Review Article. ISSN (Print) *Corresponding author Roma Sharma

Size: px

Start display at page:

Download "Review Article. ISSN (Print) *Corresponding author Roma Sharma"

Juliet Phillips
5 years ago
Views:

1 Scholars Joural of Egieerig ad Techology (SJET) Sch. J. Eg. Tech., 2014; 2(2A): Scholars Academic ad Scietific Publisher (A Iteratioal Publisher for Academic ad Scietific Resources) ISSN X (Olie) ISSN (Prit) Review Article Localizatio Techiques i WSN: A Review Roma Sharma 1, Prof. Dr. (Mrs.) Rama Sushil 2 1 Faculty of Computer Sciece & Applicatio,M.Tech. CSEstudet, DIT Uiversity, Dehradu, Uttarakhad, Idia 2 Facultyof Computer Sciece & Applicatio, DIT Uiversity, Dehradu, Uttarakhad, Idia *Correspodig author Roma Sharma Abstract: Wireless Sesor Network (WSN) is a kid of wireless etwork, which is composed of groups of very small devices kow as Sesor Nodes. It is oe of the fastest evolvig fields. Applicatios of WSN comprise of search, disaster relief, remote evirometal moitorig, target trackig etc. I all such applicatios, data gathered by the etwork is ot useful, without locatio iformatio. As odes locatio is required to report the origi of evet, about which the data is gathered. Node localizatio is highly importat, for large sesor etworks where users desire, to kow about the exact locatio of the odes to kow the evet locatio. This paper reviews differet localizatio techiques, used for odes locatio discovery i WSN ad presets their parametric compariso i tabular form. This compariso is useful for better uderstadig of existig techiques ad for the developmet of other ovel techique. Keywords: Wireless Sesor Network (WSN), approaches, sesor odes INTRODUCTION A WSN is a collectio of odes orgaized ito a cooperative etwork [1]. Each ode is capable of processig with the help of oe or more microcotrollers, CPUs or DSP chips, cotais multiple types of memory (program, data ad flash memories), have a RF trasceiver (usually with a sigle omidirectioal atea), ad have a power source (e.g., batteries ad solar cells). The odes commuicate wirelessly ad ofte self-orgaize after beig deployed i a ad hoc fashio. Such odes are called sesor odes. Each sesor ode is capable of oly a limited amout of processig. But whe coordiated with the iformatio from a large umber of other odes, they have the ability to measure a give physical eviromet i great detail. Thus, a sesor etwork ca be described as a collectio of sesor odes which coordiate to perform some specific actio. Various ovel applicatios are emergig i WSNs such as: habitat moitorig, smart buildig failure detectio ad reportig, ad target trackig. Oe commo feature shared by all of these critical applicatios is the vitality of sesor locatio. The core fuctio of a WSN is to detect ad report evets which ca be meaigfully assimilated ad respoded to oly if the accurate locatio of the evet is kow. A straightforward solutio is to equip each sesor with a GPS receiver that ca accurately provide the sesors with their exact locatio. This, however, is ot a feasible solutio from a ecoomic perspective sice sesors are ofte deployed i very large umbers ad maual cofiguratio is too cumbersome ad hece ot feasible. Therefore, localizatio i sesor etworks is very challegig. Over the years, may protocols have bee devised to eable the locatio discovery process i WSNs to be autoomous ad able to fuctio idepedetly of GPS ad other maual techiques.this paper gives a overview of various localizatio techiques. The rest of the paper is orgaized as follows: Sectio II gives the classificatio of Localizatio Techiques. I sectio III various Localizatio Techiques have bee compared ad sectio IV gives the coclusio. CLASSIFICATION OF LOCALIZATION TECHNIQUES I sesor etworks, odes are deployed ito a uplaed ifrastructure where there is o a priori kowledge of locatio. The problem of estimatig spatial-coordiates of the ode is referred to as localizatio. A immediate solutio which comes to mid, is GPS or the Global Positioig System. However, there are some strog factors agaist the usage of GPS. For oe, GPS ca work oly outdoors. Secodly, GPS receivers are expesive ad ot suitable i the costructio of small cheap sesor odes. A third factor is that it caot work i the presece of ay obstructio like dese foliage etc. Thus, sesor odes would eed to have other meas of establishig their positios ad orgaizig themselves ito a coordiate system without relyig o a existig ifrastructure. 173

