research article sensor networks hierarchical...
Post on 05-Jul-2020
0 Views
Preview:
TRANSCRIPT
Research ArticleSensor Networks Hierarchical Optimization Model for SecurityMonitoring in High-Speed Railway Transport Hub
Zhengyu Xie1,2 and Yong Qin2
1School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China2State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing 100044, China
Correspondence should be addressed to Yong Qin; yqin@bjtu.edu.cn
Received 21 November 2014; Revised 31 March 2015; Accepted 7 April 2015
Academic Editor: Fei Yu
Copyright © 2015 Z. Xie and Y. Qin. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
We consider the sensor networks hierarchical optimization problem in high-speed railway transport hub (HRTH). The sensornetworks are optimized from three hierarchies which are key area sensors optimization, passenger line sensors optimization, andwhole area sensors optimization. Case study on a specific HRTH in China showed that the hierarchical optimization method iseffective to optimize the sensor networks for security monitoring in HRTH.
1. Introduction
With the rapid development of high-speed railway in China,many modern HRTHs have been built to match the develop-ing demands. HRTHs become the crossing and interface ofmultitransportation which include high-speed railway, civilaviation, highway, waterway, urban rail transit, public trans-port, motor vehicle, and taxi. As a vital node of passengertransport net, HRTH is an important distribution place ofmassive passenger flow. With the increase of high-speedrailway operation mileage, the distribution quantity of pas-sengers will be sustained to increase sharply, which leadsHRTHs to confront severe challenges in passenger flow secu-rity monitoring.
At present, the video surveillance system is the mainsecurity monitoring approach used in HRTH. The managerscan detect the congestion of passenger flow, abnormal behav-iors of passengers, abandoned objects, and so forth by usingsurveillance systems. The basic workflow of system includesthe following: (i) data acquisition: distribute surveillancesensors and develop a sensor network; (ii) data transmission:choose suitable approaches to transmit data acquired fromthe sensor networks; (iii) data processing: utilize efficientimage processing method to process the data acquired fromthe sensor networks and obtain processing result based onthe demands of security monitoring; (iv) data dissemination:
select various channels to disseminate the security moni-toring information. Currently, the studies related to videosurveillance system in HRTHmainly focused on (ii) and (iii)to improve detection accuracy and speed; specific study on (i)is scarce. As a foundation of other parts, the sensor networkshave important influences on other parts. So it is necessary forHRTH security monitoring to optimize the sensor networks.
The rest of this paper is organized as follows:The relevantliterature is reviewed in the next section.The sensor networkshierarchical optimization problem is described in Section 3and Section 4 proposes a sensor networks hierarchical opti-mization model. A case study is reported in Section 5 andfinally Section 6 covers the conclusion.
2. Literature Review
The sensor networks optimization problem for security mon-itoring in HRTH belongs to the art gallery problem (AGP)which was first proposed in 1973 in a conversation betweenKlee and Chvatal [1]. Based on the conversation, Chvatalproofed [𝑛/3] cameras are always sufficient and sometimesnecessary. This conclusion is called the Art Gallery The-orem, or Watchman Theorem [2]. Fisk used triangulationtechniques and staining methods and got the conclusion“any simple polygon after triangulation, the correspondingdiagram can 3- stain,” and the same type of colored dots
Hindawi Publishing CorporationJournal of SensorsVolume 2015, Article ID 951242, 9 pageshttp://dx.doi.org/10.1155/2015/951242
2 Journal of Sensors
Table 1: Data acquisition demands of security monitoring in HRTH.
Level Description Concern Data acquisition demands
First level Key area monitoringFocus on the security ofkey, important, andsensitive areas
The data of key areas in HRTHmust becontinuously acquired and can meet theanomaly detection of key areas
Second level Passenger linemonitoring
Focus on the security ofpassenger input and outputlines
The data of entire passenger line in HRTHmust be continuously acquired and can meetthe forecast demands of post node inpassenger line
Third level Complete coveragemonitoring
Focus on the security ofwhole HRTH
The data of whole HRTH can beinconsecutively and optionally acquired andmust ensure all function areas arecompletely covered
can cover the entire simple polygon [3]. Avis, Toussaint, andChazelle gave different algorithms for the simple polygontriangulation. For any simple polygon with given point, wecan determine the location of monitors in a simple polygonwithin time, making any point in this simple polygon ableto see at least one monitor [4, 5]. Lee and Lin proved thatthe algorithm of solving any simple polygon which requiredminimum number of guards is NP-hard [6].
