Barja

Computer Communications 34 (2011) 985–997
Contents lists available at ScienceDirect
Computer Communications
journal homepage: www.elsevier.com/locate/comcom
Review
An overview of vertical handover techniques: Algorithms, protocols and tools

Johann Márquez-Barja ⇑, Carlos T. Calafate, Juan-Carlos Cano, Pietro Manzoni
Universidad Politécnica de Valencia, Camino de Vera, S/N, 46022 Valencia, Spain
a r t i c l e i n f o a b s t r a c t
Article history: Wireless technologies, under the ‘‘Anywhere, Anytime’’ paradigm, offer users the promise of being always
Received 16 February 2010 attached to the network. Mobile devices enabled with multiple wireless technologies make possible to
Received in revised form 15 September maintain seamless connectivity in highly dynamic scenarios such as vehicular networks (VNs), switching
2010
from one wireless network to another by using vertical handover techniques (VHO). In this paper we
Accepted 26 November 2010
Available online 4 December 2010
present an overview of VHO techniques, along with the main algorithms, protocols and tools proposed
in the literature. In addition we suggest the most appropriate VHO techniques to efficiently communicate
in VN environments considering the particular characteristics of this type of networks.
Keywords:
Vertical handover
Ó 2010 Elsevier B.V. All rights reserved.
Vehicular networks
IEEE 802.21
Decision algorithms
1. Introduction antee the Quality of Service (QoS) of the services, the On-Board Unit
(OBU) in the car must be able to perform VHO in a seamless manner.
In developed countries the user demand for mobile services is Handover, also know as Handoff, is an event taking place when-
increasing due to the need to access information anywhere, any- ever a mobile node moves from one wireless cell to another, aban-
time. The ever-growing communications infrastructure allows doning the connection with the first base station and getting
connectivity through diverse wired and wireless technologies in attached to the second one. When a handover occurs within the
different environments. domain of a single wireless access technology the process is known
The adoption of wireless technologies is increasing at a very fast as horizontal handover; in contrast, vertical handover is a term that
rate. This trend is basically due to factors such as: (i) the miniaturi- refers to handover among heterogeneous wireless access network
zation of devices such as laptops, Personal Digital Assistants (PDA), technologies. Fig. 1 illustrates both horizontal and vertical hand-
and netbooks, (ii) the multiple networking interfaces available in over events.
most devices, (iii) the availability of several wireless technologies Handover techniques have been widely studied in the cellular
such as Wireless Fidelity (Wi-Fi), Worldwide interoperability for communications domain, and their popularity is increasing among
Microwave Access (WiMAX), and Universal Mobile Telecommunica- IP-based wireless networks [1]. Handover is considered ‘‘seamless’’
tions System (UMTS), and (iv) the emerging mobile applications, when it is able to maintain the connectivity of all applications run-
such as those based on the Web 2.0 paradigm, car navigation, and ning on the mobile device, providing a continuous end-to-end data
location based services. service within the same session during the switchover, offering
The integration of different wireless network technologies is both low latency and minimal packet loss.
needed to provide a ‘‘seamless’’ interoperability, integration and One of the first approaches in order to provide seamless connec-
convergence among these heterogeneous technologies and, there- tivity among heterogeneous networks appeared during the late
fore, the use of vertical handover (VHO) techniques are required. nineties. Stemm and Katz [2] presented an implementation that
Nowadays, the automobile industry is taking advantage of the gathered Mobile IP and routing aspects, together in an application
wireless technologies to improve security and comfort offering to developed to manage the handover process. This implementation
users diverse services such as alert messages, toll payments, info- was based on the technology available in those days that consid-
tainment, etc. In order to avoid any underperformance and to guar- ered the IBM Infrared Wireless LAN, the AT&T WaveLAN and the
Metricom Ricochet Network as in-building, campus, and wide area
underlaying wireless technologies, respectively. These first studies
evidenced that vertical handover is a must when dealing with a
⇑ Corresponding author. Tel.: +34 963 877 000x85720; fax: +34 96 387 7579.
variety of technologies.
