CN104349356B - Video sensor network coverage enhancement implementation method based on differential evolution - Google Patents
Video sensor network coverage enhancement implementation method based on differential evolution Download PDFInfo
- Publication number
- CN104349356B CN104349356B CN201310335322.7A CN201310335322A CN104349356B CN 104349356 B CN104349356 B CN 104349356B CN 201310335322 A CN201310335322 A CN 201310335322A CN 104349356 B CN104349356 B CN 104349356B
- Authority
- CN
- China
- Prior art keywords
- node
- video
- coverage
- video sensor
- perceived direction
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/18—Network planning tools
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/02—Arrangements for optimising operational condition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The video sensor network coverage enhancement implementation method based on differential evolution that the invention discloses a kind of, to monitor areal coverage as object function, using the global optimization ability corrdinated adjustment video node perceived direction of differential evolution algorithm to reduce network blind area of coverage.It is related to multimedia sensor network covering control and two fields of intelligence computation.First according to monitoring region actual environment, by continuous two dimensional surface discretization, and coordinate information is uploaded to control centre;Then, coordinate information, perceived direction are sent to aggregation node by a certain number of video nodes of random placement, each node;Aggregation node summarizes forwarding information to control centre, using video node perceived direction angle as decision variable, monitoring areal coverage is object function, node direction control instruction is issued to video sensor network to adjust node perceived direction to obtain the maximization of coverage rate by three differential evolution algorithm variation, intersection, selection operation Step wise approximation optimal node perceived directions.
Description
Technical field
The present invention relates to multi-media sensor covering control and two fields of intelligence computation, and in particular to one kind being based on difference
The video sensor network coverage enhancement implementation method of evolution.
Background technology
The communication technology, the rapid development of embedded technology and sensor technology and increasingly mature, have pushed modern wireless biography
The generation and development of sensor network.It is increasingly sophisticated changeable with monitoring of environmental, by simple acquired in traditional sensors network
Data cannot meet comprehensive demand of the people to environmental monitoring further, and there is an urgent need to by matchmakers such as the audio of informative, videos
Body is introduced into the environmental monitoring activity based on sensor network, and multimedia sensor network comes into being as a result,.Depending on
The one kind of video sensor network as multimedia sensor network, by the surrounding enviroment of built in video Sensor monitoring place
Image, video information, typical video sensor network is usually by structures such as video sensor node, aggregation node, control centres
At.Video sensor node is at pleasure deployed in specified monitoring region, and the image of acquisition, video information are passed along other videos
Sensor node hop-by-hop is transmitted to aggregation node, and control centre is reached finally by Internet network or telecommunication satellite.User is logical
It crosses control centre video sensor network is configured and managed, issue monitoring task and collects monitoring data.
Covering control is to ensure a critical issue of video sensor network target detection performance, it is intended to optimize network sky
Between resource preferably to complete environment sensing, acquisition of information be the basis that entire Detection task is able to continue.Video passes
The perception of sensor nodes is limited by directionality, and sensing range is one using node as the center of circle, and radius is its perceived distance
Sector region, sensor network disposition of the tradition based on omnidirectional's sensor model and coverage optimization method have been no longer desirable for having
The video sensor node of directional perception characteristic.Completely new oriented perception mould is established according to the perception characteristics of video node
Type.According to directional sensing model, the efficient video sensor network dispositions method of reasonable design and node control strategy are studied simultaneous
The network coverage algorithm for caring for Energy Efficient and reliably monitoring is an important directions in current multi-media network research field.Usually
Node is disposed using extensive random putting mode, but this deployment way is difficult disposably to sense numerous video
The unreasonable of monitoring region overlay is placed in place and caused to device node, leads to the target monitoring ability that network provides not
Uniformly.Therefore, it after video sensor network initial deployment, needs to take effective measures with the uniformly disorderly and unsystematic video of script
The distribution of sensor node so that network is with larger probability covering monitoring region.The covering for studying video sensor network increases
Strong strategy improves monitoring quality and is of crucial importance for the Monitoring Performance of raising whole network.
Domestic and foreign scholars have carried out the research of sensor network coverage enhancement problem in succession, and achieve certain work into
Exhibition, but most coverage enhancement Study on Problems is both for the sensor network expansion based on omnidirectional's sensor model, usually
Traditional sensors network covering property is realized using conventional methods such as the position distributions or addition new node for readjusting node
Enhancing, such method essence are to be moved using sensor node position to enhance the areal coverage of sensor network.Actually answer
With under environment, it is contemplated that the video sensor node of video sensor network lower deployment cost, all deployment all has locomotivity
It is unpractical, in addition to this, the locomotivity of sensor node leads to the dynamic change of network topology, it is easy to cause part
The failure of sensor node, so that the maintenance cost of network increases, network lifetime shortens.