After Art Gallery Theorem is proved, more and morequestions of AGP are proposed, including the following: themonitor can bemoved at the edge, the monitor can bemovedat the diagonal, at least two monitors are required that canbe guarded by each other, one guard is removed while theother guards could know, and the walls of the gallery shouldbe vertical [7, 8].
In computational geometry, the gallery can be abstractsimple polygon; put a monitor abstraction for a point insimple polygon, and then the problem can be abstracted as anart gallery plane geometry problem; gallery guards problemcan be abstracted as how many points can cover the entiresimple polygon. Variant problem can be abstracted as jointguards, side cover, diagonal coverage, coguards, orthogonalgallery guards, moving guard, limited perspective guards,moving guard with limited perspective, orthogonal polygonsmobile guards, and other issues [9, 10].
For unrealistic assumptions of monitor in the solvingof AGP and its variant problem, such as magnifying themonitoring range of single monitor, expanding the depth offield, and not limiting the recognition accuracy and speed,lead to research in art galleries and related issues are hard tobe good application in the actual layout of video surveillancecapture point.
Applied researches of monitor sensors layout mainlyput video monitor sensors layout problem into set coveringproblem. Chakrabarty and Bulusu used the method of linearprogramming to obtain the minimum activity to maintaincoverage node set [11, 12]. Meguerdichian et al. made morecomplex coverage model which, from the perspective ofminimizing the uncovered area of the start, considers theproblem of network coverage uniformity runtime based onthe degree of coverage [13]. Erdem and Sclaroff proposedan efficient algorithm to calculate the radial scans of eachcollection point in the visual range of the camera, so thatthe total layout costs are optimized while the collection pointlayout constraints can be met [14].
Dataacquisitiondemands of
key areamonitoring
Data acquisition demands ofpassenger line monitoring
Data acquisition demands of completecoverage monitoring
Deman
ds co
vere
d are
a
Timeliness and precision
PriorityGradu
ally c
ontai
n
Figure 1: Demands relationship among three levels.
3. Problem Description
In this section, the sensor networks hierarchical optimizationproblem is described in three aspects. Firstly, the data acquisi-tion demands of security monitoring in HRTH are analyzed.Secondly, the hierarchical organization of sensor networks isdescribed. Based on the previous two parts, the basic processof security monitoring based on multilayer sensor networksis designed in the last part.
3.1. Data Acquisition Demands of Security Monitoringin HRTH. According to the different safety forewarningfocuses, the securitymonitoring can be divided into three lev-els, and each level has its specific data acquisition demands.The data acquisition demands of security monitoring inHRTH are shown in Table 1.
The demands relationship among three levels is shown inFigure 1.The demands covered areas are gradually increasingfrom the first level to the third level, and the timeliness andprecision of data acquisition are gradually increasing from theopposite direction.
3.2. Hierarchical Organization of Sensor Networks. Based onthe data acquisition demands analysis above, the sensor net-works for security monitoring in HRTH are classified intothree hierarchies, which are one-to-one correspondence tothe data acquisition demands levels. The hierarchical orga-nization of sensor networks is shown in Table 2.
Journal of Sensors 3
Table 2: Hierarchical organization of sensor networks.
Sensornetwork
First hierarchy Key areamonitoring sensors
(i) Sensors in different key areas are independent and do not haveany relevance(ii) Sensors do not need adjustment after setting(iii) Sensors have front-end event detecting software
Second hierarchy Passenger linemonitoring sensors
(i) Sensors should be set following the passenger line(ii) Sensors in same passenger line have association(iii) Sensors do not need adjustment after setting(iv) The data acquired by sensors should be continuouslytransferred to the control center to process
Third hierarchy Complete coveragemonitoring sensors
(i) Sensors should cover all the function areas in HRTH(ii) The monitoring areas of sensors should reduce overlaps asmuch as possible(iii) Sensors can adjust monitoring areas after setting(iv) The data acquired by sensors should be continuouslytransferred to the control center to be stored
Keyareas
Passengerline
WholeHRTH
Sensornetwork
Anomalydetectiondatabase
Passengerflow
database
HRTHarea monitoring
database
Controlcenter
...
...
...
First
Secondhierarchy
Thirdhierarchy
hierarchy
Figure 2: Basic structure of security monitoring.
3.3. Basic Structure of SecurityMonitoring Based onMultilayerSensor Networks. According to the above analysis in thissection, a basic process of security monitoring is designedbased on multilayer sensor networks. The structure is shownin Figure 2.
As observed in Figure 2, in the first hierarchy, anomaliesin key areas are detected by monitoring sensors, and thenthe anomaly detection data are transmitted to control centerand inform monitoring personnel to respond. In the secondhierarchy, passenger flow data are acquired by monitoringsensors and transmitted to control center. According to theincidence relation among sensors, the passenger flow datacan be processed to obtain real-time passenger flow status,search the post node in passenger line, and forecast thevariation trend of passenger flow. The monitoring personnelcan forewarn passenger flow congestion andmake emergency
response based on the processing results. The third hierarchymainly focuses on the overall safe state of HRTH. Themonitoring personnel need to use the monitoring sensors toobserve function areas in HRTH when the first or secondhierarchy has safety forewarning. This hierarchy is a supple-ment for the previous two hierarchies.
4. Sensor Networks HierarchicalOptimization Model
According to the problem description in Section 3, a sensornetworks hierarchical optimization model is proposed inthis section. Sensor networks for security monitoring inHRTH are optimized from three hierarchies based on thehierarchical organization mentioned above. The hierarchicaloptimization framework is shown in Figure 3.
4.1. Key Area Sensors Optimization. The core concern ofkey area sensors optimization is to determine the key areasin HRTH. According to the different area characteristic,the key areas can be mainly divided into congestion areasand sensitive areas. Each area has its specific determinationmethod.
4.1.1. Congestion Areas Determination Method. The conges-tion areas are mainly determined by the computation result.There are three main methods to calculate the relationbetween passenger flow and facilities capacity, which aredescribed as follows.
(1) Capacity Method. Capacity method is used to determinefacilities congestion. This method divides the passenger lineinto several units and calculates the capacity balance ofunits. When the facility design capacity is less than practicalcapacity, this facility is considered a key area. The facilitydesign capacity is calculated by
𝐶 = 𝑊 ⋅ 𝑞 ⋅ 𝜑. (1)
𝑊 is the width of the facility. 𝑞 is the predicted passenger flowvolume. 𝜑 is the peak period coefficient.
4 Journal of Sensors
Key areasensors
optimization
Passengerline sensors
optimization
Whole areasensors
optimization
Congestion areasdetermination
Sensitive areasdetermination
Sensor layout ofkey areas
Passenger linetype determination
Passenger linegeneration
Sensor layout ofpassenger lines
Whole area sensorsoptimization model
A heuristicalgorithm
Sensor layout ofwhole area
Sensornetworks
hierarchicaloptimization
Figure 3: Hierarchical optimization framework.
Table 3: Calculation of three behaviors.Delay behaviors Generation mechanism Calculation Range of application
Queuing delay The facility service capability isless than passenger arrival rate
𝜆
(𝜇 − 𝜆)2
Ticket entrance, wicket,baggage check entrance,and so forth
Congestingdelay
The facility cannot be used whenthe passengers arrive ∫
𝑡𝑛
𝑡0
𝑞 (𝑡) 𝑑𝑡 − 𝑛∫𝑛𝑇+(𝑛−1)(𝑘
1+𝑘2)𝑞(𝑡)
𝑛𝑇+(𝑛−1)𝑞(𝑡)
𝑞 (𝑡) 𝑑𝑡Ticket entrances delaycheck caused by train late
Waiting delay
The facility capacity isinsufficient, which leads to highdensity and low speed ofpassenger flow
𝐿𝑘𝑗
𝑘𝑢𝑓
The service capability ofinterface channel betweenservice nodes is insufficient
(2) Delay Method. Delay is an important judging parameterfor the congestion of passenger line. The passenger delay inHRTHmainly results from queuing, congesting, and waitingbehaviors. The calculation of three behaviors is shown inTable 3.
