ArticlePDF Available

An Evolutionary Scheme for Secondary Virtual Networks Mapping onto Cognitive Radio Substrate

Authors:

Abstract and Figures

The Fifth Generation (5G) of wireless communication is envisioned to comprise heterogeneous applications, different radio access technologies (RATs), and a large demand for mobile traffic. In this respect, Wireless Virtualization (WV) and Cognitive Radio (CR) are put forward as 5G enablers for providing additional spectrum resources through dynamic spectrum access (DSA) techniques, besides dealing with heterogeneity with no hardware modification. By empowering the synergy between CR and WV, we visualize an environment denoted as Cognitive Radio Virtual Networks Environment (CRVNE) that encompasses VWNs with different access priorities, called Primary Virtual Networks (PVNs) and Secondary Virtual Networks (SVNs) that may be deployed in an overlay manner. In this scenario, the SVNs users (SUs) access the resources opportunistically, which naturally raises challenges towards the SVN mapping. In this paper, we revisit our previous letter that models the interactions between PUs and SUs in a CRVNE and analyzes a proposed formulation for collision probability during the SVN mapping process. The current work is pioneer as it presents a comprehensive approach to the SVNs mapping problem; models, validates, and analyzes additional performance metrics such as SU blocking and SU dropping probabilities and joint utilization; formulates the SVNs mapping as a multiobjective problem; and proposes an evolutionary scheme based on Genetic Algorithms (GAs) to solve it. The results show that the proposed scheme outperforms the alternative method in terms of collision, SU dropping, SU blocking probabilities, and joint utilization under different primary and secondary loads.
This content is subject to copyright. Terms and conditions apply.
Research Article
An Evolutionary Scheme for Secondary Virtual Networks
Mapping onto Cognitive Radio Substrate
Andson Balieiro ,1Marcos Falcão ,2and Kelvin Dias 2
1Departamento de Computac¸˜
ao(DC),UniversidadeFederalRuraldePernambuco,R.ManueldeMedeiros,Recife,Pernambuco,Brazil
2Centro de Informatica (CIn), Universidade Federal de Pernambuco (UFPE), Av. Jornalista Anibal Fernandes, Recife,
Pernambuco, Brazil
Correspondence should be addressed to Kelvin Dias; kld@cin.ufpe.br
Received 4 December 2018; Revised 20 February 2019; Accepted 28 February 2019; Published 21 March 2019
Academic Editor: Davide Mattera
Copyright ©  Andson Balieiro et al. is is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
e Fih Generation (G) of wireless communication is envisioned to comprise heterogeneous applications, dierent radio access
technologies (RATs), and a large demand for mobile trac. In this respect, Wireless Virtualization (WV) and Cognitive Radio (CR)
are put forward as G enablers for providing additional spectrum resources through dynamic spectrum access (DSA) techniques,
besides dealing with heterogeneity with no hardware modication. By empowering the synergy between CR and WV, we visualize
an environment denoted as Cognitive Radio Virtual Networks Environment (CRVNE) that encompasses VWNs with dierent
access priorities, called Primary Virtual Networks (PVNs) and Secondary Virtual Networks (SVNs) that may be deployed in an
overlay manner. In this scenario, the SVNs users (SUs) access the resources opportunistically, which naturally raises challenges
towards the SVN mapping. In this paper, we revisit our previous letter that models the interactions between PUs and SUs in a
CRVNE and analyzes a prop osed formulation for collis ion probability during the SVN m apping process . e current work is pioneer
as it presents a comprehensive approach to the SVNs mapping problem; models, validates, and analyzes additional performance
metrics such as SU blocking and SU dropping probabilities and joint utilization; formulates the SVNs mapping as a multiobjective
problem; and proposes an evolutionary scheme based on Genetic Algorithms (GAs) to solve it. e results show that the proposed
scheme outperforms the alternative method in terms of collision, SU dropping, SU blocking probabilities, and joint utilization
under dierent primary and secondary loads.
1. Introduction
e Fih Generation (G) of wireless communication is
envisioned to comprise three service categories—enhanced
mobile broadband (eMBB), massive machine-type commu-
nication (mMTC), and ultra-reliable and low-latency com-
munication (URLLC)—that have dierent requirements in
terms of Quality of Service (QoS), Quality of Experience
(QoE), and security []. To tackle this heterogeneity, Wireless
Virtualization (WV) is put forward as a key technology
[], since virtual wireless networks (VWNs) with dierent
services may share the same wireless infrastructure.
Wireless Virtualization comprises both spectrum and
infrastructure sharing (e.g., base stations and access points).
Spectrum sharing focuses on the air interface virtualization,
i.e., how to schedule the spectral resources for VWNs. Some
works such as [, ] address spectrum sharing but consider
a strict resource allocation; i.e., the resources allocated to a
VWN are not shared with another during operation, which
may cause resource underutilization and revenue losses for
the Mobile Network Operator (MNO), the owner of the
physical resources. In this respect, opportunistic resource
sharing has been raised in [, ] as an alternative for solving
such problem as it enables multiple ows from dierent
VWNs to share common resources. However, little exibility
atthePHYandMAClayershasbeenachieved,sincethe
VWNs mapping is tied to specic radio access technologies
(RATs) []; that is, the RAT adopted by VWNs is limited
to that employed by the wireless infrastructure. In addition,
there is no dierence among VWNs in terms of access level to
Hindawi
Wireless Communications and Mobile Computing
Volume 2019, Article ID 1872765, 19 pages
https://doi.org/10.1155/2019/1872765
Wireless Communications and Mobile Computing
the resources (e.g., high and low priority VWNs). erefore,
these approaches do not address VWNs with dierent access
priorities or RATs on the same wireless infrastructure.
Due to the mobile trac increase, by  it is expected
thatthedemandshouldbetwohundredtimesgreaterthan
the current moment []. Hence, the scarce electromagnetic
spectrum must be made available and used eciently to allow
attending our future needs. Cognitive Radio (CR) has been
envisioned as an enabler for the deployment of G systems
as it focuses on the smart use of the spectrum through the
dynamic spectrum access (DSA) techniques [].
By combining Wireless Virtualization and CR, the deep-
est level of Wireless Virtualization can be achieved (spectrum
based virtualization []). is combination provides isolation
among VWNs at a low level [] and better resource uti-
lization through DSA and grants VWNs with dierent RATs
coexisting on the same wireless infrastructure with no hard-
ware modication. By empowering the synergy between CR
andWV,wevisualizeanenvironmentdenotedasCognitive
Radio Virtual Networks Environment (CRVNE), in which
VWNs with dierent access priorities to the resources, called
Primary Virtual Networks (PVNs) and Secondary Virtual
Networks (SVNs), may be deployed in an overlay manner
andsharethesamecognitiveradiosubstrate.eSVNsusers
known as secondary users (SUs) only have access to the
resources when the PVNs users, i.e., primary users (PUs), are
not using them, avoiding to cause harmful interference to the
PVN communication.
e PVNs are managed by Primary Service Providers
(PSPs) and could oer any application type supported by
wireless substrate such as multimedia and real time appli-
cations. Since the SVNs, which are managed by Secondary
Service Providers (SSPs), may suer preemption, they show
some limitations on the supported types of application. In
this respect, delay sensitive (e.g., URLLC services) or real
time services might not work as expected on this type
of network. On the other hand, best-eort services such
as PP downloading and web browsing could be oered.
is scenario raises new challenges, ranging from mapping
to operation, with this paper focusing on the rst (SVNs
mapping).
Mapping VWNs onto wireless substrate (i.e., reserv-
ing/allocating physical resources from MNO to the VWNs)
is a NP-hard problem [] and current approaches only
consider VWNs with homogeneous access priorities to the
resources [–] or do not take into account opportunistic
resource sharing [, ]. Our problem is more challenging as
it involves an environment composed of PVNs and SVNs
that share common resources. In order to provide reasonable
Quality of Service (QoS) to the SUs and avoid interference
to the PVN transmission, the SVN mapping must not only
consider the SVNs demand (e.g., bandwidth) but also the
PVN activity. Dierently from nonvirtualized scenarios in
which the resources are shared among individual users[]
and focus on channel selection/channel-user assignment
in the network operation phase, the current environment
addresses the resource allocation to multiple virtual wireless
networks (group of users) during the dimensioning stage.
is paper deals with the SVN mapping problem. It
revisits our previous letter [] that models the interac-
tions between PUs and SUs in a CRVNE and analyzes the
proposed formulation for collision probability during the
SVN mapping process. However, the current work is pioneer
as it (1) presents a comprehensive approach to the SVNs
mapping problem; (2) formulates, validates, and analyzes
additional performance metrics such as SU blocking and SU
dropping probabilities and joint utilization (to be used in
the SVNs mapping); (3) formulates the SVNs mapping as
amultiobjectiveproblem;and(4) proposes an evolutionary
scheme based on Genetic Algorithm (GA) to solve this
problem and evaluates it in terms of collision, SU dropping,
SU blocking probabilities, and joint utilization. Due to its
versatility, scalability, and computational simplicity, GA has
been widely adopted for solving optimization problems in
wireless networks [–] and solutions to reduce its conver-
gence time have been proposed []. In this work, we assume
that the GA is adopted for mapping SVNs, during the CRVNE
dimensioning, before the network becomes fully operational.
e results show that our scheme outperforms the alternative
method based on the First-Fit strategy.
is paper is organized as follows. Related works are
discussed in Section . Section  presents the Cognitive Radio
Virtual Network Environment and the challenges that emerge
in the SVNs mapping, highlighting the events/situations
that impair the primary and secondary communications and
that must be taken into account in the mapping process.
CRVNE modeling and formulations for the SU blocking,
SU dropping, and collision probabilities, as well as, joint
utilization are presented in Section . e model validation
and analysis on the SVN mappingare conducted in Section .
e formulation of the SVN mapping as a multiobjective
problem and a scheme based on GA to solve it are presented
in Section . e performance results are discussed in
Section . Section  concludes this paper and highlights the
future works.
2. Related Works
Studies have been proposed for Wireless Virtualization in
homogeneous [] and heterogeneous wireless networks []
or without specifying any network technology [, ], but
assuming that the resources allocated to a VWN cannot be
shared during operation. is restriction may cause resource
underutilization when the VWNs experience low trac load
periods, hindering new VWN deployments. Opportunistic
resource sharing has been raised in [, ] to solve the problem
as it considers the workload in a VWN to be the combination
of a permanent and a variable subworkload (following a
given probability). us, multiple ows from dierent VWNs
may share common resources, which may cause collision
and should be managed. Unlike our proposal combining CR
andWV,previousworksonlyreckonhomogeneousaccess
priority VWNs (e.g., no high and low priority VWNs) besides
providing little exibility at the PHY and MAC layers, since
the VWNs mapping is tied to specic RAT.
