on prediction, dynamic gateway deployment method, it is possible to ensures that in specifying duration gateway coverage area maximizes, extends to the bit error rate when reducing Information Communication.In the gateway deployment method of the present invention, first based on prediction, dynamic gateway deployment method is proposed, and it is carried out mathematical modeling, then introduce prediction algorithm and linear optimization method solves the gateway deployment method optimization problem that the present invention is put forward, and then the maximization of gateway coverage area in realization appointment duration, extend to the bit error rate when reducing Information Communication.

Background technology

Water sound sensor network is by having the underwater monitoring network system that the acoustic communication sensor node with computing capability is constituted, it is deployed in the underwater environments such as ocean, realizing the contamination monitoring of underwater environment, aquatic organism sample collection, the aspects such as natural disaster is prevented, assisting navigation possess wide application prospect.

In water sound sensor network, need to monitor and detect institute's event under water by disposing by the way of underwater sensor node, such as environmental pollution and biological specimen collection etc., and by water sound sensor network, the information collected is transferred to mixed node.It it is the gateway node (as shown in Figure 1) being in the water surface than more typical mixed node, the data of collection can be forwarded in the control website on bank base or boats and ships by the way of radio wave by it, when disposing multiple gateway node, sensor node is only capable of forwarding packet to its adjacent gateway node simultaneously.Owing to the spread speed of radio wave is significantly larger than the spread speed of underwater sound ripple, and bandwidth is also much larger than the bandwidth of underwater sound ripple, then overall transmission time and propagation delay will greatly reduce.Further, since the energy that underwater sound transmission is consumed is significantly larger than the energy that radio wave transmissions is consumed, then wastage in bulk or weight energy also will reduce.All of gateway node is combined with control website can be as virtual groove node.

Be obtained in that propagation delay and energy loss gain by disposing gateway node, but how to obtain optimum gain when disposing multiple gateway node and will be affected by factors, as gateway node deployed position, dispose how many gateway nodes etc..The deployment of gateway node simultaneously also needs to meet certain performance matrix to obtain optimum gain, such as life cycle, average end-to-end time delay, coverage rate and bonding ratio etc..Although current existing static network articulare deployment scheme can be by assuming to describe channel status to the prior information of network and statistical model, MAC protocol, Routing Protocol and mobility model etc., but owing to it does not accounts for the possible change of environment (as task changes, mobility and sensor node state etc.), it is likely to result in gateway node deployment scheme and lost efficacy.

In order to obtain the maximum gain disposing gateway node, then need the overlay area allowing gateway node be obtained in that maximum, so that its sensor node covered only needs to nearest gateway node for transmitting packet.In Underwater Acoustic Environment, mobility due to underwater sensor node, overlay area is moment change, if then gateway node is disposed according to the current location of sensor node, after continuing for some time, the overlay area of gateway node may diminish (as shown in Figure 2), as can be seen from the figure, in 210, the sensor node of (example 1) will remove the overlay area of gateway node after continuing for some time, the overlay area making gateway node diminishes, and for gateway deployment scheme 211 (example 2), even if sensor node there is also same mobility model, but within continuing for some time, within sensor node is still in the overlay area of gateway node.The present invention proposes and how to dispose gateway node it can be allowed to obtain maximal cover region within the persistent period, and then reduces propagation delay and the bit error rate.

If the quantity of gateway node is given, then for obtaining the maximal cover region problem of gateway node, it is converted to how to find the optimum position optimization problem of gateway node.The present invention carries out mathematical modeling, and the method proposing to use optimized algorithm to solve this optimization problem to the optimum position optimization problem how finding gateway node.

Summary of the invention

The purpose of the present invention:

In water sound sensor network, need to monitor or detect the event of underwater environment institute by the way of using and disposing underwater sensor node, in order to reduce propagation delay and reduce the energy that sensor node is consumed, the mode of disposing gateway node is usually taken, and the deployment scheme of gateway node affects the performance of water sound sensor network.

The present invention is by proposing based on prediction, dynamic gateway node dispositions method and gateway node dispositions method carrying out mathematical modeling and uses the optimum position of linear optimization algorithm searching deployment gateway node to improve the performance of gateway node dispositions method in water sound sensor network, and then extends to the bit error rate when reducing Information Communication.

Technical scheme:

In water sound sensor network, for gateway node is disposed, its scheme has multiple choices, selects the impact produced by different gateway node deployment scheme water sound sensor network performances can be different.

The present invention proposes based on prediction, dynamic gateway node dispositions method, it can maximize the overlay area of gateway node in specifying duration, and give a kind of mathematical modeling scheme, to find the optimum position that gateway node is disposed, and then reduce propagation delay time and the bit error rate.

The present invention gives a kind of method that gateway node deployment scheme is modeled as linear optimization problem, then this optimization problem is solved by Branch-and-Cut algorithm, and then the optimum position of acquisition gateway node, reduce propagation delay time and the bit error rate of water sound sensor network.