(3) Density Method. Passenger flow density is an effectiveindicator tomeasure congestion level.Thehigher density pas-senger flow has, the more congestion in passenger line arises.This density is named congestion density and calculated by
𝜌𝑖𝑗(𝑡) =𝑛𝑖𝑗(𝑡)
𝑀𝑖𝑗
. (2)
𝜌𝑖𝑗(𝑡) is congestion density of the 𝑗th segment in the 𝑖th
passenger line. 𝑀𝑖𝑗is the facility available area of the 𝑗th
segment in the 𝑖th passenger line. 𝑛𝑖𝑗(𝑡) is passenger amount
of the 𝑗th segment in the 𝑖th passenger line.
4.1.2. Sensitive Areas Determination Method. Sensitive areasdetermination, compared with congestion areas determina-tion, is relatively simple and does not have specific calculatingmethod. Most of sensitive areas are determined based onthe actual demand of security monitoring in HRTH. Thecommon sensitive areas include distribution facility areas,fireproofing facility areas, office areas, and security checkareas.
4.2. Passenger Line Sensors Optimization. The core concernsof passenger line sensors optimization are to determine thepassenger line type and generate the passenger line underestablished facility layout.
Transferhall
Waitinghall
Ticket entrance Platform
Arriving Entering Waitingtrain
Gettingin train
Passenger inputline
Passenger inputprocedure
Purchasingtickets Shopping
Securitychecking Checking
Entrance Securitycheck point
Ticket office
Business service
area
Figure 4: Passenger input line and procedure.
4.2.1. Passenger Line Type Determination. The passenger linein HRTH can be mainly divided into passenger output line,passenger input line, and passenger transfer line. These threetypes of passenger lines are described as follows.
(1) Passenger Input Line. The passenger input line begins atpassenger arriving at HRTH and finishes after passenger getsin trains. In the period between passenger arriving and leav-ing, there are several events happening, such as purchasingtickets, shopping, dining, andwaiting in trains.Thepassengerinput line and procedure are shown in Figure 4.
(2) Passenger Output Line. The passenger output line beginsat train arriving at HRTH and finishes after passenger leavesHRTH. Compared with passenger input line, passengeroutput line has few events and is relatively simple.The outputpassengers flow has characteristics of being concentrated,
Journal of Sensors 5
Platform Exit Ticket exit Transfer hall
Getting offtrain Exiting Checking Leaving
Business service area
Shopping Passenger output
procedure
Passenger output line
Figure 5: Passenger output line and procedure.
high density, and short stay time. The passenger output lineand procedure are shown in Figure 5.
(3) Passenger Transfer Line. Passenger transfer line is similarto passenger input line and relatively simple, so we do notintroduce it in detail.
4.2.2. Passenger Line Generation. After determining the pas-senger line type, passenger line is generated based on theestablished facility layout of HRTH. Generation steps areshown as follows.
Step 1. Mark the geometric center of facilities in HRTHfunction areas and use these geometric centers as the origin-destination points.
Step 2. Use directed line segments to link geometric centersbased on passenger moving tracks in different type passengerlines.
Step 3. Classify the directed line segments, use different colorto denote different type passenger lines, and use differentthicknesses lines to denote the amount of passenger flow.
4.3. Whole Area Sensors Optimization. In order to ensurethat all function areas in HRTH are covered by sensors, awhole area sensors optimization framework is proposed inthis section. After space two-dimension, space partition andvisibility analysis, we change the whole area sensors opti-mization problem into set covering problem and developa set covering model. The whole area sensors optimizationframework is shown in Figure 6.
Step 1 (HRTH space two-dimension). HRTH space two-dimension is tomake the three-dimensional space into a two-dimension ichnography and scale down the layout of facilitiesand instruments. After HRTH space two-dimension, we canobtain a schematic representation of whole HRTH.
Step 2 (HRTH space partition). After obtaining the sche-matic representation, we abstract the facilities and instru-ments into square or rectangle and lengthen the sides of
InputSpace information of HRTH
HRTH space two-dimension
HRTH space partition
Whole area sensorsoptimization model
A heuristic
Visibility analysis
OutputSensor layout of whole area
algorithm
Figure 6: Whole area sensors optimization framework.
square and rectangle. A space partition sample is shown inFigure 7.
Step 3 (visibility analysis). Based on HRTH space partition,we analyze the visibility of each region in schematic repre-sentation. The visibility analysis includes two parts.
