INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 10 | NUMBER: 4 | 2012 | SPECIAL ISSUE
© 2012 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 211
TOA NODE DISTANCE ESTIMATION ENHANCEMENT IN MANET
LOCALIZATION ALGORITHM BASED ON COOPERATIVE
TRILATERATION
Vladimi CIPOV1, Lubomi DOBOS1, Jan PAPAJ1
1Depa men o Elec onics and Mul imedia Communica ions, Facul y o Elec ical Enginee ing and In o ma ics,
Technical Uni e si y o Kosice, Le na 9, 042 00 Kosice, Slo ak Republic
ladimi .cipo @ uke.sk, lubomi .dobos@ uke.sk, jan.papa[email p o ec ed]
Abs ac . This pape p o ides a b ie su ey o he Time
o A i al (ToA) anging me hod o en used in he Range-
based localiza ion o node dis ance es ima ion. I is
ocused especially on he Non Line-o -Sigh (NLoS)
channel iden i ica ion – one o he dominan ac o s ha
nega i e a ec he loca ion accu acy. I deals wi h he
imp o emen o he ecei ed signal a i al ime
de e mina ion in he case o NLoS communica ion wi h
unde ec ed di ec pa h. The low complexi y NLoS
mi iga ion me hod is p oposed in his pape ha
es ima es he ime o a i al o he missing di ec pa h as
a mean o he ime delays o e lec ed pa hs. This me hod
is implemen ed in he coope a i e posi ioning algo i hm
based on he ci cula ila e a ion. The IEEE 802.15.4a
s a is ical channel model is used o simula e he mobile
node communica ion.
Keywo ds
LoS/NLoS iden i ica ion, NLoS mi iga ion,
posi ioning, ToA, ila e a ion.
1. In oduc ion
In he ield o mobile communica ions, he wi eless nodes
loca ion p oblem has gained inc easing in e es . Loca ion
es ima ion is a complex p oblem. I has been in es iga ed
ex ensi ely in he li e a u e in he las ew yea s, whe e
au ho s sol ed a ious p oblems. I is gene ally ue ha
he posi ion de e mina ion p oblem consis s o wo s eps.
A i s , localiza ion pa ame e s mus be es ima ed. In
acco d wi h his ac posi ioning sys ems a e
di e en ia ed o he Range-based and Range- ee [13].
Finaly, ob ained pa ame e s a e used o posi ion
calcula ion ia one o he localiza ion algo i hms, which
a e classi ied as Ancho -based o Ancho - ee [12]. In his
pape , we concen a e on Range-based localiza ion. A
a ie y o me hods, which a e based on Recei ed Signal
S engh (RSS), Time o A i al (ToA) o Angle o A i al
(AoA), [2], a e a ailable o pe o m he ange
measu emen s. We p e e he ToA anging echnique in
combina ion wi h Ul a Wide-Band (UWB)
communica ion echnology, which is one o he mos
p omising app oaches o he mobile localiza ion and has
a g ea po en ial o accu a e anging especially in he
indoo en i onmen . Due o a e y wide bandwid h o he
UWB sys ems, hese a e capable o esol e indi idual
mul ipa h componen s in he ecei ed channel impulse
esponse and o de e mine he ime o hei a i al e y
p ecisely. Ka eh Pahla an and his eam o esea che s
engage in he a ea o ToA es ima ion. The p oblem o
ToA es ima ion in he case o unde ec ed di ec pa h and
i s iden i ica ion can be ound in [6], [14]. The impac on
he node dis ance es ima ion accu acy using a di e en
sys em bandwid h is also analyzed. Mo e ela ed wo ks
which deal wi h ToA can be ound in sec ion 2. We ha e
p oposed he low complexi y me hod o ToA node
dis ance es ima ion imp o emen and u ilized i in he
p oposed coope a i e posi ioning algo i hm based on
ci cula ila e a ion.
2. ToA Node Dis ance Es ima ion
and LoS/NLoS Iden i ica ion
Time o A i al is he echnique whe e he node dis ance
can be es ima ed as he p oduc o he measu ed signal
p opaga ion delay and he speed o ligh [1]:
cd 
, (1)
whe e d ep esen s he node dis ance, c he speed o ligh
and he es ima ed ime o a i al o he ecei ed signal.
Le he Channel Impulse Response (CIR) o he
ecei ed signal be ep esen ed as [5]:
   



L
lll A h
1

, (2)
INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 10 | NUMBER: 4 | 2012 | SPECIAL ISSUE
212 © 2012 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING
whe e L ep esen s he numbe o mul ipa h componen s,
Al and τl a e he signal ampli ude and he ime delay o
he l- h mul ipa h componen , espec i ely. In gene al, he
ime o a i al o he i s a i ing pa h, he i s CIR
componen , is conside ed as he ime o a i al o he
ecei ed signal. The p oblem occu s when he di ec pa h
is missing in he channel impulse esponse because i is
blocked o a enua ed unde he de ec ion h eshold o he
ecei e . Wi eless communica ion channel p o ile can be
classi ied o [6]:
 Line-o -Sigh (LoS) wi h Dominan -Di ec -Pa h
(DDP). In his case, he di ec pa h is he s onges
and he i s de ec ed pa h in he channel p o ile
and i can be de ec ed by he ecei e . The e a e no
conside able obs acles be ween he ansmi e and
he ecei e .
 Non Line-o -Sigh (NLoS) wi h Non Dominan -
Di ec -Pa h (NDDP). Line-o -Sigh communica -
ion be ween a ansmi e and a ecei e does no
exis . In NDDP channel p o ile, he di ec pa h is
no he s onges pa h o he CIR, howe e , i is
s ill de ec able because i is ecei ed abo e he
de ec ion h eshold o he ecei e .
 Non Line-o -Sigh (NLoS) wi h Unde ec ed-
Di ec -Pa h (UDP). In UDP channels, he di ec
pa h goes below he de ec ion h eshold o he
ecei e while o he pa hs a e s ill de ec able. The
ecei e assumes he i s de ec ed peak as a di ec
pa h which causes conside able dis ance
es ima ion e o s.
The pe o mance o he Range-based localiza ion
sys ems based on ToA node dis ance es ima ion
echniques mainly depends on he p esence o Line-o -
Sigh communica ion be ween he ansmi e - ecei e
pai s. NLoS adio p opaga ion is one o he mos
dominan ac o s ha a ec he posi ioning accu acy [3].
The absence o a di ec pa h in he ecei ed signal esul s
in an ex a-delay o he ToA because he signal a els an
ex a dis ance ha leads o he inco ec loca ion [1].
Se e al LoS/NLoS iden i ica ion me hods ha e been
p oposed [3], [5]. One o he mos basic LoS/NLoS
iden i ica ion me hods a e based on he p io knowledge
o s a is ics o he signal ecei ed in LoS condi ions.
Du ing he eal- ime localiza ion, measu emen s a e aken
and he s a is ics o hese measu emen s a e calcula ed.
When he calcula ed s a is ics a e signi ican ly di e en
om he e e ence s a is ics, he ecei ed signal is
assumed o be co up ed by NLoS communica ion. Tha
can be exp essed wi h simple bina y de ec ion scheme,
whe e he de ec ion is pe o med by ex ac ing a ce ain
numbe o ecei ed signal ea u es such as Ku osis,
Mean Excess Delay o RMS Delay Sp ead o he
Mul ipa h Channel and applying he classical decision
heo y [1], [5]:
 
 
 
 
LoSp
NLoSp
Np
pH0
LoS|
LoS| 


, (3)
 
 
 
 
LoSp
NLoSp
Np
pH1
LoS|
LoS| 


, (4)
whe e p(γ|LoS) and p(γ|NLoS) ep esen , he PDFs o he
se o measu ed ecei ed signal ea u es γ = {γ1, γ2,…, γN}
unde LoS and NLoS condi ions, p(LoS) and p(NLoS) a e
he p io p obabili ies o he LoS and NLoS e en s, and
Н0 and Н1 deno e he LoS and NLoS signal p opaga ion
condi ions.
2.1. Dis ibu ion Tes s
In he case o LoS condi ions he samples o he ecei ed
signal en elope ha e dis ibu ions di e en om hose
ecei ed in NLoS condi ions. I he dis ibu ion o he
LoS is known, his ac can be used o iden i ica ion o
he communica ion channel cha ac e is ics. The aim o
he dis ibu ion es s is o es , whe he he samples o he
ecei ed signal en elope belong o a speci ic dis ibu ion,
so ha he LoS o NLoS condi ion can be iden i ied. A
numbe o well-known es s o iden i y he communica -
ion channel cha ac e is ics a e b ie ly desc ibed in [3]. In
he ollowing ex , he mos commonly used es
Kolmogo o -Smi no Tes is explained.
The objec i e o he Kolmogo o -Smi no Tes is
o iden i y whe he he ecei ed signal samples o igina e
om he known dis ibu ion ela ed o he LoS
condi ions. Suppose ha he e a e L measu emen s
o de ed as 1, 2,… L. The empi ical cumula i e
dis ibu ion unc ion o hese obse a ions (measu -
emen s) is de ined as [3]:
   
Li
L
i
iF ,...2,1;
0
, (5)
whe e (i) is he numbe o obse a ion poin s. The
ollowing hypo hesis can be o mula ed [9]:
     
iFiFiH exp00 ;: 
, (6)
     
iFiFiH exp01 ;: 
, (7)
whe e Fexp(i) is he cumula i e dis ibu ion unc ion o he
known expec ed p obabili y dis ibu ion ypical o he
signal samples ecei ed in he LoS condi ions. The ask
o he Kolmogo o -Smi no Tes is o ind he D-s a is ic,
which is de ined as he g ea es disc epancy be ween he
measu ed and he expec ed comula i e dis ibu ion
unc ions as [3]:
   
iFiFD Li exp0
1
max  
. (8)
The compu ed D-s a is ic is compa ed agains he
c i ical D´-s a is ic o he same sample size, which can
be compu ed based on he gi en le el o signi icance.
When he le el o signi icance equals 0,05 he c i ical D´-
s a is ic is [3]:
L
1,358
´D
. (9)
I D ≤ D´, he hypo hesis is alida ed. The
INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 10 | NUMBER: 4 | 2012 | SPECIAL ISSUE
© 2012 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 213
dis ibu ion unc ion o he measu ed ecei ed signal
samples co esponds wi h he expec ed dis ibu ion
unc ion ela ed o he LoS condi ions. O he wise, i
D > D´, he dis ibu ion unc ion o he measu ed samples
is no conside ed as he expec ed dis ibu ion and, hus,
hypo hesis is ejec ed. The ecei ed signal samples
o igina e in NLoS communica ion [3], [9].
2.2. Ku osis Pa ame e
The ku osis is a s a is ical pa ame e ha indica es he
a iance a ia ion o he signal ampli udes. I is de ined
as he a io o he ou h o de momen o he da a o he
squa e o he second o de momen [5]:
 
 
 
 
 
 
 
 
 
4
4
2
2
4
h
h
h
h hE
hE
hE










, (10)
whe e μ/h/ and σ/h/ a e he mean and s anda d de ia ion o
he ecei ed signal samples h( ), espec i ely. Ku osis is
cha ac e ized by he high alue o he signal ecei ed in
LoS condi ions [1], and by low alue o signal ecei ed
in NLoS condi ion ( he signal id mo e noise-like),[8].
Two ways can be used o LoS/NLoS decision.
Kolmogo o -Smi no goodness-o - i es can be used o
analyze how well he measu ed da a cha ac e ize he p io
dis ibu ion which ep esen s he LoS channel [5] o he
decision is simply aken as [1]:
NLoSi 


, (11)
LoSi 


, (12)
whe e λκ is he h eshold alue be ween he LoS and
NLoS condi ions. I Ku osis akes he alues abo e he
de ined h eshold i goes o LoS condi ion and ice e sa,
i i akes he alues less han he de ined h eshold i
goes o NLoS condi ion. The ad an age o Ku osis is
ha i can be calcula ed di ec ly on he ecei ed signal,
no es ima ion algo i hm is needed. This keeps easy and
quick he iden i ica ion p ocess and i implies ha no
es ima ion e o s o he es ima ion algo i hm a e
in oduced [8].
The simila analysis as discussed in he p e ious
pa can be pe o med using o he pa ame e s named
Mean Excess Delay τm and RMS Delay Sp ead τ ms.
While he Ku osis p o ides in o ma ion abou he
ampli ude s a is ics o he ecei ed signal samples, hese
wo pa ame e s p o ide he s a is ic ha cha ac e izes he
delay in o ma ion o he mul ipa h channel ob ained om
i s Powe Delay P o ile [5]. These pa ame e s a e
exp essed as [10]:
 
 



kk
kkk
mP
P



, (13)
 
2
2mm ms


, (14)
whe e P(τk) is he powe o he k- h pa h o he Powe
Delay P o ile wi h ime delay τk and
 
 



kk
kk
mP
P



2
2
. (15)
The comp ehensi e desc ip ion o he LoS/NLoS
iden i ica ion me hods he eade can ound in [3]. To
achie e sa is ac o y node dis ance es ima ion, i is
desi able o de e mine whe he he communica ion pa h
be ween a ansmi e - ecei e pai is LoS o NLoS. I i
is iden i ied as NLoS, he impac on localiza ion accu acy
can be supp essed ei he by he simple disca ding o he
measu emen s iden i ied as NLoS o by using some o he
NLoS mi iga ion echniques [4]. The i s men ioned
app oach is possible only when he e a e enough
measu emen s iden i ied as LoS, he e o e he app oach
o using he NLoS mi iga ion echniques seems o be
p e e able.
3. B ie O e iew o he
Localiza ion Sys em Based on
Coope a i e T ila e a ion
Philosophy o he posi ioning sys em is based on he
ila e a ion p inciple used as a undamen al localiza ion
me hod o cellula mobile ne wo ks whe e each mobile
node, which has he connec ion wi h minimum h ee
e e ence nodes, ancho s, can be loca ed. Th ee ancho s
wi h known posi ion a e su icien o wo-dimensional
localiza ion. This me hod belongs o he Range-based
posi ioning sys ems, whe e one o he localiza ion
pa ame e s mus be de e mined in o de o calcula e he
inal posi ion o he localized node. Cu en ly we use he
measu emen o he Time o A i al pa ame e
de e mined om he wi eless channel impulse esponse.
The posi ioning sys em also belongs in o he Ancho -
based sys ems he e o e i assumes he exis ence o a
couple o eal e e ence nodes wi h known loca ion. The
algo i hm consis s o mo e phases. Each node si ua ed in
he icini y o minimum h ee e e ence nodes wi h
known loca ion, is able o calcula e i s own posi ion. In
he i s phase all mobile nodes which a e in he ange o
a leas 3 s eady eal beacons a e localized. The second
and e e y o he phase a e like he i s one. Di e ence is
ha all mobile nodes al eady localized in he p e ious
phase become i ual beacons. The algo i hm is inished
when:
 posi ions o all mobile nodes we e de e mined
al eady,
 i is no possible ano he mobile nodes posi ion
de e mine (unde e mined nodes ha e no su icien
numbe o ancho s in ange).
INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 10 | NUMBER: 4 | 2012 | SPECIAL ISSUE
214 © 2012 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING
E e y phase o he p oposed algo i hm consis s o
he ollowing ou s eps:
 selec ion o a mobile node sui able o localiza ion
p ocess,
 he calcula ion o he dis ances be ween he
mobile node which should be loca ed and all
ancho s in i s ange,
 selec ion o he app op ia e ancho s in o de o
calcula e he unknown node loca ion,
 ila e a ion and posi ion calcula ion o unknown
mobile node.
A he beginning o he each phase he nodes,
which ha e no been al eady calcula ed he loca ion a e
selec ed. Each o he nodes includes he da abase o i s
neighbou s, he dis ances o hem acco ding o he
measu ed ToA pa ame e and also he in o ma ion i hese
neighbou s a e ancho s o no . F om his da abase o he
a ailable e e ence nodes only one iple is selec ed
using he ollowing p ocedu e:
 e e y ancho iple ha is conside ed is c ea ed,
 only hose iple s whe e all pai s o ci cles a e
in e sec ed a e conside ed,
 he iple which include he ancho s wi h s eady
posi ion a e p e e ed,
 his nea es iple (acco ding o ToA pa ame e ) is
p e e ed om all. This iple c ea es he iple o
ancho s sui able o calcula ion o he inal
posi ion o unknown node by ila e a ion.
I a sui able iple o ancho s in he ange o
unknown node exis s, he localiza ion p ocess can be
ca ied ou . As i is s a ed abo e, he p ocess o
ila e a ion is used o calcula ion o he posi ion o
unknown nodes whe e [x,y] ep esen he coo dina es o
he unknown node (which should be calcula ed),
[xi,j,k,yi,j,k] ep esen he loca ion o he e e ence nodes
and i,j,k ep esen he es ima ed node dis ances be ween
unknown node and he e e ence nodes sepa a ely:
   
2,,
2
,,
2
,, kjikjikji yyxx 
. (16)
Ideally, he inal posi ion is ep esen ed by he
only common in e sec ion o h ee ci cles. The nega i e
impac o wi eless communica ion channel causes
inaccu a e calcula ion o dis ances be ween nodes. Hence
he inal posi ion is calcula ed as a cen e o he a ea
which is ep esen ed by a iple o sui able in e sec ions.
4. Simula ions and Resul s
In ou simula ions we ha e used he exis ing model o he
wi eless channel impulse esponse gene a ion o IEEE
802.15.4a ul a wideband echnology, which is desc ibed
in mo e de ails in [11]. I is he esul o he esea ch
ac i i y o he membe s o he Mi subishi Elec ic
Resea ch Labo a o ies and p o ides he gene a ion o he
channel impulse esponse o di e en ypes o he
en i onmen such as esiden ial, o ice, open ou doo and
indus ial. In ou simula ions we ha e used he CIR o
“LoS and NLoS OFFICE” en i onmen .
In mos cases he ime o a i al o he ecei ed
signal is assumed as he ime o a i al o he i s
de ec ed pa h. The p oblem occu s, when he di ec pa h
is missing in he ecei ed signal channel impulse
esponse. This ac causes conside able dis ance
es ima ion e o s, which leads o he inco ec
localiza ion. Fo NLoS mi iga ion p oblem, especially in
case o NLoS communica ion wi h unde ec ed di ec
pa h, we ha e p oposed he low complexi y NLoS
mi iga ion echnique o ToA es ima ion desc ibed in he
ollowing pa . In Fig. 1, he diag am o he ToA
es ima ion o di e en ypes o communica ion channel
is in oduced.
Fig. 1: Diag am o he ToA es ima ion o di e en ypes o
communica ion channel.
In pa o LoS/NLoS iden i ica ion we ha e used
he Ku osis pa ame e “KP” calcula ion wi h a de ec ion
h eshold be ween he LoS and NLoS communica ion
condi ions o 100. The desi ed h eshold alue o KP has
been ob ained om he expe imen al obse a ions o
se e al assumed h esholds om 60 o 220, whe e 100
has b ough he mos success ul esul s o LoS/NLoS
iden i ica ion, Fig. 2. Than in he simula ions, acco ding
o he calcula ed KP he algo i hm decided he
communica ion condi ions. I KP ≥ 100 i iden i ied he
“LoS OFFICE” en i onmen , else i KP < 100 he “NLoS
OFFICE” en i onmen was iden i ied. Using 1000- imes
andomly gene a ed CIR o “OFFICE” en i onmen he
LoS/NLoS iden i ica ion by using Ku osis calcula ion
was success ul in 84,7 %.
The pa o IDENTIFICATION o di ec pa h
p esence was no he aim o his pape . I will be he aim
o ou esea ch ac i i y in he nex pe iod. In his pape
we ha e ocused on LoS es ima ion p oblem in he case
o NLoS channel wi h Unde ec ed Di ec Pa h (UDP). We
p oposed he low complexi y UDP mi iga ion me hod in
o de o imp o e he dis ance es ima ion, which is nex
explained using he conc e e example o he ansmi e -
ecei e pai 10 m dis an . The ecei e ecei ed he
signal as is depic ed in Fig. 3 ( he ime componen s om
1 o 9). I can be seen ha in he ecei ed channel
INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 10 | NUMBER: 4 | 2012 | SPECIAL ISSUE
© 2012 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 215
impulse esponse he di ec pa h is missing. We need o
es ima e he assumed ime o a i al o he blocked di ec
pa h. The co ec ion ToA pa ame e can be ob ained using
he ollowing p ocedu e:
1
)(
21






k
C
k
iii
LoS
, (17)
whe e i is he ime o a i al o he i- h pa h ecei ed
abo e he le el o signi icance, which ep esen s he
pa hs wi hin 10 dB om maximum peak, k is he numbe
o signi ican pa hs and CLoS is he mean o he ime
di e ences be ween pai s o hese neighbou ing pa hs,
named as “ToA co ec ion pa ame e ”.
Fig. 2: Dependence o he success ul a e on he KP h eshold.
Fig. 3: LoS es ima ion explana ion.
The assumed ime o a i al o he es ima ed LoS
pa h is de i ed as:
LoSLoSes ima ed C ToA 
1
. (18)
This example is explained using Tab. 1, which
con ains he imes o a i al o he men ioned signi ican
pa hs wi h he di e ences be ween hem. As i can be
seen in he Tab. 2 and Tab. 3, he p oposed NLoS
mi iga ion me hod p oduced be e esul s o he
es ima ed ToA pa ame e and inal dis ance es ima ion o
1,414 m, which ep esen s he imp o emen o dis ance
es ima ion e o o 41 %. This p ocedu e has been es ed
on 1000- imes andomly gene a ed CIRs o ansmi e -
ecei e pai s dis an a 10 m. Gene ally, he dis ance
es ima ion e o has been educed by 22,6 %, Tab. 4.
Tab.1: Time o a i al o he signi ican pa hs and he ime di e ences
be ween hem.
i
i [ns]
i- i-1 [ns]
1
41,3
---
2
44,0
2,7
3
49,0
5,0
4
53,7
4,7
5
56,3
2,6
6
62,7
6,4
7
69,3
6,6
8
74,3
5,0
9
79,0
4,7
Tab.2: ToA and dis ance es ima ion by using he i s de ec ed peak.
ToA 1 [ns]
Es ima ed dis ance [m]
41,3
12,4
Tab.3: ToA and dis ance es ima ion by using he p oposed NLoS
mi iga ion me hod o ToA es ima ion enhancemen .
CLoS [ns]
ToAes ima ed-LoS [ns]
Es ima ed
dis ance [ns]
4,71
36,62
10,986
Tab.4: The compa ison o he dis ance es ima ion e o using bo h
ToA es ima ion echniques.
ToA es ima ion
Dis ance es ima ion e o [m] o
nodes dis an a 10 m
Fi s de ec ed pa h
1,541
NLoS mi iga ion
me hod
1,193
Finally, he p oposed ToA es ima ion me hod has
been es ed also in coope a i e ila e a ion based
algo i hm desc ibed in he sec ion 3. We simula ed only
he ha sh communica ion condi ions be ween pai s o
mobile nodes – NLoS wi h UDP. We ha e s udied he
impac o he p oposed NLoS mi iga ion me hod on he
po en ial enhancemen o he dis ance es ima ion e o .
Ou simula ions we e ca ied ou in MATLAB
p og amming en i onmen and o s a is ical ele ance
we e epea ed one hund ed imes. We ha e used he
ne wo k which consis ed o he 40 mobile nodes
deployed on he a ea o 30  30 m wi h all ecei e s’
adio ange o 10 m. Th ee s eady iangle-shaped beacon
nodes we e loca ed in he middle o localiza ion a ea. We
assumed p ecise ime synch oniza ion be ween nodes.
I we used he i s algo i hm, whe e he ToA
pa ame e is de e mined acco ding o he ime o a i al
o he i s de ec ed peak, he localiza ion e o was
29,96 % in compa ison wi h he size o a ea side. By
using he p oposed NLoS mi iga ion me hod he
localiza ion e o was educed o 27,77 %. In bo h cases
he beha iou o he algo i hm was success ul enough
wi h a numbe o he localized nodes abo e 90 %, Tab. 5.

INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 10 | NUMBER: 4 | 2012 | SPECIAL ISSUE
216 © 2012 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING
Tab.5: Posi ioning algo i hm e alua ion a e p oposed NLoS
mi iga ion echnique implemen a ion.
Used ToA
es ima ion me hod
Localiza ion
e o [%]
Numbe o
success ully
localized nodes [%]
Fi s de ec ed
pa h
29,96
91,65
NLoS mi iga ion
me hod
27,77
93,30
5. Conclusion
This pape p esen s he low complexi y NLoS mi iga ion
echnique and i s applica ion in he p oposed coope a i e
localiza ion algo i hm based on ci cula ila e a ion. In
gene al, he p oposed me hod o ToA es ima ion
enhancemen imp o es he ime o a i al accu acy, bu i
has se e al d awbacks. In some cases, i he e a e less
signi ican pa hs in he channel powe delay p o ile which
a e mo e dis an om each o he , he ToA es ima ion can
be inaccu a e. The me hod, how o imp o e he p oposed
NLoS mi iga ion me hod will be he aim o he u u e
esea ch ac i i y.
Acknowledgemen s
This wo k has been pe o med pa ially in he amewo k
o he EU ICT P ojec INDECT (FP7-218086).
Re e ences
[1] DECARLI, N., D. DARDARI, S. GEIZICI and A. A.
D'AMICO. LOS/NLOS De ec ion o UWB Signals: A
Compa a i e S udy Using Expe imen al Da a. In: IEEE
In e na ional Symposium on Wi eless Pe asi e Compu ing
(ISWPC). Modena: IEEE, 2010. pp. 169–173. ISBN 978-1-
4244-6855-3. DOI: 10.1109/ISWPC.2010.5483704.
[2] VOSSIEK, M., L. WIEBKING, P. GULDEN, J.
WIEGHARDT, C. HOFFMANN and P. HEIDE. Wi eless local
posi ioning. Mic owa e Magazine. 2003, ol. 4, iss. 4, pp. 77-
86. ISSN 1527-3342. DOI: 10.1109/MMW.2003.1266069.
[3] YU, K., I. SHARP and J. GUO. G ound-Based Wi eless
Posi ioning. Chiches e : John Wiley & Sons, 2009. ISBN 978-
0-470-74704-9.
[4] GUVENC, I. and Ch.-Ch. CHONG. A Su ey on TOA Based
Wi eless Localiza ion and NLOS Mi iga ion Techniques.
Jou nal: IEEE Communica ions Su eys & Tu o ials. 2009,
ol. 11, iss. 3, pp. 107-124. ISSN 1553-877X. DOI:
10.1109/SURV.2009.090308.
[5] GUVENC, I., Ch.-Ch. CHONG and F. WATANABE. NLOS
Iden i ica ion and Mi iga ion o UWB Localiza ion Sys ems.
In: Wi eless Communica ions and Ne wo king Con e ence,
2007. WCNC 2007. 2007. IEEE. Kowloon: IEEE, 2007, pp.
1571-1576. ISSN 1525-3511. ISBN 1-4244-0658-7. DOI:
10.1109/WCNC.2007.296.
[6] HATAMI, Ahmad. Applica ion o Channel Modeling o Indoo
Localiza ion Using TOA and RSS. Wo ces e , 2006.
Disse a ion Thesis. Wo ces e Poly echnic Ins i u e.
[7] ALSINDI, Naye , C. DUAN, J. ZHANG and T. TSUBOI.
NLOS Channel Iden i ica ion and Mi iga ion in Ul a
Wideband ToA-based Wi eless Senso Ne wo ks. In:
P oceedings o 6 h Wo kshop on Posi ioning, Na iga ion and
Communica ion (WPNC). Hanno e : IEEE, 2009. pp. 59–66.
ISBN 978-1-4244-3292-9. DOI:
10.1109/WPNC.2009.4907804.
[8] MUCCHI, L. and P. MARCOCCI. A new UWB indoo channel
iden i ica ion me hod. In: C ownCom 2007. 2nd In e na ional
Con e ence on Cogni i e Radio O ien ed Wi eless Ne wo ks
and Communica ions, 2007. O lando: IEEE, pp. 58–62. ISBN
978-1-4244-0814-6. DOI:
10.1109/CROWNCOM.2007.4549772.
[9] TOMAN, Edua d. S a is icka in e e encia caso ych adoch.
B a isla a, 2007. Diploma Thesis. Comenius Uni e si y in
B a isla a.
[10] DOBOS, L. and F. JAKAB. Mul imedia in mobile ne wo ks.
Kosice: TU-FEI, 2005. ISBN 80-8086-000-9.
[11] MOLISCH, And eas F., Kannan BALAKRISHNAN, Dajana
CASSIOLI, Chia-Chin CHONG, Shah ia EMAMI, And ew
FORT, Johan KAREDAL, Jue gen KUNISCH, Hans
SCHANTZ, Ul ich SCHUSTER and Kai SIWIAK. IEEE
802.15.4a Channel Model – Final Repo . In: Documen IEEE
802.1504-0062-02-004a [online]. 2005. A ailable a :
h p://g oupe .ieee.o g/g oups/802/15/pub/04/15-04-0662-02-
004a-channel-model- inal- epo - 1.pd .
[12] POVENDRAN, R., C. WANG and R. SUMIT. Secu e
Localiza ion and Time Synch oniza ion o Wi eless Senso and
Ad hoc Ne wo ks. New Yo k: Sp inge Science+Business
Media, 2007. ISBN 978-0-387-32721-1.
[13] SRINIVASAN, A. and J. WU. A Su ey on Secu e Localiza ion
in Wi eless Senso Ne wo ks. Boca Ra on: CRC P ess, Taylo
and F ancis G oup, 2008.
[14] HEIDARI, M., F. O. AKGUL and K. PAHLAVAN.
Iden i ica ion o he Absence o Di ec Pa h in Indoo
Localiza ion Sys ems. In: IEEE 18 h In e na ional Symposium
on Pe sonal, Indoo and Mobile Radio Communica ions, 2007.
PIMRC 2007. A hens: IEEE, 2007. pp. 1–6. ISBN 978-1-4244-
1144-3. DOI: 10.1109/PIMRC.2007.4394450.
Abou Au ho s
Vladimi CIPOV was bo n in Banska Bys ica, Slo ak
epublic in 1985. He ecei ed his M.Sc. om Technical
Uni e si y o Kosice, Facul y o Elec ical Enginee ing
and In o ma ics in 2009. He g adua ed wi h majo in
Elec onics and Telecommunica ions Enginee ing. Today,
he is Ph.D. s uden and his scien i ic esea ch a ea is:
Localiza ion in Wi eless Mobile Ad-hoc Ne wo ks.
Lubomi DOBOS was bo n in V ano n/Toplou, Slo ak
Republic in 1956. He ecei ed he Ing. (M.Sc) deg ee and
Ph.D. deg ee in Radioelec onics om he Facul y o
Elec ical Enginee ing, Technical Uni e si y o Kosice, in
1980 and 1989, espec i ely. He de ended his habili a ion
wo k - B oadband Wi eless Ne wo ks o Mul imedia
Se ices in 1999. His scien i ic esea ch a ea is: b oad-
band in o ma ion and elecommunica ion echnologies,
mul imedia sys ems, elecommunica ions ne wo ks and
se ices, mobile ne wo ks, ou ing p o ocols.
Jan PAPAJ was bo n in Lip o sky Mikulas, Slo ak
Republic in 1977. He is g adua ed a he Depa men o
Compu e s and In o ma ics in 2001, Facul y o Elec ical
INFORMATION AND COMMUNICATION TECHNOLOGIES AND SERVICES VOLUME: 10 | NUMBER: 4 | 2012 | SPECIAL ISSUE
© 2012 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 217
Enginee ing and In o ma ics a Technical Uni e si y in
Kosice. He ob ained his Ph.D. deg ee in Telecommunica-
ions om he Facul y o Elec ical Enginee ing, Techni-
cal Uni e si y o Kosice in 2010. His esea ch in e es s
co e he ields o MANETs, QoS and secu i y, ou ing
p o ocols, oppo unis ic and senso ne wo ks.