Beneficial effects of the present invention:

In water sound sensor network, the quality of gateway node deployment scheme can have a strong impact on the performance of sensor network, which determine the propagation delay of network, the bit error rate and energy expenditure, and the deployment scheme of gateway node can be affected by factors, such as position, the mobility etc. of sensor node of gateway node.

1) present invention is by dynamic gateway dispositions method based on prediction, and it can maximize the overlay area of gateway node in specifying duration, and then reduce propagation delay and the bit error rate of water sound sensor network.

2) present invention passes through Mathematical Modeling Methods, linear optimization problem will be modeled as based on the dynamic gateway dispositions method of prediction, then this optimization problem is solved by Branch-and-Cut algorithm, and then the optimum position of acquisition gateway node, reduce propagation delay and the bit error rate of water sound sensor network.

Accompanying drawing explanation

Fig. 1 schematic network structure

Fig. 2 x time optimizes schematic diagram

In Fig. 1,101 represent gateway node, 102 represent sensor node, 103 represent control website, 104 represent water surface, 105 overlay areas representing gateway node, 106 represent gateway node and control by radio wave transmissions between website, and 107 represent between gateway node and sensor node by water sonic transmissions.

In Fig. 2,210 is example 1, and 211 is example 2, and 201 represent gateway node, 202 represent sensor node, 203 represent control website, and 204 represent water surfaces, and 205 represent the overlay area of gateway nodes, 206 represent gateway node and control to pass through radio wave transmissions between website, 207 represent by water sonic transmissions between gateway node and sensor node, and 208 represent the anticipated movement position of sensor node, and 209 represent the shift position of sensor node.

Detailed description of the invention

Shown below is one and gateway node deployment issue is converted into linear optimization problem, and solve this linear optimization problem by Branch-and-Cut algorithm, to obtain the mathematical model scheme of the optimum position of gateway node, and then it is effectively reduced propagation delay and the bit error rate in water sound sensor network.

The present invention is not limited only to this mathematical model scheme; it is every when the gateway node carrying out water sound sensor network disposes research; the gateway node utilizing the present invention disposes thinking; i.e. use dynamic gateway node deployment scheme based on prediction; and this gateway node deployment scheme is modeled as linear optimization problem; and then use Branch-and-Cut algorithm to solve this optimization problem, all within protection scope of the present invention.

The mathematical model scheme detailed description of the invention that gateway node is disposed is described as follows:

(1) assumed condition

The positional information of sensor node i is expressed as p_i=[p_i ⁰, p_i ¹, p_i ²..., p_i ⁿ], wherein, p_i ⁰For the current location of sensor node i, its value is by accessed by positioning service；p_i ¹, p_i ²..., p_i ⁿFor the desired location of sensor node i, its value can be obtained by IMM algorithm predicts.Assuming total " m " the individual sensor node of layer 1 at wireless sensor network, and can predict " n " individual time slot of each sensor node, definition " Y " is:

All of gateway node and sensor node, its transmission range is all defined as " d ", and the sensor node being also equivalent to be covered by gateway node is retrained by formula (1-2):

$\left|\right|C_q-p_i^k\left|\right|\·Y_{\mathrm{kiq}}\≤d---(1-2)$

Wherein, C_qPosition for gateway node q.From formula (1-2) if it can be seen that C_qWithBetween distance more than d, then Y_kiqShould be " 0 "；If C_qWithBetween distance less than d, then Y_kiqValue can be " 0 ", it is possible to for " 1 ".Being need to be maximized yet with Y, it should tend to value for " 1 ".Therefore, introducing based on Y matrix, if sensor node is capped, it will be easier to be defined and go out with formula comparison.

(2) object function

Dynamic gateway deployment scheme is modeled as the object function involved by optimization problem by formula (1-3) expression, it can be interpreted: for all nodes, all time points, by finding the accurate location disposed needed for gateway, it is desirable that node can be capped as far as possible under water, no matter it by which water surface gateway is covered.This is also equivalent to, and in fig. 2, we allow to cover the maximum area of all " dummy node ".

$\underset{C_q}{\max }\left(\underset{k=0}{\overset{n}{\mathrm{\Σ}}}\underset{i=0}{\overset{m}{\mathrm{\Σ}}}\underset{q}{\max }{Y}_{\mathrm{kiq}}\right)---(1-3)$

(3) constraints

In dynamic gateway node deployment scheme of the present invention, for the overlay area of gateway node, the restriction of 4 aspects need to be considered: gateway node number, heavily dispose energy loss, smallest coverage area and connectivity.Separately below constrained is described:

1. gateway node number

In the dynamic gateway node deployment scheme of the present invention, gateway node number is limited and unmodifiable, and it meets the constraints of formula (1-4).

1≤q≤N (1-4)

The most heavily dispose energy loss

During the heavily deployment of gateway node, each gateway node is when its initial position moves to certain new position, it is required for extra consumed energy, and the energy entrained by gateway node is limited and gateway node needs to retain certain energy so as to be moved back on bank or boats and ships to re-start control.Based on above-mentioned consideration, all there is maximum effectively displacement in each gateway node q, if it is " λ_q", also assume that the initial position of gateway node q isThen it need to meet formula (1-5).

