enhanced mutual authentication scheme for cloud of thingsenhanced mutual authentication scheme for...
TRANSCRIPT
Enhanced Mutual Authentication Scheme for Cloud of
Things
Bhuvaneshwari S1, Anantha Narayanan V
2,
1,2 Dept. of Computer Science and Engineering, Amrita School of Engineering, Coimbatore,
Amrita Vishwa Vidyapeetham, India [email protected], [email protected]
Abstract. In recent years, technologies play a vital role in day to day activities,
of which cloud computing and Internet of Things form a primary role. The
integration of these two technologies would be a huge advancement wherein
there is a large pool of embedded devices communicating with the cloud.
However, there are a lot of issues when we go for this integration. One of the
main concerns is security, where device authentication and data privacy plays a
vital role. In this paper, a mutual authentication scheme using Elliptic Curve
Cryptography has been proposed. The proposed scheme has been analyzed
against the work of Kalra et al. and the possible drawbacks with resolutions
have been presented. The proposed scheme is verified using AVISPA and is
proved to be strong against several attacks which prove it is more secure,
powerful and efficient.
Keywords: Authentication, Cloud Computing, Elliptic Curve Cryptography,
Internet of things.
1 Introduction
Internet of things (IoT) includes not only physical objects that are embedded with
software and sensors but also includes humans in the network who are connected with
these objects. IoT exchanges data among them to achieve a common goal in a smart
way [1]. In recent years IoT has embraced different technologies and services in
different fields such as healthcare, automation, manufacturing and household making
it smart by undergoing multi fold advances. Unquestioningly the strength of IoT has
been increased because of its involvement in everyday activity and since they operate
without any human intervention.
The large scale of IoT projects has been doubled in the past 12 months as the
number of organizations using IoT has increased with more than 50 thousand
connected devices. These devices produce data which could be analyzed using cloud
technologies, which has infinite storage and processing power. One of the biggest
challenges in IoT is storage of multi-source high heterogeneity data with low
semantics, so we move forward towards combining IoT with Cloud [2].
The users of cloud can be relaxed on the management, maintenance and storage of
resources and in terms of cost, its pay-as-you-use policy has been an advantage. The
usage of public cloud for IoT usage has been on increase for the past 2 years. Cloud
International Journal of Pure and Applied MathematicsVolume 119 No. 15 2018, 1571-1583ISSN: 1314-3395 (on-line version)url: http://www.acadpubl.eu/hub/Special Issue http://www.acadpubl.eu/hub/
1571
providers such as AWS (Amazon Web Service), Microsoft Azure, Google cloud and
IBM cloud has shown an increase in usage from 2017 to 2018. The trend shows more
usage in future as there is a constant increase in the usage of embedded things which
provides huge volumes of data.
Fig. 1. Cloud of Things
The two different worlds, cloud and IoT complement each other well and has been
used to obtain a lot of benefits in different applications [3][4][5]. Cloud can give an
effective solution when it comes to management and also build real-time applications
[19] and assist in the processing of the data produced [6]. This integration of cloud
computing and IoT is known as “Cloud of Things”.
On integration, the cloud will act as a layer in the middle of the things and the
applications cover the complexities in its functionalities to realize them. The cloud
offers connectivity through different means such as IoT applications in mobiles to
access the data stored in it.
Security is a major concern when it comes to integration of these two technologies,
where authentication plays a major role. There is a possibility of a lot of attacks on the
cloud side causing an issue in data privacy. There are different encryption techniques
used to address data integrity, data confidentiality and authenticity [7].
The embedded devices which will act as an HTTP (Hyper Text Transfer Protocol)
client can communicate to a cloud server (an HTTP server) as shown in Fig. 1.
Embedded devices can access data storage and computational facilities from the cloud
storage. Here mutual authentication of the device and the cloud plays an important
role, before further processing. Since the embedded devices are constrained in all
categories, an asymmetrical cryptographic technique called Elliptic Curve
Cryptography (ECC) is recommended [8].
The rest of the paper is organized as follows. Section 2 deals with related works,
followed by the proposed method in Section 3. Section 4 and 5 provides the security
analysis and cost analysis respectively. The concluding remarks are presented in
Section 6.
