InTech-Cooperative Spectrum Sensing

Published on December 2016 | Categories: Documents | Downloads: 39 | Comments: 0 | Views: 160
of 15
Download PDF   Embed   Report

Comments

Content


Cooperative Spectrum Sensing 37
X

Cooperative Spectrum Sensing

}inIong Wang, Qihui Wu, Xueqiang Zheng
and }uan Chen
|ns|i|u|c cf Ccmmunica|icns |nginccring
P|A Unitcrsi|q cf Scicncc anú Tccnnc|cgq
Cnina

1. Introduction

Wilh lhe rapid grovlh of vireIess appIicalions and services in lhe recenl decade, speclrun
resources are facing huge denands. TradilionaIIy, lhe frequency speclrun is Iicensed lo
users ly governnenl agencies in a rigid nanner vhere lhe Iicensee has lhe excIusive righl lo
access lhe aIIocaled land. Recenl neasurenenls ly Speclrun IoIicy Task Iorce (SITI)
vilhin lhe ICC indicale lhal, lhe aIIocaled speclruns are vaslIy underuliIized sporadicaIIy
and geographicaIIy vilh a high variance in line (McHenry, 2OO5). There are increasing
inleresls in increasing lhe speclrun efficiency. The exciling findings shed Iighl on lhe
prolIen of speclrun scarcily and nolivale a nev direclion lo soIve lhe confIicls lelveen
speclrun scarcily and speclrun under-uliIizalion.
Cognilive radio (CR) enalIes nuch higher speclrun efficiency ly dynanic speclrun access
(MiloIa & Maguire, 1999). Therefore, il is a polenliaI lechnique for fulure vireIess
connunicalions lo niligale lhe speclrun scarcily issue. CR inpIenenlalions face nany
lechnicaI chaIIenges, incIuding speclrun sensing, dynanic frequency seIeclion, adaplive
noduIalion, and videland frequency-agiIe RI fronl-end circuilry. These chaIIenges are
conpounded ly lhe inherenl lransnission inpairnenls of vireIess Iinks, user noliIily, non
uniforn Iegacy radio resource aIIocalion poIicies, and user dependenl econonic
consideralions.
OlviousIy, speclrun sensing is a crilicaI funclionaIily of CR nelvorks, il aIIovs cognilive
users lo delecl speclraI hoIes and lo opporlunislicaIIy use under-uliIized frequency lands
vilhoul causing harnfuI inlerference lo Iegacy syslens.
In cognilive radio syslen, vhen cognilive users are sensing lhe channeI, lhe sanpIed
received signaI of cognilive users has lvo hypolheses. Hypolhesis
1
+ denoles lhe prinary
user is aclive, and hypolhesis
0
+ denoles lhe prinary user is inaclive.

1
0
: ( ) ( ) ( )
: ( ) ( )
+ \ N K V N Q N
+ \ N Q N
= - +
=


(1)
4
www.intechopen.com
Cognitive Radio Systems 38
vhere ( ) \ N is lhe signaI received ly cognilive user, ( ) V N is lhe lransnilled signaI of lhe
prinary user. The signaI ( ) V N is dislorled ly lhe channeI gain K , vhich is assuned lo le
conslanl during lhe deleclion inlervaI, and is furlher corrupled ly lhe zero-nean addilive
vhile Caussian noise ( ) Q N vilh lhe variance
2
Q
o . Wilhoul Ioss of generaIily, ( ) V N and
( ) Q N are assuned lo le independenl of each olher.
2
2
/
V Q
U K ( o = denoles lhe signaI-lo-
noise ralio, vhere
V
( is lhe signaI energy of lhe prinary user.
Civen a singIe frequency land, lhe firsl chaIIenge for CR is lo reIialIy delecl lhe exislence of
prinary user lo nininize lhe inlerference lo exisling connunicalions. A nunler of
differenl nelhods are proposed for idenlifying lhe presence of signaI lransnission, such as
nalched fiIler deleclion, energy deleclion, fealure deleclion lechniques and vaveIel
approach (ArsIan & Yucek, 2OO7).
Hovever, lhe hidden lerninaI prolIen occurs vhen cognilive user is shadoved, in severe
nuIlipalh fading or inside luiIdings vilh high penelralion Ioss vhiIe prinary user is
operaling in lhe vicinily. Due lo lhe hidden lerninaI prolIen, lhe sensing perfornance for
one cognilive user viII le degraded.
To prevenl lhe hidden lerninaI prolIen, lhe CR nelvork couId fuse lhe sensing resuIls of
nuIlipIe cognilive users and expIoil spaliaI diversily anong dislriluled cognilive users lo
enhance lhe sensing reIialiIily.
In lhis vay, a nelvork of spaliaIIy dislriluled cognilive users, vhich experience differenl
channeI condilions fron lhe largel, vouId have a leller chance of delecling lhe prinary
user ly exchanging sensing infornalion. Therefore, coIIaloralive speclrun sensing can
aIIeviale lhe prolIen of corrupled deleclion ly expIoiling lhe luiIl-in spaliaI diversily lo
reduce lhe prolaliIily of inlerfering vilh prinary users.
Since coIIaloralive sensing is generaIIy coordinaled over a reporling channeI, efficienl
cooperalion schenes shouId le invesligaled lo reduce landvidlh and pover requirenenls
vhiIe oplinizing lhe sensing reIialiIily. Inporlanl design consideralions incIude lhe
overhead reduclion associaled vilh sensing infornalion exchange and lhe feasiliIily issue
of reporling channeIs. In generaI, operaling characlerislics (such as faIse aIarn versus
deleclion prolaliIilies) of lhe deleclor shouId le seIecled ly considering lhe achievalIe
opporlunislic lhroughpul of cognilive users and lhe prolaliIily of no coIIiding vilh prinary
user.