Invention content
In view of this, the object of the present invention is to provide a kind of, the video sensor network coverage enhancement based on differential evolution is real
Existing method, this method assume that deployed position no longer changes all the sensors node, and from the sense of video sensor node
Know that characteristic is set out, make full use of the perceived direction and direction adjustability of node, reaches to eliminate and feel in video sensor network
Know the purpose of overlay region and blind area.The present invention is to monitor areal coverage as object function, with video sensor node perceived side
It is decision variable to angle, the global optimization ability corrdinated adjustment video node perceived direction of differential evolution algorithm is used to be covered to reach
Cover the purpose of enhancing.
In order to realize above-mentioned method, the present invention provides a kind of, and the video sensor network covering based on differential evolution increases
Strong implementation method, it is characterised in that:Including following three basic procedures:
(1) first according to monitoring region scene actual environment, by continuous two dimensional surface discretization, i.e., with series of parallel
Plane is divided into uniform grid in the straight line of x, y-axis, straight parallel wire spacing is known as granularity, the meter of the smaller coverage rate of granularity
Calculation precision is higher, and vice versa, and the coordinate information of grid element center point is uploaded to control centre;
(2) monitoring region in a certain number of video sensor nodes of random placement, video node be it is fixed, once
It disposes its coordinate position no longer to change, video node obtains the position coordinates of itself by GPS or location algorithm, then each section
The coordinate information of itself, perceived direction are sent to aggregation node by point by multi-hop relay routing, and aggregation node, which summarizes, receives letter
It ceases and is transmitted to control centre by external networks such as Internet;
(3) control centre summarizes the information of upload, and using each video node perceived direction angle as decision variable, monitoring region is covered
Lid rate is object function, by differential evolution algorithm variation, is intersected, three operation Step wise approximation optimal node perceived directions of selection
To obtain the maximization of coverage rate, control centre's publisher node direction control command to video sensor network, aggregation node will
Each node direction control instruction is sent to video node and adjusts its perceived direction, reduces network aware overlay region and perceives blind
Area realizes coverage enhancement.
Typical video sensor network is usually made of video sensor node, aggregation node, control centre, and video passes
Sensor node is integrated with the micro embedded node of video sensor, data processing unit, communication module, is passed by built in video
Image, video information in surrounding enviroment where sensor monitoring, and simple process is carried out, while being also responsible for forwarding other nodes
Data fusion forwarding;Aggregation node can both be regarded as having more powerful video sensor node, can also be not supervise
Survey ability, the only gateway node of communication capacity, aggregation node are responsible for connecting the outside such as video sensor network and Internet
Network, the monitoring task of release management video sensor node, and it forward the data to external network;Control centre is then responsible for inquiry
Or the monitoring information of video sensor network is collected, it can also release news to video sensor network.Under being suitable for of the present invention
The application scenarios stated:The perception radius of all video sensor nodes in network, sensitive zones visual angle is all identical and perceives
Direction is adjustable;All nodes no longer change once deployed position in network, while the perception radius and sensitive zones visual angle are also protected
Hold constant, only perceived direction is adjustable.
In practical application, monitoring region is a continuous two dimensional surface, in order to which the calculating for simplifying areal coverage is usual
Sliding-model control is done to the two dimensional surface, i.e., monitoring region is divided into several uniform grids, these grids and grid with
Monitor region boundary composition monitoring subregion, when monitor sub-zone dividing enough to it is small when, we were both believed that in subregion
The coverage rate of heart point is exactly the coverage rate of the subregion.Specific discrete operations are as follows:It will with the series of parallel straight line in x, y-axis
Monitoring region Ω is divided into K uniform grids, and straight parallel wire spacing Δ d is known as granularity, then the specification of grid is Δ d × Δ
The central point of the rectangle of d, grid is known as pixel, and the set of pixel is denoted as | | ΛK| |, monitor areal coverage calculating
When, just with | | ΛK| | instead of continuous level Ω, the precision of the smaller calculating of granularity Δ d is higher, and vice versa.It simultaneously will monitoring
The information in region returns to control centre, i.e., by the coordinate information q of each pixelj(xj, yj), j ∈ { 1,2 ..., K } are sent to control
Center processed.