2 Localizatio algorithms ca be classified ito followig categories. 1) Cetralized vs. distributed 2) Achor based vs. Achor less 3) Rage based vs. Rage free 1. Cetralized Localizatio: I certai etworks where a cetralized iformatio architecture already exists, such as road traffic moitorig ad cotrol, evirometal moitorig, health moitorig, ad precisio agriculture moitorig etworks, the measuremet data of all the odes i the etwork are collected i a cetral processor uit. I such a etwork, it is coveiet to use a cetralized localizatio scheme. Cetralized localizatio is basically migratio of iter-ode ragig ad coectivity data to a sufficietly powerful cetral base statio ad the the migratio of resultig locatios back to respective odes. The advatage of cetralized algorithms are that it elimiates the problem of computatio i each ode, at the same time the limitatios lie i the commuicatio cost of movig data back to the base statio. Distributed Localizatio: I the process of Distributed localizatio each ode idepedetly determies its locatio with oly limited commuicatio with oe hop or multi hops eighbor odes. It has the characteristics of small traffic, equal calculatio burde of each ode, little storage requiremets ad good scalability. However due to lack of global iformatio locatio accuracy is susceptible to umber f beaco odes ad the distributio of odes. 2. Achor Based: Achor-based algorithms operate o a ad-hoc etwork of sesor odes where a small percetage of the odes (achors) are aware of their positios either through maual cofiguratio or usig GPS. Achor odes broadcast their locatios iformatio to their eighbors. The goal is to estimate the positios of as may ukow odes as possible usig achor ode iformatio. Achor-based algorithms usually produce a absolute locatio system where absolute ode positio is kow, for example, latitude, logitude, ad altitude. However, the accuracy of the estimated positio is highly affected by the umber of achor odes ad their distributio i the sesor field [2]. Achor Free: Achor-free algorithms do ot make ay assumptios regardig ode positios. I this case, istead of computig absolute ode positios, the algorithm estimates relative positioig, i which the coordiate system is established by a referece group of odes. I some cases kowig the relative positios of the odes compared to each other is eough, for example, locatio aided routig [3]. Moreover, a relative coordiate system ca be trasformed to a absolute coordiate system if the coordiates of three separate o-coliear odes are kow i case of 2D (or four i case of 3D). Achorfree schemes have the disadvatage that whe the referece ode moves, positios have to be recomputed for odes that have ot moved. This is cosidered a mior problem i sesor etworks where sesor odes are usually assumed to be statioary. 3. Rage Based: Rage Based methods deped upo distace or agle betwee odes to obtai ukow ode s locatio. I the first stage, a ode merely estimates its distace to other odes i its viciity usig oe or more features of the received sigal. I the secod stage, a ode uses all the distace estimates to compute its actual locatio. The method employed i stage two to compute the actual locatio depeds o the sigal feature used i stage oe, such as: a. Received Sigal Stregth Idicator (RSSI) :This category of distace related measuremet techiques estimates the distace betwee eighborig sesors from the received sigal stregth measuremets [4]-[8].These techiques are based o a stadard feature foud i most wireless devices, a received sigal stregth idicator(rssi).they are attractive because they require o additioal hardware, ad are ulikely to sigificatly impact local power cosumptio,sesor size ad thus cost. I free space, the received sigal stregth at a receiver is give by Friis equatio [9] P r (d)= P tg t G r λ 2 (4π) 2 d 2 L Where P t is the trasmitted power, P r (d) is the received power at a distace d from the trasmitter, G t is the trasmitter atea gai,g r is the receiver atea gai, L is a system loss factor ot related to propagatio ad λ is the wavelegth of the trasmitter sigal. I a real eviromet, the propagatio of a electromagetic sigal is affected by reflectio, diffractio ad scatterig. Moreover the eviromet will chage depedig o particular applicatio(e.g., idoor versus outdoor).it is very difficult,if possible to obtai the received sigal stregth usig aalytical methods.however measuremets have show that at ay value of d,the received sigal stregth P r (d), at a particular locatio is radom ad distributed logormally about the mea distace-depedet value is, 174