International Journal of Pure and Applied Mathematics Special Issue
1572
2 Related Works
Recent years have seen a lot of authentication schemes being proposed, which uses
integration of IoT and cloud technologies. Table 1 gives a summary of the most
relevant techniques to our work. The table also shows the different vulnerabilities for
each of the authentication schemes.
Table 1. Authentication Schemes
Proposed
Methodology
Authentication Scheme Description and vulnerabilities
Wu et al. [9] ID-based remote authentication with
smart cards on open distributed
system from elliptic curve
cryptography
All the client information is warehoused
in the server, so that only the server can
authenticate the client. This leads to lot
of threats such as man in the middle and
insider attack.
Tian et al.
[10]
Analysis and improvement of an
authenticated key exchange protocol
for sensor networks
They use communal authentication
using certificates which upsurge the
cost.
Abichar et al.
[11]
A fast and secure elliptic curve
based authenticated key agreement
protocol for low power mobile
communications
Certificate based authentication scheme
is used, which is not that cost effective.
Yang et al.
[12]
An id-based remote mutual
authentication with key agreement
scheme for mobile devices on
elliptic curve cryptosystem
ECC centered authentication aimed at
smart devices is proposed.
Yao et al. [13] A lightweight attribute-based
encryption scheme for the internet of
things
A feature based simple encryption
intended for IoT, built using ECC is
proposed.
Moosavi et al.
[14]
An elliptic curve-based mutual
authentication scheme for RFID
implant system
An authentication method centered on
ECC for RFID is proposed
Liao et al.
[15]
A secure ECC- based RFID
authentication scheme integrated
with ID-verifier transfer protocol
ID verifier transfer protocol using ECC
is proposed.
Muhamed et
al. [16]
A novel user authentication and key
agreement scheme for
heterogeneous ad hoc wireless
sensor networks based on the
Internet of Things notion
Authentication along with key
agreement for heterogeneous system has
been proposed.
Persson et al.
[17]
Merging cloud and IoT A model for unification IoT and cloud
in an incorporated framework is
presented.
Sood et al.
[18]
Dynamic identity-based single
password anti-phishing protocol
Single password anti-phishing method
using cookie is presented.
International Journal of Pure and Applied Mathematics Special Issue
1573
3 Proposed Method
The proposed method provides an enhancement of the necessary security features for
mutual authentication of Cloud of Things platform. In the proposed methodology, the
mutual authentication scheme runs with HTTP cookies. There are 3 phases designed,
namely, Registration, Pre-computation and log in, and Authentication. The whole
scheme here is centered on the ECC, which is present in the Initialization phase [8].
Fig. 2 shows the proposed authentication scheme and Table 2 shows the different
representations used in the proposed protocol. This section briefs the proposed
algorithm.
Table 2. Representations used in the protocol
Device Embedded device
D_IDi Identity of the device
Cloud_server Cloud server
Pwi Password of the device
RS First random number produced in the server
ran1, ran2 Random numbers used in ECC
Hash() One-way hash function
X Server’s private key
G Base point of ECC
| Concatenation
Ep Finite field group
Cookie HTTP cookie information
SK Session key
⊕ XOR operation
EXP_TIME Cookie’s expiration time
3.1 Registration phase
Step R1: For recording with Cloud_server, the client directs a unique D_IDi and Pwi
password to the server. Here the embedded device used is considered to be intruder
proofed so the password is created by itself. The embedded device directs the
password to the Cloud_server as a hashed value N=Hash (D_IDi|Pwi).
Step R2: The server generates random number RS and with private key of server, it
computes Bi,
⊕ .
and security parameters,
⊕ ⊕ Step R3: Cookie’ is passed from the server to the device and is stored there.
International Journal of Pure and Applied Mathematics Special Issue
1574
3.2 Login phase
Step L1: The device generates a random number ran1 and uses the generator point G,
Num1=ran1*G.
Step L2: Num2=Hash (ran1*Cookie’) is calculated with already stored Cookie’
value.
Step L3: {N,Num1,Num2} is passed from the device to the server.
Step L4: The server computes Num2’ = Hash (Num1 .Cookie) with the values stored
in it.
