Dynamic Resource Allocation and Priority Based Scheduling for Heterogeneous Services in Cognitive Radio Networks

Abstract

As a futuristic solution to the serious scarcity of spectrum and radio resources in today's scenario, cognitive radio offers great solutions. Cognitive radio is a groundbreaking approach in improving the utilization of precious natural resources. Cognitive radio network (CRN) is an intelligent wireless communication system that is aware of its environment. The problem of spectrum sensing by heterogeneous nodes, with varying computing power and sensing range, and spectrum allocation in CRN where there exist multiple primary users and secondary users have been considered. Focus has been laid on optimum resource allocations and scheduling for primary user (PU) and secondary user (SU) to maximize the performance of the whole network by reducing the interference at PU and also maximizing the throughput and minimizing the risk of overlapping the coverage of CRNs. The effectiveness of DRAPBS-CRN algorithm is compared with DRA-CRN and is found to be 26% more efficient using ns-2 simulation.

Full text

127
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
Dynamic Resource Allocation and Priority Based Scheduling for Heterogeneous
Services in Cognitive Radio Networks
Santhamurthy Tamilarasan1* Kumar Parasuraman 2
1Manonmaniam Sundaranar University, Tirunelveli, Tamil Nadu, India
2Manonmaniam Sundaranar University, Tirunelveli, Tamil Nadu, India
* Corresponding author’s Email: stamilarasan0606@gmail.com
Abstract: As a futuristic solution to the serious scarcity of spectrum and radio resources in today’s scenario,
cognitive radio offers great solutions. Cognitive radio is a groundbreaking approach in improving the utilization of
precious natural resources. Cognitive radio network (CRN) is an intelligent wireless communication system that is
aware of its environment. The problem of spectrum sensing by heterogeneous nodes, with varying computing power
and sensing range, and spectrum allocation in CRN where there exist multiple primary users and secondary users
have been considered. Focus has been laid on optimum resource allocations and scheduling for primary user (PU)
and secondary user (SU) to maximize the performance of the whole network by reducing the interference at PU and
also maximizing the throughput and minimizing the risk of overlapping the coverage of CRNs. The effectiveness of
DRAPBS-CRN algorithm is compared with DRA-CRN and is found to be 26% more efficient using ns-2 simulation.
Keywords: Cognitive radio; MIMO; Femtocell; Nash equilibrium; SINR.
1. Introduction
1.1. Cognitive Radio and Intelligent Antenna
In recent years, there has been a hike in a
number of internet users (accessing in a variety of
ways) [1]. The licensed users (primary users), use
the licensed spectrum for special purposes [2]. In
practical application, the licensed spectrum has very
low utilization compared to the non-licensed band.
Cognitive Radio (CR) is an intelligent radio [3], that
can be programmed in any given network, by using
the best wireless channel available in its vicinity [4,
5, 6, 7] as one key technology to activate the
utilization of spectrum resource [8, 9]. Hence, such
a Dynamic Spectrum Access improves the
performance of the communication network and
makes the network more reliable and robust even
though the nodes are mobile and has fewer power
sources [10, 11, 12, 13].
In a cognitive radio with intelligent antenna, a
transmitter/receiver is designed to intelligently
detect whether a particular segment of the radio
spectrum is currently in use or not and to access
(and out of, as necessary) the temporarily unused
spectrum very rapidly, without interfering with the
transmissions of other authorized users [14, 15].
Smart Antenna (also known as adaptive array
antenna, multiple antennas and MIMO) are antenna
arrays with smart processing algorithm used to
identify special signal signature such as the direction
of arrival of the signal to track and locate the
antenna beam on the target [16, 17, 18]. Hence
determines the best signal to be obtained with high
SNR and optimum power utilization.
1.4 Metropolitan Network and Femtocell
A metropolitan area network (MAN) [19] is a
network that interconnects users with computer
resources in a geographic area or region larger than
that covered by even a large local area network
(LAN) but smaller than the area covered by a wide
area network (WAN) [20, 21, 22]. The term is
applied to the interconnection of networks in a city
into a single larger network (which may then also
offer efficient connection to a wide area network).
[23] To increase capacity and exploit physical
entities, reduction of cell size is entertained. But in
recent years, it reached its limit of sustainability by
the introduction of femtocell [24, 25].