3 P r (d)[dbm] =P o (d 0 )[dbm]-10 p log 10 d d o +X σ Where P 0 (d 0 )[dbm] is a referece power i db mill watts at a close-i referece distace d 0 from the trasmitter, p is the path loss expoet which idicates the rate at which the received sigal stregth decreases with distace ad the value of p depeds o the specific propagatio eviromet, X σ is a zero mea Gaussia distributed radom variable with stadard deviatio σ ad it accouts for the radom effect of shawdowig. b. Time of Arrival: This method -called also time of flight"-, exploits the relatioship betwee the beaco-ode distace ad the trasmissio time that a sigal has to travel betwee seder ad beaco. If the velocity of the sigal is kow ad assumig that the seder ad receptor kow the time whe a trasmissio starts, the the time of arrival of the sigal is a idicative of the beacoode distace ad this distace ca be computed usig the propagatio time by either the beaco or the ode [10]. The two mai disadvatages of this method are: it is ecessary to have a sychroized seder ad receiver. Depedig o the trasmissio medium that is used, a high clock resolutio is required to produce results of acceptable accuracy. For example, for acoustic waves, this requiremet is modest (about microsecods) but for radio, a very high resolutio is ecessary (about aosecods). c. Time Differece of Arrival (TDoA): There is a category of the localizatio algorithms utilizig TDoA measuremets of the trasmitter s sigal at a umber of receivers with kow locatio iformatio to estimate the locatio of the trasmitter. Fig.1 shows a TDoA localizatio sceario with a group of four receivers at locatios r 1, r 2, r 3, r 4 ad a trasmitter at r t. The TDoA betwee a pair of receivers i ad j Fig.1: Localizatio usig Time differece of arrival measuremets is give by: Δt ij t i t j = 1 C ( r i r t - r j - r t ), i j..3.3 where t i ad t j are the time whe the sigal is received at receivers i ad j respectively, c is the propagatio speed of the sigal ad deotes the euclidea orm. d. Agle of Arrival (AOA):This method cosists of the agle obtaied betwee a referece ode ad the ode which wats to kow its positio. The AOA is typically gathered usig radio-chips or microphoes array, which allows a listeig ode to determie the agular directio of a trasmittig ode [11]. Oe ca also obtai AOA data from optical commuicatio [12].This method is implemeted by spatially separated microphoes that hear a sigle trasmitted sigal. By aalyzig the phase or time differece betwee the sigal arrivals at differet microphoes, it is possible to discover the agle of arrival of the sigal. Ufortuately, AOA's hardware teds to have a big form factor ad is more expesive tha the distace estimatio methods, sice each ode must have oe speaker ad several microphoes. 4. Rage Free: Rage-free localizatio ever tries to estimate the absolute poit to poit distace based o received sigal stregth or other features of the received commuicatio sigal like time, agle, etc. This greatly simplifies the desig of hardware, makig rage-free methods very appealig ad a cost effective alterative for localizatio i WSNs. a. Cetroid: The Cetroid scheme was proposed by Bulusu et al. i [13]. I Cetroid based Algorithm all achors first seds their positios to all sesor odes withi their trasmissio rage. Each ukow ode listes for a fixed time period t ad collects all the beaco sigals it receives from various referece poits. Secodly, all ukow sesor odes positios are calculated by a cetroid determiatio from all positios of the achors i rage. The cetroid localizatio algorithm is simple but the locatio error is high due to the cetroid formula. It is the most basic scheme that uses achor beacos, cotaiig locatio iformatio (X i,y i ) to estimate ode positio. After gettig these beacos, the ukow ode use the followig cetroid formula to computes its locatio, give as: (X',Y') = ( i=1 X i, i=1 Y i )

