Machine Learning-Based Malicious User Detection in Energy Harvested Cognitive Radio-Internet of Things

Abstract

Our simulation data is generated from energy vector using ED technique.

Full text

Machine Learning-Based Malicious User Detection in EnergyMachine Learning-Based Malicious User Detection in Energy
Harvested Cognitive Radio-Internet of ThingsHarvested Cognitive Radio-Internet of Things
This paper was downloaded from TechRxiv (https://www.techrxiv.org).
LICENSE
CC BY 4.0
SUBMISSION DATE / POSTED DATE
01-02-2021 / 03-02-2021
CITATION
MIAH, MD SIPON; Hossain, Mohammad Amzad; Ahmed, Kazi Mowdud; Rahman, Md. Mahbubur; Calhan, Ali;
Cicioglu, Murtaza (2021): Machine Learning-Based Malicious User Detection in Energy Harvested Cognitive
Radio-Internet of Things. TechRxiv. Preprint. https://doi.org/10.36227/techrxiv.13681672.v1
DOI
10.36227/techrxiv.13681672.v1

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 1
Machine Learning-Based Malicious User Detection in
Energy Harvested Cognitive Radio-Internet of
Things
Md. Sipon Miah, Mohammad Amzad Hossain, Kazi Mowdud Ahmed, Md. Mahbubur Rahman, Ali Calhan,
and Murtaza Cicioglu
Abstract—The Internet of things (IoT) is a network of
interconnected objects that are connected and controls
autonomous machines in the world. The cognitive radio
based Internet of things (CR-IoT) concept is a revolu-
tionary technology for the future of IoT that mitigates
the spectrum scarcity problem. However, each CR-IoT
user does not obtain a better sensing gain, an enhanced
sum rate, and a prolonged network lifetime in con-
ventional CR-IoT networks under the existing energy
harvesters due to both (i) underutilizing the reporting
framework and (ii) without separating normal and
abnormal (malicious) CR-IoT users. For these reasons,
we proposed machine learning (e.g. logistic regression
(LR), support vector machine (SVM) and k-nearest
neighbors (k-NN)) based malicious user detection in
energy harvested CR-IoT networks, where each CR-
IoT user will be powered by finite capacity batteries
and energy harvesters. The main contributions of this
paper: First, we reviewed the technological attributes
and platforms proposed in the current literature for
the sensing, sum rate, and network lifetime with se-
curity threats; Second, the proposed classification al-
gorithms using machine learning are divided into two
groups between normal and abnormal (malicious) CR-
IoT users; Third, this scheme is utilized the reporting
framework by only normal CR-IoT users where each
normal CR-IoT user is obtained a longer sensing time
slot; Fourth, as a proof-of-concept, the performance of
the proposed scheme is evaluated through numerical
experiment; Finally, this proposed scheme is greatly
achieved better sensing gain, enhanced sum rate, and
prolonged network lifetime, in comparison to the exist-
ing conventional schemes.
M. S. Miah is with the Department of Information and Commu-
nication Technology, Islamic University, Kushtia-7003, Bangladesh,
and also with the Department of Computer Science, National
University of Ireland Galway, Galway, Republic of Ireland e-mail:
m.miah1@nuigalway.ie (www.nuigalway.ie).
M. A. Hossain is with the Department of Computer Science,
National University of Ireland Galway, Galway, Republic of Ireland
e-mail: m.hossain3@nuigalway.ie (www.nuigalway.ie).
K. M. Ahmed is with the Department of Information and Commu-
nication Technology, Islamic University, Kushtia-7003, Bangladesh
e-mail: mowdud@ice.iu.ac.bd(www.iu.ac.bd).
M. M. Rahman is with the Department of Information
and Communication Technology, Islamic University, Kushtia-7003,
Bangladesh e-mail: mrahman@ice.iu.ac.bd(www.iu.ac.bd).
A. Calhan is with the Department of Computer
Engineering, Duzce University, Duzce, Turkey e-mail:
alicalhan@duzce.edu.tr(www.duzce.edu.tr).
M. Cicioglu is with Information Technologies Department, Min-
istry of National Education, Bolu, Turkey e-mail: murtazaci-
cioglu@gmail.com(www.bolu.meb.gov.tr).
Manuscript received January 19, 2021; revised August 26, 2021.
Index Terms—Cognitive radio, internet of things,
machine learning, spectrum sensing, detection perfor-
mance, sum rate, network lifetime.
I. Introduction
INTERNET of things (IoT) consists of objects or things
that are meant to sense their surrounding environ-
ments, exchange, and communicate information or data
with others devices (e.g., computers, mobiles, cordless
phones, any wired or wireless devices) with different ap-
plications [1], [2]. The number of devices connected to
either public or private networks via wired or wireless has
recently increased significantly [3]. However, this dramatic
increase in IoT causes spectrum shortage that will be
resulting in degradation in detection performance because
of the massive number of IoT devices are connecting to
the licensed band.
Since the last decade, the idea of cognitive radio (CR)
for the design of wireless communications networks has
been developed to alleviate the shortage problem of insuf-
ficient radio spectrum by enhancing spectrum utilization
[4]–[6]. The integration of CR and IoT i.e., CR based IoT
(CR-IoT) is in the developing field of future wireless com-
munications which have the great potential to opportunis-
tically access the allocated spectrum bands (licensed) and
support new IoT services (e.g., traffic related to disaster
management, banking, response planning, security, health-
care, agriculture, and education) that can profoundly
impact our lives in a positive way. Therefore, CR-IoT
networks are becoming an attractive solution for spectrum
scarcity problem, low sum rate, and a shorter network life-
time [7]–[14]. Furthermore, energy harvesting is becoming
increasingly important to complement existing battery-
powered wireless communication networks, extending their
longevity, and keeping them more environmental-friendly.
In addition, it can prolong network lifetime by apply-
ing energy harvesting techniques to a machine learning-
based malicious user detection in energy harvesting CR-
IoT networks [15]. Despite the potential benefits of CR-
IoT networks, the major challenges faced by network
researchers and engineers of IoT networks are to better
sensing gain, enhance sum rate, and prolong network
lifetime of a typical CR-IoT users or secondary users (SUs)
(e.g., consisting of malicious users) with accommodating
increasingly growing new applications and services more