128
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
Femtocell [3] is a wireless access point that
improves cellular reception inside a home or office
building. In telecommunication, a Femtocell acts a
small low-power cellular base station. In
telecommunication industry use of small cell, low-
power access points operating in licensed and
unlicensed spectrum, with femtocells as the subset
has been used widely. A femtocell allows the
service provider to extend service coverage at cell
edges or indoor, especially where access would
otherwise be limited or unavailable. The range of a
femtocell is in the order of 10 meters.
Femtocells were originally called access point
base station. They are compatible with CDMA2000,
WiMAX and UMTS mobile telephony devices,
using the provider's own licensed spectrum to
operate. Consumer-oriented femtocells will support
no more than 4 active users while enterprise-grade
femtocells can support up to 16 active users. The use
of femtocell in metropolitan area network (MAN)
[7] can increase the efficiency of spectrum usage
and power allocation.
The rest of the paper is organized as follows: In
section 2, literature review, the background and
related works are described, in Section 3 the
problem identification along with the solution to
overcome the drawbacksis described(DRAPBS-
CRN) and Section 4, describes the proposed
DRAPBS-CRN approach. Section 5 shows the
simulation result and compares DRAPBS-CRN with
the existing DRA-CRN approach, and Section 6
concludes the article.
2. Literature Review
P. Cheng et al [1] have designed a game
approach for the problem of resource allocation in
an OFDMA based cell as a two-tier game. In RA-
Game, each PU dynamically accesses the sub-
channels according to his payoff while in PS-Game;
each SU buys a radio from some PU to maximize
his utility. Furthermore, they propose two
algorithms to allocate resource among PUs and SUs
in a distributed manner. Their algorithms can
converge to Nash equilibrium automatically, which
maximizes the utilization of the whole network
resource. Besides, their relay scheme greatly
improves the throughput of SUs.
Y. Tachwali et al [2] have proposed a novel
resource allocation framework based on the
bandwidth-power product minimization, which is an
effective metric in evaluating the spectral resource
consumption in a cognitive radio environment. The
framework takes into consideration the challenges
aforementioned. The proposal shows a significant
enhancement in spectral efficiency.
R. Xie et al [3] have proposed a spectrum
sharing and resource allocation for energy-efficient
heterogeneous cognitive radio networks with
femtocells. They use the price of interference to
model the interference between femtocells and
macrocells. They formulate the problem of
interference management and power allocation as a
Stackelberg game. In addition, an iteration algorithm
based on price updating is proposed to obtain the
Stackelberg equilibrium solution to the resource
allocation problem for energy efficiency.
Y. Teng et al [4] have proposed energy-efficient
resource allocation in the multi-secondary user (SU)
cognitive radio networks with network coding based
cooperative transmission (NcCT). They set up a
framework for multi-SU resource allocation game
with Nash bargaining solution (NBS) under the
cognitive radio scenario (CR-MSU-NBS game).
These are the sum of pairwise NBS function with
pairing strategy is exploited as the network
optimization objective and context conditions as
constraints.
J. M. Alfonso and L. B. Agudelo [7] present a
new approach of Spectrum Sensing using a
Compressive Sensing technique named Finite Rate
of Innovation in a Cognitive Radio Network with
centralized Spectrum Management based Spectrum
Broker in the next generation wireless
communications networks. The use of compressive
sensing techniques improves the performance of the
control channel in cognitive radio due the traffic
control protocol requires smaller packet sizes. The
performance of the cognitive network in function of
the control packet size was determinate by analysis
of collisions when different secondary users trying
to access spectrum resources and they make the
request to the Spectrum Broker. They observed that
there are fewer collisions between control packets
and collision probability is smaller if the
compressive technique is used, thus improving the
performance in a fair resource allocation for
cognitive radio networks.
C. Gao et al [8] have mathematically modeled
the opportunities and constraints for CR network
with the objective of maximizing the weighted
network throughput. They proposed a centralized
algorithm and a distributed algorithm to flexibly
assign spectrum channel or spatial DoF exploiting
the multiuser diversity, channel diversity and spatial
diversity for a higher performance in a practical
network. The algorithm further supports different
transmission priorities, reduces transmission delay
and ensures fair transmissions among nodes by