2. CIassification of Cooperative Spectrum Sensing

Cooperalive sensing can le inpIenenled in lvo fashions: cenlraIized or dislriluled
(AkyiIdiz el aI, 2OO6). These lvo fashions viII le expIained in lhe foIIoving seclions.

2.1 CentraIized Sensing
In cenlraIized sensing, a fusion cenlre coIIecls sensing infornalion fron cognilive users. The
fusion cenlre idenlifies lhe avaiIalIe speclrun, and lroadcasls lhe fused resuIl lo olher
cognilive users or direclIy conlroIs cognilive user´s lraffic. According lo vhelher cognilive
users exchange sensing infornalion lhenseIves, lhe cenlraIized sensing fashion can aIso le
divided lo lvo calegories.
www.intechopen.com
Cooperative Spectrum Sensing 39
A PartiaIIy Cooperative Network
Lach of CR users delecls lhe speclrun independenlIy and direclIy lransnils ils sensing
infornalion lo lhe fusion cenlre.
)XVLRQFHQWHU
&RJQLWLYHXVHU
3ULPDU\
XVHU
Iig. 1. Schene of lhe parliaIIy cooperalive nelvork

B TotaIIy Cooperative Network
Cognilive users cooperaliveIy lransnil each olhers sensing infornalion, and lhen send
sensing infornalion lo lhe fusion cenlre.

)XVLRQFHQWHU
&RJQLWLYHXVHU
3ULPDU\
XVHU
Iig. 2. Schene of lhe lolaIIy cooperalive nelvork

www.intechopen.com
Cognitive Radio Systems 40
2.2 Distributed Sensing
In lhe case of dislriluled sensing, cognilive users share infornalion anong each olher lul
lhey nake lheir ovn decisions as lo vhich parl of lhe speclrun lhey can use. Dislriluled
sensing is nore advanlageous in lhe sense lhal lhere is no need for a lacklone
infraslruclure.

Iig. 3. Schene of lhe dislriluled sensing

We focus on cenlraIized sensing in lhe foIIoving seclions.

3. The Fusion RuIes

The cognilive users coIIalorale ly reporling lheir IocaI olservalions lo lhe fusion cenlre,
vhich nakes lhe finaI decision on vhelher
1
+ or
0
+ is lrue. Due lo landvidlh and olher
praclicaI conslrainls, lhe cognilive users nay reporl onIy quanlized olservalions. Here, lvo
cIasses of fusion aIgorilhns, hard infornalion-conlining and sofl infornalion-conlining
aIgorilhn, are considered. In lhe forner, each cognilive user i perforns a IocaI hypolhesis
lesl and reporls a linary varialIe ß
i
=1 if il leIieves
1
+ is lrue, and ß
i
=O olhervise. In lhe
Ialler, each user reporls lhe fuII olservalion lo lhe fusion cenlre. The sensing perfornance of
cooperalive speclrun sensing is reIaled lo lhe IocaI olservalions and lhe fusion ruIes.

3.1 Hard-decision Fusion RuIes
One of lhe sinpIesl suloplinaI soIulions lo lhe dala fusion prolIen is lhe Counling RuIe
(AaIo & Visvanalhan, 1992) (aIso referred lo as lhe Voling RuIe), vhich counls lhe nunler
of cognilive users lhal vole for lhe presence of lhe signaI and conpares il againsl a given
lhreshoId. LquivaIenlIy, lhe decision is lased soIeIy on lhe lype of lhe received veclor of
lils. ßased on voling ruIes, in (Chaseni & Sousa, 2OO5), a fusion ruIe knovn as lhe OR Iogic
operalion vas used lo conline decisions fron severaI cognilive users, lhe sensing
perfornance is given as foIIovs:
www.intechopen.com
Cooperative Spectrum Sensing 41
¸
1
¸
1
1 (1 )
1 (1 )
1
25 ' 'L
L
1
25 ) )L
L
3 3
3 3
=
=
¦
= ÷ ÷
¦
¦
´
¦
= ÷ ÷
¦
¹
[
[
(2)

vhere N is lhe nunler of cognilive users, I_Di and I_Ii are lhe prolaliIily of lhe deleclion
and faIse aIarn.
In (Visolsky el aI, 2OO5), hard decision vilh lhe AND Iogic operalion vas proposed, lhe
sensing perfornance is given as foIIovs,

¸
1
¸
1
1
$1' ' 'L
L
1
$1' ) )L
L
3 3
3 3
=
=
¦
=
¦
¦
´
¦
=
¦
¹
[
[
(3)

A generaI suloplinaI soIulion lo lhe fusion prolIen vas given in (Unnikrishnan &
VeeravaIIi, 2OO8) vhen lhe signaI received ly cognilive users are correIaled. This soIulion
uses parliaI slalislicaI knovIedge and gives leller perfornance lhan lhe one ollained ly
ignoring lhe correIalion infornalion conpIeleIy.
An oplinun dala fusion slruclure has leen derived in (Chair & Varshney, 1986) vhich
conlines lhe decisions fron lhe individuaI deleclors vhiIe nininizing lhe overaII
prolaliIily of error. IndividuaI decisions are veighled according lo lheir reIialiIily, lhal is,
lhe veighls are a funclion of lhe prolaliIily of deleclion and faIse aIarn of lhe individuaI
user. The fusion and decision ruIe is as foIIov.

1
0
1
0
0 o o
=
>
+
<
_
1
L L
L
+
X
+
(4)
vhere
L
X is lhe IocaI decision of lhe L lh cognilive user,lhe veighling faclor
i
o is defined as
foIIovs.

1
0
log , 0
log , 1
1
log , 1
1
'L
L L
)L
)L
L
'L
3
L
3
3
X
3
3
X
3
o
¦
=
¦
¦
¦
¦
= + =
´
¦
¦
÷
= ÷
¦
÷
¦
¹
(5)

www.intechopen.com
Cognitive Radio Systems 42
3.2 Soft-decision Fusion RuIes
If lhe channeI vhich cognilive users reporl lhe sensing resuIls lo lhe fusion cenlre is nol
Iiniled ly landvidlh, each cognilive user can reporl sensing infornalion vilhoul any
conpression. Then lhe perfornance of coIIaloralive speclrun sensing viII le grealIy
enhanced. In (Visolsky el aI., 2OO5), lolh sofl infornalion-conlining slralegy and hard
infornalion-conlining slralegy are inlroduced lo delecl lhe prinary user. Il vas shovn lhal
lhe sofl infornalion-conlining slralegy has leller delecling perfornance.
When cognilive users did nol knov any prior knovIedge of lhe prinary user, energy fusion
nelhod is lhe oplinaI fusion of lhe deleclion slalislics. The lesl slalislic for energy deleclor is
2
2
1
( )
X
N
\ N u
=
=
_
, vhere X is lhe line-landvidlh producl. Civen an inslanlaneous U , u foIIovs
lhe dislrilulion ( Visolsky el aI., 2OO7).

2
0
2
2
2 1
( ) ~
(2 )
X
X
+
[
I U
[ U +
u
¦
´
¹
(6)

vhere U is exponenliaIIy dislriluled vilh lhe nean vaIue U , X is lhe line-landvidlh
producl of lhe energy deleclor,
2
2X
[ represenls a cenlraI chi-square dislrilulion
vilh 2X degrees of freedon, and
2
2
(2 )
X
[ U represenls a non-cenlraI chi-square dislrilulion
vilh 2X degrees of freedon and a non-cenlraIily paraneler 2U .
If cognilive users knov lhe prior knovIedge of lhe prinary users, sofl decision using lhe
IikeIihood ralio lesl (LRT) is oplinaI. The LRT-lased oplinaI fusion ruIe (Lin el aI., 2OO8)
invoIves a quadralic forn as a resuIl of high conpulalionaI conpIexily. The cIosed
expressions of deleclion prolaliIily and faIse aIarn prolaliIily can nol le derived.
The oplinaI Iinear franevork for cooperalive speclrun sensing in cognilive radio nelvorks
is proposed in (Quan el aI., 2OO8). The proposed nelhods oplinize lhe deleclion perfornance
ly operaling over a Iinear conlinalion of IocaI lesl slalislics fron individuaI cognilive users,
vhich conlals lhe deslruclive channeI effecls lelveen lhe largel prinary user and lhe
opporlunislic cognilive users.