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 2
than a limited available spectrum band. For example, the
sensing gain of the CR-IoT network drops significantly
when multiple malicious users are sensing the licensed
spectrum band. Sensing gain and classification algorithm
of these networks are required for efficient detection and
deployment of such systems.
The machine learning (ML) based malicious user detec-
tion in energy harvested CR-IoT networks has recently at-
tracted significant interest in the wireless research commu-
nity [16]–[22]. The authors are concerned with separating
malicious users in energy harvested CR-IoT networks, as
specified in the ML model [23]. Specifically to be addressed
is the potential of the machine learning to make a CR-IoT
network better to identify malicious CR-IoT users. The
main contributions of this paper are
•We proposed a novel classification algorithm based
on the machine learning, i.e., logistic regression (LR),
support vector machine (SVM), and k-nearest neigh-
bors (k-NN) for separating normal CR-IoT users and
malicious CR-IoT users.
•The sensing performance is performed by the energy
detection technique, the proposed ML algorithms are
employed on the data set and then it makes group
for normal CR-IoT users and abnormal CR-IoT users
(malicious users).
•After the grouping normal CR-IoT users and mali-
cious CR-IoT users at the fusion centre (FC), the FC
needs to employ the Dempster-Shafer (DS) theory to
evaluate the sensing performance of the proposed ML
algorithms.
•The sum rate of a PU network and the CR-IoT
network, including only normal CR-IoT users, is eval-
uated on the basis of the sensing performance of the
proposed ML algorithms.
•Moreover, for the proposed ML algorithms, the pro-
longed network life of the energy harvested CR-IoT
network, including only normal CR-IoT users, is eval-
uated based on the sensing performance and sum rate.
•The results of the simulation demonstrate that the
proposed scheme is achieved an improved sensing per-
formance, a better sum rate, and a prolonged network
lifetime of the energy harvested CR-IoT networks
compared to other conventional schemes.
The remainder of this paper is structured as follows.
In Section II, the related works with contributions are
addressed. The proposed system model with explanation
is presented in Section III. Analysis of different network
metrics based on machine learning algorithms e.g., LR,
SVM and k-NN is discussed in Section IV. The results
of the simulation of the proposed scheme is validated in
Section V. In Section VI, closing notes for possible works
are point-outs.
II. Related Works
The CR is in the developing field of cognitive radio
network (CRN) [6] which have great potential to allow
CR users to access the allocated spectrum bands which
are temporally idle due to each CR user being equipped
with energy harvesting. The IoT is a modern machine-
to-machine (M2M) communication paradigm that allows
machines, computers, devices, apps, or appliances without
human interference to communicate with each other [7].
In CR-IoT networks, each CR-IoT device/user shares
information of different objects like the things-oriented,
Internet-oriented and semantic-oriented over the Internet
using different communication technologies [9]–[11], [13],
[14]. In the current era, the dramatic growth of the number
of CR-IoT users that are needed more available spectrum.
However, the scarcity of the available spectrum in a CR-
IoT network is a major challenge due to the spectrum
allocation techniques in a CR-IoT network which are
totally under utilized by CR-IoT users. Therefore, for CR-
IoT networks, efficient spectrum sensing is essential. The
cooperative spectrum sensing (CSS) approach is that CR-
IoT users share the sensing results with the fusion center
(FC) to make decisions. With this approach, hidden PU
offers a higher and more reliable sensing opportunity for
the problem compared to individual sensing performance.
The CSS approach is sensitive to attacks by malicious
users who transmit incorrect sensing results to the FC
[15]. However, the sum rate and the network lifetime
are not analyzed in energy harvested CR-IoT networks.
Robust machine learning (ML) based spectrum sensing
in cognitive radio networks (CRNs) proposed by Shah
et al. [16], where each CR user contrasts their current
sensing data to existing sensing class and measures dis-
tance vectors. However, the malicious user detection, sum
rate, and network lifetime were not analyzed. Jan et al.
[17] proposed a spectrum sensing architecture with multi-
class hypotheses in CRNs, where each CR user enhances
throughput using an support vector machine (SVM). How-
ever, the malicious user detection and network lifetime
were not analyzed. Zhu et al. [18] proposed a new Q-
learning based transmission scheduling framework using
deep learning in a CR-IoT network, where each CR-IoT
user is maximizing the system throughput using the re-
quired strategy to transfer packets in different buffers over
multiple channels. However, the malicious user detection
and network lifetime were not analyzed. Rahman et al.
[19] proposed reinforcement learning based on efficient
transmission mode selection for cooperative CRNs, where
the proposed scheme is to analyze the energy efficiency,
time delay, and PU interference. However, the malicious
user detection, sum rate, and network lifetime were not
analyzed. Thilina et al. [20] proposed a ML technique
for CSS in CRNs, where the proposed scheme is analysis
of the detection performance based on unsupervised ML
and supervised ML. However, the malicious user detection,
sum rate, and network lifetime were not analyzed. Mustafa
et al. [21] proposed a survey of ML algorithms and their
CR applications, where many ML algorithms and their
CR applications are discussed in the proposed scheme
with regards of the detection performance, classification
of modulation, and allocation of power. Hung et al. [22]
proposed a fuzzy SVM algorithm for CSS with noise

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 3
uncertainty, where the parametric tests obtained by the
secondary users are arranged into a feature vector instead
of being combined through weighted sum. However, the
malicious user detection, sum rate, and network lifetime
were not analyzed. Li et al. [23] proposed an improved
CSS model based on ML for CRNs that utilizes user
classification methods to minimize overhead cooperation
and efficiently enhance detection performance. However,
the sum rate and the network lifetime were not ana-
lyzed. Moreover, the harvested energy from the CR-IoT
users were not analyzed. In [24], the authors proposed
a ML framework for detecting the PU emulation attack
(PUEA) in CRNs, where various classification algorithms
used to differentiate between malicious users (MUs) and
legitimate users (LUs). However, the sensing performance,
sum rate, and network lifetime were not evaluated. In
[25], the authors proposed ML approach for CRNs, where
unsupervised ML, supervised ML, semi-supervised ML,
deep learning algorithms used to detect the PUEA and
improve detection performance. However, the sum rate
and the network lifetime were not analyzed. In [26], the
authors proposed SVM algorithm based malicious primary
user emulation signal detection model, which classifies the
PU and the malicious PU signal while using the signal-to-
noise ratio (SNR) and energy signal entropy. This model
is enhanced the detection performance of the existence
of MUs in low SNR even without a threshold calculation
and detection probability of the legitimate PU. However,
the sum rate and the network lifetime were not analyzed.
In summary, the current research has some drawbacks as
shown in Table I: (i) a typical CR-IoT network in which
all CR-IoT users including normal and malicious CR-IoT
users are participating to sense the PU licensed channel
which not to enhance the sensing gain; (ii) an enhanced
the sum rate and a prolonged network lifetime has not
been analyzed regarding the detection probability, the
probability of false alarm, and the energy harvested from
the PU licensed channel. The proposed scheme overcome
these drawbacks.
III. Proposed System Model
In this system model, we proposed the ML algorithms
based malicious CR-IoT users detection in energy har-
vested CR-IoT networks as shown in Fig. 1 where to
separate the normal CR-IoT users and malicious CR-
IoT users of a CR-IoT network operating in both with
fading and without fading phenomena. As in Fig. 1, we
consider CR-IoT networks with Nnormal CR-IoT users
and Mmalicious CR-IoT users. In the training phase as
shown in Fig. 1, the operating environment is discovered
by measuring the action of the CR-IoT user with the
changing of the PU activities. Each CR-IoT user sends
their signal to noise ratio (SNR) to the FC and then the FC
separates the normal and malicious CR-IoT users based on
the ML algorithms i.e., LR, SVM, and k-NN during the
classification phase where all malicious CR-IoT users are
dropout due to this malicious CR-IoT users can severely
reduce the sensing gain. In sensing phase, only each normal
Fig. 1. The proposed system model here the normal CR-IoT users
(N)and the malicious CR-IoT users (M).
CR-IoT user (e.g., ith CR-IoT user) generates a sensing
report during the longer sensing time (due to utilizing the
reporting framework) which makes a local decision, and
transmits this local decision during the fixed reporting
time to the FC which will be combined to make a global
decision about the PU activities on the licensed channel. In
the secondary CR-IoT network, there is a control channel
that is used to communicate data from CR-IoT users (e.g.,
normal and abnormal CR-IoT users) to FC. It is assumed
that the reporting signal is fading or error-free. In data
transmission phase, the source CR-IoT user (e.g. CR-IoT
Tx) uses a direct connection to transmit the data to the
destination CR-IoT user (CR-IoT Rx) during the rest
of time slot. Various factors are analyzed, including the
number of CR-IoT users (e.g., normal users and malicious
users) and their traffic load distribution, the changing
activities of the PU, and wireless channel conditions (e.g.,
without fading and with fading) that affect the detection
performance of an energy harvested CR-IoT network.
A. Primary Users Traffic Model
The PU traffic is modeled as a basic mechanism of ON
and OFF involving two instantaneous states e.g. st= 1 is
for active and st= 0 is for inactive at the time slot t[2].
B. Cooperative Energy Vectors
Each CR-IoT user executes the local spectrum sensing
independently in the PU and CR-IoT user link to identify
the activities of PU signal and transmits its local report
based on a binary hypothesis that is either H0or H1; here,
H0and H1are the absence of PU and the presence of PU,

