Design and Evaluation of an Intelligent Static Analysis Framework for Detecting Access-Control Vulnerabilities in DeFi Smart Contracts PDF Free Download

1 / 10
0 views10 pages

Design and Evaluation of an Intelligent Static Analysis Framework for Detecting Access-Control Vulnerabilities in DeFi Smart Contracts PDF Free Download

Design and Evaluation of an Intelligent Static Analysis Framework for Detecting Access-Control Vulnerabilities in DeFi Smart Contracts PDF free Download. Think more deeply and widely.

I. *
Received(10. 15. 2025), Modified(11. 18. 2025),
Accepted(11. 21. 2025)
* 2025
(RS-2024-00415520)
ICT4.0 (II
TP-2022-RS-2022-00156310)
(RS-2024-00437252) (RS-
2025-00518150) .
, gyqlswh1109@naver.com
, iglee@sungshin.ac.kr(Corresponding author)
1601
Journal of The Korea Institute of Information Security & Cryptology
VOL.35, NO.6, Dec. 2025
ISSN 1598-3986(Print)
ISSN 2288-2715(Online)
https://doi.org/10.13089/JKIISC.2025.35.6. 1601
DeFi
*
,1
,2 3
1,2,3 (, , )
Design and Evaluation of an Intelligent Static Analysis Framework for
Detecting Access-Control Vulnerabilities in DeFi Smart Contracts*
Hyo-been Cho,1
Nam-ryeong Kim,2Il-gu Lee3
1,2,3Sungshin Womens University (Undergraduate student, Graduate student, Professor)
DeFi .
·
,
. IR(Intermediate Representation) ,
, .
·
·
·
, IR Semantics
. 2020
2024 20 95% 1.72
.
, DeFi
.
ABSTRACT
The growth of the DeFi ecosystem has intensified financial losses caused by access-control vulnerabilities in smart
contracts. Existing static and dynamic tools detect code-level issues but struggle to capture logical flaws tied to protocol
structures and authorization flows. This paper introduces an intelligent static analysis framework that uses an IR to integrate
operation flow, state transitions, and authorization-verification patterns. By aggregating and normalizing permission, call, and
state-change patterns and applying a semantics-based method with automatically adjusted statistical thresholds, the framework
identifies access-control weaknesses more effectively. Evaluation on 20 major hacking incidents from 2020
2024 shows a
95% detection rate and 1.72-second average latency, outperforming existing tools and substantially improving the precision of
DeFi access-control vulnerability analysis.
Keywords: DeFi, Lending Protocol, Smart Contract, Access Control, Vulnerability Detection
1602
DeFi
I.
(Decentralized Finance, DeF
i) ,
,
[1].
,
[2].
DeFi 2020 (Total Value
Locked, TVL) 10 2022 2
40
, 2025 5
156
[3].
,
.
Ta
b
le 1. 2025 DeFi
,
B
y
B
it
( 2 ), Coin
b
ase
( 5,600 ), Cetus
( 3,000 ),
N
o
b
ite
x
( 1,200 )
.
DeFi
,
[4-9].
.
,
·
, , ,
,
·
.
,
DeFi
.
,
,
·
,
.
DeFi
Semantics .
, ,
(entry poin
t) ,
.
IR
.
,
.
. 2 DeFi
, 3
. 4
, 5
.
6
.
II.
·
.
,
,
·
[10-11].
N
.
R. Kim
,
[11].
, Slither, SmartChec
k, Securify
[12].
Z
hou
,
[13]. Chaliasos
, 127
11( 8%)
Date
Victim
Category
Fe
b
21
B
y
B
it
C
EX
Mar 9
A
b
racada
b
ra
Lending
May 11
Coin
b
ase
C
EX
May 22
Cetus (Sui)
D
EX &
Liquidity
Jun 18
N
o
b
ite
x
C
EX
Jun 26
Resupply
Lending
Table 1. Major DeFi hacking incidents in 2025.
(2025. 12)
1603
[14]. Murala
G
A
N
,
[15].
