Content uploaded by Suyel Namasudra
Author content
All content in this area was uploaded by Suyel Namasudra on Dec 01, 2023
Content may be subject to copyright.
1
Securing IoT-based Smart Healthcare Systems by
using Advanced Lightweight Privacy-Preserving
Authentication Scheme
Sangjukta Das, Member, IEEE, Suyel Namasudra, Member, IEEE, Suman Deb, Member, IEEE, Pablo Moreno Ger, and Ruben
Gonzalez Crespo, Senior Member, IEEE
Abstract—In the healthcare network, the Internet of Things
(IoT) devices are connected to the network for enabling remote
monitoring of patients’ health. IoT device security, however, is a
serious concern because typical security measures might not be
appropriate for IoT devices, making them naturally vulnerable to
physical and copying attacks. Therefore, device authentication is
a very essential security concern for IoT networks. Additionally,
the storage and processing power of these devices are
constrained. To address all these requirements, Physically
Unclonable Functions (PUFs) for device authentication is a
potential strategy. In this paper, an advanced lightweight
authentication scheme for IoT devices is proposed by using PUF.
This scheme provides robust authentication without storing any
sensitive information on the device’s memory and establishes the
session key exchange process simultaneously. Moreover, this
scheme preserves device privacy by including a temporary
identity, which is updated at the end of each session. The
effectiveness of this novel model is assessed, and results
demonstrate that it is more effective and secure than many
existing schemes.
Index Terms— Untracebility, Key Agreement, Anonymity, PUF.
I. INTRODUCTION
HE Internet of Things (IoT) has recently become one of
the most popular research topics in both industry and
academia. IoT refers to a network of objects, such as
sensors, actuators, embedded technology, and smartphones,
which are connected by Internet connections. Nowadays, IoT
devices are being used by households, workplaces, major
corporations, etc., to have network connectivity and to
exchange data. One of the numerous applications of IoT is the
smart home, which uses IoT along with machine learning
techniques to get cost-efficient solutions for energy
management with great accuracy [1]. Intelligent transportation
system is another application that uses this technology for
traffic management and sustainable transportation planning
S. Das is with the Department of Computer Science and Engineering,
National Institute of Technology Patna, Bihar, India. Email:
sangjukta24@gmail.com
S. Namasudra and S. Deb are with the Department of Computer Science
and Engineering, National Institute of Technology Agartala, Tripura, India.
Email: suyelnamasudra@gmail.com, sumandeb.cse@nita.ac.in.
P. M. Ger and R. G. Crespo are with the Universidad Internacional de La
Rioja, Logroño, Spain. Email: {pablo.moreno, ruben.gonzalez}@unir.net
[2]. IoT is also extensively being used in the healthcare
domain as well as in the agricultural domain and
environmental ecosystems for digital monitoring [3, 4].
However, connecting these devices to the cloud raises data
security risks and makes it possible for any unauthorized user to
access data available on an IoT network. The implementation of
IoT devices on a greater scale also causes many security attacks
on both physical and network levels [5]. For example, if an
attacker is able to access the devices, s/he can perform a variety
of physical attacks to learn the secrets stored in devices and
corrupt both devices and the system as a whole. IoT devices are
vulnerable to cyberattacks as they lack strong security protocols
because of their limited resources to process complex
operations. Therefore, while deploying IoT devices on a
healthcare network, it is crucial to take security and privacy
requirements and problems into account. Different facets of
security for IoT applications have been studied by researchers.
Yet, IoT devices have significant challenges in maintaining data
security and privacy due to device heterogeneity, resource-
constrained nature, etc. To solve these security issues in an IoT-
enabled healthcare network, numerous studies have designed
and deployed effective solutions based on digital certification,
access control, and authentication [6]. In [7], one lightweight
protocol for user authentication is proposed by using
cryptographic operations and biometric information. A mutual
authentication scheme is suggested in [8] for a heterogeneous
IoT environment. This scheme maintains user privacy by
providing anonymity, and it cannot resist security attacks like
impersonation attacks. Although conventional authentication
methods are considered as secure, an attacker can use a variety
of physical attacks to fraudulently capture the data stored on
an IoT device [9]. Here, a two-factor authentication system
can solve the aforementioned issues by ensuring layered
defense, thus, making it more difficult for unauthorized users
to access IoT devices [10].
In the above context, the PUF is one of the most
dependable and robust security functions used to secure IoT
devices [11]. PUFs are designed based on digital logic and
Integrated Circuits (ICs) are acknowledged as a promising
primitive for hardware security. They are hardware modules
that operate as one-way operations. It is difficult to replicate
these operations since they provide different outputs for the
same inputs [12]. The PUFs are very helpful for enhancing
security in devices that cannot support complicated
T
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
2
cryptographic operations. As a result, they are quite beneficial
in IoT networks with limited resources. The PUF can be
carefully embedded inside the IoT device to make it
unclonable and uniquely identifiable. In an area like the
healthcare sector, IoT technology is extensively being used for
both pre and post-operational monitoring, medication, and
medical alerting. Here, PUF-based security can provide
privacy and authorized access to patient data. PUF-based
authentication can support some factors, such as fingerprints
or biometrics along with generic authentication factors like
passwords and tokens [13]. In [14], a PUF-based lightweight
mutual authentication protocol is designed for IoT
applications. The authors have implemented and analyzed the
performance of this scheme in terms of energy, memory, and
power utilization. Another PUF-based mutual authentication
protocol is proposed in [15] for IoT systems. Here, the IoT
device only uses PUF and does not use secret keys for the
authentication process. However, these schemes cannot ensure
the privacy of IoT devices. To solve the privacy issue, one
two-factor lightweight authentication scheme using PUF is
proposed in [16] that preserves the privacy of IoT devices.
However, PUF-based authentication protocols are often
vulnerable to many security threats like message tampering,
mutual authentication threat, key-agreement attacks, physical
and side-channel attack, impersonation attacks, etc. Till date,
many PUF-based protocols are proposed and validated using
various logical methods to address these issues and to protect
IoT devices. As IoT devices in a healthcare system generate
critical and sensitive data, it requires security mechanisms in
order to maintain the anonymity and privacy of its user or
patients.
To address all the above-mentioned issues, a lightweight
and privacy-preserving authentication scheme for the IoT-
based healthcare environment is proposed in this paper. Here,
the pseudonym identity of each device is generated by using
random integers and PUF outputs are used for authentication
purposes. PUFs provide a distinctive hardware fingerprint to
the devices by taking advantage of the natural random changes
present in an integrated circuit. As a result, the proposed work
is anonymous and safe against user identity profiling attacks.
Below are the main contributions of this work:
1) This work presents a lightweight device authentication
scheme by using PUFs to provide security to the data in
healthcare systems.
2) It provides privacy-preserving, anonymous identity and
untraceability to the devices, and enables devices to
authenticate themselves without revealing their details.
3) Here, at the end of each communication, this protocol
updates its credentials, as well as anonymous identities,
which improves data security against cyberattacks.
The rest of the paper is organized as follows. Related works
and preliminary studies are presented in sections II and III.
Sections IV and V present the overview and the construction
of the proposed scheme, respectively. Sections VI and VII
discuss the security and performance analysis of the proposed
technique, respectively. Finally, the entire work is concluded
with future works in section VIII.
II. RELATED WORKS
In the literature, some works related to the proposed work
are proposed to authenticate and generate a session key for
protecting data. In [17], one authentication scheme is proposed
for low-power mobile devices. This scheme is susceptible to
password brute-force attacks due to the lack of fundamental
security requirements. Masud et al. [18] have proposed a
lightweight mutual authentication scheme to create a secure
channel between the device and the user. Although this secure
channel between the user and device prevents unauthorized
users from getting access to network data, this scheme cannot
resist attacks like device capture attacks. Another
authentication scheme for network nodes based on biometric
data is proposed by Koya et al. [19]. This gives better security
by combining the patient’s electrocardiogram signals with the
authentication protocol. However, this scheme faces
untraceability and key-escrow issues. This scheme is
improved by Gupta et al. [20] by including an anonymous
authentication and key agreement technique. Still, scalability
issues exist in the scheme of Gupta et al. [20] because of high
communication and computation overheads. Also, there exist
many schemes based on mutual authentication [21-23].
Many Radio Frequency Identification (RFID) systems and
wireless sensor networks use PUFs to achieve secure
authentication methods [24]. A double PUF-based RFID
identity authentication protocol is proposed by Liang et al.
[25]. This scheme is vulnerable to denial-of-service attacks
because the messages in this scheme are not authenticated.
Alladi et al. [26] have designed a mutual authentication
scheme using the Challenge Response Pair (CRP) of the PUF
for IoT-enabled healthcare systems. It is a two-phase
authentication process to increase physical security against
node tampering and node replacement attacks in the healthcare
system. However, this scheme is vulnerable to many attacks as
the CRPs are stored in the database of the device during the
registration process. Another PUF-based authentication for the
Internet of Medical Things (IoMT) is proposed by Yanambaka
et al. [27]. In this scheme, both the server and the IoMTs are
PUF-equipped and the gathered CRPs are stored in a third-
party database. However, the messages are not encrypted in
this scheme, when they are exchanged between different
entities. Due to this reason, this scheme is simple to undertake
modeling attacks. Additionally, PUF response noise correction
is not considered in this scheme. Another authentication and
key agreement technique is suggested by Wang et al. [28] to
ensure secure communication between IoT nodes. This is a
PUF-based technique that uses lightweight cryptographic
operations and the reverse fuzzy extractor to re-produce
responses correctly in a noisy environment. However, the
higher number of cryptographic processes used in this scheme
increases overall computation time. Many PUF-based
authentication schemes are also proposed by researchers that
use computationally expensive public key cryptography [29-
32]. Most of these schemes do not provide the anonymity
feature of IoT security protocol. Many other alternative mutual
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
3
authentication systems based on advanced technologies are
found in the literature [33-37].
III. PRELIMINARY STUDIES
A. Physical Unclonable Function
Physical Unclonable Function is an IC that can output an
arbitrary string of bits called the response from a string of bits
known as a challenge. PUF provides a unique CRP due to the
random variances during the fabrication process of ICs [38].
In a CRP, , i.e., PUF P's response to a challenge
identifies a PUF. Every PUF responds uniquely to the same
challenge, indicating that each PUF is unique. However, the
output of a PUF may be impacted by environmental
conditions, including temperature and voltage. The use of
fuzzy extractors may circumvent this issue and provide robust
PUF replies suitably for security applications [39].
B. Fuzzy Extractor
The Fuzzy Extractor generates probabilistic keys by
using two algorithms, namely key Generation (FE.Gen) and
key Reconstruction (FE.Rec) [39]. FE.Gen generates a key
and helper data , from an input bit string as
. FE.Rec can reconstruct from a noisy input
by using as ata).
IV. OVERVIEW OF THE PROPOSED SCHEME
The system model and design goals of the proposed scheme
are discussed in this section.
A. System Model
The system model considered in the proposed scheme is
similar to the system model used in [40]. Here, the goal is to
create mutual authentication and direct communication between
any two communicating entities. In Fig. 1, a simplified
representation of the system model is shown. The two entities
considered in the proposed system model are IoT Device
(IoTD) and Central System (CS). Here, it is also assumed that
IoT devices have resource limitations, while the server does
not have any such limitations and the server is trusted. The
role of each entity is defined below:
1) IoT Device: An IoT device is associated with a patient’s
body, collects real-time data, and sends them to the
server via a gateway device. The device must validate its
authenticity before sending it to the CS. Here, the device
is a resource-constrained node. Through the Internet, IoT
devices interact and send data to the server. It is
considered that every device has a PUF. Any effort to
tamper with the PUF causes the device to behave
functionally differently.
2) Central System: The device shares the collected data
with the CS. The CS initializes the entire system at the
beginning and registers all other entities in the system.
The CS authenticates each device before creating a
secure channel for data exchange.
B. Design Goals
While designing the proposed scheme, a few goals are
considered, which are mentioned below:
1) Privacy Preservation: Sensitive parameters like identity-
related information can be used by attackers to carry out
impersonation attacks, Man-In-The-Middle (MITM)
attacks, or physical attacks. Thus, it is crucial to maintain
privacy, when exchanging data over a network.
2) Lightweight: IoT devices' capacities for computation are
constrained. Therefore, to support the processing
capacity of devices, any security mechanism must be
designed by using lightweight cryptography processes.
3) Message Integrity: If any malicious user changes the
healthcare data, the entire system may be compromised.
So, the integrity of healthcare data must be preserved
through some advanced and unbreakable procedures.
4) Mutual Authentication: The healthcare network has
many devices, users, and associated equipment. To
establish a secure communication channel, each device
and user must authenticate themselves with the central
server and agree on a session key.
Fig. 1. System model of the proposed scheme
TABLE I
SYMBOL DESCRIPTION
Symbol
Description
Temporary identity
Device identity
{, }
Initial CRP
, }
Additional CRP set
{ }
Random nonce
PUF operation
Hash operation
Valid time range
Current timestamp
New temporary identity
Intregrity checker for meassage
Secret session key
, ,
Messages exchanged during registration
, ,
Messages exchanged during Authentication
V. CONSTRUCTION OF THE PROPOSED SCHEME
The proposed scheme using a PUF is presented in this
section for mutual authentication and key exchange. Along
with PUF, the fuzzy extractor is also employed in the
proposed protocol to reproduce the same keys. The proposed
scheme is constructed in four phases, namely system
initialization, device registration, authentication, and update
phase. The workflow during different phases of the proposed
IoT Device
IoT Device
IoT Device
Data Unit
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
4
Fig. 2. Proposed scheme’s workflow
scheme is illustrated in Fig. 2 and Table I lists the main
symbols used in this protocol.
A. System Initialization
The central server initializes the system and selects its key
values like a master key, private and public key, and other
parameters. The CS uses this master key during the device
registration process to generate temporary identities of
devices.
B. Device Registration
Initially, each device enrolls itself with the CS by
generating CRP, temporary identity, and other random values.
At the end of this process, the CS stores the primary CRP,
,
and for each device. Along with this information, the CS
stores another set of CRP to prevent an emergency situation
like DoS attacks. However, the device stores its
, ,
primary Challenge , and Challenge set . The device
does not store its response string corresponding to and .
One of the most important aspects of this scheme is that the
device does not store any secrets, which prevents physical
attacks. The registration process initiated by the device is
described in the below steps. For simplicity, this process is
also shown in Fig. 3, where the messages exchanged during
this process are represented by , where .
Step 1: The device sends its identity to the CS through a
registration request , which is the first message during the
registration process.
Step 2: The CS generates
by calculating the hash of the
string , where is the master key of the CS. The
CS selects and Challenge set . Then, the CS sends the
second message containing
, , and to the device.
Step 3: The device extracts and from to generate
and . Finally, the device stores
, , and , and sends the third registration-related
message to the CS.
Step 4: On receiving the response strings corresponding to the
challenges sent through , the CS finally stores ,
, the
primary CRP, and the additional CRP set for the device.
The device registration is an offline process, and the device
does not store and . Both of these factors reduce the
possibility for an attacker to obtain crucial information about
the device.
Fig. 3. Device registration process
C. Authentication
In this subsection, the device authentication process is
explained in detail. To create a secure communication channel
for data transfer, the device initiates the authentication process
by sending an authentication request message. The pictorial
representation of the interaction between the device and the
central system is given in Fig. 4. In Fig. 4, the messages
exchanged during the authentication phase are represented by
, where . The device authentication process
is executed by the following steps.
Step 1: The IoT device takes its
and from its secure
memory to generate the PUF response corresponding to as
. Then, it encrypts by using and forms a
message
. The integrity checker of
is also computed as , where is the
current timestamp. The device sends the authentication request
to the server.
Step 2: The CS receives and checks the validity of .
Then, the CS searches for the
in its database and takes the
R1. Registration request
R2. Challenge, Temporary Id
R3. Response
1. Device registered
IoT Device
IoT Device
IoT Device
IoT Device
A1. Authentication request
A2. Response parameters
A3. Confirmation
2. Device Authenticated
Device
CS
Store
Store
,{ ,
{
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
5
Fig. 4. Authentication process
primary CRP, i.e., {, } pair. Again, the CS checks the
integrity of the message by calculating the integrity checker
of . If any of these validation processes are failed, the
subsequent steps are not executed and the entire process is
terminated. The CS selects random and calculates the
message after encrypting { } by . The CS forms
the second message after computing the
integrity checker as and sends it to the
device. Here, the used in and is a new timestamp
selected by the CS.
Step 3: The device verifies the validity of
. If
is valid, the
device obtains the random nonce from the message by
using , and again, checks the integrity message by
calculating . Then, the device calculates the
secret session key ,
, , and . The device sends
the third authentication message to the CS.
Step 4: After receiving , the CS, checks the validity of
.
Then, it verifies the third integrity by verifying
and
=
. If
is valid, the CS stores
and of the device.
Here, is the shared secret session key between the device
and the server. This key is valid for only this session. Once the
data exchange is completed, the session ends and is
discarded. Thus, the mutual authentication and key agreement
processes are completed, which are also shown in Fig. 4.
D. Credential Update
To maintain the freshness assurance of the authentication
protocol [34], the server may update the related CRPs for each
device by acquiring new CRPs. The CRP updating process is
shown in Fig. 5 and the steps are discussed below:
Step 1: The server sends an update request with the new
challenge to the device to change the current CRP. During
this phase,
of the device can also be updated.
Step 2: On receiving the update request , the device
calculates a new response corresponding to and
second update message . Then, the device sends and
to the CS. At this stage, the device changes the challenge
stored in its memory.
Fig. 5. Credential update phase
Step 3: On receiving and , the CS verifies the
timestamp and recalculates . If it matches with the received
, then, the CS updates its CRP list with { .
VI. SECURITY ANALYSIS
This section begins with briefly introducing a number of
attacks against PUF-based IoT authentication methods. Then,
the proposed protocols’ informal and formal security analyses
are presented. The robustness of the proposed scheme, i.e.
Advanced Lightweight Privacy-Preserving Authentication
(ALPAS) is assessed against some well-known attacks.
A. Formal Analysis
This section formally analyses the security of ALPAS by
using Dolev-Yao (DY) and Canetti-Krawczyk (CK) adversary
models, and their assumptions mentioned in [40]. ALPAS has
two main entities, namely Device and CS. According to the
threat model, an adversary has capability to capture,
corrupt, alter, delete, or replay all messages sent over the
Device
CS
Verify
Verify
Verify
Verify
Verify
Verify
Store
,
Data exchange by using
Device
CS
Store
{,}
Verify-
Store {
{, }
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
6
communication channel. can perform the following queries:
1) : can eavesdrop over the
communication channel between a device and the CS to
get all the messages exchanged by executing this query.
2) : By executing this query, can send
messages to a device and the CS, and can also receive a
reply from them.
3) can execute query
to capture all the parameters stored in the device and
CS’s memory. However, can execute only a limited
number of query.
4) runs this query to reveal the secrets stored
in the device’s memory using a physical attack.
Considering is the event, where wins a game, then,
’s advantage to break ALPAS is .
If , ALPAS is secure, where .
Lemma 1. The output of a PUF cannot be guessed.
Proof. According to [29], a PUF cannot be replicated and it
generates unique responses. A PUF produces a response of n
bits for a challenge of m bits such that . A
security game between an adversary and challenger can be
performed as follows. can send a query to the PUF using
challenge polynomial times. At first, sends a challenge
to . reveals as PUF() to and sends another
challenge to get as PUF(). Then, the
adversary wins the game, if its guess response
for
is the same as . This indicates ’s advantage in this game
is
. can only guess the output of a
PUF to a given challenge. Therefore,
.
Lemma 2. The secrets used in ALPAS cannot be revealed
by the Reveal oracle. The temporary identities of devices also
cannot be correlated by the Reveal oracle.
Proof: In ALPAS, IoT devices do not store any secret or
sensitive data in their local memory. The device stores only its
, , and . According to the assumptions in the threat
model [40], cannot get by using or invoking the
Reveal oracle. Initially, the temporary identity
of a device
is calculated as , which is updated at each new
round. Each
is therefore only valid for a single session. As
a result, cannot correlate the temporary identities unless
is able to receive the secret response, which is impossible. In
this case, ’s advantage is
, where is the correlation coefficient.
Theorem 1. Mutual Authentication: This protocol can be
successfully executed between the device and the CS only if
both entities are legitimate.
Proof: By impersonating an authentic device, can try to
establish authentication with the server. To simulate this
attack, a security game between and can be performed as
follows. Initially, to perform the proposed authentication with
the CS, selects any legitimate device, namely . can
send a polynomial number of query requests to the CS and .
attempts to authenticate itself as a valid device to the CS. If
successfully completes the authentication step of ALPAS,
wins the game. If is able to produce the third integrity
checker
, only then, it can properly
authenticate itself. can try to reveal embedded within
the . Suppose is able to reveal bits of , where
. Then, the advantage
is
.
Therefore, ’s advantage of the successful authentication
process with the CS is
.
However, can only randomly guess , i.e., by
Lemmas 1 and 2 and
. So,
.
Theorem 2. Privacy: The proposed ALPAS maintains the
anonymity of the device.
Proof: The ALPAS protocol is said to be untraceable if
cannot correlate two executions of ALPAS by the same
with the CS. The following security game can be used to
analyze this attack. Initially, to perform the proposed scheme
with the CS, selects two valid devices and . can
send polynomial numbers of query requests to the CS and
devices and . Then, selects one of the device’s
identity randomly. sends a query to the CS and
polynomial times. Then, guesses the identity . Now, if
==, wins the game.
Here, ’s advantage of guessing successfully can be
presented as
=
. As IoTD’s
s cannot be correlated, ’s advantage of correlating
can be presented as
.
’s advantage of winning this game can be presented as
-
. If
guesses randomly, then, s/he has no advantage. By
Lemmas 1 and 2, it can be concluded that
.
B. Informal Analysis
To analyze ALPAS informally, a few attack scenarios are
considered in this paper.
1) Replay Attack: Since a valid timestamp is assigned to each
transmitted message, even if an attacker replays old
messages, it cannot counterfeit the current transmitted
message. Furthermore, every parameter, including the
temporary identity of devices is updated after every new
session. As a result, replay attacks are successfully
avoided. In this context, to detect replay attacks,
maintaining the freshness of each exchanged message is an
important requirement. To fulfill this requirement, the
ALPAS uses the timestamp concept, and also, allows the
entities to use different credentials during different
sessions, which is done via the credential update phase.
2) Message Analysis Attack: In Message Analysis attacks, an
attacker tries to intercept the transmitted information
between the communication entities. Despite the
possibility of intercepting authentication communications,
in the ALPAS technique, secret keys, session keys, and
responses are private and inaccessible to an attacker. This
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
7
is accomplished by not keeping the transferred messages
locally, but, encrypting and hashing them.
3) DOS Attack: In DOS attacks, a device is targeted by an
attacker to temporarily or permanently interrupt its
functionality by overloading it with service requests. In
ALPAS, only two entities are present, namely CS and
IoTD. On the CS’s side, DOS attacks are unfeasible
because of the server's high computing capabilities.
Therefore, only the device is considered for this attack.
The device verifies the integrity of each message after
receiving any message. Due to the secret key contained in
each integrity checker message, the possibility of random
guessing of the hash values to pass the verification
procedure is very less. The DoS attack is therefore
impractical in ALPAS.
4) Physical Attack: In the proposed model, IoT devices do
not keep any secret or sensitive data in their local memory.
Additionally, in the system model, one assumption is that
the communication between the PUF IC and the device is
secure. Therefore, if an adversary gets a device, ALPAS is
secure against physical attacks.
C. Formal Verification
Fig. 6. Summary produced by the AVISPA tool’s two back
ends (OFMC and CL-AtSe)
In this section, the proposed protocol is simulated using the
widely used protocol security analysis and verification tool
AVISPA (Automated Validation of Internet Security Protocols
and Applications). There are four back-ends (OFMC, CL-
AtSe, SATMC, and TA4SP) integrated into AVISPA to create
a single platform for protocol verification. To define the roles
and goals of the proposed protocol, AVISPA uses the formal
language HLPSL (High-Level Protocol Specification
Language). The backends cannot directly detect HLPSL,
therefore, it is converted into Intermediate Format (IF) using
the platform’s HLPSL2IF translator. Then, the IF is directly
built and executed in backends to verify the protocol’s
security.Two different entities, namely device and central
server, are included in the proposed protocol. As a result, two
roles in AVISPA using the HLPSL specification are defined.
The role definition codes contain corresponding operations,
states, and parameters. The role of the session and
environment are also considered. Fig. 6 depicts the output of
CL-AtSe and OFMC backends. It indicates that ALPAS is
SAFE against many attacks, including MITM attacks, replay
attacks, and impersonation attacks, and the confidentiality of
the session key is maintained.
VII. PERFORMANCE ANALYSIS
The performance analysis of ALPAS is shown in this
section in comparison to the relevant protocols [15, 16, 27, 28]
in the literature. These protocols are based on some features,
such as mutual authentication and error correction. Since
SHA-2 is currently used in well-known security applications
like Transport Layer Security (TLS) and Secure Sockets Layer
(SSL), and a number of integrated circuits for commercial
security, ALPAS is implemented by using SHA-2.
Additionally, SHA-2 is easier and quicker to implement than
SHA-3 since a wider range of hardware and software is
supported by it. At first, the security properties are compared.
Then, the storage requirement, computation, and
communication complexities of ALPAS are assessed.
A. Security Feature Comparison
Table II compares ALPAS’s security properties to various
existing schemes. The schemes proposed in [27, 28] do not
provide anonymity and un-traceability properties. Similarly,
the scheme [16] does not provide resistance to reply attacks.
However, ALPAS has all of the properties mentioned in the
table.
B. Storage Requirement Comparison
The overall cost of storage of each entity in ALPAS is
determined based on the size of the parameters given in Table
III. Table IV lists the cost of storage in ALPAS along with
other existing schemes. Each device stores {
, , } and
the CS stores ,
, CRP} parameters for each device in
their memory. In Fig. 7, the total storage cost and the
individual storage costs for the CS and the device are shown.
C. Discussion on Experimental Results
In this section, the performance of ALPAS is evaluated in
terms of computational cost and communicational cost.
1) Computation Cost: The computational cost of the
authentication process of ALPAS is computed by
executing a number of operations. The performance of
ALPAS is compared with the existing schemes by
considering the same conditions and operations. The basic
operations that are used in ALPAS and other related
schemes are hash, XOR, concatenation (||), PUF, and noise
correction. For cost evaluation, only PUF and hash
operations are considered since the other operations are
comparatively very less time-consuming. Table V
represents the time taken to execute each operation by
device and server. Then, the number of operations used in
the existing protocols and the proposed scheme is
compared, which is listed in Table VI. The computation
cost comparison is graphically represented in Fig. 8. It can
be seen that ALPAS uses comparatively less operations
than the other schemes, thus, the computation cost is
10+1
. It is seen that ALPAS takes comparatively less
execution time than the other protocols.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
8
TABLE II
COMPARISONS OF SECURITY FEATURES
Security Feature
Aman et al. [15]
Gope et al. [16]
Yanambaka et al. [27]
Wang et al. [28]
Proposed Scheme
Resistance to replay attack
√
˟
˟
√
√
Anonymity
√
√
˟
˟
√
Traceability
√
√
˟
˟
√
PUF security
˟
√
√
˟
√
Resistance to physical attack
√
√
√
˟
√
Noise consideration in PUF
˟
√
˟
˟
√
TABLE III
SIZE OF THE PARAMETERS
Parameter
Length in Bits
Parameter
Length in Bits
Secret keys
160
Timestamp
32
Hash
256
Challenge
160
Identity parameters
160
Response
160
Random nonce
160
-
-
TABLE IV
STORAGE COST COMPARISON (BITS)
Scheme
Device
CS
[16]
1576
1792
[28]
576
480
Proposed scheme
420
1116
Fig. 7. Storage cost in bits
TABLE V
RUNTIME OF EACH OPERATIONS IN MILLISECONDS
Operation
CS
Device
PUF
-
0.14ms
Hash
0.012ms
0.028ms
XOR
0.003ms
0.005ms
FE.Gen
-
2.7ms
Fe.Rec
-
4.23ms
TABLE VI
COMPARISON OF COMPUTATION COST
Scheme
Device
CS
Total
[16]
7+2
7
14+2
[28]
7+2
7
14+2
Proposed scheme
5+1
5
10+1
2) Communication Cost: Here, the communication cost
means the total number of bits sent and received
throughout the authentication procedure. The length of
each message being delivered is determined by using Table
III, which provides the size of the parameters used in these
messages. In Table VII, the communication costs of
ALPAS and the existing schemes are compared. This table
also shows the total number of messages sent and received
by the communicating entities, namely IoT device and CS.
From Tables III and VII, it can be seen that the transmitted
bits in ALPAS are 1504 bits that are less than [16] and
[28].
Fig. 8. Computation cost in time (ms)
TABLE VII
COMPARISON OF COMMUNICATION COST
Scheme
Total no. of Messages
Total no. of Bits
[16]
3
1568
[28]
3
1568
Proposed scheme
3
1504
VIII. CONCLUSIONS AND FUTURE WORK
In this paper, a device-to-central server mutual
authentication and key exchange protocol has been developed
for IoT devices in healthcare systems. The proposed protocol
aims to establish a secure channel between the communicating
healthcare entities for the secure exchange of sensitive and
confidential healthcare data. Here, each device equipped with
PUF must register itself with the central server system. This
scheme eliminates the requirement to store CRPs in the
device’s local memory, which not only satisfies the resource
limitation of IoT devices, but also reduces the security risk of
device node attacks due to the accessibility to these devices.
The proposed ALPAS also creates a session key, and securely
exchanges it between the device and server at the end of each
successful authentication phase. Furthermore, it is
demonstrated that ALPAS is secure against many advanced
cyber attacks, namely replay, MITM, DoS, and impersonation
attacks. Most importantly, it resists physical attacks by using
the PUF-based authentication technique. As a future work, it is
planned to deploy this PUF-based authentication protocol with
a blockchain architecture scheme to provide security and
automation to IoT-based healthcare systems.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
9
REFERENCES
[1] W. Li, T. Logenthiran, V. -T. Phan, and W. L. Woo, “A novel smart
energy theft system (SETS) for IoT-based smart home”, IEEE Internet
of Things Journal, vol. 6, no. 3, pp. 5531-5539, 2019.
[2] Z. Huang et al., “Survey on vehicle map matching techniques”, CAAI
Transactions on Intelligence Technology, vol. 6, no. 1, pp. 55-71, 2021.
[3] P. Gangwani, A. Perez-Pons, T. Bhardwaj, H. Upadhyay, S. Joshi, and
L. Lagos, “Securing Environmental IoT Data Using Masked
Authentication Messaging Protocol in a DAG-Based Blockchain: IOTA
Tangle,” Future Internet, vol. 13, no. 12, pp. 312, Dec. 2021
[4] S. Das and S. Namasudra, “Multi-authority CP-ABE-based access
control model for IoT-enabled healthcare infrastructure”, IEEE
Transactions on Industrial Informatics, vol. 19, no. 1, pp. 821-829,
2023.
[5] A. Gutub, “Boosting image watermarking authenticity spreading secrecy
from countingbased secretsharing”, CAAI Transactions on Intelligence
Technology, 2022. DOI: 10.1049/cit2.12093
[6] S. Das and S. Namasudra, “A lightweight and anonymous mutual
authentication scheme for medical big data in distributed smart
healthcare systems”, IEEE/ACM Transactions on Computational
Biology and Bioinformatics, 2022. DOI: 10.1109/TCBB.2022.3230053.
[7] X. Li, J. W. Niu, J. Ma, W. D. Wang, and C. L. Liu, “Cryptanalysis and
improvement of a biometrics-based remote user authentication scheme
using smart cards”, Journal of Network and Computer Applications, vol.
34, no. 1, pp. 73–79, 2011.
[8] T. Muhamed, B. Boštjan, and H. Marko, “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, vol.
20, pp. 96–112, 2014.
[9] C. Chang, and H. Le, “A provably secure, efficient, and flexible
authentication scheme for ad hoc wireless sensor networks”, IEEE
Transactions on Wireless Communications, vol. 15, no. 1, pp. 357–366,
2016
[10] M. K. Khan and K. Alghathbar, “Cryptanalysis and security
improvements of twofactor user authentication in wireless sensor
networks”, Sensors, vol. 10, no. 3, pp. 2450–2459, 2020.
[11] A. Lakhan, M. A. Mohammed, J. Nedoma, R. Martinek, P. Tiwari, and
N. Kumar, “Blockchain-enabled cybersecurity efficient IIOHT cyber-
physical system for medical applications”, IEEE Transactions on
Network Science and Engineering, 2022, DOI:
10.1109/TNSE.2022.3213651
[12] C. Herder, M.-D. Yu, F. Koushanfar, and S. Devadas, “Physical
unclonable functions and applications: A tutorial,” Proc. IEEE, vol. 102,
no. 8, pp. 1126–1141, Aug. 2014.
[13] O. Günlü, O. İşcan, V. Sidorenko, and G. Kramer, “Code constructions
for physical u nclonable functions and biometric secrecy systems”, IEEE
Transactions on Information Forensics and Security, vol. 14, no. 11, pp.
2848-2858, 2019.
[14] M. Hossain, S. Noor, R. Hasan, HSC-IoT: A Hardware and software co-
verification based authentication scheme for internet of things, in: The
Proceedings of 2017 5th IEEE International Conference on Mobile
Cloud Computing, Services, and Engineering (MobileCloud), San
Francisco, CA, 2017, pp. 109–116,
[15] M. N. Aman, K. C. Chua, and B. Sikdar, “Mutual authentication in IoT
systems using physical unclonable functions,” IEEE Internet of Things
Journal, vol. 4, no. 5, pp. 1327–1340, 2017.
[16] P. Gope and B. Sikdar, “Lightweight and privacy-preserving two-factor
authentication scheme for IoT devices,” IEEE Internet of Things
Journal, vol. 6, no. 1, pp. 580–589, 2019
[17] X. Li, J. Peng, M. S. Obaidat, F. Wu, M. K. Khan, and C. Chen, “A
secure three-factor user authentication protocol with forward secrecy for
wireless medical sensor network systems”, IEEE Systems Journal, vol.
14, pp. 39–50, 2019.
[18] M. Masud, G. S. Gaba, K. Choudhary, M. S. Hossain, M. F. Alhamid,
and G. Muhammad, “Lightweight and anonymity-preserving user
authentication scheme for IoT-based healthcare”, IEEE Internet of
Things Journal, vol. 9, no. 4, pp. 2649-2656, 2022.
[19] A. M. Koya, and D. P. P., “Anonymous hybrid mutual authentication
and key agreement scheme for wireless body area network”, Computer
Networks, vol. 140, pp. 138–151, 2018.
[20] A. Gupta, M. Tripathi, and A. Sharma, “A provably secure and efficient
anonymous mutual authentication and key agreement protocol for
wearable devices in WBAN”, Computer Communications, vol. 160, pp.
311–325, 2018
[21] S. Das and S. Namasudra, “Lightweight and efficient privacy-preserving
mutual authentication scheme to secure Internet of Things-based smart
healthcare”, Transactions on Emerging Telecommunication Technology,
2023. DOI: 10.1002/ett.4716.
[22] Y. K. Huang, “Design of a smart cabin lighting system based on internet
of things”, Cloud Computing and Data Science, vol. 4, no. 2, pp. 112-
121, 2023.
[23] Z. Chen, “Research on internet security situation awareness prediction
technology based on improved RBF neural network algorithm”, Journal
of Computational and Cognitive Engineering, vol. 1, no. 3, pp. 103-108,
2022
[24] S. Karda¸s, S. Çelik, M. Yıldız, and A. Levi, “PUF-enhanced offline
RFID security and privacy,” Journal of Network and Computer
Applications, vol. 35, no. 6, pp. 2059–2067, 2012.
[25] W. Liang, S. Xie, J. Long, K. Li, D. Zhang, K. Li, “A double PUF-based
RFID identity authentication protocol in service-centric internet of
things environments”, Information Sciences, vol. 503, pp. 129-147,2019.
[26] T. Alladi, V. Chamola and Naren, “HARCI: A Two-Way Authentication
Protocol for Three Entity Healthcare IoT Networks”, IEEE Journal on
Selected Areas in Communications, vol. 39, no. 2, pp. 361-369, 2021.
[27] V. P. Yanambaka, S. P. Mohanty, E. Kougianos, and D. Puthal, “Pmsec:
Physical unclonable function-based robust and lightweight
authentication in the internet of medical things”, IEEE Transactions on
Consumer Electronics, vol. 65, no. 3, pp. 388–397, 2019.
[28] H. Wang, J. Meng, X. Du, T. Cao, and Y. Xie, “Lightweight and
anonymous mutual authentication protocol for edge iot nodes with
physical unclonable function,” Security and Communication Networks,
vol. 2022, 2022.
[29] U. Chatterjee, R. S. Chakraborty and D. Mukhopadhyay, “A PUF-based
secure communication protocol for IoT”, ACM Transactions on
Embedded Computing Systems, vol. 16, no. 3, pp. 1-25, 2017
[30] H. Jennath, V. S. Anoop, and S. Asharaf, “Blockchain for healthcare:
securing patient data and enabling trusted artificial intelligence”,
International Journal of Interactive Multimedia and Artificial
Intelligence, vol. 6, pp. 15-23, 2020.
[31] Kumar, and S. Chand, “A provable secure and lightweight smart
healthcare cyber-physical system with public verifiability”, IEEE
Systems Journal, 2021. DOI: 10.1109/JSYST.2021.312055.
[32] S. Khasim and S. S. Basha, “An improved fast and secure CAMEL
based authenticated key in smart health care system”, Cloud Computing
and Data Science, vol. 3, no. 2, pp. 77-91, 2022.
[33] A. Kishor, C. Chakraborty, and W. Jeberson, “A novel fog computing
approach for minimization of latency in healthcare using machine
learning”, International Journal of Interactive Multimedia and Artificial
Intelligence, vol. 6, pp. 7-17, 2020.
[34] K. Lam and D. Gollmann, “Freshness assurance of authentication
protocols”, Proceedings of the European Symposium on Research in
ComputerSecurity, pp. 261-272, 1992.
[35] A. Lakhan et al., “Restricted boltzmann machine assisted secure
serverless edge system for Internet of Medical Things”, IEEE Journal of
Biomedical and Health Informatics, vol. 27, no. 2, pp. 673-683, 2023.
[36] Y. Guo, Z. Mustafaoglu, and D. Koundal, “Spam Detection Using
Bidirectional Transformers and Machine Learning Classifier
Algorithms”, Journal of Computational and Cognitive Engineering, vol.
2, no. 1, pp. 5–9, Apr. 2022.
[37] A. Lakhan et al., “Federated-learning based privacy preservation and
fraud-enabled blockchain IoMT system for healthcare”, IEEE Journal of
Biomedical and Health Informatics, vol. 27, no. 2, pp. 664-672, 2023.
[38] Dimitrios Schinianakis. Lightweight security for the internet of things: A
soft introduction to physical unclonable functions. IEEE Potentials,
38(2):21–28, 2019.
[39] Y. Dodis, J. Katz, L. Reyzin, and A. Smith, “Robust fuzzy extractors
and authenticated key agreement from close secrets,” in Advances in
Cryptology (CRYPTO) (Lecture Notes in Computer Science), vol. 4117.
Heidelberg, Germany: Springer, 2006, pp. 232–250
[40] Y. Zheng, W. Liu, C. Gu, and C. H. Chang, “PUF-based Mutual
Authentication and Key Exchange Protocol for Peer-to-Peer IoT
Applications”, IEEE Transactions on Dependable and Secure
Computing, 2022, DOI:10.1109/TDSC.2022.3193570.
This article has been accepted for publication in IEEE Internet of Things Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JIOT.2023.3283347
© 2023 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: National Institute of Technology Patna. Downloaded on June 13,2023 at 09:34:50 UTC from IEEE Xplore. Restrictions apply.