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 4
TABLE I
Comparison of existing ML algorithms to proposed ML algorithms based CR-IoT networks
Approaches Proposed idea ML Sensing
performance
Sum rate Expected
lifetime
A. Mustafa et al. (2015)
[21]
A survey of ML algorithms and their applica-
tions in CR
SVM 3 7 7
Y. D. Huang et al. (2017)
[22]
A fuzzy SVM algorithm for CSS with noise
uncertainty
SVM X7 7
Z. Li et al. (2018) [23] An enhanced CSS model based on ML for
CRNs
SVM 3 7 7
A. Albehadili et al.
(2019) [24]
ML based PU emulation attack detection in
CRNs using pattern described link-signature
(PDLS)
ML 3 7 7
A. Albehadili et al.
(2019) [25]
Semi-supervised ML for PU emulation attack
detection and prevention through core based
analytic for CRNs
ML 3 7 7
E. C. Munoz et al. (2020)
[26]
Detection of Malicious PU emulation Based
on a SVM for a mobile CRN using software
defined radio
SVM 3 7 7
M. S. Khan et al. (2020)
[27]
SVM based classification of MUs in CRNs SVM 3 7 7
M. S. Khan et al. (2020)
[28]
A genetic algorithm based soft decision fusion
scheme in cognitive IoT networks with MUs
GA 3 7 7
Y. Zhang et al. (2020)
[29]
Ensemble learning based robust CSS in full-
duplex CRNs
SVM,
LR
3 7 7
M. S. Miah et al. (pro-
posed)
ML based MU detection in energy harvested
CR-IoT
SVM, k-
NN, LR
3 3 3
respectively. The spectrum sensing problem of the CR-
IoT user can be computed based on a binary hypothesis
as follows:
(H0:if PU is for inactive at time slot (st= 0);
H1:if PU is for active at time slot (st= 1); (1)
The signals received of the ith CR-IoT user can be com-
puted according to the packet transmission of the PU [30]
as follows:
zi(t) = (yi(t) ; H0
hi(t)x(t) + yi(t) ; H1
(2)
where t= 1,2,3, ..., L, here Ldenotes the number of
samples of the signals received that defines as L= 2τsfs;τs
denotes the flexible sensing time slot in sec, and fsdenotes
the sampling frequency. We can consider that the flexible
sensing time,τsis proportional to the energy harvested, es
with a constant, ζi.e., τs=ζes. Moreover, zi(t)denotes
the signal received by the ith CR-IoT user, x(t)denotes
the PU transmitted signal, i.e., x(t)∼ ℵ0, σ2
x, and yi(t)
denotes the additive white Gaussian noise of the ith CR-
IoT user, i.e., yi(t)∼ ℵ(0, σ2
y,i). Also, hi(t)denotes the
channel gain of the ith CR-IoT user, H1denotes the active
of PU signal, and H0denotes the inactive of PU signal.
For the CSS of the CR-IoT users, the Energy Detec-
tor (ED) technique is commonly used because it can be
applied effectively without any previous PU signal infor-
mation being acquired. The sensing result zi(t)obtained
by the ith CR-IoT user transmitter is the signal power
in the time domain at a given frequency; a band-pass
filter is added to the received signal, then an analog-to-
digital converter (ADC) converts the output of this filter,
which is independently averaged and squared using the
conventional ED technique to determine its own calculated
energy, Eias follows [2]:
Ei=1
L
L
X
t=1 kzi(t)k2(3)
Based on the central limit theorem (CLT), when Lis
relatively high, the signal received at the ith CR-IoT user,
Eiin (3) can be estimated as a Gaussian random variable
under a binary hypotheses i.e., H0and H1with mean and
variance, respectively as follows [30]:
Ei∼(ℵµi(H0), σ2
i(H0)
ℵµi(H1), σ2
i(H1)(4)
where
µi(H0)=2τsfsσ2
y,i,
σ2
i(H0)=4τsfsσ4
y,i,
µi(H1)=2τsfs1 + |hi|2γiσ2
y,i,
σ2
i(H1)=4τsfs1+2|hi|2γiσ4
y,i,
where γidenotes a signal to noise ratio (SNR) which
defines as γi=p2
x
σ2
y,i
; here p2
xis the signal power of x(t).
IV. Analysis of different network metrics
based on ML algorithms
ML is a sub-branch of computer science that was de-
veloped in 1959 from computational learning experiments
and model recognition in artificial intelligence. ML is a
framework that can learn as a structural function and
investigate the work and construction of algorithms that
can make predictions about data. ML algorithms operate
by constructing a model from sample inputs to make

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 5
Training Module
Classification Module Channel
Availability
Training
Energy Vector
Test
Energy Vector
Fig. 2. Modular framework of the proposed CSS model [23].
data-based predictions and decisions, rather than simply
following classic instruction set. There are three types
of learning strategies in ML. ML algorithms are often
used in model classification. The relevant feature vector
is extracted from a model and given to the classifier so
that it can be categorized. In this context, we considered
an estimated energy level (energy vector) of the CR device
as a feature vector. Next, the classifier transfers the energy
vector to one of two classes: “CR-IoT/SU user” and “MU”.
First, the classifier must go through a training phase where
it learns from the training of feature vectors i.e., energy
vector as shown in Fig. 2.
In supervised ML, data is collected from interactive
systems and arranged in a certain order. Unsupervised ML
is another ML method with no data set and no outputs.
It is to investigate common points by interpreting the
data in the data set and to obtain meaningful data by
clustering them. Semi-supervised ML is both learning that
is supervised ML and unsupervised ML. It consists of
using too much untagged data and small-sized data tagged
together. In this paper, we use supervised ML algorithms
such as LR, SVM, and k-NN for classification of a CR-
IoT/SU users is either normal CR-IoT users or malicious
users. The dataset for ML algorithms are constructed from
sensing energies of CR-IoT/SU users and MUs. Owing
to energy detection based spectrum sensing the related
energy levels of users can be detected. The range of sensing
energies at the valid CR-IoT/SU users is between 90 and
108 and the values outside this range can belong to invalid
CR-IoT users as MUs. Benefits of the proposed solution
as follows:
•Thanks to the optimized classifier trained with en-
ergy vectors, a structure capable of more comfortable
adaptation to dynamic radio environments has been
created. Since the training process does not require
any prior knowledge and parameter setting about the
environment, an environment-independent model was
created.
•The proposed MU detection technique also performs
better in terms of false alarm and detection probabil-
ities, as it can make optimized decisions compared to
traditional approaches.
A. Logistic Regression (LR) Algorithm
The LR algorithm in ML framework is a widely used
efficient meta learner that automatically learns the opti-
mal weights of the results of basic learner prediction [29].
LR is a regression method which is used for classification
problems based on the concept of probability. It is used for
the classification categorical or numerical data. It operates
only if the dependent variable, i.e. the result can be
taken two different values. In LR algorithm, the logistic
regression function is defined as follows:
y=logit(p) = ln p
p−1=wTx+b(5)
where logistic (logit)is the ratio of class probabilities, x
is the data vector, y∈ {−1,+1}for two classes, wand b
are the weight parameters.
After training phase with the aim of finding parameters
for class probabilities i.e., Class1and Class2are defined
as follows:
p(y=−1|x)=1−p(y= +1|x)(6)
and
p(y= +1|x)=1−p(y=−1|x)(7)
Similarly, for class probability of Cis defined as follows:
p(y=C|x) = 1
1 + e(−y(wTx+b)) (8)
The benefits of using LR include its flexibility, reliability
and the ability to resist over-fitting without any hyper-
parameter tuning in small-scaled data sets. The LR based
MUs classification ML algorithm in an energy harvested
CR-IoT network is shown in Algorithm 1 which consists
of four stages: data generation, sensing, classification, and
area under curve (AUC).
B. Support Vector Machine (SVM) Algorithm
The SVM can be defined as a ML approach based on
vector space that finds a boundary decision between two
classes that are furthest from any point in the training
dataset. It is a supervised ML algorithms which can be
used for classification. In classification problem, it creates
a set of hyper-planes in a high dimensional feature space
by analyzing data for binary classification [31]. In the
CR-IoT network, we consider CR-IoT networks with N
normal CR-IoT users, Mmalicious CR-IoT users and total
number of user is represent by K=N+M.
The notation of training data set (X)of the proposed
SVM algorithm is defined as follows [27]:
X=(xi, yi)|xi∈RZ, yi∈[+1,−1]Z
i=1 (9)
Here, (xi, yi)represents the data set for normal and mali-
cious users which is defined as follows
(xi, yi) = ((x1, y1),(x2, y2), ..., (xZ, yZ)) (10)
where xirepresents the energy vector/training data of
Z(i= 1,2, ..., Z)users, yi∈[+1,−1] is the class vec-
tor/target output, and class “+1” and “-1” represent
normal CR-IoT users and malicious CR-IoT users, respec-
tively.

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 6
Algorithm 1 LR based MUs classification ML algorithm
in an energy harvested CR-IoT network.
Input: θ,Land 
Output: Normal CR-IoT users and Malicious Users
Initialisation :
1: Sensing the data
2: for i= 1 to Ndo
3: for t=lto Ldo
4: Energy reported by the ith CR-IoT users based
on Eq. 3
5: end for
6: end for
7: for i= 1 to Mdo
8: for t=lto Ldo
9: Energy reported by the ith malicious users based
on Eq. 3
10: end for
11: end for
Data processing for LR:
12: Combining the both users data in to one data set
13: Find the number of features from data set
14: Extract the features matrix Xand label vector Y
15: Scaling the features data set
16: Divided the data set into training and testing group
Classification:
17: for i= 1 to Zdo
18: Mapping functions (inputs)
19: Update Augmented Weight Matrix(θ)
20: Cost function or Average Cost J(θ)
21: if J(θ)≤∈k (N== Nmax)then
22: Optimum weights(θ)
23: else
24: Update Augmented Weight Matrix(θ)
25: end if
26: end for
27: Find Normal CR-IoT users and Malicious Users
28: Plot the normal CR-IoT user data
29: Plot the malicious CR-IoT user data
30: Calculate the confusion matrix
31: Plot the confusion matrix value
32: Plot the AUC for LR
In SVM, the following optimization problem is defined
which is maximising the classifier margin as follows:
min 1
2kwk2
s.t. yi(w.xi+b)≥1
(11)
where k.krepresents the norm which is defined as kwk2=
w.w, and bdenotes the bias that shifting the hyperplane
away from it’s origin.
Now, the decision function of the hyperplane can then
be written as follows:
g(x) = δ X
i=1:αi>0
yiαi(x.xi) + b!(12)
Algorithm 2 SVM based MUs classification ML algo-
rithm in an energy harvested CR-IoT network.
Input: XN,Mand L
Output: Normal CR-IoT users set and Malicious Users
set
Initialisation :
1: Sensing the data
2: for i= 1 to Ndo
3: for t=lto Ldo
4: Energy reported by the ith CR-IoT users based
on Eq. 3
5: end for
6: end for
7: for i= 1 to Mdo
8: for t=lto Ldo
9: Energy reported by the ith malicious users based
on Eq. 3
10: end for
11: end for
Data processing for SVM:
12: Combining the both users data in to one data set
13: Find the number of features from data set
14: Extract the features matrix Xand label vector Y
15: Scaling the features data set
16: Divided the data set into training and testing group
Classification:
17: Selected the linear support vector classifier (SVC)
18: Fit the data in the linear SVC
19: Find Normal CR-IoT users and Malicious Users
20: Plot the normal CR-IoT user data
21: Plot the malicious CR-IoT user data
22: Calculate the confusion matrix
23: Plot the confusion matrix value
24: Plot the AUC for SVM
where, g(x)is the decision function of the hyperplane, P
is the sum of support vector/data, δis the sign function,
x.xiis the kernel function and the sum is over support
vectors. The classification process compares the new in-
stance xwith each of the support vectors. Moreover, the
(xi.x)measures how similar the new instance xis to the
training instance xi. Here, αiare called support vectors
that measure the contribution of xi, i.e., αimeasures
how important the given support vector and the decision
function are defined by these vectors. We multiply by yito
take into account the influence of the given support vector
on the classification.
The SVM based MUs classification ML algorithm in an
energy harvested CR-IoT network is shown in Algorithm
2 which consists of four stages: data generation, sensing,
classification, and AUC.
C. k-Nearest Neighbors (k-NN) Algorithm
The k-NN algorithm is one of the easy-to-implement
supervised ML algorithms. Although it is used in the
solution of both regression and classification tasks, it used
mainly in the solution of classification tasks in industry.

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 7
Algorithm 3 k-NN based MUs classification ML algo-
rithm in an energy harvested CR-IoT network.
Input: D N,Mand L
Output: Normal CR-IoT users set and Malicious Users
set
Initialisation :
1: Sensing the data
2: for i= 1 to Ndo
3: for t=lto Ldo
4: Energy reported by the ith CR-IoT users based
on Eq. 3
5: end for
6: end for
7: for i= 1 to Mdo
8: for t=lto Ldo
9: Energy reported by the ith malicious users based
on Eq. 3
10: end for
11: end for
Data processing for k-NN:
12: Combining the both users data in to one data set
13: Find the number of features from data set
14: Extract the features matrix Xand label vector Y
15: Scaling the features data set
16: Divided the data set into training and testing group
Classification:
17: Choose the value of K
18: for i= 1 to Zdo
19: Compute standard Euclidean distance, d(x, y)be-
tween ith data point and each row of training data
point by using the Eq. 13
20: Sort the computed distances
21: Find closest k-nearest neighbors point
22: Vote for labels
23: end for
24: Find normal CR-IoT users and abnormal CR-IoT
users (MUs)
25: Plot the normal CR-IoT user data
26: Plot the malicious CR-IoT user data
27: Calculate the confusion matrix
28: Plot the confusion matrix value
29: Plot the AUC for k-NN
In k-NN, the Euclidean distance is calculated based on the
following equations:
d(x, y) = v
u
u
t
n
X
i=1
ci(xi−yi)2(13)
where, ciis the weight, and drepresents distance between
xiand yi.
The k-NN based MUs classification ML algorithm in an
energy harvested CR-IoT network is shown in Algorithm
3 which consists of four stages: data generation, sensing,
classification, and AUC.
D. Accuracy of the ML Algorithms
In this section, we can calculate the accuracy of ML
algorithms as follows:
η=(T N +T P )
(T P +F P +T N +F N )(14)
where, ηis accuracy of the ML algorithms, and T P ,F P ,
T N and F N indicates True Positives,False Positives,
True Positives and False Negatives, respectively.
E. Global Decision
When the classification is done through the ML algo-
rithms, all normal CR-IoT users execute local decision
independently based on the sequential manner, and then
transmit their decisions to the FC during the reporting
time slot. After that, the FC makes a global decision about
the PU activities like the absence and presence of the PU
signal using the DS evidence theory under the sequential
manner [2] as follows:
m(H0) = ω1m1(H0)⊕ω2m2(H0)⊕···⊕ωNmN(H0)
=PΓ1∩Γ2∩···∩ΓN=H0ΠN
i=1ωimi(Γi)
1−PΓ1∩Γ2∩···∩ΓN=∅ΠN
i=1ωimi(Γi)
(15)
m(H1) = ω1m1(H1)⊕ω2m2(H1)⊕···⊕ωNmN(H1)
=PΓ1∩Γ2∩···∩ΓN=H1ΠN
i=1ωimi(Γi)
1−PΓ1∩Γ2∩···∩ΓN=∅ΠN
i=1ωimi(Γi)
(16)
where
mi(H0) = Z∞
Ei
1
√2πσi(H0)exp −(Xi−µi(H0))2
2σ2
i(H0)!
mi(H1) = ZEi
∞
1
√2πσi(H1)exp −(Xi−µi(H1))2
2σ2
i(H1)!
here, mi(H0),mi(H1)and mi(Ω) are the basic probability
assignment (BPA) hypotheses of ith normal CR-IoT users
under H0,H1, and Ω, respectively. Moreover, Γiis an
element of the set {H0,H1,Ω}and ωiis the weight factor
of the ith CR-IoT user.
At the FC, the final combination result m(H0)and
m(H1)by each normal CR-IoT user is obtained, then it
makes a global decision gdf|das follows:
gdf|d=(0; if m(H0)> m (H1)
1; if m(H1)≥m(H0)(17)
F. Sum Rate Analysis
After calculating the detection performance at the FC
based on the sequential manner in the previous sub-
section, now the sum rate is evaluated when considering
numerous premises. During the transmission phase, the
CR-IoT Txtransmits its own relevant information towards
the respective CR-IoT Rxbased on round robin scheduling
approach [30]. In the case of the non-false alarm, if the

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 8
PU is absent, each unlicensed CR-IoT user is correctly
sensed the absence of the PU; then each unlicensed CR-
IoT user is likely to allow the licensed spectrum of the
PU for a certain amount of time , as described by the
probability of (1 −gdf). In another side, in the case of
detection, the CR-IoT users should not interfere with the
PU transmission. Consequently, the sum rate based on
the round robin scheduling approach of the proposed ML
algorithms is defined as follows:
Rsum =ρgddCP U + (1 −ρ)gdfCi,C R−I oT (18)
where CP U is the channel capacity of the PU link,
Ci,CR−I oT is the channel capacity of the ith CR-IoT link,
and ρ∈[0,1] indicates the primary activity factor which
means the probability of the PUs transmitting in a given
frame.
The CP U and Ci,CR−I oT are given as follows:
CP U =log2(1 + S N RP U )(19)
Ci,CR−I oT =T−τs−τr
Tlog2(1 + SNRi,CR−IoT )(20)
SN RP U is the SNR of the PUs link, SN Ri,C R−I oT is the
SNR of the ith CR-IoT link, and Tis the total frame
length.
G. Network Lifetime Analysis
Now, we can measured the average energy consumption
of the proposed ML algorithms in this section as follows:
Eavg =esτs+etTT(ρgdd+ (1 −gdf) (1 −ρ)) (21)
where esdenotes the energy consumed for the sensing
duration, TTdenotes the transmission time, i.e., TT=
T−τs−τr, and τs,etdenotes the energy consumed for
the transmission duration,
The expected network lifetime (ξ)of the proposed ML
algorithm can be calculated as follows:
ξ=ec+eh
s
Eavg
(22)
where ecis the capacity of battery and eh
sis the energy
harvested during the sensing phase (the flexible sensing
time duration).
V. Simulation Results and Discussions
In this section, we validate the theoretical findings,
and analyze the detection performance of the proposed
the ML algorithms. This is achieved through numerical
simulations via Matlab. Monte-Carlo simulations were
carried out using the simulation parameters listed in Table
II below which are based on the rationale of the other
researchers [26], [27], [31]. The performance of the pro-
posed scheme based on the ML algorithm is compared with
similar schemes, such as SVM based classification of MUs
in CRNs [27], the detection of malicious PU emulation
based on a SVM for a mobile CRN using software-defined
60 70 80 90 100 110 120 130
Energy Level
60
70
80
90
100
110
120
130
Energy Level
Original data set values
SU
MU
Fig. 3. Data set values.
radio [26], and CSS algorithm based on SVM against
spectrum sensing-data-falsification (SSDF) attack [31].
TABLE II
Simulation parameters with values
Parameters Value
The number of the CR-IoT users 43
The number of malicious users 28
The sampling frequency, fs300kHz
The sensing time slot, τs1 ms
The reporting time slot, τr1 ms
The time slot length, T10 ms
The energy consumption in sensing phase, es1 J
The energy consumption in transmission phase, et3 J
The capacity of battery, ec301 J
The probability of the absence of the PU, ρ0.5
The probability of the presence of the PU, (1 −ρ)0.5
The data set consists of MUs and CR-IoT users/SUs
classes categorized according to energy level. This dataset,
consisting of 71 data, was first optimized and then training
tests were passed. Fig. 3 shows the MU and CR-IoT user
distribution chart in our data set.
After the data set was trained with classifiers, the
accuracy scores of the obtained models were compared.
The performances were checked using class predictions,
confusion matrix, and Receiver Operating Characteristic
(ROC) curve. The validation accuracy score predicts the
performance of a model on new data compared to training
data. The best model was chosen based on this score. k-
fold cross-validation was used to calculate the accuracy
points using the observations in kvalidation folds and to
determine the mean cross-correct error and was considered
as 5. It calculates the confusion matrix and ROC curve
based on these estimates by making predictions on k-fold
cross-validation observations.
In Fig. 4, the LR prediction values are shown. SU and
MU correct predictions are given with circles and the miss
predictions are given with x shape. Also in Fig. 5, the con-
fusion matrix are shown. The confusion matrix is the most
important metric commonly used to evaluate classification
patterns. Some metrics can be derived from the confusion

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 9
60 70 80 90 100 110 120 130
Energy Level
60
70
80
90
100
110
120
130
Energy Level
Predictions: (Logistic Regression)
MU - Incorrect
MU - Correct
SU - Correct
Model predictions
Fig. 4. The prediction value of the LR algorithm.
MU SU
Predicted Class
MU
SU
True Class
Confusion Matrix (Logistic Regression)
2.3%
100.0%
97.7% 2.3%
100.0%
97.7%
TPR FNR
Fig. 5. The confusion matrix of the LR algorithm.
matrix like accuracy. Accuracy is the proposed model’s
overall predicted accuracy and it can be formalized using
the Eq. 14.
As can be seen from confusion matrix, True Positive
Rate (TPR) indicates positive values that have been cor-
rectly predicted and False Positive Rate (FPR) shows
negative values that have been incorrectly predicted. True
Negative Rate (TNR) gives negative values that have been
correctly predicted and False Negative Rate (FNR) indi-
cates positive values that have been incorrectly predicted.
The results of LR, SVM, KNN classifiers are given in
the scatter plot given in Fig. 4, Fig. 7 and Fig. 10. After
the classifiers are trained, the estimates of the models are
shown with the help of these distribution charts. When
these three figures are compared, the best model is LR
with its predictions.
The confusion matrix plot was used to understand how
the classifiers performed in each class and in which areas
they performed poorly. In these graphs, the rows show the
actual class and the columns the estimated class. In Fig.
5, Fig. 8 and Fig. 11 , confusion matrices obtained for LR,
SVM and k-NN are given.
In Fig. 6, the ROC of the LR is given in detail. The
ROC determines the correctness of a classification model
at a user-defined threshold. As can be seen from Fig. 6,
AUC gives the model’s accuracy by ROC. The aim of the
algorithm is to push the line towards one and maximize
0 0.2 0.4 0.6 0.8 1
False positive rate
0
0.2
0.4
0.6
0.8
1
True positive rate
ROC Curve (Logistic Regression)
AUC = 1.00
(0.00,0.98)
Positive class: SU
ROC curve
Area under curve (AUC)
Current classifier
Fig. 6. ROC of the LR algorithm.
60 70 80 90 100 110 120 130
Energy Level
60
70
80
90
100
110
120
130
Energy Level
Predictions: (Quadratic SVM)
MU - Incorrect
MU - Correct
SU - Incorrect
SU - Correct
Model predictions
Fig. 7. The prediction value of the SVM algorithm.
the under the curve.
To see how classifiers perform per class, TPR and FNR
are given. TPR expresses the proportion of correctly classi-
fied observations per actual class, while FNR indicates the
proportion of misclassified observations per actual class.
The SVM prediction values are shown in Fig. 7. The
means of the points are discussed for LR, and it can be
seen from Fig. 7, the wrong predictions are increased in
SVM. The confusion matrix is also given in Fig. 8, values of
some areas are decreased in SVM confusion matrix. When
these results are compared, LR obtained 100% to 97.7%,
SVM 96.4% to 97.7%, and k-NN 100% to 81.4% in terms
of determining MU and CR-IoT user, respectively.
In the light of these results, it has been observed that
the LR model is more successful. In Fig. 6, Fig. 9 and
Fig. 12, the ROC curves are given for LR, SVM and k-NN
which show the correct and false positive rates after the
models are trained. Here, true positive and false positive
rates for the trained classifier are shown.
As Fig. 11 the k-NN confusion matrix shows the afore-

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 10
MU SU
Predicted Class
MU
SU
True Class
Confusion Matrix (Quadratic SVM)
2.3%
3.6%96.4%
97.7%
3.6%
2.3%
96.4%
97.7%
TPR FNR
Fig. 8. The confusion matrix of of the SVM algorithm.
0 0.2 0.4 0.6 0.8 1
False positive rate
0
0.2
0.4
0.6
0.8
1
True positive rate
ROC Curve (Quadratic SVM)
AUC = 1.00
(0.04,0.98)
Positive class: SU
ROC curve
Area under curve (AUC)
Current classifier
Fig. 9. ROC curve of the SVM algorithm.
60 70 80 90 100 110 120 130
Energy Level
60
70
80
90
100
110
120
130
Energy Level
Predictions: (Cosine KNN)
MU - Incorrect
MU - Correct
SU - Correct
Model predictions
Fig. 10. The prediction value of the k-NN algorithm.
mentioned areas for accuracy calculation.
ROC curve of the k-NN algorithm is given in Fig. 12,
and the AUC is reduced as 0.89 with positive class of SU.
The k-NN is the most unsuccessful classification algorithm
for the proposed system.
MU SU
Predicted Class
MU
SU
True Class
Confusion Matrix (Cosine KNN)
18.6%
100.0%
81.4% 18.6%
100.0%
81.4%
TPR FNR
Fig. 11. The confusion matirx of the k-NN algorithm.
0 0.2 0.4 0.6 0.8 1
False positive rate
0
0.2
0.4
0.6
0.8
1
True positive rate
ROC Curve (Cosine KNN)
AUC = 0.89
(0.00,0.81)
Positive class: SU
ROC curve
Area under curve (AUC)
Current classifier
Fig. 12. ROC curve of the k-NN algorithm.
Fig. 13 demonstrates the sum rate of the ML algorithms
(LR, SVM, k-NN). The sum rate is a function of the
probability of false alarm. For example, the sum rate of
the proposed LR algorithm is 3.45Hz compared to other
SVM algorithm and the k-NN algorithm are 3.38Hz and
3.25Hz, respectively when the probability of false alarm is
0.3. Therefore, the proposed LR algorithm is an enhanced
sum rate when compared to both the SVM algorithm and
the k-NN algorithm.
The energy consumption for the ML algorithms (LR,
SVM, k-NN) shows in Fig. 14. The average energy con-
sumption is a function of the probability of false alarm.
The energy consumption of the proposed ML algorithms
with the LR algorithm achieved a better energy efficient
when compared to other ML algorithms because of its
higher detection performance. For example, the energy
consumption of the LR algorithm is 1.29Jcompared to
other SVM algorithm and the k-NN algorithm are 1.33J
and 1.49J, respectively when the probability of false alarm
is 0.2. Therefore, the proposed LR algorithm is a better
than other ML algorithms with SVM and k-NN for any
value of the probability of false alarm.
The expected lifetime of the proposed ML algorithms
with the LR, SVM, and k-NN shows in Fig. 15. It is

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 11
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
gdf: Probability of false alarm at CR-IoT user
0.5
1
1.5
2
2.5
3
3.5
Rsum: Sum rate (bps/Hz) at CR-IoT user
sum rate of the ML algorithm (k-NN)
sum rate of the ML algorithm (SVM)
sum rate of the ML algorithm (LR)
Fig. 13. The sum rate of the ML algorithms (k-NN, SVM, LR).
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
gdf: Probability of false alarm at the CR-IoT user
0
0.5
1
1.5
2
2.5
3
Eavg: Energy consumption at the CR-IoT user in J
Energy consumption of the ML algorithm (k-NN)
Energy consumption of the ML algorithm (SVM)
Energy consumption of the ML algorithm (LR)
Fig. 14. The energy consumption of the ML algorithms (k-NN, SVM,
LR).
clearly seen that the expected lifetime of the proposed
ML algorithms with the LR algorithm has extended when
compared to both SVM and k-NN algorithms. For exam-
ple, the sum rate of the LR algorithm is 3.48Hz com-
pared to other SVM algorithm and the k-NN algorithm
are 3.37Hz and 3.28Hz, respectively; when the expected
life time at the CR-IoT user in round is 400. Moreover,
with regard to the expected lifetime, the sum rate of the
proposed ML algorithms are decreased when the expected
lifetime in round is greater than 400. Moreover, the sum
rate of the proposed ML algorithm is a same when the
expected lifetime in round is greater than 850. Therefore,
the proposed LR algorithm is a prolonged the expected
lifetime when compared to both the SVM algorithm and
the k-NN algorithm.
Accuracy scores for LR, SVM, and k-NN are given in
Table III. In the light of these values, the LR model was
preferred for the architecture we recommend.
0 500 1000 1500 2000 2500 3000
:Expected lifetime at the CR-IoT user in round
0.5
1
1.5
2
2.5
3
3.5
Rsum:Sum rate (bps/Hz) at the CR-IoT user
Expected lifetime of the ML algorithm (k-NN)
Expected lifetime of the ML algorithm (SVM)
Expected lifetime of the ML algorithm (LR)
Fig. 15. The expected lifetime of the ML algorithms (k-NN, SVM,
LR).
TABLE III
Accuracy of the machine learning algorithms
Machine learning algorithms Accuracy
LR 98.60%
SVM 97.20%
k-NN 60.60%
VI. Conclusion
Malicious user detection is very important issue for
efficient use of spectrum in CR-IoT networks. With using
improved malicious user detection schemes, more robust
spectrum utilization will be provided. For detecting mali-
cious users in CR-IoT networks, classification algorithms
can be used. For this reason, we utilized machine learning
based classification algorithms such as LR, SVM, and k-
NN. And also, the proposed schemes have been applied
in CR-IoT networks, and enhanced results have been
obtained in terms of sensing gain, sum rate, and network
lifetime compared to the conventional schemes. For future
works, various datasets and classification algorithms can
be considered for different CR-IoT networks scenarios.
Acknowledgment
This research was supported in part by the Islamic
University (IU), Kushtia-7003, Bangladesh (Ref. No.
141/EDU/IU-2020/634 and by the Department of Infor-
mation and Communication Technology, Islamic Univer-
sity, Kushtia, Bangladesh.
References
[1] S. Chen, H. Xu, D. Liu, B. Hu, and H. Wang, “A vision
of iot: Applications, challenges, and opportunities with china
perspective,” IEEE Internet of Things journal, vol. 1, no. 4, pp.
349–359, 2014.
[2] M. S. Miah, “An energy efficient spectrum sensing scheme for
the cognitive radio based internet of things,” Ph.D. dissertation,
NUI Galway, 2020.

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 12
[3] H. Afzal, M. Rafiq Mufti, A. Raza, and A. Hassan, “Performance
analysis of qos in iot based cognitive radio ad hoc network,”
Concurrency and Computation: Practice and Experience, p.
e5853.
[4] M. A. Hossain, M. Schukat, and E. Barrett, “Enhancing the
spectrum utilization in cellular mobile networks by using cogni-
tive radio technology,” in 2019 30th Irish Signals and Systems
Conference (ISSC). IEEE, 2019, pp. 1–6.
[5] M. R. Amin, M. M. Rahman, M. A. Hossain, M. K. Islam, K. M.
Ahmed, B. C. Singh, and M. S. Miah, “Unscented kalman filter
based on spectrum sensing in a cognitive radio network using
an adaptive fuzzy system,” Big Data and Cognitive Computing,
vol. 2, no. 4, p. 39, 2018.
[6] J. Mitola, “Cognitive radio for flexible mobile multimedia
communications,” in 1999 IEEE International Workshop on
Mobile Multimedia Communications (MoMuC’99)(Cat. No.
99EX384). IEEE, 1999, pp. 3–10.
[7] A. A. Khan, M. H. Rehmani, and A. Rachedi, “Cognitive-radio-
based internet of things: Applications, architectures, spectrum
related functionalities, and future research directions,” IEEE
wireless communications, vol. 24, no. 3, pp. 17–25, 2017.
[8] M. A. Hossain, M. Schukat, and E. Barrett, “Enhancing the
spectrum sensing performance of cluster-based cooperative cog-
nitive radio networks via sequential multiple reporting chan-
nels,” Wireless Personal Communications, pp. 1–23, 2020.
[9] A. A. Khan, M. H. Rehmani, and A. Rachedi, “When cog-
nitive radio meets the internet of things?” in 2016 interna-
tional wireless communications and mobile computing confer-
ence (IWCMC). IEEE, 2016, pp. 469–474.
[10] A. Ali, L. Feng, A. K. Bashir, S. H. A. El-Sappagh, S. H.
Ahmed, M. Iqbal, and G. Raja, “Quality of service provisioning
for heterogeneous services in cognitive radio-enabled internet of
things,” IEEE Transactions on Network Science and Engineer-
ing, 2018.
[11] D. T. Otermat, I. Kostanic, and C. E. Otero, “Analysis of the
fm radio spectrum for secondary licensing of low-power short-
range cognitive internet of things devices,” IEEE Access, vol. 4,
pp. 6681–6691, 2016.
[12] Y. B. Zikria, F. Ishmanov, M. K. Afzal, S. W. Kim, S. Y. Nam,
and H. Yu, “Opportunistic channel selection mac protocol for
cognitive radio ad hoc sensor networks in the internet of things,”
Sustainable Computing: Informatics and Systems, vol. 18, pp.
112–120, 2018.
[13] Y. Gu, H. Chen, C. Zhai, Y. Li, and B. Vucetic, “Minimizing age
of information in cognitive radio-based iot systems: Underlay or
overlay?” IEEE Internet of Things Journal, vol. 6, no. 6, pp.
10 273–10 288, 2019.
[14] H. A. B. Salameh, S. Al-Masri, E. Benkhelifa, and J. Lloret,
“Spectrum assignment in hardware-constrained cognitive radio
iot networks under varying channel-quality conditions,” IEEE
Access, vol. 7, pp. 42 816–42 825, 2019.
[15] A. Sultan, “Sensing and transmit energy optimization for an en-
ergy harvesting cognitive radio,” IEEE wireless communications
letters, vol. 1, no. 5, pp. 500–503, 2012.
[16] H. A. Shah and I. Koo, “Reliable machine learning based spec-
trum sensing in cognitive radio networks,” Wireless Communi-
cations and Mobile Computing, vol. 2018, 2018.
[17] S. U. Jan, V.-H. Vu, and I. Koo, “Throughput maximization
using an svm for multi-class hypothesis-based spectrum sensing
in cognitive radio,” Applied Sciences, vol. 8, no. 3, p. 421, 2018.
[18] J. Zhu, Y. Song, D. Jiang, and H. Song, “A new deep-q-learning-
based transmission scheduling mechanism for the cognitive in-
ternet of things,” IEEE Internet of Things Journal, vol. 5, no. 4,
pp. 2375–2385, 2017.
[19] M. A. Rahman, Y.-D. Lee, and I. Koo, “An efficient transmission
mode selection based on reinforcement learning for cooperative
cognitive radio networks,” Human-centric Computing and In-
formation Sciences, vol. 6, no. 1, p. 2, 2016.
[20] K. M. Thilina, K. W. Choi, N. Saquib, and E. Hossain, “Ma-
chine learning techniques for cooperative spectrum sensing in
cognitive radio networks,” IEEE Journal on selected areas in
communications, vol. 31, no. 11, pp. 2209–2221, 2013.
[21] M. Alshawaqfeh, X. Wang, A. R. Ekti, M. Z. Shakir, K. Qaraqe,
and E. Serpedin, “A survey of machine learning algorithms and
their applications in cognitive radio,” in International Confer-
ence on Cognitive Radio Oriented Wireless Networks. Springer,
2015, pp. 790–801.
[22] Y.-D. Huang, Y.-C. Liang, and G. Yang, “A fuzzy support
vector machine algorithm for cooperative spectrum sensing with
noise uncertainty,” in 2016 IEEE Global Communications Con-
ference (GLOBECOM). IEEE, 2016, pp. 1–6.
[23] Z. Li, W. Wu, X. Liu, and P. Qi, “Improved cooperative spec-
trum sensing model based on machine learning for cognitive
radio networks,” IET Communications, vol. 12, no. 19, pp.
2485–2492, 2018.
[24] A. Albehadili, A. Ali, F. Jahan, A. Y. Javaid, J. Oluochy, and
V. Devabhaktuniz, “Machine learning-based primary user emu-
lation attack detection in cognitive radio networks using pattern
described link-signature (pdls),” in 2019 Wireless Telecommu-
nications Symposium (WTS). IEEE, 2019, pp. 1–7.
[25] S. Srinivasan, K. Shivakumar, and M. Mohammad, “Semi-
supervised machine learning for primary user emulation at-
tack detection and prevention through core-based analytics for
cognitive radio networks,” International Journal of Distributed
Sensor Networks, vol. 15, no. 9, p. 1550147719860365, 2019.
[26] E. Cadena Muñoz, L. F. Pedraza Martínez, and J. E. Ortiz Triv-
iño, “Detection of malicious primary user emulation based on a
support vector machine for a mobile cognitive radio network
using software-defined radio,” Electronics, vol. 9, no. 8, p. 1282,
2020.
[27] M. S. Khan, L. Khan, N. Gul, M. Amir, J. Kim, and S. M.
Kim, “Support vector machine-based classification of malicious
users in cognitive radio networks,” Wireless Communications
and Mobile Computing, vol. 2020, 2020.
[28] M. S. Khan, N. Gul, J. Kim, I. M. Qureshi, and S. M. Kim, “A
genetic algorithm-based soft decision fusion scheme in cognitive
iot networks with malicious users,” Wireless Communications
and Mobile Computing, vol. 2020, 2020.
[29] Y. Zhang, Q. Wu, and M. Shikh-Bahaei, “Ensemble learning
based robust cooperative sensing in full-duplex cognitive radio
networks,” in 2020 IEEE International Conference on Commu-
nications Workshops (ICC Workshops). IEEE, 2020, pp. 1–6.
[30] M. S. Miah, M. Schukat, and E. Barrett, “Sensing and through-
put analysis of a mu-mimo based cognitive radio scheme for the
internet of things,” Computer Communications, 2020.
[31] H. Zhu, T. Song, J. Wu, X. Li, and J. Hu, “Cooperative
spectrum sensing algorithm based on support vector machine
against ssdf attack,” in 2018 IEEE International Conference on
Communications Workshops (ICC Workshops). IEEE, 2018,
pp. 1–6.
Dr. Md. Sipon Miah received his B.Sc.,
M.Sc. and Ph.D. in Information and Commu-
nication Technology (ICT) from the Islamic
University (IU), Kushtia-7003, Bangladesh, in
2006, 2007 and 2016, respectively. Dr. Sipon
received a Structured Ph.D. in the School of
Computer Science from the National Univer-
sity of Ireland Galway (NUIG), Galway, Ire-
land, in 2020. In 2013, Dr. Sipon was awarded
the prestigious ICT Scholarship (Bangladesh).
In 2016, Dr. Sipon was awarded the prestigious
Hardiman Scholarship (Ireland). Since 2010, he has been with the
Department of Information and Communication Technology (ICT),
in the Islamic University (IU), Kushtia-7003, Bangladesh. He is cur-
rently an Associate Professor in the same Department. His research
interests include Spectrum Sensing, Energy Harvesting, Internet
of Things, MU-MIMO based Cognitive Radio Networks, Massive
MIMO based Cognitive Radio Networks and Machine Learning based
Cognitive Radio based IoT.

JOURNAL OF L
A
T
E
X CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 13
Mohammad Amzad Hossain received his
B.Sc., and M.Sc. in Information and Com-
munication Technology (ICT) from the Is-
lamic University (IU), Kushtia, Bangladesh,
in 2010 and 2011, respectively. He is currently
an Assistant Professor in the Department of
Information and Communication Engineering
(ICE), Noakhali Science and Technology Uni-
versity, Noakhali, Bangladesh. Amzad is cur-
rently pursuing a Structured Ph.D. in the
School of Computer Science, National Univer-
sity of Ireland Galway (NUIG), Galway, Ireland. In 2018, Amzad has
awarded the prestigious College of Science and Engineering postgrad-
uate research Scholarship. His research interests include Spectrum
Sensing, MIMO based Cognitive Radio Networks, Cognitive Radio
based Internet of Things (CR-IoT) Networks and Deep Learning.
Kazi Mowdud Ahmed received his B.Sc.
(Hon’s) and M.Sc. in Information and Com-
munication Technology (ICT) from the Islamic
University (IU), Kushtia-7003, Bangladesh, in
2012 and 2013, respectively. Since 2018, he
has been with the Department of Informa-
tion and Communication Technology (ICT),
in the Islamic University (IU), Kushtia-7003,
Bangladesh. He is currently a lecturer in the
same department. He worked as a research
assistant in Multimedia Communication Sys-
tems Lab. (MCSL) and Computer Vision and Intelligent Interfacing
Lab.(CVIIL) from 2011 to 2016 in the same department. Since 2018,
Mowdud is working as a research assistant in Wireless Communica-
tions Lab (WCL) in the department of Information and Communica-
tion Technology, Islamic University, Kushtia-7003, Bangladesh. His
research interests includes Cloud Computing, Machine Learning and
Deep Learning, AI-enabled Cognitive Radio based IoT.
Dr. Md. Mahbubur Rahman is a Profes-
sor, Department of Information and Commu-
nication Technology, Islamic University, Kush-
tia, Bangladesh. He received the B.Sc. and
M.Sc. degrees in Physics, Rajshahi University,
Rajshahi, Bangladesh. In 1997, Rahman re-
ceived his Ph.D. in Computer Science & En-
gineering from Rajshahi University, Rajshahi,
Bangladesh. He worked as a dean, faculty
of Applied Science and Technology, Islamic
University, Kushtia, Bangladesh. Since 1998,
Mabubur is working as a director in Wireless Communications Labo-
ratory (WCL) in the department of Information and Communication
Technology, Islamic University, Kushtia-7003, Bangladesh. He has
published fifty reputed journal papers. His main research interests
include mobile communications, wireless sensor networks, Internet
of things, cognitive radio networks and AI-enabled networking.
Dr. Ali Calhan received his M.Sc. and Ph.D.
degrees from the University of Kocaeli, Turkey
in 2006 and 2011. Science 2011, he has been
a member of the Computer Engineering De-
partment of Duzce University. Currently, he
is an Associate Professor in the Department
of Computer Engineering, Duzce University
(DU). His research interests are wireless com-
munications, cognitive radio networks, body
area networks and software-defined networks.
Dr. Murtaza Cicioglu received his Ph.D.
degree in Electrical-Electronic and Computer
Engineering from Düzce University, Turkey in
2020. Since 2009, he has been a member of
the Information Technologies Department of
Ministry of National Education, Turkey. He
is a member of the Software Defined Net-
works Community, IEEE. His research inter-
ests include software-defined networking, wire-
less communications, 5G, body area networks,
cognitive radio, and Riverbed Modeler (OP-
NET) simulation software.