, DeFi
Tainter[16], DeFiRanger[17]
(tai
nt) semantic lifting
,
.
,
, Mythril,
E
chidna, Manticore
[12]. Liao
·
96.5%
, [18].
DeFi
, R
B
AC(Role-
B
ased Access Control), ACL
(Access Control List), Permission Manager
.
,
,
.
W
u
Remi
x
,
[19].
,
,
.
,
, DeFi
.
III. DeFi
DeFiLlama 2025 TVL
5 (lending)
,
. T
a
b
le 2. 2025 8
, 10
TVL
.
,
Fig. 1.
6
.
,
.
1.
U
ser Interface Layer: (supply),
(
b
orrow), (repay), (withdraw)
,
.
2. Core Protocol Layer:
,
,
,
,
.
3.
O
racle Layer:
.
4.
G
overnance/
G
uardian Layer: LTV(Loan-
to-Value),
,
,
,
.
5. Liquidator Layer:
,
·
.
.
Protocol
TVL
(25.08)
TVL
(25.10)
Category
Chains
(25.10)
Aave V3
$
38.814
b
$
45.185
b
Lending
19
Compound
Comet
$
2.454
b
$
3.012
b
Lending
9
SparkLend
$
1.625
b
$
6.163
b
Lending
2
Maple
$
2.016
b
$
2.972
b
Lending
2
Fluid
$
1.32
b
$
2.262
b
Lending
5
Table 2. Analyzed DeFi protocols.
Fig. 1. Common structure of major lending
protocols
1604
DeFi
6. Token Layer :
,
·
, Core Proto
col
.
.
,
.
delegatecall ,
.
,
.
,
(t
ime-lock)
.
,
.
Aave V3, Spar
kLend
ACL Manager
R
B
AC
,
ACL
. Compound Comet governor,
pause
G
uardian
,
. Maple Fluid Permi
ssion Manager
,
.
,
Ta
b
le 3. . I
nitialization Protection
, Access Control
.
O
racle C
onfiguration Pro
x
y
U
pgrade
,
.
,
.
,
,
, ,
·
,
,
7
.
e
x
ternal
pu
b
lic set,
update, grant, revoke, withdraw, transfer
.
.
IV.
4.1
(Interm-
ediate Representation, IR) DeFi
.
.
Category
Description
Initialization
Protection
Protection Against
Re-Initialization Privilege
E
scalation
Access
Control
Missing Access Checks / Faulty
Comparison Detection
Input
Validation
Insufficient
B
oundary/Type
Validation
E
RC20 Secure
H
andling
Missing Failure
H
andling for
N
on-Standard Tokens
O
racle
Configuration
U
nverified Source / Missing
Multisig
&
Timestamp Checks
Pro
x
y
U
pgrade
Centralized
U
pgrade Authority /
N
o Timelock
Liquidation
Parameters
Liquidation Risk due to
W
eak
Validation on Threshold
U
pdates
L2-Specific
O
ptimization
Am
b
iguous Authority
B
oundaries in Rollup/
B
ridge
O
perations
Table 3. DeFi protocols security, validation
pattern.
(2025. 12)
1605
,
,
, modifier ,
·
.
.
,
(category),
( , ),
0.3(
) 0.9(
)
.
,
,
.
, IR
,
퀀
,
.
.
.
1.
:
, modifier,
.
2.
: IR
,
, ,
.
3.
: modifier
,
,
0.0
1.0
.
4.
:
,
,
.
.
,
.
4.2
Fig. 2.
,
IR
.
,
Slither IR
, ,
.
,
.
/
,
,
,
(transfer/send/low-level call)
Fig. 2. Intelligent static analysis system archi-
tecture.
1606
DeFi
.
E
ntry Point Validation
pu
b
lic/e
x
ternal (
N
o
n-view/pure,
N
on-constructor)
.
,
E
R
C20/
,
getter
·
event e
mitter
.
.
O
peration Risk Assessment
IR
,
,
, ,
.
.
,
·
·
·
value transfer
.