The first part is to analyze the geometric visibility. Assumethe region center is the laying position. If the link linebetween two regions is not interrupted by facilities or instru-ments, the two regions are considered geometric visibility.Figure 8(a) is a geometric visibility analysis sample. The geo-metric visibility set of 𝑅
12is {𝑅3, 𝑅4, 𝑅5, 𝑅7, 𝑅8, 𝑅9, 𝑅10, 𝑅11,
𝑅13, 𝑅14, 𝑅15, 𝑅16, 𝑅17}.
The second part is visual range analysis. We set thecoverage of one sensor as a circle whose radius is the visualrange of sensor. The regions which are covered by the circleare the visual regions. Figure 8(b) is a visual range analysissample. The visual visibility set of 𝑅
12is 𝑉12= {𝑅7, 𝑅8, 𝑅10,
𝑅11, 𝑅13, 𝑅14, 𝑅17}.
Step 4 (a set covering model for whole area sensors opti-mization). In this step, we develop a set covering model todescribe the whole area sensors optimization problem.
(1) Notations and Variables. Consider the following:
𝑖, 𝑗: region index,𝑚: total amount of regions,𝑅: visual range of sensor,dis(𝑖, 𝑗): the distance between region 𝑖 and region 𝑗,𝑊𝑖: weighted values,𝑥𝑖: 0-1 variable; if sensor is set in region 𝑖, 𝑥
𝑖= 1;
otherwise, 𝑥𝑖= 0,
cov(𝑖, 𝑗): 0-1 variable; if the visual range of region 𝑖 cancover region 𝑗, cov(𝑖, 𝑗) = 1; otherwise, cov(𝑖, 𝑗) = 0.
6 Journal of Sensors
Facilities and instruments
(a)
Facilities and instruments
R1 R6 R10 R15
R16
R17
R18
R11
R12
R13
R14
R7
R8
R9
R2
R3
R4
R5
(b)
Figure 7: A space partition sample.
R1 R6 R10 R15
R2 R11 R16
R3
R4 R8
R7 R12 R17
R13
R5 R9 R14 R18
(a)
R1 R6 R10R15
R2
R11
R16
R3
R4
R8
R7R12 R17
R13
R5 R9R14
R18
(b)
Figure 8: Visibility analysis samples.
(2) Objective Function. The objective function of whole areasensors optimization model is written as follows:
Minimize𝑍 =𝑚
∑𝑖=0
𝑊𝑖𝑥𝑖. (3)
The objective function minimizes the amount of sensorsto cover whole function areas in HRTH.
(3) Constraints. The constraints of whole area sensors opti-mization model are introduced as follows to ensure thepractical feasibility of the solution:
𝑚
∑𝑖1=0
cov (𝑖, 𝑗) 𝑥𝑖1
≥ 1, 0 ≤ 𝑗 ≤ 𝑚, (4)
cov (𝑖, 𝑗) ⋅ (𝑅 − dis (𝑖, 𝑗)) ≥ 0,
0 ≤ 𝑖 ≤ 𝑚, 0 ≤ 𝑗 ≤ 𝑚,(5)
(1 − cov (𝑖, 𝑗)) ⋅ (dis (𝑖, 𝑗) − 𝑅) ≥ 0,
0 ≤ 𝑖 ≤ 𝑚, 0 ≤ 𝑗 ≤ 𝑚,(6)
𝑥𝑖∈ {0, 1} , (7)
cov (𝑖, 𝑗) ∈ {0, 1} . (8)
Constraint (4) represents that each region in HRTHshould at least be covered by one sensor. Constraint (5)ensures that the distance between sensor and covered regioncannot be larger than the visual range of sensor. Constraint(6) represents that the region whose distance is larger thanthe visual range of sensor cannot be covered by this sensor.Constraint (7) and Constraint (8) are 0-1 variable constraints.
Step 5 (solution algorithm). In order to solve the opti-mization model developed above, a heuristic algorithm isproposed in this section. 𝑋 is the area set of two-dimensiondivision. 𝑚 is the elements amount in 𝑋. 𝐶 is the area set of
Journal of Sensors 7
Start
Yes
No
No
Yes
Calculate the element amount in set C
Yes
No
No
Yes
Sensor layout of whole area
Set C and S empty
Randomly select Rn in X
Rn ∈ S
Obtain the visual visibility region Vn of Rn
Add Rn to set C and S;add Vn to set S
S = X
Set qn = qn−1 qn − qn−1 ≥ 0
Set n = n + 1
n = m
Figure 9: Algorithm implement process.
sensors layout. 𝑞 is the elements amount in 𝐶. The algorithmimplement process is shown in Figure 9.