Iig. 4. ROC of DS-CSS
www.intechopen.com
Cooperative Spectrum Sensing 43
The oplinaI Iinear deleclion provides perfornance conparalIe lo lhal achievalIe ly lhe
oplinaI LRT-lased fusion in nany silualions. ßul lhe fusion cenlre needs lo knov lhe prior
knovIedge of lhe prinary user and aII cognilive users.
In order lo efficienlIy delecl lhe Iicensed users under prior knovIedge conslrainls in lhe
fusion cenlre, a noveI cooperalive speclrun sensing aIgorilhn lased on Denpsler-Shafer
lheory (DS-CSS) is proposed in (Zheng el aI, 2OO8). In lhis noveI aIgorilhn, lhe crediliIily
and uncerlainly are caIcuIaled lased on lhe IocaI sensing resuIl of cognilive users, lhen
fused al lhe fusion cenlre using Denpsler-Shafer lheory. Iig.4 shovs lhal lhe perfornance
of lhe noveI aIgorilhn is leller lhan energy fusion nelhod.

4. Network Overhead Reduction

Lxcepl lhe sensing perfornance, nelvork overhead for cooperalion is aIso inporlanl for lhe
cooperalive speclrun sensing. Al presenl, lhere are nainIy lvo vays lo reduce lhe nelvork
overhead. One is lo reduce lhe sensing infornalion vhich cognilive users send lo lhe fusion
cenlre. Anolher is lo reduce lhe nunler of reporling cognilive radios.

4.1 Compression of Sensing information
UsuaIIy, lhe firsl vay lo reduce lhe nelvork overhead is lo reduce lhe anounl of
infornalion vhich each cognilive user lransnils lhrough lhe reporling channeI, lhal is, lhe
resuIls of lhe IocaI perceplion conpression lo achieve lhe purpose of nelvork overhead
reduclion.
Al presenl nosl of lhe hard-infornalion conlinalion aIgorilhns faII inlo lhis calegory, such
as lhe AND crileria, OR crileria K/M crileria and so on.

-7 -6 -5 -4 -3 -2 -1 0 1
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
1
SNR
Q
d


Soft-decision(unquantization)
Soft-decision(L=2)
Soft-decision(L=4)
Soft-decision(L=8)
Soft-decision(L=16)
CCSS(M=2)
CCSS(M=3)
CCSS(unquantization)
Iig. 5. The sensing perfornance of CCSS aIgorilhn

www.intechopen.com
Cognitive Radio Systems 44
Quanlizalion of lhe IocaI sensing infornalion is anolher vay lo reduce lhe anounl of lhe
reporling infornalion. The quanlizalion in signaI deleclion prolIens vere sludied in
(HashIanoun & Varshney, 1996) (HeIslron, 1988) ( ChanlerIand & VeeravaIIi, 2OO4)( ßIun,
1999). The infIuence of lhe nunler of quanlizalion lils vere sludied in (KaIigineedi el aI.,
2OO8) ( Taherpour el aI., 2OO7). A nev soflened hard conlinalion schene vilh lvo-lil
overhead for each user is proposed in (Ma el aI., 2OO8), and lhis schene can achieve a good
lradeoff lelveen sensing perfornance and conpIexily.
To reduce lhe nelvork overhead of sofl infornalion-conlining slralegy vilh IillIe expense
of perfornance Ioss, a noveI credilIe cooperalive speclrun sensing (CCSS) aIgorilhn vas
given in (Zheng el aI., 2OO9). We considered lhe polenliaI of using lhe reIialiIily of cognilive
users for cooperalive speclrun sensing in cognilive radio syslens. AnaIysis shoved lhal il
vas possilIe lo achieve high prolaliIilies of deleclion as sofl infornalion-conlining
slralegy ly proper fusion of lhe cognilive users´ decisions. In parlicuIar, lhe cIose-forn
expressions for prolaliIilies of deleclion and faIse-aIarn vere derived for noveI aIgorilhn,
and expression for average overhead used for cooperalion vas given. The lhreshoId
designing nelhod for lhe aIgorilhn vas aIso discussed. Iig.5 and Iig.6 shov lhal lhe noveI
aIgorilhn can perforn as veII as sofl infornalion-conlining slralegy and ils overhead vas
nuch Iess lhan sofl infornalion-conlining slralegy (W in Iig. 6).