Proposals for opportunistic sharing in nonvirtualized
Cognitive Radio Networks (CRNs) have been widely
Wireless Communications and Mobile Computing
presented in literature; each CRN usually has its own
physical infrastructure and it is generally employed over
a primary network in a one-to-one relationship, such as
[, ]. In terms of resource allocation in nonvirtualized
CRNs, the focus is on channel selection/channel-user
assignment; the resources are shared among multiple parties
that are individual users [] and this process takes place
during the network operation. By applying virtualization to
CRNs, VWNs with dierent services/RATs/priorities can
bemappedontothesamesubstratenetworkandeasingthe
no one-to-one mapping restriction. us, channels that are
allocatedtodierentPVNscanbeusedbythesameSVN,
providing better resource utilization. In this environment,
the resources are also shared among multiple parties, but
unlike nonvirtualized scenarios, these are virtual wireless
networks (groups of users).
eworkdevelopedin[]isbasedonopportunistic
resource sharing. However, it does not include PUs or SUs.
In addition, there are other factors in CRVNE (apart from the
collision probability) that must be considered during the SVN
mapping process, such as the SU blocking and SU dropping
probabilities, both neglected in [].
Moreover, platforms for end-to-end network virtualiza-
tion that consider cognitive radio as a component have been
proposed in [, ]. Yet, the authors only provide a schematic
illustration of the interaction between the elements. In [],
ahypervisor-basedarchitectureforintra-andinter-node
resource scheduling in virtualization-based CR networks is
presented. Similarly to [, ], neither the VWN mapping
nor the evaluation in terms of SU blocking, SU dropping,
and collision probabilities and joint utilization is addressed.
Dierently, this work focuses on the multiobjective formu-
lation for the SVNs mapping problem and the design of an
evolutionary scheme to solve it. In addition, our scheme is
envisioned to act on the dimensioning stage, i.e., resource
allocation from infrastructure provider to VWNs.
An approach denoted as spectrum demand access as a
service is proposed in []. It dynamically oers spectrum
services to users and enables these to set up dynamic virtual
topologies to meet the needs of a specic application. e
authors adopt the dynamic spectrum allocation approach for
DSA, which does not distinguish between PUs and SUs. us,
each user has an exclusive spectrum band within a certain
time period, e.g., in the order of minutes. In addition, besides
considering only homogeneous requests, i.e., all virtual
topologies requesting the same spectrum amount (not always
thecaseinrealscenarios),theyfailtodrawonanyproposalin
theliteraturetomakeacomparativeevaluation.Unlike[],
our study adopts the opportunistic spectrum access (OSA)
approach for DSA, which dierentiates between PU and SU.
In OSA, the SUs dynamically search and access idle PUs
spectrum bands through spectrum sensing or databases. In
view of this, we take into account the existence of the PVNs
and the heterogeneous requests on the SVN mapping process.
Asitcanbeseen,workshaveproposedmechanisms
for wireless network virtualization by using strict resource
allocation, opportunistic resource sharing or spectrum
demand as a service. But, they fail with regard to exibil-
ity at MAC/PHY layer, resource eciency, or support to
the users/networks with heterogeneous priorities. Although
other studies have adopted cognitive radio in the wireless part
for providing end-to-end network slicing, they do not address
the virtual network mapping or schemes to map virtual
networks. Dierently, we combine CR and WV to tackle these
and dene a new virtual environment, in which the SVN
mapping is addressed and a GA-based scheme is proposed.
e next section describes this new environment and the
challenges that emerge in the SVNs networks mapping.
3. Cognitive Radio Virtual Network
Environment (CRVNE)
e Cognitive Radio Virtual Network Environment
(CRVNE) is made up of three wireless networks types:
substrate network, PVNs, and SVNs. e substrate networks
are managed by the MNO and consist of channels, spectrum
bands, base stations, and other features that compose wireless
environment []. e PVNs have higher access priority to
theresourcesthanSVNandareusuallymappedwithout
taking into account the SVNs existence [, ], hence, not
supporting the concept of opportunistic sharing.
Owing to the existence of low trac periods in the
PVNs, SVNs can be embedded through the opportunistic
access to the resources. ese networks have lower access
priority to the resources and will only use them when the
PVNs are idle. e adoption of SVNs can provide better
resource utilization (e.g., spectrum) and increase revenue for
the provider infrastructure, as more VWNs can be admitted.
e introduction of cognitive radio in Wireless Virtual-
ization allows new players to emerge in the business model.
Without CR, game is basically composed of two players:
the service provider (SP), which leases the virtual wireless
networks, programs them, and oers end-to-end services
to users, and the Mobile Network Operator (MNO), which
owns the network infrastructure (e.g., radio access networks,
backhaul, transmission networks, licensed spectrum and core
networks) []. When CR is considered, the SP is split into
two players: Primary Service Provider (PSP) and Secondary
Service Provider (SSP) []. e former oers its services
via PVNs, which have higher access priority. Hence, they
could oer any type of application supported by wireless
substrate such as voice service, multimedia, and real time
applications. e second manages SVNs that could oer best-
eort services such as PP download and web browsing.
Figure  illustrates the CRVNE.
e resources allocation in CRVNE raises several chal-
lenges; for instance, since the SVNs perform opportunis-
tic access, their mapping must take into account both
the demand requested by the SVN (e.g., the number of
users/requested bandwidth) and the primary activity. In this
respect, both protecting the primary communication from
the SU interference and meeting the SVN requirements are
relevant, requiring awareness of the resources usage pattern
by the PVNs and which situations/events could impair the
primary and/or secondary communications in the CRVNE.
e collision between PU and SU is one of these events. It
happens when a PU returns to a channel that is being used by
Wireless Communications and Mobile Computing
Cognitive Radio Virtual Networks Environment (CRVNE)
Channels(resources) Primary Virtual Networks Secondary Virtual Networks
Secondary Virtual Network Mapping Primary Virtual Network Mapping
Primary Service Provider (PSP)
Secondary Service Provider (SSP)
Mobile Network Operator (MNO)
(
bl
36.136.2
36.336.N
06.106.2
06.306.L
F : Cognitive Radio Virtual Networks Environment (CRVNE).
a SU. PU collides with SU and both communications suer
degradation. Moreover, as PU has higher priority of access to
the resources than SU, the SU has to vacate the channel and
ndanotheravailablechanneltoresumeitscommunication.
AcollisionbetweenPUandSUisshowninFigure.Inthis
example, a PVN was mapped onto channels (Ch) , , and
 and shares these channels with the SVN. Ch  is occupied
by PU (i.e., the channel is in ON state). us, the channel 
cannot be used by users from the SVN at this moment. SU is
performing its communication in the channel Ch , which is
denoted as OFF state due to the fact that the PU is not using it.
At this moment, when PU arrives at PVN and accesses Ch ,
which is occupied by SU, a collision occurs and the primary
communication suers interference from the secondary one
and vice-versa. Consequently, SU has to vacate channel  and
ndanotheravailablechanneltoresumeitscommunication
(e.g., Ch ). Avoiding or keeping this interference below a
thresholdisafeaturethatmustbetakenintoaccountinthe
SVNs mapping.
When a SU tries to access the SVN and there are
not enough resources to its communication, it is blocked/
rejected, which damages the secondary communication.
us, admitting as many SU as possible dimensioned for
each SVN, i.e., reducing the SUs blocking probability is an
important goal in the SVNs mapping. A situation in which a
SU is blocked due to resource scarcity is shown in Figure .
PVN shares channels  and  with the SVN. Users PU and
PU are occupying channels  and , respectively. Here, there
are two secondary users (SU and SU) arriving at SVN,
but there is only one channel available (Ch) to be used for
communication. us, only one SU can be admitted while the
other is rejected.
A third situation that can aect the quality of service of
the secondary communication is when the SU is dropped
from SVN, due to a returned PU to the channel occupied by
an SU while an extra channel is not available in its current
SVN. An example where the SU dropping happens is depicted
in Figure . ere are two primary users (PU and PU) in
the PVN, which are using channels  and , represented as
channels ON (in ON state). Ch  is not being used by PVN,
which is represented as a channel OFF (in OFF state), but
theSUisusingitinanopportunisticway.Whenanew
PU arrives at PVN, SU is preempted from channel  and
it searches for another one to resume its communication.
However, as there is no available channel in its SVN, SU is
dropped from the SVN. is event forces the termination of
the secondary communication prematurely.
In the next section, we model the interactions between
PUs and SUs in CRVNE by using queuing theory and
formulate the probabilities for the collision (Section .), SU
blocking (Section .), and SU dropping (Section .) events,
as well as the resource joint utilization (Section .), which
are considered in the SVNs mapping problem.
4. CRVNE Model
In a CRVNE, the PSP requests the creation of and manages
Primary Virtual Networks (PVNs). Given that the substrate
network is composed of channels and that the mapping
algorithm divides the resources between the PVNs according
to percentage 𝑗,with = 1,2,3,...,,0≤
𝑗≤1
and 𝐿
𝑗=1 𝑗=1, then for each PVN is allocated |𝑗|=
𝑗or  𝑗channels, where 𝑗means the set of
channels allocated to PVN ,and are the ceil and oor
functions, respectively.
We assume that the PUs arrive at channel (𝑖)ofvirtual
network ,with𝑖∈
𝑗, following a Poisson process with
arrival rate 𝑃𝑈,𝑖,𝑗, and the user holding time is given by an
exponential distribution with mean 1/𝑃𝑈,𝑖,𝑗. In addition, we
consider each channel has capacity to satisfy one PU.
Wireless Communications and Mobile Computing
Channel ON (2)
Secondary Virtual Network Primary Virtual Network
SVN Mapping PVN Mapping
Wireless Substrate Network
Ch2
Ch1
Ch3
Channel OFF (1, 3)
Arrival of a PU1
PU2
SU1
F : An example of collision when the PU arrives in a channel occupied by SU.
Channel ON (2, 3)
Secondary Virtual Network Primary Virtual Network
SVN Mapping PVN Mapping
Wireless Substrate Network
Ch2
Ch1
Ch3
Channel OFF (1)
SU2
SU1
PU2
PU3
F : An example of SU blocking when there is no enough available resource in the SVN to admit a new secondary user.
Channel ON (2,3)
Secondary Virtual Network Primary Virtual Network
SVN Mapping PVN Mapping
Wireless Substrate Network
Ch2
Ch1
Ch3
Channel OFF (1)
Arrival of a PU
PU2
PU3
SU1
F : An example of SU dropping when the PU returns and there is no available resource in the SVN for resuming the SU communication.
Wireless Communications and Mobile Computing
Given that channels were allocated to the PVN , i.e.,
|𝑗|=,thetotalPUarrivalratemaybeobtainedby().
𝑃𝑈,𝑗 =
𝐶𝑖∈𝑄𝑗𝑃𝑈,𝑖,𝑗 ()
e SVNs provide their services by opportunistically
using the resources. In CRVNE, there is no one-to-one
mapping between PVNs and SVNs as in nonvirtualized CRN
[, ]. us, channels allocated to dierent PVNs can be
allocated to the same SVN.
We consider that the SSPs requests SVNs to be mapped
onto cognitive radio substrate. In each SVN (𝑙), with =
1,2,...,, the SUs arrival follows a Poisson distribution with
rate 𝑆𝑈,𝑙 users per second (users/s) and the SU holding time
is exponentially distributed with mean 1/𝑆𝑈,𝑙 seconds [].
Similarly to PVN, we consider that the bandwidth requested
by each SU can be satised by one channel. Hence, the average
number of SUs in the 𝑙and the amount of resources
requested by SUs considering each channel with bandwidth
bps are calculated by () and (), respectively.
𝑙=𝑆𝑈,𝑙 1
𝑆𝑈,𝑙 ()
𝑟𝑒𝑞,𝑙 =𝑆𝑈,𝑙 1
𝑆𝑈,𝑙 ∗=𝑙∗ ()
Given that the mapping of the SVN onto CR substrate
adopted a set of channels, 𝑙={
1,2,...,𝑁},with
𝑙⊂
𝑗𝑗,and𝑙∩
𝑢=ø, for all  =,where
,=1,2,...,are SVN identiers, and that the PU service
rate is homogeneous and denoted as 𝑃𝑈,𝑙, i.e., 𝑃𝑈,𝑙 =𝑃𝑈,𝑖,𝑙 =
𝑃𝑈,𝑑,𝑙,∀𝑖,𝑑∈𝑙, the interaction between PVN and SVN
may be modeled as an M/M/N/N queue with preemptive-
priority service, where PUs and SUs compete for channels
[]. In this system, resources are limited and no queue is
allowedtoform.Moreover,aSUcanbeforciblyterminatedif
a PU arrival occurs when there is no other available channel
in the 𝑙.
In our model, each state (,),with0≤,≤and
0≤+≤, means that there are PUs and SUs in
the system (𝑙). e states (,)denote a full system,
where all resources are being used by PUs or SUs. Specically,
when −1, these states model situations where the SU is
droppedfromSVNduetoPUarrivalandthereisnoavailable
channel to resume its communication. Figure  presents the
state transition diagram of our CRVNE model. Horizontal
ows to right (le) mean PU arrival (departure) and vertical
ows to top (down) represent SU arrival (departure).
4.1. Formulation for Collision Probability. In the SVNs map-
ping, it is important to consider other factors apart from the
demand for these networks. As the channels adopted by the
SVNs are shared with the PVNs, it is necessary to ensure
minimum interference to PVNs, which can be dened as a
threshold, based on the service level agreement (SLA) from
the PVNs, for example. A collision (between PU and SU)
happens when a PU returns to a channel that is being used by
SU, damaging both primary and secondary communications.
It is noted that the PU arrival in the 𝑙certainly leads to
a collision with SU when the 𝑙is full and there is at least
one SU active, i.e., for states (,),with>0.us,the
probability sum of these states (see ()) bounds the collision
probability.
𝑖𝑛𝑓,𝑙 =𝑁
𝑗=1, ()
When the 𝑙is not full and at least one active SU is
present, the PU arrival does not necessarily lead to a collision,
since the PU may have returned to a channel that is not being
occupied by SU. States (,),with>0and (+)< ,model
this situation and the probabilities sum of these states (𝑗,
in ()) denotes the probability of such event taking place.
𝑙=(𝑖+𝑗)<𝑁
𝑖=0,𝑗=1, ()
In order to calculate the collision probability, it is nec-
essarytoknowwhichchannelsthePUsandSUsareusing.
However, this specic information is not available in the
mapping process, since it just deals with the allocation of a set
of channels to each VWN. Generally, this kind of information
may be obtained during the network operation, because it
involves channel-user allocation, which is not represented by
this model.
For states (,0),with≥0, the PU arrival does not trigger
acollisionbecausetherearenoSUsinthe𝑙.Inthis
respect, a collision might only take place in the previous two
cases and we may use () to estimate its probability. It uses
() as an inferior bound and () multiplied by a factor as an
increment. e factor expresses how likely a collision may
occur when the 𝑙is in the states (,),with>0,and
(+)<.
𝑙=
𝑖𝑛𝑓,𝑙 +∗𝑙()
e factor is given by (), which represents the average
probability that the PU returns to the channel while the SU
is using it. So, for each channel allocated to 𝑙,the
probability that a PU returns to the channel during the SU
communication (,) is computed, i.e., the probability
of the OFF time (time in which the PU is absent) being lower
than the SU service time.
=𝑁
𝑖=1 𝑏𝑎𝑐𝑘,𝑖
()
Given that the PU arrival rate in a channel is modeled
as a Poisson process with rate 𝑃𝑈,𝑖 and that the PU service
time follows an exponential distribution with rate 𝑃𝑈,𝑖,the
channelsmeanOFFperiodisgivenby().
𝑂𝐹𝐹𝑖 =1
𝑃𝑈,𝑖 1
𝑃𝑈,𝑖 ()
Wireless Communications and Mobile Computing
0,N
0,N-1
0,1
0,0 N,0N-1,0
1,0
1,1
1,N-1
N-1,1
2,0
2,1
0,N-2 1,N-2 2,N-2
0,2 1,2 2,2
N-2,1
N-2,0
N-2,2
PU,l
PU,l
PU,l
PU,l
PU,l
PU,l
PU,l
PU,l
PU,l
SU,l
SU,l
SU,l
SU,l
SU,l
SU,l
SU,l
(N−2)SU,l (N−2)SU,l (N− 2)SU,l
(N −1)SU,l
(N −1)SU,l
3SU,l 3SU,l
3PU,l
SU,l
SU,l 3SU,l
SU,l
2SU,l 2SU,l
2PU,l
SU,l 2SU,l SU,l 2SU,l
2PU,l
2PU,l
(N − 2)PU,l
(N−2)PU,l
(N−2)PU,l
(N−1)PU,l
(N − 1)PU,l
NSU,l
PU,l PU,l PU,l
PU,l 3PU,l
3PU,l
2PU,l
PU,l PU,l PU,l PU,l
PU,l
PU,l
PU,l
PU,l
PU,l
PU,l PU,l PU,l PU,l PU,l PU,l
SU,l
SU,l
SU,l SU,l SU,l SU,l SU,l
PU,l NPU,l
SU,l SU,l
SU,l SU,l
F : State transition diagram of the CRVNE model.
Assuming that the SU service time is exponentially
distributed with rate 𝑆𝑈,the,is given by (), where
𝑂𝐹𝐹𝑖 =1/𝑂𝐹𝐹𝑖.eprooffor()isgivenin[].
𝑏𝑎𝑐𝑘,𝑖 =𝑂𝐹𝐹𝑖
𝑂𝐹𝐹𝑖 +𝑆𝑈 ()
4.2. Formulation for SU Blocking Probability. As well as
protecting the PUs, the SVNs mapping process must provide
reasonable quality of service for the SUs. us, it must provide
high level of SU admission, i.e., low SU blocking probability.
As stated in Section , the SU blocking occurs in the SVN
when all channels are busy during a SU arrival. Hence, the SU
blocking probability on the 𝑙isgivenby(),whichisthe
probability sum of the states that represent the full system.
𝑆𝑈,𝑙 =𝑁
𝑖=0(−,)()
4.3. Formulation for Joint Utilization. As well as seeking to
admit as many users as possible, providing better resource
utilization is also a goal of both SVN mapping and cogni-
tive radio technology through opportunistic access to the
resources, besides increasing the revenue for the MNO. Given
the set of channels 𝑙=1,2,...,𝑁used in the 𝑙
mapping, the joint utilization (primary and secondary usage)
of these channels is given by (), which is the ratio of the
average number of channels occupied by PUs or SUs to the
number of channels allocated to 𝑙.
𝑙=𝑙+𝑙
()
where 𝑙is the average number of PU in channels
shared with 𝑙,calculatedby()and𝑙is the average
number of SUs, which can be obtained by ().
𝑙=𝑁
𝑖=0
𝑁−𝑖
𝑗=0, ()
𝑙=𝑁
𝑗=0
𝑁−𝑗
𝑖=0 , ()
4.4. Formulation for SU Dropping Probability. e SU block-
ing probability is the rejection level of new SUs in the
SVN. Once the SUs are admitted, some events triggered by
PU activity can aect their QoS. Among these events, the
SU dropping happens as a result of the PU return to the
channel occupied by SU and inexistence of available channel
in its SVN. e SU dropping causes great degradation to
secondary communication, as the SU has its communication
abruptly terminated. With a full network, each collision
between PU and SU leads to a SU preemption. us, the SU
preemption rate from 𝑙is numerically equal to the rate
of PUs that suered collision, which is given by (), where
Wireless Communications and Mobile Computing
the summation considers the states that represent the full
network and there is at least one active SU.
𝑆𝑈,𝑙 =𝑃𝑈,𝑙 𝑁
𝑗=1, ()
By dividing the rate of SU preempted from 𝑙by the
rate of admitted SUs, the SU dropping probability is given by
().
𝑆𝑈,𝑙 =𝑆𝑈,𝑙
1𝑆𝑈,𝑙∗𝑆𝑈,𝑙 ()
e next section presents the model validation and ana-
lyzes the behavior of the collision, SU blocking, SU dropping
probabilities, and the joint utilization regarding a mapping
scenario with dierent PU and SU loads.
5. Model Validation
A Matlab simulation model was used to validate the work
presented in Section . We considered a scenario with two
channels, which were shared by a SVN and PUs (from PVNs).
e PU and SU service rates were dened as  and .
(users/s), respectively. PU arrival rate (in users/s) in each
channel was varied (from . to .) in order to analyze the
model behavior when in dierent PU loads. Similarly, the
model was evaluated considering dierent SU arrival rates
(ranging from . to . users/s), i.e., under dierent SU loads.
We performed  simulation instances for each evaluated
point. e simulation time was , seconds and the
average results are presented considering a % condence
level,whichwereobtainedbytheBootstrapmethod[],
with ‘resample’ size and number of (re)samplings equal to 
and,respectively.InFigures,,,,and,‘Model’and
‘Sim’ mean results obtained through the analytical model and
simulation, respectively.
e results for the SU blocking probability are presented
in Figure , where the analytical model followed the simula-
tion. It is noted that initially the SU blocking probability tends
to decrease when the PU arrival rate increases. is behavior
occursuntilacertainPUarrivalratevalue.Beyondthis,the
PU arrival rates increase enables more SU blocking events.
ConsideringanSUarrivalrateequalto.,wenotethatthe
SU blocking probability changes its behavior (decreasing to
increasing) when PU arrival rate is about .. Similarly, when
we consider SU arrival rates equal to . and ., the change
happens in the points where the PU arrival rate is about .
and.,respectively.Inthesepoints,weobservethatthereisa
rangewherethePUarrivalrateincreasedoesnotnecessarily
lead to a SU blocking probability boost. is initial decrease
occurs due to cases where the SU is dropped from the SVN,
andthePUonlyusesthechannelsforashorttime,releasingit
aerwards. erefore, a new SU can be accepted in the SVN,
which might experience the same situation.
For the cases where the SU arrival rate is equal to ., .,
., or ., it was observed that the SU blocking probability
decreases when the PU arrival rate increases in the interval
[0.1,0.9]. For this region, as the SU arrival rate is higher than
PU arrival, i.e., the SU interarrival time is shorter than the
PU inter-arrival time, the SU can access the channel when
it is idle (in OFF state). But, as its service time is larger (on
average) than the channel’s OFF period, it is preempted due
to the PU return.
In turn, a PU arrival raise causes the SU blocking increase
and the consequent opportunistic access reduction. is is
observed when SU arrival rate is equal to ., ., or ..
Moreover, it is shown in Figure  that when SU arrival rate
increases, the SU blocking increases as well because the same
resource amount is considered to satisfy an increasing SU
demand.
Figure  depicts the joint utilization results for both the
model and simulation, which present the same behavior. e
joint utilization is similar to the SU blocking probability.
Initially, it decreases and later it starts to increase with the
PU arrival rate escalation. In order to analyze such behavior,
it is highlighted that the joint utilization is obtained by taking
into account the primary and secondary utilization. As the
primary user has higher access priority to the channels, the
resource utilization provided by the primary communica-
tion is not inuenced from the secondary communication.
erefore, it increases when the PU arrival rate increases, as
showninFigure.Onthecontrary,thesecondaryutilization
is impacted by the PU communication; hence, when the
PU arrival rate increases, the secondary utilization decreases
once there are fewer chances for opportunistic access to the
channels.
According to Figure , the secondary utilization is depen-
dent of the PU arrival rate. In some cases, the secondary
utilization reduction is compensated by the primary utiliza-
tion, leading to a joint utilization increase. is can be noted
inthecaseswheretheSUarrivalrateisequalto.,.,
or ., for example, (see Figure ), whereas in other cases,
the primary utilization cannot compensate the secondary
utilization reduction. Hence, the joint utilization tends to
decrease, which is noted in the cases where the SU arrival
rate is higher than . as shown in Figure . Moreover, in
Figure , the joint utilization increases together with the SU
arrival rate, which is expected since the SU load is higher,
leading to a secondary utilization increase (see Figure ).
e SU dropping probability results are depicted in
Figure , which illustrates that both model and simulation
present a similar behavior. We note that the SU dropping
increases when the PU arrival rate increases. is is expected,
since a higher PU arrival rate implies a shorter PU interarrival
time, which boosts the chances of an admitted SU to be
preempted. Moreover, it is observed that the SU dropping
probabilityalsoincreasesduetotheSUarrivalrateraise(see
Figure ), while resource amount is kept the same.
e results for the collision probability are presented
in Figure , and although () expresses an approximation
for such metric, the results obtained are similar to those
ones from our simulation. In addition, the gure shows that
when the PU arrival rate increases, the collision probability
decreases. At rst, it seems as an odd conclusion, once when
thePUarrivalrateincreases,thePUloadalsoincreasesand
therefore we would expect that the collision probability would
also increase. is is true when we are addressing collision in
Wireless Communications and Mobile Computing
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
PU Arrival Rate in each Channel
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
SU Blocking Probability
(PU Service Rate = 1 user/s; SU Service Rate = 0.1 users; Number of Channels =2)
Model (35=0.2)
Sim (35=0.2)
Model (35=0.4)
Sim (35=0.4)
Model (35=0.6)
Sim (35=0.6)
Model (35=1.0)
Sim (35=1.0)
Model (35=1.5)
Sim (35=1.5)
Model (35=2.0)
Sim (35=2.0)
Model (35=2.5)
Sim (35=2.5)
F : Results obtained by model and simulation in terms of SU blocking probability.
0.5
0.55
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
Joint Utilization
(PU Service Rate = 1 user/s; SU Service Rate = 0.1 users; Number of Channels =2)
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.90.1
PU Arrival Rate in each Channel
Model (35=0.2)
Sim (35=0.2)
Model (35=0.4)
Sim (35=0.4)
Model (35=0.6)
Sim (35=0.6)
Model (35=1.0)
Sim (35=1.0)
Model (35=1.5)
Sim (35=1.5)
Model (35=2.0)
Sim (35=2.0)
Model (35=2.5)
Sim (35=2.5)
F:Resultsobtainedbymodelandsimulationintermsofjoint
utilization.
a media access control situation, where the users compete for
channel access, and a higher user arrival rate leads to a greater
collision amount. However, for a collision between PU and
SU to occur in a CRVNE (Section ), the following condition
has to be satised: the SU is using the channel to which the PU
will return during this period. From this condition, we note
that SU needs access opportunities for a collision to happen.
If these are reduced, the collision number also tends to be
reduced. erefore, in Figure , when the PU arrival rate
increases, implying less opportunity for the SUs, the collision
probability decreases. Moreover, as the SU service time is
higher(onaverage)thanthechannelsOFFtime(thePUisnot
currently using it), when the SU gets the access to the channel,
it is very likely that the SU will still be using the channel when
the PU returns.
6. Formulation Problem and Proposed Scheme
is section presents the formulation of the SVN mapping
asamultiobjectiveproblem,whichtakesintoaccountthe
objectives discussed in the previous section. Moreover, it
proposes a scheme based on Genetic Algorithms to solve
the problem, detailing its structure, parameters values, and
operation.
6.1. Formulation of the SVN Mapping Optimization Problem.
Asshowninprevioussections,severalobjectivesmustbe
considered in the SVN mapping process. e following
optimization problem (see ()) is formulated as follows:
given a set of SVNrequestsandthechannelusagepattern
for the primary networks, to perform the SVNs mapping,
i.e., to determine the set of channels to be allocated to each
SVN, in order to minimize the average collision (), SU
blocking (𝑆𝑈 ), and SU dropping (𝑆𝑈)probabilitiesand
maximize the average joint utilization (). Two constraints
must be satised: the resource amount allocated to each SVN
cannot be less than the requested demand and a common
channel cannot be allocated to dierent SVNs. is last
 Wireless Communications and Mobile Computing
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Utilization
(PU Service Rate = 1 user/s; SU Service Rate = 0.1 user/s; Number of Channels =2)
Primary
Secondary (txCHSU=0.2)
Secondary (txCHSU=0.4)
Secondary (txCHSU=0.6)
Secondary (txCHSU=1)
Secondary (txCHSU=1.5)
Secondary (txCHSU=2)
Secondary (txCHSU=2.5)
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.90.1
PU Arrival Rate in each Channel(user/s)
F : Primary and secondary utilization obtained by model under dierent loads of PU and SU.
(PU Service Rate = 1 user/s; SU Service Rate = 0.1 users; Number of Channels =2)
0.3
0.4
0.5
0.6
0.7
0.8
0.9
SU Dropping Probability
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.90.1
PU Arrival Rate in each Channel
Model (35=0.2)
Sim (35=0.2)
Model (35=0.4)
Sim (35=0.4)
Model (35=0.6)
Sim (35=0.6)
Model (35=1.0)
Sim (35=1.0)
Model (35=1.5)
Sim (35=1.5)
Model (35=2.0)
Sim (35=2.0)
Model (35=2.5)
Sim (35=2.5)
F:ResultsobtainedbymodelandsimulationintermsofSU
dropping probability.
constraint aims to provide interslice isolation among SVNs
[]. Formally,
 ,𝑆𝑈,𝑆𝑈
  
 :𝑙𝑟𝑒𝑞,𝑙, =1,2,...,
𝑙∩𝑢=ø,
 =, ,=1,2,...,
()
where 𝑙and 𝑙are the amount of resources and the
set of channels allocated to 𝑙, respectively. A challenging
problem arises if the SVN mapping is focused on a specic
objective: it may deteriorate other SVN’s performance goals.
To mitigate such eect, our evolutionary scheme (based on
genetic algorithms) is proposed in the next subsections.
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
PU Arrival Rate in each Channel
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Collision Probability
(PU Service Rate = 1 user/s; SU Service Rate = 0.1 users; Number of Channels =2)
Model (35=0.2)
Sim (35=0.2)
Model (35=0.4)
Sim (35=0.4)
Model (35=0.6)
Sim (35=0.6)
Model (35=1.0)
Sim (35=1.0)
Model (35=1.5)
Sim (35=1.5)
Model (35=2.0)
Sim (35=2.0)
Model (35=2.5)
Sim (35=2.5)
F : Results obtained by model and simulation in terms of
collision probability.
6.2. Chromosome Structure and Fitness Function. GA is a
search algorithm based on the principles of natural selection.
It relies upon evolving a set of solutions, represented by the
so-called chromosomes. Eventually, through the GA opera-
tors (selection, crossover, and mutation) a good solution will
be found by combining dierent solutions [].
SomeGAcharacteristicssuchasversatility,scalability,
andcomputationalsimplicityaresuitablefortheSVNmap-
ping problem. GA handles many solutions simultaneously
at each interaction and evolves them to achieve better
solutions. us, many possible mappings are evaluated at
each interaction. In general, GA is exible enough to tackle
manyobjectivesorconstraintsandcanbecombinedwith
classicalapproaches[]todealwiththiskindofproblem.In
addition, GA has been widely adopted for solving optimiza-
tion problems in wireless networks [–] and solutions to
reduce its convergence time have been proposed []. In this
Wireless Communications and Mobile Computing 
1
Channel 1
0 0 …
1
K bits
Channel 2 Channel 2 Channel k
F : Chromosome Structure.
work,itisassumedthattheGAisadoptedfortheCRVNE
dimensioning phase,i.e., before the network becomes fully
operational.
In our GA-based proposed scheme, the individual or
chromosome 𝑗
𝑙is represented by a sequence of bits (see
Figure ), where isthenumberofavailablechannels
thatcanbeusedtomeettherequested𝑙,and=
1,2,3,...,,whereis the population size. Each gene (bit) in
the chromosome refers to an available channel for allocation.
Ifthegenevalueisequalto,thentherespectivechannelis
selected for the SVN mapping. Otherwise, the channel is not
selected.IntheindividualshowninFigure,channelsand
k are selected for mapping the requested SVN. Our proposal
considers the intrinsic parallelism of the GA when seeking to
nd an optimal or suboptimal mapping for each SVN.
It should be noted that there is an equivalence between
an individual in the GA and a set of channels. us, given
an individual 𝑗
𝑙,itispossibletoknowthesetofchannels
represented by it (𝑗
𝑙) and vice-versa. () describes how to
obtain the set of channels represented by an individual 𝑗
𝑙in
the GA, where 𝑙isthesetofavailablechannelsformapping
the SVN and 𝑖is the channel associated with the gene 𝑖.
e gene (bit) 𝑖composes the chromosome 𝑗
𝑙.
𝑗
𝑙=𝑖∈𝑙|𝑖∈𝑗
𝑙and 𝑖=1()
In our scheme, the SVNs are mapped sequentially, so a
population is created for each SVN and evolves to obtain a
nal solution represented by a set of channels that can be
allocated to the respective SVN.
To evaluate the individuals (solutions), we dened the
tness function given in (). In order to handle the multiob-
jective problem of SVNs mapping and reduce its complexity,
we adopted the classical approaches such as weighted sum,
-constraint, and programming method []. In this respect,
two expressions dened our tness function. e weighted
summethodwasadoptedtocomposetherstpartofthe
tness function, where the collision and SU dropping prob-
abilities were taken into account, which are mainly related
to primary and secondary communications, respectively. In
addition, the -constraint method was also used in this
part, where 𝑏𝑙𝑜𝑐𝑘 and 𝑢𝑡𝑖𝑙 were the constraints dened to
SU blocking probability and joint utilization, respectively.
us,therstpartofourtnessfunctionaimstoreduce
thecollisionandSUdroppingprobabilitieswhilekeeping
the joint utilization and SU blocking within certain limits.
e -constraint method should have constraints dened
in a feasible region, otherwise no solution will be found.
However, it is hard to do so for all possible SVN mapping
scenarios, since the constraint values may or may not be in
a feasible region. erefore, the second part of our tness
function adopts the goal programming method along with
the expression given by weighted sum (𝑚𝑎𝑖𝑛).
us, target values were dened for SU blocking prob-
ability and joint utilization, which are also represented by
𝑏𝑙𝑜𝑐𝑘 and 𝑢𝑡𝑖𝑙, respectively. e second expression (in ())
adopts the average between 𝑏𝑙𝑜𝑐𝑘 and 𝑢𝑡𝑖𝑙 as a penalization
factor for 𝑚𝑎𝑖𝑛.𝑏𝑙𝑜𝑐𝑘 is the relative dierence between the
SU blocking probability and its target value. Similarly, the
relative dierence between the joint utilization and its target
value is 𝑢𝑡𝑖𝑙. ey are expressed in () and (), respectively.
In our approach, 𝑏𝑙𝑜𝑐𝑘 and 𝑢𝑡𝑖𝑙 were set up as . and .,
respectively.
𝑗
𝑙=
𝑚𝑎𝑖𝑛,
𝑆𝑈,𝑙 ≤
𝑏𝑙𝑜𝑐𝑘  𝑙≥
𝑢𝑡𝑖𝑙
1𝑏𝑙𝑜𝑐𝑘 +𝑢𝑡𝑖𝑙
2∗
𝑚𝑎𝑖𝑛,()
where 𝑚𝑎𝑖𝑛 is dened in ().
𝑚𝑎𝑖𝑛 =100∗1−𝑙𝑗
𝑙+1𝑆𝑈,𝑙 𝑗
𝑙 ()
𝑏𝑙𝑜𝑐𝑘 𝑗
𝑙=
𝑆𝑈,𝑙 −𝑏𝑙𝑜𝑐𝑘
1𝑏𝑙𝑜𝑐𝑘,
𝑆𝑈𝑙>
𝑏𝑙𝑜𝑐𝑘
0,  ()
𝑢𝑡𝑖𝑙 𝑗
𝑙=
𝑙
𝑢𝑡𝑖𝑙 ,
𝑙<
𝑢𝑡𝑖𝑙
0,  ()
It is noted that () and () have values dierent from
zero when the collision probability or joint utilization values
provided by a given mapping (individual) are worse than the
targets. In addition, the greater the collision and utilization
are from the target values, the greater the penalty in the
individual’s tness of the individual will be. Both expressions
in () aim to achieve a good tradeo between the objectives
of the optimization problem presented in Section ..
Although there are GA approaches for solving multiob-
jective problems such as Multiobjective Generic Algorithms
(MOGA) and Nondominated Sorting Genetic Algorithm
(NSGA) [], for instance, in our scheme (as shown in
()), the classical methods were used to deal with multiple
 Wireless Communications and Mobile Computing
64
10 20 30 40 50 601
Test Ca se
114
115
116
117
118
119
120
Average Fitness
F : Test case results for GA.
objectives. With this approach, the complexity of the multi-
objectiveproblemisreducedandjusthandledinthetness
function. Moreover, it does not require any changes to the
basic GA mechanism. In addition, the literature includes
studies that have successfully used this approach in a multi-
objective optimization [, ]. e next section presents the
GA operators and parameters adopted in our evolutionary
scheme.
6.3. Genetic Operators and Parameters. We a d o p t e d t h e
roulette wheel as the selection operator, which involves
selecting individuals for the crossover process based on their
tness values. It simulates the natural selection mechanism,
which acts on biological species []. Hence, individuals
with the highest tness values are best suited for the next
generation.
A uniform operator was employed for the crossover oper-
ation, which selects genes (bits) from parents’ chromosomes
and creates a new ospring. A bit mutation was used as the
mutation operator ; i.e., it randomly changes the new ospring
[].
Two key parameters are the crossover (pc) and mutation
(pm) probabilities since they express the frequency with
which the crossover and mutation operations are carried out,
which have great impact on the GA performance [].
In this way, multiple tests were conducted to dene
the probability values (pc and pm) that could be used in
our GA-based scheme. We have employed  test values
within the interval [0.10.8] for the crossover probability
(pc) and  test values within the interval [0.01 0.8] for the
mutation probability (pm), as shown in Table . Combined,
these meant  test cases. For each test case,  simulation
instances were performed. e test scenario was composed
of  channels; PU service rate equals  for all channels
andPUarrivalrateineachchanneldenedwithinthe
interval [0 1].Moreover,theSUarrivalrateandaverageSU
service time were uniformly distributed in [1 3] and [1 4],
respectively.
We have selected the highest average tness value for
the last generations population (Figure ). Test case 
displayed the best performance for the GA, having crossover
and mutation probabilities equal to . and ., respectively.
In addition to crossover and mutation probabilities, the
population size (S) and number of generations (G) were
dened as being equal to  and , respectively; all GA
parameter values are summarized in Table .
T  : Te s t c a s e v a lues.
Pc ./././././././.
Pm ./././././././.
T : GA parameters.
Parameter Value
Number of generations (G) 
Population size (L) 
Crossover rate (Pc) .
Mutation rate (Pm) .
6.4. Execution Flow of the Scheme. e execution ow of our
SVN mapping GA-based scheme is as follows (see Figure ).
GivenaSVNrequest,thepopulation(whichmightbeused
for mapping the SVN) is randomly generated so as to
provide candidate solutions. e information about channel
availability is considered to create feasible solutions. en,
the individuals are evaluated in accordance with the adopted
tness function that takes into account the SU blocking
probability, joint utilization, SU dropping probability, and
collision probability for computing each individuals tness
value. ereupon, the individuals are submitted for selection,
together with the crossover and mutation operators, and
moreover, an elitist strategy is employed to ensure the best
tness individuals will not be lost during the selection
process. Finally, a new generation of candidate mappings
will be created and the stop criterion, which is determined
by the number of generations (G), is evaluated. If the stop
criterion is not satised, the process is repeated in the tness
evaluation stage. Otherwise, the best individual is chosen as
thenalsolution.isrepresentsthesetofchannelsthatwill
be allocated to the requested SVN. In this way, the pool of
available channels is updated.
7. Scheme Evaluation
is section presents the scenarios and metrics adopted in the
performance evaluation of the proposed scheme. It analyzes
the results achieved in comparison to those of the First-Fit
strategy,aswellastheGAconvergence.
7.1. Evaluated Metrics. For evaluation purposes, four metrics
were employed: collision probability, SU blocking probability,
SU dropping probability, and joint utilization (see Section ).
ey aim to show the mapping impact that was carried out,
on both primary and secondary communications and on the
resource utilization.
7.2. Evaluation Scenarios. ree evaluation scenarios were
dened to analyze the proposed scheme. Each scenario has an
evaluation focus, which are presented in Table . In the rst
scenario, the SU arrival rate varied from  to  with a step of
, which shows the behavior of the proposed scheme under
dierent SU loads. e 𝑃𝑈,𝑚𝑖𝑛 and 𝑃𝑈,𝑚𝑎𝑥 were dened as 
Wireless Communications and Mobile Computing 
Initial Population:
Bit Mutation
New Population:
Uniform Crossover
Roulette Wheel
Selection
Numberof
Generation >
threshold?
Fitness Evaluation
SU Blocking Probability
SU DroppingProbability
No
Elitism
Ye s
Best Individual is chosen:
SVN embedding is performed.
Individuals (SVN Mappings) generated randomly
New Individuals
(SVN mappings)
Joint Utilization
SVN Request
Pool of available
channels
Update of
available channels
CollisionProbability
F : Execution ow of the scheme based on GA.
T : Evaluation scenarios.
Scenario Focus
To evaluate the schemes considering
dierent SU loads
To evaluate the schemes considering
dierent PU load
To evaluate the schemes considering
more than one SVN request
and , respectively, and the substrate network was composed
of  channels.
To evaluate the performance of the scheme in cases where
the environment has dierent PU loads and, hence, distinct
possibilities of opportunistic use, in the second scenario, two
cases (intervals for 𝑃𝑈,𝑖) were dened. In the rst, 𝑃𝑈,𝑚𝑖𝑛
and 𝑃𝑈,𝑚𝑎𝑥 were set as  and ., respectively. ereupon,
the channels that compose this scenario are more susceptible
to opportunistic use because of the low primary activity. In
thesecondcase,anewthePUrateintervalof[0.51]was
specied. It represents scenarios with high PU load, which
reduces the possibility of opportunistic use. Additionally, the
SUarrivalratewaschangedtousers/s.echannelnumber
and SVN requests were the same as the previous case and the
remaining parameters are as described in the rst paragraph
of this section.
e goal in the two rst scenarios was to analyze the
proposed scheme under dierent PU and SU loads. us, the
number of SVNs to be mapped was one SVN in both cases.
T : Parameter values adopted in the scenarios.
All Scenarios
Parameter Value
𝑃𝑈,𝑖 U[𝑃𝑈𝑚𝑖𝑛,𝜆𝑃𝑈𝑚𝑎𝑥]
𝑃𝑈,𝑖 user/s
𝑆𝑈,𝑙 user/s
Scenario  Scenario  Scenario 
[𝑃𝑈𝑚𝑖𝑛,𝜆𝑃𝑈𝑚𝑎𝑥 ][01] [00.5]and [0.51] [01]
𝑆𝑈,𝑙 ///  U[2 4]
Channels   
SVN 
On the contrary, the third scenario evaluated the schemes
performance when more than one SVN had to be mapped.
Hence,thenumberofSVNrequestswassetupto.For
this experiment, the SU arrival rate was uniformly distributed
between  and  users/s, the substrate network was composed
of  channels and similarly to the rst scenario, the PU
arrival rates were uniformly distributed within [0 1].
In all scenarios, the primary (for all channels) and
secondary service rates were dened as  user/s, along with
the PU arrival rate of each channel (𝑃𝑈,𝑖)thatwasset
within the interval [𝑃𝑈,𝑚𝑖𝑛,𝑃𝑈,𝑚𝑎𝑥],with𝑃𝑈,𝑚𝑖𝑛 ≤𝑃𝑈,𝑚𝑎𝑥.
Table  show s t he c h o s e n p a r a m e t e r v a l u e s .
A First-Fit strategy (similar to that used in []) was
also built for further comparison. It maps the SVNs sequen-
tially, in the same way the GA scheme does. First, all the
 Wireless Communications and Mobile Computing
T : Average number of channels adopted by the schemes to
map the SVN considering the scenario .
𝑆𝑈,𝑙 First-Fit GA
. .
. .
 . .
 . 
50 100 150 200 250
133
134
135
136
137
138
139
Generation
Average Fitness
F : Average tness of the population over the generations.
SVN requests are sorted in a descending order in terms
of requested demand and are placed in a queue. Next,
the First-Fit strategy is employed to sequentially allocate
channels for each SVN in the sorted queue. It takes the
nextavailablechannel(notbeingusedbyotherSVNs)to
map the current SVN, aiming to achieve the lowest collision
probability. Similarly to the GA scheme, the First-Fit strategy
encompasses the restrictions dened in ().
Next, the GA convergence and a comparison between the
GA-based and the First-Fit are drawn. For each evaluated
point,  instances were performed and the average results
are presented considering a % condence level, which were
obtained by using the Bootstrap method [], with ‘resample
size and number of (re)samplings equal to  and ,
respectively. No bars were drawn due to a small dierence
between upper and lower bounds.
7.3. GA Convergence. Before evaluating the GA considering
all the scenarios and metrics dened in previous subsections,
its convergence was examined with regard to the population
average tness. To this end, the scenario  with 𝑆𝑈,𝑙 equal
to  was taken into account. Moreover, we extended the
GA’s evolution process by adopting  generations to verify
whether the average tness would change signicantly aer
the number of generations dened in Table .
Figure  shows the evolution of the populations average
tness. It can be noted that in the rst  generations the
average tness increases sharply as consequence of the GAs
exploration process in which it searches for new solutions
and explores the search space. In the next  generations,
the population tness rises soly, indicating that the GA is
rening already existing solutions to improve their tness
(exploitation). About generation , the average tness of
the population becomes stable and no signicant changes
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
SU Blocking Probability
GA Scheme
First-Fit Scheme
6 8 10 12 14 164
SU Arrival Rate ( users/s)
F : Average blocking probability when the SU arrival rate
varies.
take places, denoting that the individuals have similar tness
values. It is worth observing that other GA-based works such
as [], which performs resource allocation during network
operation and, thus, has a critical time scale, converge
signicantly later than our GA-based scheme.
7.4. Results for Scenario 1. is section presents the results
obtained by the schemes when the SVN experiences dierent
SU loads. In terms of SU blocking probability, they achieved
similar performance (there are intersections between the
condence intervals), with slight superiority for the First-Fit,
when SU arrival rate is , , or  (see Figure ). On the other
hand, the GA-based scheme showed a stable performance
under dierent SU loads in the SVN. Besides, it selected the
appropriate channels to meet each demand, which allowed
the SU blocking probability to be lower than ..
As shown in Figure , both approaches presented similar
SU admission levels; however, the First-Fit adopted more
channelsthanourGA-basedscheme(seeTable).isshows
that the proper selection of channels is more important than
thenumberofchannelstobeallocated,sincetheyhave
dierent primary usage patterns. Likewise, an intersection
was observed between the two schemes when the SU arrival
rate varies from  to  (see Figure ). For the rst value,
the First-Fit adopts channels that provide a higher user per
channel density, which leads to a higher SU blocking proba-
bility than our GA-based scheme. When the SU arrival is ,
the First-Fit allocates much more channels (sequentially) in
order to reduce the collision probability. With more channels,
the possibility of opportunistic access by SU increases and,
consequently, the SU blocking probability reduces.
Forcollisionprobability,theGA-basedschemeoutper-
formed the First-Fit for all SU arrival rates (see Figure ).
In the rst three cases (𝑆𝑈,𝑙 equals , , or ), the average
reductioninthecollisionprobabilitywas.%(onaverage)
andevenwheninahighSUload(𝑆𝑈,𝑙 equals ), our
GA-based scheme obtained a performance gain of .%.
In brief, the GA approach was superior to the First-Fit for
various SU loads in up to .%; i.e., it largely reduced the
interference caused to PU communication, providing better
protection to the PU.
Wireless Communications and Mobile Computing 
GA Scheme
First-Fit Scheme
0.25
0.3
0.35
0.4
0.45
0.5
0.55
0.6
Collision Probability
6 8 10 12 14 164
SU Arrival Rate (users/s)
F : Average collision probability when SU arrival rate varies.
GA Scheme
First-Fit Scheme
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
SU Dropping Probability
6 8 10 12 14 164
SU Arrival Rate(users/s)
F : Average SU dropping probability when SU arrival rate
varies.
In order to achieve a low collision probability in this
scenario, the First-Fit scheme allocated more channels to
the SVN than the GA scheme (see Table ). However, this
did not ensure a low collision probability. e PU load of
the selected channels must be observed to properly select
the channels that will provide less interference. As the GA
evaluates multiple solutions in each generation and composes
the nal solution by using building blocks, it provided a lower
collision probability even adopting fewer channels.
e average SU dropping probability got by the schemes
is illustrated in Figure . Similarly to the previous metric,
our GA-based scheme also outperformed First-Fit. It signif-
icantly reduced the SU dropping probability in up to .%
comparedtotheFirst-Fit.eseresultswereachievedwhen
𝑆𝑈,𝑙 was  but even in the worst case, when 𝑆𝑈,𝑙 was , it
reduced the SU dropping probability by .%. On average,
the GA-based scheme reduced the SU dropping probability
by .% compared to the First-Fit. erefore, a reasonable
QoS could be achieved by enabling the admitted SUs a better
chance to nish their communications.
Results for the channel’s joint utilization are shown in
Figure . As opposed to previous evaluations, the First-
Fit had a slightly better performance than the GA scheme,
mainly where the SU arrival rate is considered low, such as for
 and  users/s. In these cases, it achieved a performance gain
of .% and .%, respectively. On the other hand, both
0.55
0.6
0.65
0.7
0.75
0.8
0.85
Joint Utilization
6 8 10 12 14 164
SU Arrival Rate (users/s)
GA Scheme
First-Fit Scheme
F : Average joint utilization when SU arrival rate varies.
T : Average number of channels adopted by the schemes to
map the SVN considering the scenario .
PU Arrival Rate First-Fit GA
[0 0.5] . .
[0 5 1] .
behaved similarly when there was a high load of SUs, such
as with 𝑆𝑈,𝑙 equal to  or , where the absolute dierence
between them was .% and .%, respectively.
Although First-Fit achieves better joint utilization, this
does not imply that the secondary utilization is higher
than that provided by the GA-based scheme. Since both
approaches are similar in terms of SU blocking and our
scheme is better than First-Fit (when the SU dropping is con-
sidered), the secondary utilization achieved by GA strategy
is probably higher than that obtained by the competitor. In
this scenario, the First-Fit achieves a high joint utilization
by selecting high PU load channels, which reduces the
possibility of opportunistic access and, consequently, highly
impacts on the secondary utilization.
7.5 . R e sult s f o r S ce n a r io 2. is section analyzes the results
obtained by the schemes when the substrate networks are
composed of channels with high or low PU load, leading to
dierent opportunistic access possibilities.
Figure  presents our ndings considering a low PU load.
e First-Fit scheme achieved approximately zero blocking
and dropping probabilities for the SUs, by allocating more
channels than the GA scheme (see Table ). As the channels
have low PU load (PU arrival rate dened between  and
.), the opportunistic access possibilities increased and, as a
consequence, the SU blocking and SU dropping probabilities
were greatly reduced. On the other hand, although the Fist-
Fit scheme had achieved a low collision probability, our
GA scheme has outperformed it, by reducing the collision
probability by .%.
e probability of having active SUs increases when the
First-Fit is adopted as more SUs are admitted in the SVN.
Also, when more channels are allocated and each one has
aPUload,thetotalPUarrivalrateincreases.However,
as noted, lower collision probability cannot be achieved by
 Wireless Communications and Mobile Computing
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Collision Probability SU Blocking
Probability
SU Dropping
Probability
Joint Utilization
GA Scheme
First-Fit Scheme
F : Results obtained by schemes when the PU arrival rates
are within the [0.00.5].
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Collision Probability SU Blocking
Probability
SU Dropping
Probability
Joint Utilization
GA Scheme
First-Fit Scheme
F : Results obtained by schemes when the PU arrival rates
are within the [0.51.0].
simply allocating more channels to SVN. Moreover, the way
of mapping used by the First-Fit signicantly impairs the
joint utilization, as shown in Figure , where it had .%.
Contrarily, the GA-based scheme provided a reasonable
tradeo between the adopted metrics. It has selected the most
appropriate channels to map the SVN and although both
SU blocking and SU dropping probabilities are higher than
those oered by the First-Fit, the values obtained by GA-
basedschemewerealsolowanditachievedahigherjoint
utilization, with a gain of .%.
e GA-based scheme also presented better results
towards the blocking probability, considering the PU arrival
rate dened between . and . (high PU load), as shown
in Figure . With GA, the blocking probability was reduced
by .% compared to the First-Fit scheme, which has
a high blocking probability under heavy PU load. Such
behavior occurs because the First-Fit scheme allocates fewer
channels for the SVNs in order to achieve lower collision
probability, thus reducing the SU access to the SVN. On
the other hand, it increases the user per channel density,
which impacts on the blocking probability and, conse-
quently, on the SU admission. is shows that the First-
Fit scheme is not able to provide a reasonable service to
the SUs when the channels are submitted to high primary
loads.
In terms of collision probability, Figure  shows that the
First-Fit had a slightly superior performance in comparison
to our scheme. e absolute dierence between their perfor-
mances was .%, which means a reduction of .% in the
collision probability. Although this may suggest the First-Fit
scheme’s superiority, it achieved a lower collision probability
because it allocated fewer channels to SVN (see Table )
and its SU blocking probability was high (see Figure ),
similar to what was described in the previous paragraph.
With fewer admitted SUs, the collision probability tends to
be lower. Hence, the result obtained by the First-Fit scheme
was masked by its high blocking probability and this way
of providing protection to PU communication signicantly
impairs the secondary network. e GA-based scheme, in
turn, provided a similar protection to PU and a reasonable
servicetoSVN.
e results for the SU dropping probability are also shown
in Figure , where the GA scheme surpassed the First-Fit,
with gain of .%, on average. Hence, our scheme provided
better service to the secondary communication, by admitting
more SUs and by reducing the chances of communication
awsduetotheforcedterminationprocess.
In terms of joint utilization, the First-Fit strategy outper-
formed the GA by .% (in absolute value). It has selected
fewerchannelstomaptheSVN(seeTable)and,therefore,
presented higher user per channel density. However, this
caused an adverse eect in the secondary communication,
since more SUs were rejected or dropped. Briey, the joint
utilization achieved by the First-Fit does not mean a higher
secondary usage. On the contrary, our scheme achieved
a similar joint utilization but provided more secondary
opportunities. In addition, it reached the goal dened for SU
blocking probability and a close result to that dened for joint
utilization (see Section .).
7.6 . R e sult s f o r S cen a r io 3. e third scenario’s challenge was
map multiple SVNs with maximum eciency. As shown in
Figure , the GA-based scheme had better lower blocking
probability value, ensuring the secondary access to the SVNs.
Wireless Communications and Mobile Computing 
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Collision Probability SU Blocking
Probability
SU Dropping
Probability
Joint Utilization
GA Scheme
First-Fit Scheme
F : Results obtained by schemes when  SVNs are mapped.
In brief, when  SVNs were mapped, the First-Fit achieved
an average blocking probability of . while the GA had
.. In other words, the GA approach reduced blocking
probability by .%.
In terms of collision probability, the GA-based scheme
also outperformed First-Fit (see Figure ). It has been found
that the GA reduced the collision probability by .%
when compared to the First-Fit scheme; i.e., the GA-based
scheme provides more protection for PU communication.
Furthermore, it should be stressed that our scheme can
provide a higher protection degree for the PU and ensure a
low SU rejection rate for the SVN simultaneously, which is
not the case for the First-Fit.
e SU dropping probability results are also drawn in
Figure . For such metric, our scheme outplayed the First-
Fit with an impressive reduction of .%. us, our scheme
does not only admit more secondary users in the SVN, but
also provides better quality for secondary communication, as
the possibility of the secondary communication dropping is
largely mitigated.
For the joint utilization (see Figure ), it was noted
that the First-Fit had the best value, with performances
gap of .% (absolute value). However, this does not
necessarily mean higher secondary utilization or QoS. Again,
it is possible to infer that our scheme in fact provides
higher secondary utilization, as the achieved SU blocking
and SU dropping probabilities were lower than the First-
Fit. erefore, as in the rst scenario, in order to achieve a
higher joint utilization, the First-Fit selected channels with
higher primary load to map the SVNs, meaning that the
primary utilization was dominant on the joint utilization
result.
8. Conclusion
We have combined two key technologies for G networks
(Cognitive Radio and Wireless Virtualization) in order to
provide enhanced resource utilization besides dealing with
heterogeneous applications and wireless technologies with no
hardware modication. By empowering the synergy between
them, a new scenario, denoted as CRVNE, has been presented
and modeled to address the SVNs mapping onto cognitive
radio substrate as a multiobjective problem. A GA-based
scheme was suggested as an alternative to a known solution
named First-Fit strategy. It was found that our approach
provided reasonable protection to the primary communica-
tion and an ecient tradeo between the SU blocking and
dropping and joint resource utilization. Besides, we have also
few situations where the First-Fit had apparently beaten our
scheme, such as for the joint utilization metric.
e GA’s intrinsic parallelism and the use of tness
functions that encompass multiple objectives enable many
solutions to be handled, improved, and evaluated simulta-
neously.Duringthemappingprocess,theproposedscheme
not only deals with the collision probability, SU dropping
probability, and SVN demands, but also takes a broad view of
the SU blocking probability and joint utilization, which also
composes the dened objectives for the mapping problem.
Several practical implications emerge from our results; for
instance,itwasfoundthatitispossibletosharespectrum
resources (e.g., channels) between dierent access priority
users while meeting individual demands. us, a business
model that oers a layered service type (e.g., primary and
secondary) can be deployed, for example, by assigning a price
menu to virtual network communications depending on their
access level, being managed by dierent service providers:
PSPs and SSPs.
When sharing resources between virtual networks (PVNs
and SVNs) with dierent access levels, it is important to
ensure that their demands and restrictions are met, for
example, a PVN that imposes inexible restrictions regarding
interference to its communication. is condition must be
satised during the mapping process and in a similar way,
during the SVN mapping, so that the SVN can opportunis-
tically access the resources, preventing starvation, despite the
PUload.Inbrief,wehavesoughttoshowthatourGA-
based scheme can simultaneously meet the restrictions and
demands from PVNs and SVNs.
Regarding the future development of this work, some
possibilities may be explored. e rst involves designing
other bio-inspired approaches to the same problem while the
second relates to the model extension through the support
for heterogeneous secondary user scenarios (e.g., SVNs with
dierent QoS requirements), considering SU handover and
channel aggregation technology, and the combination of
spectrum access approaches such as opportunistic spectrum
access (OSA) and spectrum leasing.
Data Availability
edatausedtosupportthefndingsofthisstudyareavailable
from the corresponding author upon request.
 Wireless Communications and Mobile Computing
Conflicts of Interest
On behalf of all authors, the corresponding author states that
there are no conicts of interest.
Acknowledgments
e author Andson Balieiro would like to thank the Pernam-
buco State Research Foundation (FACEPE) for the nancial
support through Grant IBPG --./. e author Kelvin
DiaswouldliketothanktheNationalCouncilforScientic
and Technological Development (CNPq) for the nancial
support through Grant /-. e authors would
like to thank the Universidade Federal de Pernambuco
(UFPE), a public Brazilian university, where this work was
performed as part of the rst author’s Ph.D. thesis. e thesis
document is available at the UFPE’s repository.
References
[] H. Ji, S. Park, J. Yeo, Y. Kim, J. Lee, and B. Shim, “Ultra-reliable
and low-latency communications in G downlink: physical
layer aspects,IEEE Wireless Communications Magazine,vol.,
no. , pp. –, .
[]G.Liu,F.R.Yu,H.Ji,V.C.M.Leung,andX.Li,“In-band
full-duplex relaying for G cellular networks with wireless
virtualization,IEEE Network,vol.,no.,pp.,.
[] M.Kalil,A.Moubayed,A.Shami,andA.Al-Dweik,“Ecient
low-complexity scheduler for wireless resource virtualization,
IEEE Wireless Communications Letters,vol.,no.,pp.,
.
[] B. Fan, H. Tian, and X. Yan, “A generic framework for hetero-
geneous wireless network virtualization: Virtual MAC design,
in Proceedings of the 2016 IEEE Wireless Communications and
Networking Conference, WCNC 2016,Qatar,April.
[] J. K. Park, H. S. Jeon, M.-C. Ko, and S. H. Noh, “R-Map: a hybrid
map for mobile robot in ubiquitous computing environment,
INFORMATION-An International Interdisciplinary Journal,vol.
, no. , pp. –, .
[] S. Zhang, Z. Qian, J. Wu, and S. Lu, “An opportunistic resource
sharing and topology-aware mapping framework for virtual
networks,” in Proceedings of the IEEE Conference on Computer
Communications, INFOCOM 2012, pp. –, USA, March
.
[] W.Xiang,K.Zheng,andX.Shen,5G Mobile Communications,
Springer International Publishing, Z¨
urich, Switzerland, .
[] X. Hong, J. Wang, C.-X. Wang, and J. Shi, “Cognitive radio in
G: a perspective on energy-spectral eciency trade-o,” IEEE
Communications Magazine,vol.,no.,pp.,.
[] H. Wen, P. K. Tiwary, and T. Le-Ngoc, “Current trends and
perspectives in wireless virtualization,” in Proceedings of the
2013InternationalConferenceonSelectedTopicsinMobileand
Wireless Networking, MoWNeT 2013,pp.,Canada,August
.
[] K. Zhu, Z. Cheng, B. Chen, and R. Wang, “Wireless virtual-
ization as a hierarchical combinatorial auction: An illustrative
example,” in Proceedings of the 2017 IEEE Wireless Communi-
cations and Networking Conference, WCNC 2017,USA,March
.
[] A. Belbekkouche, M. M. Hasan, and A. Karmouch, “Resource
discovery and allocation in network virtualization,IEEE Com-
munications Surveys & Tutorials, vol. , no. , pp. –,
.
[] A. Balieiro, M. Falcao, and K. Dias, “Secondary virtual network
mapping onto cognitive radio substrate: a collision probability
analysis,” IEEE Communications Letters,vol.,no.,pp.
, .
[] A. Balieiro, P. Yoshioka, K. Dias, D. Cavalcanti, and C. Cordeiro,
A multi-objective genetic optimization for spectrum sensing in
cognitive radio,Expert Systems with Applications,vol.,no.,
pp. –, .
[] J. Jia, Y. Deng, J. Chen, A. H. Aghvami, and A. Nallanathan,
Achieving high availability in heterogeneous cellular networks
via spectrum aggregation,IEEE Transactions on Vehicular
Tech n o l o g y ,vol.,no.,pp.,.
[] D.-T.Huynh,X.Wang,T.Q.Duong,N.-S.Vo,andM.Chen,
“Social-aware energy eciency optimization for device-to-
device communications in G networks,Computer Communi-
cations, vol. , pp. –, .
[] B. Han, J. Lianghai, and H. D. Schotten, “Slice as an evolutionary
service: genetic optimization for inter-slice resource manage-
ment in G networks,IEEE Access, vol. , pp. –, .
[] C.-C. Lin, H.-H. Chin, and W.-B. Chen, “Balancing latency
and cost in soware-dened vehicular networks using genetic
algorithm,JournalofNetworkandComputerApplications,vol.
, pp. –, .
[]S.Chen,T.R.Newman,J.B.Evans,andA.M.Wyglinski,
“Genetic algorithm-based optimization for cognitive radio
networks,” in Proceedings of the 33rd IEEE Sarno Symposium
2010,USA,April.
[]S.M.A.Kazmi,N.H.Tran,T.M.Ho,andC.S.Hong,
“Hierarchical matching game for service selection and resource
purchasing in wireless network virtualization,IEEE Communi-
cations Letters,vol.,no.,pp.,.
[] C. Wang, K. Sohraby, R. Jana, L. Ji, and M. Daneshmand,
“Network selection in cognitive radio systems,” in Proceedings of
the 2009 IEEE Global Telecommunications Conference, GLOBE-
COM 2009, USA, December .
[] J. Lai, E. Dutkiewicz, R. P. Liu, R. Vesilo, and G. Fang, “Network
selection in cooperative cognitive radio networks,” in Proceed-
ings of the 11th International Symposium on Communications
and Information Technologies, ISCIT 2011, pp. –, China,
October .
[] K.Nakauchi,K.Ishizu,H.Murakami,A.Nakao,andH.Harada,
AMPHIBIA: A cognitive virtualization platform for end-to-
end slicing,” in Proceedings of the 2011 IEEE International
Conference on Communications, ICC 2011, Japan, June .
[] A. Popescu, H. Kim, F. R. Davoli et al., “ENVIRAN: Energy
ecient virtual radio access networks,” in Proceedings of the 2013
16th International Symposium on Wireless Personal Multimedia
Communications, WPMC 2013 - Co-located with Global Wireless
Summit 2013,USA,June.
[]M.Al-Ayyoub,Y.Jararweh,A.Doulatetal.,“Virtualization-
based cognitive radio networks,e Journal of Systems and
Soware,vol.,pp.,.
[] C. Xin and M. Song, “Dynamic spectrum access as a service,
in Proceedings of the IEEE Conference on Computer Communi-
cations, INFOCOM 2012, pp. –, USA, March .
[] C. Yang, J. Li, M. Guizani, A. Anpalagan, and M. Elkashlan,
Advanced spectrum sharing in G cognitive heterogeneous
Wireless Communications and Mobile Computing 
networks,IEEE Wireless Communications Magazine,vol.,
no.,pp.,.
[] L. Akter, B. Natarajan, and C. Scoglio, “Modeling and fore-
casting secondary user activity in cognitive radio networks,” in
Proceedings of the 17th International Conference on Computer
Communications and Networks, ICCCN 2008, pp. –,
Virgin Islands, August .
[] K. Singh and M. Xie, Bootstrap: A Statistical Method,Rut-
gers University, , http://www.stat.rutgers.edu/home/mxie/
rcpapers/bootstrap.pdf.
[] K. Deb, Multi-objective Optimization Using Evolutionary Algo-
rithms, Wiley, New York, NY, USA, .
[]C.A.C.Coello,G.B.Lamont,andD.A.vanVeldhuizen,
Evolutionary Algorithms for Solving Multi-Objective Problems,
Springer,NewYork,NY,USA,.
[] W.Yuan,X.You,J.Xu,H.Leung,T.Zhang,andC.L.P.Chen,
“Multiobjective optimization of linear cooperative spectrum
sensing: Pareto solutions and renement,IEEE Transactions on
Cybernetics,vol.,no.,pp.,.
[] D. E. Goldberg, “Genetic algorithms in search, optimization,
and machine learning,Choice Reviews Online,vol.,no.,
pp. -  -  , .
... Diferente de [11], adotamos Algoritmos Genéticos (AGs) e Otimização por Enxame de Partículas (PSO) para definir os coeficientes dos modelos. Devido a vesartilidade, escalabilidade e simplicidade computacional, os AGs e o PSO têm sido amplamente utilizados para resolver problemas de otimização [1] [12] [18]. ...
... AGs e PSOs têm sido adotados na resolução de problemas de otimização [1] [12] [18] [9] [16]. Em [9], os autores formulam o problema do consumo de energia de aplicação multimídia em comunicação dispositivo a dispositivo (D2D). ...
... Através dos operadores genéticos (seleção, cruzamento e mutação), as diferentes soluções são combinadas a fim de que uma boa solução seja encontrada. Os AGs têm sido amplamente empregados com sucesso na resolução de diversos problemas na literatura [1]. ...
Conference Paper
In the last years, the multimedia traffic has increased significantly and the mobile devices (e.g. smart phones and tablets) have been widely used to consume this content type. Video applications demand high energy consumption of the device because they perform complex operations and deal with a large data amount. Although hardware improvements in the mobile devices have been achieved, the advances in battery technology have not kept the same pace. In this respect, the combination of video applications with the limited battery capacity of the mobile devices has challenged the academia and industry in the development of techniques for energy management and provision of quality of experience (QoE) to the user. Energy consumption estimation models may assist these techniques, as well as, the decision made process when the computational offloading from the mobile device to the cloud is considered. This paper presents an evolutionary approach based on Genetic Algorithms (GAs) and Swarm Particle Optimization (PSO) for energy consumption estimation in mobile devices running video applications. The proposal is directly applicable to different model types (linear and non-linear ones), without the linearization cost, and it is evaluated in terms of mean squared error (MSE), using energy consumption measurement data of videos with different configurations. Results show the superiority of our proposal in comparison to the literature that adopts the Ordinary Least Squares method.
... The fifth generation (5G) has started its operation in some countries and is aimed at meeting demands beyond the current system capabilities such as the huge amount of connected devices from IoT applications (e.g., smart cities) and device-to-device communication (e.g., factory automation), the explosive growth of highspeed mobile data traffic (e.g., ultrahigh definition video streaming and virtual reality applications), and ultrareliable and low latency communication (e.g., telesurgery and autonomous vehicle) [1]. To address these needs, the electromagnetic spectrum must be made available [2]. ...
... Figure 2 shows the interactions between the architecture components when a test is performed. First, the user requests the list of test cases from the central CRT (1), which queries the test cases database for getting this list (2). Once the list is ready (3), the central CRT shows it to the user (4). ...
Article
Full-text available
The fifth generation (5G) of mobile networks has started its operation in some countries and is aimed at meeting demands beyond the current system capabilities such as the huge amount of connected devices from IoT applications (e.g., smart cities), explosive growth of high-speed mobile data traffic (e.g., ultrahigh definition video streaming), and ultrareliable and low latency communication (e.g., autonomous vehicle). To attend to these needs, the electromagnetic spectrum must be made available, but the static spectrum allocation policy has caused a spectrum shortage and impaired the employment/expansion of the wireless systems. To overcome this issue, the dynamic spectrum access (DSA) has been promoted in 5G/6G networks, which is enabled by the cognitive radio (CR) technology. Although diverse mechanisms have been developed to tackle the challenges that emerge in different CR layers/functionalities, a standardized testing methodology and system for CR is still immature. Existing standards or methodologies and systems for CR only focus on the definition of network technologies (e.g., IEEE 802.22 and IEEE 802.11af), performance evaluation of CR algorithms/mechanisms, or definition of the device cognition level via performance results or psychometric approaches, not covering systems/methodologies to verify if the device meets the CR capabilities and regulatory policies, neglecting the conformance testing. In this respect, this paper proposes a flexible methodology and system for CR conformance testing under two perspectives, functionalities and limits. We instantiate it by using the Universal Software Radio Peripheral (USRP) software-defined radio platform and present a proof-of-concept with a conformance metric. The results show the feasibility of our proposal.
... GA has been widely adopted for solving wireless network optimization problems [35] and multiple solutions to reduce its convergence time have been proposed [36]. In this work, we assume that the GA is adopted for the dimensioning phase, i.e., before proper operation. ...
Article
Full-text available
In Ultrareliable and Low Latency Communications (URLLC), balancing trade-offs between energy consumption, service availability, and strict reliability and latency requirements is a significant challenge, especially in unmanned aerial vehicle (UAV)-enabled multi-access edge computing (MEC) environments. The constraints imposed by the size, weight and power limitations of UAVs further complicate this task. This study addresses optimizing resource allocation in such environments to meet URLLC demands while minimizing power consumption and maximizing service availability. We explore the virtualization layer of the network function virtualization (NFV)-MEC architecture, incorporating node availability and power consumption alongside conflicting URLLC reliability and latency demands. We introduce an energy-aware model based on continuous-time Markov chain (CTMC) with an embedded virtual resource scaling scheme for Dynamic Resource Allocation (DRA). To solve the optimization problem related to MEC-enabled UAV node dimensioning, we propose a genetic algorithm (GA)-based solution. Our results demonstrate that the proposed GA-based approach achieves a superior balance, with up to a 44% reduction in power consumption compared to the first fit with maximum resources strategy, while also improving service availability and meeting URLLC requirements. This work provides a comprehensive analysis of key virtualization parameters and their impact on critical services within a single NFV-MEC over a UAV node, offering a robust framework for future 6 G network applications.
... The work of Wang et al. [11] has considered the usage of artificial intelligence, where the resource spectrum is targeted to be improved. Another recent work of Balieiro et al. [12] has a virtual networking system has been improved for better access using a unique mapping model. The work of Bega et al. [13] [14] has used a deep learning approach over a cognitive radio targeted for predicting the capacity to hold an upcoming traffic system. ...
Article
Full-text available
The adoption of cognitive radio technology is characterized by various beneficial characteristics that can facilitate better spectrum sensing performance in a 5G network and thereby acting as a boosting element towards a high data transmission rate. However, it is also characterized by various challenges that limit the significant development in resource utilization in 5G. Therefore, this paper introduces a novel and simplified mechanism that facilitates the 5G network to perform better in data transmission and its associated quality of it. The proposed system also performs modeling using practical constraints associated with the usage of cognitive radio over 5G networks using a convex optimization approach. The model is simulated using practical environmental parameters to prove that the proposed system excels better performance in faster processing and quality signal in contrast to the existing resource allocation scheme exercised in 5G networks.
Article
Fast growth of wireless devices in IoT (Internet of Things), industry 5.0 and 6 G cellular networks, is dealing with the challenge of spectrum scarcity and uneven utilization of available resources. To achieve Ambient Intelligence (AmI) in IoT communication network, dynamic utilization of available network resources, is a challenge faced by resource constraint IoT devices. Cognitive radio technology enables these devices to intelligently sense the free spectrum hole and smartly manage available channels opportunistically. To jointly deal with the challenge of spectrum scarcity and application specific network service provision, we propose a novel multi-attribute based two-tier framework called Cognitive Radio Efficient Spectrum Utilization (CR-ESU) which integrates the concept of Network Function Virtualization (NFV) as Virtual Utility Functions (VUFs) with cognitive radio enabled IoT devices to achieve the dynamic and flexible network services as well as spectral efficiency. The proposed framework is one of the pioneer works that incorporate the two ground breaking technologies, NFV and CRN to create a smart and flexible CR-enabled IoT network. CR-ESU also ensures the best channel selection using Dynamic Channel Reservation (DCR) scheme to decrease the channel switching rate. Continuous Time Markov Chain (CTMC) model is used for network modeling and mathematical expression formation. Promising results show the potential of CR-ESU framework to form a reconfigurable and efficient network.