Optimization of Base Station and Maximizing the Lifetime of Wireless Sensor Network

Transcription

1 Optimizatio of Base Statio ad Maximizig the Lifetime of Wireless Sesor Network P.Parthiba 1, G.Sudararaj 2, K.A.Jagadheesh 3, P.Maiiarasa 4 SS1 Research Scholar, P.S.G College of Techology, Coimbatore, Idia 2,3 Associate Professor, P.S.G College of Techology, Coimbatore, Idia 4 Professor,Nehru Istitute of Egieerig ad Techology, Coimbatore, Idia 1 parthibamephd@gmail.com, 2 gsraj_557@yahoo.com, 3 jagu_psgtech@yahoo.com, 4 maiiarasa@gmail.com Abstract Maximizig the Wireless Sesor Networks (WSNs) lifetime is oe of the most remarkable studies i the academic ad the idustrial research. I the existig Base statio locatios, the sesor etwork lifetime is short due to higher eergy cosumptio. Higher eergy cosumptio is due to the ooptimized locatio of base statio. No optimized locatio icreases the distace betwee the sesor odes to the base statio, so it will ot trasfer the data directly to the base statio. Each ode trasfer the data to the eighbor ode ad fially it reaches the base statio. Due to trasferrig of the data to the eighbor ode icreases the data rate step by step ad also the umber of sesor odes icreases the optimal locatio of base statio to become complex. So the mai objective is to fid optimal base statio locatio to reduce eergy cosumptio ad to maximize the etwork s lifetime i wireless sesor etworks ad to fid the optimal path for all sesor odes to the base statio to aalyze its lifetime. I this paper, a algorithm based o Gravity Locatio Model ad data routig based o particle swarm optimizatio (PSO) are thus proposed for efficiet eergy cosumptio ad data routig. Keyword-Wireless sesor etwork, Base statio, Gravity locatio, etwork lifetime, particle swarm Optimizatio. I. INTRODUCTION A wireless sesor etwork (WSN) cosists of wireless sesors odes that are distributed accordigly to their fuctioality, to moitor physical or evirometal coditios, such as temperature, soud, pressure, etc. These sesors pass their data through the etwork to a data ceter via base statios cosecutively. Durig the past decades, the developmet of wireless sesor etworks were motivated by military applicatios but today such etworks are used i may idustrial ad cosumer applicatios which icludes idustrial process moitorig ad cotrol, machie health moitorig, ad so forth. The WSN is made of odes, from a few to several hudred, where each ode is coected to oe (or sometimes several) sesors. Each such sesor etwork ode has typically several parts such as a radio trasceiver, a microcotroller, a electroic circuit for iterfacig with the sesors ad a eergy source, usually a battery eabled with a solar pael form of eergy harvestig. A sesor ode might vary i size, depedig o the fuctios. The cost of sesor odes is similarly variable, depedig o the complexity of the idividual sesor odes. A importat performace metric for wireless sesor etworks is the so-called etwork lifetime, which is highly depedet upo the physical topology of the etwork. This is because eergy expediture at a ode to trasmit data to aother ode ot oly depeds o the data bit rate, but also o the physical distace betwee these two odes. Cosequetly, it is importat to uderstad the impact of locatio related issues o etwork lifetime performace ad to optimize topology durig etwork deploymet stage. This article cosiders the importat base statio placemet problem for a give sesor etworks such that etwork lifetime ca be maximized. The followig problem has bee specifically cosidered ad aalyzed. Give a sesor etwork with each ode I producig sesig data at a rate of r, where should we place the base statio i this sesor etwork such that all the data ca be forwarded to the base statio such that the etwork lifetime is maximized. ISSN : Vol 6 No 2 Apr-May

2 Fig 1. A Wireless Sesor Network The Sectio II provides the related work. The Sectio III provides the existig system. Followed by the Gravity locatio model, it deals with the locatio of the optimum poit usig gravity locatio algorithm i the sectio IV. Sectio V provides the simulatio of the algorithm ad fidig the ew optimum locatio for the base statio. The eergy cosumed for data trasfer amog the sesor odes is also discussed i the previous sectio. The sectio VI deals with the Particle swarm optimizatio ad the optimized data routig. II. RELATED WORK The followig are the relevat literature pertaiig to the process placig base statio i a optimized locatio. Yi Shi et al [20] this paper deals with the approximatio algorithm that ca guaratee optimal etwork lifetime performace for base statio placemet problem. Cerulli.R et al [3] author explaied most power dissipatio occurs durig commuicatio, thus routig protocols i wireless sesor etworks (WSNs) maily aim at power coservatio. A routig protocol should be scalable, so that its effectiveess does ot degrade as the etwork size icreases. Kostatios Kalpakis et al [7] discussed the maximum lifetime data gatherig problem. Mig Yu et al [4-5] addressed how to dyamically orgaize these sesors ito wireless commuicatio etwork ad effectively route the iformatio amog sesors to a remote collectio statio. Mohamed Youis [13] proposed Reliable Eergy Aware Routig (REAR), which is a distributed, o-demad, reactive routig protocol that is iteded to provide a reliable trasmissio eviromet for data packet delivery. A eergy-aware QoS routig protocol for sesor etworks which ca also ru efficietly with best-effort traffic [8]. The problem ivolvig clusterig of wireless sesor odes ad coverage area is give i [9-11]. The optimized routig techiques i the wireless sesor etworks ad the lifetime maximizatio are discussed i the literatures i [12-20]. III. Existig System I the existig system the sesor odes are spatially distributed ad the base statio locatios were ot optimized. Due to this the eergy cosumptio was very high due to data trasmissio distace. This causes the degradatio i the lifetime of the sesor odes which i-tur causes the etire etwork to cosume more eergy. The followig table idicates the sesor ode locatio, data trasmissio rate ad the total eergy cosumed by each sesor ode i the existig system. ISSN : Vol 6 No 2 Apr-May

3 S.NO TABLE I Locatio Of The Sesor, Data Trasmissio Rate Ad The Eergy Cosumed LOCATION X Y DATA RATE(Kbps) ENERGY(J) ISSN : Vol 6 No 2 Apr-May

4 IV. Gravity Locatio Algorithm Gravity locatio model is oe of the supply chai maagemet techiques usig for locatig the ware house optimally. The same techique is used for locatig the base statio optimally. Data rate ad eergy is the mai factor ifluecig the locatio the base statio. The mai problem is 100 sesor odes are placed radomly i the sesible area 100m x 100m accordig to their fuctioality, all odes havig the differet load ad differet eergy level. So this work locates the base statio optimally to all odes. The gravity locatio model is very useful method to fid out the optimal locatio. I supply chai, cost ad load (how much quatity is trasferred) are cosidered, to locate the ware house. Here the eergy ad data rate is cosidered to locate the base statio optimally. Gravity models are used to fid locatios that miimize the eergy of trasmittig sesed data from sesor ode to the base statio. All distaces are calculated as the geometric distace betwee two poits o a plae. Gravity locatio model is used to fid the optimal locatio of base statio ad the total eergy spet for trasmittig data to base statio. Locatio, data rate (load) ad eergy is the iput data for the gravity locatio model. The obtaied data are plotted ad the locatio of the sesor odes is show i the figure 2. ISSN : Vol 6 No 2 Apr-May

5 Fig 2. Locatio of the sesor odes The locatio that miimizes the total eergy E j is obtaied by iteratig through the followig three steps where (x,y)is the locatio of the base statio to begi each iteratio. 1. statio, evaluate usig the formula d 2. Obtai a ew locatio (x, y ) for the base statio. X = D F X k =1 d D F k =1 d 4.1 Y = D F X k =1 d D F k =1 d If the ew locatio (x, y ) is almost the same as (x, y), stop. 4. Otherwise, set (x, y) = (x, y )ad go to 1 The placemet of the base statio should be optimized amog the dispersed locatio of the wireless sesor odes. The optimum locatio ca be obtaied by usig the gravity locatio model. Usig the gravity locatio model, the base statio locatio ca be idetified ad also the total eergy also ca be obtaied. A. The Total Eergy Calculatio Total eergy speds for trasmittig data from sesor ode to base statio. E j = k = 1 d D F 4.3 F - Eergy D - of trasmittig oe bit for oe meter betwee the sesor odes to base statio Load (data) to be trasmitted betwee sesor odes to base statio x, y - Coordiate locatio of either a sesor ode or base statio d Euclidea Distace If (x, y) is the locatio selected for the sesor ode, the distace d facility at locatio (x, y) d = ( x x)2 + ( y y)2 4.4 ISSN : Vol 6 No 2 Apr-May

6 I this method, the optimal locatio of base statio is obtaied by evaluatig the above formulae ad the collected data. S.No Eergy Spet (f) j Table 2 Calculated Data With Euclidea Distace Data rate (d) kbps X Coordiates Y Equlidia distace (d) m ISSN : Vol 6 No 2 Apr-May