After obtaining monitoring area information, N number of video sensor node, the number of node deployment are dispensed at random in monitoring region
Mesh N and monitoring region Ω, the perception radius r, sensor offset angleRelationship between areal coverage p is:It can refer to the interstitial content N disposed needed for inequality estimation.The video sensor node disposed
For oriented sensor node, the directional perception ability of oriented sensing node can be abstracted as to a sector, fan-shaped center of circle S is
Node, fan-shaped radius r are the perception radius of node, and fan-shaped central axes are the perceived direction of node, between central axes and x-axis
Angle theta be video node perceived direction angle, and θ ∈ [0,2 π), the i.e. available unit vector of perceived directionIt indicates, the angle of central axes and sector boundaryFor the sensor offset angle of node, the sensing unit of video node
Domain visual angle isAfter video sensor network initial deployment, each sensor obtains itself by GPS or location algorithm
Position coordinates Si(xi, yi), node SiPerceived direction perceived direction angle θiIt indicates, θiBetween perception unit vector and x-axis
Angle, the set of all nodesSet of nodeCorresponding perceived direction collection is combined into Θ={ θ1, θ2...,
θN, each video sensor node is by the position coordinates of itself, perceived direction angle (xi, yi, θi) be sent to by multihop routing
Aggregation node, aggregation node forward the data to control centre by external networks such as Internet again.
Control centre summarizes the information of upload, including monitoring region pixel collection | | ΛK| | in each pixel seat
Mark information qj(xj, yj), the coordinate S of j ∈ { 1,2 ..., K } and N number of nodei(xi, yi), i ∈ { 1,2 ..., N } and perception side
To angle set Θ={ θ1, θ2..., θN, differential evolution algorithm is using each video node perceived direction angle Θ as decision variable, monitoring
Areal coverage is object function, pixel number/pixel sum K that monitoring areal coverage p=sets of node are covered.Pixel
Point qj(xj, yj) by node Si(xi, yi) covering condition be:Pixel qjWith node SiBetween Euclidean distance d (Si, qj)
In the perception radius of node and pixel qjIn sensitive zones angular field of view in video node, if pixel qjIt is arbitrary
One coverage is considered as the pixel and is covered by set of node, pixel qjThe probability P covered by set of nodeCOV(j) it is:Monitor areal coverageTherefore it regards
Video sensor network coverage enhancing can be regarded as single-object problem, and object function to be optimized is to monitor areal coverage,
Decision variable is the perceived direction angle of set of node, then the present invention solves the problem using differential evolution optimization algorithm.
Differential evolution algorithm is a kind of bionic intelligence computational methods based on Evolution of Population, and algorithm has memory individual optimal
The feature that solution and population internal information are shared, seeks optimization problem to realize by the cooperation and competition between individual in population
Solution.Differential evolution algorithm includes variation, intersects, selects three basic operations, in calculating iteration, passes through variation and crossover operation
Generate new candidate individual, selection operation be then by being at war with one to one between parent individuality and newly generated individual, it is excellent
Win bad eliminate so that offspring individual is always better than or is equal to parent individuality, to make population evolve always to the direction of optimal solution.
The concrete operations of differential evolution algorithm are as follows:Population scale is denoted as NP, and decision vector space dimensionality is D, with X (t)
Indicate to evolve to t for when population, each dimension component for randomly generating i-th of initial population individual individual in the decision vector space of problem first can generate as the following formula:xJ, max、xJ, minThe respectively bound of decision space jth dimension;In calculating iteration, calculate
Method randomly choose two different individual vectors subtract each other generation difference vector, then by difference vector assign weights after with it is another with
The addition of vectors that machine is selected to generate variation individual be from parent population with
The individual vector for the inequality that machine is chosen, F is a constant between [0,2], the shadow for controlling difference vector
It rings;Variation individual and target individual carry out dimension and mixs to intersect, obtain intersection individual intersection to
Each element is measured by following formulaIt generates, randb is the random number between [0,1], is intersected
Constant CR is constant of the range in [0,1], randjIt is in [1, D] randomly selected integer, it ensures that intersecting individual at least wants
An element is obtained from variation individual avoids the evolution of population from stagnating to ensure to have new individual to generate;Differential evolution algorithm
Selection operation is a kind of greedy selection mode, and the target function value and if only if new vector individual is individual better than object vector
When, it can just be received by population, otherwise will be retained in follow-on population, it is assumed that problem to be optimized is then
Selection operation can be described as,So that offspring individual is always better than or is equal to parent individuality, to make
Population is always as the direction of optimal solution is evolved.
