Content uploaded by Adnan Mahmood
Author content
All content in this area was uploaded by Adnan Mahmood on Feb 01, 2018
Content may be subject to copyright.
DOI: 10.4018/IJBDCN.2018010101
Volume 14 • Issue 1 • January-June 2018
Copyright © 2018, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
1
Adnan Mahmood, Faculty of Computer and Information Technology, Al-Madinah International University, Malaysia &
Faculty of Engineering, Universiti Malaysia Sarawak, Malaysia
Shadi M. S. Hilles, Faculty of Computer and Information Technology, Al-Madinah International University, Malaysia
Hushairi Zen, Faculty of Engineering, Universiti Malaysia Sarawak, Malaysia
Vertical Handover Decision (VHD) algorithms are indispensable components of forthcoming 4G
heterogeneous wireless networks architecture – so as to provide requisite Quality of Service to
an assortment of applications anywhere at any time, while allowing seamless roaming in highly
dynamic scenarios (i.e. multitude of access network technologies that vary in bandwidth, latency,
monetary cost, etc.) using Mobile Terminals (MTs) enabled with multiple access interfaces. In this
article, a critical review of the existing VHD algorithms has been carried out as an effort to update
the previous studies. To offer a methodical contrast, recently published VHD algorithms have been
classified into four major classes depending on the key handover decision criterion used, i.e. RSS
based algorithms, bandwidth based algorithms, cost function based algorithms, and the combination
algorithms. Moreover, operational fundamentals, advantages, and disadvantages of exemplary VHD
algorithms for each class have been presented to assess the tradeoffs between their intricacy of
implementation and the efficacy.
4G, Network Selection, Vertical Handover, Wireless Networks
The Beyond Third Generation (B3G) wireless communication systems intends to offer the end-users
with the appropriate global information access competences and personalized wireless communication
services (Chandavarkar and Reddy, 2012; Kassar et al., 2008; Assouma et al., 2006). Their architecture
aims to integrate an assortment of heterogeneous wireless networks over an Internet Protocol (IP)
backbone. The recently sanctioned / ratified IEEE 802.21 Media Independent Handover (MIH) standard
intends to support seamless roaming amongst various wireless access technologies, comprising of the
GSM, UMTS, WLAN, WiMAX and the Bluetooth, through different handover techniques. Several
leading world’s operators have already started deploying this approach. In January 2009, 4G network
Volume 14 • Issue 1 • January-June 2018
2
CLEAR was launched through the collaboration of Clearwire and Intel in Portland, Oregon, USA.
Similarly, major carriers such as AT&T are in process of converting their existing networks into 4G
by using a successor of UMTS – 3rd Generation Partnership Project (3GPP) Long Term Evolution
(LTE) standards (Kassar et al., 2008).
The International Telecommunication Union (ITU) Radiocommunication Sector (ITU-R) has
explicitly specified a globally accepted and agreed definition of the 4G in consultation with its
diverse stakeholder groups, as to what should be encompassed in the nucleus of a 4G System (or
International Mobile Telecommunication Advanced – IMT Advanced as defined by ITU), so that
technologies could earn the right to be categorized into this group. At present, there are only two
families of standards that fit in the bill – LTE Advanced and WiMAX Release 2 (Assouma et al.,
2006). Though High-Speed Packet Access (HSPA) and Evolved High Speed Packet Access (HSPA+)
are marketed by various network operators as 4G services in various parts of world, but they are not
considered 4G in the true sense (Kassar et al., 2008; Zhang et al., 2007).
The convergence of internet and wireless mobile communication accompanied by massive
growth in number of cellular subscribers has led mobility management to emerge as a significant and
challenging domain for wireless mobile communication over the internet (Chandavarkar and Reddy,
2012; Kassar et al., 2008; Assouma et al., 2006; Zhang et al., 2007). Mobility management enables
serving networks to locate roaming terminals for call delivery (i.e. location management) and ensures
a seamless connection as the MT enters into the new service area (i.e. handover management). The
hierarchy of the mobility management in a heterogeneous network environment is depicted in Figure 1.
Location management facilitates the serving network to trace the location of the MT; and
constitute(s) two stages, i.e. location registration and call paging. In location registration, the MT
periodically advises specific signals to inform the network of its present location (or of its newest
access point ~ AP) in order to keep the location database updated (Vuong, 2008). Call paging,
which is invoked after the location registration procedure, is for querying the network about the
Figure 1. Hierarchy of mobility management in a heterogeneous network environment
Volume 14 • Issue 1 • January-June 2018
3
MT’s location profile in order to deliver any calls successfully (Zekri et al., 2012; Misra et al., 2008;
George et al., 2008).
Handover management involves the entire gamut / breadth of issues and actions by which a MT
keeps its connection active, as it migrates from one AP to another and comprises of three stages – first,
the instigation of handover is caused by a MT, network agent, or changing network circumstances;
second is establishment of a new connection, wherein, the network must find new resources for
successful handover connection and execute certain additional routing procedures; lastly, data flow
control needs to sustain the data delivery from the old connection path to a new one based on agreed
QoS guarantees (Sen, 2010; Pahlavan and Krishnamurth, 2013).
Handover is a process of maintaining a user’s active sessions, as MT traverses from the air
interface served by one base station (BS) to air interface served by another BS (also referred to as
Point of Attachment - PoA). Depending on the access network that each PoA belongs to, the handovers
can be either classified as Horizontal or Vertical (Mahmood et al., 2013; Shih and Chen, 2010; Han
et al., 2007). Figure 2 illustrates the graphical representation of horizontal and vertical handovers.
Horizontal (or an intrasystem) handover takes place amongst the PoAs supporting the same
network technology, i.e. between any two geographically neighboring IEEE 802.11 APs (Shaikh
and Nerkar, 2010); whereas, vertical (or an intersystem) handover occurs amongst PoAs supporting
different network technologies, i.e. between IEEE 802.11 AP and a 3G cellular network (Rao et al.,
2014; Dutta and Schulzrinne, 2014).
A vertical handover is executed across the heterogeneous cells of access systems, which differ(s)
in numerous characteristics, i.e. bandwidth, data rate, frequencies of operation, etc., thus making its
implementation quite challenging as compared to the horizontal handovers. Vertical handovers are
further categorized as being either Upward or Downward (Bhuvaneswari and Raj, 2012; Lin, 2011).
Upward vertical handover roams an overlay with larger cell coverage and lower bandwidth per
unit area. This makes the MT disconnect from a network providing faster but smaller coverage (i.e.
WLAN) to a new network possessing slower but broader coverage (i.e. WCDMA or UMTS or cellular
networks) (Kirsal et al., 2013; Kim and Jan, 2002). Upward vertical handovers are time-critical, as
duration of the small cell layer is time constrained. Contrarily, downward vertical handover roams an
overlay with a smaller cell size and higher bandwidth per unit area. This makes the MT disconnects
from a network providing slower but broader coverage to a new network providing a limited coverage
with higher access speeds (Wozniak et al., 2011; Dixit and Prasad, 2003).
Handovers could also be distinguished depending on the number of engaged frequencies as either
Intrafrequency or Interfrequency (Rao et al., 2014; Das, 2013; Holma and Toskala, 2010; Braithwaite
and Scott, 2004). Intrafrequency handovers occurs amongst the PoAs operating on same carrier
frequency. This type of handover is a feature of CDMA networks with frequency division duplex
(FDD). Whereas, interfrequency handovers occurs amongst the PoAs operating on different carrier
frequency. This type of handover is a characteristic of CDMA networks with time division duplex
(TDD). In order to execute interfrequency handovers, efficient schemes needs to be envisaged for
Figure 2. Graphical illustration of horizontal and vertical handovers (Yan, 2010)
Volume 14 • Issue 1 • January-June 2018
4
performing measurements on the other frequencies, while still having the connection being operated
on the current frequency.
In addition to the classification schemes highlighted above, handovers could be categorized
as Hard and Soft, wherein, a hard handover transpires if a MT remains associated with only one
AP at a time (i.e. break-before-make handover) and a soft handover emerges if a MT communicate
with more than one AP during handover (i.e. make-before-break handover). Other categorizations
include the mobile-assisted handover, network-controlled handover, mobile-controlled handover,
intra-administrative handover, inter-administrative handover, obligatory handover, and voluntary
handover, etc. (Rao et al., 2014; Berndt, 2008).
In this manuscript, a critical review of the existing VHD algorithms has been carried out as an
effort to update the previous studies. To offer a methodical contrast, recently published VHD algorithms
have been classified into four major classes depending on the key handover decision criterion used,
i.e. RSS based algorithms, bandwidth based algorithms, cost function based algorithms, and the
combination algorithms. Moreover, operational fundamentals, advantages, and disadvantages of
exemplary VHD algorithms for each class have been presented to assess the tradeoffs between their
intricacy of implementation and the efficacy.
The vertical handover process can be segmented into three stages – Network Discovery, Handover
Decision, and Handover Triggering (Akyildiz and Wang, 2002).
Network Discovery is the process wherein a MT with multiple interfaces searches for the
accessible wireless networks. This is achieved by activating all the critical interfaces of a MT, so that
the service advertisements broadcasted by the different wireless technologies can be heard. However,
by keeping interfaces continuously active, consumes power even without receiving or sending some
packets. The ideal approach is to periodically activate the MT interfaces, so as to continuously receive
the service advertisements. Discovery time should also be kept low, so that MT can benefit faster
from the new wireless networks. As the activating frequency directly influences a system’s discovery
time; the MT activating interfaces with a higher frequency may discovers the reachable network
much more quickly, but its battery may also get consumed quickly. Therefore, there always exists a
tradeoff between the power efficiency and system discovery time (Cao and Zhang, 2014; Habeib et
al., 2011; Chen et al., 2004).
Handover Decision is the competence to decide about the targeted PoA and the exact time of
handover. This decision depends on several issues and policies pertinent to the network to which
a MT is already connected, and to the one that it is going to perform handover (i.e. RSS, available
bandwidth, network connection time, power consumption, handover latency, network security,
MTs mobility, monetary costs, and MTs preferences). VHD schemes generally comprises of three
close integrated processes – Handover Necessity Estimation (HNE) ascertaining that if a particular
handover would be essential to an accessible network; Handover Target Selection (HTS) opting for
the most optimal network amongst the accessible candidates depending on a fixed set of criteria;
and Handover Triggering Condition Estimation (HTCE) determining the right moment in order to
commence handover out of the currently connected network. Figure 3 suggests an in-depth algorithm
underlining detailed steps involved in a VHD process (Yahiya, 2011; Zhang et al., 2009).
Handover Triggering requires transfer of data packets to a new wireless link, so as to re-route
the MT’s connection path to a selected PoA. Since 4G heterogeneous networks operate in a multi-
network / multi-standards environment, the transfer of packets to a new wireless link is augmented
with the contextual information. This is done to enable MT to move through diverse networks so as
to minimize any delay in re-establishing its traffic flows. However, if the context transfer delay is
as large as having the impact of a complete re-establishment or large enough to increase the overall
Volume 14 • Issue 1 • January-June 2018
5
handover call drop rate, the advantages of context transfer are completely lost (Wei et al., 2013; Bae
et al., 2011; Hasswa et al., 2006).
Existing literature reveals that a considerable number of studies have previously surveyed the design
/ performance issues of vertical handoffs in an envisaged 4G heterogeneous network environment,
along with analysis and comparison of several VHD algorithms (Navarro and Wong, 2006; Zahran
and Liang, 2005; McNair and Zhu, 2004). However, these studies now stand outdated in terms of the
presented VHD algorithms (Rajule et al., 2013; Yussuf et al., 2012; Yan et al., 2010; Gaikwad and
Bhute, 2014) or scope of these remain quite constricted with some comparing just the operational
fundamentals of various VHD algorithms and some comparing performances based on applicable
networking technologies, throughput and packet delay.
Figure 3. Vertical handover decision process
Volume 14 • Issue 1 • January-June 2018
6
Rajule et al. 2013, Yussuf et al. 2012, and Yan et al. 2010 illustrated a comparison of 15, 12 and
12 VHD algorithms / schemes respectively; with majority of compared algorithms proposed before
year 2008 in the associated literature.
In this research, a critical review of the existing VHD algorithms has been carried out as an effort
to update the previous studies. To offer a methodical contrast, the recently published VHD algorithms
have been classified into four major classes depending on the key handover decision criterion used,
i.e. RSS based algorithms, bandwidth based algorithms, cost function based algorithms, and the
combination algorithms. Moreover, operational fundamentals, advantages, and disadvantages of
exemplary VHD algorithms for each class have been presented to assess the tradeoffs between their
intricacy of implementation and the efficacy.
Several parameters have been previously deliberated in the literature for consideration in the VHD
algorithms (Rao et al., 2014; Gaikwad and Bhute, 2014; Ramya and Poornima, 2013; Sinan et al.,
2008). Some of them are briefly explained as follows:
• Received Signal Strength: One of the most critical and widely used parameter in VHD
algorithms. RSS is easy to determine and has a close correlation with the link quality. There
always exist(s) a sharp relationship between RSS measurements, and the distance amongst a
MT and its PoA;
• Available Bandwidth: A transparent and crucial indicator for assessing the traffic performances
in an access network, and is closely associated to the link capacity, i.e. total number of channels,
or a measure of accessible data communication resources in bits per second (bps);
• Network Security: Significant for the applications demanding confidentiality and integrity
of the transmitted information. Thus, a network with the higher encryption schemes would be
preferred / opted over network with lower level of data security;
• Cost of Service / Monetary Costs: Can affect users decision for network selection in
heterogeneous networking environment, as charging policies (encompassing both traffic costs and
roaming costs amongst diverse networks) can significantly vary for the different service providers;
• Handover Latency: For an MT, it is depicted in terms of the time duration / interval that has
elapsed between the arrival of the first packet along the new access router (after a handover
is performed) and last packet received via the previous access router. Handover latency plays
decisive role in the interactive cellular applications, as it could be significantly different amongst
different technologies;
• Power Utilization: Becomes especially crucial if MT’s battery is low. In such sort of
circumstances, handover is desirable to a PoA that may help extend battery’s life;
• Network Connection Time: Refers to a duration for which MT remains associated to a specific
network. Measuring network connection time is critical for opting the exact moment for handover
initiation, so as to maintain QoS at a satisfactory level. For instance, a handover performed too
early for a MT traversing from WLAN to 3G cellular network would cause network resource
wastage, and a too delayed one would ultimately result in handover failure;
• User Personal Preferences: For a particular access network can result in selection of one sort
of network over the other candidate networks.
VHD algorithms may be quantitatively evaluated by assessing their performances in terms of total
number of handovers, number of handover failures caused due to the erroneous decisions, handover
latency, and the aggregate session’s throughput (Chandralekha and Nehera, 2010; Tripathy and
Acharjya, 2014). These performance metrics are further deliberated as under:
Volume 14 • Issue 1 • January-June 2018
7
• Total Number of Handovers: Critical as too frequent handovers would result in network resource
wastage. A handover is regarded as superfluous if a handover back to original PoA is required
inside a particular time frame (ping-pong effect), and such type of handovers should be avoided;
• Handover Failure: Transpires if the handover has been instigated, but the target network is
unable to accommodate it due to inadequate resources (due to channel’s non-availability), or if
MT passes out of the coverage area of the target network at very high speed consequently not
letting the handover process to be finalized;
• Handover Latency: The time interval in between the initiation and execution of a handover
request. Minimizing of handover latency is critical for managing the time- sensitive voice and
multimedia applications;
• Aggregate Throughput: Depicted as the sum of data rates delivered to the MTs in the network. It
is always desirable to perform handover to the candidate networks possessing higher throughput.
VHD algorithms can be segregated in numerous ways (Bhosale and Daruwala, 2013; Arabmakki et al.,
2012). In this manuscript, VHD algorithms have been grouped into four major categories depending
on the key handover decision criterion adopted:
• RSS Based Algorithms: Assesses RSS of the present PoA with that of candidate PoA in order
to execute a handover. Due to the simplicity of requisite tools for the RSS measurements, a
significant proportion of research has already been carried out in this particular domain;
• Bandwidth Based Algorithms: Regards available bandwidth to be one of the prime decision
criterion in this group;
• Cost Function Based Algorithms: Amalgamates various network parameters, such as network
security, monetary costs, handover latency, available bandwidth, power consumption, etc. within
a cost function; and handover decision is then finalized via assessing the result of this function
with that of candidate’s network;
• Combination Algorithms: Attempts to employ a richer set of the input parameters as compared
to other schemes in order to make the handover decisions. This makes it quite complicated for
the researchers to devise analytical formulations of handover decision processes, and machine
learning techniques (i.e. fuzzy logic and artificial neural networks) are thus widely employed.
This section portrays a critical review of representative groups (as per the classifications outlined in
Section 3.3) of existing VHD algorithms. The operational fundamentals, advantages, and disadvantages
of exemplary VHD algorithms for each group have been presented to assess the tradeoffs between
their intricacy of implementation and the efficacy.
4.1.1. A Dwell Time Prediction Based Heuristic
In order to minimize the number of handover failures and unnecessary handovers for MTs traversing
through a WLAN cell, Hussain et al. (2013) suggested a VHD algorithm estimating the dwell time
(i.e. time for which a MT’s connection is maintained over a particular cell and is thus closely related
to the MT’s speed) and computation of certain time threshold values. A handover failure transpires
if the handover latency into ( )
τ
i the WLAN cell is greater than the dwell time; and an unnecessary
handover occurs when the sum
τ
( )
of handover latency into ( )
τ
i and out ( )
τ
o of the WLAN cell
exceeds the predicted dwell time.
Volume 14 • Issue 1 • January-June 2018
8
A geometrical model has been envisaged for the said scheme (as depicted in Figure 4), wherein,
an AP is assumed at point A (0, 0) possessing a coverage radius of a. The MT traverses in a straight
line through the WLAN coverage region with P
i and P
o as the entry and exit points, and covering
the distance l PP
i o
= = 2
a
sin
φ
, where
φ
is the acute angle that MT makes at point P
i with the
tangent. The MT further observes its distances d1 and d2 from the AP at two sample points P
1 and
P
2 respectively along its trajectory path, which can be computed by using the Log-distance Path
Loss Model:
d d i
i ref
P PL RSS X
Tx ref P
= =
− − +
10 1 2
10
( )
, ,
σ
β
(1)
where, di is the distance of MT from the AP, dref is reference distance, PTx is transmitted power of
WLAN’s AP in dBm, PLref refers to path loss at reference distance, RSSp is the received signal strength
at reference distance along the MT’s trajectory,
β
is the path loss exponent, while X
σ
signifies
attenuation. The dwell time for a MT is evaluated as:
Ta
v
az
v z
Ta
v
=
=+≤ ≤
2 2
1
02
2
sin
φ
, (2)
wherein, v represents MT’s velocity; and:
Figure 4. Hussain’s et al. (2013) proposed VHD geometrical model
Volume 14 • Issue 1 • January-June 2018
9
zm m
m m
l t
l t
=−
+
1
with ml as slope of the line of trajectory of MT and mt denotes the slope of tangent line at point Pi. The
time threshold values for keeping the probability of handover failures (M) and unnecessary handovers
(N) within desirable bounds is calculated based on various network parameters as:
Mak
v k
M
f
f
i
=+< ≤
2
1
0
2,
τ
(3)
kv
a v
P
f
i
i
f
=−
−
tan arctan
τ
τ
π
42
2 2 2 (4)
N
ak
v k
N
u
u
T
=+< ≤
2
1
0
2,
τ
(5)
kv
a v
P
u
T
T
u
T i o
=−
−
= +tan arctan ,
τ
τ
πτ τ τ
42
2 2 2 (6)
where, kf and ku are introduced as handover failure and unnecessary handover variable substituted
to simplify the mathematical expressions of M and N respectively; whereas, Pf and P
u refers to
probability of handover failures and unnecessary handovers correspondingly. Figure 5 depicts the
Hussain’s et al. (2013) heuristic.
Hussain’s et al. (2013) heuristic though considerably reduced the number of handover failures
and unnecessary handovers for MTs velocities of up to approximately 30 km/h in contrast to the
traditional VHD heuristics (as throughput gain has also been considered in order to optimize the
handover decision process); however, it has proven quite ineffective for MTs traversing at high speeds
(Mahmood et al., 2014).
4.1.2. A Travelling Distance Estimation Based Heuristic
Yan et al.1 (Yan et al., 2008) suggested a 3G to WLAN handover decision mechanism that predicts
a MT’s travelling distance in the WLAN coverage area by employing the change rate of RSS,
and estimates distance threshold parameters in order to reduce the probability of handover failures
or unnecessary handovers. A handover procedure is initiated if the predicted travelling distance
d is greater than the distance threshold parameter L. As the MT penetrates in the WLAN coverage
boundary at any point Pi, the stretch between AP (O) and point Pi ( )lOP
i is given through the
path loss model:
Volume 14 • Issue 1 • January-June 2018
10
lE
RSS
OP
t
P
i
i
=
1010
1
εβ
(7)
where, Et is the transmit power of AP,
RSS
P
i is the received signal strength at the WLAN cell’s
entrance point Pi,
β
signifies the path loss exponent, and
ε
refers to Gaussian distributed random
variable with a zero mean and standard deviation up to 12 dB. The change rate of RSS is as follows:
∆ = −
−
−
−
RSS RdR E
dv
t
2
22
10
410
2
β
β
ε
(8)
where,
∆RSS
is computed by taking two sample points S1 and S2 along the MT’s trajectory, and
travelling speed v is computed by using the VEPSD algorithm due to its low estimation errors
(Mohanty, 2006). Once
∆RSS
and v has been obtained, the travelling distance d can be predicted
by using equation (9). The distance threshold parameter L for reducing the number of handover
failures is calculated as:
L R v
RPf
=
−
−
22 2
1
sin sin
τ π
(9)
Figure 5. Hussain’s et al. (2013) heuristic
Volume 14 • Issue 1 • January-June 2018
11
where,
τ
represents handover latency into the WLAN cell and Pf signifies the tolerable handover
failure probability. Figure 6 demonstrates the Yan’s et al. (2008) heuristic.
The benefit of Yan’s et al. (2008) algorithm is its capability to reduce the number of
handover failures and connection breakdowns at high speeds (i.e. over 30 km/h) as compared
to the other similar VHD mechanisms, but it remains quite critical at the pedestrian speeds
(Mahmood et al., 2014). Furthermore, absolute interior difference between the MT’s angle of
arrival and departure in the proposed geometrical model has been erroneously assumed between
0 2,
π
( )
which has led to significant mathematical imperfections in the computation of
threshold values, as it could only fall in the range of 0,
π
( )
. Moreover, as the RSS fluctuates
due to the shadow fading, but no deliberation has been undertaken in the said algorithm to
minimize its impact.
4.2.1. A Received Signal to Interference and Noise Ratio
(SINR) and Back-Haul Bandwidth Based Heuristic
Wang & Kao (2011) suggested a received signal to interference and noise ratio (SINR) and back-haul
bandwidth based VHD algorithm in order to perform handovers from the Wireless Metropolitan Area
Network (WMAN)2 to WLAN. The scheme ascertains the data rate of WLAN and compares it with
WMAN to ensure seamless service continuity without any disruption. Based on the computation of
Figure 6. Yan’s et al. (2008) heuristic
Volume 14 • Issue 1 • January-June 2018
12
maximum available back-haul bandwidth among candidate APs, the best target AP is selected for
handover.
In order to determine the available back-haul bandwidth, each AP estimates its available back-
haul bandwidth (R) with utilization (U) and back-haul bandwidth (B) as:
R U B= −
( )
×1 (10)
Higher utilization rates results in lower available back-haul bandwidth thus consequently leading
to increase in transmission delay. Since the value of R would be periodically broadcasted by each
individual AP; the array of available back-haul bandwidth for all APs RAP
( )
is given as:
R R R R R
AP AP AP AP AP
n
= …
1 2 3
, , ,, (11)
In order to opt for the best candidate AP for a MT to perform handover to, the maximum available
downlink bandwidth ( )
θ
i for APiis computed as follows:
θ
i
t T
T
WMAN t AP AP AP
AP
i
i i
C R RT T= + × −
( )
=
∑
'
,
' (12)
where, CWMAN t, denotes the data rate of WMAN at time t (in order to keep an uninterrupted connection,
it is expected that the data rate of WLAN would be greater than WMAN at the time of handover.
Therefore, the point of coincident data rates for WLAN and WMAN – referred as X T C
',′
( )
, has
been estimated by authors); RTAP is expected residence time inside the AP and can be computed
through d
v
with d as the radius of AP and
v
as the MT’s average velocity; and TAP
i
' as the time of
both coincident data rates. The APi with the maximum
θ
i would be opted as the target AP for
handover. Figure 7 demonstrates the Wang & Kao (2011) heuristic.
By regarding maximum available back-haul bandwidth as the foremost VHD criterion, Wang
& Kao (2011) heuristic was able to achieve a consistently high system throughput and low handover
latency (i.e. small packet delay) in contrast to RSS based VHD schemes. Nevertheless, acquiring
available back-haul bandwidth information in cellular networks for VHD Decision is quite difficult
(Lee et al., 2005).
4.2.2. A Combined SINR Based Heuristic
Bathich et al. (2013) proposed a bandwidth based VHD algorithm amongst WLAN (IEEE
802.11) and WiMAX (IEEE 802.16) networks using the received SINR. The SINR of WiMAX
γ
BS
( )
is transformed into an equivalent SINR ( )
γ
AP
Equ to be evaluated with SINR calculations
of WLAN signals
γ
AP
( )
. The maximum achievable downlink date rate, i.e. RBS and RAP for the
MTs connected with WiMAX and WLAN respectively is determined through the Shannon
Capacity formula as:
Volume 14 • Issue 1 • January-June 2018
13
R W log
BS BS
BS
BS
= +
21
γ
Γ (13)
R W
AP AP
AP
AP
= +
log21
γ
Γ (14)
Letting RBS = RAP, the relationship between requisite
γ
BS (equivalent) and
γ
AP (in case of same
downlink data rate being offered to MTs by the WiMAX and WLAN) can be computed as:
γγ
AP AP
BS
BS
W
W
Equ
BS
AP
= +
−
ΓΓ
1 1 (15)
where,
Γ
refers to dB gap amongst the uncoded Quadrature Amplitude Modulation (QAM) and the
channel capacity, minus the coding gain; and
W
BS and
W
AP refers to carrier bandwidths. Through
combined effects of both SINRs being taken into contemplation, handover is initiated if MT is
receiving higher equivalent SINR from another access network as depicted in Figure 8.
SINR based VHD heuristics can provide MTs with a higher overall throughput in contrast to RSS
based algorithms, as available throughput is directly reliant on SINR. The said scheme has resulted
in a balanced load amongst the WiMAX and WLAN networks, nevertheless, such sort of algorithms
results in excessive handovers with significant SINR variation(s) thus causing MTs to handover back
and forth frequently among two networks, also regarded as the ping-pong effect (Pollini, 1996).
Figure 7. Wang and Kao (2011) heuristic
Volume 14 • Issue 1 • January-June 2018
14
4.3.1. An Adaptive Cost Function Based VHD Algorithm
Li & Chen’s (2013) VHD algorithm relies on a cost function that calculates and compares the cost
function value of each candidate network, and assigns higher priority to the network with a lower cost
function and regards it as an optimal choice for the MT to perform handover to. The cost function
is calculated as:
Cost Value w delay w RSS w
n delay n RSS
n
power
_ log log log=
( )
+
+
1PPower
wSecurity w Cost
n
security
n
cost n
( )
+
+
( )
log log
1
(16)
where, Cost Valuen
_ refers to the cost function value of network n, delayn is propagation delay
from network n to MT, RSSn is the received signal strength from network n, Powern regards to
power consumption required to access network n, securityn is network’s n security level, Costn
is the requisite monetary cost to access the network n; and w w w w w
delay RSS power security cost
, , , , are
the corresponding weights for the said network parameters satisfying the following equation:
w w w w w
delay RSS power security cost
+ + + + =1 (17)
This implies that if RSS from a particular network n is higher, the network delay is lower, power
consumption is lower, network security is higher, monetary costs are lower, and thus accordingly, the
cost of accessing the network n is lower.
Figure 8. Bathich’s et al. (2013) heuristic
Volume 14 • Issue 1 • January-June 2018
15
An adaptive function has also been incorporated in order to enhance the MT’s experience by
dynamically adjusting weight factors of certain network parameters as per the requirements of different
network environments. The authors have illustrated a scenario, wherein, if the residual power of
MT deteriorates, MT can enhance the weight factor associated with power consumption gradually
by employing the adaptive algorithm thereby increasing the priority of network with a lower power
consumption. The adaptive function is computed as:
if Power Limit then
wp
w
p
ww
power
power
security
s
>
( )
= −
+
=
,
11
eecurity
cost
cost
RSS
RSS
delay
delay
p
ww
p
ww
p
ww
p
=
=
=
(18)
where, p refers to the step factor and is a positive integer.
The primary benefit of Li & Chen’s (2013) algorithm is the increased user experience (i.e.
user preference / satisfaction) due to the dynamic adjustment of the weight factors. However,
the authors didn’t discussed that how the weights for QoS factors were obtained; and since
each network parameter has different units, information pertinent to normalization of QoS
factors has not been deliberated too. Furthermore, the VHD heuristic result in an overall short
handover delay due to introduction of a hysteresis threshold of 0.3 among cost function values
of both networks in order to avoid ping-pong effect. The flowchart of Li & Chen’s (2013)
algorithm is depicted in Figure 9.
Figure 9. Li and Chen’s (2013) heuristic
Volume 14 • Issue 1 • January-June 2018
16
4.3.2. A Simple Additive Weighting (SAW) Algorithm
Garzon et al. (2013) demonstrated a Simple Additive Weighting (SAW) algorithm based on the
assessment of information pertinent to the QoS parameters and preferences of active critical
services for the 3GPP / LTE, WiFi, and WiMAX networks. SAW is a form of cost function based
VHD algorithm that acquires network parameters as inputs, and is referred to as the weighted
sum of normalized values of the selected parameters (Miyim et al., 2014). The QoS parameters
are divided into two main groups – ascending parameters deemed to possess highest values
such as RSS (rss) and availability (av); and the descending parameters deemed to possess lower
values including packet loss (pl), delay (d), load (l), jitter (j), and monetary cost (c). A decision
maker is employed to assign respective weights for each parameter based on the priority / needs
of each service, thus allowing each operator to state the most important parameters in order to
select the most apposite network. The suggested SAW mathematical model is described in terms
of following equation:
V W d
d
Wj
j
Wpl
pl
Wav
av
i d
m
j
m
pl
m
av
n
=
+
+
+
⋅ ⋅ ⋅ ⋅
1 1 1
+
+ −
+ −
⋅
⋅ ⋅
Wrss
rss
Wc
cWl
l
rss
n
c
n
l
n
1 1
(19)
where, nth and mth value of a parameter refers to the highest and lowest value amongst the available
network interfaces respectively; with W W W W W W
d j pl av rss c
, , , , , and Wl as the assigned weights
for the respective parameters.
Once, requisite information pertaining to all of the available interfaces has been obtained,
the said algorithm compares the information of each interface in order to establish the highest
and lowest values; and then proceeds to normalize all of them and thus obtains SAW weighted
values. The selected handover target is the one that scores the highest value, provided it is
different from the present interface. Figure 10 illustrates the flowchart for Garzon’s et al.
(2013) heuristic.
One of the advantages of Garzon’s et al. (2013) algorithm is its holistic consideration of the
various network parameters, and the implementation of load balancing by considering a channel’s
burden as the foremost decision criterion before opting for it. However, information pertinent to the
weights assignment and normalization of QoS parameters has not been deliberated. Similarly, the
said scheme requires an extra cooperation amongst the MT and PoAs of candidate networks, which
may result in increased network latency.
4.4.1. A Fuzzy Logic Control Based Heuristic
Feng et al. (2013) developed a low-complexity fuzzy logic control based VHD algorithm on the
basis of rough set theory in order to reduce fuzzy logic decision rules and opt for the key parameters
(amongst the available network parameters) as the input criterion. The key parameters are passed
through a fuzzy logic controller in order to assess the Access Network Candidacy Value (ANCV),
and thus selected access network is the one possessing maximum ANCV. Figure 11 portrays the
flowchart of Feng’s et al. (2013) heuristic.
Volume 14 • Issue 1 • January-June 2018
17
The input parameters are initially mapped into three sets as per corresponding triangular fuzzy
membership function, wherein, membership degree can be represented as
µ µ µ µ
c L M H
=
[ ]
, , ,
with L, M and H referring Low, Medium, and High values – a process referred to as fuzzification.
Hereafter, fuzzy rule base establishment and reduction process is executed to establish fuzzy rule
base (fuzzy rules specifies relationships between input sets and output fuzzy sets, and are critical for
the fuzzy inference systems). Since numerous parameters are considered for the VHD process and
the size of fuzzy rule base is also quite considerable, the authors have illustrated a fuzzy rule reduction
scheme based on rough set theory in order to eliminate the redundant parameter
α
( )
in the process
of fuzzy rule base establishment, with the importance of
α
in a discernibility matrix
M S mij n n
( )
=
× computed as:
fm
m
m
i
n
j
n
ij
ij
ij
ij
ij
αλλα
α
( )
= = ∉
∈
= =
∑∑
1 1
0
1
,
(20)
Figure 10. Garzon’s et al. (2013) heuristic
Volume 14 • Issue 1 • January-June 2018
18
where,
S
refers fuzzy rule base as a decision information system / decision table represented
by
S U C D= ∪
( , ) with U x x x xn
= …
{ }
1 2 3
, , , , as finite set of objects – universe, and
C D&
as the set of reduction and decision attributes with C a i m
i
= = …{ , , , }1 2 , where, ai denotes
the
i
th input parameter and D d=
{ }
correspondingly. The discernibility function of M S
( )
is defined as:
f a k n
M S k
( )
= ∨ ∧ ≤ ≤
{ }
1 (21)
Once the redundant parameter
α
with least importance has been eradicated from the set of
reduction attributes, the value of P1 / P0 is compared against β% with P0 as number of fuzzy rules
before reduction operation and P1 as the number of fuzzy rules whose handover decisions are not
altered by reduction operation. β% signifies reduction of
C
with regards to
D
and changes with
the user requirements. If P1 / P0 is greater than β%, fuzzy rule reduction process is said to have
been established.
The ANCV of any network j is computed in the process of defuzzification as:
Y f x y x
x
j
l
M
li
p
i
l
M
i
p
i
=
( )
=
( )
( )
′
′
= =
= =
∑ ∏
∑ ∏
1 1
1 1
'
µ
µ
(22)
Figure 11. Feng’s et al. (2013) heuristic
Volume 14 • Issue 1 • January-June 2018
19
where, M signifies the number of fuzzy rules, yl
' symbolizes the output linked to rule Rl, ′
( )
µ
xi
refers to input membership value corresponding to xi, and
p
denotes number of elements in vector
x
. Furthermore, the handover would not be enabled until
Y Y
j th
>
, where
Y
th is ANCV’s threshold
and Yth ∈
[
)
0 5 1. , . The candidate access network with highest ANCV is the MT’s selected target:
A Y
j N j
*=∈
arg max (23)
where, N represents the set of candidate access networks of a MT.
4.4.2. An Adaptive Neuro-Fuzzy Based Heuristic
Çalhan & Çeken (2010) suggested an adaptive neuro-fuzzy based heuristic comprising of an Adaptive
Fuzzy Logic (AFL) based and an Adaptive Network Fuzzy Inference System (ANFIS) based VHD
algorithm for wireless heterogeneous networks comprising of WiFi, WiMAX, GSM / GPRS, and
UMTS technologies. Three parameters – RSS, data rate, and the monetary cost – are processed as inputs
of the AFL based system in order to acquire the training elements as output, which are subsequently
fed in as the ANFIS inputs. Trim and trapezoid are chosen as the fuzzy membership functions for the
parameters due to their potential of achieving better performance in especially real-time applications.
The block diagram of AFL based VHD algorithm is depicted in Figure 12.
The ANFIS scheme is devised in order to allow if-then rules and membership functions to
be constructed depending on the earlier obtained input and output metrics data of the AFL based
algorithm. A fuzzy inference system maps input characteristics into input membership functions,
input membership functions to rules, rules to set of output characteristics, output characteristics to
output membership functions, and output membership functions as single-valued output generally
associated with a decision. The proposed architecture of ANFIS is depicted in Figure 13.
The ANFIS is classified as a pattern recognition algorithm with amalgamation of fuzzy logic and
artificial neural network (ANN). The ANN provides the training elements to the fuzzy logic algorithm
in order to adjust the membership functions and rules. The training data with the learning algorithm is
allowed to rules to adapt by ANFIS. The fuzzy inference system produces an output (APCV – Access
Point Candidacy Value) in accordance with available APs parameters thus illustrating their candidacy
Figure 12. Block diagram of adaptive fuzzy logic based VHD algorithm (Çalhan & Çeken, 2010)
Volume 14 • Issue 1 • January-June 2018
20
level. As soon as the scan process is accomplished, the APCV of each candidate AP is compared with
that of the current AP. If the difference of the compared APCV values is greater or equal to that of
the handover resolution, the new AP is chosen as the serving interface / node.
The said heuristic is able to achieve an enhanced performance by reducing the number of
handovers, especially due to its shorter learning duration as compared to the conventional neural
networks. However, it lacks aspects on how the neuro-fuzzy network has been trained, and also results
in an increased algorithm complexity.
Table 1 summarizes the overall comparison of the representative VHD algorithms (as per
classifications outlined in the Section 3.3) in terms of their salient applicable internetworking
technologies, advantages, and disadvantages.
High service availability in order to access the network infrastructure under Anywhere, Anytime
and Always Best Connected paradigm has long been the user requirement. In this manuscript, a
comprehensive critique on the existing VHD algorithms reported in literature has been presented,
Figure 13. Proposed architecture of adaptive network fuzzy inference system (Çalhan & Çeken, 2010)
Volume 14 • Issue 1 • January-June 2018
21
Table 1. Comparison of representative vertical handover decision algorithms
Group Heuristic
Applicable
Internetworking
Technologies
Advantages Disadvantages
RSS Based
Hussain’s et al.
(2013)
Between 3G and
WLAN
i) Significantly reduced the
number of handover failures and
unnecessary handovers for MTs
velocities up to approximately 30
km/h in contrast to the traditional
VHD heuristics.
ii) Throughput gain has been
taken into consideration in order
to optimize the handover decision
process.
The algorithm has proven ineffective for the
MTs traversing at high speeds (i.e. above 30
km/h).
Yan’s et al.
(2008)
Between 3G and
WLAN
Reduction in the number of
handover failures and connection
breakdowns for the MTs traversing
at high speeds (i.e. above 30 km/h)
as compared to other similar VHD
mechanisms.
i) Number of handover failures and connection
breakdowns remains quite critical at pedestrian
speeds.
ii) Absolute interior difference between MT’s
angle of arrival and departure in the geometrical
model is erroneously assumed between (0,2
π
),
which had led to considerable mathematical
imperfections in the computation of threshold
values.
iii) No deliberations were made for minimizing
RSS fluctuations caused due to impact of
shadow fading.
Bandwidth Based
Wang and Kao
(2011)
Between WMAN and
WLAN
Consistently high system
throughput and low handover
latency (i.e. small packet delay) in
contrast to traditional RSS based
VHD schemes.
Acquiring of the available back-haul bandwidth
information in 3G cellular networks for the
VHD Decision is quite difficult.
Bathich et al.
(2013)
Between WLAN and
WiMAX
i) Higher overall throughput in
contrast to RSS based algorithms,
as available throughput is directly
reliant on the SINR.
ii) Balanced network load amongst
the WiMAX and WLAN networks.
Results into excessive handovers with significant
SINR variation, thus causing the MTs to
handover back and forth frequently among the
two networks (i.e. ping-pong effect).
Cost Function Based
Li and Chen
(2013)
Between any two
Heterogeneous
Wireless Networks
Increased user experience (i.e.
user satisfaction) due to dynamic
adjustment of the weight factors.
i) Missing information pertinent to the weights
assignment for QoS factors.
ii) Since each network parameter has different
units, the normalization mechanism for the QoS
factors has not been deliberated.
iii) Results in short handover delay due to
introduction of hysteresis threshold of 0.3
among the cost function values of both networks
in order to avoid the ping-pong effect.
Garzon’s et al.
2013)
Between 3GPP /
LTE, WiFi, and
WiMAX
i) Holistic consideration of the
various network parameters.
ii) Implementation of load
balancing by considering a
channel’s burden as the foremost
decision criterion.
i) Missing information pertinent to the weights
assignment and the normalization of QoS
parameters.
ii) Requires an extra cooperation among the MT
and the PoAs of candidate networks, resulting in
an increased network latency.
Combination Based
Feng’s et al.
(2013)
Between any two
Heterogeneous
Wireless Networks
High success ratio in opting for the
best / optimal candidate network. Increased system complexity.
Çalhan and
Çeken (2010)
Between any two
Heterogeneous
Wireless Networks
Reduction in number of handovers
due to shorter learning duration
as compared to the conventional
neural networks.
i) Increased system complexity.
ii) Training delay.
iii) Lack of details pertinent to training process
of neuro-fuzzy network.
Volume 14 • Issue 1 • January-June 2018
22
wherein, an effort has been carried out in order to update the earlier studies. To offer a methodical
contrast, VHD algorithms have been segregated into four major classes depending on the key handover
decision criterion used, i.e. RSS based algorithms, bandwidth based algorithms, cost function based
algorithms, and combination algorithms. Moreover, the operational fundamentals, advantages,
and disadvantages of exemplary VHD algorithms for each group have been presented to assess
the tradeoff between their intricacy of implementation and efficacy. The survey illustrates that the
presently proposed VHD algorithms lacks an extensive deliberation of various network parameters,
and thus the challenge is to formulate a scheme that encompass wide-ranging network conditions
and user preferences. One potential solution is enhancement in the computational power of handsets,
implementation of numerous VHD algorithms in handsets and adaptation of adaptive techniques that
opt for algorithms intelligently based on a fixed set of parameters.
Volume 14 • Issue 1 • January-June 2018
23
Akyildiz, I. F., & Wang, W. (2002). A Dynamic Location Management Scheme for Next-Generation Multitier
PCS Systems. IEEE Transactions on Wireless Communications, 1(1), 178–179. doi:10.1109/7693.975456
Arabmakki, E., Rashad, S., & Krijestorac, S. (2012). A Comparison of Different Vertical Handoff Algorithms
Between WLAN and Cellular Networks. International Journal of Engineering Research and Innovation, 4(1), 5–12.
Assouma, A. D., Beaubrun, R., & Pierre, S. (2006). Mobility Management in Heterogeneous Wireless Networks.
IEEE Journal on Selected Areas in Communications, 24(3), 638–648. doi:10.1109/JSAC.2005.862407
Bae, S. J., Chung, M. Y., & So, J. (2011). Handover Triggering Mechanism Based on IEEE 802.21 in
Heterogeneous Networks with LTE & WLAN. In Proceedings of the International Conference on Information
Networking (pp. 399-403). doi:10.1109/ICOIN.2011.5723117
Bathich, A. A., Baba, M. D., Rahman, R. A., & Ibrahim, M. (2013), ‘Performance of SINR Based Handover
Among Heterogeneous Networks in MIH Environments’, in Proceedings of the IEEE International Conference
on System Engineering and Technology, pp. 136-141, Shah Alam, Malaysia.
Berndt, H. H. (Ed.). (2008). Towards 4G Technologies: Services with Initiative. Chichester, UK: John Wiley &
Sons Ltd. doi:10.1002/9780470010334
Bhosale, S., & Daruwala, R. (2013). Investigations on IEEE 802.21 Based Media Independent Handoff Algorithm
for Access Network Selection between WiFi and WiMAX. International Journal of Scientific and Engineering
Research, 4(5), 2287–2292.
Bhuvaneswari, A., & Raj, E. G. D. P. (2012). An Overview of Vertical Handoff Decision Making Algorithms.
International Journal of Computer Network & Information Security, 4(9), 55–62. doi:10.5815/ijcnis.2012.09.07
Braithwaite, C., & Scott, M. (2004). UMTS Network Planning and Development: Design and Implementation
of 3G CDMA Infrastructure. MA: Newnes (An Imprint of Elsevier).
Çalhan, A., & Çeken, C. (2010), ‘An Adaptive Neuro-Fuzzy Based Vertical Handoff Decision Algorithm for
Wireless Heterogeneous Networks’, in Proceedings of the IEEE 21st International Symposium on Personal
Indoor and Mobile Radio Communications, pp. 2271-2276, Istanbul, Turkey. doi:10.1109/PIMRC.2010.5671693
Cao, J., & Zhang, C. (2014). Seamless and Secure Communications over Heterogeneous Wireless Networks
~Springer Briefs in Computer Science. New York, United States: Springer. doi:10.1007/978-1-4939-0416-7
Chandavarkar, B. R., & Reddy, G. R. M. (2012). Survey Paper: Mobility Management in Heterogeneous Wireless
Networks. Procedia Engineering, 30, 113–123. doi:10.1016/j.proeng.2012.01.841
Chandralekha, M., & Behera, P. K. (2010). Minimization of Number of Handoff Using Genetic Algorithm in
Heterogeneous Wireless Networks. International Journal of Latest Trends in Computing, 1(2), 24–28.
Chen, W. T., Liu, J. C., & Huang, H. K. (2004). An Adaptive Scheme for Vertical Handoff in Wireless Overlay
Networks. In Proceedings of the Tenth International Conference on Parallel and Distributed Systems, Newport
Beach, CA (pp. 541-548).
Das, S. K. (2013). Mobile Handset Design. Singapore: John Wiley & Sons (Asia) Pte Ltd.
Dixit, S., & Prasad, R. (Eds.). (2003). Wireless IP and Building the Mobile Internet’, Massachusetts. United
States: Artech House.
Dutta, A., & Schulzrinne, H. (2014). Mobility Protocols and Handover Optimization: Design, Evaluation &
Application. West Sussex, UK: Wiley. doi:10.1002/9781119945505
Feng, Y., Xie, L., Chen, H., & Wang, K. (2013), ‘A Low-Complexity Fuzzy Logic Control Based Vertical
Handoff Decision Algorithm with Rough Set Theory’, in Proceedings of the IEEE International Conference
on Wireless Communications & Signal Processing, pp. 1-6, Hangzhou, P.R.China.
Gaikwad, K. D., & Bhute, H. A. (2014). A Survey on Vertical Handover Decision Making in Next Generation
Heterogeneous Wireless Networks. International Journal of Research Studies in Science, Engineering and
Technology, 1(7), 55–64.
Volume 14 • Issue 1 • January-June 2018
24
C. L. Garzon, N. O. Gonzalez, and Y. Donoso (2013), ‘A Proactive VHD Algorithm in Heterogeneous Wireless
Networks for Critical Services’, Vol. 8, No. 3, pp. 425-431.
George, A., Kumar, A., Cavalcanti, D., & Agrawal, D. P. (2008). Protocols for Mobility Management in
Heterogeneous Multi-hop Wireless Networks. Pervasive and Mobile Computing, 4(1), 92–116. doi:10.1016/j.
pmcj.2007.03.002
Habeib, M. Z., Elsayed, H. A., Elramly, S. H., & Ibrahim, M. M. (2011). Heterogeneous Networks Handover
Decision Triggering Algorithm Based on Measurements Messages Transfer Using IP Option Header. In
Proceedings of the Seventh International Conference on Wireless and Mobile Communications (pp. 254-259).
Han, Y. H., Jang, H., Choi, J., Park, B., & McNair, J. (2007). A Cross Layering Design for IPv6 Fast Handover
Support in an IEEE 802.16e Wireless MAN. IEEE Network, 21(6), 54–62. doi:10.1109/MNET.2007.4395111
Hasswa, A., Nasser, N., & Hassanein, H. (2006). Tramcar: A Context-Aware Cross-Layer Architecture for
Next Generation Heterogeneous Wireless Networks. In Proceedings of the IEEE International Conference on
Communications, Istanbul, Turkey (Vol. 1, pp. 240-245). doi:10.1109/ICC.2006.254734
Holma, H., & Toskala, A. (2010). WCDMA for UMTS – HSPA Evolution and LTE. Chichester, UK: John Wiley
& Sons Ltd. doi:10.1002/9780470669501
Hussain, R., Malik, S. A., Abrar, S., Riaz, R. A., Ahmed, H., & Khan, S. A. (2013). Vertical Handover Necessity
Estimation Based on a New Dwell Time Prediction Model for Minimizing Unnecessary Handovers to a WLAN
Cell. Wireless Personal Communications, 71(2), 1217–1230. doi:10.1007/s11277-012-0870-5
Kassar, M., Kervella, B., & Pujolle, G. (2008). An Overview of Vertical Handover Decision Strategies
in Heterogeneous Wireless Networks. Computer Communications, 31(10), 2607–2620. doi:10.1016/j.
comcom.2008.01.044
Kim, S. H., & Jang, Y. M. (2002). Soft QoS-Based Vertical Handover Scheme for WLAN and WCDMA Networks
Using Dynamic Programming Approach. In J. Lee & C. H. Kang (Eds.), Mobile Communications (pp. 134–142).
Berlin, Germany: Springer-Verlag. doi:10.1007/3-540-36555-9_15
Kirsal, Y., Ever, E., Mapp, G., & Gemikonakli, O. (2013). Enhancing the Modeling of Vertical Handover in
Integrated Cellular / WLAN Environments. In Proceedings of the IEEE International Conference on Advanced
Information Networking and Applications (pp. 924-930).
Lee, C. W., Chen, L. M., Chen, M. C., & Sun, Y. S. (2005). A Framework of Handoffs in Wireless Overlay
Networks Based on Mobile IPv6. IEEE Journal on Selected Areas in Communications, 23(11), 2118–2128.
doi:10.1109/JSAC.2005.856833
Li, X., & Chen, R. (2013), ‘Adaptive Vertical Handover Algorithm Based on User Experience for Heterogeneous
Network’, in Proceedings of the International Congress on Image and Signal Processing, Vol. 3, pp. 1540-1544,
Hangzhou, P.R.China doi:10.1109/CISP.2013.6743920
Lin, J. C. (2011). Recent Advances in Wireless Communications and Networks. Rijeka, Croatia: InTech Europe.
doi:10.5772/777
Mahmood, A., Zen, H., & Othman, A. K. (2013). Vertical Handover Decision Processes for Fourth Generation
Heterogeneous Wireless Networks. Asian Journal of Applied Sciences, 1(5), 229–235.
Mahmood, A., Zen, H., Othman, A. K., & Siddiqui, S. A. (2014), ‘A Novel Travelling Time Estimation
Mechanism for Minimizing Unnecessary Handovers from Cellular Networks to WLANs’, in Proceedings of
the IEEE International Conference on Information Technology & Electronic Commerce, pp. 38-42, Dalian,
P.R.China. doi:10.1109/ICITEC.2014.7105567
McNair, J., & Zhu, F. (2004). Vertical Handoffs in Fourth-Generation Multinetwork Environments. IEEE Wireless
Communications, 11(3), 8–15. doi:10.1109/MWC.2004.1308935
Misra, A., Roy, A., & Das, S. K. (2008). Information Theory Based Optimal Location Management Schemes
for Integrated Multi-System Wireless Networks. IEEE/ACM Transactions on Networking, 16(3), 525–538.
doi:10.1109/TNET.2007.901067
Miyim, A. M., Ismail, M., & Nordin, R. (2014). Vertical Handover Solutions Over LTE -Advanced Wireless
Networks: An Overview. Wireless Personal Communications, 77(4), 3051–3079. doi:10.1007/s11277-014-1695-1
Volume 14 • Issue 1 • January-June 2018
25
Mohanty, S. (2006). A New Architecture for 3G and WLAN Integration and Inter-system Handover Management.
Wireless Networks, 12(6), 733–745. doi:10.1007/s11276-006-6055-y
Navarro, E. S., & Wong, V. W. S. (2006). Comparison Between Vertical Handoff Decision Algorithms for
Heterogeneous Wireless Networks. In Proceedings of 63rd IEEE Vehicular Technology Conference, Melbourne,
Australia (pp. 947-951).
Pahlavan, K., & Krishnamurthy, P. (2013). Principles of Wireless Networks: A Unified Approach. Chichester,
Sussex, United Kingdom: John Wiley & Sons, Ltd.
Pollini, G. P. (1996). Trends in Handover Design. IEEE Communications Magazine, 34(3), 82–90.
doi:10.1109/35.486807
Q.-T. N.-Vuong. (2008). Mobility Management in 4G Wireless Heterogeneous Networks [Ph.D. Dissertation].
Université Evry Val d’Essonne, Évry, France.
Rajule, N., Ambudkar, B., & Dhande, A. P. (2013). Survey of Vertical Handover Decision Algorithms.
International Journal of Innovations in Engineering and Technology, 2(1), 362–368.
Ramya, M., & Poornima, U. (2013). QOS Oriented Vertical Handoff for WIMAX / WLAN in Adhoc Wireless
Network. International Journal of Research in Computer and Communication Technology, 2(4), 201–205.
Rao, K. R., Bojkovic, Z. S., & Bakmaz, B. M. (2014). Wireless Multimedia Communication Systems – Design,
Analysis, and Implementation. FL: CRC Press, Taylor & Francis Group. doi:10.1201/b17059
Sen, J. (2010). Mobility & Handoff Management in Wireless Networks. In C.J. Bouras (Eds.), Trends in
Telecommunication Technologies (pp. 457-484). Rijeka: Croatia: InTech Europe. doi:10.5772/8482
Shaikh, A. N., & Nerkar, M. H. (2010). Handoff Issues in 4G – An Overview. In Proceedings of National
Conference on Advances in Recent Trends in Communication & Networks (ARTCON), Ashta, Madhya Pradesh,
India (pp. 239-243).
Shih, C. H., & Chen, Y. C. (2010). A FMIPv6 Based Handover Scheme for the Real-Time Applications in Mobile
WiMAX. Journal of Networks, 5(8), 929–936. doi:10.4304/jnw.5.8.929-936
Sinan, S., Ismail, M., & Jumari, K. (2008). A Comparison of Mobile Node’s Handoff Between Mobile IPV6
and Fast Handover Protocol. Journal - The Institution of Engineers, 69(2), 27-30.
Thiyagarajan, S., & Sedhu, S. (2014). Fuzzy Normalized Handoff Initiation Algorithm for Heterogeneous
Networks. International Journal of Hybrid Information Technology, 7(5), 155–166. doi:10.14257/ijhit.2014.7.5.14
Tripathy, B. K., & Acharjya, D. P. (2014). Advances in Secure Computing, Internet Services and Applications.
Pennsylvania, United States: Information Science Reference (An Imprint of IGI Global). doi:10.4018/978-1-
4666-4940-8
Wang, H. L., & Kao, S. J. (2011), ‘A Vertical Handover Scheme from WMAN to WLAN By Taking Into Account
the Maximum Available Resource’, in Proceedings of the International Conference on Computer Science &
Education, pp. 1373-1378, Singapore. doi:10.1109/ICCSE.2011.6028887
Wei, H. Y., Rykowski, J., & Dixit, S. (2013). WiFi, WiMAX, and LTE Multi-Hop Mesh Networks. NJ: John Wiley
& Sons, Inc. doi:10.1002/9781118571026
Wozniak, J., Machan, P., Gierlowski, K., Hoeft, M., & Lewczuk, M. (2011). Comparative Analysis of IP-Based
Mobility Protocols and Fast Handover Algorithms in IEEE802.11 Based WLANs. In A. Kwiecień, P. Gaj, & P.
Stera (Eds.), Communications in Computer and Information Science (pp. 87-101). Berlin, Germany: Springer-
Verlag.
Yahiya, T. A. (2011). Understanding LTE and Its Performance. New York: Springer. doi:10.1007/978-1-4419-6457-1
Yan, X. (2010). Optimization of Vertical Handover Decision Processes for Fourth Generation Heterogeneous
Wireless Networks [Ph.D. Dissertation]. Monash University, Australia.
Yan, X., Mani, N., & Şekercioğlu, Y. A. (2008). A Traveling Distance Prediction Based Method to Minimize
Unnecessary Handovers from Cellular Networks to WLANs. IEEE Communications Letters, 12(1), 14–16.
doi:10.1109/LCOMM.2008.071430
Volume 14 • Issue 1 • January-June 2018
26
Adnan Mahmood (BEng., MSc., MEng.) is presently associated with the Faculty of Computer and Information
Technology, Al-Madinah International University, Malaysia and Faculty of Engineering, Universiti Malaysia Sarawak,
Malaysia. He has previously remained associated with the COMSATS Institute of Information Technology, Islamabad,
Pakistan and University of Jinan, Shandong, P.R.China. His particular research interests are in ‘Optimization of
Next Generation Heterogeneous Wireless Networks, Software-Defined Networking, Intelligent Transportation
Systems, and IoT Networks’.
Shadi M. S. Hilles is currently associated in the capacity of Associate Professor / Head of Department with the
Faculty of Computer and Information Technology, Al-Madinah International University, Malaysia.
Hushairi Zen is currently the Deputy Dean (Industry and Community Engagement) with the Faculty of Engineering,
Universiti Malaysia Sarawak, Malaysia.
Yan, X., Şekercioğlu, Y. A., & Narayanan, S. (2010). A Survey of Vertical Handover Decision Algorithms in
Fourth Generation Heterogeneous Wireless Networks. Computer Networks, 54(11), 1848–1863. doi:10.1016/j.
comnet.2010.02.006
Yussuf, A. A., Hassan, W. H., & Issa, S. (2012). A Review of VHD Approaches in Next Generation Wireless
Networks. In Proceedings of the Second International Conference on Digital Information and Communication
Technology and its Applications, Bangkok, Thailand (pp. 363-367). doi:10.1109/DICTAP.2012.6215376
Zahran, A. H., & Liang, B. (2005). Performance Evaluation Framework for Vertical Handoff Algorithms in
Heterogeneous Networks. In Proceedings of the IEEE International Conference on Communications, Seoul,
Korea (Vol. 1, pp. 173-178). doi:10.1109/ICC.2005.1494342
Zekri, M., Jouaber, B., & Zeghlache, D. (2012). A Review on Mobility Management and Vertical Handover
Solutions Over Heterogeneous Wireless Networks. Computer Communications, 35(17), 2055–2068. doi:10.1016/j.
comcom.2012.07.011
Zhang, H., Olariu, S., Cao, J., & Johnson, D. B. (Eds.). (2007). Mobile Adhoc and Sensor Networks, LNCS (Vol.
4864). Berlin, Germany: Springer-Verlag. doi:10.1007/978-3-540-77024-4
Zhang, Y., Yang, L. T., & Ma, J. (Eds.). (2009). Unlicensed Mobile Access Technology: Protocols, Architectures,
Security, Standards, and Applications ~ Wireless Networks and Mobile Communication Series. Florida, United
States: Taylor & Francis Group, LLC.
1 Yan et al.’s (2008) study has been regarded as the first serious effort for the inter-technology (vertical)
handovers. However, certain erroneous assumptions led to a number of mathematical imperfections in
their proposed scheme.
2 WMANs as opposed to WLANs possesses properties such as higher construction cost and smaller
bandwidth, but with a wider transmission range and higher mobility.