7 X AXIS VALUE FOR BASESTATION LOCATION X = 4.71 Y AXIS VALUE FOR BASESTATION LOCATION Y = 4.15 TOTAL ENERGY CONSUMPTION= J B. The Optimal Base Statio Locatio There are 100 odes placed i 100m x100m, each ode is placed accordig to its x, y co ordiates value. Optimal base statio is placed betwee the 100 odes. It is deoted 101 th ode show i fig 3 (4.71, 4.15) represets the optimal locatio of base statio. Newly located base statio is the optimum locatio from each ode. ISSN : Vol 6 No 2 Apr-May

8 Fig 3. The Optimized Base Statio Locatio V. Particle Swarm Optimizatio The Particle swarm optimizatio is a computatioal method that optimizes a problem by iteratively tryig to improve the solutio. PSO optimizes a problem by havig a populatio of solutios, the particles are moved aroud i the search space with simple mathematical formulae over the particle's positio ad velocity. Each particle's movemet is iflueced by its local best kow positio. It is also guided toward the best kow positios i the search-space, which are updated as better positios are foud by other particles. This is expected to move the swarm toward the best solutios. The PSO algorithm works by havig a populatio, termed as a swarm of particles. The movemets of the particles are guided by their ow best kow positio i the searchspace as well as the etire swarm's best kow positio. Whe improved positios are beig discovered these will the come to guide the movemets of the swarm. The process is repeated ad by doig so it is hoped to result i a best solutio. Each particle keeps track of its coordiates i the solutio space which are associated with the best solutio (fitess) that has achieved so far by that particle. This value is called persoal best, pbest. Aother best value that is tracked by the PSO is the best value obtaied so far by ay particle i the eighborhood of that particle. This value is called gbest. The basic cocept of PSO lies i acceleratig each particle toward its pbest ad the gbest locatios, with a radom weighted acceleratio at each time step. I this work the weighted acceleratio represets the data rate. Fig 4. Particle swarm optimizatio velocity updatig ISSN : Vol 6 No 2 Apr-May

9 sk : curret searchig poit. sk+1 : modified searchig poit. Vk : curret velocity. Vk+1 : modified velocity. Vpbest : velocity based o pbest. Vgbest : velocity based o gbest Figure 4 represets how the velocity is modified from curret search space to ew search space. pbest ad gbest also idicated for ew positio. A.FORMULA FOR CALCULATING THE ENERGY E ij = (e t x r x d ) +(e r x r) 5.1 Eij - total eergy take for trasmittig data from i th ode to j th ode et - Eergy take for trasmittig the data from i th ode to j th ode i j/bit er - Eergy take for receivig the data from i th ode to j th ode i j/bit r - Data rate i Kbps. d - Euclidia distace i m d = ( x x)2 + ( y y) X,Y- Coordiate locatio of either i th sesor ode X, Y - Coordiate locatio of either jth sesor ode B. FORMULA FOR FITNESS CALCULATION Fitess = j= 1 Eij E ij - total eergy take for trasmittig data from i th ode to j th ode C.MODEL CALCULATION Eergy calculatio for 87 th ode Eij = (e t x r x d) +(e r x r) Possible eighbor odes are 48, 41, ad 25 Each ode takig for trasmittig oe bit to oe meter is 0.1 j/bit/m Each ode takig for receivig oe bit is 5 j/bit/m E 1 E = (0.1 x 8 x o.583) + (0 x5) = 466 j/bit/m E2 E87-41= (0.1 x 8 x.50) = 1200 j/bit/m 5.3 E 3 E = (0.1 x 8 x 1.923) = 1538 j/bit/m D. FITNESS CALCULATION Fitess = mi Fitess = mi (E 1, E 2,E 3 ) Eij j= ISSN : Vol 6 No 2 Apr-May

10 = mi (466, 1200, 1538) Fitess = 466 j/bit/m Node E 48 is selected for ext iteratio Further search is cotiued util the data reach the optimal base statio. Same steps are repeated for all odes, fially will get the shortest path from each ode to base statio. E. OPTIMAL ROUTING USING PARTICLE SWARM OPTIMIZATION The data routig is a importat parameter i maximizig the life time of the wireless sesor etwork. Here the Particle swarm optimizatio algorithm is employed to optimize the data routig from each sesor ode to the base statio, so that each ode seds the data to base statio through the miimum distace. These optimized data routes provide the efficiet cosumptio of eergy. The data routes are show i the Table 3. It represets the distace travelled i each optimal route from the source to destiatio ad the total data that are trasferred i each route, amout of eergy sped for trasmittig data from the source ode to the base statio. TABLE 3 Optimal Route for the data trasmissio S.No Routig Total data (kbps) Distace travelled (m) Eergy Spet(j/bit) ISSN : Vol 6 No 2 Apr-May