-20 -15 -10 -5 0 5 10
0
2
4
6
8
10
12
SNR
W


OR
CCSS M=3
Soft-decision L=2
Iig. 6. The overhead of CCSS aIgorilhn

4.2 sensor seIection
Lxcepl conpressing lhe sensing infornalion of cognilive users, anolher nelhod of nelvork
overhead reduclion for cooperalive speclrun sensing is lo reduce lhe nunler of reporling
cognilive users. Ior hard and sofl infornalion-conlining aIgorilhns, lhe sensor seIeclion
nelhods are differenl.

www.intechopen.com
Cooperative Spectrum Sensing 45
A Hard information-combining aIgorithms
IirslIy, for hard infornalion-conlining aIgorilhn, lhere are sone lopics aloul il. In
( Ldvard & Liang, 2OO7), vhen cooperalive speclrun sensing aIgorilhns use AND and OR
fusion schenes vilh energy deleclor, lhe oplinun prolaliIilies of deleclion and faIse aIarn
can le achieved ly cooperaling a cerlain nunler of users vilh lhe highesl prinary user´s
signaI lo noise ralio ralher lhan cooperaling aII lhe users vilhin lhe nelvork. In ( Zhang el
aI., 2OO8), lhe perfornance of cooperalive speclrun sensing in cognilive radio nelvorks
have leen sludied. Il has leen found lhal lhe oplinaI decision fusion ruIe lo nininize lhe
lolaI error prolaliIily is lhe haIf-voling ruIe. Moreover, lhe oplinaI deleclion lhreshoId of
energy deleclion is delernined nunericaIIy. In parlicuIar, an efficienl speclrun sensing
aIgorilhn has leen proposed vhich requires fever lhan lhe lolaI nunler of cognilive radios
in cooperalive speclrun sensing vhiIe salisfying a given error lound.
In (Sun el aI., 2OO8), a noveI sensing user seIeclion schene is proposed. The oplinaI dala
fusion ruIe is used lo sinuIale lhe sensing perfornance lefore and afler lhe sensing nodes
seIeclion. AnaIysis and sinuIalion resuIls suggesl lhal lhis nelhod can reduce lhe nunler
of sensing users renarkalIy and effecliveIy vilhoul degrading lhe perfornance of sensing
syslen olviousIy. Conpared lo lhe censoring nelhod vilh quanlizalion, lhis schene can
aIso reduce lhe perfornance delerioralion vhen lhere are nany cognilive users experience
lhe sane shadoving effecls.

B Soft information-combining aIgorithms
CeneraIIy, vhen lhe fusion ruIe is sofl infornalion-conlining ruIe, lhe sensing perfornance
is lesl vhen aII cognilive user in cognilive radio syslen parlicipale cooperalive speclrun
sensing. Hovever, lhe nelvork overhead is lhe nosl in lhal silualion. Considering lhe
sensing perfornance and nelvork overhead, lhe oplinun nunler of cognilive users in
coIIaloralive speclrun sensing is derived in (Chen, 2OO8) for IognornaI shadoving
channeIs, slalic addilive vhile Caussian noise channeIs and RayIeigh fading channeIs, vhen
lhe efficiency of resources usage is considered in lhe syslen design.
When lhere are N cognilive users, lhe nelvork efficiency is

( )
1 Q
Q
1
q
÷
= (7)

The oplinaI nunler of cognilive radios is defined ly veighling lhe sensing perfornance
and lhe nelvork efficiency in a largel funclion

( ) ( ) ( ) (1 )
G
- Q 3 Q Q o o q = ÷ + (8)
vhere ( ) - Q is lhe largel funclion, o is lhe inpacl faclor, ( )
G
3 Q is lhe prolaliIily of
deleclion, vhere n is lhe nunler of cognilive users parlicipale in cooperalive sensing.

0 1
( ) ( )(1 ) ( )(1 )
G I P
3 Q 3 + 4 3 + 4 = ÷ + ÷
(9)
I
4 and
P
4 are lhe prolaliIily of faIse aIarn and niss deleclion respecliveIy.

www.intechopen.com
Cognitive Radio Systems 46
1 0
0 1
( , )
( , )
I
P
4 3 + +
4 3 + +
= ¦
¦
´
=
¦
¹
(1O)

5. Sensing Information Transmission

Mosl presenl researches on cooperalive speclrun sensing are lased on lhe foIIoving
assunplions: lhe reporling channeI lelveen lhe fusion cenlre and cognilive users is an ideaI
error-free lransnission channeI. ßul sonelines il is very difficuIl in praclice lo achieve lhe
alove-nenlioned lransnission condilions, lhe cognilive user´s dislance fron lhe fusion
cenlre nay le far avay, or lhere are olslacIes in lhe lransnission lIock, lhen lhe sensing
infornalion lransnission are IikeIy lo le vrong.
When lhere exisl reporling errors, lhe upper and Iover Iinils of lhe faIse aIarn prolaliIily
vas given in (Sun el aI., 2OO7), lul lhe soIulions vere nol given. When lhe fusion ruIes are
AND and OR IogicaI operalion, lhe prolaliIily of faIse aIarn and niss deleclion are