4 Where is the umber of achor odes, (X i,y i ) are their coordiates ad (X',Y') are estimated positio coordiates of the achor ode. (X 1, Y 1 ) (X 2, Y 2 ) (Xꞌ, Yꞌ) (X 3, Y 3 ) (X 4,Y 4 ) Fig. 2:Cetroid Method b. DV-Hop Localizatio: DV-Hop localizatio is proposed by D. Niculescu ad B. Nath i the Navigate project [14]. DV-Hop localizatio uses a mechaism that is similar to classical distace vector routig. I this work, oe achor broadcasts a beaco to be flooded throughout the etwork cotaiig the achors locatio with a hop-cout parameter iitialized to oe. Each receivig ode maitais the miimum couter value per achor of all beacos it receives ad igores those beacos with higher hop-cout values. Beacos are flooded outward with hop-cout values icremeted at every itermediate hop. Through this mechaism, all odes i the etwork (icludig other achors) get the shortest distace, i hops, to every achor. I order to covert hop cout ito physical distace,the system estimates the average distace per hop without without rage based techiques. Achors perform this task by obtaiig locatio ad hop cout iformatio for all other achors iside the etwork. The average sigle hop distace is the estimated by achor i usig the followig formula: Hopsize i = x i x j 2 + y i y j 2 h j 4.2 where, (xj,yj) is the locatio of achor j, ad hj is the distace, i hops, from achor j to achor i. Oce calculated, achors propagate the estimated HopSize iformatio out to the earby odes. Oce a ode ca calculate the distace estimatio to more tha 3 achors i the plae, it uses triagulatio (multilateratio) to estimate its locatio. Theoretically, if errors exist i the distace estimatio, the more achors a ode ca hear the more precise localizatio will be. c. APIT: [15] is a area-based rage-free localizatio scheme. It assumes that a small umber of odes, called achors, are equipped with high-powered trasmitters ad kow their locatio, obtaied via GPS or some other mechaism.usig beacos from these achors, APIT employs a ovel area-based approach to perform locatio estimatio by isolatig the eviromet ito triagular regios betwee achor odes.a ode's presece iside or outside of these triagular regios allows a ode to arrow dow the area i which it ca potetially reside. By utilizig differet combiatios of achors, the size of the estimated area i which a ode resides ca be reduced, to provide a good locatio estimate. Fig 3: A Area-based APIT Algorithm Overview The actual algorithm is as follows: Step 1 Receive beaco positios from hearable beacos. Step 2 Iitialize iside-set to be empty. Step 3 For each triagle Ti i possible triagles formed over beacos, add T i to iside-set if ode is iside T i. Goto Step 4 whe accuracy of iside-set is sufficiet. Step 4 Compute positio estimate as the ceter of mass of the itersectio of all triagles i iside-set. The theoretical method used to arrow dow the possible area i which a target ode resides is called the Poit-I-Triagulatio Test (PIT). For three give achors: A(ax; ay);b(bx; by);c(cx; cy), the Poit-I- Triagulatio test determies whether a poit M with a ukow positio is iside triagle ΔABC or ot. APIT repeats this PIT test with differet achor combiatios util all combiatios are exhausted or the required accuracy is achieved. At this poit, APIT calculates the ceter of gravity (COG) of the itersectio of all of the triagles i which a ode resides to determie its estimated positio. 176

5 d. Amorphous Localizatio: The Amorphous Localizatio algorithm [16][17], proposed idepedetly from DV-Hop, uses a similar algorithm for estimatig positio. First, like DV-Hop, each ode obtais the hop distace to distributed achors through beaco propagatio. Oce achor estimates are collected, the hop distace estimatio is obtaied through local averagig. Each ode collects eighborig odes hop distace estimates ad computes a average of all its eighbors values. Half of the radio rage is the deducted from this average to compesate for error caused by low resolutio. The Amorphous Localizatio algorithm takes a differet approach from the DV-Hop algorithm to estimate the average distace of a sigle hop. This work assumes that the desity of the etwork, local, is kow a priori, so that it ca calculate HopSize offlie i accordace with the Kleirock ad Silvester formula [18]: Hopsize = r(1 + e local - ) local arccos t t 1 t e 2 dt Parametric Compariso Table 1: Performace Compariso Parameters Cetroid DV-Hop Amorphous APIT Accuracy Fair Good Good Good Node Desity >0 >8 >8 >6 Achor Heard >10 >8 >8 >10 ANR >0 >0 >0 >3 DOI Good Good Fair Good GPS Error Good Good Fair Good Overhead Smallest Largest Large Small Table 1 provides a overview of performace compariso of differet rage free schemes, ad it ca be used as a desig guide for applyig rage-free schemes i WSN systems. A descriptio of these parameters follows: Node Desity (ND): Average umber of odes per ode radio area. Achors Heard (AH): Average umber of Achors heard by a ode ad used durig estimatio. Achor to Node Rage Ratio (ANR): The average distace a achor beaco travels divided by the average distace a regular ode sigal travels. Achor Percetage (AP): The umber of achors divided by the total umber of odes (1000~3000 odes). This value ca be derived from the three parameters described above usig the formula: AP=AH/(AH+ND*ANR 2 ). Degree of Irregularity (DOI): DOI is defied as a idicator of radio patter irregularity. GPS Error: This parameter is defied as the maximum possible distace from the real achor positio to the GPS estimated achor positio iuits of ode radio rage (R). CONCLUSION Localizatio i wireless sesor etworks is a importat issue. Great efforts have bee made by may researchers ad a variat of algorithm also have bee proposed. I this paper, we proposed a ew classificatio for localizatio techiques. I this classificatio, localizatio algorithms were classified based o differet key features like Cetralized, Distributed, Achor Based, Achor Less, Rage Based ad Rage Free. This classificatio is useful for better uderstadig of the operatios of various localizatio methods ad workig as a aidig tool for the developmet of a ovel localizatio algorithm. Refereces 1. Hill J, Szewczyk R, A, Woo, Hollar S, Culler D, Pister K; System Architecture Directios for Networked Sesors, ASPLOS Nirupama B, Heidema J, Bychkovskiy V, Estri D; Desity-adaptive beaco placemet algorithms for localizatio i ad hoc wireless etworks. I IEEE Ifocom Stojmeovic, Li X; Loop-free Hybrid Sigle-path Floodig Routig Algorithms with Guarateed Delivery for Wireless Networks, IEEE Trasactios o Parallel ad Distributed Systems, 2001: 12(10): Bergamo P, Mazzii G; Localizatio i sesor etworks with fadig ad mobility. 13th IEEE Iteratioal Symposium o Persoal, Idoor ad Mobile Radio Commuicatios. 2002; 2:

6 5. Elahrawy E, Li X, Marti R; The limits of localizatio usig sigal stregth: a comparative study. First Aual IEEE Coferece o Sesor ad Ad-hoc Commuicatios ad Networks. 2004; Madiga D, Eiahrawy E, Marti R, Ju WH, Krisha P, Krishakumar A; Bayesia idoor positioig systems. IEEE INFOCOM. 2005; 2: Niculescu D, Nath B; Localized positioig i ad hoc etworks. IEEE Iteratioal Workshop o Sesor Network Protocols ad Applicatios. 2003: Patwari N, Hero A, Perkis M, Correal N, O Dea R; Relative locatio estimatio i wireless sesor etworks, IEEE Trasactios o Sigal Processig. 2003;51(8): Rappaport TS; Wireless Commuicatios: Priciples ad Practice, 2d ed. Pretice Hall PTR Bible S, Zyda M, Brutzma D; Usig Spread Spectrum Ragig Techiques for Positio Trackig i a Virtual Eviromet. Secod IEEE Workshop Networked Steggles P, Gschwid S; The Ubisese Smart Space Platform. I Proceedigs of the Third Iteratioal Coferece o Pervasive Computig Romer K.; The Lighthouse Locatio System for Smart Dust. Proceedigs of the 1st iteratioal coferece o Mobile systems, applicatios ad services. 2003; Bulusu N, Heidema J, Estri D; GPS-less low cost outdoor localizatio for very small devices. IEEE Persoal Commuicatios Magazie Nicolescu D, Nath B; Ad-Hoc Positioig Systems. I Proceedigs of IEEE GLOBECOM. 01 November He T, Huag C., Blum B, Stakovic JA, Abdelzaher T; Rage-Free localizatio schemes i large scale sesor etworks. I ACM Iteratioal Coferece o Mobile Computig ad Networkig (Mobicom) Nagpal R; Orgaizig a Global Coordiate System from Local Iformatio o a Amorphous Computer, A.I. Memo1666, MIT A.I. Laboratory Nagpal R, Shrobe H, Bachrach J; Orgaizig a Global Coordiate System from Local Iformatio o a Ad Hoc Sesor Network, I IPSN '03, Palo Alto Kleirock L, Slivester J; Optimum trasmissio radii for packet radio etworks or why six is a magic umber, I proceedigs of atioal Telecomm coferece. 1978;

A New Space-Repetition Code Based on One Bit Feedback Compared to Alamouti Space-Time Code

A New Space-Repetition Code Based on One Bit Feedback Compared to Alamouti Space-Time Code Proceedigs of the 4th WSEAS It. Coferece o Electromagetics, Wireless ad Optical Commuicatios, Veice, Italy, November 0-, 006 107 A New Space-Repetitio Code Based o Oe Bit Feedback Compared to Alamouti