L
OW~
CRITICA
L Risk Level .
Access Control
E
valuation mod
ifier , require , msg.sender
,
,
0.0
~
1.0
(AC Strength)
. Risk Score
×
(1
AC Strengt
h) .
Dynamic Threshold
E
valuation
.
.
Report
,
E
nd
.
0.6 ,
.
,
.
,
.
V.
2020
2024
.
,
·
·
20
.
(S
lither), (
E
chidna),
,
, F1-Score
.
Ta
b
le 4.,
Fig. 3. .
, 20 19
95%
.
1.72
. IR
·
. , Slither
,
3
15%
. Slither 1.86
,
·
.
E
chidna
Category
Precision
Recall
F1-Score
Proposed
0.95
0.95
0.95
Slither
0.50
0.26
0.34
E
chidna
0.53
0.40
0.46
Table 4. Comparison of access control
vulnerability detection performance.
Fig. 3. Detection Rate and Latency Comparison
of Analysis Methods.
(2025. 12)
1607
Slither
45% .
, 15.24
.
,
,
,
Slither
E
chidna
.
VI.
DeFi
.
(IR)
,
,
,
.
,
95%
1.72
.
,
.
.
·
,
.
References
[1]
Y
.A. Shin,
G
.T.
N
oh, and J.
Y
. Chun,
A survey of leader election technique
s without oracles in pu
b
lic
b
lockchain
systems,
Journal of Future Society, 1
6(2), pp.170-185, Jun. 2025.
[2] T. Feng,
X
.
Y
u,
Y
. Chai and
Y
. Liu,
Smart contract model for comple
x
rea
lity transaction,
in International Jour
nal of Crowd Science, vol. 3, no. 2, p
p. 184-197, July. 2019.
[3] M.
B
rigida,
The surprising irrelevanc
e of total-value-locked on cryptocurre
ncy returns,
ar
X
iv preprint ar
X
iv:250
6.03287, Jun. 2025.
[4] Federal
B
ureau of Investigation,
N
ort
h Korea responsi
b
le for
$
1.5
B
illion
B
y
b
it hack,
Internet Crime Complaint
Center (IC3), PSA-250226, Fe
b
. 2025.
[5] TRM La
b
s,
The
B
y
b
it hack: Followin
g
N
orth Korea
s largest e
x
ploit,
TRM
B
log, https://www.trmla
b
s.com/resour
ces/
b
log/the-
b
y
b
it-hack-following-nort
h-koreas-largest-e
x
ploit,
O
ct. 2025.
[6] Chainalysis,
Colla
b
oration in the wak
e of record-
b
reaking
B
y
b
it theft,
Chai
nalysis
B
log, https://www.chainalysis.
com/
b
log/
b
y
b
it-e
x
change-hack-fe
b
ruar
y-2025-crypto-security-dprk/ ,
O
ct. 2025.
[7] CC
N
,
Crypto hacks 2025: Full list of
scams, e
x
change e
x
ploits
&
DeFi vuln
era
b
ilities,
CC
N E
ducation, https://w
ww.ccn.com/education/crypto/crypto-h
acks-e
x
ploits-full-list-scams-vulnera
b
i
lities/,
O
ct. 2025.
[8]
H
al
b
orn,
Ex
plained: The A
b
racada
b
ra
Money
H
ack (March 2025),
H
al
b
orn
B
log, https://www.hal
b
orn.com/
b
log/p
ost/e
x
plained-the-a
b
racada
b
ra-money-
hack-march-2025 ,
O
ct. 2025.
[9] Protos,
DeFi lending protocol A
b
raca
da
b
ra e
x
ploited for
$
13M of
E
T
H
,
Pro
tos, https://protos.com/defi-lending-pr
otocol-a
b
racada
b
ra-e
x
ploited-for-13m-
of-eth/ ,
O
ct. 2025.
[10] P.
Q
ian, R. Cao,
Z
. Liu,
W
. Li, M. L
i, L.
Z
hang,
Y
.
X
u, J. Chen, and
Q
.
