Documents

A Survey on Mobile Anchor Node Assisted Localizati

Description
A Survey on Mobile Anchor Node Assisted Localization in WSN
Categories
Published
of 25
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  1553-877X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/COMST.2016.2544751, IEEECommunications Surveys & Tutorials 1 A Survey on Mobile Anchor Node AssistedLocalization in Wireless Sensor Networks Guangjie Han,  Member, IEEE,  Jinfang Jiang, Chenyu Zhang, Trung Q. Duong,  Senior Member, IEEE,  MohsenGuizani,  Fellow, IEEE,  George Karagiannidis,  Fellow, IEEE   Abstract —Localization is one of the key technologies in Wire-less Sensor Networks (WSNs), since it provides fundamentalsupport for many location-aware protocols and applications.Constraints on cost and power consumption make it infeasibleto equip each sensor node in the network with a Global PositionSystem (GPS) unit, especially for large-scale WSNs. A promisingmethod to localize unknown nodes is to use mobile anchor nodes(MANs), which are equipped with GPS units moving amongunknown nodes and periodically broadcasting their currentlocations to help nearby unknown nodes with localization. Aconsiderable body of research has addressed the Mobile AnchorNode Assisted Localization (MANAL) problem. However to thebest of our knowledge, no updated surveys on MAAL reflectingrecent advances in the field have been presented in the pastfew years. This survey presents a review of the most successfulMANAL algorithms, focusing on the achievements made in thepast decade, and aims to become a starting point for researcherswho are initiating their endeavors in MANAL research field. Inaddition, we seek to present a comprehensive review of the recentbreakthroughs in the field, providing links to the most interestingand successful advances in this research field.  Index Terms —wireless sensor network, localization, mobileanchor node, mobility model, path planning. I. I NTRODUCTION Wireless Sensor Networks (WSNs) consist of a set of physically small sensor nodes deployed in a given monitoringarea (region), namely, in two-dimensional (2D) or three-dimensional (3D) environments, to fulfill tasks such as surveil-lance, biological detection, home care, object tracking, etc.,[1], [2], [3]. The monitoring information is sent to sink nodesvia multi-hop communication [4], [5], [6]. The sink collectsthe sensing data from the sensor nodes and then processes thisinformation as required by the specific applications [7].In WSNs, determining unknown nodes’ locations is acritical task since it provides fundamental support for manylocation-aware protocols and applications, such as location-based routing protocols, where the location information is Guangjie Han and Jinfang Jiang are with the Department of Communica-tion & Information Systems, Hohai University, Changzhou, China. E-mail:hanguangjie@gmail.com, jiangjinfang1989@gmail.com.Chenyu Zhang is in the State Grid Wulumuqi Electric Power Sup-ply Company, Beijing South Road, No.1,Wulumuqi, 830000, China ,zhangchenyu.hhu@gmail.com.Trung Q. Duong is with Queen’s University Belfast, UK, E-mail:trung.q.duong@gmail.com.Mohsen Guizani is the Department of Electrical and Computer Engineering,University of Idaho, USA, E-mail: mguizani@ieee.org.G. K. Karagiannidis is with the Department of Electrical and ComputerEngineering, Khalifa University, PO Box 127788, Abu Dhabi, UAE and withthe Department of Electrical and Computer Engineering, Aristotle Universityof Thessaloniki, 54 124, Thessaloniki, Greece. E-mail: geokarag@ieee.org. critical for sensor nodes to make optimal routing decisions[8], [9]. The problem of localization is the process of findinglocation information of the sensor nodes in a given coordinatesystem. To localize a WSN in the global coordinate system,some special nodes should be aware of their positions inadvance either from Global Position System (GPS) or byvirtue of being manually placed, which are called anchors(beacons). Other nodes, which are usually called unknownnodes, calculate their positions by using special localizationalgorithms [10]. Sensor nodes localization usually consistsof two steps: (i) distance measurement between neighboringnodes, and (ii) geometric calculation based on measureddistances. Based on the distance measurement techniquesused, localization algorithms can be classified into range-basedlocalization algorithms and range-free localization algorithms.Range-based localization means that distances between sensornodes are estimated by using some physical properties of communication signals, i.e., Received Signal Strength Indi-cator (RSSI), Time of Arrival (ToA), Time Difference of Arrival (TDoA) and