E-mail addresses: [email protected] (J. Márquez-Barja), calafate@dis-
ca.upv.es (C.T. Calafate), [email protected] (J.-C. Cano), [email protected]
More recently, various works appeared covering the vertical
(P. Manzoni). handover (VHO) among heterogeneous technologies including:
0140-3664/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.comcom.2010.11.010
986 J. Márquez-Barja et al. / Computer Communications 34 (2011) 985–997
The research community has been making significant efforts

towards the convergence of the different wireless networking
technologies. As a consequence, there are different proposals
addressing heterogeneous scenarios, protocols, handover tech-
niques and algorithms, network technologies, metrics, and proce-
dures. In addition, since 2004, the IEEE 802.21 Working Group
has been working in the Media Independent Handover Services
WiMAX
Protocol [46] whose purpose is to provide an homogeneous
(802.16) function-interface between heterogeneous network technologies,
offering standard handover services between lower and upper lay-
ers. The IEEE 802.21 standard was finally approved in November
2008. Nevertheless, there have been proposals and studies
addressing the performance of 802.21 and offering improvements
in terms of VHO effectiveness [47,37,48].
In this work we survey the most significant proposals found in
the literature concerning VHO techniques, including both particu-
UMTS (3GPP)
lar proposals and standards. In addition, we emphasize on the VHO
process from a vehicular networks (VNs) perspective, highlighting
those techniques and algorithms that fit better to this type of net-
works. This survey can be useful to the research community since
most proposals in the literature, merely evaluate specific VHO
techniques per se. Notice that we use the acronym VNs referring
to wide area networks where vehicles are interconnected through
a base station, and not referring to Vehicular Ad-hoc Networks (VA-
Wi-Fi (802.11) NETs) where the vehicles are connected among them using a short-
range wireless technology.
Vertical Handover The rest of this paper is organized as follows. Sections 2 and 3
Horizontal Handover present a general overview of VNs and the VHO process, respec-
tively. In Section 4 the handover information gathering phase is de-
Fig. 1. Illustration of the horizontal and vertical handover processes. scribed. The handover decision phase and the diverse handover
algorithms are presented in Section 5. Section 6 refers to the hand-
UMTS and Wi-Fi [3], Wi-Fi and Wireless Broadband (WiBro) [4], over execution phase and the different processes involved in the
Wi-Fi and WiMAX [5], WiMAX and UMTS [6], WiBro and UMTS handover execution. The main testbeds, simulators and tools for
[7], WiMAX, Wi-Fi and UMTS [8], Bluetooth (BT) and Wi-Fi [9], VHO evaluation found in the literature are collected and presented
broadcast communication technologies like Digital Video Broad- in Section 7. Section 8 presents some guidelines and open research
casting (DVB) and Multi media Broadcast/Multicast Service issues for the development of efficient VHO systems in vehicular
(MBMS) [10], or even UMTS and Low Earth Orbit (LEO) satellite environments. Finally, the main conclusions are presented in
[11]. Considering networking technologies, Table 1 shows the tech- Section 9.
nologies used by the different proposals. As observed, most pro-
posals (72.9%) evaluate the VHO using only two technologies, 2. Vehicular network overview
being that the remaining 27.1% of the proposals have considered
three technologies instead. In addition, about 53.6% of the propos- VNs offer many application branches which can be classified as:
als focus on evaluating the VHO viability between Wi-Fi and UMTS. (i) safety such as emergency warning systems for vehicles, transit
The main drawback of this broad variety of solutions stands in the or emergency vehicle priority signaling, and (ii) non-safety applica-
fact that none proposes a unique homogeneous approach that can tions. As examples of the latter we have cooperative adaptive
be adapted to all the available wireless technologies. cruise control, electronic parking and toll payments, infotainment
Table 1
Networking technologies used in VHO studies.
Proposals % Networking technologies

Wi-Fi WiMAX WiBro UMTS GPRS Bluetooth Ethernet
Two technologies
[12–31,3,32] 53.6 w w
[33] 2.5 w w
[9] 2.5 w w
[6,8] 4.9 w w
[34] 2.5 w w
[4] 2.5 w w
[35] 2.5 w w
[36] 2.5 w w
Three technologies
[37–40,5,41,42] 14.7 w w w
[43,44] 4.9 w w w
[1] 2.5 w w w
[7] 2.5 w w w
[45] 2.5 w w w
J. Márquez-Barja et al. / Computer Communications 34 (2011) 985–997 987
services, and content delivery, among others. Within VNs a vehicle so VHO strategies should focus mostly on technologies such as
is considered as a node of the network, being equipped with multi- UMTS, WiMAX and WiBro.
ple interfaces that provide access to different technologies such as
Global Positioning System (GPS), Wi-Fi, WiMAX and UMTS. Vehi-
cles are able to communicate among them and with their base sta- 3. Vertical handover overview
tions (access points or point of attachments) under the Ad-hoc or
the infrastructure mode [49], respectively. Notice that we focus An accurate VHO process should take into account and care
our study on access-point-based networks (VNs) and not on the about the service continuity, network discovery, network selection,
ad hoc mode (VANETs). Further information about VANETs could security, device’s power-management and QoS issues [9,33,36],
be found in [50]. focusing mostly on the latter. Several proposals [25,26,58] split
The main specific characteristics of VNs can be summarized in: the VHO process into three parts: (i) Handover information gather-
(i) topology restrictions, (ii) mobility patterns, (iii) power consump- ing, (ii) Handover decision, and (iii) Handover execution. Fig. 2
tion, (iv) scalability, (v) node reliability and (vi) speed [49]. VNs are shows the interactions among the three phases required to imple-
geographically constrained to the roadmap layout; consequently, ment handover in heterogeneous networks.
the topology is heavily dependent on the roadmap or the vehicular
scenario. Nevertheless, this restriction allows to improve network 3.1. Handover information gathering
performance or VHO decisions allowing to predict mobility.
Concerning power consumption, wireless devices are traditionally The handover information gathering phase collects not only
limited by the battery autonomy. However, in vehicular network network information, but also information about the rest of the
(VN) environments, the devices are continuously powered by the components of the system such as network properties, mobile de-
vehicle’s energy system; therefore, there are no energy constrains vices, access points, and user preferences. For that reason this
limiting their performance. It is important to mention that VNs phase receives different names: handover information gathering
can reach a high number of nodes, which requires them to be very [24], system discovery [59], system detection [60], handover initiation
scalable. Finally, node reliability must be taken into account since a [61,18] or simply network discovery [62,48,43]. In this phase, the
node may quickly join or leave the network coverage area depend- information is collected to be used and processed for making deci-
ing on the speed and the destination of the vehicle. sions in the handover decision phase. The information typically
Several worldwide projects are being developed and promoted collected is the following:
in order to provide solutions to reinforce safety and to extend
the use of non-safety applications on cars and highways through Availability of neighboring network links by offering informa-
the use of wireless technologies. Industry, researchers and govern- tion such as throughput, cost, packet loss ratio, handoff rate,
ments are working towards a standardization for those solutions: Received Signal Strength (RSS), Noise Signal Ratio (NSR), Carrier
Europe through projects and consortiums such as Car2Car Commu- to Interference Ratio (CIR), Signal to Interference Ratio (SIR), Bit
nication Consortium, Network on Wheels, and Safespot; USA Error Ratio (BER), distance, location, and QoS parameters.
through the Vehicle Safety Consortium (VSC), Vehicular Infrastruc- The Mobile device’s state by gathering information about bat-
ture Integration (VII) and Intelligent Vehicle Highway Systems tery status, resources, speed, and service class.
(IVHS); finally, Japanese projects Advanced Safety Vehicle (ASV) User preferences information such as budget and services
and Advanced Highway Systems (AHS) focus their efforts on set- required.
ting common issues such as frequency allocation, protocol defini-
tion and infrastructure deployment [51].
Wireless Access in Vehicular Environments (WAVE) [52] is a
standard developed for VNs, to satisfy the communication needs
of a large class of applications using the Dedicated Short Range
Communication (DSRC) spectrum [53]. The WAVE protocol archi-
tecture encompasses the IEEE P802.11p standard [54] for the
PHY/MAC layers and the IEEE P1609 [52] family of protocols for
the application layers. These protocols must be efficient and reli-
able in order to provide safety and comfort services to passengers
via context sensitive applications [55,56], as well as low latencies
to deliver contents within acceptable time bounds.
The International Organization for Standardization (ISO) is also
working towards a seamless connectivity solution for a wide vari-
ety of wireless access technologies. This proposal, called Continu-
ous Air Interface for Long and Medium Range (CALM M5), has
been specifically designed to reduce the end-to-end latency com-
munication in mobile environments such as VNs [57], and also
works on top of IEEE 802.11p.
Referring to VN environments, we consider that VHO strategies
must be designed and applied depending on the underlying
network technologies geographically available and the expected
lifetime of each possible connection. Therefore, at least two
scenarios might be evaluated: (i) Urban scenarios and (ii) Highway
scenarios. Concerning urban scenarios, multiple short and wide
range wireless technologies such as BT, Wi-Fi, General Packet
Radio System (GPRS), UMTS and WiMAX may be considered in
the design of the VHO strategy. In highway scenarios the predom-
inant wireless technologies are those that cover a wide range, and Fig. 2. Handover management procedure.
Section 4 describes this phase along with the techniques used tocol stack in order to cover all the possible information sources.
by the different proposals to perform the data gathering task. Moreover, Table 3 presents diverse parameters used by different
authors in their proposals and works.
3.2. Handover decision Different proposals addressing this phase are based not only in
monitoring different layers, but also by implementing events and
The handover decision phase is one of the most critical pro- triggered notifications. Attaullah et al. [8] present a trigger man-
cesses during the handover. This phase is also know as system selec- agement system that monitors and collects multi-level parameters.
tion [26], network selection [25,43] or handover preparation [61]. Similarly, works [16,4] present modules called ‘‘Link Layer State’’
Based on the gathered information, this phase is in charge of decid- and ‘‘Network Connection Module’’, respectively, where both per-
ing When and Where to trigger the handover. The When decision re- form the same monitoring task. An xml-based process is used by
fers to the precise instant in time to make an optimal handover, a CORBA communication middleware to gather information and
while the Where refers to selecting the best network fulfilling our to notify events in [64]. Seigneur et al. [65] use different operating
requirements for the switching. system’s Application Programming Interface (API) to monitor the
In a homogeneous network environment deciding When to device and the network in order to evaluate the electrosmog expo-
handover usually depends on RSS values, while the Where is not sure to trigger the vertical handover (VHO).
an issue since we use the same networking technology (horizontal Several proposals [48,58,21,66–72] rely on the Media Indepen-
handover). In heterogeneous networks the answer to these ques- dent Information Service (MIIS) mechanism offered by the IEEE
tions is quite complex. To make the best decisions the information 802.21 standard for this phase, which allows interaction with a
gathered must be evaluated taking into account many parameters wide set of wireless technologies in a unified manner. The 802.21
obtained from the different information sources, i.e. network, mo- addresses the optimization of both network detection and selec-
bile devices, and user preferences. Vertical Handover Decision tion by providing a source of extensible and semantically defined
Algorithm (VHA) are used to weight up and evaluate the parame- information to facilitate optimized handover decision making
ters involved under each specific criteria. As examples of algo- [61]. As shown in Fig. 3, the IEEE 802.21 offers a middleware pro-
rithms allowing to evaluate cross-layer multi-parameters we tocol called Media Independent Handover Function (MIHF) that is
have techniques such as fuzzy logic, neural networks, and pattern able to encapsulate the different underlaying network technologies
recognition, among others. Section 5 presents a brief description (e.g. 802.3, 802.11, 802.16, 3GPP, and 3GPP2) to the upper layers,
and a taxonomy of the VHA found in the literature. allowing the handover management process to operate indepen-
dently of the physical and data link layers.
3.3. Handover execution The MIHF protocol defines the messages exchanged between
peer MIH entities for handover, offering a common message pay-
This phase performs the handover itself; besides performing the load across different media (802.3, 802.11, 802.16, Cellular). The
handover, the phase should also guarantee a smooth session tran- standard refers as lower layers to the technology dependent
sition process. In order to perform the VHO different handover components, and as upper layers to the requesting modules. These
strategies cooperate with control signaling, and the IP manage- lower layers can be accessed by different functions to retrieve
ment protocols. This phase is usually known as Handover execution information to detect, prepare and execute the vertical handover
[59,24,58], but it also receives the name of VHO assessment [18] or (VHO), while the upper ones demand that information; therefore,
Handoff implementation [62]. A detailed description of the execu- they are also referred to as Media Independent Handover User
tion phase is presented in Section 6. (MIHU). The MIIS allows the MIHF to discover its network environ-
Concerning VNs, the performance of each phase must be fo- ment by gathering information that the upper layers use to make
cused on the distinctive characteristics and features of such type decisions. Selected information refers to the list of available net-
of networks. The information gathering phase must consider the works, location of the point of attachment (PoA), operator ID,
dynamism of the available information at the devices and the net- roaming partners, cost, security, QoS, PoA capabilities, and vendor
work. Making decisions based on highly dynamic information with specific information, among others.
a given degree of the device’s mobility requires a quick and reliable In heterogeneous vehicular networks it is important to consider
decision algorithm. Finally, the execution of the VHO must be care- the environmental conditions when performing the vertical hand-
fully controlled to achieve accuracy by considering the geographi- over (VHO). A high mobility degree as well as the mobility patterns
cal location, the selected network and the precise time. will significantly affect the reliability of measurements and the
decision-time. Therefore, vehicular mobility patterns and speed
4. Handover information gathering phase
Table 2
In order to perform an ‘‘always best connected’’ handover [63], a Information parameters pertinent to the VHO process.
full set of information is gathered and provided to the decision Layers Parameters
phase. To collect the available information from different sources,
Application User preferences (e.g. cost, provider)
the mobile device surveys the surrounding networks in order to Context information (e.g. speed)
discover services, data rates, and power consumption. As a comple- QoS parameters (e.g. bw offered, delay, jitter)
ment to the information gathered through scanning, networks may Security alerts (e.g. notifications)
also advertise their supported services and QoS parameters, while Transport Network load (e.g. bw available)
the device information is also collected, i.e., speed, battery status, Network Available foreign agents
features, and so on. Finally, information concerning user prefer- Network pre-authentication
ences is also a relevant element to the decision-making process, Network configuration
Network topology
mostly due to its impact on the end user’s satisfaction.
Routing information
Gathering information reliably is crucial for the VHO process
Data-link Radio access network conditions
since the decision-making procedure relies on that data. Table 2
Link parameters
presents the information that should be taken into account in order Link status
to maximize the benefits of decision-making. It clearly shows that
Physical Available access media
information should be collected at each and every layer of the pro-
Table 3
Parameters and metrics used for VHO.
Source Domain Metric Description Used as: input parameter Used as: evaluation metric
Network Latency Latency e2e, End-to-end latency, from packet [11,38,44,58] [38,48,44,6,40,31,20]
Average delay generation to packet reception
Network delay Time taken to deliver a packet [24,44,40,25,98,5,62,99–101] [9,36,102]
Coverage Network Related to the area covered by the [48,24,45,4,73,99,81,69]
availability, network
coverage
Performance RSS Received signal strength [37,1,58,22,24,17,23,44,39,29–31,
7,43,14,45,76,35,21,20,18,19,
8,59,66,79,81,70,103,101]
CIR Carrier interference ratio [22,44]
RTT Round trip time [22,58,44,39,16,98,99,81]
Retransmissions Packets retransmitted [44,16]
BER Bit error rate [38,24,22,44,9,76,25,5]
SINR Signal to interference and noise ratio [28,9,76,18,81,101]
Security The security level deployed in the [38,44,75]
network
Packet loss Packets unsuccessfully delivered [24,44,40,98,62,81] [6,98,36,104,100–102]
Throughput, data Throughput in the network [11,105,8,5,41] [1,28,39,30,9,31,32,14,76,
rate 16,98,36,18,15,81]
Bandwidth The bandwidth available at a precise [58,38,11,30,77,40,31,25, [21,69]
available instant in the network 21,8,59,62,81,100,101]
Bandwidth The bandwidth offered by a candidate [24,33,40,43,19]
offered, link network
capacity
Network Jitter Jitter introduced by the network [24,44,75,40,25,98,8,5,81] [36,102]
Network overhead Extra management packets in the [3]
network
Billing Cost/money The price for using the network [38,24,44,75,30,9,77,40,43, [38,40,12]
25,59,41,62]
Miscellaneous Number of users Amount of users connected in the [11,12]
network
User preferences Budget User budget to spent in using [24,25,62]
networks
Preferred network User choice [24,48,11,25,73,62,81]
Mobile terminal Location Mobility, location Geolocation or AP-based location [24,17,39,29,7,18,73,62,66,70]
Movement The change of movement direction [7,17,62,66]
Speed, velocity The speed while moving the mobile [24,7,11,75,31,43,62,66,81,101]
terminal
Capabilities Network Technologies available in the device All proposals referred
interfaces
Battery Power consumption [24,11,44,75,77,43,25,21,73,62,101] [21]
consumption
VHO Occurrence Number of VHO VHO events performed [58,7,39,105,29,30,9,77,
7,43,45,76,18,79]
VHO success rate The ratio of VHO events successfully [37] [101]
performed
Latency VHO Latency Time associated with the VHO [1,37,27,14,13,34,4,66,
process 103,100,106,107]
Gathering latency Time associated with the VHO [37,27,34]
Gathering phase alone
Decision latency Time associated with the VHO [37,27,34]
Decision phase alone
Execution latency Time associated with the VHO [37,27,34]
execution phase alone
Dwell time The period during which a VHO [22,58,14,20]
process remains halted in order to
allow another process to take place
Network VHO packet loss Undelivered packets during the VHO [48,23,103]
process
must be carefully taken into consideration. Additionally, context requirements, thus providing the desired seamless communi-
and location information are also important to perform an accurate cations.