For video sensor network coverage enhancement optimization problem, differential evolution algorithm is with video node Ji Ganzhifangxiangjiao
As decision variable, since the dimension D of the vector that decision variable is made of the perceived direction angle of N number of node, decision space is
It generates the individual that NP dimension is D in initialization of population for the number N of node and forms initial population X (0), due to perception side
To angle [0,2 π) in range, then xJ, max、xj,min2 π, 0 are taken respectively;Object function is monitoring areal coverage p, which is
Maximum value optimization problem, in order to differential evolution algorithm problem to be optimizedIt is consistent, enablesAlgorithm is pre-
First set greatest iteration cycle-index G max, then stop iteration when iterations t reaches G max, otherwise repeat variation,
Intersect, three operations of selection, generates population at individual of new generation and population is evolved to the direction of optimal solution always;Work as iteration ends
When, algorithm exports globally optimal solutionIt can be according to formulaObtain global optimum coverage rate p*, simultaneouslyInstitute is right
The decision vector answeredFor the optimal perceived direction angle of video node collection.The optimal perceived direction control of control centre's publisher node refers to
Node direction control instruction is sent to each video node, video node by order to video sensor network, then aggregation node
The perceived direction of itself, which is adjusted, according to the instruction received realizes video sensor network coverage enhancement to reduce network blind area of coverage.
The present invention is a kind of video sensor network coverage enhancement implementation method based on differential evolution, all video sensings
Its deployed position of device node one no longer changes, herein under the premise of, pass through differential evolution algorithm adjust video sensor node sense
Deflection is known with optimization monitoring areal coverage, and the present invention has the following advantages:In terms of video sensor network life cycle angle,
The body operation of the present invention is all responsible for by control centre, and control centre is calculated by Different iterative and generates optimal perceived direction angle,
In the process, each video node need not execute any instruction and operation, be in suspend mode low power consumpting state, this is effectively reduced
The energy expenditure of node simultaneously extends Network morals;From the point of view of video sensor coverage enhancement implementation method, we
The complexity of case is low, takes full advantage of oriented sensor node sensor model and is only covered with region with differential evolution intelligent algorithm
The evolution of lid rate guiding solution is continued to optimize by self study in algorithm independent of the strict mathematical property of optimization problem itself
The target function value of individual.
Description of the drawings
Fig. 1 is video sensor network coverage enhancement implementation method information flow block diagram of the present invention
Fig. 2 is video sensor network structural schematic diagram
Fig. 3 is monitoring region discretization schematic diagram
Fig. 4 is video sensor node direction sensor model
Fig. 5 is the program structure process of differential evolution algorithm
Fig. 6 (a), (b), (c) are respectively to affix one's name to 110 sensitive zones visual angles inside 500 × 500 monitoring regions to beVideo
The initial random deployment coverage diagram of sensor node, after coverage enhancement optimizes network Landfill covering figure and be based on differential evolution
The network coverage iterativecurve figure of algorithm
Specific implementation mode
In order to make the statement of technical scheme of the present invention, implementing procedure be more clear, below in conjunction with attached drawing to being based on difference
The specific implementation step for the video sensor coverage enhancement implementation method evolved is divided to be described in further detail:
Referring to Fig.1, the block diagram of video sensor network coverage enhancement implementation method is described in the form of information flow, it is first
First according to monitoring region actual environment, by continuous two dimensional surface discretization, and coordinate information is uploaded to control centre;So
Afterwards, coordinate information, perceived direction are sent to aggregation node by a certain number of video nodes of random placement, each node;Convergence section
Point summarizes forwarding information to control centre, and using video node perceived direction angle as decision variable, monitoring areal coverage is target
Function, by three differential evolution algorithm variation, intersection, selection operation Step wise approximation optimal node perceived directions to be covered
The maximization of rate, and node direction control instruction is issued to video sensor network to adjust node perceived direction.
With reference to Fig. 2, the structure composition of video sensor network is described, video sensor network is usually by video sensor
The compositions such as node, aggregation node, control centre.Video sensor node is at pleasure deployed in specified monitoring region, acquisition
Image, video information are transmitted to aggregation node along other video sensor node hop-by-hops, finally by Internet network or
Telecommunication satellite reaches control centre.User is configured and is managed to video sensor network by control centre, publication monitoring
Task and collection monitoring data.
With reference to Fig. 3, in practical application, monitoring region is a continuous two dimensional surface, in order to simplify areal coverage
Calculating usually does sliding-model control to the two dimensional surface, i.e., monitoring region is divided into several uniform grids, these grids with
And grid with monitoring region boundary composition monitoring subregion, when monitor sub-zone dividing enough to it is small when, we were both believed that
The coverage rate of subregion central point is exactly the coverage rate of the subregion.Specific discrete operations are as follows:With series of parallel in x, y-axis
Straight line be divided into K uniform grids by region Ω is monitored, straight parallel wire spacing Δ d is known as granularity, then the specification of grid
Central point for the rectangle of Δ d × Δ d, grid is known as pixel, and the set of pixel is denoted as | | ΛK| |, calculating monitoring section
When domain coverage ratio, just with | | ΛK| | instead of continuous level Ω, the precision of the smaller calculating of granularity Δ d is higher, and vice versa.Together
When will monitor region information return control centre, i.e., by the coordinate information q of each pixelj(xj, yj), j ∈ { 1,2 ..., K }
It is sent to control centre.