5. A Case Study
To illustrate the proposed model and algorithm for sensornetworks hierarchical optimization problem, a case study isperformed by using the actual data from a specific HRTH inChina.We choose comprehensive transfer layer of the HRTHas optimization object.
This layer is composed by transfer hall, parking area,passenger output system, and passenger input system. Thereare six entrances, six exits, and four ticket offices in this layer.The transfer hall connects with metro, taxi, and bus. Thewhole layer has various kinds of passenger lines and crossoveramong passenger lines.
We use the hierarchical optimization method mentionedin Section 4 to optimize the sensor networks in this layer.Thehierarchical optimization is shown as follows.
5.1. Key Area Sensors Optimization. According to the actualpassenger flow data in this layer, we use the key area deter-mination method mentioned in Section 4.1 to determine keyareas. The distribution of key areas in this layer is shown inFigure 10.
Figure 10: Distribution of key areas.
Figure 11: Passenger lines in the layer.
Figure 12: Space partition process.
5.2. Passenger Line Sensors Optimization. Through analysisof origin-destination points and passenger moving tracks, wegenerate the passenger lines in this layer. The passenger linesare shown in Figure 11.
5.3. Whole Area Sensors Optimization. Follow the whole areasensors optimization framework mentioned in Figure 6; thespace is partitionedwhich is shown in Figure 12 and the spacepartition result is shown in Figure 13.
After space partition, the layer is divided into 58 regions.Through the visibility analysis, we can obtain the visualvisibility sets of 58 regions and use the heuristic algorithmto find a solution for the whole area sensors optimizationmodel. The final sensors layout region set is {𝑅
2, 𝑅5, 𝑅6,
𝑅9, 𝑅12, 𝑅14, 𝑅21, 𝑅22, 𝑅29, 𝑅30, 𝑅34, 𝑅36, 𝑅41, 𝑅49, 𝑅50, 𝑅57}. The
solution obtained by the heuristic algorithm is shown inFigure 14.
According to the hierarchical optimization, the finalsensor networks for security monitoring in HRTH are shownin Figure 15.
6. Conclusion
In this paper, we considered the sensor networks hierarchicaloptimization problem in HRTH. A hierarchical optimizationframework is proposed, and the problem is solved from
8 Journal of Sensors
R1
R2
R3
R4
R6
R7
R8 R12
R11
R17 R25 R33R29
R28
R34
R35
R36 R46
R47
R43
R55
R54R50
R51
R58
R57
R56R53R45R40
R41
R42
R44
R39
R52R48
R49
R16 R20
R21
R24 R32
R13
R14
R15
R18 R26R22
R23
R37R30
R31 R38R27R19
R9
R10
R5
Figure 13: Space partition result.
R1
R2
R3
R4
R6
R7
R8 R12
R11
R17 R25 R33R29
R28
R34
R35
R36 R46
R47
R43
R55
R54R50
R51
R58
R57
R56R53R45R40
R41
R42
R44
R39
R52R48
R49
R16 R20
R21
R24 R32
R13
R14
R15
R18 R26R22
R23
R37R30
R31 R38R27R19
R9
R10
R5
R4
R5
R6
R7
R16 R20
R21
R
R29
R284 R3
R25
R24
R22
R23 RR 9
R30
R31 R27
R26
R
R
R36
R
R4
R37
R46
R47 R55
R54R50
R51
R42
3
R34
R
R399
R3329
R
R535
R44 R52R48
R49
R
R58
R57
R56R12
RR11
R17
R41R35
R4 RR42
34
R
R4
R35
R4540
R14
R155
10
R18
R1
R2
R3
R9
5
R9
R1R8
R13
R
99
R10
R
Figure 14: Solution obtained by the heuristic algorithm.