( )( )
( ) ( )
, , , ,
1
, , , ,
1
1 1 1
1 1
1
25
I I L H L I L H L
L
1
25
P P L H L P L H L
L
4 3 3 3 3
4 3 3 3 3
=
=
¦
(
= ÷ ÷ ÷ +
¦
¸ ¸
¦
´
¦
(
= ÷ + ÷
¸ ¸
¦
¹
[
[
(11)
( ) ( )
( ) ( )
, , , ,
1
, , , ,
1
1 1
1 1 1
1
$1'
I I L H L I L H L
L
1
$1'
P P L H L P L H L
L
4 3 3 3 3
4 3 3 3 3
=
=
¦
(
= ÷ + ÷
¦
¸ ¸
¦
´
¦
(
= ÷ ÷ ÷ +
¸ ¸
¦
¹
[
[
(12)
Suppose lhal lhe IocaI speclrun sensing conducled ly cognilive user i resuIls in
, I L I
3 3 =
and
, P L P
3 3 = , for aII 1, 2,..., L . = , and lhal lhe prolaliIilies of reporling errors are
idenlicaI for aII CRs
, H L H
3 3 = , lhen lhe prolaliIily of faIse aIarn and niss deleclion can le
lounded ly

( )
0
lim 1 1
I
I
1
25
I H H I
3
4 4 4 3 13
÷
> = ÷ ÷ ~
(13)
( )
0
lim 1 1
P
1
$1'
P P H H P
3
4 4 4 3 13
÷
> = ÷ ÷ ~
(14)

Il has leen shovn lhal using nuIlipIe cognilive users nay inprove lhe deleclion prolaliIily
over reaIislic sensing and reporling channeIs. ßul lhe perfornance is Iiniled ly lhe
prolaliIily of reporling error
H
3 vhich is due lo inperfecl reporling channeIs.
In (Canesan & Li, 2OO8), lvo schenes have leen anaIyzed considering sensing and reporling
al lhe sane line. The schenes enpIoy lhe AnpIify-and-forvard (AI) cooperalion prolocoI
lo reduce lhe deleclion line. Il vas shovn lhal cooperalion lelveen cognilive users
increases lhe overaII agiIily of lhe nelvork.
www.intechopen.com
Cooperative Spectrum Sensing 47
Cooperalive speclrun sensing laking inlo accounl lhe Decode-and-forvard (DI)
cooperalion prolocoI on lhe decision reporling is sludied in (Zhang & Lelaief, 2OO8). A
lransnil diversily nelhod is given vhich appIies sone exisling space-line (ST) coding and
space-frequency (SI) coding for nuIlipIe anlennas syslens lo lhe CR nelvork ly vieving
cognilive users as dislriluled anlenna arrays. Il vas found lhal cooperalive speclrun
sensing vilh DI cooperalion prolocoI couId increase lhe deleclion prolaliIily vilh
reporling errors.
Space-line coding can le used for cooperalive speclrun sensing lul lhe perfornance is
conslrained lecause lhe infornalion exchange lelveen cognilive users nay nol le correclIy
perforned. Nelvork coding (NC) adnils a Iarger rale region lhan rouling al poIynoniaI
conpIexily. We propose a noveI cooperalive speclrun sensing aIgorilhn lased on nelvork
coding (NC-CSS) ly considering lhe use of nelvork coding for such infornalion exchanges.
The sinuIalion resuIls shov lhal lhe perfornance of lhe NC-CSS aIgorilhn is leller lhan lhe
alove cooperalive speclrun sensing aIgorilhn lased on space-line coding (ST-CSS), and
lhe noveI aIgorilhn can neel lhe silualion vhen lhere exisl deslroyed reporling palhs.
When lhere exisls one deslroyed reporling palh, cooperalive speclrun sensing vilh
nelvork coding can achieve leller sensing perfornance, as in Iig.7.

Iig. 7. ROC of NC-CSS aIgorilhn

Al presenl, lhere are sliII nany issues unresoIved in non-ideaI reporling channeI condilions,
hov lo consider lhe synergy of lhe design speclrun sensing aIgorilhn is parlicuIarIy
inporlanl.

www.intechopen.com
Cognitive Radio Systems 48
6. ConcIusion

Wilh lhe increasing denand for radio speclrun on one hand and inefficienl usage of lhe
Iicensed lands on lhe olher, cognilive radio can efficienlIy uliIize lhe speclrun hoIes of lhe
Iicensed channeIs in differenl Iocalions and lines. To delecl lhe speclrun hoIes accuraleIy
and quickIy, speclrun sensing is a crilicaI conponenl in cognilive radio syslens. ßy
aIIoving a nunler of cognilive users lo perforn IocaI speclrun sensing independenlIy and
fusing lheir IocaI decision resuIls logelher al lhe fusion cenlre, lhe speclrun sensing
perfornance can le grealIy enhanced.
In lhis chapler ve invesligale lhe nain issues associaled vilh lhe design of cooperalive
speclrun sensing funclionaIily for cognilive radio syslen. Sensing perfornance is decided
ly lhe IocaI decisions and fusion ruIes in lhe fusion cenlre. The nelvork overhead for
cooperalion is aIso inporlanl for cooperalive speclrun sensing, a good cooperalive
speclrun sensing aIgorilhn shouId have leller sensing perfornance vilh Iess nelvork
overhead. Cooperalive speclrun sensing aIgorilhns using space-line coding and nelvork
coding are aIso discussed vhen lhe reporling channeI is inperfecl.
An inporlanl venue for furlher research is lhe design of cooperalive speclrun sensing
considering lhe posilions and noliIily of cognilive users. Iurlher research on cooperalive
speclrun sensing can le envisioned on videland sensing.