After obtaining monitoring area information, N number of video sensor node, video sensor net are dispensed at random in monitoring region
After network initial deployment, each sensor obtains the position coordinates S of itself by GPS or location algorithmi(xi, yi).With reference to figure
4, the directional perception characteristic of video sensor node is described, the sensing region of video sensor node S is to be with node coordinate
The center of circle, sensing range r are the sector of radius, and the angle of the sector is that the sensitive zones visual angle of video sensor node is
The perceived direction angle θ of video sensor node is the angle of fan-shaped central axes and positive direction of the x-axis, since video sensor node is
Random placement, so perceived direction is also arbitrary, therefore θ ∈ [0,2 π), perceived direction unit vector can be expressed asNode SiPerceived direction perceived direction angle θiIt indicates, the set of all nodesSection
Point setCorresponding perceived direction collection is combined into Θ={ θ1, θ2..., θN}.Pixel qj(xj, yj) by node Si(xi, yi) covering
Condition is:Pixel qjWith node SiBetween Euclidean distance d (Si, qj) in the perception radius of node and pixel qjIn regarding
In the sensitive zones angular field of view of frequency node, if pixel qjThe pixel is considered as by node by any one coverage
Collection covering, pixel qjBy set of nodeThe probability P of coveringCOV(j) it is:Therefore, monitoring areal coverage can be expressed asVideo sensor network coverage enhancement can be regarded as single-object problem, and object function to be optimized is to supervise
Areal coverage is surveyed, decision variable is the perceived direction angle of set of node, and the problem is solved using differential evolution optimization algorithm.
With reference to Fig. 5, the program structure of differential evolution algorithm is described in a flowchart, to solve object function
Minimum problems for, specific implementation steps are as follows:
Step1 initiation parameters enable iterations t=0, setting greatest iteration cycle-index Gmax, population scale NP, contracting
It puts factor F, intersect constant CR;
Step2 generates initial population at random in problem search spaceAnd calculate each individual
Target function value;
Step3 cycle-indexes t ← t+1;
Step4 enables the call number i=1 of target individual;
Step5 randomly chooses the different individual of the other three except target individualIt is performed simultaneously variation behaviour
Make, generates variation individual
Step6 is to target individualAnd variation individualCrossover operation is executed, generates and intersects individual
What Step7 was calculatedTarget function value retains more preferably individual according to " greediness " selection operation;
Step8 target individual call number i ← i+1 return to Step5 until i=NP (is carried out all individuals in population
Variation intersects, selection operation);Otherwise, Step9 is executed;
If Step9 meets termination condition, i.e., if iterations t >=Gmax, recycles end and export result of calculation, it is no
It then jumps to Step3 and continues next iteration.
For video sensor network coverage enhancement optimization problem, differential evolution algorithm is with video node Ji Ganzhifangxiangjiao
As decision variable, since the dimension D of the vector that decision variable is made of the perceived direction angle of N number of node, decision space is
It generates the individual that NP dimension is D in initialization of population for the number N of node and forms initial population X (0), due to perception side
To angle [0,2 π) in range, then xJ, max、xJ, min2 π, 0 are taken respectively;Object function is monitoring areal coverage p, which is
Maximum value optimization problem, in order to differential evolution algorithm problem to be optimizedIt is consistent, enablesAlgorithm is pre-
First set greatest iteration cycle-index G max, then stop iteration when iterations t reaches G max, otherwise repeat variation,
Intersect, three operations of selection, generates population at individual of new generation and population is evolved to the direction of optimal solution always;Work as iteration ends
When, algorithm exports globally optimal solutionIt can be according to formulaObtain global optimum coverage rate p*, simultaneouslyInstitute is right
The decision vector answeredFor the optimal perceived direction angle of video node collection.
The optimal perceived direction control instruction of control centre's publisher node to video sensor network, then aggregation node will save
Point direction control command is sent to each video node, and video node adjusts the perceived direction of itself to subtract according to the instruction received
Realize video sensor network coverage enhancement in few network blind area of coverage.
