LIS: Localiza ion based on an in elligen dis ibu ed uzzy
sys em applied o a WSN
D.F. La ios , J. Ba bancho, F.J. Molina, C. León
Depa men o Elec onic Technology, Uni e si y o Se ille, Escuela Poli écnica Supe io , C/Vi gen de
Á ica S/N, Se ille, Spain
Keywo ds:
Fuzzy sys em
WSN
Localiza ion
RSSI
Cen oid
CL
abs ac
The localiza ion o he senso nodes is a undamen al p oblem in wi eless senso ne wo ks.
The e a e a lo o di e en kinds o solu ions in he li e a u e. Some o hem use ex e nal
de ices like GPS, while o he s use special ha dwa e o implici pa ame e s in wi eless
communica ions.
In applica ions like wildli e localiza ion in a na u al en i onmen , whe e he powe a ail-
able and he weigh a e big es ic ions, he use o hung y ene gy de ices like GPS o ha d-
wa e ha add ex a weigh like mobile di ec ional an enna is no a good solu ion.
Due o hese easons i would be be e o use he localiza ion’s implici cha ac e is ics in
communica ions, such as connec i i y, numbe o hops o RSSI. The measu emen ela ed
o hese pa ame e s a e cu en ly in eg a ed in mos adio de ices. These measu emen
echniques a e based on he beacons’ ansmissions be ween he de ices.
In he cu en s udy, a no el acking dis ibu ed me hod, called LIS, o localiza ion o
he senso nodes using mo ing de ices in a ne wo k o s a ic nodes, which ha e no addi-
ional ha dwa e equi emen s is p oposed.
The posi ion is ob ained wi h he combina ion o wo algo i hms; one based on a local
node using a uzzy sys em o ob ain a pa ial solu ion and he o he based on a cen alized
me hod which me ges all he pa ial solu ions. The cen alized algo i hm is based on he
calcula ion o he cen oid o he pa ial solu ions.
Ad an ages o using uzzy sys em e sus he classical Cen oid Localiza ion (CL)
algo i hm wi hou uzzy p ep ocessing a e compa ed wi h an ad hoc simula o made o
es ing localiza ion algo i hms.
Wi h his simula o , i is demons a ed ha he p oposed me hod ob ains less localiza-
ion e o s and be e accu acy han he cen oid algo i hm.
1. In oduc ion
A Wi eless Senso Ne wo k (WSN) consis s o a lo o
small de ices deployed in a physical en i onmen . E e y
de ice, called a node (Fig. 1), has special capabili ies, such
as communica ing wi h i s neighbo s, sensing and da a
s o age and p ocessing. The nodes can make a mesh ne -
wo k o de ices in such a way ha hey can collabo a e
amongs hemsel es can collabo a ing amongs hem-
sel es. These ea u es pe mi he implemen a ion o
dis ibu ed solu ions o sol e complex p oblems.
The main componen s o a node o WSN a e a mic ocon-
olle , a wi eless anscei e , a powe sou ce and Inpu
and Ou pu modules (I/O) o in e ac wi h he en i onmen
(Fig. 2).
The mic ocon olle p ocesses he da a and con ols he
unc ionali y o he o he componen s in he senso node.
Low powe mic ocon olle s a e ypically used in WSN.
Wi eless modules use ISM bands which allow ansmis-
sions o e some special license- ee equencies wi h
Co esponding au ho . Tel.: +34 954 55 28 38.
E-mail add esses: dfl[email p o ec ed] (D.F. La ios), [email p o ec ed] (J.
Ba bancho), [email p o ec ed] (F.J. Molina), [email p o ec ed] (C. León).
global a ailabili y (like he 173, 433, 868, 915 MHz and
2.4 GHz band). The unc ionali y o bo h he ansmi e
and he ecei e a e combined in o a single de ice known
as anscei e s. T anscei e s o WSN o en lack unique
iden ifie s and ha e low powe consump ion. Recen ly,
s anda d p o ocols, such as IEEE 802.15.4 [1] and ZigBee
[2] ha e specially de eloped o hese de ices.
The powe sou ce on a WSN is p o ided by ba e ies o
capaci o s [3]. Ba e ies can be echa geable and non-
echa geable depending on he applica ion. Sola panels
a e he ypical ene gy sou ces o he efill ba e ies.
Typically, a WSN is o med by e y small de ices wi h
se e al es ic ions: low powe consump ion, low weigh
(especially o mobile de ices), low cos , low da a s o age
and p ocessing and low adio ange.
Gene ally, amongs all he node componen s, he g ea -
es ene gy consump ion is dissipa ed in he adio ans-
cei e . As an example, he TelosB pla o m consumes
41 mW in ac i e mode. The mic ocon olle consumes only
5 mW and he emainde powe consump ion is caused by
he adio anscei e ha equi es 38 mW in he ecei e
mode and 35 mW in ansmission [4].
I is impo an o poin ou ha he powe consump ion
is e y high ei he in ansmission o in he ecep ion mode.
The e o e, o educe he powe consump ion i is necessa y
o educe he numbe o message exchanges, and also s op
all he node ac i i y enabling low powe modes by swi ch-
ing o he adio anscei e . The e o e, a sui able ac i i y
manage wi h hibe na ion pe iods is needed.
Se e al p o ocols o hibe na ion ha e been p oposed
[5–9] looking o a educ ion in he powe consump ion.
The main p oblem o hese p o ocols consis s in se ing
he clock synch oniza ion in e e y node; o he wise a node
can send a message when all i s neighbo s a e in idle s age
and losing in o ma ion.
WSN has been widely used in many a eas [10], such as
en i onmen al moni o ing [11] and con ol [12], heal h-
ca e and medical esea ch [13], na ional de ense and
mili a y a ai s [14,15], e c.
Fo all o hese applica ions which ha e mobile o
au onomous nodes, i is usually necessa y o implemen
a hibe na ion mechanism o sa e powe . Mo eo e , many
o hese applica ions equi e he posi ion knowledge o a
significan amoun o nodes.
In some o hese applica ions he in o ma ion ga he ed
om he nodes is no ele an wi hou he knowledge o
he associa ed posi ion, o example, desc ibing a sys em
o wildfi e acking based [16] on he de e mina ion o
specific si ua ions ( empe a u es, humidi y, wind di ec-
ion, e c.). The posi ion o he senso s ha desc ibes hese
si ua ions should be known.
In o he applica ions he in o ma ion equi ed is he po-
si ion i sel , o example, in ehicle acking [17].
Thus, i can be seen why localiza ion is one o he c u-
cial issues in wi eless senso ne wo ks esea ch. In a lo
o cases i is impossible o use specific localiza ion de ices,
like a GPS, because hese de ices ha e huge ene gy con-
sump ion and educe au onomy. In o he applica ions i
is necessa y o ha e nodes in he in e io o buildings,
whe e GPS echnology does no wo k co ec ly [18].
The main con ibu ion o he cu en s udy is he p e-
sen a ion o a no el localiza ion algo i hm, based on uzzy
logic p ocessing, and implemen ed o e small wi eless
senso nodes. As he de ices un a uzzy logic algo i hm
o p ocessing he in o ma ion, i was conside ed ha he
de ices we e p o ided wi h some kind o in elligence.
Thus, he sys em has been named as LIS, i.e. Localiza ion
based on In elligen Sys em.
The abo e men ioned localiza ion algo i hm is designed
specifically o acking small mobile de ices wi h e y low
weigh , he eby pe mi ing o ob ain he posi ion o small
animals. The e o e, no addi ional ha dwa e is used.
The p oposed p o ocol akes in o accoun he necessi y
o keep he powe consump ion low. In his sense, he p e-
sen ed algo i hm implemen s a hibe na ion p o ocol o
he non-ancho nodes ha can sa e powe ene gy on hese
de ices.
The es o he cu en s udy is o ganized as ollows:
Sec ion 2sums up he s a e o he a abou localiza ion.
Sec ion 3desc ibes LIS. The ou come o LIS pe o mance
is de eloped in Sec ion 4. Finally, Sec ion 5p esen s he
concluding ema ks and p o ides he discussion o u u e
s udies.
2. Localiza ion echniques
In localiza ion applica ions, he e a e wo ypes o
nodes:
Ancho Nodes: si ua ed on fixed and known posi ions.
Non-Ancho Nodes ( ags): nodes wi h unknown posi ion.
These nodes a e usually called ags.
Fig. 1. Example o a WSN node.
Fig. 2. A chi ec u e o a WSN node.
Fo many applica ion, i is impo an o conside ha
ancho nodes a e less ene gy cons ained han non-ancho
nodes. This is because in ancho nodes sys ems o echa ge
he ba e ies, such as sola panels, can be used. Fo many
applica ions (as in acking small animals) hese echa ge
sys ems can no be used in non-ancho nodes, due o he
cons ain s o maximum size and weigh .
Fig. 3 shows he ypical WSN used o localiza ion p o-
poses. The essence o localiza ion consis s in ob aining he
posi ion o he non-ancho nodes, using he in o ma ion
p o ided by he ancho nodes.
The localiza ion esea ch o e WSN is ocused on he
ollowing poin s: de e mining a new localiza ion algo-
i hm, modi ying he p oposed solu ion in o de o
imp o e accu acy o es ing classical me hods [19]. Tes ing
can be ocused on specific applica ions, such as indoo uses
[20,21], o cen e ed on specific cha ac e is ics, such as en-
e gy e ficiency [22].
Localiza ion algo i hms p esen ed in he li e a u e can
be classified in o wo ca ego ies, as gi en below:
2.1. Range-based echniques
These echniques es ima e, poin - o-poin , he dis ance
be ween each pai o nodes.
Wi h his in o ma ion and using echniques, such as
mul ila e a ion [23], iangula ion o o he me hods, he
absolu e posi ion o he non-ancho nodes can be es i-
ma ed. The mos common ones a e Recei ed Signal
S eng h Indica ion (RSSI) [24], Time O A i al (TOA)
[25], Time Di e ence O A i al (TDOA) [26] and Angle O
A i al (AOA) [27,28].
Nowadays, new algo i hms con inue o appea based on
hese classical me hods o he imp o emen o accu acy,
o example, based on AOA [29–33], based on TOA
[34,35] and based on TDOA [36,37].
O he au ho s p opose new ange based algo i hms, like
Su [38] ha uses a likelihood calcula ion o de e mining
he dis ance.
In gene al, he ange-based ones o e good accu acy,
bu addi ional ha dwa e is o en needed. The e o e, he
weigh , he cos and he powe consump ion o he node
de ices inc ease and make hese echniques unsui able
o he p oposed applica ion. RSSI ange-based echniques
a e an excep ion o his because mos o he cu en ans-
cei e s p o ide his measu emen by de aul . Howe e ,
RSSI echniques a e e y sensi i e o noise and in e e -
ences. The p oblem o he RSSI measu emen is desc ibed
in Sec ion 4.1.
These RSSI echniques equi e calib a ion and a model
o he en i onmen . Mo eo e , he calib a ion could
change acco ding o he en i onmen al condi ions. Nowa-
days, he de e mina ion o how o imp o e he accu acy o
he RSSI echniques is an impo an esea ch a ea [39–41].
In o de o a oid he RSSI p oblems, he p oposed me h-
od uses a uzzy RSSI model, ins ead o a ma hema ical
model, as desc ibed in Sec ion 3.
2.2. Range- ee echniques
In he ange- ee localiza ion algo i hms he posi ion o
he non-ancho nodes is ob ained om he beacons ex-
change among nodes (ancho and non-ancho nodes).
Beacons can con ain di e en in o ma ion, such as:
Radio co e age membe ship: An ancho node de ec s
whe he a non-ancho node is in i s adio co e age.
Using his in o ma ion, he sys em can es ima e he
non-ancho posi ion as a unc ion o he in e sec ion
o he co e age a eas o e e y ancho node ha a e in
i s adio co e age. Fig. 4a shows an example o he
in e sec ion o he co e age a eas be ween he wo
ancho nodes.
Numbe o hops o an ancho -node: I he e is no connec-
i i y wi h an ancho node, a non-ancho node can es i-
ma e i s posi ion knowing he numbe o hops o e e y
ancho node. An example o ha is ep esen ed in
Fig. 4b. Node B is a a dis ance o wo hops o ancho
node one, h ee hops o ancho node wo and wo hops
Fig. 3. Example o WSN o localiza ion.
(a) (b)
Fig. 4. Range- ee echniques: (a) Radio co e age membe ship. (b)
Numbe o hops.
o ancho node h ee. Acco ding o his in o ma ion and
conside ing he mean dis ance be ween he nodes, he
absolu e posi ion can be calcula ed applying algo-
i hms, such as iangula ion.
The mos common ange- ee echniques a e Cen oid
(CL) [42], DV-Hop [43], Con ex [44], APIT [45], e c.
CL is based on he es ima ion o he posi ion o a non-
ancho node by he cen oid o he posi ion o all o i s
ancho node neighbo s. Fig. 5 ep esen s an example
whe e a beacon o non-ancho node is ecei ed by ancho
nodes 3, 4 and 6, he eby ob aining he es ima ed posi ion
as he cen oid o he coo dina es o hese nodes.
Nowadays, modifica ions o he CL algo i hm a e an
impo an a ea o esea ch. Many au ho s con inue p opos-
ing modifica ions ha o e be e accu acy by weigh ob-
ained wi h RSSI [46] o wi h he Link Quali y Indica o
(LQI) [47]. Some app oaches ha e ocused on he educ ion
o ene gy consump ion. Behnke [48], p oposed a modified
CL algo i hm wi hou he use o complex ma hema ical
ope a ions o low esou ces mic ocon olle like he
squa e oo . O he app oaches ha e s udied he specifica-
ions o he CL algo i hm in o de o de e mine he a eas
whe e i o e s bigge e o s [49].
New ange- ee algo i hms con inue o appea in he li -
e a u e [50]. Some o hem a e ocused on de e mining he
new aspec s o he ne wo ks ha can gi e in o ma ion
abou he localiza ion, such as he in e sec ions o a simpli-
fied co e age a ea [51], o modi ying he powe ansmis-
sion o he beacons in o de o ob ain he smalles a ea
whe e i is mo e p obable o loca e he non-ancho node
[52]. O he s a e ocused on imp o ing any kind o a cha ac-
e is ic, such as ene gy e ficiency [53,54].
Many ange- ee algo i hms, such as cen oid and mos
o i s a ia ions, a e ully cen alized algo i hms, i.e. all he
in o ma ion, use ul o no , is sen o a Base S a ion, whe e
he posi ion o he non-ancho nodes is ob ained. This ea-
u e cons i u es a disad an age, because i equi es he
deli e y o a la ge amoun o messages o he Base S a ion,
which would was e ene gy. This is especially impo an in
wide ne wo ks.
Agains his ocus, he es o he au ho s p opose he
use o ully dis ibu ed algo i hms, ypically execu ed o e
he non-ancho nodes. Howe e i cons i u es a disad an-
age because he non-ancho nodes mus main ain he
adio anscei e du ing long pe iods o ime. These
de ices ha e he highes ene gy cons ain in he ne wo k.
In o he wo ds, i is impo an o emembe ha o his
acking applica ion, he Base S a ion needs o know he
posi ion o he non-ancho nodes. Using a ully dis ibu ed
algo i hm equi es he deli e y o he addi ional messages
om he non-ancho nodes o he Base S a ion, he eby
losing a pa o he ad an ages o dis ibu ed algo i hms.
The p oposed algo i hm ac s in a di e en manne . I
uses wo algo i hms looking o wa d o ob ain a pa o
he ad an ages o hese wo ocuses: a dis ibu ed
algo i hm, bu only execu ed o e he ancho nodes and
a cen alized algo i hm o ob ain he final posi ion by he
Base S a ion.
3. LIS algo i hm
Despi e he ac ha he ange- ee and ange-based
echniques ha e been ex ensi ely s udied, nowadays he e
a e some aspec s ha con inue o be a challenge:
The use o addi ional ha dwa e o lo s o beacons o
inc ease powe consump ion.
Fully cen alized p ocessing (i.e. on Base S a ions)
equi es a la ge amoun o messages. Con e sely, p o-
cessing in he ags’ nodes educes he ba e y o hese
de ices significan ly.
Scalabili y. Many localiza ion algo i hms a e ha d o
ex end o big senso ne wo ks.
LIS has been especially designed o phase ou all o he
abo e men ioned p oblems. As a esul , he p oposed algo-
i hm is scalable and he powe consump ion and ne wo k
au onomy a e op imized. As usual in a acking sys em, he
non-ancho nodes in LIS do no need o know hei loca-
ion. In his case, only he Base S a ion wan s o know i .
The p oposed localiza ion algo i hm is a ange- ee
acking echnique ha uses RSSI. Howe e , ins ead o
using a ma hema ical model, LIS p oposes a uzzy-logic-
based sys em o de i e he dis ance om he RSSI le el.
Fuzzy logic is a o m o mul i- alued logic; i deals wi h
easoning ha is app oxima e a he han fixed and exac .
To do his, Fuzzy logic uses a iables ha ha e a u h
alue in a [0–1] ange, commonly called membe ship de-
g ee. A uzzy sys em o e s many ad an ages, such as i s
obus ness in noisy and complex scena ios.
The use o compu a ional in elligence in localiza ion is
no a no el idea, as could be seen in p e ious wo ks, such
as [55] ha uses p obabilis ic neu onal ne wo ks, [56] ha
applies a uzzy sys em and [57] ha uses uzzy neu ons.
Fig. 5. CL algo i hm.
In gene al, all hese a e dis ibu ed algo i hms ha exe-
cu e mos pa o he localiza ion algo i hm o e he non-
ancho nodes. Howe e , none o hem conside he p ob-
lem o powe consump ion in he non-ancho nodes. Mo e-
o e , he algo i hms wi h Compu a ional In elligence
gene ally ack down he cu en posi ions based on he
es ima ed posi ion changes, needing an ini ializa ion o
he non-ancho nodes. These sys ems ail i he ags ( he
animal) go ou o he co e age o he WSN, and e u n in o
i a e a while. As desc ibed in Sec ion 5, he au ho s a e
cu en ly conside ing a me hod o imp o e he accu acy
based on hese es ima ion changes, bu wi hou he neces-
si y o an ini ializa ion.
3.1. Ne wo k p ocessing s age
LIS combines: (1) a uzzy sys em o es ima e (ac ually
o quali y) he dis ance be ween he ansmi e and he
ecei e om RSSI measu es, (2) a dis ibu ed algo i hm
execu ed in ecei e ancho nodes in o de o de e mine
he ela i e posi ions o hem and fil e useless in o ma-
ion, and (3) a cen alized algo i hm o de i e he mos
likely loca ion unning a he Base S a ion.
LIS consis s o ou s ages:
S1: Ancho nodes wai o non-ancho node beacons.
S2: The ag node b oadcas s a beacon.
S3: Recei e ancho nodes measu e RSSI, and execu e
bo h, he uzzifica ion algo i hm and he ubiqui ous
p ocessing o ela i e and pa ial posi ioning.
S4: Ancho nodes send pa ial solu ions o he Base
S a ion, whe e he loca ion is finally de e mined.
Fig. 6 illus a es hese s ages. When a non-ancho node
b oadcas s a beacon o any o he so o message, he local-
iza ion p ocess s a s (Fig. 6a). Jus a ha momen he
ecei e ancho nodes pa icipa e in he p ocess. The es
o he nodes can swi ch o he adio anscei e o hold
i in a low powe s a e.
The beacon equency can be de e mined by he appli-
ca ion, and can be modified by ex e nal condi ions, such
as he emaining ba e y in he non-ancho node, a dia y
schedule o wi h in o ma ion egis e ed by ex e nal senso ,
such as an accele ome e . In his way, he ba e y au on-
omy can be imp o ed.
3.1.1. Dis ibu ed p ocessing
LIS uses he measu es o RSSI o a non-ancho node e-
cei ed by an ancho node and i s neighbo s ancho s o
de e mine he a ea whe e he non-ancho node could be
loca ed. This algo i hm is based on a uzzy sys em dis ib-
u ed on e e y ancho node o he ne wo k (Fig. 7).
Acco ding o he algo i hm s ages, once an ancho node
ecei es a beacon, i es ima es he posi ion o he non-an-
(a) (b)
(c) (d)
Fig. 6. S ages o LIS algo i hm. (a) S2. (b) S3.A – uzzyfica ion. (c) S3.B – Pa ial posi ioning (d) S4.

cho nodes. The localiza ion algo i hm has been designed
o dis ibu e he compu a ion consump ion o e he ne -
wo k. The a ea whe e he non-ancho node could be local-
ized wi h a ce ain p obabili y is called he Rep esen a i e
A ea. A ‘‘sec o ’’ is he minimum a ea o med by h ee an-
cho -node neighbo s. A Rep esen a i e A ea can be made
up o one o mo e sec o s. Ancho nodes mus execu e
he dis ibu ed uzzifica ion algo i hm o e e y su ound-
ing sec o . Fig. 8 shows an example wi h fi e sec o s, in
which, he uzzy algo i hm is execu ed fi e imes.
E e y ancho node ha ecei es a beacon measu es and
b oadcas s he RSSI le el o i s ancho neighbo s (Fig. 6b).
In his way, he closes ancho nodes elabo a e a able wi h
he RSSI measu ed by hemsel es and hei ancho
neighbo s.
The RSSI able is p ocessed by he uzzy sys em o e al-
ua e he Rep esen a i e A ea, i espec i e o he numbe o
sec o s. This a ea can be o med by he union o one o
mo e sec o s (Fig. 6c). A sec o is conside ed as a pa o
he Rep esen a i e A ea i i s membe ship deg ee (i.e. he
ou pu o he uzzy sys em) is highe han a h eshold. This
alue is adjus ed expe imen ally. The cu en simula ions
show ha a h eshold o 0.1 manages a good ade-o
be ween he noise immuni y and localiza ion pe o mance.
The esul s o he Rep esen a i e A eas a e sen om he
ancho nodes o he Base S a ion o compu e he final solu-
ion (Fig. 6d).
A Rep esen a i e A ea is emp y i i does no con ain
significan sec o , i.e. i he membe ship deg ee o all o
hem is lowe han he h eshold. In his case, o sa e en-
e gy, he esul is disca ded and he algo i hm will finish
un il he nex beacon a i es. This is especially impo an
in huge ne wo ks, whe e he ene gy needed o mul i-
hop ansmissions is high and is a disad an age o he
cen alized localiza ion algo i hm.
3.1.2. The uzzy sys em inpu s
RSSI ables ep esen he signal le el ecei ed in ei he
he local o he neighbo ing ancho nodes. These RSSI mea-
su emen s a e used as he inpu s o he uzzy sys em. In
hese so o sys ems, i is necessa y o define se e al uzzy
se s. E e y uzzy se is desc ibed as a con inuous ma he-
ma ical unc ion ha ep esen s how an inpu signal
belongs o his se . This is called ‘‘membe ship deg ee’’ in
he li e a u e. This deg ee is ep esen ed in ange [0,1],
whe e 0 ep esen s 0% o membe ship and 1 ep esen s
100% o membe ship o his se . Fo example, Fig. 9 shows
he membe ship deg ee o he HIGH uzzy se , conside ing
he RSSI alue as he inpu signal.
In ou sys em, h ee uzzy se s quali y he RSSI as HIGH,
MEDIUM and LOW o each inpu (Fig. 10).
Fig. 7. In e ence uzzy sys em.
Fig. 8. Example o node wi h 5 neighbo s.
Fig. 9. HIGH uzzy se .
The LOW RSSI uzzy se is ep esen ed by a apezoid.
The maximum membe ship deg ee ( alue 1) is assigned
i he powe alls bellow he sensibili y h eshold o he
emi e node (RSSI
sens
). As he powe inc eases, he mem-
be ship deg ee dec eases linea ly un il i eaches ze o a
medium RSSI alue (RSSI
med
).
The MEDIUM RSSI uzzy se is ep esen ed by a iangle
whe e he maximum membe ship deg ee co esponds o
he medium RSSI alue (RSSI
med
). The ze o membe ship is
eached o he powe RSSI alues lowe han he sensibil-
i y h eshold o close o he maximum ansmission (RSSI
TX
powe
). In he cu en s udy, he medium RSSI alue mus be
compu ed o e e y sec o using he F iis model equa ion
and assuming he emi e ag is loca ed a he cen e . This
compu a ion only needs o be execu ed once because he
ancho nodes a e loca ed a fixed posi ions.
The use o F iis is a ade-o be ween he accu acy and
he in o ma ion equi ed. Mo e ealis ic models equi e
ha ing mo e ini ial in o ma ion o he en i onmen , a p i-
o i unknown, such as he posi ion o he obs acles. The F iis
app oxima ion simplifies he p oblems o he saw- oo h o
he dis u bances wi h a smoo h unc ion. Acco ding o a
eal scena io, his assump ion could no be a good app ox-
ima ion o he eali y. Howe e , he e o s assumed wi h
his app oxima ion a e compensa ed wi h he noise immu-
ni y o LIS, which assumes he dis u bances as noise. The
au ho s conside ha he use o an in si u calib a ion, mea-
su ing he eal ecei ed alue in he cen e o he sec o
would imp o e he accu acy, bu his alue would change
i he en i onmen changes (i.e. i i is aining o i he g ass
g ows). As men ioned in Sec ion 5, he au ho s hink ha
measu ing he RSSI be ween he ancho nodes, using he
b oadcas messages o LIS (i does no inc ease he o al
numbe o messages), would be a be e op ion, because
he calib a ion would be upda ed a un ime i he en i on-
men changes. Cu en ly he au ho s a e e alua ing his
op ion.
Fuzzy se o HIGH RSSI alues is a apezoid wi h a lin-
eal inc easing om 0 o 1 o he RSSI powe alues anging
be ween RSSI
med
and RSSI
TX powe
.
Fo example, le ’s conside a sys em based on a adio
ecei e wi h a sensibili y o 95 dBm, a TX powe ans-
mission o 0 dBm and a medium powe o 60 dBm. I
he inpu has a RSSI o 77.5 dBm, he membe ship deg ee
is 0, 0.5 and 0.5 o each se (HIGH, MEDIUM and LOW). I
he inpu has a RSSI alue o 30 dBm, we ob ain a mem-
be ship deg ee o 0.5, 0.5 and 0 o each uzzy se .
3.1.3. The uzzy sys em ou pu s
The Fuzzy Sys em o e s an ou pu o each and e e y
sec o . The ou pu associa ed o a sec o is a [0,1] anged
alue ha ep esen s he confidence deg ee (i.e. he u h
alue ob ained as ou pu o he ules o he uzzy sys em)
in which he ag is ac ually loca ed in ha sec o .
As Fig. 11 shows, he LOW ou pu uzzy se is a iangle
wi h he cen al poin a ze o and he co ne s a 0.5 and
0.5.
The MEDIUM ou pu is ep esen ed by a iangle wi h
he cen al poin a 0.5 and co ne s a 0 and 1.
The HIGH ou pu qualifie is also defined by a iangle
wi h he cen al poin a 1 and he co ne s a 0.5 and 1.5.
3.1.4. In e ence engine
The in e ence engine is he Mandani’s ules, wi h a cen-
oid de uzzifica ion me hod and a single on inpu uzzifi-
ca o . The uzzy engine e alua es he an eceden o e e y
ule by he in e sec ion o he uzzy inpu s, using he min-
imum unc ion o he AND ope a o , and he maximum
unc ion o he OR ope a o . In he p ocess o ob aining
he ou pu se s as a unc ion o he inpu se s is called
implica ion. In ou sys em we use he minimum unc ion
as he implica ion p ocess. The implica ion be ween he
inpu s and ou pu s applies he minimum unc ion.
As men ioned, he ules mus be e alua ed o e e y
single sec o o es ima e he confidence deg ee, aking in o
accoun he uzzy qualifica ions o RSSI alues o ei he he
cu en sec o nodes o he su ounding ones. The ules
summed up in Table 1 ha e been de i ed om mul iple
simula ions in o de o ob ain he bes ade-o be ween
p ecision and noise immuni y.
I is impo an o conside ha in he applica ion de-
sc ibed, i is necessa y o know he localiza ion in o ma-
ion o he nodes in he Base S a ion, whe e i can be
accessible o i s use. Due o i , ob aining he localiza ion
in o ma ion o e he non-ancho nodes, as gene ally de-
sc ibed on ypical ully dis ibu ed localiza ion algo i hms,
is no in e es ing. This is because hese de ices do no use
he abo e men ioned in o ma ion. Mo eo e , ob aining
his in o ma ion could inc ease ba e y consump ion o
he non-ancho nodes. In ac , hey a e he de ices wi h
highe ene gy cons ain s.
LIS akes his issue in o accoun . Also, he ancho nodes
ha e mo e powe supply esou ces han he ags. Conse-
quen ly, he algo i hm has been designed o be execu ed
mainly in he ancho nodes. Fu he mo e, he adio ans-
cei e o he ag is ac i a ed o a sho ime, which is jus
Fig. 10. Se s o he uzzy inpu s.
Fig. 11. Se s o he uzzy ou pu .
enough o b oadcas he beacon. In he emaining pe iod o
ime, he ag will be in an idle s a e and i s adio ans-
cei e would be o .
LIS also educes he powe consump ion in he ancho
nodes. I implemen s a ubiqui ous and dis ibu ed algo-
i hm ha sp eads he localiza ion p ocessing amongs
he nodes su ounding he ag. In a cen alized-only
algo i hm, all he in o ma ion ecei ed by he ancho
nodes mus be deli e ed o he Base S a ion. By con as ,
he p oposed algo i hm sa es he powe ene gy because
only significan in o ma ion is deli e ed.
Addi ional sa ings can be managed by clus e ing he
ne wo ks, and using he clus e heads as he Base S a ions,
i.e. ecei ing he p ocessing pa ial es ima ions om i s
clus e nodes.
3.1.5. Cen alized p ocessing s age
The Base S a ion collec s he pa ial solu ions om he
ancho nodes, and p ocesses hem cyclically as ollows:
C1: The Base S a ion wai s o ecei ing he fi s pa ial
solu ion.
C2: On a i al, he pa ial solu ion is sa ed and a ime
s a s unning.
C3: While he ime is unning, he nex pa ial solu ions
a e sa ed in a able as hey we e ecei ed.
C4: When he ime expi es, he sys em will compu e
he final posi ion as he cen oid o all hese pa ial
solu ions ( iangle sec o s). The cen oid compu a-
ion o a fini e se o poin s P
1
!;P
2
!;... P
N
!can be
simplified as:
Posi ion
!¼P
N
i¼1
P
i
!
Nð1Þ
Wi h his algo i hm, he Base S a ion needs o always be
ac i e, while wai ing o ecei e any message om he
nodes. I de e mines whe he he Base S a ion is going o
consume mo e han he o he nodes. This is no c i ical be-
cause he Base S a ion ac s as a ga eway be ween he sen-
so ne wo k and he ex e nal ne wo k. Thus, i needs o be
placed in an a ea wi hou powe es ic ions.
The p e ious algo i hm can be easily ex ended o loca -
ing mul iple ags, by simply associa ing a ag iden ifie o
he ansmi ed beacons. The final es ima ed posi ion is
ime s amped and sa ed in he Base S a ion o make i
accessible h oughou he In e ne .
4. Expe imen al esul s
The localiza ion sys em was designed o be de eloped
in a Wi eless Senso Ne wo k, called ICARO, in ended o
es an en i onmen al moni o iza ion algo i hm. ICARO
was placed in he Doñana Biological Rese e.
The Doñana Biological Rese e is a pa o he Doñana
na u al pa k, a pa k and wildli e e uge in sou hwes e n
Spain. I is loca ed in Andalusia, in he p o inces o Huel a,
Se illa and Cadiz, and co e s 543 km
2
, o which 135 km
2
a e a p o ec ed a ea. The pa k is an a ea o ma shes, shal-
low s eams, and sand dunes. Doñana Biological S a ion
(DBS) is a Resea ch Ins i u e o he Spanish Council o Sci-
en ific Resea ch (CSIC). Some o i s main goals a e conse -
a ion and imp o ing he quali y a esea ch in he Doñana
Biological Rese e (DBR), which was decla ed humani y
pa imony by UNESCO in 1994 and conside ed as one o
he mos impo an na u al p o ec ed landscapes in he
wo ld. In ac , his yea , DBR was included inside he g ea
scien ific in as uc u es o Eu opean Union. In addi ion, in
Ap il 2006, he Spanish In e minis e ial Comission o Sci-
ence and Technology (CICYT) om he Spanish Science
and Educa ion Depa men ecognized he Doñana Scien-
ific Rese e as a Singula Scien ific and Technological
In as uc u e (ICTS).
ICARO consis s o en wi eless senso s deployed be-
ween he zone o ‘‘El Ojillo’’ and ‘‘El Zacallon’’, wo king
on 2.4 GHz ISM band wid h IEE 802.15.4 P o ocol. Fig. 12
shows a map wi h he cu en deploymen o he sys em.
This ne wo k is based on wo kinds o de ices gi en
below:
Base s a ion: This is he de ice ha ac s as a ga eway
be ween he emo e measu emen senso and he
communica ion in as uc u e o Doñana. I pe mi s
he collec ion o in o ma ion and allows cen alized
p ocessing and da a usion. This de ice is shown in
Fig. 13. This sys em is based on a wi eless senso
a ached o an indus ial PC.
Remo e measu emen Senso s: These a e he de ices ha
pe mi he acquisi ion o en i onmen al in o ma ion.
They a e powe ed by sola panels. These de ices pe mi
o execu e dis ibu ed and collabo a i e algo i hms.
These de ices pe mi he u iliza ion o da a usion and
agg ega ion o educing he usage o he bandwid h.
These de ices a e shown in Fig. 14.
Table 1
Rules o he in e ence engine.
Cen al RSSI
node
RSSI neighbo s Ou pu
HIGH All MEDIUM HIGH
LOW All LOW LOW
MEDIUM All MEDIUM HIGH
MEDIUM All LOW LOW
HIGH All HIGH MEDIUM
MEDIUM MEDIUM in cu en sec o HIGH
LOW in he es
MEDIUM HIGH in any sec o excep he cu en
one
LOW
LOW in he es
HIGH High in a neighbo o he cu en sec o MEDIUM
LOW in he es
HIGH HIGH in a neighbo , excep on he
cu en sec o
LOW
LOW in he es
MEDIUM MEDIUM in a neighbo o he cu en
sec o
MEDIUM
LOW in he es
MEDIUM MEDIUM in a neighbo , excep on he
cu en sec o
LOW
LOW in he es
The cu en goal consis s o he use o he ICARO in a-
s uc u e as he ancho nodes, de eloping a low weigh
and low cos mobile de ice, wi h high au onomy as a
non-ancho node o acking animals in he Doñana Na u-
al Pa k, such as he de ice shown on Fig. 15. To do his, he
Base s a ion can be used o execu e he cen alized algo-
i hm o he localiza ion, while he emo e measu emen
senso execu es he dis ibu ed algo i hm as he ancho
nodes.
4.1. Rela ionship be ween RSSI and dis ance
Fig. 16 shows he expe imen s o e alua e he ela ion-
ship be ween RSSI and he dis ance in di e en si ua ions:
ee-space wi hou obs acles and long dis ance wi h
obs acles.
These g aphics a e ob ained by inc easing he dis ance
be ween a node ac ing as an emi e and ano he node
ac ing as a ecei e , wai ing o he ime o s abiliza ion
o he sys em and acqui ing a minimum o 50 alues pe
poin ep esen ed on he g aphic. These da a shows he
medium alue, ying o educe he e ec o he noise in
o de o ob ain he ela ionship be ween dis ance (in
[5–50 m] ange) and RSSI (in dBm).
The measu ed s anda d de ia ion has been app oxi-
ma ely abou 3 dBm. I was de ec ed ha he measu e-
men fluc ua ions a e highe in sho dis ances ha in a
o dis ances.
The ‘‘ ee space’’ esul s a e measu ed in he pa k, in a
desse zone, wi hou ees o g ass in he a ea ha could
p oduce dis u bances. The esul s wi h ‘‘obs acles’’ a e
he measu emen s in a co ido be ween he ees wi h
a ound 15 m o heigh , whe e he e a e small plan s and
g ass.
The esul s wi h obs acles do no ma ch wi h any alid
ma hema ical model ha could di ec ly pe mi o ob ain
he dis ance, using only he RSSI.
Ob aining models wi h he ela ionship be ween RSSI
and dis ance is cu en ly an impo an a ea o esea ch.
In any case, i is possible o ob ain his model wi h a much
s udied en i onmen . Fig. 17 shows an example de eloped
by Texas Ins umen s [58], whe e he RSSI alue was
Fig. 12. Deploymen o he p ojec ICARO.
Fig. 13. Base s a ion o he p ojec ICARO.
Fig. 14. Remo e measu emen senso o he p ojec ICARO.
Fig. 15. Example o non-ancho node o acking animals.
ma sh zones o he na u al pa k o Doñana. Cu en ly, he
au ho s a e s udying he inc ease in he numbe o ancho
nodes, o ob ain a eliable benchma k o localiza ion
algo i hms.
5. Conclusions and u u e wo k
LIS is a new uzzy algo i hm o localiza ion designed o
educe powe consump ion, especially bu no limi ed o,
he ag nodes whe e he powe cons ain s a e highe . LIS
fil e s he useless in o ma ion a e being p ocessed in
he ancho nodes. I also implemen s a hibe na ion
mechanism. All hese mechanisms inc ease he ba e y
au onomy.
LIS has been es ed by simula ions and in a eal sce-
na io. The ob ained esul s showed ha he p oposed
me hod ob ains less localiza ion e o s han he CL
algo i hm wi hou highe compu a ion equi emen s o
an ex ensi e use o adio.
The localiza ion sys em LIS is being applied o loca ing
and acking o wild animals in na u al pa ks.
Cu en ly, he aim is o inc ease he accu acy o he sys-
em e sus he noise, ac ing along he ollowing lines:
Fig. 25. Numbe o e o bigge han 100 m.

Using a fil e o educe he a iabili y o he measu e-
men : The cu en s udy a emp s o wo k on he e al-
ua ion o he ad an ages o using fil e s o RSSI
measu es o inc easing he accu acy.
Use addi ional in o ma ion o RSSI be ween nodes: Cu -
en adio de ices can p o ide RSSI in o ma ion o
e e y message, which is mo e han wha he adio can
in e cep . Fo example, he adio message sende
be ween he ancho nodes can be used in he localiza-
ion o model he en i onmen wi hou adding a fic
be ween he de ices.
Use Tempo al in o ma ion: The cu en p oposed
me hod only uses he in o ma ion ob ained in a p ede-
e mined ins an ; i does no ha e pas memo y. The
cu en s udy a emp s o wo k on a modifica ion o
he cu en algo i hm, which imp o es i s pe o mance
using his in o ma ion.
Acknowledgmen s
This esea ch has been suppo ed by he ‘‘Conseje ı
´ade
Inno ación, Ciencia y Emp esa’’, ‘‘Jun a de Andalucı
´a’’,
Spain, h ough he excellence p ojec ARTICA ( e e ence
numbe : P07-TIC-02476) and by he ‘‘Cá ed a de Tele ó-
nica, In eligencia en la Red’’, Se ille, Spain, h ough he
P ojec ICARO.
The au ho s hank he Biological S a ion o he na u al
pa k o ‘‘Doñana’’ and he esea che s o i s Biological S a-
ion Cen e, o hei collabo a ion and suppo .
Re e ences
[1] J.A. Gu ie ez, M. Nae e, E. Callaway, M. Bou geois, V. Mi e , B. Heile,
IEEE 802.15.4: a de eloping s anda d o low-powe low-cos
wi eless pe sonal a ea ne wo ks, IEEE Ne wo k 15 (2001) 12–19.
[2] P. Ba on i, P. Pillai, V.W.C. Chook, S. Chessa, A. Go a, Y.F. Hu,
Wi eless senso ne wo ks: a su ey on he s a e o he a and he
802.15.4 and ZigBee s anda ds, Compu . Commun. 30 (2007)
1655–1695.
[3] Z. Hu, B. Li, On he undamen al capaci y and li e ime limi s o
ene gy-cons ained wi eless senso ne wo ks, in: P oc. IEEE Real
Time Embedded Technol. Symp., ol. 10, 2004, pp. 2–9.
[4] J. Polas e, R. Szewczyk, D. Culle , Telos: enabling ul a-low powe
wi eless esea ch, in: Fou h In e na ional Symposium on
In o ma ion P ocessing in Senso Ne wo ks, P oc. IPSN, 2005, pp.
364–369.
[5] W. Ye, J. Heidemann, D. Es in, An ene gy-e ficien MAC p o ocol o
wi eless senso ne wo ks, P oc. IEEE INFOCOM 3 (2002) 1567–1576.
[6] V. Raghuna han, C. Schu ge s, S. Pa k, M.B. S i as a a, Ene gy-awa e
wi eless mic osenso ne wo ks, IEEE Signal P ocess. Mag. 19 (2002)
40–50.
[7] B. Chen, K. Jamieson, H. Balak ishnan, R. Mo is, Span: an ene gy-
e ficien coo dina ion algo i hm o opology main enance in ad hoc
wi eless ne wo ks, in: P oc. Annu. In . Con . Mobile Compu .
Ne wo king, 2001, pp. 85–96.
[8] G. Sudha Anil Kuma , G. Manima an, Z. Wang, Ene gy-awa e
scheduling wi h p obabilis ic deadline cons ain s in wi eless
ne wo ks, Ad Hoc Ne wo ks 7 (2009) 1400–1413.
[9] Y. Wu, S. Fahmy, N.B. Sh o , Sleep/wake scheduling o mul i-hop
senso ne wo ks: non-con exi y and app oxima ion algo i hm, Ad
Hoc Ne wo ks 8 (2010) 681–693.
[10] I.F. Akyildiz, W. Su, Y. Sanka asub amaniam, E. Cayi ci, Wi eless
senso ne wo ks: a su ey, Compu . Ne wo ks 38 (2002) 393–
422.
[11] J. Yick, B. Mukhe jee, D. Ghosal, Wi eless senso ne wo k su ey,
Compu . Ne wo ks 52 (2008) 2292–2330.
[12] J.A. López Riquelme, F. So o, J. Sua dı
´az, P. Sánchez, A. Ibo a, J.A.
Ve a, Wi eless senso ne wo ks o p ecision ho icul u e in
Sou he n Spain, Compu . Elec on. Ag ic. 68 (2009) 25–35.
[13] W. Chung, Y. Lee, S. Jung, A wi eless senso ne wo k compa ible
wea able U-heal hca e moni o ing sys em using in eg a ed ECG,
accele ome e and SpO2, in: P oc. Annu. In . Con . IEEE Eng. Med.
Biol. Soc., EMBC – Pe s. Heal hc. h ough Technol., 2008, pp. 1529–
1532.
[14] T. He, S. K ishnamu hy, J.A. S anko ic, T. Abdelzahe , L. Luo, R.
S ole u, e al., Ene gy-e ficien su eillance sys em using wi eless
senso ne wo ks, in: MobiSys Second In . Con . Mobile Sys . Appl.
Se ., 2004, pp. 270–283.
[15] A. Bouke che, H.A.B. F Oli ei a, E.F. Nakamu a, A.A.F. Lou ei o, Secu e
localiza ion algo i hms o wi eless senso ne wo ks, IEEE Commun.
Mag. 46 (2008) 96–101.
[16] T. An oine-San oni, J. San ucci, E. de Gen ili, X. Sil ani, F. Mo andini,
Pe o mance o a p o ec ed wi eless senso ne wo k in a fi e, Anal.
Fi e Sp ead Da a T ans., Senso s 9 (2009) 5878–5893.
[17] M. Tubaisha , Z. Peng, Q. Qi, S. Yi, Wi eless senso ne wo ks in
in elligen anspo a ion sys ems, Wi eless Commun. Mobile
Compu . 9 (2009) 287–302.
[18] W. Ren, A apid acquisi ion algo i hm o WSN-aided GPS loca ion, in:
P oc. In . Symp. In ell. In . Technol. Secu . In o ma ics, IITSI., 2009,
pp. 42–46.
Fig. 26. (a) Mesh ep esen a ion o he links in he ICARO ne wo k, and (b) de ailed a ea.
[19] E. Elnah awy, X. Li, R.P. Ma in, The limi s o localiza ion using signal
s eng h: a compa a i e s udy, in: Fi s Annu. IEEE Commun. Soc.
Con . Senso Ad Hoc Commun. Ne w. IEEE SECON., 2004, pp. 406–
414.
[20] G. Zanca, F. Zo zi, A. Zanella, M. Zo zi, Expe imen al compa ison o
RSSI-based localiza ion algo i hms o indoo wi eless senso
ne wo ks, in: REALWSN – P oc. Wo kshop Real-Wo ld Wi el.
Senso Ne w., 2008, pp. 1–5.
[21] S.A. Mi ilineos, J.N. Gou as, O.E. Segou, S.C.A. Thomopoulos, WAX-
ROOM: an indoo WSN-based localiza ion pla o m, in: P oc. SPIE
In . Soc. Op . Eng., ol. 7697, 2010.
[22] F. Reichenbach, J. Blumen hal, D. Timme mann, Compa ing he
e ficiency o localiza ion algo i hms wi h he Powe -E o -P oduc
(PEP), in: P oc. In . Con . Dis ib. Compu . Sys ., 2008, pp. 150–155.
[23] E. Cassano, F. Flo io, F. De Rango, S. Ma ano, A pe o mance
compa ison be ween ROC-RSSI and ila e a ion localiza ion
echniques o WPAN senso ne wo ks in a eal ou doo es bed,
Wi el. Telecommun. Symp. WTS. (2009).
[24] A. Awad, T. F unzke, F. D essle , Adap i e dis ance es ima ion and
localiza ion in WSN using RSSI measu es, in: P oc. – Eu omic o Con .
Digi . Sys . Des. A chi ., Me hods Tools, DSD., 2007, pp. 471–478.
[25] S. Wu, N. Zhang, Two-s ep TOA es ima ion me hod o UWB based
wi eless senso ne wo ks, Ruan Jian Xue Bao. 18 (2007) 1164–1172.
[26] S. Xiaoyan, L. Jiandong, H. Pengyu, P. Jiyong, To al leas -squa es
solu ion o ac i e a ge localiza ion using TDOA and FDOA
measu emen s in WSN, in: P oc. In . Con . Ad . In . Ne w. Appl.
AINA., 2008, pp. 995–999.
[27] D. Niculescu, B. Na h, Ad hoc posi ioning sys em (APS) using AOA,
P oc. IEEE INFOCOM 3 (2003) 1734–1743.
[28] P. Rong, M.L. Sichi iu, Angle o a i al localiza ion o wi eless senso
ne wo ks, senso and ad hoc communica ions and ne wo ks, 2006,
in: SECON ’06. 2006 3 d Annual IEEE Communica ions Socie y on.,
ol. 1, 2006, pp. 374–382.
[29] S. Wang, K. Shih, C. Chang, Dis ibu ed di ec ion-based localiza ion
in wi eless senso ne wo ks, Compu . Commun. 30 (2007) 1424–
1439.
[30] K. Kucuk, A. Ka ak, H. Yigi , C. Ozdemi , A no el localiza ion
echnique o wi eless senso ne wo ks using adap i e an enna
a ays, IEEE Radio Wi el. Symp., RWS. (2008) 483–486.
[31] D. Niculescu, B. Na h, Localized posi ioning in ad hoc ne wo ks, Ad
Hoc Ne wo ks 1 (2003) 247–259.
[32] D. Niculescu, B. Na h, Posi ion and o ien a ion in ad hoc ne wo ks,
Ad Hoc Ne wo ks 2 (2004) 133–151.
[33] K. Kucuk, A. Ka ak, Scalable loca ion es ima ion using sma
an ennas in wi eless senso ne wo ks, Ad Hoc Ne wo ks 8 (2010)
889–903.
[34] T. Sa hyan, M. Hedley, Coope a i e localiza ion – algo i hms o
ini ial posi ion es ima ion, in: ISSNIP – P oc. In . Con . In elligen
Sens., Sens. Ne w. In . P ocess., 2008, pp. 13–18.
[35] A. Cheh i, P. Fo ie , P.M. Ta di , UWB-based senso ne wo ks o
localiza ion in mining en i onmen s, Ad Hoc Ne wo ks 7 (2009)
987–1000.
[36] R. Eickho , F. Ellinge , R. Mosshamme , R. Weigel, A. Zi o , M.
Hueme , 3D-accu acy imp o emen s o TDoA based wi eless local
posi ioning sys ems, IEEE Globecom Wo kshops, GLOBECOM (2008).
[37] J. Hwang, T. He, Y. Kim, Secu e localiza ion wi h phan om node
de ec ion, Ad Hoc Ne wo ks 6 (2008) 1031–1050.
[38] F. Su, W. Ren, H. Jin, Localiza ion algo i hm based on di e ence
es ima ion o wi eless senso ne wo ks, in: P oc. In . Con .
Commun. So w. Ne w., ICCSN., 2009, pp. 499–503.
[39] H. Miu a, K. Hi ano, N. Ma suda, H. Taki, N. Abe, S. Ho i, Indoo
localiza ion o mobile node based on RSSI, Lec . No es Compu . Sci.
4694 LNAI (2007) 1065–1072.
[40] S. Tian, X. Zhang, P. Liu, P. Sun, X. Wang, A RSSI-based DV-hop
algo i hm o wi eless senso ne wo ks, In . Con . Wi el Commun.
Ne wo king Mob. Compu . (2007) 2555–2558.
[41] Z. Shan, T.-P. Yum, P ecise localiza ion wi h sma an ennas in ad-
hoc ne wo ks, GLOBECOM IEEE Global Telecommun. Con . (2007)
1053–1057.
[42] N. Bulusu, J. Heidemann, D. Es in, GPS-less low-cos ou doo
localiza ion o e y small de ices, IEEE Pe s. Commun. 7 (2000)
28–34.
[43] G.Q. Gao, L. Lei, An imp o ed node localiza ion algo i hm based on
DV-HOP in WSN, P oc. IEEE In . Con . Ad . Compu . Con ol, ICACC. 4
(2010) 321–324.
[44] L. Dohe y, K.S.J. Pis e , L. El Ghaoui, Con ex posi ion es ima ion in
wi eless senso ne wo ks, P oc. IEEE INFOCOM 3 (2001) 1655–1663.
[45] Y. Zhou, X. Ao, S. Xia, An imp o ed APIT node sel -localiza ion
algo i hm in WSN, in: P oc. Wo ld Cong . In elligen Con ol Au om.
WCICA., 2008, pp. 7576–7581.
[46] J. Blumen hal, R. G ossmann, F. Gola owski, D. Timme mann,
Weigh ed cen oid localiza ion in Zigbee-based senso ne wo ks,
IEEE In . Symp. In elligen Signal P ocess., WISP. (2007).
[47] S. Schuhmann, K. He mann, K. Ro he mel, J. Blumen hal, D.
Timme mann, Imp o ed weigh ed cen oid localiza ion in sma
ubiqui ous en i onmen s, Lec . No es Compu . Sci. 5061 LNCS (2008)
20–34.
[48] R. Behnke, D. Timme mann, AWCL: Adap i e weigh ed cen oid
localiza ion as an e ficien imp o emen o coa se g ained
localiza ion, in: Wo kshop Posi ioning, Na ig. Commun., WPNC.,
2008, pp. 243–250.
[49] R. Behnke, J. Salzmann, R. G obßmann, D. Lieck eld , D.
Timme mann, K. Thu ow, S a egies o o e come bo de a ea
e ec s o coa se g ained localiza ion, in: P oc. – Wo kshop
Posi ioning, Na ig. Commun., WPNC., 2009, pp. 95–102.
[50] H. Lee, J.L. Welch, N.H. Vaidya, Loca ion acking using quo ums in
mobile ad hoc ne wo ks, Ad Hoc Ne wo ks 1 (2003) 371–381.
[51] H. Chu, R. Jan, A GPS-less, ou doo , sel -posi ioning me hod o
wi eless senso ne wo ks, Ad Hoc Ne wo ks 5 (2007) 547–557.
[52] C. Wang, K. Liu, N. Xiao, A ange ee localiza ion algo i hm based on
es ic ed-a ea o wi eless senso ne wo ks, in: P oc. – In . Mul i-
Con . Compu . Global In . Technol., ICCGI Conjunc ion ComP2P: In .
Wo kshop Compu . P2P Ne w.: Theo y P ac., 2008, pp. 97–101.
[53] C. You, P. Huang, H. Chu, Y. Chen, J. Chiang, S. Lau, Impac o senso -
enhanced mobili y p edic ion on he design o ene gy-e ficien
localiza ion, Ad Hoc Ne wo ks 6 (2008) 1221–1237.
[54] Y. Zou, K. Chak aba y, Ta ge localiza ion based on ene gy
conside a ions in dis ibu ed senso ne wo ks, Ad Hoc Ne wo ks 1
(2003) 261–272.
[55] S. Rajaee, S.M.T. Almoda esi, M.H. Sadeghi, M. Aghabozo gi, Ene gy
e ficien localiza ion in wi eless ad-hoc senso ne wo ks using
p obabilis ic neu al ne wo k and independen componen analysis,
In . Symp. Telecommun., IST. (2008) 365–370.
[56] F. Xiu ang, G. Zhanqiang, Y. Mian, X. Shibo, Fuzzy dis ance measu ing
based on RSSI in wi eless senso ne wo k, in: P oc. In . Con . In ell.
Sys . Knowl. Eng., ISKE., 2008, pp. 395–400.
[57] S. Chiang, J. Wang, Localiza ion in wi eless senso ne wo ks by
uzzy
logic sys em, Lec . No es Compu . Sci. 5712 LNAI (2009)
721–728.
[58] T.I. K aks ud, Range Measu emen s in an Open Field En i onmen ,
Design No e DN018 – SWRA169A, Texas In umen . (2008) 1–14.
D.