Performance enhancing techniques in cognitive radio networks

Full text

Performance Enhancing Techniques in
Cognitive Radio Networks
Sumi.M.S, R.S.Ganesh,
Department of Electronics and Communication Engineering,
Noorul Islam University, Tamil Nadu, India,
sumirhn@gmail.com
r_s_ganesh@rediffmail.com
Abstract—$QLPSRUWDQWIXQFWLRQDOLW\RI
FRJQLWLYH UDGLR QHWZRUN LV VSHFWUXP VHQVLQJ ,W LV
SHUIRUPHGWRGHWHUPLQHWKHLGOHIUHTXHQF\EDQGVWKDWDUH
QRW EHLQJ RFFXSLHG E\ D OLFHQVHG SULPDU\ XVHU DW WKDW
SDUWLFXODU WLPH (YHQ WKRXJK FRRSHUDWLYH VSHFWUXP
VHQVLQJ LV IRXQG WR EH KLJKO\ UHOLDEOH FRPSDUHG ZLWK
VLQJOH XVHU VHQVLQJ LQFUHDVHG HQHUJ\ FRQVX PSWLRQ LV D
QRWDEOHGUDZEDFN9DULRXVPHWKRGVDQGDOJRULWKPVKDYH
EHHQ SURSRVHG VR IDU WR LPSURYH DQG PDLQWDLQ HQHUJ\
HIILFLHQF\ LQ WKH QHWZRUN 7KLV SDSHU JLYHV D EULHI
GHVFULSWLRQDERXW FRRSHUDWLYH VSHFWUXP VHQVLQJDQG WKH
YDULRXV WHFKQLTXHV IROORZHG LQ WKH H[LVWLQJ ZRUNV WR
LPSURYHWKHQHWZRUNSHUIRUPDQFHDQGFRQVHUYHHQHUJ\
Keywords—Cognitive radio, cooperative
sensing [2]. This paper gives a brief overview about
spectrum sensing and its classification in section II. In
section III various representations for cooperative
spectrum sensing are discussed. Section IV explains
about the various metrics important for cognitive radio
networks. Advanced methods adopted in previous
works for obtaining energy conservation are explained
in section V along with conclusion in section VI.
II. SPECTRUM SENSING
Spectrum sensing is an important
functionality of CR. Based on the sensing result,
secondary user (SU) can transmit data via the free
channel [3] and if the channel is not free, the SU can
wait until the channel is free or sense for another free
channel availability. When a SU is accessing a
spectrum band from a PU, it has to make sure that the
SU doesn’t create any interference to the PU. This can
be achieved by two methods. The SU can transmit with
limited power so that it can coexist with the PU and
share the same spectrum band or the SU has to perform
spectrum sensing to detect the availability of free
channel [4]. With respect to the distance of the SU from
the PU, the better of the above two methods can be
considered. That is for SUs far away from the PU,
power control method can be adopted and if the SU is
nearer to the PU, a small interference power can even
affect the PU, hence spectrum sensing is necessary to
check whether the channel is free or not. Also in certain
cases, where the SU is not too far as well as too near to
the PU, both sensing as well as power control can be
used to make sure that the PU is not affected [4] [5].
A. Spectrum sensing methods
The main classification of spectrum sensing
are energy detection method, cyclostationary detection
method and matched filter detection method [6].
Energy detection method is most frequently used due
to its simplicity as it does not need any preceding
knowledge of the PU. The measured signal energy of
the primary user will be compared with a threshold
value to determine the presence of the PU [7].
Energy detection method shows good
performance at high signal-to-noise ratio (SNR).
However certain problems like fading, shadowing and
noise uncertainty leads to signal degradation and SNR
spectrum sensing, energy efficiency, clustering, relay.
I.INTRODUCTION
Due to the advances in the field of wireless
communication and the tremendous growth in the
usage of mobile and other wireless equipments, the
demand for spectrum resources has reached a great
extent. Since our wireless era is moving towards the
commencement of 5G technology, this demand will
increase further. Simultaneously as the need for
spectrum is increasing, scarcity begins. That is as the
number of users increases, the demand for frequency
spectrum exceeds the amount of available spectrum.
But still as per Federal Communications Commission
report, the licensed spectrum is underutilized at
different durations. About 15 to 85% of the spectrum
are only being used [1]. Hence this underutilized
spectrum bands termed as spectrum holes can be
permitted to be accessed by the unlicensed users
(termed as Secondary Users or SUs) until the licensed
user (termed as primary user or PU) is back.
To provide better availability of frequency
spectrum among various users and to maintain the
quality of data communication, some intelligent
technology has to be adopted. One such is cognitive
radio (CR) technology which is an advanced version of
software defined radio. CR can change its transmitter
parameters with respect to its operating environment.
To obtain this dynamic spectrum access, the
availability of free spectrum bands can be recognized
either by making use of the information about primary
user schedule that is already stored in the database or
by means of a real time activity called spectrum
Proceedings of 2017 IEEE International Conference on Circuits and Systems (ICCS 2017)
978-1-5090-6480-9/17/$31.00©2017 IEEE 172

secondary user detects the presence of a primary user,
then the probability of detection increases. And if the
PU is falsely detected to be active while it is actually
absent, then the probability of false alarm increases
[12].
In a centralized cooperative spectrum sensing,
initially the fusion center has to decide which spectrum
band has to be selected for sensing and informs the SUs
to carry out local sensing individually. Then all the SUs
transmit the sensed data to the fusion center. Finally the
fusion center makes the final decision about the
primary user status [13]. Various fusion rules are
followed at the fusion center to make the final decision.
One such is the OR rule, in which the FC decides that
the PU is active if atleast one SU informs PU is present,
otherwise PU is considered to be absent. Whereas in
AND rule, the FC decides the channel is free, if atleast
one SU reports PU is absent, otherwise PU is active.
Also in certain cases, where the SU cannot make a local
decision (i.e.) PU neither present nor absent, the FC
considers such decisions as PU is present (i.e.) the
channel is busy even though such decisions are
censored off [14]. Another fusion rule being applied is
the majority rule where, out of total N SUs, if K SUs
decide channel is free, then FC decides channel is idle
and PU is absent, where the value of K has to be
carefully selected. Chair-Varshney fusion rule is an
optimal rule developed for global decision fusion. Here
the weighted sum of local decisions are calculated
where the functions of probability of detection and
probability of false alarm are taken as weights [15].
In the case of distributed cooperative
spectrum sensing, there is no central control (FC). But
the SUs sense the spectrum band and share the local
sensing information among all users and each
individual user makes its own final decision about the
channel status combining the received decisions from
other secondary users as well as its own decision.
Apart from the above two approaches, relay
based cooperative spectrum sensing can also be used.
If among a group of secondary users, few users may
have strong sensing channels and weak reporting
channels while vice versa in other users, then the
secondary users with strong reporting channels can act
as relay nodes for those SUs with weak reporting
channels to transmit their sensed information. This
relay based approach can work as either a centralized
or distributed method [16].
III. SYSTEM MODEL
A. Representation of cooperative spectrum sensing
Fig. 1 represents a simple example for
cooperative spectrum sensing, where the system has a
single PU and multiple secondary users with a centrally
located fusion center. Each SU performs spectrum
sensing individually and transmits the sensed
information to the FC. The fusion center combines the
gets reduced. Hence energy detection results are not
reliable [6]. Cyclostationary detection makes use of the
periodicity conditions in the received primary signal to
find out whether the PU is active or not. This property
is mainly obtained using pulse trains, cyclic prefixes,
sinusoidal carriers, hopping sequences as well as
spreading code of the PU signals. This cyclostationary
detector is robust to noise uncertainty and hence able
to detect primary user even when the SNR is low.
However it requires more detection time [8]. This is
mainly due to the increase in computational complexity
[9]. Matched filter detection is another method which
also requires preceding knowledge of the primary user.
This information is considered to be as the known
signal. This known signal will be correlated with the
received signal (unknown signal) to find whether the
primary user signal is present in the received signal
[10], which indicate that the primary user is active in
the channel.
Among the above three methods, energy
detection has least complexity as it does not require
prior information about primary user signal and least
accuracy especially at low SNR, because it is unable to
differentiate between signal and noise. At the same
time, matched filter method is found to have highest
accuracy even at low SNR and highest complexity as it
requires prior information about primary user. Based
on the channel noise conditions and primary user
information availability, the most suitable detection
method can be selected [11].
B.Cooperative spectrum sensing
Single user spectrum sensing is not always
effective due to the presence of problems like noise
uncertainty, shadowing and fading. Hence to improve
the reliability of sensing even at low SNR and to avoid
hidden node problem, cooperative spectrum sensing
(CSS) is considered [6] [7]. Cooperative spectrum
sensing can be carried out with multiple secondary
users involved in spectrum sensing with or without the
use of fusion center (FC). Subject to the presence or
absence of the fusion center, CSS can be classified as
either centralized or distributed. In the case of
centralized cooperative spectrum sensing, the
secondary users sense the spectrum at a particular time
slot during the sensing phase and transmit the sensed
information during the reporting phase to the fusion
center in which the final decision about the existence
of primary user is determined by following any fusion
rule. Based on the local sensing result, the secondary
user can make a single bit decision indicating the
presence or absence of PU and transmit it to the fusion
center for final decision. This method is termed to be
as decision fusion. And if the secondary user directly
transmits the received signal to the FC, then it is said
to be as data fusion [7]. Cooperative spectrum sensing
can be justified by two important parameters. One
among them is the probability of detection (P
d
) while
the other one is the probability of false alarm (P
f
). If a
Proceedings of 2017 IEEE International Conference on Circuits and Systems (ICCS 2017)
173

local sensing information and decides whether the
channel is idle or not. Energy detection based spectrum
sensing can be performed using a hypothesis testing
method. If the received signal r(n) is measured as
r(n) = w(n) (1)
it represents the absence of the PU signal and if
r(n) = h(n)x(n) + w(n) (2)
then PU is active, where w(n) is the noise in the channel
and x(n) is the PU signal and h(n) denote the impulse
response of the channel [17]. Fig. 2 represents a
network for centralized cooperative spectrum sensing
with multiple primary and secondary users. These SUs
are involved in performing spectrum sensing and report
to the centralized fusion center [18]. Similarly Fig. 3
represents a distributed approach for cooperative
spectrum sensing, where no centralized fusion center is
required [18].
Fig. 1. Example model representing cooperative spectrum sensing
Fig. 2. Centralized cooperative spectrum sensing with multiple PUs
Fig. 3. Distributed cooperative spectrum sensing with multiple PUs
B. Time frame in cognitive radio spectrum sensing
Generally in cognitive radio networks
performing cooperative spectrum sensing, each time
frame T
l
can be given as a summation of the sensing,
reporting and data transmission time i.e.
T
l
= T
ss
+ T
d
(3)
as in Fig. 4,
where T
ss
is the combination of location
sensing and reporting time for N cooperating
secondary users and T
d
is the data transmission time
slot as in [19]. That is,
T
ss
= T
s
+ NT
r
(4)
Where T
s
is the common sensing time for N secondary
users and T
r
is the reporting time slot for a single user.
If the local sensing time is increased, then more
accurate detection can be made, but the time available
for data transmission is reduced which lowers data
transmission. But at the same time as detection is
improved, unnecessary missed detections and
collisions can be avoided. Hence it is found to be
essential that sensing time is optimized. The data
transmission time slot can be either flexible or fixed. If
T
d
is fixed, then variation in the sensing or the reporting
time will not have any impact on data transmission.
Fig. 4. Time frame for cognitive radio network
Proceedings of 2017 IEEE International Conference on Circuits and Systems (ICCS 2017)
174

IV. CERTAIN METRICS CONSIDERED IN while simultaneously optimizing their respective
transmission powers in order to achieve higher
throughput.
B. Energy efficiency concerns
In order to maintain proper spectrum usage
and to avoid interference from the secondary users to
the primary user due to misdetections, proper and
efficient spectrum sensing has to be made. More
energy is consumed while performing spectrum
sensing. Also as the primary user may reappear in the
band at any time, spectrum sensing must be done
continuously which further increase energy
consumption. Hence some energy conservation
methods have to be considered so that spectrum
efficiency is maintained without creating interference
to PUs and other SUs in the network. About 25% of the
carbon dioxide foot print is mainly from the
information and communications field in which the
mobile industry is noted to be the highest energy
consumer. Various mobile communication industries
aim at reducing this carbon dioxide emission by 2020.
Research works supporting energy efficiency such as
EARTH (Energy Aware Radio and network
Technologies), Green Radio, Green Touch Consortium
etc. have been developed [25].
Energy consumption increases not only by
spectrum sensing, but also due to channel switching
and transmission of data [3]. Sensing time and
spectrum access are optimized to reduce energy
consumption without affecting the system reliability
and throughput. Energy consumption is increased due
to the rise in sensing time. But still due to longer
sensing time, more error free transmissions are
possible which saves energy. Hence an optimized
sensing time is obtained along with optimizing the
probability whether the user decide to wait in the
channel or switch to new channel when a channel is
found to be busy. A two phase cooperative spectrum
sensing is carried out in [17], where if the SNR is large
and also PU is absent, then only one stage of sensing is
required. In other conditions, one more stage of sensing
is required so that sensing results can be improved.
Also the local sensing results are transmitted to the
fusion center using a single bit which as a whole saves
energy. The above method is further modified by
another algorithm in which, the global decision is taken
at the FC using the first stage local decisions and if no
appropriate result is obtained, another stage of sensing
is performed. Here energy efficiency increases with
decrease in detection reliability. In [12], energy
efficiency can be improved by following an iterative
algorithm to optimize the number of secondary users.
Also a suboptimal method is used to improve energy
efficiency without degrading the system throughput.
As a result, it is noted that obtaining energy efficiency
is related to the number of sensing users. The final
decision threshold required for the fusion center is
optimized in [6] so that energy efficiency can be
COGNITIVE RADIO NETWORKS
A.Throughput maximization
An important performance measure of any
network is throughput which gives the actual rate at
which data is being transmitted. At higher SNR value,
detection probability is increased and also the sensing
time can be reduced which simultaneously increases
the transmission time resulting in an increase in system
throughput. Reference [4] performs optimized sensing
along with power control with respect to the location
information between the secondary user transmitter
and the primary user transmitter to improve the
throughput. If the secondary user can transmit with
more power, then maximum throughput can be
achieved. For an intermediate distance, the SU can
transmit only with minimum power which results in
reducing the throughput. But at the same time since
collision is avoided, throughput can be increased. In
[12], a throughput maximization problem is
formulated. When more number of secondary users are
involved in spectrum sensing, more accurate sensing
can be achieved and hence throughput is increased. But
at the same time, increasing the number of sensing
users increases the reporting time which reduces the
number of samples received by the SUs and as a whole
the throughput is affected. Hence the number of
sensing users has to be optimized such that the sensing
and reporting time can be properly divided to obtain
throughput maximization. Here the data transmission
time slot is considered to be fixed [12]. Also it is
observed in [20] that as the number of SUs are getting
increased, there is a decrease in the reporting error with
respect to the FC. That is if more number of SUs
identify that PU is active, then the bit error rate is
reduced and the SNR is increased, finally enhancing
the capacity of the system.
In [21] as the number of sensing users
increases, the system throughput increases. With
further increase in the number of sensing users, the
throughput decreases and becomes constant. For lower
SNR values, more number of sensing users are required
for maximizing the throughput. Also the complexity of
the sensing method depends on the complexity of
throughput estimation as well as finding the parameters
to be optimized [22]. If the sensing and reporting time
are increased to obtain better sensing results, then the
data transmission time will be reduced which further
reduces the throughput of the SUs. To overcome this,
a limited reporting method is proposed in [23] in which
the maximum number of reporting users as well as the
frequency bands to be reported are adjusted vigorously
according to the total number of SUs. Combined
optimization of sensing and reporting time is
performed to improve the throughput of the SUs. In
[24], combined optimization of sensing and
transmission time of the source and the relay nodes
assisting the cognitive radio networks is
performed
Proceedings of 2017 IEEE International Conference on Circuits and Systems (ICCS 2017)
175

increased. This optimal threshold value changes with
respect to the SNR as well as the type of channel being
used. Combined optimization of sensing time, sensing
threshold and number of sensing users is performed in
[26] to minimize energy consumption. A weighting
factor is introduced and based on its value, either
energy efficiency or throughput is given importance.
In order to perform cooperative spectrum
sensing, in [7] double threshold values (high and low)
are used. These values are jointly optimized along with
the sensing time to improve system throughput and
energy efficiency. Energy efficiency which can be
measured in bit/s/joule is inversely proportional to the
total power consumed by the SUs. An iterative
algorithm is followed in [27], in which both the sensing
time and transmission power are optimized to obtain
higher energy efficiency. The base stations present in
cognitive radio networks have power amplifiers which
are accountable for most of the total energy consumed.
Hence in [28], a sleeping mode is followed so that
certain base stations that are not utilized are turned off
and hence energy can be saved. Especially the base
stations which are far away from the network can be
chosen to be switched off.
V. ADVANCED METHODS FOR
where a group of SUs form a cluster with a centralized
cluster head
Fig. 5. Clustering approach for cooperative spectrum sensing
Three steps are involved in the clustering approach in
[32]. Initially the SUs that cannot predict the PU status
are identified and are not included in the cluster. Later
among the remaining SUs, the highly reliable user is
chosen as the cluster head. Also the users whose
sensing results are not related are placed in the same
cluster. FC is not present in this system and there is a
reduction in the number of transmissions and hence
energy is saved. Even when the size of the cluster is
larger, accuracy is not affected and energy is
conserved.
B. Censoring in cognitive radio networks
Censoring is another method of minimizing
energy consumption in a cognitive radio system where
the secondary users’ decisions which fall within the
upper and lower threshold values cannot be used to
predict the status of primary users. Such decisions are
not transmitted to the fusion center [7]. Censoring has
been performed along with clustering in [33], where
the number of bits transmitted to the FC is reduced. As
a result, there is a reduction in the transmission power
of the cognitive users paving way for energy
efficiency. Censoring is combined along with sleeping
method in [14], where each SU is inactive for a
particular period of time and when the sleeping period
expires, the SU performs spectrum sensing following
the censoring policy. Both sensing and reporting
energy are reduced without degrading the performance
of the system.
C. Relay supported cognitive radio networks
With the help of relay in cognitive radio
network, both the sensing results as well as the system
throughput can be improved. Also it results in
minimizing energy consumption. For such a network
with relay node as in [24], both the cognitive user and
the relay sense the spectrum band and the results are
transmitted to the fusion center by forwarding methods
like amplify and forward (AF), decode and forward
(DF) or hybrid method. Fig. 6 represents a cognitive
radio network with a relay node. Here the source sense
the channel and transmit the sensed data to the relay
IMPROVING ENERGY EFFICIENCY
Even though energy efficiency can be
obtained by optimizing various sensing parameters, the
structure of cooperating SUs can also be modified to
minimize energy consumption. Clustering, censoring,
relay based spectrum sensing etc. are few techniques
followed in cooperative spectrum sensing for
improving energy efficiency.
A.Clustering in cognitive radio networks
The secondary users involved in cooperative
spectrum sensing can be arranged in a form of clusters
where in each cluster, a single user is selected as the
cluster head. The local sensing results from the
individual SUs can be conveyed to the fusion center via
the cluster head. As a result energy consumption can
be minimized. In such cluster based approach, the size
of the cluster and its setup can be varied with respect to
the channel to be sensed [29]. In [30], a clustering
scheme is used where the cluster head is determined
based on sensing performance of the users for each
duration. Frequency division method is followed to
report the decisions from each cluster to the FC, which
can reduce collisions. Energy consumption can also be
reduced as cluster head alone transmits to the FC. Two
stages of fusion are employed in [31], where the cluster
is divided in to groups and sub groups. First the data
from the sub groups are fused by the group head and
the group decisions are fused again by the cluster head
from where the decisions are transmitted to the FC.
Since reporting distance is reduced, reporting energy is
saved compared with the earlier methods. Fig. 5
represents a system model with clustering approach
Proceedings of 2017 IEEE International Conference on Circuits and Systems (ICCS 2017)
176

which decodes and forwards the information to the
destination [34].
Fig. 6. Cognitive radio network with relay node
Sensing time and SNR are jointly optimized in [21] to
improve energy efficiency of a cognitive radio network
with relay node. Also the number of SU and the
threshold of the fusion rule are properly selected to
obtain energy conservation. When a free spectrum
band is detected, the cognitive source node forward the
information to the destination and the relay node at the
same time during a particular time slot and then during
the next slot, the relay alone transmits the information
to the destination, where both the received signals are
combined to produce highly reliable detection results.
Table I gives a brief overview of the enhanced
performance in cognitive radio networks as a result of
different schemes in cooperative spectrum sensing.
TABLE I
OVERVIEW OF DIFFERENT SCHEMES IN COOPERATIVE
SPECTRUM SENSING
$OJRULWKPPHWKRGV
IROORZHG
7HFKQLTXHV
HPSOR\HG
0HULWVDFKLHYHG
Cooperative
spectrum sensing
with limited
reporting method
[23]
Combined
optimization of
sensing and
reporting time
Throughput is
improved.
Optimization
process for energy
utility maximization
[26]
Joint
optimization of
sensing
threshold,
sensing time and
number of
secondary users
High throughput
and better energy
conservation are
achieved.
Bi-threshold scheme
for cooperative
spectrum sensing [7]
Joint
optimization of
detection
threshold and
sensing time
Better throughput
and energy
efficiency trade-
off is achieved for
different values of
SNR.
Cluster based
cooperative
spectrum sensing
[29]
Clustering Network
performance is
improved as
cluster size
increases.
Optimizing the
number of clusters
reduces the false
alarm probability,
thereby increasing
the throughput for
each user.
Cluster based
optimal selective
method [30]
Clustering Sensing efficiency
increases with rise
in the cluster size
(nearly 20).
Reporting time
and energy
consumption are
reduced.
Multilevel
hierarchical
structure algorithm
[31]
Clustering Improves
detection
performance and
throughput.
Reporting energy
and overhead are
reduced.
Three stage
clustering scheme
[32]
Clustering,
censoring
Performs better
under low SNR
condition.
Reduces sensing
overhead and
achieves better
sensing results
Cluster based
weighted sensing
with frequency
detection [33]
Clustering with
censoring
Sensing time is
reduced.
Transmission
power of SUs is
minimized
resulting in better
battery
performance.
Combined censoring
and sleeping scheme
for distributed
spectrum sensing
[14]
Censoring and
sleeping method
Energy
consumption is
minimized.
Joint spectral
sensing and
secondary data
transmission method
[24]
Relay Throughput is
maximized by
optimizing sensing
and transmission
time. Decode and
forward relay
gives improved
throughput than
amplify and
forward method.
Sensing with joint
optimization of
sensing time and
signal to noise ratio
[21]
Relay Energy efficiency
is maximized. By
optimizing the
threshold for
fusion rule and the
number of sensing
users, more energy
conservation can
be obtained.
VI. CONCLUSION
In order to deal with the destructive effects of
shadowing, fading etc. in spectrum sensing,
cooperative spectrum sensing is performed. Even
though cooperative spectrum sensing results in highly
reliable sensing results, it is also associated with
considerable drawbacks which cannot be ignored. This
paper gives a brief overview about spectrum sensing
methods and various types of cooperative spectrum
sensing. Also different optimization methods adopted
previously for achieving throughput maximization and
increasing energy efficiency are discussed. Various
approaches like clustering, censoring and relaying that
Proceedings of 2017 IEEE International Conference on Circuits and Systems (ICCS 2017)
177

are introduced in cooperative spectrum sensing in order
to obtain energy conservation are also learnt.
REFERENCES
[1] N.Muchandi and R.Khanai, “Cognitive radio spectrum
sensing: A survey”, in Proc. IEEE International Conference on
Electrical, Electronics, and Optimization Techniques
(ICEEOT), pp. 3233-3337, Mar 2016.
[2] K.Cichon, A.Kliks, and H.Bogucka, “Energy-efficient
cooperative spectrum sensing: A survey”, IEEE
Communications Survey and Tutorials, Vol. 18, No. 3, pp.
1861-1886, April 2016.
[3] S.Wang, Y.Wang, J.P.Coon and A.Doufexi, “Energy-efficient
spectrum sensing and access for cognitive radio networks”,
IEEE Transactions on Vehicular Technology, Vol. 61, No. 2,
pp. 906-912, Feb 2012.
[4] E.C.Y.Peh, Y.C.Liang and Y.Zeng, “Sensing and power
control in cognitive radio with location nnformation”, in Proc.
IEEE ICCS, pp. 255-259, Nov 2012.
[5] H.Hu, H.Zhang and N.Li, “Location-information-assisted joint
spectrum sensing and power allocation for cognitive radio
networks with primary-user outage constraint”, IEEE
Transactions on Vehicular Technology, Vol. 65, No. 2, pp.
658-672, Feb 2016.
[6] H.Hu, H.Zhang, H.Yu, Y.Chen and J.Jafarian, “Energy-
efficient design of channel sensing in cognitive radio
networks”, Computers and Electrical Engineering, Vol. 42,
Issue C, pp. 207-220, Feb 2015.
[7] M.Moradkhani, P.Azmi and M.A.Pourmina, “Optimized
energy limited cooperative spectrum sensing in cognitive radio
networks”, Computers and Electrical Engineering, Vol. 42,
Issue C, pp. 221-231, Feb 2015.
[8] K.W.Choi, W.S.Jeon and D.G.Jeong, “Sequential detection of
cyclostationary signal for cognitive radio systems”, IEEE
Transactions on Wireless Communications, Vol. 8, No. 9, pp.
4480-4485, Sep 2009.
[9] D.Ghosh and S.Badchi, “Cyclostationary feature detection
based spectrum sensing technique of cognitive radio in
Nakagami-m fading environment”, Computational Intelligence
in Data Mining, Vol. 2, pp. 209-219, Dec 2014.
[10] S.Kapoor, S.Rao and G.Singh, “Opportunistic spectrum
sensing by employing matched filter in cognitive radio
network”, in Proc. IEEE International Conference on
Communication Systems and Network Technologies, pp. 580-
583, June 2011.
[11] F.Salahdine, H.E.Ghazi, N.Kaabouch and W.F.Fihri, “Matched
filter detection with dynamic threshold for cognitive radio
networks”, in Proc. IEEE International Conference on
Wireless Networks and Mobile Communications (WINCOM),
pp. 1-6, Oct 2015.
[12] S.Althunibat, M.D.Renzo and F.Granelli, “Cooperative
spectrum sensing for cognitive radio networks under limited
time constraints”, Computer Communications, Vol. 43, pp. 55-
63, May 2014.
[13] R.Gao, Z.Li, P.Qi and H.Li, “A robust cooperative spectrum
sensing method in cognitive radio networks”, IEEE
Communications Letters, Vol. 18, No. 11, pp. 1987-1990, Nov
2014.
[14] S.Maleki, G.Leus, S.Chatzinotas and B.Ottersten, “To AND or
To OR: On energy-efficient distributed spectrum sensing with
combined censoring and sleeping”, IEEE Transactions on
Wireless Communications, Vol. 14, No. 8, pp. 4508-4521, Aug
2015.
[15] E.C.Y.Peh, Y.C.Liang, Y.L.Guan and Y.Zeng, “Cooperative
spectrum sensing in cognitive radio networks with weighted
decision fusion schemes”, IEEE Transactions on Wireless
Communications, Vol. 9, No. 12, pp. 3838-3847, Dec 2010.
[16] I.F.Akyildiz, B.F.Lo and R.Balakrishnan, “Cooperative
spectrum sensing in cognitive radio networks: A survey”,
Physical Communication, Vol. 4, No. 1, pp. 40-62, Mar 2011.
[17] N.Zhao, F.R.Yu, H.Sun and A.Nallanathan, “Energy-efficient
cooperative spectrum sensing schemes for cognitive radio
networks”, Eurasip Journal on Wireless Communications and
Networking, May 2013.
[18] A.Zakaria, M.Tahir, N.Ramli, H.Mohamad and M.Ismail,
“Performance evaluation of centralized and decentralized
cooperative spectrum sensing in cognitive radio networks”, in
Proc. IEEE International Conference on Computer and
Communication Engineering (ICCCE 2012), pp. 283-288, July
2012.
[19] M.A.Sarijari, M.S.Abdullah, G.J.M.Janssen and A.J.Veen,
“On achieving network throughput demand in cognitive radio-
based home area networks”, EURASIP Journal on Wireless
Communications and Networking, Dec 2015.
[20] K.Kalimuthu and R.Kumar, “Capacity maximization in
spectrum sensing for cognitive radio networks thru outage
probability”, International Journal of Electronics and
Communications, Vol. 67, No. 1, pp. 35-39, 2013.
[21] S.Yaolian, Z.Fan, S.Yubin and L.Hua, “Energy efficiency
optimization of cognitive relay network based on cooperative
spectrum sensing”, The Journal of China Universities of Posts
and Telecommunications, Vol. 22, No. 3, pp. 26-34, June 2015.
[22] J.So, “Energy-efficient cooperative spectrum sensing with a
logical multi-bit combination rule”, IEEE Communications
Letters, Vol. 20, No. 12, pp. 2538-2541, Dec 2016.
[23] J.So and T.Kwon, “Limited reporting-based cooperative
spectrum sensing for multiband
cognitive radio n etworks”, International Journal of Electronics
and Communications, Vol. 70, No. 4, pp. 386-397, April 2016.
[24] S.Chatterjee, T.Acharya and S.P.Maity, “On optimized decode
and forward relay assisted CR system design for throughput
maximization”, Digital signal processing, Vol. 34, pp. 92-100,
Nov 2014.
[25] Y.Chen, S.Zhang and S.Xu, “Fundamental trade-offs on the
design of green radio networks”, Green Radio Communication
Networks, pp. 3-23, Cambridge University Press, 2012.
[26] X.Wu, J.L.Xu, M.Chen and J.B.Wang, “Optimal energy-
efficient sensing in cooperative cognitive radio networks”,
EURASIP Journal on Wireless Communications and
Networking 2014, Dec 2014.
[27] D.Das and S.Das, “Optimal resource allocation for soft
decision fusion-based cooperative spectrum sensing in
cognitive radio networks”, Computers and Electrical
Engineering, Vol. 52, pp. 362-378, May 2016.
[28] E.F.Orumwense, T.J.Afullo and V.M.Srivastava, “On
increasing the energy efficiency of cognitive radio network
base stations”, in IEEE 7th Annual Computing and
Communication W orkshop and Conference (CC WC), Jan 2017.
[29] S.Hussain and X.Fernando, “Approach for cluster-based
spectrum sensing over band-limited reporting channels”, IET
Communications, Vol. 6, No. 11, pp. 1466-1474, July 2012.
[30] N.N.Thanh and I.Koo, “A cluster-based selective cooperative
spectrum sensing scheme in cognitive radio”, EURASIP
Journal on Wireless Communications and Networking 2013,
2013.
[31] F.A.Awin, E.A.Raheem and M.Ahmadi, “Optimization of
multi-level hierarchical cluster-based spectrum sensing
structure in cognitive radio networks”, Digital Signal
Processing, Vol. 36, pp. 15-25, Jan 2015.
[32] Y.Jiao, P.Yin and I.Joe, “Clustering scheme for cooperative
spectrum sensing in cognitive radio networks”, IET
Communications, Vol. 10, No. 13, pp. 1590-1595, Sep 2016.
[33] K.G.Smitha and A.P.Vinod, “Cluster based power efficient
cooperative spectrum sensing under reduced bandwidth using
location information”, International Jou rnal of Electronics and
Communications, Vol. 66, No. 8, pp. 619-624, Aug 2012.
[34] J.Zhu, J.Huang and W.Zhang, “Optimal One-Dimensional
Relay Placement in Cognitive Radio Networks”, in Proc. IEEE
International Conference on Wireless Communications and
Signal Processing, Oct 2010.
Proceedings of 2017 IEEE International Conference on Circuits and Systems (ICCS 2017)
178