$\left|\right|C_q-C_q^0\left|\right|\≤{\mathrm{\λ}}_q,\∀q---(1-5)$

3. smallest coverage area

For " dummy node " (shown in Fig. 2), it is not actual sensor node, and any instant, only m sensor node is in esse.In order to avoid due to a certain gateway node specifying the moment to dispose more, and ignore other moment real gateway node and cover, then for any instant, all specify smallest coverage area, it is assumed that represent with γ, then can obtain formula (1-6).

$\underset{i=0}{\overset{m}{\mathrm{\Σ}}}\underset{q}{\max }{Y}_{\mathrm{kiq}}\≥\mathrm{\γ},\∀k---(1-6)$

4. connectivity

When disposing gateway node, in order to ensure the connectivity of all the sensors node, sensor node will be by cluster at any one time.Each cluster is made up of one group of interconnective sensor node, with the sensor node of other clusters without being connected.If more than one cluster, in arbitrary cluster, it is necessary at least a sensor node is covered by gateway node, it is assumed that φ (v) represents cluster v, then should meet the constraints of formula (1-7).

$\underset{i\∈\mathrm{\φ}\left(v\right)}{\mathrm{\Σ}}\underset{q}{\mathrm{\Σ}}{Y}_{\mathrm{kiq}}\≥1,\∀k,v---(1-7)$

(4) linear transformation

Above-mentioned formula (1-2), formula (1-3), formula (1-6) are nonlinear equation, need to be translated into linear equality, then use Branch-and-Cut algorithm to solve to dispose the position problems of gateway node.

For formula (1-2), formula (1-3) and formula (1-6), in order to be translated into linear equality, introduce new variables X, and be defined as,

It is thus possible to the relational expression obtained between X and Y (1-9),

$X_{\mathrm{ki}}\≤\underset{q=0}{\overset{N}{\mathrm{\Σ}}}{Y}_{\mathrm{kiq}}---(1-9)$

Wherein, in inequality (1-9), N is effective gateway node number.Thus object function expression formula (1-3) can be converted into formula (1-10),

$\underset{C_q}{\max }\left(\underset{k=0}{\overset{n}{\mathrm{\Σ}}}\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{X}_{\mathrm{ki}}\right)---(1-10)$

Formula (1-6) can be converted into formula (1-11),

$\underset{i=0}{\overset{m}{\mathrm{\Σ}}}{X}_{\mathrm{ki}}\≥\mathrm{\γ},\∀k---(1-11)$

For each Y in formula (1-2), introduce independent binary variable Z, thus formula (1-2) can conversion formula (1-12),

Y≤M·(1-Z) (1-12)

In formula (1-12), M is a bigger positive number, and Z formula (1-13) represents,

And formula (1-13) can be converted into formula (1-14),

$Z_{\mathrm{kiq}}\≥\left(\right||C_q-p_i^k||\·Y_{\mathrm{kiq}}-d)/M---(1-14)$

In formula (1-14), if meetingThen the right-hand component of formula (1-14) should be the decimal number of 0～1, and then can to obtain Z be 1.If meetingThen the right-hand component of formula (1-14) is negative, and exist Z be 0 or be 1 two kinds may.

Based on constrained described above, for formula (1-12), it is possible to obtain, if meetingThen Z is 1, and Y is 0.If meetingThen Z is unfixed value, and Y is also unfixed value, but due to maximization problems, Y will tend to be chosen for 1.

After above-mentioned flow path switch, all of constrained function and object function are all linear functions, then the optimization problem disposing gateway node just can be solved by Branch-and-Cut algorithm.

Claims (4)

1. a water sound sensor network dynamic gateway node deployment method based on prediction, it is characterised in that: propose based on prediction , dynamic gateway deployment method, it is possible to ensureing that gateway coverage area maximizes in duration specifying, the information that reduces passes Sowing time extends to the bit error rate.

2. water sound sensor network dynamic gateway node deployment method based on prediction, present invention proposition as claimed in claim 1 Introduce prediction algorithm and linear optimization method solves gateway deployment method optimization problem, and then realize specifying gateway in duration The maximization of node coverage areas, extends to the bit error rate when reducing Information Communication.

3., as claimed in claim 1 based on the water sound sensor network dynamic gateway node deployment method predicted, the present invention is given A kind of based on prediction, dynamic gateway deployment method mathematics modeling scheme, can be converted by this data model scheme For the optimization problem of gateway node deployed position, when reducing Information Communication, extend to the bit error rate.

4., as claimed in claim 2 based on the water sound sensor network dynamic gateway node deployment method predicted, the present invention is given A kind of mathematical model using prediction algorithm and linear optimization method to solve gateway deployment scheme optimization problem, by this data mould Gateway deployment scheme can be converted into linear optimization problem by type, extends to the bit error rate when reducing Information Communication.