11 VI. Coclusio A wireless sesor etwork has bee cosidered ad the gravity locatio optimizatio ad particle swarm optimizatio are employed to miimize the Eergy cosumptio i this paper. The proposed approach ca search for early optimal Base Statio Locatios ad the optimized routig of data i the wireless sesor etworks, where each sesor odes may ow differet data trasmissio rates, iitial eergies ad parameter values. Experimets ad simulatios have also bee made to show the performace of the proposed approach ad the effects of the parameters o the results. This helps i aalyzig the proximity of the sesors which miimizes the eergy sped o the sesor ode to trasmit the data. As a result of the simulatios, it ca be cocluded that the proposed Gravity locatio algorithm ad the PSO algorithm coverges very fast ad also to reduce the data traffic amog the odes with the help of optimal routig, whe compared to the exhaustive search. This work ca be exteded to optimize wireless sesor etwork with multiple base statios. REFERENCES [1] Adre Rossi et al A exact approach for maximizig the lifetime of sesor etworks with adjustable sesig rages, Computers & Operatios Research,39,pp ,2012. [2] Arvid Sakar ad Zhe Liu Maximum Lifetime Routig i Wireless Ad-hoc Networks, IEEE joural, pp ,2012 [3] Cerulli.R et al, Exact ad heuristic methods to maximize etwork lifetime i wireless sesor etworks with adjustable sesig rages, Europea Joural of Operatioal Research, 220,pp 58 66,2012. [4] Chadresh Pratap Sigh, A Overview of Routig Protocols of Sesor Networks, CIMCA, pp , 2008 [5] Hui Qu, Co-desiged achor-free localizatio ad locatio-based routig algorithm for rapidly-deployed wireless sesor etworks, Iformatio Fusio 9, pp , [6] Kemal Akkaya, A survey o routig protocols for wireless sesor etworks, Ad Hoc Networks, 3, pp ,2005. [7] Kostatio Kalpakis et al, Efficiet algorithms for maximum lifetime data gatherig ad aggregatio i wireless sesor etworks, Computer Networks, 42, pp , [8] Li Hu et al, A New Eergy-Aware Routig Protocol for Wireless Sesor Networks, IEEE joural, , [9] Lifeg Yua et al, A eergy-efficiet real-time routig protocol for sesor etworks, Computer Commuicatios, 30, pp ,2007. [10] Luqu Li, Mile Zuo, A Dyamic Adaptive Routig Protocol for Heterogeeous Wireless Sesor Network, Networks Security, pp , 2007 [11] Marcel Busse, Eergy-efficiet forwardig i wireless sesor etworks, Pervasive ad Mobile Computig,4, pp 3 32,2008. [12] Mohammed Tarique et al, Miimum eergy hierarchical dyamic source routig for Mobile Ad Hoc Networks, Ad Hoc Networks, 7, pp , [13] Mohamed Youis et al, O Hadlig QoS Traffic i Wireless Sesor Networks, Hawaii Iteratioal Coferece o System Scieces pp 1-10, [14] Nikolaos A et al, Eergy efficiecy i wireless sesor etworks usig sleep mode TDMA schedulig, Ad Hoc Networks, 7, pp ,2009 [15] Pig Yua et al, Optimal Multicast Routig i Wireless Ad Hoc Sesor Networks, Joural of etworks, pp ,2009. [16] Qigchu Re, Eergy ad quality aware query processig i wireless sesor database systems, Iformatio Scieces, 7, pp ,2009. [17] Sajjad Zarifzadeh et al, Joit rage assigmet ad routig to coserve eergy i wireless ad hoc etworks, Computer Networks,53, pp ,2009. [18] Tai-Jug Chag et al, A color-theory-based eergy efficiet routig algorithm for mobile wireless sesor etworks, Computer Networks, 52, pp , [19] Weifa Liag et al, O-lie disjoit path routig for etwork capacity maximizatio i eergy-costraied ad hoc etworks, Ad Hoc Networks,5,pp ,2008. [20] Yi Shi ad Y. Thomas Hou, Approximatio Algorithm for Base Statio Placemet i Wireless Sesor Net works, IEEE joural, pp ,2007. ISSN : Vol 6 No 2 Apr-May