Step L5: In server, If Num2 == Num2’
Then it selects a random number, ran2
Calculates, Num3 =ran2*G*RS
Num4 = Num3 * Ai
Step L6: Server passes {Num3,Num4,Ti} to the device.
Fig. 2. Proposed authentication scheme
International Journal of Pure and Applied Mathematics Special Issue
1575
3.3 Authentication phase
Step A1: The device calculates Ai’ with already stored value from the previous
phase.
Ai’=Hash(Bi ⊕ Cookie’ ⊕ Hash (D_IDi|Pwi)
Step A2: Num4’ = Num3*Ai’ is calculated by the device.
Step A3: In the device, If Num4 == Num4’ is true,
Then it calculates
S=ran1*ran2
V = Hash ((Num1*Cookie’)|S)
Step A4: The calculated V is passed to the server from the device.
Step A5: In the server, S =Num1*ran2*RS is calculated. Using the new calculated S,
the server computes V’=Hash ((Num1*Cookie)|S)
Step A6: In server, If V==V’ is true, then
Session key SK=Hash (X|S) is generated and shared with the device.
Else,
Reject the device.
4 Security Analysis
The proposed work is simulated using HLPSL in AVISPA. The simulation is used to
demonstrate that the proposed method is secure against various threats. Figures 3-5
show the HLPSL code developed for simulation. The different threats observed in
mutual authentication are:
Fig. 3. HLPSL code for role specification of device
International Journal of Pure and Applied Mathematics Special Issue
1576
Fig. 4. HLPSL code for role specification of server
Insider attack: This attack is impossible because the Cloud_server does not have
the vice details as the device identity and the password is hashed and before it is
passed to the server from the intruder proofed device.
Replay attack: Here the intercepted messages are retransmitted to perform the
attack but in the system proposed, if an attacker tries and intercept the message
{Num1,Num2,N} which is transmitted from D_IDi to Cloud_server. But even if
the login phase computation continues the device will not be authenticated in the
authentication phase where the verification takes place only if Num4 == Num4’,
then proceeds to calculate S=ran1*ran2 using which the mutual authentication
variable V = Hash ((Num1*Cookie’)|S) is computed but when it is passed to the
sever side it is again verified with the values stored in the server only if I holds
true the device would be authenticated. Hence a false device cannot compute the
V and SK session key. So the replay attack is impossible.
Man-in-the-middle attack: Since here there is joint authentication among the
D_IDi and Cloud_server there can be no possibility of Man-in-the –middle attack.
International Journal of Pure and Applied Mathematics Special Issue
1577
Fig. 5. HLPSL code for role specification of session, goal and environment
Brute force attack: Only if the invader is able to take out all the security
constraints {Bi,Num1,Num2,Num3,Num4} calculated, he can test with the brute
force method but even then guessing the password Pwi and also the random
numbers ran1 and ran2 is impossible and to access the attacker should know{X},
so brute force attack cannot be performed.
Cookie-theft attack: Embedded device D_IDi is intruder proofed, the cookie
stored in the smart device cannot be stolen.
Offline password guessing attack: Hence the device id and password are only
hashed and passed to the seer, there is no way to perform this attack, even if
password would be guessed the attacker will not be able to find the identity of the
device.
International Journal of Pure and Applied Mathematics Special Issue
1578
4.1 System Security Requirements
In order to propose an authentication scheme or any security measures, there are few
security requirements that need to be achieved.
Mutual authentication: The device and server is mutually authenticated using the
variable Num4 and also V, which is exchanges during authentication phase only if
it is valid then the SK is calculated and shared else the device is rejected.
Confidentiality: During all the transmission no important variable is directly
transmitted to the other side without hashing and even throughout the
communication the device details and the private key of the server is kept very
confidential.
Anonymity: The illegitimate server cannot know any device details in the
proposed scheme as because the password and identity of the device are always
hashed before exchanging any messages between the server and the device.
4.2 Comparison with Kalra et al.
Recently, Kalra et al. presented a mutual authentication method in order to link the
cloud and embedded devices using ECC and appealed that their method gains all the
security necessities and is very resilient to several types of threats. But the
authentication method suffers from certain drawbacks [8] and is deliberated in detail
in Table 3. The solutions for these shortcomings are also provided. The shortcomings
have been addressed in the proposed method.