129
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
providing all nodes with certain transmission
probability. Results show that their algorithm is very
effective and can significantly increase the network
throughput while reducing the delay.
D. G¨oz¨upek and F. Alag¨oz [9] have proposed
a very general scheduling model achieving goals
such as making frequency, time slot, and data rate
allocation to secondary users with possibly multiple
antennas, in a heterogeneous multi-channel and
multi-user scenario. Their schedulers ensure that
reliable communication between the cognitive base
station and secondary users. They also propose a
heuristic algorithm for the fair schedulers. Their
performance is in terms of both total throughput and
fairness for varying number of secondary users,
frequencies, antennas, and window size.
3. Problem Identification and Solution
From the literature review, it can be concluded
that there is no work which jointly provides resource
allocation and scheduling in CRN. Considering
networks with heterogeneous services would be
difficult in allocating the resources and assign the
channel to each node in the given network. In [12],
only the secondary users with different rate
requirements are considered ignoring the delay
requirements. Moreover, while estimating the
channel condition for efficient communication, the
channel quality indicator (CQI) has to be considered.
In this proposal, we propose to develop a
dynamic resource allocation and priority-based
scheduling for heterogeneous services in cognitive
radio network (CRN). The network consists of
primary users (PU) and secondary users (SU) with
different service requirements in a heterogeneous
network. The secondary base station (BS) is
responsible for resource allocation for different SU’s
[12] present in the network. The SU’s can be
categorized as: SU with minimum-rate guarantee
(MRG) [12], SU with minimum delay guarantee
(MDG), SU with minimum rate and delay guarantee
(MRDG) and SU with best effort service (BE).
While allocating resources to the set of SU, the
objective function should satisfy the following
constraints:
(i) Total power constraint [12] in which the
total transmit power of SU on all channels
should be within the power budget at the
secondary BS.
(ii) For MRG, the transmission rate for SU
should be greater than the minimum-rate
threshold [12]
(iii) For MDG, the transmission delay for SU
should be less than the deadline threshold.
(iv) For MRDG, the transmission rate should
be greater than the minimum-rate threshold
and the transmission delay should be less
than the deadline threshold.
(v) For BE, fairness constraint should be
satisfied.
Here a channel quality indicator (CQI) is used as
a utility function [13] for each channel. The CQI is
estimated in terms of the signal-to-interference-plus-
noise ratio (SINR). For each incoming stream, the
priority of packets is determined depending on the
service type and queuing delay [8]. Then an
objective function is estimated for each stream by
multiplying the priority with channel gain. Then the
streams can be sorted in the descending order of this
objective values, then allocated to the respective
type of SU. The stream with highest objective
function is allocated to the MRDG, followed by
MRG and MDG. The stream with least objective
function is assigned to BE.
4. Proposed Solution
4.1. System Model
In cognitive radio communication systems, users
are dived into primary users and secondary users
(secondary users). The secondary users (cognitive
users) have a relative low spectrum access authority.
Secondary users must continuously sense the
environment and must be aware of any changes in
their wireless network and make adjustments to
communication changes accordingly without
causing harmfulinterference to the primary users.
Here as already know the primary users represent
the licensed spectrum users whereas the intelligent
cognitive users (secondary users) represent the users
accessing the unlicensed spectrum for
communication. The main objective of the spectrum
sensing is to find available free spectrum resources
(un-accessed channels), estimate the channel quality,
transmit the data packets and design self-adaptive
transmitting waveform which fit spectrum
characteristics based on communication needs.
When the channels are not accessed by the primary
users the secondary users can access the idle channel
optimally. And periodically check the ability of the
PU such that the Cognitive users’ handovers the
channel to the PU as soon as a PU is detected.
Hence, the SU doesnot interferes with the PU and
only access the idle channel when the PU is not
accessing the given channel. PUs operations are
given more priority as during the channel utilization
there must not be any interference experienced by