Finally, it with reference to Fig. 6 (a), (b), (c), specifically describes a video sensor network based on differential evolution and covers
Cover the simulation example of enhancing.110 sensing radius r=60m of random placement, sensitive zones in 500 × 500 monitoring region
Visual angleVideo sensor node complete the Detection task to target area.Each ginseng of differential evolution optimization algorithm
Number setting is as follows:NP=40, F=0.6, CR=0.1, Gmax=200.Emulation uses Matlab emulation platforms, figure (a) to have recorded
Initial random deployment coverage diagram, initial time, monitoring areal coverage is only 0.6570, after differential evolution algorithm optimizes,
After 200 iteration, the coverage rate of network is promoted to 0.7972, after coverage enhancement optimizes network Landfill covering figure referring to figure (b),
Network coverage iterativecurve figure based on differential evolution algorithm can be found in figure (c).Therefore, optimize by differential evolution algorithm
Afterwards, the perception overlapping region between multiple adjacent video sensor nodes significantly reduces, and effectively enhances entire video sensor
The covering performance of network.
Claims (1)
1. a kind of video sensor network coverage enhancement implementation method based on differential evolution, it is characterised in that:Including following three
A basic procedure:
(1) first according to monitoring region scene actual environment, by continuous two dimensional surface discretization, i.e., with it is series of parallel in x,
Plane is divided into uniform grid by the straight line of y-axis, and straight parallel wire spacing is known as granularity, the calculating essence of the smaller coverage rate of granularity
Degree is higher, and vice versa, and the coordinate information of grid element center point is uploaded to control centre;
(2) a certain number of video sensor nodes of random placement in monitoring region, video node is fixed, and one is deployed
Its coordinate position no longer changes, and video node obtains the position coordinates of itself by GPS or location algorithm, and then each node will
The coordinate information of itself, perceived direction are sent to aggregation node by multi-hop relay routing, and aggregation node, which summarizes, receives information simultaneously
It is transmitted to control centre by Internet external networks;
(3) control centre summarizes the information of upload, using each video node perceived direction angle as decision variable, monitors areal coverage
For object function, Step wise approximation optimal node perceived directions are operated to obtain by differential evolution algorithm variation, intersection, selection three
The maximization of coverage rate is obtained, control centre's publisher node direction control command to video sensor network, aggregation node will be each
Node direction control instruction is sent to video node and adjusts its perceived direction, reduces network aware overlay region and perception blind area, real
Existing coverage enhancement;The video sensor network is made of video sensor node, aggregation node, control centre, depending on
Video sensor node is dispersed in specified monitoring region, and the data of acquisition are transmitted along other video sensor node hop-by-hops
To aggregation node, control centre is reached finally by Internet network or telecommunication satellite, user is by control centre to video
Sensor network is configured and is managed, and is issued monitoring task and is collected monitoring data;The video sensor node is
The directional perception ability of oriented sensing node, can be abstracted as a sector by oriented sensor node, and fan-shaped center of circle S is section
Point, fan-shaped radius r are the perception radius of node, and fan-shaped central axes are the perceived direction of node, between central axes and x-axis
Angle theta is the perceived direction angle of video node, and θ ∈ [0,2 π), the i.e. available unit vector of perceived direction
It indicates, the angle of central axes and sector boundarySensitive zones visual angle for the sensor offset angle of node, video node is
The flow (1) further comprises following operation content:Region Ω will be monitored with the series of parallel straight line in x, y-axis
K uniform grids are divided into, straight parallel wire spacing Δ d is known as granularity, then the specification of grid is the rectangle of Δ d × Δ d, net
Center of a lattice point is known as pixel, and the set of pixel is denoted as | | ΛK| |, calculate monitor areal coverage when, just with | | ΛK|
| instead of continuous level Ω, by the coordinate information q of each pixelj(xj, yj), j ∈ { 1,2 ..., K } are sent to control centre;
The flow (2) further comprises following operation content:
The N number of video sensor node of random placement in monitoring region, each sensor obtain itself by GPS or location algorithm
Position coordinates Si(xi, yi), node SiPerceived direction perceived direction angle θiIt indicates, θiBetween perception unit vector and x-axis
Angle, the set of all nodesSet of nodeCorresponding perceived direction collection is combined into Θ={ θ1,