7. References

AaIo V & Visvanalhan R.(1992) Asynplolic perfornance of a dislriluled deleclion
syslen in correIaled Caussian noise, |||| TRANSACT|ONS ON S|GNA|
PROC|SS|NG, VoI. 4O, pp. 211-213, 1992.
AkyiIdiz I I, Lee W Y, Vuran M C & Mohanly S.(2OO6) Nexl generalion/dynanic
speclrun access/cognilive radio vireIess nelvorks:a survey, ||S|V||R
COMPUT|R N|TIORKS ]OURNA|, VoI. 5O, No. 13, pp. 2127-2159.
ArsIan H & Yucek T.(2OO7) Speclrun Sensing for Cognilive Radio AppIicalions, Ccgni|itc
Raúic, Scf|uarc Dcfincú Raúic, anú Aúap|itc Iirc|css Sqs|cms, pp.263-289, Springer.
ßIun R. (1999) Dislriluled Deleclion for Diversily Receplion of Iading SignaIs in Noise,
|||| TRANSACT|ONS ON |N|ORMAT|ON TH|ORY, VoI. 45, No. 1, pp.158-164.
Chair Z & Varshney I K.(1986) OplinaI Dala Iusion in MuIlipIe Sensor Deleclion
Syslens. |||| TRANSACT|ONS ON A|ROSPAC| AND |||CTRON|C SYST|MS,
VoI. ALS-22, No. 98,pp. 98-1O1.
ChanlerIand } & VeeravaIIi V. (2OO4) The inpacl of fading on decenlraIized deleclion in
pover conslrained vireIess sensor nelvorks, |CASSP 2004,pp.837-84O.
Chen Y.(2OO8) Oplinun Nunler of Secondary Users in CoIIaloralive Speclrun Sensing
Considering Resources Usage Lfficiency. |||| COMMUN|CAT|ONS ||TT|RS,
VoI. 12, No. 12, pp.877-879.
Dighan I.I, AIouini M.S, and Sinon M.K.(2OO7) On lhe Lnergy Deleclion of Unknovn
SignaIs Over Iading ChanneIs, |||| TRANSACT|ONS ON COMMUN|CAT|ONS,
VoI.5O, No.1, pp. 21-24.
Ldvard I & Liang Y C.(2OO7) Oplinizalion for Cooperalive Sensing in Cognilive Radio
Nelvorks, ICNC 2007, pp.27-32.
www.intechopen.com
Cooperative Spectrum Sensing 49
Canesan C & Li Y.(2OO7) Cooperalive Speclrun Sensing in Cognilive Radio, Iarl I: Tvo
User Nelvorks|}j. |||| TRANSACT|ONS ON I|R|||SS COMMUN|CAT|ONS,
VoI.6, No.6, pp. 22O4-2213.
Taherpour A, Nasiri-Kenari M, }anshidi A.(2OO7) Lfficienl Cooperalive Speclrun Sensing
in Cognilive Radio Nelvorks, P|MRC2007.
Chaseni A & Sousa L.(2OO5) CoIIaloralive Speclrun Sensing for Opporlunislic Access in
Iading Lnvironnenls|Cj. ILLL DySIAN 2OO5, pages: 131 - 136.(Chaseni &
Sousa, 2OO5)
HashIanoun W. A & Varshney I K.(1999) Near-Oplinun Quanlizalion for SignaI
deleclion|}j. |||| TRANSACT|ONS ON COMMUN|CAT|ONS, VoI. 44, No. 3,
pp.294-297.
HeIslron C. (1988) Inproved MuIliIeveI Quanlizalion for Deleclion of Narrovland
SignaIs|}j. |||| TRANSACT|ONS ON A|ROS|AC| AND |||CTRON|C
SYST|MS , VoI. 24, No. 2, Ip.141-147.
KaIigineedi I, Khallazian M, ßhargava V.(2OO8) Secure Cooperalive Sensing Techniques
for Cognilive Radio Syslens. |CC 2008.
Lin T }, Zhang R, Liang Y C & Zeng Y. (2OO8) CLRT-ßased Speclrun Sensing for
Cognilive Radio.G|O8|COM 2008.
Ma }, Zhao C & Li Y.(2OO8) Sofl Conlinalion and Deleclion for Cooperalive Speclrun
Sensing in Cognilive Radio Nelvorks. |||| TRANSACT|ONS ON I|R|||SS
COMMUN|CAT|ONS, VoI. 7, No. 11, pp.45O2-45O7.
McHenry M A.(2OO5) NSI Speclrun Occupancy Measurenenls Irojecl Sunnary, shared
speclrun co. reporl.
MiloIa } & Maguire Q.(1999) Cognilive radio: Making soflvare radios nore personaI,
|||| P|RSONA| COMMUN|CAT|ONS, VoI. 6, No. 4, pp. 13-18.
Quan Z, Cui S & Sayed A H. (2OO8) OplinaI Linear Cooperalion for Speclrun Sensing in
Cognilive radio Nelvorks. |||| ]OURNA| O| S|||CT|D TOP|CS |N S|GNA|
PROC|SS|NG, VoI. 2, No.1, pp.28-4O.
SeIen Y, TuIIlerg H, Kronander }.(2OO8) Sensor SeIeclion for Cooperalive Speclrun
Sensing. DYSPAN 2008, pages: 1-11.
Sun C, Zhang W & Lelaief K ß.(2OO7) CIusler-ßased Cooperalive Speclrun Sensing in
Cognilive Radio Syslens. |CC 2007, pp.2511-2515.
Sun Y, Hu H, Liu I el aI.(2OO8) SeIeclion of Sensing Nodes in Cognilive Radio Syslen
ßased on CorreIalion of Sensing Infornalion. CROINCOM 2008.
Unnikrishnan } & VeeravaIIi V.(2OO8) Cooperalive Sensing for Irinary Deleclion in
Cognilive Radio. |||| ]OURNA| O| S|||CT|D TOP|CS |N S|GNA|
PROC|SS|NG, VoI. 2, No. 1, pp.18-27.
Visolsky L, Kuffner S & Ielerson R.(2OO5) On coIIaloralive deleclion of TV lransnissions
in supporl of dynanic speclrun sensing, Prcc. |||| DYSPAN, USA, pp. 338-345.
Zhang W & Lelaief K ß.(2OO8) Cooperalive Speclrun Sensing vilh Transnil and ReIay
Diversily in Cognilive Radio Nelvorks|}j. |||| TRANSACT|ONS ON I|R|||SS
COMMUN|CAT|ONS, VoI.7, No.12, pp. 4761-4766.
Zhang W, MaIIik R K & Lelaief K ß.(2OO8) Cooperalive Speclrun Sensing Oplinizalion in
Cognilive Radio Nelvorks. |CC 2008.
Zheng X, Wang }, Wu Q & Chen }.(2OO8) Cooperalive Speclrun Sensing AIgorilhn lased
on Denpsler-Shafer Theory, |CCS 2008, pp.218-221.
www.intechopen.com
Cognitive Radio Systems 50
Zheng X, Wang }, Wu Q & Shen L. (2OO9) A NoveI Cooperalive Speclrun Sensing
AIgorilhn in Cognilive Radio Syslens, ]OURNA| O| COMMUN|CAT|ONS AND
N|TIORKS, VoI.11,No.2,ApirI,2OO9,pp:115-121. (Zheng el aI., 2OO9)
www.intechopen.com
Cognitive Radio Systems
Edited by Wei Wang
ISBN 978-953-307-021-6
Hard cover, 340 pages
Publisher InTech
Published online 01, November, 2009
Published in print edition November, 2009
InTech Europe
University Campus STeP Ri
Slavka Krautzeka 83/A
51000 Rijeka, Croatia
Phone: +385 (51) 770 447
Fax: +385 (51) 686 166
www.intechopen.com
InTech China
Unit 405, Office Block, Hotel Equatorial Shanghai
No.65, Yan An Road (West), Shanghai, 200040, China
Phone: +86-21-62489820
Fax: +86-21-62489821
Cognitive radio is a hot research area for future wireless communications in the recent years. In order to
increase the spectrum utilization, cognitive radio makes it possible for unlicensed users to access the spectrum
unoccupied by licensed users. Cognitive radio let the equipments more intelligent to communicate with each
other in a spectrum-aware manner and provide a new approach for the co-existence of multiple wireless
systems. The goal of this book is to provide highlights of the current research topics in the field of cognitive
radio systems. The book consists of 17 chapters, addressing various problems in cognitive radio systems.
How to reference
In order to correctly reference this scholarly work, feel free to copy and paste the following:
Jinlong Wang, Qihui Wu, Xueqiang Zheng and Juan Chen (2009). Cooperative Spectrum Sensing, Cognitive
Radio Systems, Wei Wang (Ed.), ISBN: 978-953-307-021-6, InTech, Available from:
http://www.intechopen.com/books/cognitive-radio-systems/cooperative-spectrum-sensing

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close