130
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
the PU (licensed user like-Satellite communication,
Medical Equipments).
Assume that there are common control units
(SU-BS) (Secondary Base station) to coordinate the
resource allocation for SUs. As soon as SUs
requests the SU-BS for a channel. The SU-BS
would determine “which channel to access” and
“which SU to be provided the channel?”.
Here in this paper the heterogeneous SUs are
considered:
 SU with minimum-rate guarantee (MRG)
[12],
 SU with minimum delay guarantee (MDG),
 SU with minimum rate and delay guarantee
(MRDG)
 SU with best effort service (BE).
We simply assume K1SUs with minimum-rate
guarantee (MRG), K2SUs with minimum delay
guarantee (MDG), K2SUs with minimum rate and
delay guarantee (MRDG) and K2SU with best effort
service (BE). We also assume N ideal sub-channels
in a given time slot. The number K1, K2, K3, K4 and
N can vary dynamically in different time slots.
In this paper to reduce the complexity, we have
considered k1, k2, k3, and k4 and relation defined for
our approach as, K1=K2=K3=K4. Hence we consider
that there are equal number of heterogeneous SUs
with MRG, MDG, MDRG and BE.
The following assumptions are made for optimal
resource allocation in the given network with
heterogeneous SUs and multiple PUs:
 Each sub-channels occupancy must be
represented as a binary index γk,n ϵ {0, 1}. γk, n
denotes the sub channel n accessed by SU k. If
the sub-channel is free γk,n=0 whereas if the
channel is busy γk,n=1.
 Total power constraint: Let PTOTAL denote the
total power available for the SUs. Pk,n denote the
transmit power for SU k in sub-channel n.
TOTALnk
kkkk
k
N
n
nk PP 
 

 
,
1 1
,
4321

(1)
(For example: -consider a license-free
frequency band
865-867 MHz
Use: low power RFID equipment or any
other low power wireless device or
equipment power
Max. Output power: 1Watt
Carrier Bandwidth: 200 KHz)
 Minimum rate guarantee: Let denote the
minimum rate requirement for SU k.
)2(
4321Kk kkkkkRR min 
Figure.1 Heterogeneous cognitive radio network with
multiple PU and SU with idle PU2 Channel (pr=0)
At the SU-BS, the presence of the primary user
is estimated periodically using the utility function
Channel Quality Indicator (CQI). CQI also
quantifies the utility that each channel offers to the
CR in terms of average information transfer capacity
and the interference. As soon as the CQI value
decreases beyond a threshold value, it denotes the
presence of PU and hence the SU handsover the
sub-channel back to the PU.
Let us consider

number of sub-channels in
this cell. Then the set of sub-carries can be
represented as, C={C1,….Cφ}. The Total bandwidth
of the cell is
B
and each sub contains the same
quality of sub-carriers. Hence, the bandwidth of
each sub-channel can be given by


.
CQI of any sub-channel is defined using the
Signal-to-Interference-Plus-Noise Ratio (SINR)
detected by the SU-BS during channel allocation
and during channel accessing.
Signal-to-Interference-Plus-Noise Ratio (SINR)
is a quantity used to represent the channel capacity
(or rate of information being transferred) in a
wireless communication system such as networks.
SINR is defined as the power of interested signal
divided by the sum of interference power (from all
the interfering signals) and the power of arbitrary
background noise.
SINR
NOISE
PceInterferen
PSignal

(4)
PInterference represents the interference power of
other signals in the network. Psignal represents the
power of the incoming signal of interest. If the
power of background noise term PInterference=0 then
the SINR is reduced to Signal to Signal-to-Noise
Ratio (SNR).

131
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
Channel Quality Indicator can be measured at SU-
BS by,
CQI
 
SINR1log2
(5)
By threshold value at the SU-BS, the value of
CQI estimated can be monitored and as soon the SU
request for a channel the CQI value of the channel is
estimated and the assigned to the SU only if the
CQI>threshold, else the channel is not assigned to
the SU. After the channel is assigned to SU in
between the data transmission SU-BS finds
degradation in CQI showing the presence of PU, the
SU hands-off the channel back to the PU as shown
in fig. 2.
Based on Dynamic Spectrum Allocation for
cognitive radio, when a primary user (PU) returns to
the band area which cognitive users are
using,cognitive radio have to pause the work and
switch to another band, this is called spectrum
handoff. The switching needs to be done as smooth
as possible because the sudden transition of working
spectrum may lead to decreased transmission
reliability or even transmission failure.
As shown in fig. 2, The spectrum handoff takes
place from PU to the secondary user and vice versa.
For example in the time slot t1 frequency range f1to
f2 is vaccant and hence the SU which requests the
spectrum slot is allocated the same by the SU-BS
and the reset of the frequency slots (e.g- f3 to f4)
are being accesed by some other PUs and hence
those chennels are unintrupted.Whereas in the time
slot t2to t3both the frequency slots f2to f3 and f3 to
f4is vacant and hence can be accesed by the SU.
(Illustration for section 3.1)
Figure.2 SU accessing the idle channel of PU through
SU-BS
Figure.3 Illustration of spectrum Handoff
For each incoming stream, the priority of
packets is determined depending on the service type
and queuing delay [8]. Then an objective function is
estimated for each stream by multiplying the priority
with channel gain. Then the streams can be sorted in
the descending order of this objective values, then
allocated to the respective type of SU. The stream
with highest objective function is allocated to the
MRDG, followed by MRG and MDG. The stream
with least objective function is assigned to BE.
Hence, as shown in fig. 2, value is assigned to
each SU.
Existing Algorithm:
Step1: Define n sub-channels SUs and
Step2: Sense the presence of PU
if yes
Hand-off
else
Step3: Free channel sensing and allocation to
SU
Step4: Sense the presence of PU
if yes
Hand-off
else
Step5: Power allocation to channel given to
SU
Step6: Sense the presence of PU
if yes
Hand-off
else
Step7: Data transmission takes place
Step 8: Sense the presence of PU
if yes
Hand-off
else
Step 9: Channel allocated to other SU
Step 10: Goto Step 2

132
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
5. Simulation Result
We use ns-2 simulation package to simulate the
proposed algorithm using the cognitive radio. At the
MAC layer, we use IEEE 802.11. The network field
considered for the simulation is 1000m X 1000m
over a flat region for 70 seconds of simulation time.
All nodes have the same transmission range of 250
meters. The simulated traffic is Constant Bit Rate
(CBR) with a packet size of 512 B. The simulation
settings and parameters are summarized in Table1.
Figure.4 Proposed algorithm

133
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
Figure.5 Screen shot of the out.nam for the proposed
approach
Table 1. Simulation Environment
No. of Nodes
20,40,60,80 and 100
Area Size
1000m X 1000m
Mac
IEEE802.11
Transmission Range
250m
Simulation Time
70 sec
Traffic Source
CBR
Packet Size
1000B
Sources
4
Attackers
2
Nodes Speed
5,10,15,20 and 25 m/s
As shown in fig.1, six cognitive radio networks
are considered with 6 SUs in each CRN. PU-6 is
idle as shown in fig.5 and has channel-6 as idle.
The Proposed DRAPBS-SRN Algorithm is
compared with DRA-CRN algorithm. As DRAPBS-
SRN employs dynamic resource allocation; it
outperforms the traditional approaches in cognitive
radio.
Scenario 1: Quality and robustness of
Channel allocation in DRAPBS
Fig.6 shows the delay DRAPBS-SRN in
comparison to DRA-CRN. With varying number of
idle channels present at the PU-BS, the proposed
approach is found to experience less delay since
each SU is prioritized in accessing the idle channel
without any interference occurring at the channel.
Figure.6 Channels Vs Delay
Figure 7. Speed Vs Delay
Since each SU senses the channel before
accessing the channel and also periodically checks
for the existence of the PU, the interference
introduced due to the sudden appearance of PU is
reduced. Prioritization of SU’s at the SU-BS leads to
idle channel allocation robustly. DSAPBS-SRN
experiences 17% less delay compared to DRA-CRN
approach. From fig. 7 it is also found that the delay
experienced, due to varying range of mobility of
nodes, is 16% less than DRA-CRN approach.
From fig. 8, the interference experienced by any
idle channel for communication (along with
tradeoff) is 9% less than DRA-SRN approach. From
fig. 9, it can be observed that at varying scenario of
mobility of nodes DSAPBS-SRN is found to be
more stable compared to DRA-CRN. Interference in
the proposed DRA-SRN approach is reduced due to
the Channel Quality Indicator can be measured at
SU-BS. Here channel is estimated and assigned to
the SU only if the CQI>threshold. At mobility of 10
m/s, 20m/s and 30m/s the interference experienced
by the network due to the utilization of DSAPBS-
SRN is negligible compared to DRA-CRN by 58%.

134
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
Figure.8 Channels Vs Interference
Figure.9 Speed Vs Interference
The inefficiency in resource allocation by the BS
(for a given frequency band used by the PU) leads to
interference at the channel leading to collision of
data sent using the defined channel and hence the
network experiences more drop of data packets in
comparison to an efficient network with robust
channel allocation and high delivery ratio. With total
power constraints (as calculated in eq. (1)),
minimum rate guarantee ( as calculated in eq. (2))
and fairness constraints (as calculated in eq. (3)), our
proposed approach employs an efficient channel
allocation procedure (as discussed in section 3),
experimentally it is observed that DRAPBS-SRN is
efficient in comparison to DRA-CRN.
Scenario 2: Network Performance
Evaluation
In Fig. 10 and 11, drop refers the average
number of packets dropped during the transmission.
Each channel in the CRN experiences drop at each
channel with respect to the number of idle channels
and mobility of nodes present in the network.
Figure.10 Channel Vs Drop
Figure.11 Speed Vs Mobility
As the number of idle channels in the network
increases from 20 to 30 channels, the packet drop
due to the employment of DRAPBS-SRN is found
to be negligible as compared to DRA-CRN by 47%.
Due to prioritization of SU at the SU-BS, each idle
channel in the network experiences less delay and
more throughput. From fig. 12, it can be observed
that DRAPBS-SRN is found to experience 19% of
less drop in data packets in comparison to DRA-
CRN.
Network throughput is the rate of successful
message delivery over a communication channel. It
is usually measured in bits per second (or data
packets per second or data packets per time slot).
Hence for an efficient and reliable network, the
throughput must be maximum and it should
experience less drop in the packet. As already
observed from fig. 10 and fig. 11, drop of data
packets is negligible in our proposed approach with
soft spectrum handoff. Similarly, from fig 12,
throughput in DRAPBS-SRN is 41% more efficient
with varying the number of idle channels from 20
channels to 100 channels in the given network.

135
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
Figure.12 Channel Vs Throughput
Figure.13 Speed Vs Throughput
As shown in Fig.13, with variations in the
mobility of nodes from 10 m/s to 50 m/s the
throughput of DRAPBS-CRN is found to be 3%
efficient in comparison to DRA-CRN, with an
efficient calculation of CQI.
6. Conclusion
However, when cognitive radio theories are
becoming mature gradually and start to turn an
interesting topic among researchers, lots of
problems confronted in practical application are still
challenging the realization of the Cognitive radio
network. Our focus has been laid on optimum
resource allocations and scheduling for primary user
(PU) and secondary user (SU) to maximize the
performance of the whole network by reducing the
interference at PU and also maximizing the
throughput and minimizing the risk of overlapping
the coverage of CRNs. As a future work we would
like to interface the proposed approach with
artificial intelligent and human like decision making
qualities, to enhance the performance during
handoff and channel accessing. Prioritization of SUs
can be made efficient by using a best optimization
approach with higher convergence.
Reference
[1] P. Cheng, A. Guo, Y. Xu, X. Wang, and X. Gao, “A
game approach for dynamic resource allocation in
cognitive radio networks”, Wireless Communications
and Signal Processing (WCSP), International
Conference on. IEEE, pp. 1-6, 2010.
[2] Y. Tachwali, B. F. Lo, I. F. Akyildiz and R. Agust,
“Multiuser Resource Allocation Optimization Using
Bandwidth-Power Product in Cognitive Radio
Networks”, IEEE journal on selected areas in
communications, Vol. 31, No. 3, pp. 451-463, 2013.
[3] R. Xie, F. R. Yu, and H. Ji, “Spectrum Sharing and
Resource Allocation for Energy-Efficient
Heterogeneous Cognitive Radio Networks with
Femtocells”, IEEE ICC 2012 - Cognitive Radio and
Networks Symposium, pp. 1661-1665, 2012.
[4] Y. Teng, Y. Liu and Y. Zhang, “An Energy Efficient
Resource Allocation in Cognitive Radio Networks
with Pairwise NBS Optimization for Multi-
Secondary Users”, IEEE Wireless communication, pp.
744-749, 2013.
[5] M. G. Adian and H. Aghaeinia, “Optimal resource
allocation for opportunistic spectrum access in
multiple-input multiple-output orthogonal frequency
division multiplexing based cooperative cognitive
radio networks”, IET Signal Process, Vol. 7, No. 7,
pp. 549-557, 2013.
[6] J. L. Xu, M. Chen and N. Wang, “Optimal
Cooperative Sensing and Resource Allocation in
Cognitive Radio Networks”, Wireless
Communications & Signal Processing (WCSP),
International Conference on. IEEE, pp. 1-6, 2013.
[7] M. Alfonso, Jeison, and L. Betancur Agudelo.
"Centralized Spectrum Broker and Spectrum Sensing
with Compressive Sensing techniques for resource
allocation in Cognitive Radio Networks."
Communications (LATINCOM), 2013 IEEE Latin-
America Conference on. IEEE, pp. 1-6, 2013.
[8] C. Gao, S. Chu, and X. Wang, “Distributed
Scheduling in MIMO Empowered Cognitive Radio
Ad Hoc Networks”, IEEE transactions on mobile
computing, Vol. 13, No. 7, pp. 1456-1468, 2014.
[9] D. Gozupek and F. Alagoz, “A Fair Scheduling
Model for Centralized Cognitive Radio Networks”,
arXiv preprint arXiv: 1309.2233, 2013.
[10] Y. Li and A. Nosratinia, “Hybrid Opportunistic
Scheduling in Cognitive Radio Networks”, IEEE
transactions on wireless communications, Vol. 11,
No. 1, 328-337, 2012.
[11] H. Li, X. Cheng, Keqiu, X. Xing and T. Jing,
“Utility-Based Cooperative Spectrum Sensing
Scheduling in Cognitive Radio Networks”,
INFOCOM, 2013 Proceedings IEEE. IEEE, pp. 165-
169, 2013.
[12] R. Xie, F. R. Yu and H. Ji, “Dynamic Resource
Allocation forheterogeneous Services in Cognitive
Radio Networks with Imperfect Channel Sensing”,
IEEE transactions on vehicular technology, Vol. 61,
No. 2, pp. 770-780, 2012.

136
International Journal of Intelligent Engineering and Systems, Vol.9, No.3, 2016 DOI: 10.22266/ijies2016.0930.13
[13]P. M. R. Santos, M. A. Kalil, O. Artemenko, A.
Lavrenko, and A. M. Thiel, “Self-Organized
Common Control Channel Design for Cognitive
Radio Ad Hoc Networks”, IEEE 24th International
Symposium on Personal, Indoor and Mobile Radio
Communications: Mobile and Wireless Networks, pp.
2419-2423, 2013.
[14]S. Byun, "TCP over scarce transmission opportunity
in cognitive radio networks", Computer Networks,
Vol. 103, pp. 101-114, 2016.
[15]K. Berbra, M. Barkat, F. Gini, M. Greco and P.
Stinco, "A fast spectrum sensing for CP-OFDM
cognitive radio based on adaptive thresholding",
Signal Processing, vol. 128, pp. 252-261, 2016.
[16]U. Premarathne, I. Khalil and M. Atiquzzaman,
"Trust based reliable transmissions strategies for
smart home energy consumption management in
cognitive radio based smart grid", Ad Hoc Networks,
vol. 41, pp. 15-29, 2016.
[17]V. Kumar, S. Minz and V. Kumar, "Performance
analysis of cognitive radio networks under spectrum
sharing using queuing approach", Computers &
Electrical Engineering, 2016.
[18]C. Charan and R. Paney, "Eigenvalue based double
threshold spectrum sensing under noise uncertainty
for cognitive radio", Optik - International Journal for
Light and Electron Optics, Vol. 127, No. 15, pp.
5968-5975, 2016.
[19]M. Khasawneh, S. Alrabaee, A. Agarwal, N. Goel
and M. Zaman, "Power trading in cognitive radio
networks", Journal of Network and Computer
Applications, Vol. 65, pp. 155-166, 2016.
[20]O. Ergul, A. Bicen and O. Akan, "Opportunistic
reliability for cognitive radio sensor actor networks
in smart grid", Ad Hoc Networks, Vol. 41, pp. 5-14,
2016.
[21]M. Moon and V. Gulhane, "Appropriate Channel
Selection for Data Transmission in Cognitive Radio
Networks", Procedia Computer Science, Vol. 78, pp.
838-844, 2016.
[22]V. Esmaeelzadeh, E. Hosseini, R. Berangi and O.
Akan, "Modeling of rate-based congestion control
schemes in cognitive radio sensor networks", Ad Hoc
Networks, Vol. 36, pp. 177-188, 2016.
[23]A. Sharifi, M. Sharifi and M. Niya, "Secure
cooperative spectrum sensing under primary user
emulation attack in cognitive radio networks: Attack-
aware threshold selection approach", AEU -
International Journal of Electronics and
Communications, Vol. 70, No. 1, pp. 95-104, 2016.
[24]G. Zhang, X. Li, M. Cui, G. Li and Y. Tan,
"Transceiver design for cognitive multi-user MIMO
multi-relay networks using imperfect CSI", AEU -
International Journal of Electronics and
Communications, Vol. 70, No. 5, pp. 544-557, 2016.
[25]S. Maity, S. Chatterjee and T. Acharya, "On optimal
fuzzy c-means clustering for energy efficient
cooperative spectrum sensing in cognitive radio
networks", Digital Signal Processing, Vol. 49, pp.
104-115, 2016.