H
e,
Comprehensive review of smart c
ontract and DeFi security: Attack, vu
lnera
b
ility detection, and automated r
1608
DeFi
epair,
Ex
pert Systems with Applicati
ons, vol. 291, Art. no. 128431, 2025.
[11]
N
. R. Kim, D.J. Ryu, and I.
G
. Lee,
Code similarity-
b
ased framework for
smart contract attack surface analysi
s,
Convergence Security Journal, 24
(5), pp. 161-169, Dec. 2024.
[12] J. Feist,
G
.
G
rieco, and A.
G
roce,
Sli
ther: a swtatic analysis framework fo
r smart contracts,
I
EEE
/ACM Interna
tional
W
orkshop on
E
merging Trends
in Software
E
ngineering for
B
lockchai
n (
WE
TS
EB
), pp. 8
15, May. 2019.
[13] L.
Z
hou,
X
.
X
iong, J.
E
rnst
b
erger, S.
Chaliasos,
Z
.
W
ang,
Y
.
W
ang, K.
Q
in,
R.
W
attenhofer, D. Song, A.
G
ervais ,
SoK: decentralized finance (DeFi) att
acks,
2023 I
EEE
Symposium on Secur
ity and Privacy (S
&
P), pp. 2444
246
1, May. 2023.
[14] S. Chaliasos, M.A. Charalam
b
ous, L.
Z
hou, R.
G
alanopoulou, A.
G
ervais,
D. Mitropoulos, and
B
. Livshits,
Sma
rt contract and DeFi security tools:
o they meet the needs of practitioner
s
?
,
Proceedings of the I
EEE
/ACM
 
46
th International Conference on Softwar
e
E
ngineering (ICS
E
), pp. 1
13, 2024.
[15] D. K. Murala, S. Loucif, K.V.P. Rao,
H
.
H
amam,
E
nhancing smart contrac
t security using a code em
b
edding an
d
G
A
N
-
b
ased approach,
Scientific Re
ports, vol. 15, no. 1, Art. no. 99267,
May. 2025.
[16]
Q
. Kong, J. Chen,
Y
.
W
ang,
Z
. Jiang,
and
Z
.
Z
heng,
DeFiTainter: Detecting
price manipulation vulnera
b
ilities in
DeFi protocols,
Proceedings of the 32
nd ACM SI
G
S
O
FT International Symp
osium on Software Testing and Analy
sis (ISSTA 23), pp. 1144
1156, Jul. 2
023.
[17] S.
W
u,
Z
.
Y
u, D.
W
ang,
Y
.
Z
hou, L.
W
u,
H
.
W
ang,
X
.
Y
uan,
Q
.
H
e, and
K. Ren,
DeFiRanger: Detecting DeFi
price manipulation attacks,
I
EEE
Tra
nsactions on Dependa
b
le and Secure
Computing, vol. 21, no. 4, pp. 3370-33
85, Aug. 2024.
[18]
X
. Liao,
Smart contract vulnera
b
ility
detection
b
ased on dynamic and static
com
b
ination,
Proceedings of the Inter
national Conference on Digital
E
cono
my,
B
lockchain and Artificial Intellige
nce (D
EB
AI 24). Association for Comp
uting Machinery, pp. 412
416, Dec. 2
024.
[19] J.
W
u, L.
X
ie, and
X
. Li,
Security vu
lnera
b
ilities in
E
thereum smart contr
acts: A systematic analysis,
ar
X
iv pr
eprint ar
X
iv:2504.05968, Apr. 2025.
(2025. 12)
1609
< >
(
H
yo-
b
een Cho)
2023 3
월~현
:
<
>
, ,
W
e
b
3.0
김남령
(
N
am-ryeong Kim)
2023 2
:
2024 9
월~현
:
<
>
, ,
W
e
b
3.0 , AI
(Il-gu Lee)
2016 2
: KAIST
2005 2
~
2017 2
:
2017 3
월~현
:
/
2024 3
월~현
:
2025 3
월~현
:
<
>
, ,