Table 3. Possible vulnerabilities of Kalra’s method with possible solutions
Drawbacks Definition Reason Solution
Phase Message
Absence of
device
anonymity
Anonymity is to
secure the device
specific information
so the data in cloud
server would remain
secure.
Login phase D_IDi is passed
in plain text, so
attacker can
easily get the
D_IDi by
monitoring
requests.
The ID_I s not
passed at any
phase in the
system. N only
the hashed
value of PW
and D_IDi are
passed.
Mutual
authentication
is impossible
Mutual
authentication
between cloud serer
and the embedded
device.
Registration
phase
Authentication
phase
While registering
the device, the
cloud server
creates the pwi
but it is never
shared to the
device.
The pwi is
created by
device itself and
hashed with
D_IDi and
passed to the
server.
International Journal of Pure and Applied Mathematics Special Issue
1579
Offline
password
guessing
attack
When other details for
authentication is
known, the password of
the device can be easily
guessed by which the
attacker can enter into
system.
Login phase
Authentication
phase
Cookie’ can be
obtained from
device.
D_IDi , val3,val4
can be obtained
from
authentication
phase.
Using the above
values attacker ca
guess the pwi of
the device and
enter the system.
Cookie’ cannot
be obtained by
making the
device intruder
proofed.
D_IDi of the
device will not be
known as it is
never passed. N
only the hashed
value of PW and
D_IDi are passed
in the system.
Insider
attack
Insider of the cloud
server will know the
pwi and he/she can
misuse it.
Registration
phase
Pwi is created by
cloud server and
stored in it.
Pwi is created by
the device itself
and is never
passed in the
message directly.
5 Cost Analysis
The computation and storage cost is discussed in Table 4 and Table 5 respectively.,
where, CC is the computational cost (hash operation) and CCecc is the computational
cost (ECC point multiplication operation). The experiment results of Kilinc et al.
shows the computation costs i.e. execution time of CC and CCecc are 2.3μs and 22.26
× 102 μs. Since all the operations are lightweight the system is inexpensive when
compared to other methods as shown in Table 4.
Table 4. Computational cost comparison
Schemes Embedded device Cloud server Total
Kalra et al. 4 CC+3CCecc 5CC +4CCecc 9CC +7CCecc =15.603×103 µs
Kumari et al. 3CC +4CCecc 4CC +4CCecc 7CC +8CCecc=17.824×103 µs
Proposed Scheme 3CC +3CCecc 5CC +3CCecc 8CC +6CCecc=13.374×103 µs
Table 5. Storage cost comparison
Schemes Number of messages Number of bits Storage cost (bits)
Kalra et al. 3 1760 320
Kumari et al. 3 1760 480
Proposed Scheme 3 1760 320
The communication and storage cost is lesser when compared to the Kumari et al.
method and is equal to the Kalra et al. method but with lesser computational cost. The
simulation results in AVISPA tool is given in Fig. 6 were both the results using
OFMC back-end and using CL-AtSe back-end are shown. Both the results show that
the system is safe even when the intruder has full access to the network and the
different types of attacks are tried out but it is impossible to break the authentication
scheme.
International Journal of Pure and Applied Mathematics Special Issue
1580
Fig 6. OFMC and ATSE output of the proposed scheme
6 Conclusion
The proposed ECC based mutual authentication scheme is proved to be protected
against several threats. Moreover this method shows that it satisfies all the security
requirements such as mutual authentication, anonymity and confidentiality. The
results of the implemented system shows that the computational cost is efficient and it
can be implemented on any system which has HTTP protocol and the system is
proved to be safe. The future work of this paper is to use the same authentication
scheme on a real-time Cloud of Things system.
References
1. Atzori, Luigi, Antonio Iera, and Giacomo Morabito.: The internet of things: A survey.
Computer networks 54.15. 2787-2805 (2010).
2. Cai, H., Xu, B., Jiang, L. and Vasilakos, A.V .: IoT-based big data storage systems in
cloud computing: perspectives and challenges. IEEE Internet of Things Journal, 4(1),
pp.75-87 (2017).
International Journal of Pure and Applied Mathematics Special Issue
1581
3. N. Alhakbani, M.M. Hassan, M.A. Hossain, M. Alnuem.: A framework of adaptive
interaction support in Cloud-based Internet of Things (IoT) environment, Internet and
Distributed Computing Systems, Springer, pp.136–146.( 2014).
4. Aitken, R., Chandra, V., Myers, J., Sandhu, B., Shifren, L. and Yeric, G.: Device and
technology implications of the internet of things. VLSI Technology Digest of Technical
Papers pp. 1-4. IEEE (2014).
5. Gomes, Márcio Miguel, Rodrigo da Rosa Righi, and Cristiano André da Costa.: Future
directions for providing better IoT infrastructure. ACM International Joint Conference on
Pervasive and Ubiquitous Computing: Adjunct Publication (2014).
6. Lee K, Murray D, Hughes D, Joosen W.: Extending sensor networks into the cloud using
amazon web services. Networked Embedded Systems for Enterprise Applications
(NESEA), pp. 1-7. IEEE.European commission, Definition of a research and innovation
policy (2010).
7. Kumar S., Pradeep P., Kj S.: Detection of SPAM Attacks in the Remote Triggered WSN
Experiments. In: Kim K., Joukov N. (eds) Information Science and Applications (ICISA)
2016. Lecture Notes in Electrical Engineering, vol 376. Springer, Singapore (2016)
8. Kalra S, Sood SK .: Secure authentication scheme for IoT and cloud servers. Pervasive
Mob Comput 24:210–223 (2015).
9. Wu ST, Chiu JH, Chieu BC.: ID-based remote authentication with smart cards on open
distributed system from elliptic curve cryptography. IEEE International Conference on
Electro Information Technology. IEEE, pp 5 (2005).
10. Tian X, Wong DS, Zhu RW .: Analysis and improvement of an authenticated key
exchange protocol for sensor networks. IEEE Commun Lett 9(11):970–972 (2005).
11. Abi-Char PE,Mhamed A, El-Hassan B.: A fast and secure elliptic curve based
authenticated key agreement protocol for low power mobile communications. The 2007
International Conference on Next Generation Mobile Applications, Services and
Technologies (NGMAST’07). IEEE, pp 235–240 (2007).
12. Yang JH, Chang CC.: An id-based remote mutual authentication with key agreement
scheme for mobile devices on elliptic curve cryptosystem. Comput Secur 28(3):138–143
(2009).
13. Yao X, Chen Z, Tian Y.: A lightweight attribute-based encryption scheme for the internet
of things. Future Gener Comput Syst 49:104–112 (2015).
14. S.R. Moosavi, E. Nigussie, S. Virtanen, J. Isoaha.: An elliptic curve-based mutual
authentication scheme for RFID implant system, in: International Conference on Ambient
Systems, Network and Technologies, 32, pp. 198-206 (2014).
15. Y.P. Liao, C.M Hsiao.: A secure ECC- based RFID authentication scheme integrated
with ID-verifier transfer protocol, Ad Hoc Networks, 18 133-146 (2014).
16. M. Turkanovic, B. Brumen, M. Holbl.: A novel user authentication and key agreement
scheme for heterogeneous ad hoc wireless sensor networks based on the Internet of
Things notion, Ad Hoc Networks, 20 96-112 (2015).
17. P. Persson, O. Angelsmark, Calvin.: Merging cloud and IoT. International Conference on
Ambient Systems, Network and Technologies 52 pp. 210-217 (2015).
18. S.K. Sood, A.K. Sarje , K. Singh.: Dynamic identity-based single password anti-phishing
protocol, Security and Communication Networks. 4(4) 418-427 (2009).
19. P. Poornachandran, Ashok, A., AU, P. Shankar, VG, S., Krishnamoorthy, S.,
Bharataraju, M., Panchagnula, S. K., and Khyati, D., “MobAware’-Employing Context
Awareness, Sensors and Cloud for Automatic Emergency Help Requests”, in
International Conference on Ambient Assisted Living Technologies based on Internet of
Things (2016).
20. Kumari S, Karuppiah M, Das AK, Li X, Wu F, Kumar N.: A secure authentication
scheme based on elliptic curve cryptography for IoT and cloud servers. The Journal of
Supercomputing, 1-26 (2017).
International Journal of Pure and Applied Mathematics Special Issue
1582
1583
1584