Sensors for key areasSensors for passenger linesSensors for whole areas
Figure 15: Final sensor networks for security monitoring in HRTH.
three hierarchies which are key area sensors optimization,passenger line sensors optimization, and whole area sensorsoptimization. In the third hierarchy, a whole area sensors
optimization model is developed and a heuristic algorithm isdesigned. Case study on a specific HRTH in China showedthat the hierarchical optimization method is effective tooptimize the sensor networks for security monitoring inHRTH. In the future, considering the layout costing inoptimization method is a possibility for further research.
Conflict of Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper.
Acknowledgments
This research was supported by the Fundamental ResearchFunds for the Central Universities (Grant no. 2015JBM044),the National Natural Science Foundation of China (Grant no.61374157), and the Talented Faculty Funds of Beijing JiaotongUniversity (Grant no. 2014RC005).
Journal of Sensors 9
References
[1] A. Aggarwal,The art gallery theorem: its variations, applicationsand algorithmic aspects [Ph.D. thesis], JohnsHopkinsUniversity,Baltimore, Md, USA, 1984.
[2] V. Chvatal, “A combinatorial theorem in plane geometry,”Journal of CombinatorialTheory Series B, vol. 18, pp. 39–41, 1975.
[3] S. Fisk, “A short proof of Chvatal’s watchmen theorem,” Journalof Combinatorial Theory Series B, vol. 24, no. 3, 374 pages, 1978.
[4] D. Avis and G. T. Toussaint, “An efficient algorithm for decom-posing a polygon into star-shaped polygons,”The Journal of thePattern Recognition Society, vol. 13, no. 6, pp. 395–398, 1981.
[5] B. Chazelle, “A theorem on polygon cutting with applications,”in Proceedings of the 23rd Annual Symposium on Foundationsof Computer Science (SFCS ’82), pp. 339–349, Chicago, Ill, USA,November 1982.
[6] D. T. Lee and A. K. Lin, “Computational complexity of art gal-lery problems,” IEEE Transactions on Information Theory, vol.32, no. 2, pp. 276–282, 1986.
[7] J.-I. Doh andK.-Y. Chwa, “An algorithm for determining visibil-ity of a simple polygon from an internal line segment,” Journalof Algorithms, vol. 14, no. 1, pp. 139–168, 1993.
[8] Y. Ke, “Detecting the weak visibility of a simple polygon andrelated problems,” Tech. Rep., Johns Hopkins University, Balti-more, Md, USA, 1987.
[9] J. O. Rourke,Art GalleryTheorems and Algorithms, Oxford Uni-versity Press, New York, NY, USA, 1987.
[10] T. C. Shermer, “Recent results in art galleries,” Proceedings of theIEEE, vol. 80, no. 9, pp. 1384–1399, 1992.
[11] K. Chakrabarty, S. S. Iyengar, H. Qi, and E. Cho, “Grid cov-erage for surveillance and target location in distributed sensornetworks,” IEEE Transactions on Computers, vol. 51, no. 12, pp.1448–1453, 2002.
[12] N. Bulusu, J.Heidemann, andD. Estrin, “Adaptive beaconplace-ment,” in Proceedings of the 21st IEEE International Conferenceon Distributed Computing Systems, pp. 489–498, April 2001.
[13] S. Meguerdichian, F. Koushanfar, G. Qu, and M. Potkonjak,“Exposure in wireless ad-hoc sensor networks,” in Proceedingsof the 7th Annual International Conference onMobile Computingand Networking, pp. 139–150, July 2001.
[14] U. M. Erdem and S. Sclaroff, “Automated camera layout to sat-isfy task-specific and floor plan-specific coverage require-ments,” Computer Vision and Image Understanding, vol. 103, no.3, pp. 156–169, 2006.
International Journal of
AerospaceEngineeringHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
RoboticsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Active and Passive Electronic Components
Control Scienceand Engineering
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
RotatingMachinery
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporation http://www.hindawi.com
Journal ofEngineeringVolume 2014
Submit your manuscripts athttp://www.hindawi.com
VLSI Design
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Shock and Vibration
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Civil EngineeringAdvances in
Acoustics and VibrationAdvances in
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Electrical and Computer Engineering
Journal of
Advances inOptoElectronics
Hindawi Publishing Corporation http://www.hindawi.com
Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
SensorsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Modelling & Simulation in EngineeringHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Chemical EngineeringInternational Journal of Antennas and
Propagation
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Navigation and Observation
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
DistributedSensor Networks
International Journal of
top related