Angle of Arrival (AoA) [11], [12]. Range-free localization algorithms estimate sensor node’s coordinatesusing connectivity information between sensor nodes withoutranging (i.e., distance or angle) information. Unknown NodeMobile Anchor NodeAnchor Point  R R R Fig. 1. Mobile anchor node assisted localization. It is costly to equip each sensor node with a GPS unit,especially for large-scale WSNs. A feasible method to localizeunknown nodes is to use several mobile anchor nodes (MANs)which are equipped with GPS units moving among unknownnodes and periodically broadcasting their current locations (an-chor points) to help nearby unknown nodes with localization[13], [14], [15], as shown in Fig. 1. A considerable bodyof research has addressed the Mobile Anchor Node Assisted  1553-877X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/COMST.2016.2544751, IEEECommunications Surveys & Tutorials 2 Localization (MANAL) problem. This kind of architectureoffers significant practical benefits, since the mobile anchornode is not as energy constrained as an unknown node andthe localization accuracy also can be improved by carefullydesigning movement trajectory of the mobile anchor nodes.Moreover, the size of a robot is much larger than the size of asensor node and thus it is much easier to install a GPS unit onthe robot [16]. Therefore, this is a viable solution that couldbe used.A fundamental research issue of MANAL algorithms is todesign movement trajectories that mobile anchor nodes shouldmove along in a given monitoring area (region) in order toimprove localization performances of WSNs.Another research issue of MANAL algorithms is the lo-calization methods by which unknown nodes calculate theirpositions based on the beacon packets received from location-aware mobile anchor nodes, as they move through the monitor-ing area (region). These algorithms employ either only mobileanchor nodes (a single mobile anchor node or a group of mobile anchor nodes) or mobile anchor nodes together withreference nodes to help unknown nodes with localization.Generally, MANAL algorithms involve three stages: (i)mobile anchor nodes traverse the monitoring area (region)while periodically broadcasting beacon packets which includetheir current positions; (ii) unknown nodes within the com-munication ranges of the anchors receive the beacon packetsand estimate distances to the anchors by using the physicalproperties of communication signals when needed; and (iii)unknown nodes calculate their positions if they fall inside theoverlapping communication ranges of at least three (four) non-collinear (non-coplanar) anchor nodes by the use of appropri-ate localization algorithms in 2D (3D) WSNs.There is plenty of literature discussing the MANAL prob-lem. However, there is no recent review discussing MANALalgorithms in the past few years. This paper’s objective is tofill this gap and provide a comprehensive review of the recentbreakthroughs in the field, focusing on the achievements madein the past decade, and aims to become a starting point forresearchers who are initiating their endeavors in the MANALresearch field. In addition, we seek to present a comprehensivereview of the recent breakthroughs in the field, providing linksto the most interesting and successful advances in this researchfield. We survey the current works on the above two issues,namely, movement trajectories and localization methods.Resulting from these considerations, the remainder of thisarticle is organized as follows. Section II introduces back-grounds and basic localization methods for WSNs. SectionIII presents related work about existing survey paper forlocalization algorithms in WSNs. Section IV introduces theclassification of MANAL algorithms. Section V and VI re-views two categories of MANAL algorithms in detail. InSection VII illustrates existing problems and future researchissues in the MANAL research field. Finally, Conclusionsincluding a summary table is given in Section VIII.II. B ACKGROUND KNOWLEDGE AND BASIC LOCALIZATIONMETHODS  A. Basic Terminologies  Anchor (Beacon) Node : To localize a WSN in the globalcoordinate system, some special sensor nodes should be awareof their positions in advance either from GPS or by virtueof being manually placed, which are called  anchor nodes  or beacon nodes . Unknown (Ordinary) Node : Sensor nodes that do not knowtheir positions and need to calculate them with the help of anchor nodes. Static Anchor (Beacon) Node : The anchor (beacon) nodewhich cannot move automatically after initial deployment.  Mobile Anchor (Beacon) Node : The anchor (beacon) nodewhich can move automatically after initial deployment.  