θ2..., θN, each video sensor node is by the position coordinates of itself, perceived direction angle (xi, yi, θi) sent out by multihop routing
Aggregation node is given, aggregation node forwards the data to control centre by Internet external networks again;
The flow (3) further comprises following operation content:
(31) control centre summarizes the information of upload, including monitoring region pixel collection | | ΛK| | in each pixel coordinate
Information qj(xj, yj), the coordinate S of j ∈ { 1,2 ..., K } and N number of nodei(xi, yi), i ∈ { 1,2 ..., N } and perceived direction
Angle set Θ={ θ1, θ2..., θN, differential evolution algorithm is using each video node perceived direction angle Θ as decision variable, monitoring section
Domain coverage ratio is object function, pixel number/pixel sum K that monitoring areal coverage p=sets of node are covered;
(32) pixel qj(xj, yj) by node Si(xi, yi) covering condition be:Pixel qjWith node SiBetween it is European away from
From d (Si, qj) in the perception radius of node and pixel qjIn sensitive zones angular field of view in video node, if picture
Vegetarian refreshments qjThe pixel is considered as by any one coverage to be covered by set of node, pixel qjBy set of nodeWhat is covered is general
Rate PCOV(j) it is:Monitor areal coverageTherefore, video sensor network coverage enhancement can be regarded as single-object problem, and object function is to supervise
Areal coverage is surveyed, decision variable is the perceived direction angle of set of node, and the present invention solves this using differential evolution optimization algorithm and asks
Topic, by the variation of differential evolution algorithm, intersection, the optimal node perceived direction of three basic operation Step wise approximations of selection, to obtain
Obtain the maximization of coverage rate;
(33) video sensor network coverage enhancement problem is made with video node Ji Ganzhifangxiangjiao according to operation (32)
For decision variable, since the dimension D of the vector that decision variable is made of the perceived direction angle of N number of node, decision space is
The number N of node generates the individual that NP dimension is D and forms initial population X (0), due to perceived direction in initialization of population
Angle [0,2 π) in range, then the bound of decision vector takes 2 π, 0 respectively;Object function is monitoring areal coverage p, this is asked
Entitled maximum value optimization problem, in order to differential evolution algorithm problem to be optimizedIt is consistent, enables
Algorithm presets greatest iteration cycle-index Gmax, then stops iteration when iterations t reaches Gmax, otherwise repeats change
Three different, intersection, selection operations generate population at individual of new generation and make population always to the evolution of the direction of optimal solution;
(34) when iteration ends, algorithm exports globally optimal solutionIt can be according to formulaGlobal optimum is obtained to cover
Lid rate p*, simultaneouslyCorresponding decision vectorFor the optimal perceived direction angle of video node collection, control centre's publication section
Node direction control instruction is sent to each video section by point direction control command to video sensor network, then aggregation node
Point, video node adjust the perceived direction of itself to reduce network blind area of coverage according to the instruction received.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310335322.7A CN104349356B (en) | 2013-08-05 | 2013-08-05 | Video sensor network coverage enhancement implementation method based on differential evolution |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310335322.7A CN104349356B (en) | 2013-08-05 | 2013-08-05 | Video sensor network coverage enhancement implementation method based on differential evolution |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104349356A CN104349356A (en) | 2015-02-11 |
CN104349356B true CN104349356B (en) | 2018-10-30 |
Family
ID=52503965
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310335322.7A Active CN104349356B (en) | 2013-08-05 | 2013-08-05 | Video sensor network coverage enhancement implementation method based on differential evolution |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104349356B (en) |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104796915B (en) * | 2015-05-08 | 2018-11-09 | 北京科技大学 | A kind of optimization method of two dimension aeoplotropism sensor network covering |
CN105959912B (en) * | 2016-02-23 | 2019-03-15 | 重庆工商大学 | Based on the aggregation node localization method for improving discrete differential algorithm |
CN109600710B (en) * | 2018-12-10 | 2020-10-30 | 浙江工业大学 | Multi-moving-target monitoring method based on difference algorithm in video sensor network |
CN109451529A (en) * | 2019-01-02 | 2019-03-08 | 成都华日通讯技术有限公司 | A kind of miniature spectral sensor net system and its synergistic data processing method |
US20240072893A1 (en) * | 2019-10-09 | 2024-02-29 | Signify Holding B.V. | Optical wireless communication system and device |
CN111132182B (en) * | 2019-12-16 | 2022-08-23 | 北京农业信息技术研究中心 | Method and system for enhancing wireless multimedia network coverage |
CN111698656B (en) * | 2020-05-21 | 2023-02-07 | 江苏海洋大学 | Multi-target dynamic scheduling method for underwater mobile wireless sensor network |
CN112288167A (en) * | 2020-10-30 | 2021-01-29 | 南京农业大学 | Position-limited solar insecticidal lamp node deployment method |
CN114329855B (en) * | 2020-11-06 | 2023-05-12 | 北京航空航天大学 | Sensor layout optimization and rapid deployment method of wireless vision sensing network |
CN113784361B (en) * | 2021-05-13 | 2022-07-19 | 中国地质大学(武汉) | Dynamic adjustment method and device for sensor node and storage medium |
CN113453183B (en) * | 2021-05-31 | 2023-03-21 | 西南电子技术研究所(中国电子科技集团公司第十研究所) | Remote perception monitoring global target space coverage optimization method |
CN117062092B (en) * | 2023-10-11 | 2023-12-08 | 成都大学 | Wireless sensor network deployment method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6577906B1 (en) * | 1999-08-05 | 2003-06-10 | Sandia Corporation | Distributed optimization system and method |
CN101009821A (en) * | 2007-01-31 | 2007-08-01 | 北京邮电大学 | Implementing method for the enhanced cost-effective video sensor network coverage |
CN102014398A (en) * | 2010-09-21 | 2011-04-13 | 上海大学 | Optimal deployment method of large-scale industrial wireless sensor network based on differential evolution algorithm |
CN102984731A (en) * | 2012-12-06 | 2013-03-20 | 重庆工商大学 | Adjustment method of heterogeneous wireless sensor network nodes based on multiple covering |
-
2013
- 2013-08-05 CN CN201310335322.7A patent/CN104349356B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6577906B1 (en) * | 1999-08-05 | 2003-06-10 | Sandia Corporation | Distributed optimization system and method |
CN101009821A (en) * | 2007-01-31 | 2007-08-01 | 北京邮电大学 | Implementing method for the enhanced cost-effective video sensor network coverage |
CN102014398A (en) * | 2010-09-21 | 2011-04-13 | 上海大学 | Optimal deployment method of large-scale industrial wireless sensor network based on differential evolution algorithm |
CN102984731A (en) * | 2012-12-06 | 2013-03-20 | 重庆工商大学 | Adjustment method of heterogeneous wireless sensor network nodes based on multiple covering |
Non-Patent Citations (3)
Title |
---|
D-S theory based on an improved PSO for data fusion;ZHU Peiyi,等;《JOURNAL OF NETWORKS》;20120228;全文 * |
传感器网络中节点感应模型与相关工作的比较;秦宁宁,等;《计算机应用研究》;20100930;全文 * |
基于差分算法的异构移动节点分布优化策略;李明,等;《华中科技大学学报》;20110430;第2.1 * |
Also Published As
Publication number | Publication date |
---|---|
CN104349356A (en) | 2015-02-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104349356B (en) | Video sensor network coverage enhancement implementation method based on differential evolution | |
Shahzad et al. | Multi-objective optimization for a reliable localization scheme in wireless sensor networks | |
CN105531963B (en) | The system and method that multiple target cell is closed in wireless network | |
Dou et al. | Optimizing sensor network coverage and regional connectivity in industrial IoT systems | |
CN106792750B (en) | A kind of node deployment method based on guiding particle swarm algorithm | |
Ahmadi et al. | An efficient routing algorithm to preserve k-coverage in wireless sensor networks | |
CN103281704B (en) | Based on the wireless sensor network certainty spatial portion arranging method of three-dimensional perception | |
CN105208569B (en) | Directional sensor network covering method under complex three-dimensional scene environment | |
Udgata et al. | Sensor deployment in irregular terrain using artificial bee colony algorithm | |
Zhou et al. | Sensor deployment scheme based on social spider optimization algorithm for wireless sensor networks | |
Yu et al. | Connectivity-constrained placement of wireless chargers | |
CN106162663B (en) | A kind of sensing node covering method based on improvement ant colony algorithm | |
Priyadarshi et al. | Area coverage optimization in three-dimensional wireless sensor network | |
CN105898765A (en) | Region coverage optimization method suitable for video sensor network with adjustable sensing direction | |
CN112637860B (en) | Three-dimensional wireless sensor network coverage method and system | |
Xia | Retracted: Coverage Optimization Strategy of Wireless Sensor Network Based on Swarm Intelligence Algorithm | |
Xia et al. | AI-driven and MEC-empowered confident information coverage hole recovery in 6G-enabled IoT | |
Temesgene et al. | Distributed deep reinforcement learning for functional split control in energy harvesting virtualized small cells | |
CN106209261B (en) | The mobile data collection method of three-dimensional UASNs based on probability neighborhood grid | |
CN108768472A (en) | A kind of near space High Altitude Platform space-polarity diversity mimo channel modeling method | |
Abdulrab et al. | Hybrid harris hawks with sine cosine for optimal node placement and congestion reduction in an industrial wireless mesh network | |
Navin et al. | Solving coverage problem in wireless camera-based sensor networks by using genetic algorithm | |
Park et al. | A grid-based self-deployment schemes in mobile sensor networks | |
Fu et al. | Research on the three-dimensional perception model and coverage-enhancing algorithm for wireless multimedia sensor networks | |
Xu et al. | Optimizing coverage of 3D Wireless Multimedia Sensor Networks by means of deploying redundant sensors |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |