TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 12, No. 10, Octobe
r 20
14, pp. 7343
~ 735
2
DOI: 10.115
9
1
/telkomni
ka.
v
12i8.520
2
ļ®
7343
Re
cei
v
ed
De
cem
ber 6, 20
13; Re
vised Ju
ly 18, 20
14; Acce
pted Au
gust 2, 201
4
Provably Secur
e
and Efficient ID-based Strong
Designated Verifier Signature Scheme with Messag
e
Recovery
Min Li
Sichu
an Norm
al Univ
ersit
y
of
Chin
a, Che
n
g
du, Sichu
a
n
Univers
i
t
y
of Electronic Sci
enc
e and T
e
chno
l
o
g
y
of Chi
n
a
email: lm_t
urni
p@1
26.com
A
b
st
r
a
ct
Many ID-bas
e
d
strong des
ig
nated ver
i
fier
sign
at
ure sche
m
es h
a
ve b
e
e
n
prop
osed i
n
recent
years. How
e
ve
r, most of the
m
d
i
d
not
giv
e
the rig
o
rous s
e
curity pro
o
fs a
nd d
i
d n
o
t sati
sfy the strong
n
e
ss
prop
erty that anyo
ne exc
ept
the des
ig
nate
d
verifier ca
nn
ot check the
valid
ity of a desig
nate
d
verifie
r
sign
ature, In
a
dditi
on, c
onsi
d
erin
g so
me s
p
ecial
a
p
p
licati
o
ns, these
sch
emes
hav
e l
a
r
ger
data
si
z
e
of
communic
a
tio
n
.
T
o
overc
o
me thos
e
prob
l
e
ms,
exp
l
oiti
n
g
mess
age
r
e
covery
tech
n
i
qu
es w
h
ic
h
are
regar
ded
as
a
useful
metho
d
to shorte
n ID-
base
d
si
gnat
ur
es'
si
z
e
, w
e
p
u
t forw
ard a
n
efficient ID-
bas
ed
strong des
ig
na
ted verifier si
gn
ature sche
m
es
w
i
th messag
e
recovery a
nd g
i
ve its rigor
ous
security pro
o
f i
n
the ran
d
o
m
or
acle
mod
e
l b
a
s
ed o
n
th
e h
a
rd
ness ass
u
mp
ti
ons of th
e co
mputatio
na
l Bil
i
n
ear D
i
ffie-He
ll
ma
n
prob
le
m i
n
thi
s
pap
er. T
o
th
e best
of our
know
led
ge,
it
i
s
the first ID-b
ased stro
ng
d
e
sig
nate
d
verif
i
er
sign
ature sch
e
m
es w
i
th mess
age rec
o
very a
nd rig
o
rous s
e
curity proofs. D
ue to its mer
i
ts, it can be use
d
in
some s
peci
a
l
e
n
viro
nments w
here
the
ba
nd
w
i
dth is
one
of
the
ma
in c
onc
erns, suc
h
as
PDAs, cell
p
h
o
nes,
RFID etc.
Ke
y
w
ords
:
ide
n
tity-base
d
publ
ic key cr
yptogra
phy, d
e
sig
nate
d
veri
fier sign
ature,
mess
age r
e
c
o
very,
bili
ne
ar pair
i
n
g
s
, rando
m orac
le mod
e
l
Copy
right
Ā©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
Digital
sig
nat
ure
a
s
o
ne i
m
porta
nt p
r
i
m
itiv
e
in cry
p
togra
phy, which
can pro
v
ide
data
integrity, authenticatio
n an
d non-rep
udi
ation, has m
any pra
c
tical
appl
i
c
ation
s
in the real wo
rld,
su
ch a
s
el
ectro
n
ic
co
mmerce, ele
c
troni
c g
o
ve
rnme
nt etc.
Ho
wever, i
n
som
e
sp
ecial
environ
ment
s, signatu
r
e
s
with spe
c
ial
prop
ertie
s
a
r
e always
de
sirable. F
o
r
example, in
so
me
scena
rio
s
su
ch a
s
E-votin
g
, call fo
r ten
ders an
d
software licen
sin
g
, the publi
c
verification
of an
ordin
a
ry sign
ature
i
s
n
o
t desi
r
ed, sin
c
e
the sign
er
may not wan
t
to the re
cip
i
ent of a di
gi
tal
signature to transfer
the conviction to a third party at will.
To
a
ddress
t
h
is problem
above, Cha
u
m
an
d Van
Antwerpen
in
trodu
ced
un
d
eniabl
e
sign
ature
s
[2
, 3] whi
c
h
allowe
d a
sig
n
e
r to
com
p
le
tely control h
i
s
signatu
r
e
s
.
In und
enia
b
l
e
sign
ature
s
, the verifie
r
(B
ob)
can
not
che
c
k t
he va
lidity of the signature give
n by the si
g
ner
(Alice
)
by him
s
elf. Instea
d, Alice pa
rtici
p
ates in
th
e scheme to
prov
e the validity (or i
n
validity)
of
the si
gnatu
r
e
to Bob
by
m
ean
s of
an
in
teractive
protocol.
Neve
rth
e
less, Ali
c
e
can o
n
ly de
cid
e
whe
n
to prove, but not who to verify. Hence,
the co
nviction ca
n be tran
sferre
d to anyone else.
Motivated by
the a
bove
p
r
oble
m
, Jako
bsson
et
al.
[4] introdu
ce
d the
co
ncep
t of de
sign
ated
verifier si
gnat
ure (DVS)
scheme in Eu
rocrypt 1
996.
A DVS sch
eme ma
ke
s i
t
possible for a
sign
er Alice to convin
ce a
desig
nated
verifier
Bob t
hat Alice ha
s signe
d a me
ssage in
su
ch a
way that Bob can not transfe
r
the conviction to a third party
Cindy. This is called n
on-
transfe
ra
bility, and is usu
a
l
l
y achieved b
y
enabling
Bob the cap
a
b
ility of efficiently simulatin
g
a
sign
ature
whi
c
h is in
distin
g
u
ish
able fro
m
Alice's.
In ord
e
r to e
nhan
ce
the
signer's priva
cy, Ja
kob
s
son
et al. al
so
introdu
ce
d a
stron
g
e
r
versio
n of DVS in the same wo
rk [4]. It is us
ually calle
d stro
ng
desig
nated
verifier si
gnat
ure
(SDVS) sch
e
m
e, in
whi
c
h
no
third
pa
rt
y can
eve
n
check th
e vali
dity of a
de
si
gnated
verifi
er
sign
ature, si
n
c
e the verification of the si
gnatur
e re
qui
res the
desi
g
nated verifie
r
'
s
private
key.
Evaluation Warning : The document was created with Spire.PDF for Python.
ļ®
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 734
3
ā 7352
7344
Since th
e n
o
tion of S
D
VS
prop
osed
by
Ja
kob
s
son
et
al. in [4], m
any SDVS
scheme
s
have b
een
p
u
t forward in
the literatu
r
e.
In 20
03,
Sae
ednia
et al. [
5
] firstly form
alize
d
the
not
ion
of SDVS an
d pro
p
o
s
ed
an efficient
schem
e in th
e sam
e
pap
er. In 2004,
Susilo et al.
[6
]
prop
osed the
first stron
g
desi
gnate
d
verifier
sign
ature
sch
eme
in identity-ba
sed p
ubli
c
key
crypto
system
that was first
introdu
ce
d b
y
Shamir
[1] in 1984 to
sol
v
e the probl
e
m
s of certificate
manag
eme
n
t in public
key
infrastructu
re (PKI). D
ue
to its advanta
ge in co
ntra
st
to PKI, several
new I
D
-b
ase
d
SDVS (IBSDVS) have
b
een p
r
op
osed
in the ne
w
setting re
centl
y
. In 2008, Z
hang
et al. [8] p
r
o
posed
a n
o
vel IBSDVS schem
e
by co
mbining ID-
b
as
ed public
key cr
yptos
y
s
t
em
with the de
sig
nated verifie
r
sign
ature. In their
work, the
y
claimed tha
t
their schem
e wa
s a st
ron
g
desi
gnate
d
v
e
rifier si
gnatu
r
e, that i
s
to
say,
no thi
r
d
p
a
rty can
ch
eck the
validity
of a
de
signat
ed
verifier
sign
ature
gene
rate
d by the
sign
er. In 20
09,
h
o
weve
r, Kan
g
et al. [9] found that Z
h
a
ng et
al.'s sche
me can n
o
t satisf
y the strongn
ess pro
per
ty as they claim
ed in [8]. In
the sam
e
pap
e
r
[9], they presented a ne
w IBSDVS sch
eme and I
D
-
based de
sig
n
a
ted verifier
proxy sign
atu
r
e
s
c
heme
(IBDVPS) based
on the new I
BSDVS s
c
heme.
In the meanwhile, they als
o
put forward
a novel
IBSDVS scheme [10] with security proofs
i
n
t
he random
oracle
model based on Bilinear
Diffie-Hellma
n
a
s
sumptio
n
. Unfo
rtunat
ely, Lee
et
al. [11] sho
w
ed
that Ka
ng et
al.'s
n
e
w
scheme
s
in [
9
] are u
n
iversally forgea
ble
in 2010,
that is,
anyon
e
ca
n
gen
erate a sign
ature on an
arbitrarily ch
o
s
en me
ssag
e
without the secret key
of e
i
ther the sig
n
e
r or the d
e
si
gnated verifie
r
.
To overcom
e
these flaws,
they also pres
ented a new IBSDVS and IBDVPS schem
e and give
the formal
se
curity p
r
oof
s
in the rand
o
m
ora
c
le
mo
del [14, 1
5
] in the
same
pape
r. In 20
08,
Hua
ng et al. [
7
] also
propo
sed
an effici
e
n
t IBSD
VS scheme
whi
c
h i
s
secure ba
sed on
a
stron
ger
assumption, i
.
e. Gap Bilinear Diffie-Hel
l
man, the si
ze of the
signature in
their scheme i
s
v
e
ry
sho
r
t co
mpa
r
ed to all the
existing sch
e
m
es. Howeve
r, the sig
natu
r
e in thei
r scheme i
s
short o
f
rand
omi
c
ity beca
u
se the signatures o
n
the same
me
ssage a
r
e al
ways ide
n
tica
l in every time
sign
ature g
e
n
e
ration p
r
o
c
e
dure.
As
far as
we
k
n
ow, all these exis
ting
sc
heme
s
exce
pt the schem
es
in [7, 11, 16]
can
not
sup
port
the
strongn
ess property
of the
SDVS, and
the signi
ng m
e
ssag
es
in al
l
these sche
mes
alway
s
nee
d to be tran
smi
tted together
with t
he sig
n
a
ture
s. Thu
s
, these sch
e
m
e
s have a la
rge
total commu
nicatio
n
co
st, for whi
c
h they maybe
can n
o
t efficiently used i
n
som
e
sp
e
c
ial
environ
ment
s where low- communi
catio
n
and lo
w-co
mputation co
st are u
s
ually
requi
red.
To solve th
ose
above
probl
em
s, combinin
g th
e messa
g
e
recovery t
e
ch
niqu
es
pre
s
ente
d
in
[12], we firstl
y put forwa
r
d
an e
ffic
i
ent IBSDVS s
c
heme with mess
age rec
o
very
(IBSDVSMR), and prove
its se
cu
rity in the ran
d
o
m
ora
c
le mo
del ba
sed
o
n
Bilinear
Di
ffie-
Hellma
n
a
s
su
mption. In th
e prosed
sch
e
me, t
he m
e
ssage i
s
e
m
b
edde
d in a
si
gnature an
d
can
be re
cove
re
d from the
received
sig
nature.
Hen
c
e, it results in more b
and
width sa
ving.
Obviou
sly, it
has the a
d
va
ntage of smal
l data size of comm
uni
cati
on.
Due to its m
e
rits, the pro
posed sch
e
m
e
in th
is pap
er is not onl
y quite efficient with
respe
c
t to
co
mputation
co
st, but al
so
very small
with re
sp
ect to
the total l
engt
h of the
si
gni
ng
messag
e an
d the ap
pen
ded
signatu
r
e, i.e. co
mm
unication cost.
For
these advantag
es, the
prop
osed sch
e
me with me
ssage recove
ry is very
use
f
ul for the environme
n
ts where b
and
wid
t
h
is on
e of the
main
con
c
e
r
n
s
. Fo
r in
stan
ce, on
wi
rele
ss devices (e
.g. PDAs, cel
l
phon
es,
RFI
D
chip
s and
se
nso
r
s) wh
ere battery
life
i
s
the
mai
n
limit
ation, commu
nicatin
g
eve
n
one
bit
of da
ta
usu
a
lly use
s
signifi
cantly more
power t
han exe
c
utin
g one 3
2
-bit instru
ction [1
3
]. Redu
cing t
he
numbe
r of co
mmuni
cation
bits save
s po
wer a
nd is
im
portant for in
crea
sing the b
a
ttery life.
The rest
of this pa
pe
r is org
ani
zed
as follo
ws. In se
ction
2,
we first giv
e
so
me
prelimi
nari
e
s,
inclu
d
ing bili
near
map
s
a
nd some
rela
ted hard p
r
ob
lems, the M
o
del of ID-ba
s
ed
stron
g
d
e
si
gn
ated
verifie
r
signature sche
me wi
th me
ssag
e recovery and
so
on,
and th
en
we
put
forwa
r
d
an
ef
ficient
con
c
re
te schem
e a
nd give it
s
se
curity p
r
o
o
f i
n
the
ran
dom
ora
c
le
mod
e
l
in
Section 3 an
d
sectio
n 4, re
spe
c
tively; In
Secti
on 5, we
compa
r
e o
u
r sch
eme
s
wit
h
some exi
s
ting
scheme
s
p
r
e
s
ente
d
in [8, 10, 11]; Finally, we end this paper
with a brief co
ncl
u
si
on.
2. Rese
arch
Metho
d
In this sectio
n, we b
r
iefly review
so
me
fundame
n
tal
backg
rou
n
d
s
requi
re
d thro
ugh thi
s
pape
r, in
cludi
ng bilin
ea
r m
aps,
hardne
ss a
s
sumpti
o
n
s
, the
notion
of IBSDVSMR a
nd it
s
se
curity
definitions etc.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
ļ®
Provabl
y Secure an
d Efficient ID-ba
s
e
d
Strong Desi
g
nated Verifie
r
Signature
ā¦
(Min Li)
7345
2.1. Some Nota
tions
For co
nvenie
n
ce, we
list
some notation
s
wi
th thei
r
meanin
g
s through
out thi
s
pape
r i
n
Table 1.
Table 1. Nota
tions an
d thei
r Meani
ng
s
Notation Meanin
g
||
ab
a concatenation of t
w
o strings a a
nd b
ļ
X-OR comp
utatio
n in the binar
y s
y
stem
2
[]
x
the binar
y notatio
n of
x
Z
ļ
10
[]
y
the decimal notation of
*
{0
,
1
}
y
ļ
1
||
l
ļ”
the first
1
l
bits of
ļ”
fro
m
the right side
2
||
l
ļ”
the first
2
l
bits of
ļ”
fro
m
the left side
2.2. Bilinear Pairings
Let
1
(,
)
G
ļ«
be a cyclic ad
ditive grou
p of prime o
r
de
r
q
and
2
(,
)
G
ī
be a cycli
c
multiplicative
group
of th
e same
o
r
de
r
q
,
12
||
ql
l
ļ½
ļ«
. We assu
me that the
discrete lo
ga
rithm
probl
em
s in b
o
th
1
(,
)
G
ļ«
and
2
(,
)
G
ī
are intracta
ble. Le
t
11
2
:
eG
G
G
ļ“ļ®
be a bilinear map with the
following properties
:
(1) Bilinear:
(,
)
(
,
)
,
ab
ea
Pb
Q
e
PQ
ļ½
for all
1
,
PQ
G
ļ
, and
*
,
q
ab
Z
ļ
.
(2)
Non
-
de
ge
nerate: Th
ere
exists
1
PG
ļ
su
ch t
hat
(,
)
1
eP
P
ļ¹
.
(3)
Comp
utab
le: There exi
s
ts an efficient
algorithm to
comp
uter
(,
)
eP
Q
for any
1
,
PQ
G
ļ
.
A bilinea
r ma
p satisfying t
he p
r
op
ertie
s
above i
s
na
med a
n
a
d
mi
ssi
ble bili
nea
r ma
p. It
can
be o
b
tai
ned fro
m
the
modified
Weil and T
a
te
pairin
g
s. F
o
llowe
d by are the ha
rdn
e
ss
assumptio
n
s
use
d
in the secu
rity proof
s:
Defini
tion 1.
Bilinea
r
Diffie-Hell
man
Problem:
Th
e BDH
pro
b
l
em i
s
to
compute
r
(,
)
abc
eP
P
whe
n
given
(,
,
,
)
Pa
P
b
P
c
P
for so
me un
kn
own
*
,,
q
abc
Z
ļ
.
Defini
tion 2.
Bilinea
r
Diffie-Hellma
n
a
s
sumptio
n
: Su
ppo
se th
at G
is
a B
D
H pa
ramete
r
gene
rato
r,
Adv
G
(B
) is the
advantage t
hat an algo
ri
thm
B has in solving the
BDH proble
m
.
Adv
G
(B) i
s
de
fined to be the prob
abilit
y that the algorit
hm B outputs
(,
)
abc
eP
P
when the in
puts to
algorith
m
are
12
,,
,
,
,
GG
P
a
P
b
P
c
P
, where
12
(,
,
)
GG
e
is the output of G
for sufficie
n
tly large se
cu
rit
y
para
m
eter
k
,
P
is a rando
m gene
rator
of
1
G
, and
,,
ab
c
are rando
mly ch
ose
n
from
*
q
Z
. The
BDH a
s
sump
tion is that
Adv
G
(B) is n
egli
g
ible for all ef
ficient algo
rithms B.
2.3. Definition of IBSDVS
MR
An IBSDVS schem
e with
messag
e recovery is
a tu
p
l
e of five polynomial time
a
l
gorithm
s
as
follows
:
Setup:
This
algorith
m
takes a security param
eter
k
as input an
d returns the
system
para
m
eters
Pa
r
a
ms
an
d a se
cret ma
ste
r
key
m
a
ster-key
.
Key
Extrac
tion:
T
h
is
alg
o
rit
h
m t
a
ke
s
sy
st
em
pa
ra
met
e
r
s
Params
,
m
a
ster-key
and
a
use
r
'
s
identity
i
I
D
as input, an
d then retu
rn
s the private
key
i
Sk
with res
p
ec
t to the identity
i
I
D
.
Signature
G
e
nera
tion:
This
algo
rit
h
m t
a
ke
s t
h
e
sy
st
em
pa
r
a
met
e
r
s
Params
, a
messag
e
m
, a
signe
r'
s i
dentity
A
I
D
, his
corre
s
p
ondin
g
private
ke
y
A
Sk
and the
de
sign
ated
verifier'
s
publi
c
key
B
I
D
as inp
u
ts, and then it outputs a vali
d sign
ature
ļ³
on messag
e
m
.
Signature Verification:
Th
is alg
o
rit
h
m
t
a
ke
s t
h
e
sy
st
em pa
ram
e
t
e
rs
Params
, si
gnature
ļ³
, the sign
er'
s
identity
A
I
D
, the desig
nated ve
rifier's i
dentity
B
I
D
and p
r
ivate
key
B
Sk
as in
puts. I
t
outputs
1 if the si
gnatu
r
e
is valid, the
signi
ng me
ssage can be recove
red su
cce
ssfully
in
t
h
is
ca
se. Othe
rwi
s
e outp
u
ts 0.
Evaluation Warning : The document was created with Spire.PDF for Python.
ļ®
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 734
3
ā 7352
7346
Transc
ript simulation:
The de
sig
n
a
t
ed verifier
run
s
this
al
gorithm to
prod
uce
identically distributed tra
n
script
s whi
c
h a
r
e indi
stingui
shable fro
m
the si
gn
ature g
enerated by the
sign
er.
2.4. Securit
y
propertie
s o
f
IBSDVSM
R
As define
d
in
[10, 11], the IBSDVS sche
me
with me
ssag
e re
cove
ry should
sati
sfy some
main se
cu
rity prope
rtie
s, which a
r
e de
scribed a
s
follo
ws:
a)
Corre
ctn
ess
:
A prope
rly prod
uced IBSDVSMR mu
st
be accepted
by the signature
verification al
gorithm.
b)
Non-Transferabilit
y
:
The non-transferability means
that any designated veri
fier
can n
o
t tran
sf
er the
convict
i
on to any third party,
that is, the de
sign
ated ve
rifier
can not prove
to
a third pa
rty that the sig
nat
ure
wa
s pro
d
u
ce
d by
the signer
or by hi
mself. This i
s
accompli
she
d
by a tran
scri
pt sim
u
lation
algo
rithm th
rough
wh
ich t
he d
e
si
gnate
d
verifie
r
can
produ
ce
an
in-
disting
u
ishabl
e sign
ature from the
one g
enerated by the real
sign
er.
c)
Strongn
ess
:
Given a
sign
ature, the ve
rificati
on p
r
o
c
edure requi
re
s the
se
cret key
of the desig
n
a
ted verifier, that is, any third pa
rty can n
o
t che
ck the
validity of the
sign
ature.
d)
Source hidi
ng:
Give
n a
sign
ature
on
messag
e
m
, it is
infeas
ible to tell apart the
sign
ature i
s
prod
uced by
the original
sign
er
o
r
the
desig
nated
verifier on e
a
rth even if one
kno
w
s both the se
cret key
s
.
e)
Unforgeabilit
y
:
It is com
putationally i
n
feasi
b
le to
con
s
tru
c
t
a
valid IBSDV
S
MR
without the knowl
edge of
the priv
ate key of either the sig
ner o
r
the desi
gnat
ed verifier. T
h
e
formal definiti
on of existential unforgeabilit
y of IBSDVSMR is modeled by the following game
betwe
en an a
d
versary A and a ch
allen
g
e
r C.
Game:
Th
e game is exe
c
uted b
e
twe
en a chall
e
n
ger C a
nd a
n
adaptively cho
s
e
n
-
messag
e and
cho
s
en
-ide
ntity adversa
ry A.
Setup:
Th
e
challen
ger C runs the Setu
p alg
o
ri
thm
to ge
nerate th
e sy
stem p
a
rameters
Param
s
and
the
sy
stem maste
r
key
ma
s
t
e
r
-k
ey
.
Then
he
sen
d
s
Pa
r
a
ms
to
adve
r
sa
ry A
whil
e
kee
p
s
ma
s
t
er-
k
ey
se
cret
.
Queries
:
A
a
daptively issu
es th
e foll
owi
ng q
u
e
r
ies in
a
polynomi
a
l
bou
nde
d n
u
m
ber of
times
.
1)
Ke
y
extra
cti
on qu
eries:
Whe
n
re
ceiving p
r
ivate
ke
y que
ry on
id
entity
i
I
D
, C runs
the Key Extraction alg
o
rith
m and retu
rn
s the private
key
i
Sk
.
2)
Sign queries
:
On
receiving
a sign
ature query on
m
e
ssag
e
m
for a
sign
er
i
I
D
an
d a
desi
gnate
d
verifier
j
I
D
, C run
s
the Signatu
r
e Gen
e
ratio
n
algorith
m
an
d
returns
a vali
d sig
nature
ļ³
on messa
g
e
m
.
3)
Verif
y
queries:
On
re
ceivi
ng a verify qu
ery on si
gnat
ure
ļ³
for the si
g
ner
i
I
D
and the
desi
gnate
d
verifier
j
I
D
, C run
s
the Signatu
r
e Verification
algorith
m
an
d
outputs 1 if t
he si
gnatu
r
e
is valid. In th
is case, C re
co
vers the
m
e
ssag
e from
the sig
nature
and returns it. Otherwi
se
,
outputs 0.
4)
Forgery
:
Eventually, A o
u
tputs
a tu
pl
e
**
*
(,
,
)
mh
ļ³
with th
e
sign
er
*
i
I
D
and
the
desi
gnate
d
verifier
*
j
ID
. We sa
y that A wins the game if t
he follow condi
tions are all satisfied:
(1)
*
ļ³
is a vali
d si
gnature on
messag
es
*
m
w
i
th
th
e
s
i
gn
er
*
i
I
D
and the
de
si
gnated
ver
i
fier
*
j
ID
.
(2) Duri
ng
th
e
simulation,
*
i
I
D
a
nd
*
j
ID
have n
e
v
er b
een
su
bmitted to t
he
Key
extr
action q
u
eries
.
(3)
*
m
has
never
be
en que
rie
d
d
u
ring th
e
Sign queries
wi
th the sign
er
*
i
I
D
and the
desi
gnate
d
verifier
*
j
ID
.
Definition
3. An IBSDVSMR i
s
existe
nt
ially unforg
eable
agai
nst adaptively cho
s
e
n
-
message and chosen-identity a
ttacks
if the success probabilit
y
of any polynomial bounded
adversa
ry A in the above g
a
me is ne
gligi
b
le.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
ļ®
Provabl
y Secure an
d Efficient ID-ba
s
e
d
Strong Desi
g
nated Verifie
r
Signature
ā¦
(Min Li)
7347
3. Rese
arch
Metho
d
3.1. The Efficient IBS
D
VSMR Schem
e
s
In this
part, we present the firs
t effic
i
ent IBSDVS
scheme
s
wit
h
messag
e recove
ry,
whi
c
h is not
only much
efficient but also
can
sup
port the stro
ngne
ss pro
p
e
rty. In this
new
sign
ature
sch
eme,
it can
deal with me
ssage
s
of
so
me fixed l
e
n
g
th (i
e.,
1
{0
,
1
}
l
m
ļ
for
s
o
me
fixed integer
1
l
), each al
gorith
m
of which i
s
spe
c
ified a
s
follows:
Setup:
G
i
ve
n a
s
e
c
u
r
i
ty par
a
m
e
t
er
k
, th
e KGC ge
ne
rates
a
cycli
c
additive g
r
ou
p
1
(,
)
G
ļ«
of prim
e o
r
d
e
r
q
, a multipli
cative group
2
(,
)
G
ī
of the
sam
e
pri
m
e
ord
e
r
, and
an
ad
missi
ble
bilinea
r map
11
2
:
eG
G
G
ļ“ļ®
. The KG
C al
so cho
o
se
s fou
r
cryptograp
hic
hash fun
c
tio
n
s:
*
11
:{
0
,
1
}
H
G
ļ®
,
12
*
22
:{
0
,
1
}
{
0
,
1
}
ll
HG
ļ«
ļ“ļ®
,
12
1
:{
0
,
1
}
{
0
,
1
}
ll
F
ļ®
,
21
2
:{
0
,
1
}
{
0
,
1
}
ll
F
ļ®
a rando
m
*
q
s
Z
ļ
and a
gene
rato
r
P
of
1
G
, and co
mp
utes
Pu
b
Ps
P
ļ½
, w
h
er
e
/
Pub
s
P
is the
priv
ate/publi
c
ke
y pair of
KGC, and the
n
publi
s
he
s the syste
m
pa
ramete
rs
Params
:
12
1
2
1
2
{,
,
,
,
,
,
,
,
,
}
Pub
GG
e
q
P
P
H
H
FF
Ke
y
Extracti
on:
Given a
n
identity
i
ID
, KG
C co
mpute
s
the co
rre
sp
o
nding p
r
ivate
key
1
()
ii
Sk
s
H
ID
ļ½
i
s
Q
ļ½
, where
1
()
ii
Qs
H
I
D
ļ½
, and then
send
s
it to the use
r
with ide
n
tity
i
ID
via
a
se
cure
cha
n
nel. In thi
s
scena
rio, the
sig
ner Alice
with i
dentit
y
A
ID
ha
s the
private
key
AA
Sk
s
Q
ļ½
, and the desi
gnated verifie
r
Bob ha
s his
private key
B
B
Sk
s
Q
ļ½
.
Signature
G
e
nera
tion:
T
o
sig
n
a
me
ssag
e
1
{0
,
1
}
l
m
ļ
, the signer Alice
with private
key
A
Sk
perfo
rm
s as follows (h
ere we define
(,
)
AB
g
eS
k
Q
ļ½
and it can b
e
pre
-
comput
ed):
(1)
Cho
o
se a ra
ndom elem
e
n
t
*
q
rZ
ļ
, and com
pute
12
2
(,
,
)
{
0
,
1
}
ll
r
AB
HI
D
I
D
g
ļ”
ļ«
ļ½ļ
,
whe
r
e
A
ID
and
*
{0
,
1
}
B
ID
ļ
.
(2) Comp
ute
12
12
1
()
|
|
(
(
()
)
)
{
0
,
1
}
ll
Fm
F
F
m
m
ļ¢
ļ«
ļ½ļ
ļ
an
d
10
[]
h
ļ”
ļ¢
ļ½
ļ
.
(3) Comp
ute
()
A
Vr
h
S
k
ļ½ļ
, a
nd o
u
tput
(,
)
B
eV
Q
ļ³
ļ½
.
T
he sign
ature
on
m
e
ssa
ge
m
is
(,
)
h
ļ³
.
Signature V
e
rificatio
n:
Given sy
ste
m
paramete
r
s
Params
, identity
A
ID
and
the
sign
ature
(,
)
h
ļ³
, the desig
nated v
e
rifier Bob
with private key
B
Sk
performs
as
follows
:
(1) Comp
ute
2
'(
,
,
(
,
)
)
.
h
AB
A
B
HI
D
I
D
e
Q
S
k
ļ”ļ³
ļ½ļ
(2) Comp
ute
2
'[
]
'
h
ļ¢
ļ”
ļ½ļ
.
(3) Re
cover
the
messag
e
12
2
'|
'
|
(
|
'
|
)
ll
mF
ļ¢
ļ¢
ļ½
ļ
.
(4)
Output 1
and
accept
(,
)
h
ļ³
as a
valid si
gnatu
r
e of m
e
ssa
g
e
'(
)
mm
ļ½
if
and only if
2
1
('
)
|
'
|
l
Fm
ļ¢
ļ½
.
Transc
ript s
i
mulation:
T
he desi
gnate
d
verifier Bo
b can p
r
odu
ce the disting
u
ish
able
sign
ature
Ė
ļ³
intended for hi
mself by doing the followi
ng steps:
(1) Ran
domly
choo
se
*
Ė
q
rZ
ļ
, then compute
Ė
U
ļ½
Ė
(,
)
r
BA
eS
k
Q
and
2
Ė
Ė
(,
,
)
.
AB
HI
D
I
D
U
ļ”
ļ½
(2)
Comp
ute
12
1
Ė
()
|
|
(
(
(
)
)
)
Fm
F
F
m
m
ļ¢
ļ½ļ
and
10
Ė
Ė
Ė
[]
h
ļ”
ļ¢
ļ½ļ
.
(3) Comp
ute
Ė
ĖĖ
(,
(
)
)
B
A
eS
k
r
h
Q
ļ³
ļ½ļ
. Then
Ė
Ė
(,
)
h
ļ³
is also a
valid signatu
r
e on messa
g
e
m
.
3.2. Securit Analy
s
is
In this se
ction
,
we will give se
curity analy
s
is of ou
r pro
posed sch
e
m
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ļ®
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 734
3
ā 7352
7348
a)
Corre
ctn
ess
:
Given a
si
gnature pai
r
(,
)
h
ļ³
, the corre
c
tness of the
prop
osed
scheme can be
prove
d
as follows:
(,
)
h
AB
eQ
S
k
ļ³
ļ
=
((
)
,
)
(
,
)
h
AB
A
B
er
h
S
k
Q
e
Q
S
k
ļļ
=
(,
)
(
,
)
(
,
)
rh
h
AB
AB
A
B
eS
k
Q
eS
k
Q
e
Q
S
k
ļ
ļļ
=
(,
)
(
,
)
rh
h
BA
A
B
g
eS
k
Q
e
Q
S
k
ļ
ļļ
=
r
g
If
(,
)
h
ļ³
is a
valid sig
nat
ure, the
n
we
have
2
(,
,
)
r
AB
HI
D
I
D
g
ļ”
ļ½
and
21
[]
(
)
|
|
hF
m
ļ”
ļ¢
ļ ļ½
ļ½
21
((
(
)
)
)
F
Fm
m
ļ
. Thus
, we obtain
12
2
||
(
|
|
)
ll
Fm
ļ¢
ļ¢
ļ
ļ½
.
Finally, the integrity of
m
is justified by
2
1
()
|
|
l
Fm
ļ¢
ļ½
.
b)
Non-Transferabilit
y
:
The non-tran
sferability means t
hat the design
a
ted veri
fier
Bob can n
o
t prove to any
third party that the
signa
ture wa
s p
r
o
duced by the signe
r Alice
or
himself. In
our scheme, the non-t
ransferability is
achieved throug
h the
simulati
on algorithm.
In
particula
r, su
ppo
se
('
,
'
)
h
ļ³
is a signature whi
c
h is rand
oml
y
chose
n
fro
m
the set of
all valid
sign
ature
s
int
ende
d to Bob
,
then the probability
P
r[(
,
)
(
'
,
'
)]
1
/
(
1
)
hh
q
ļ³
ļ³
ļ½
ļ½ļ
since
('
,
'
)
h
ļ³
is
gene
rated
fro
m
a rand
oml
y
cho
s
en
val
ue
*
q
rZ
ļ
. Similarly, it is easy
to get that the
probability
Ė
Ė
Pr
[
(
,
)
(
'
,
'
)
]
1
/
(
1
)
hh
q
ļ³ļ³
ļ½ļ½
ļ
. This
mea
n
s th
at tran
scripts si
mul
a
ted by Bo
b an
d the
sign
ature
s
ge
nerate
d
by Alice hav
e the i
dentical distri
bution, so
the
y
are indistin
guishable fro
m
each other.
c)
Strongn
ess
:
Since
any i
n
formatio
n a
bout the
priv
ate keys
can
not be
obtai
ned
from the tran
script
(,
)
h
ļ³
in our
scheme
and
Bob's
private
key is
req
u
i
r
ed in th
e ve
rification
pro
c
ed
ure, any third party
wit
hout the private key
will be un
abl
e to che
ck t
he validity of the
sign
ature
(,
)
h
ļ³
. Thus, the strong
ness prope
rty is
achi
eved i
n
our p
r
op
osed schem
e.
d)
Source hidi
ng:
F
r
om th
e
sign
ature ge
neratio
n a
nd
simulatio
n
al
gorithm
s, we
can
say that even
if the third party kno
w
s b
o
th t
he sig
n
e
r
and the
de
signated ve
rifier'
s
private
keys,
she still
can not identify whether
the signature i
s
generated
by
the original si
gner or
the
desi
gnate
d
verifier o
n
eart
h
. Th
is is d
u
e
to the fact th
at:
((
)
,
)
(
(
)
,
)
AB
B
A
er
h
S
k
Q
er
h
S
k
Q
ļ³
ļ½ļ
ļ½ļ
Whe
r
e
22
(,
,
(
,
)
)
(
,
,
(
,
)
)
rr
AB
A
B
AB
B
A
h
H
ID
ID
e
S
k
Q
H
I
D
I
D
e
S
k
Q
ļ½ļ½
.
Next, we wil
l
prove that the prop
osed
sch
eme is
existentially unforg
eabl
e again
s
t
adaptively ch
ose
n
- me
ssa
ge and
cho
s
e
n
-ide
ntit
y atta
cks ba
se
d on
the BDH a
s
sumption.
Theorem 1.
If there exi
s
ts
an ad
aptively cho
s
e
n
-m
essag
e
an
d ide
n
tity adversary A wh
o
can a
s
k at most
1
H
q
times
1
H
queries
,
2
H
q
times
2
H
qu
eries
,
E
q
times
Ke
y
extracti
on queries
,
S
q
times
Sign q
u
eries
,
V
q
times
Verif
y
queries
, re
spe
c
tively, and bre
a
k
the p
r
opo
se
d schem
e
in polynomi
a
l
time
t
with success
probability
2
10(
1
)
(
)
/
SH
S
qq
q
q
ļ„
ļ³ļ«
ļ«
, then there exi
s
ts a
n
algorith
m
C that can u
s
e A
to solve the BDH p
r
oble
m
with prob
abili
ty
()
BD
H
c
A
dv
k
in time s
pan
tā
, whe
r
e
11
1
1
1
2
22
2
()
(
1
)
(
1
)
(1
)
EV
S
qq
q
BD
H
c
HH
H
H
H
Ad
v
k
qq
q
q
q
ļ„
ļ«
ļ³ļ
ļ
ļļ
,
2
'
1
2
068
6
/
10
(
1
)
HS
tq
q
t
q
ļ³ļ«
21
*
()
(
3
)
(
2
)
,
H
SH
E
S
V
S
V
P
qq
q
q
q
q
t
q
q
t
ļÆ
ļÆ
ļ«ļ«
ļ«
ļ«
ļ«
ļ«
ļ«
*
t
is
the time
to
comp
ute a scala
r
multipli
cation in
1
G
, and
P
t
is the tim
e
to com
pute
a pairin
g
op
eration
on
12
(,
)
GG
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
ļ®
Provabl
y Secure an
d Efficient ID-ba
s
e
d
Strong Desi
g
nated Verifie
r
Signature
ā¦
(Min Li)
7349
Proof.
Let C be a BDH attacker. He re
ceives a ran
d
om instan
ce
(,
,
,
)
Pa
P
b
P
c
P
o
f
t
he
BDH
pro
b
lem
,
his g
oal i
s
t
o
compute
(,
)
ab
c
eP
P
after interactin
g with A i
n
th
e above
Ga
me
. In
our setting,
1
H
and
2
H
are both
regarded a
s
random o
r
a
c
le
s.
Setup:
Firs
t, C sets
Pu
b
Pc
P
ļ½
, and ge
nerates the sy
stem
paramete
r
s
Params
=
12
{,
,
,
,
,
,
Pub
GG
e
q
P
P
12
1
2
,,
,
}
HH
F
F
runni
ng the
Setup
al
gorit
hm, then he
send
s
Params
to the
adversa
ry A.
Queries
:
A
adaptively i
s
sue
s
th
e q
u
e
rie
s
to
the
followin
g
o
r
a
c
le
s in
a
pol
ynomial
boun
ded nu
mber of time
s. The
s
e ora
c
le
s are a
ll
simulate
d by C. To avoid colli
sion
s and
con
s
i
s
tently respon
d
to the qu
eries, C
maintains two li
sts
1
{,
,
}
,
Hi
i
i
LI
D
Q
r
ļ½
2
{,
,
,
}
Hi
j
LI
D
I
D
U
ļ”
ļ½
which are init
ially empty.
(1)
1
H
queries
:
Re
ceiving
an
1
H
query on
i
ID
, C firs
t sc
ans
the lis
t
1
H
L
, then retu
rns
the same a
n
s
wer in
1
H
L
if the req
u
e
s
t has bee
n asked before. Otherwi
se, C
sele
ct
s a
ran
dom
*
iq
rZ
ļ
, and answers the qu
eries a
s
follows:
,
i
f
,
,
i
f
,
,
ot
he
r
w
i
s
e
.
ii
A
ii
i
B
i
ra
P
I
D
I
D
Q
r
b
P
ID
ID
rP
ļ½
ļ¬
ļÆ
ļ½ļ½
ļ
ļÆ
ļ®
Then C a
d
d
s
(,
,
)
ii
i
ID
Q
r
to
1
H
L
. In all c
a
s
e
s
,
C returns
i
Q
a
s
the an
swer.
(2)
2
H
queries
:
Wh
en re
ceivin
g
an
2
H
query o
n
(,
,
)
ij
ID
ID
U
, C firs
t
s
c
ans
2
H
L
list ,
if the reque
st
has b
een a
s
ked b
e
fore, the sa
me an
swer i
n
the list
will be retu
rned.
Other
wis
e
, C s
e
le
cts
12
{0
,
1
}
ll
l
ļ«
ļ
at rand
om, a
nd sets
l
ļ”
ļ½
, then a
d
d
s
(,
,
,
)
ij
ID
I
D
U
ļ”
to
2
H
list
2
H
L
and
ret
u
rn
s the an
swer
ļ”
.
(3)
Ke
y
extrac
t
i
on querie
s
:
Whe
n
A issu
es a
key extractio
n
q
uery
on
i
ID
, C firs
t
make
s a
n
1
H
qu
ery on
i
ID
, rec
o
v
e
rs
the tuple
(,
,
)
ii
i
ID
Q
r
, and an
swers
the que
ry
as
follows
:
,
i
f
,
,
oth
e
rw
i
s
e.
iP
u
b
i
A
B
i
r
P
ID
ID
o
r
ID
Sk
ļ¹
ļ¬
ļ½
ļ
ļ
ļ®
Then
C returns the
co
rre
s
po
ndin
g
an
swer
i
Sk
(when
iA
ID
ID
ļ½
or
B
ID
, C abo
rts
and
outputs
ļ
).
(4)
Sign queries
:
O
n
receivin
g a
sig
nature
que
ry on
me
ssage
m
for
a si
gne
r
i
ID
and
the desi
gnate
d
verifier
j
ID
, C does the follo
wing step
s:
(a)
If
iA
B
ID
ID
o
r
I
D
ļ¹
, then
C re
covers
(,
,
)
ii
i
ID
Q
r
from the
list
1
H
L
, com
putes th
e
sign
er'
s
priva
t
e key
ii
P
u
b
i
Sk
r
P
r
c
P
ļ½ļ½
, and randomly cho
o
se
s
*
q
tZ
ļ
to generate
the sign
ature
as follo
ws:
(,
)
t
ij
Ue
r
c
P
Q
ļ½
21
2
1
1
0
[
(
,
,
)
(
()
|
|
(
(
()
)
)
)
]
ij
h
H
ID
ID
U
F
m
F
F
m
m
ļ½ļ
ļ
((
)
,
)
ij
et
h
r
c
P
Q
ļ³
ļ½
ļ
Evaluation Warning : The document was created with Spire.PDF for Python.
ļ®
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 734
3
ā 7352
7350
(b)
If
j
AB
ID
ID
or
ID
ļ¹
, C re
cov
e
rs
(,
,
)
j
jj
ID
Q
r
from the
1
H
list and
com
putes the
sign
er'
s
priva
t
e key
jj
Sk
r
c
P
ļ½
. Then
C sel
e
ct
s
*
q
tZ
ļ
at random a
nd p
r
odu
ce
s the
sign
ature a
s
follows:
(,
)
t
j
i
Ue
r
c
P
Q
ļ½
21
2
1
1
0
[
(
,
,
)
(
()
|
|
(
(
()
)
)
)
]
ij
h
H
ID
ID
U
F
m
F
F
m
m
ļ½ļ
ļ
((
)
,
)
j
i
et
h
r
c
P
Q
ļ³
ļ½
ļ
(c)
Otherwise, quits the proto
c
ol.
Eventually, C retu
rn
s
(,
)
h
ļ³
as th
e si
gnatu
r
e
o
n
me
ssage
m
with th
e
sign
er'
s
id
entity
i
ID
and the de
si
gnated verifie
r
's id
entity
j
ID
.
(5)
Verif
y
queries:
Wh
en
re
ce
iving a ve
rify
query
on
the
sign
ature
(,
)
h
ļ³
for t
he
sign
er
i
ID
and th
e de
si
gnated
verifi
er
j
ID
, C
chec
ks
w
eather
(,
)
(
,
)
ij
A
B
I
D
ID
ID
I
D
ļ½
or
(,
)
(
,
)
ij
B
A
ID
ID
ID
I
D
ļ½
h
o
ld
s. If it hold
s
, th
en
qui
ts it. Ot
herwise,
C
rec
o
vers
(,
,
)
j
jj
ID
Q
r
from the li
st
1
H
L
and
com
put
es the
de
sig
nated ve
rifier's p
r
ivate
key
jj
Sk
r
c
P
ļ½
to verify the given
sig
nature
(,
)
h
ļ³
by the alg
o
rithm
Signatur
e
Verification
.
If it is true,
C re
cove
rs th
e
messa
ge, th
en return
s it
and o
u
tputs
1; or
els
e
, C outputs
0.
Forgery
:
Fin
a
lly, A outputs a tuple
**
(,
)
h
ļ³
as the forge
d
si
g
nature o
n
me
ssage
*
m
for
the sig
ner
*
i
ID
and the
de
sign
ated verifier
*
j
ID
with n
o
n
-
negli
g
ible
prob
ability
ļ„
. If
**
(,
)
(
,
)
ij
A
B
ID
ID
ID
ID
ļ½
or
**
(,
)
(
,
)
ij
B
A
ID
ID
ID
ID
ļ½
, then
C o
u
tputs
**
(,
)
h
ļ³
an
d proceed
s.
Otherwise, C outputs
āfailā and ab
ort
s
it. Addi
tionally, it is requi
re
d
that the sign
ature
**
(,
)
h
ļ³
is
valid, and th
at in the sim
u
lation
*
i
ID
,
*
j
ID
have never b
een
submitted to
the
Key
extraction
queries
, m
o
reover,
*
m
has n
e
ver b
een
q
uerie
d d
u
rin
g
the Sign
q
uerie
s
with t
he
signe
r'
s
identity
*
i
ID
and the de
sign
ate
d
verifier'
s
ide
n
tity
*
j
ID
.
If
**
(,
)
h
ļ³
sati
sfies all the con
d
ition
s
above, C recovers the tupl
e
**
*
*
(,
,
,
)
ij
ID
ID
U
ļ”
f
r
o
m
the
2
H
list, an
d
then
re
plays A with
the
same
ran
dom
tape
but
different
choice
of the h
a
sh
function
2
H
by e
x
ploiting the
āforkin
g
ā techn
i
que fo
rmali
z
ed in
[15]. On
the
same
m
e
ssag
e
*
m
,
C get
s an
oth
e
r valid
sign
ature
('
,
'
)
h
ļ³
su
ch t
hat
*
'
hh
ļ¹
and
*
'
ļ³
ļ³
ļ¹
. The
n
, the BDH
probl
em
with ins
t
ance
(,
,
,
)
Pa
P
b
P
c
P
can be e
a
sily
solved by C
as follo
ws:
(1)
If
**
(,
)
(
,
)
ij
A
B
ID
ID
ID
ID
ļ½
, we have
*
**
*
*
*
'
(,
)
'
(,
)
hh
ji
ji
e
S
kQ
e
S
kQ
ļ³ļ³
ļļ½
ļ
That is,
*1
('
)
*
**
*
*
(,
)(
,
)
'
hh
ji
j
i
eS
k
Q
e
r
b
c
P
r
a
P
ļ³
ļ³
ļ
ļ
ļ¦ļ¶
ļ½ļ½
ļ§ļ·
ļØļø
Then, it is ea
sy to get that
**
*
1
[(
'
)
]
*
(,
)
.
'
ij
rr
h
h
abc
ePP
ļ³
ļ³
ļ
ļ
ļ¦ļ¶
ļ½
ļ§ļ·
ļØļø
(2)
If
**
(,
)
(
,
)
ij
B
A
ID
ID
ID
ID
ļ½
, simila
rly, we
obtai
n
*1
('
)
*
**
*
*
(,
)
(
,
)
'
hh
ji
j
i
e
S
k
Q
era
c
Pr
b
P
ļ³
ļ³
ļ
ļ
ļ¦ļ¶
ļ½ļ½
ļ§ļ·
ļØļø
thus
,
**
*
1
[(
'
)
]
*
(,
)
.
'
ij
rr
h
h
abc
eP
P
ļ³
ļ³
ļ
ļ
ļ¦ļ¶
ļ½
ļ§ļ·
ļØļø
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
ļ®
Provabl
y Secure an
d Efficient ID-ba
s
e
d
Strong Desi
g
nated Verifie
r
Signature
ā¦
(Min Li)
7351
Followi
ng thi
s
, we
will co
mpute the p
r
obability t
hat
C solves th
e
given in
stan
ce of th
e
BD
H
pr
o
b
l
e
m
. C
s
u
cc
ee
ds
if:
(1)
1
:
E
Durin
g
the si
mulation, C d
oes n
o
t abort
any query;
(2)
2
:
E
A outputs a valid and n
ontrivial forgery
**
(,
)
h
ļ³
on messag
e
*
m
;
(3)
3
:
E
2
E
occu
rs and
C doe
s not q
u
it in the forgery pah
se.
The
pro
babili
ty that C
su
ccee
ds to
solv
e the
BDH p
r
oblem
is
12
3
Pr[
]
E
EE
ļļ
, that is
,
these eve
n
ts
happ
en sim
u
l
t
aneou
sly.
From the a
b
o
v
e game, it is easy to get:
11
1
1
1
2
22
2
()
(
1
)
(
1
)
(1
)
EV
S
qq
q
BD
H
c
HH
H
H
H
Ad
v
k
qq
q
q
q
ļ„
ļ«
ļ³ļ
ļ
ļļ
.
4. Results a
nd Analy
s
is
In this sectio
n, we
comp
a
r
e ou
r
schem
e with
tho
s
e
pre
s
ente
d
in
[8], [10-11] from the
asp
e
ct
s of the total comm
unication cost, ie.
||
|
|
s
i
gna
t
u
r
e
m
e
s
s
age
ļ«
, and the com
putatio
n co
st
requi
re
d by t
he
signatu
r
e
gene
ration
a
nd verifi
cati
o
n
procedu
re
s, respe
c
tively. In table 2,
we
denote by
P
a
computatio
n
of the pairin
g
operation,
S
a scal
a
r multiplication in
1
G
,
E
an
expone
ntiatio
n
in
2
G
, and
H
an unefficie
n
t āMaptoPointā
hash fun
c
tio
n
. We
also
denote th
e
total sig
natu
r
e len
g
th a
n
d
the
bit len
g
t
h of a
poi
nt in
1
G
by
TS
L
ļ
an
d
1
||
G
(assum
e t
hat
12
||
|
|
GG
ļ½
), respe
c
tively. For some
operatio
ns
su
ch a
s
addi
tions in
1
G
, XO
R in the bin
a
ry
system
and t
he commo
n h
a
sh fu
nctio
n
, they are
so
efficient that th
ey all ca
n be
negle
c
ted i
n
the
comp
ari
s
o
n
.
Table 2. The
Comp
ari
s
o
n
of Several Existing Sch
e
m
e
s
Schemes
Sign Verif
y
TS-L
Strongness
Scheme in [8]
4
S+H
3
P+H
3
|G
1
|+m
Ć
Scheme in [10]
2
S+
2
P+E
P+E
2
|G
1
|+m
Ć
Scheme in [11]
2
S+
2
P+E
2
P+S
2
|G
1
|+m
ā
Our scheme
S+
2
P+E
P+E |G
1
|+|q|
ā
From th
e co
mpari
s
o
n
ab
ove, we
can
see th
at, wh
en si
gnin
g
th
e sa
me me
ssag
es, o
u
r
prop
osed scheme is mu
ch more effici
ent on the
whole, and the
total commu
nicatio
n
co
st is
much
le
ss th
an the
others sin
c
e
no me
ssage
or j
u
st
partial
me
ssage n
eed
s to
be tra
n
smitted
along
with th
e sig
natu
r
e.
What i
s
mo
re, our
sc
he
m
e
sati
sfies th
e strong
ne
ss pro
perty, wh
ile
others excep
t
[11] can n
o
t. As far as
we
kno
w
, it is the first IBSDVS schem
e with me
ssage
recovery i
n
t
he lite
r
ature,
whi
c
h
not
onl
y req
u
ir
e
s
lo
w
com
putatio
n po
we
r, b
u
t also
ha
s
sm
all
comm
uni
cat
i
on co
st
.
5. Conclusio
n
In this paper,
we put forwa
r
d the first efficient IBSDVSMR schem
e
and give its se
curity
proof
in th
e
random
o
r
a
c
le
mod
e
l b
a
se
d on
the
BDH a
s
sum
p
tio
n
s. T
he
prop
ose
d
scheme
ha
s
both lo
w co
mputation
an
d commu
nication
co
st, a
nd
can
sup
p
o
rt the
st
ron
gne
ss p
r
op
erty of
SDVS.
Evaluation Warning : The document was created with Spire.PDF for Python.
ļ®
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 734
3
ā 7352
7352
Referen
ces
[1]
A Sham
ir. Ide
n
tit
y
-b
ase
d
cr
yptos
y
stems a
n
d
si
gnat
ure sc
hemes. G.R.
Blake
l
y , D.
C
haum
Edit
ors
.
Cr
ypto 19
84, L
NCS 19
6, Spri
nger-v
er
la
g, Califor
nia, USA, 198
4: 7-53.
[2]
D Ch
aum. Z
e
ro-kno
w
l
e
dge
und
eni
ab
le si
g
natures.
Adv
a
nces i
n
Cr
ypt
o
lo
g
y
- Eurocr
ypt'
90, LNCS.
Sprin
ger-Ver
la
g.
1990; 4
73: 4
58-4
64.
[3]
D Cha
u
m, H van Ant
w
er
pe
n. Und
eni
abl
e
signat
ures.
Advanc
es in Cr
ypt
o
lo
g
y
- Cr
yp
to'
89, LNC
S
.
Springer-Ver
lag
. 1990; 4
35: 1
2
-21
6
.
[4]
M Jakobss
on,
K Sako, R
Imp
agli
a
zzo.
Desi
gnate
d
ver
i
fier proofs and
the
i
r
app
lic
ations. Advanc
es
i
n
Cr
yptol
o
g
y
-Eur
ocr
y
pt'
96,
LNC
S
. Springer-V
e
r
lag.
19
96; 10
7
0
: 143-1
54.
[5]
S Saee
dni
a, S Kramer, O Markovitch. An effi
cient stron
g
de
sign
ated ver
i
fie
r
signat
ure sch
eme.
ICISC
200
3.
Sprin
ger
-Verla
g, Berlin
.
2003: 4
0
-54.
[6]
W
Susilo, F
Z
h
ang, Y Mu. Identit
y
-
b
a
se
d st
rong d
e
sig
nate
d
verifier sig
n
a
t
ure schemes.
Lecture Notes
in Co
mputer S
c
ienc
e (LNCS)
.
2004; 3
108: 3
13-3
24.
[7]
X H
uan
g, W
Susil
o
, Y Mu, F
Z
hang. Short i
dentit
y-b
a
s
ed strong
de
sign
at
ed ver
i
fi
er sign
ature
schemes.
Lect
u
re Notes i
n
C
o
mputer Sci
e
n
c
e ( LNCS).
20
06; 390
3: 214-
225.
[8]
J Z
hang, J M
a
o. A nov
el ID-b
ased
des
ign
a
ted ver
i
fier si
gn
ature sch
eme.
Information
Sci
ences. 2
0
0
8
;
178(
3): 766-
77
3.
[9]
BB Kang, C B
o
yd, E Da
w
s
o
n
. Identit
y-bas
ed stro
n
g
des
i
gnate
d
verifi
er
signat
ure sch
emes: attacks
and n
e
w
co
nstruction.
Co
mput
ers & Electrical
Engin
eeri
n
g
. 2
009; 35(
1): 49-
53.
[10]
BB Kan
g
, C B
o
yd, E D
a
w
s
o
n
. A nov
el
ide
n
tit
y
-
base
d
str
ong
des
ign
a
te
d verifi
er si
gna
ture schem
e
.
T
he Journ
a
l of
Systems a
nd S
o
ftw
are
. 2009; 82(2): 27
0 - 27
3.
[11]
J Le
e, J C
h
a
n
g
, D
Lee. F
o
rger
y
attacks o
n
Ka
ng
et
a
l
.ā
s id
entit
y
-
bas
e
d
stron
g
d
e
si
g
nated
verifi
er
sign
ature sc
he
me an
d its
im
provem
ent
w
i
t
h
secur
i
t
y
pro
o
f.
Co
mp
uters
& Electric
al
Engi
neer
in
g,
Co
mp
uter& Ele
c
trical Eng
i
ne
e
r
ing
. 20
10; 36(
5): 948-9
54.
[12]
R T
s
o, C Gu,
T
Okamoto, et al. Efficient ID-
B
ased
Dig
ital
Sign
atures
w
i
t
h
Messa
ge R
e
cover
y
. F
.
Ba
o
et al.
Editors
. CANS 20
07, L
NCS 48
56, Spr
i
ng
er-Verl
ag. 2
007: 47-
59.
[13]
K Barr, K Asan
ovic. Ener
g
y
-
a
w
a
re l
o
ssless
data com
p
ressi
on.
Jour
nal A
C
M T
r
ansaction
on Co
mpute
r
System
s
(T
OC
S). 2006; 24(
3)
: 250-29
1.
[14]
M Bellare, P Rog
a
w
a
y.
Ra
n
d
o
m
oracl
e
s a
r
e practical: a
para
d
ig
m for d
e
sig
n
in
g efficie
n
t protocols
.
Procee
din
g
CC
Sā 93 of 1
st
AC
M conferenc
e on comp
uter a
nd commu
nicat
i
ons sec
u
rit
y
. 1
993: 62-
73.
[15]
D Pointch
e
va
l, J Stern. Securi
t
y
pro
o
fs
for signatur
e schem
es. Eurocr
ypt 1
996. LN
CS, 10
70.
Sprin
ger-
Verlag.
199
6: 387-3
98.
[16]
E Yoo
n
. An
efficient
an
d s
e
cure
id
entit
y-bas
e
d
strong
desi
g
n
a
ted
verifier s
i
g
nat
ure sc
heme.
Information
T
e
chno
logy and Contro
l
. 201
1; 40(4): 32
3-3
2
9
.
Evaluation Warning : The document was created with Spire.PDF for Python.