Reference Node : The sensor node which already knows itscoordinates and works as anchor node to help other unknownnodes with localization.  Anchor (Beacon) Packet : The data packet broadcasted bymobile anchor nodes periodically.  Anchor (Beacon) Point : Virtual coordinates broadcasted bymobile anchor nodes periodically, which is part of the anchorpacket.  Broadcast Interval  : The time period a mobile anchor nodetakes to broadcast beacon packets.  Node’s Speed  : The mobility features that capture node move-ment speed in mobility models.  Node’s Direction : The mobility features that capture nodemovement direction in mobility models.  Pause Time : The time period that a node is steady in aspecific position, i.e., the interval of time when the node’sspeed is zero or close to zero.  Inter-contact Time : The time interval between two consec-utive contacts of the same two nodes. Contact Duration : The time period two nodes attain whilewithin the same radio range.  B. Basic Methods of Calculating Sensor Nodes’ Location1) Trilateration:  Trilateration is the process of finding theposition of an unknown node based on its distances to threeanchors, as shown in Fig. 2. Assume that the coordinateof an unknown node  D  is  ( x,y ) . The coordinates of threeanchor nodes  A ,  B ,  C   are  ( x a ,y a ) ,  ( x b ,y b ) , and  ( x c ,y c ) .The distances between  D  and  A ,  B ,  C   are  d a ,  d b  and  d c ,respectively. These geometric constraints can be expressed bythe following system of equations [17],           ( x − x a ) 2 + ( y − y a ) 2 =  d a    ( x − x b ) 2 + ( y − y b ) 2 =  d b    ( x − x c ) 2 + ( y − y c ) 2 =  d c .  (1)By solving Eq. (1), we can get the matrix  AX   =  B , where X   =    x y   T  ,A  = 2    ( x a − x c ) ( y a − y c )( x b − x c ) ( y b − y c )    ,  1553-877X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/COMST.2016.2544751, IEEECommunications Surveys & Tutorials 3 b  =    x 2 a − x 2 c  + y 2 a − y 2 c  + d 2 c  − d 2 a x 2 b  − x 2 c  + y 2 b  − y 2 c  + d 2 c  − d 2 b   ,  A  B  D C  Unknown NodeAnchor Node Fig. 2. An example of trilateration. 2) Triangulation:  Triangulation, unlike trilateration, com-putes the position of an unknown node based on the angu-lar distance between three different pairs of anchor nodes.Consider the example depicted in Fig. 3, suppose that thecoordinate of an unknown node  D  is  ( x,y ) . The coordinates of three anchor nodes  A ,  B ,  C   are  ( x a ,y a ) ,  ( x b ,y b )  and  ( x c ,y c ) ,respectively. If we know the angles between the line segmentsconnecting  D  and the anchors, then the unknown node’scoordinates must be calculated using triangulation instead of trilateration. 1 1 1 ( , ) O O O x y ( , ) b b  B x y  ( , ) c c C x y ( , ) a a  A x y ( , )  D x y Unknown NodeAnchor Node Fig. 3. An example of triangulation. Let  ∠ ADB ,  ∠ ADC  ,  ∠ BDC   denote the angles betweenthe line segments connecting  D  to the anchors, respectively. D  is the intersection point of the three circles. If the angulardistance between the anchor nodes is known, the centers of thecircles can be obtained. For anchor nodes  A ,  C   and the angle ∠ ADC  , if the arc       AC   is within the scope of the   ABC  , thecircle can be uniquely identified. Assume that the center of thecircle is  O 1 ( x O 1 ,y O 1 ) , the radius is  r 1 , thus,  α  = ∠ AO 1 C   =2( π  − ∠ ADC  ) .  O 1  and  r 1  can be calculated using Eq. (2)[18],           ( x O 1  − x a ) 2 + ( y O 1  − y a ) 2 =  r 1    ( x O 1  − x b ) 2 + ( y O 1  − y b ) 2 =  r 1 ( x a − x c ) 2 + ( y a − y c ) 2 = 2 r 21 − 2 r 21  cos α.  (2)Similarly, anchor nodes  A ,  B , the angle ∠ ADB  and anchornodes  B ,  C  , the angle  ∠ ADB  can determine  O 2 ( x O 2 ,y O 2 ) , r 2  and  O 3 ( x O 3 ,y O 3 ) ,  r 3 , respectively. Thus, knowing the co-ordinates of   O 1 ( x O 1 ,y O 1 ) ,  O 2 ( x O 2 ,y O 2 )  and  O 3 ( x O 3 ,y O 3 ) ,the coordinate of   D ( x,y )  can be calculated by using of Eq. (1). 3) Maximum Likelihood Estimation:  When the numberof anchor nodes  n >  3 , we use the maximum likelihoodestimation to calculate the coordinate of the unknown node D ( x,y ) . Assume that the coordinates of anchor nodes arerespectively  ( x 1 ,y 1 ) ,  ( x 2 ,y 2 ) ,  ( x 3 ,y 3 ) , ··· ,  ( x n ,y n ) , and thedistances between  D  and the anchor nodes are  d 1 ,  d 2 ,  d 3 , ··· , d n , respectively, as shown in Fig. 4. Then the equations canbe obtained as follows [17], [18]: Unknown NodeAnchor Node 1234  D n Fig. 4. An example of the maximum likelihood estimation.             ( x − x 1 ) 2 + ( y − y 1 ) 2 =  d 21 ( x − x 2 ) 2 + ( y − y 2 ) 2 =  d 22 ( x − x 2 ) 2 + ( y − y 2 ) 2 =  d 22 ... ( x − x n ) 2 + ( y − y n ) 2 =  d 2 n .  (3)By subtracting the last equation from the first  n  −  1 equations, we can obtain             2( x 1 − x n ) x + 2( y 1 − y n ) y  = d 2 n − d 21  + x 21 − x 2 n  + y 21 − y 2 n ... 2( x n − 1 − x n ) x + 2( y n − 1 − y n ) y  = d 2 n − d 2 n − 1  + x 2 n − 1 − x 2 n  + y 2 n − 1 − y 2 n .  (4)With some proper transformations, the above equation canbe rewritten as  AX   =  b , where X   =    x y   T  ,A  = 2     ( x 1 − x n ) ( y 1 − y n ) ...... ( x n − 1 − x n ) ( y n − 1 − y n )     ,b  =     d 2 n − d 21  + x 21 − x 2 n  + y 21 − y 2 n ... d 2 n − d 2 n − 1  + x 2 n − 1 − x 2 n  + y 2 n − 1 − y 2 n     .  1553-877X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/COMST.2016.2544751, IEEECommunications Surveys & Tutorials 4 Fig. 5. Classification of mobile anchor node assisted localization algorithms. Then, we can obtain  X   = ( A T  A ) − 1 A T  b .Actually, the maximum likelihood estimation is the exten-sion of the trilateration method.III. R ELATED WORK Recently, a large number of localization techniques andalgorithms have been proposed for WSNs, and simultaneouslymany studies have been done to survey and analyze existinglocalization techniques and algorithms. For example, in [17],Mao  et al.  first provide an overview of measurement tech-niques that can be used for WSN localization, e.g., distancerelated measurements, angle-of-arrival (AOA) measurementsand RSS profiling techniques. Then the one-hop and themulti-hop localization algorithms based on the measurementtechniques are presented in detail, respectively, where theconnectivity-based or “range free” localization algorithms andthe distance-based multi-hop localization algorithms are par-ticularly discussed due to their prevalence in multi-hop WSNlocalization techniques. In addition, based on the analysis, theopen research problems in the distance-based sensor network localization and the possible approaches to these problems arealso discussed.In [18], Amundson  et al.  present a survey on localizationmethods for mobile wireless sensor networks (MWSNs). First,the authors provide a brief taxonomy of MWSNs, includingthe three different architectures of MWSNs, the differencesbetween MWSNs and WSNs, and the advantages of addingmobility. The MWSN localization discussed in [18] is consistsof three phases: 1) coordination, 2) measurement, and 3)position estimation. In the coordination phase, sensor nodescoordinate to initiate localization, including clock synchroniza-tion and the notification that the localization process is about tobegin. In the second phase, the measurement techniques, e.g.,the angle-of-arrival (AOA) and the time-difference-of-arrival(TDOA) methods are presented. The measurements obtainedin the second phase can be used to determine the approximateposition of the mobile target node based on localizationalgorithms, e.g., the Dead Reckoning, the maximum likelihoodestimation (MLE) and the Sequential Bayesian estimation(SBE). To the best of our knowledge, the reference [18] isthe first survey focusing on MWSNs localization.In [19], an overview of localization techniques is presentedfor WSNs. The major localization techniques are classifiedinto two categories: centralized and distributed based on wherethe computational effort is carried out. Based on the details of localization process, the advantages and limitations of each lo-calization technique are discussed. In addition, future researchdirections and challenges are highlighted. This paper pointout that the further study of localization technique should beadapted to the movement of sensor nodes since node mobilitycan heavily affect localization accuracy of targets. However,the localization techniques proposed for mobile sensor nodesare not discussed in [19].In [20], localization algorithms are classified into tar-get/source localization and node self-localization. In thetarget localization, Single-Target/Source Localization inWSNs, Multiple-Target Localization in WSNs and Single-Target/Source Localization in Wireless Binary Sensor Net-works (WBSNs) are mainly introduced. Then, in node self-localization, range-based and range-free methods are investi-gated. With the widespread adoption of WSNs, the localizationalgorithms are very different for different applications. There-fore, in the paper, the localization in some special scenarios arealso surveyed, e.g., localization in non-line-of-sight (NLOS)scenarios, node selection criteria for localization in energy-constrained network, cooperative node localization, schedulingsensor nodes to optimize the tradeoff between localization per-formance and energy consumption, and localization algorithm

csme1421

Jul 29, 2018
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks