A Research of RSSI-AM Localization Algorithm Based on Data Encryption in Wireless Sensor Networks

Size: px

Start display at page:

Download "A Research of RSSI-AM Localization Algorithm Based on Data Encryption in Wireless Sensor Networks"

Jack Sharp
5 years ago
Views:

1 Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp Sesors & Trasucers 04 by IFSA Publishig, S. L. A Research of RSSI-AM Localizatio Algorithm Base o Data Ecryptio i Wireless Sesor Networks * Wag Wei, Wag Zhaoba School of Iformatio a Commuicatio Egieerig, North Uiversity of Chia, Taiyua, Shaxi, Chia * Tel.: , fax: * @qq.com Receive: 6 July 04 /Accepte: 8 July 04 /Publishe: 3 July 04 Abstract: I practical applicatio of wireless sesor etworks, because of ope eviromet, sigal is easy to be attacke a traitioal RSSI locatio techology prouces errors. By aalyzig the locatio moal of RSSI, this paper proposes a ew ecryptio moulatio algorithm: RSSI-AM, which is ulike most approaches. The locatio algorithm has the followig avatages: simple calculatio, strog security, powerful ati-iterferece ability a o harware expasio require. Besies, the simulatio experimet shows the locatio precisio of ragig metho base o RSSI-AM has obvious improvemet compare with traitioal algorithm. It ca be use i the eviromet of wireless sesor etwork oes with low cost a performace of harware. Copyright 04 IFSA Publishig, S. L. Keywors: Wireless sesor etworks, RSSI, Moulatio, Ecryptio, Ati-attack.. Itrouctio WSN, Wireless Sesor Network, is a hot techology at preset. It itegrates three fiels: calculatio, commuicatio a sesor, which becomes a bra-ew acquisitio a processig iformatio techology. It is geerally ackowlege to be curret a future research irectio a it will geerate a prologe ifluece o huma beigs lifestyle. WSN is a self-orgaizig wireless etwork compose of sesor oes with ata-collectig, ataproceeig a wireless commuicatio ability, whose purpose is to cooperatively perceive, collect a process iformatio which come from perceptio objects i etwork coverage itra-regio, the trasmits to mai cotrol ceter to execute system ecisio [-3]. As may techologies, like wireless commuicatio, microprocessor a micro electromechaical system, evelop icreasigly rapily, with low cost, self-orgaizatio, small size a flexible istributio, wireless sesor etwork techology graually becomes mature. Its aaptability is becomig wier a wier a has a broa prospect i military etectio, evirometal iformatio moitorig, space exploratio, agriculture prouctio, meical health moitorig, architecture a itelliget house, iustrial prouctio cotrol, traffic cotrol a commercial fiel [4-6]. I wireless sesor etwork, locatio iformatio of sesor oes is a emphasis to whole etwork s iformatio acquisitio a processig. Curretly, there have bee may systems a algorithms committe to researchig WSN locatio techology, but because WSN has may aaptabilities i ope eviromet, locatio system is liable to be iterfere by outer sigals, icluig outer attack [7-9]. Locatio algorithm of WSN ees a 84

2 Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp safety protectio i orer to provie a safe a effective etwork. I recet years, may researchers have propose some relevat algorithms, which have strog pertiet suggestio a maily emphasize certai specific attack. Oce attack type is chage, safety protectio will prouce failure. What s more, most of algorithms maily stuy how to prevet malicious attacks, however, i the practical a applie eviromet, there are other uavoiable iterferece-sigals. It is extremely ecessary to thik about prevetive methos [0-3].. Relate Work The locatio i WSN meas how a oe i etwork obtais its geographical locatio iformatio. Most of WSN aopt the istribute localizatio techology, whose priciple is as follows: Network cotais a small quatity of achor oes, which use self-carryig GPS locatio compoet to obtai locatio iformatio, the se referetial iformatio a establish cooriate system. Other ukow-oes measure or estimate locatio relatio betwee self oes a achor oes, icluig istace, agle a relatio of regio, the aopt these relatio a specific algorithm to calculate locatio of ukow oes. The methos of locatio algorithm ofte use have trilateratio, triagulatio a maximum likelihoo estimatio metho, etc. [4-6]. Accorig to the locatio seekig mechaism, the locatio metho of WSN ca be ivie ito two types: rage-base a rage-free. The former is use to measure istace betwee oes or agle iformatio, the it applies mathematics metho to calculate oes locatio, whose methos ofte utilize are base o TOA (time of arrival), TDOA (time ifferece of arrival), AOA (agle of arrival), RSSI (receive sigal stregth iicatio) a so o. The latter ee t istace a agle iformatio. It makes use of etwork coectivity a other iformatio to realize oe locatio, whose methos ofte use are cetroi algorithm, amorphous algorithm, DV-HOP, APIT a so o. These algorithms metioe above are i ieal eviromet to locate a o t cosier that outer factors have locatio ifluece o the system [7-]. 3. RSSI Algorithm Research Uer the situatio of kow trasmissio power, through receivig oe measuremet to receive power, RSSI uses receivig power loss of receivig oe to calculate trasmissio loss, ultimately, it turs trasmissio loss ito istace by applyig theory or practical sigal trasmissio moel. After obtaiig relevat istace iformatio betwee achor oes a ukow oes, aopt trilateratio or maximum likelihoo estimatio metho to get the locatio of ukow oes. I practical eviromet a aaptability process, especially i the situatio of ese-oes istributio, reflectio, multi-path trasmissio, atea gai a other factors ca geerate ifferet effect. Because the value of RSSI has symmetry, whe measurig the value of RSSI betwee oes, task ca be complete by seig a receivig a message package, but ot a routrip. Distace value betwee the value of RSSI a oes possesses mootoicity, which will ecrease with icrease of istace. Thus, RSSI ca satisfy locatio algorithm requiremet of rage-base a rage-free simultaeously. Whe oes have commuicatio, RSSI is the value of sigal stregth from receivig sigal of oppoets. So whe wireless sesor oes have ormal commuicatio, sigal ee t ay other equipmet a ca be irectly extracte from harware. I this way, it ca ecrease etwork cost a complexity. Let us aalyze the basic priciple of RSSI from the agle of mathematics. The mathematical relatio betwee trasmissio power of wireless sigal from sesor oes a its receivig power ca be expresse as follows i Formula (). The receivig power of wireless sigal is P R ; the trasmissio power is P T, the istace betwee trasceiver uits is r, the trasmissio factor is, the trasmissio eviromet of wireless sigal has great ifluece o the value of trasmissio factor. Sigal trasmissio theory moel is geerally use to aalyze raio wave trasmissio loss moel. The establishmet of sigals empirical moels is the matchig betwee locatio of each poit a atabase of sigal stregth. Because loss affects locatio precisio of RSSI value greatly, it is cautious to select suitable loss moel. I geeral, it is reasoable to choose free space trasmissio loss moel a log ormal istributio moel to execute calculatio. Free space trasmissio loss moel is as follows: PL ) = lg 0lg f () ( I the moel, whe 0 is meter, PL( 0 ) is pass loss of wireless trasmissio istace, is the istace loss factor, f is the sigal frequecy. Log ormal istributio moel is as follows: PL ) = PL( ) + 0lg( / ) + ξ () ( 0 0 I the moel, PL() is the pass loss of trasmissio istace,, PL( 0 ) is the pass loss, whe the trasmissio istace 0 = m, is the atteuatio coefficiet of sigal, the value is betwee a 5, ξ, is the zero-mea Gaussia raom istributio fuctio. The sigal stregth of locatio oes receivig referetial oes is: RSSI = Ps + Pa PL( ) (3) I (3), SSI is the power of receivig, Ps is the 85

3 Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp power of oe trasmissio, Pa is the atea gai, a PL() is the path loss. Formula () a () are itrouce ito Formula (3), the the result is RSSI= Ps + Pa [ lg0 + 0 lg f + 0 lg( /0) + ξ] = Ps + Pa 3.44 ξ 0 (lg + lg f ) (4) Settig a parameter: M=P s +P a ξ, Formula (4) is: RSSI = M -0lgf (5) I WSN, because RSSI sigal is trasmitte i space, it ot oly possesses a feature of trasmissio, but also it will be iterfere by other sigals. Aimig at the practical problem, we pose a solutio to solve the problem by improvig algorithm to reiforce ati-iterferece ability a resistace ability of malicious attack. I the process of WSN layout, calculatio of ifferece from multi RSSI sigals is RSSI - RSSI = ( M -0lg f )( - M -0lg f ) f = M - M + 0lg f Whe, i WSN oes, trasmissio frequecy, trasmissio power, atea gai a path loss gai are the same, the result is RSSI - RSSI lg (6) = 0 (7) y = Acoswt (8) A is the sigal amplitue; w is the agular frequecy. Aoptig RSSI ifferece value to moulate amplitue o carrier wave ca realize log-istace trasmissio, a the frequecy of carrie wave is chageable, i ope eviromet, it is ot liable to be attacke maliciously. The result of moulatio is: = (9) y 0Alg coswt, where is the correspoig oe umber i WSN. I the Formula (9), / is the iscrete value. I the practical applicatio, use a timer to make the sigal keep a amplitue value uchage i a carrier circle. Because the WSN is ofte set i ope eviromet, it is easily attacke a iterfere from outsies. Accorig to the basic priciple of RSSI, ukow oe is eee at least three achor oes of kow locatio iformatio to locate. Simulatio aopts MATLAB to establish a uifie simulatio eviromet i the rage of 50 m 50 m 50 m a achor oes are istribute i the rage. O the basis of three RSSI ifferece values, =.5 m, =.7 m, 3=.9 m, A= m, f=50 Hz, the Formula (9) is simulate with MATLAB, a the result is Fig.. I the same priciple, the relatio of RSSI ifferece value betwee two oes is also gotte by usig RSSI ifferece value of relevat oes to trasmit sigal locatio iformatio. 4. Ecryptio RSSI-AM Algorithm Research Base-ba moulatio sigal a carrier start to moulate at the begiig, the obtai moulate wave (also calle moulate sigal). Moulate sigal is trasmitte to receivig termial by chael after receivig termial emoulates a restores origial base-ba sigal. Demoulatio is reverse trasform of moulatio, which is a process of extractig moulate sigal from moulate wave. Moulatio a emoulatio sigal commuicatio system ot oly optimizes harware requiremet of receivig a trasmitter, but also ecrease iterferece a attack amage urig trasportatio process. I WSN, oes receive RSSI ifferece sigal to moulate so that it ca icrease the ability of ati-iferece a ati-attack of iformatio trasmissio system, usig Formula (7) to moulate. y is the carrier sigal: Fig.. Moulatio of three RSSI ifferece values. I Fig., if a achor oe is estroye, we have o way to obtai locatio iformatio of ukow oes. Accorig to referece [7], the system ca reach the better locatio effect by te achor oes. I Fig., the relatio betwee te achor oes a locatio oes ca help to locate, accorig to te groups of RSSI ifferece value:.5 m,.7 m,.9 m,.3 m,.5 m,.9 m, 3.3 m, 3.7 m, 3.8 m, 3.9 m. 86

Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp. 84-90 iteger multiple of i the secoary ciphertext, the ata of the seco, forth, sixth, eighth, teth, is shifte to get the thir ciphertext.

Uer the circumstace that a WSN system has bee a small applicatio scale, which meas the umber of achor oes is limite.

Base o the RSSI-AM, which is stuie i the paper, system ca possess security a ati-iterferece. Firstly, all sesors are set physically. The layout process ca be regular or at raom.

All oes ca act as the role of achor oes a help the system to establish a iter-relative oe locatio relatio.

4 Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp iteger multiple of i the secoary ciphertext, the ata of the seco, forth, sixth, eighth, teth, is shifte to get the thir ciphertext. Thir ciphertext: 8.7 m, 9. m, 0. m, 9.9 m, 3.7 m,.5 m, 6.5 m, 4. m, 8.0 m, 7.9 m. The st, a 3 r ecryptios are show i Fig.3-5. Fig.. Moulatio of te RSSI ifferece values. Uer the circumstace that a WSN system has bee a small applicatio scale, which meas the umber of achor oes is limite. If the system simply epes o the achor oes to locate ukow oes, it is easily estroye especially whe there are attacks from outsies. Base o the RSSI-AM, which is stuie i the paper, system ca possess security a ati-iterferece. Firstly, all sesors are set physically. The layout process ca be regular or at raom. The, accorig to RSSI relatio, sesor etwork locates all oes by achor oes. I this situatio, the system ca get locatio iformatio of all oes, which they are accurate. All oes ca act as the role of achor oes a help the system to establish a iter-relative oe locatio relatio. If oes are estroye or attacke by oppoets, or oes have itself performace fault, the system ca realize the accurate locatio by much more kow oes. Locatio iformatio is coverte ito ciphertext a forme ecryptig key, the at receivig termial, use ecryptio key to ietify locatio iformatio. Curret ecryptio algorithm maily aopts the symmetrical key algorithm a the osymmetry key, because the calculatio complexity of two algorithms is much higher a has better security. However, because it is limite for sesor equipmets to use available resources i WSN, we cosier aoptig a substitutio algorithm to ecrypt. Accorig to the ormal sequece to the arrage ata, make them correspo with 3, 4, 5, 6, 7, 8, 9, 0,, (ecryptio key is 3) a substitutio ecryptio is implemete o the 0 RSSI ifferece value sigals metioe above. Plaitext:.5 m,.7 m,.9 m,.3 m,.5 m,.9 m, 3.3 m, 3.7 m, 3.8 m, 3.9 m. First ciphertext: 3.7 m, 3.9 m, 3. m, 4.5 m, 4.7 m, 4. m, 5.5 m, 5.9 m, 5.0 m, 5. m. Next, a the first ciphertext with 5, 6, 7, 8, 9, 0,,, 3, 4 (ecryptio key is 5) correspoetly a get the secoary ciphertext. Secoary ciphertext: 8.7 m, 9.9 m, 0. m,.5 m, 3.7 m, 4. m, 6.5 m, 7.9 m, 8.0 m, 9. m. The, accorig to 5-3= ( is ecryptio key), Fig. 3. The First Ecryptio. Fig. 4. The Secoary Ecryptio. Fig. 5. The Thir Ecryptio. 87

5 Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp The amplitue a locatio sequece of the origial ata are all chage after three ecryptios, a attacker is ifficult to ecrypt the ata. Algorithm is ot complex a it satisfies harware limit requiremet of WSN, furthermore, it oes t occupy much too large calculatio resources, eergy cosumptio is small a security is strog. The receiver of the receivig termial emoulates moulatio sigals a the moulatio of sigals termial has receive are execute a aalysis of evelope etectio. After etectio, Fig. 6 is gotte. D 3 4 Fig. 7. Maximum likelihoo estimatio metho. The formula is: ( x x) + ( y y) = x x) + ( y y) = ( First formula eucts the last formula respectively, the result is x x x ( x x) x + y y ( y y) y = y x ( x x) x+ y ( y y) y = The formula rewrites to liear equatios: AX=b, Fig. 6. The result of evelope etectio. Decrypt the emoulatio sigals a its process is a reverse process of ecryptio. Accorig to ecryptio key 3 a 5, we implemet three iverse trasformatios o ciphertext a get RSSI ifferece value iformatio of seig termial. It is the plaitext:.5 m,.7 m,.9 m,.3 m,.5 m,.9 m, 3.3 m, 3.7 m, 3.8 m, 3.9 m. RSSI ifferece sigal iformatio, which receivig termial has receive, is processe accorig to a sychroous clock. E = RSSI - RSSI = 0Alg (0) =, () E 0 where E is the RSSI ifferece value oes have receive, is the istace betwee kow oes a target oe. By the Formula (), we ca get the correspoet istace. Usig the maximum likelihoo estimatio algorithm, we get locatio iformatio of target oes. From to N oe cooriates ca be represete as (x, y ), (x, y ), (x 3, y 3 ),, (x, y ), assume kowig:,, 3,,, which are the istace from oe D to oes. Suppose the cooriate of oe D as (x, y) (Fig. 7). x X = () y x x + y y + b = (3) x x + y y + x x + y y + b = (4) x x + y y + Usig the staar optimal mea the square estimatio ca get the cooriate of oe D: T T X = ( A A) A b. The feature of Maximum likelihoo estimatio is that it ca make full use of locatio iformatio that oes have collecte a its expressio way is brief so that may locatio algorithms ofte use it. 5. Compariso of Property I orer to compare the property betwee the RSSI-AM locatio oe algorithm a the traitioal RSSI locatio algorithm, we aopt MATLAB to establish a simulatio eviromet i the rage of 88

6 Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp m 50 m 50 m. Achor oes are evely istribute i the rage. Achor oes ca eliver itself iformatio perioically a each ukow oe ca receive iformatio from all achor oes. Simulatio aopts 00 results from ukow oes evely istribute i the measure rage to get arithmetic average as the fial result. I Fig. 8, whe the ratio of achor oe is less tha 6 %, the locatio precisio of RSSI-AM is obviously higher tha the RSSI algorithm. As the umber of achor oes icreases, the locatio precisio of two algorithms gets cotiual improvemet. Whe the ratio of achor oes is early approachig to 30 %, the precisio approaches to the costat value. 6. Coclusio a Future Work The paper proposes a ew locatio ecryptio moulatio algorithm:rssi-am base o the RSSI sesor etwork. The algorithm aopts the RSSI basic ragig priciple, space trasmissio loss moel a logormal istributio moel to euce the relatio of ukow oes a achor oes. I orer to ajust to complicate eviromet for livig a malicious attack i WSN, the algorithm ecrypts ata a uses the RSSI ifferece value to moulate carrie wave amplitue. I the receivig termial, get the RSSI by the evelope etectio a the ecryptio. The the maximum likelihoo estimatio ca help get locatio iformatio of oes. The algorithm has ot higher requiremets o harware a is utilize to ragig requiremet of most of WSN locatio. I the future research work, we will aim at the techology features of WSN, icrease the umber of etecte oes a o a large quatity of simulatio experimets over ifferet istributio moels i orer to verify locatio effect whe sesors are applie i large scale. Besies, we will make a lot of simulatio aalysis a fuctioalities compariso irectig at ifferet moulatio methos, master its applicatio fuctioalities a scope a o a further research of ata ecryptio algorithm which is suitable to WSN. Fig. 8. The precisio of locatio algorithm. Fig. 9 shows the commuicatio traffic compariso of RSSI-AM ragig algorithm a RSSI ragig algorithm. Although the RSSI-AM algorithm icreases some commuicatio expese, however, the expese is ot costly a oes t brig huge bure to commuicatio. Fig. 9. Time of locatio algorithm. Ackowlegemets This work was supporte by the atioal atural sciece fous of shaxi provice of Chia uer grat a i part by the atioal atural sciece fous of Chia uer grat Refereces []. Haowe Cha, Aria Perrig, Security a Privacy i Sesor Networks, IEEE Computer, 36, 0, 003, pp []. G. J. Pottie, W. J. Kaiser, Wireless itegrate etwork sesors, Commuicatios of the ACM, 43, 5, 000, pp [3]. L. F. Akyiliz, W. Su, Y. Sakarasubramaiam, et al., A Survey o Sesor Networks, IEEE Commuicatio Magazie, August 00, pp [4]. We-Lig P., Wag Li-Na, Builig Itrusio Tolerat Software System, Wuha Uiversity Joural of Natural Sciece,, 005, pp [5]. C. Karlofa, D. Wager, Secure routig i wireless sesor etworks: attack a coutermeasure, i Proceeigs of the First IEEE Iteratioal Workshop o Sesor Network Protocols a Applicatios, May 003, pp [6]. Woo D., Stakovic J. A., Deial of Service i Sesor Networks, Computer, 35, 0, 00, pp

Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp. 84-90 [7]. L. Escheauer, V. D.

, Desig guielies for maximizig lifetime a avoiig eergy holes i sesor etworks with uiform istributio a uiform reportig, i Proceeigs of the IEEE INFOCOM, Barceloa, Spai, 006, pp. -. [9].

7 Sesors & Trasucers, Vol. 75, Issue 7, July 04, pp [7]. L. Escheauer, V. D. Gligor, A key-maagemet scheme for istribute sesor etworks, i Proceeigs of the 9 th ACM Coferece o Computer a Commuicatios Security, 9, 00, pp [8]. Olariu S., Stojmeovic I., Desig guielies for maximizig lifetime a avoiig eergy holes i sesor etworks with uiform istributio a uiform reportig, i Proceeigs of the IEEE INFOCOM, Barceloa, Spai, 006, pp. -. [9]. Zhou Ya, Li Haicheg, Space localizatio algorithm base RSSI i wireless sesor etworks, Joural o Commuicatios, 6, 009, pp [0]. Zhag Y., Liu W., Fag Y., et al., Secure localizatio a autheticatio i ultra-wieba sesor etworks, IEEE Joural o Selecte Areas i Commuicatios, 4, 4, 006, pp []. Capku S., Cagalj M., Srivastava M., Secure Localizatio with Hie a Mobile Base Statios, i Proceeigs of the 5 th IEEE Aual Joit Coferece o Computer a Commuicatios Societies (INFOCOM 06), Barceloa, Spai, April 006, pp []. Ajum F., Paey S., Agrawal P., Secure localizatio i SN usig trasmissio rage variatio, i Proceeigs of the IEEE Iteratioal Coferece o Mobile A-hoc a Sesor System (MASS 05), Washigto, DC, November 005. [3]. Lazos L., Poovera R., SeRLoc: Secure rageiepeet localizatio for wireless sesor etworks, i Proceeigs of the ACM Workshop o Wireless Security (WISE 04), Brisbae, Australia, November 004, pp [4]. Lazos L., Poovera R., Capku S., ROPE: Robust positio estimatio i wireless sesor etworks, i Proceeigs of the Iteratioal Symposium o Iformatio Processig i Sesor Networks (IPSN 05), Los Ageles, CA, April 005, pp [5]. Lazos L., Poovera R., HiRLoc: High-resolutio robust localizatio for wireless sesor etworks, IEEE Joural o Selecte Areas i Commuicatios, 4,, 006, pp [6]. Ekici E., Vural S., McNair J., et al., Secure probabilistic locatio verificatio i raomly eploye wireless sesor etworks, A Hoc Networks, 007. [7]. Du W. L., Fag L., Nig P., LAD: Localizatio aomaly etectio for wireless sesor etworks, The Joural of Parallel a Distribute Computig, 66, 7, 006, pp [8]. Liu D. G., Nig P., Du W. L., Detectig malicious beaco oes for secure locatio iscovery i wireless sesor etworks, i Proceeigs of the 5 th Iteratioal Coferece o Distribute Computig System (ICDCS 05), Columbus, Ohio, Jue 005, pp [9]. Li Z., Trappe W., Zhag Y., et al., Robust statistical methos for securig wireless localizatio i sesor etworks, i Proceeigs of the Iteratioal Symposium o Iformatio Processig i Sesor Networks (IPSN 05), Washigto, 005, pp [0]. Liu D. G., Nig P., Du W. L., Attack-resistat locatio estimatio i sesor etworks, i Proceeigs of the Iteratioal Coferece o Iformatio Processig i Sesor Networks (IPSN 05), Los Ageles, CA, April 005, pp []. Wag C., Liu A., Nig P., Cluster-base miimum mea square estimatio for secure a resiliet localizatio i wireless sesor etworks, i Proceeigs of the Iteratioal Coferece o Wireless Algorithms, Systems a Applicatios (WASA 07), Chicago, IL, August 007. []. Bras S., Chaum D., Distace-bouig protocols, i Proceeigs of the Workshop o the Theory a Applicatio of Cryptographic Techiques o Avaces i Cryptology, New York, 994, pp Copyright, Iteratioal Frequecy Sesor Associatio (IFSA) Publishig, S. L. All rights reserve. ( 90

WSN Node Localization Regularization Algorithm Based on Quasi Optimal Criterion Parameter Selection

WSN Node Localization Regularization Algorithm Based on Quasi Optimal Criterion Parameter Selection Sesors & rasducers Vol. 23 Special Issue July 203 pp. 94-98 Sesors & rasducers 203 by IFSA http://www.sesorsportal.com WSN Node Localizatio Regularizatio Algorithm Based o Quasi Optimal Criterio Parameter