handover in vehicular networks. This information can be retrieved To make an accurate decision, this phase takes advantage of
from the vehicle’s OBU. algorithms that, considering the information available, perform
an evaluation process in order to obtain the best choice for hand-
over execution. These algorithms are usually called vertical hand-
5. Handover decision phase over decision algorithms, or simply vertical handover decision
algorithm (VHA). In the literature we can find several VHA propos-
The decision phase can be considered as the core phase of the als. Some of them take into account, in a straightforward manner,
VHO since it is in charge of evaluating and deciding the most the handover decision task by considering only the lower layers’
appropriate network choices in order to fulfill both system and user information given by the media independent information service;
Upper Layers (L3 and above)
MIH Users
SIP MIPv4 MIPv6 HIP ....
Mobility Management Handover Management
MIH MIH Information

Events Commands Service
MIH_SAP
802.21 MIH Function
Smart Handover Information

Triggers Messages Service
MIH_Link_SAP
Link Link Information

Events Commands Service
Lower Layers (L2 and below)
Protocol and Device Hardware

Fig. 4. Vertical handover algorithms taxonomy.
802.3 802.11 3GPP 802.16 ....

selected for processing, these algorithms can be considered as:
basic or advanced.
Concerning advanced algorithms, Pawar et al. [73] present a
Fig. 3. IEEE 802.21 architecture. mobile service platform with a context-aware middleware. It takes
as context-sources parameters the user trip information, device
services, network services, user preferences, device specifications,
most of the proposals combine the metrics and parameters of the
and time. Once the information is collected, a QoS predictor per-
diverse components to build an accurate cross-layer handover
forms a path prediction to ensure the end-to-end QoS. The whole
algorithm.
information gathered and predicted is evaluated under the Ana-
The decision phase basically consists of three steps: (i) param-
lytic Hierarchy Process (AHP) algorithm (an algorithm of the Multi-
eter selection, (ii) parameter processing, and (iii) parameter aggre-
ple Criteria Decision-Making family. This type of algorithms will be
gation. The parameter selection step takes into account only those
described in Section 5.3). Similarly, other works [24,25] present an
parameters, that the algorithm uses, to evaluate and weight a
autonomic architecture that evaluates the context information
candidate connection. The parameter processing step allows to
through a combination of multiple criteria and policy-based algo-
normalize all the parameters; value parameters with diffuse infor-
rithms in order to perform the best decision. Lassoued et al. [38]
mation are merged using several techniques such as fuzzy logic,
present a VHO evaluation methodology which combines different
neural networks and specific functions, to extract relevant data.
algorithms to tune the decision. Despite of the diverse algorithms
Finally, to make a decision, an algorithm aggregates and evaluates
used, this methodology has a context-aware core to collect and
the weight of each parameter and, based on some decision criteria,
process the information.
selects the best candidate. Although the process described applies
The Location-Aware algorithms are a basic type of context-aware
to most proposals, authors typically modify this process slightly to
algorithms [38,73,74]. These algorithms, take as primary informa-
fit their needs.
tion the events triggered by changes in the context, focusing on
Based on the decision strategies and algorithms analyzed in the
location changes alone. The work in [18] presents a location-based
literature, we classified the algorithms according to the taxonomy
VHA approach that combines mobile location information and net-
shown in Fig. 4. We now proceed to describe each of these algo-
work information in order to avoid the ping-pong effect related to
rithms in detail.
performing predictions at wireless cells’ boundaries, thus improv-
ing the VHO latency. With this purpose the authors introduce a
5.1. Parameter selection algorithms hysteresis cycle in the decision process, evaluating the mobile de-
vice’s location through the network goodput, a metric based on the
This type of algorithms takes advantage of the context informa- allocated bandwidth to the device for the serviced requested. Joe
tion, generating knowledge to perform an accurate decision. Any and Hong [7] present a mobility prediction VHA solution based
changes in the mobile and network context trigger events and pro- on two location schemes: (i) a sector and zone scheme, which con-
cesses, which are notified to the VHO decision phase by the infor- sists in dividing the zones into several sectors and predicting the
mation gathering phase. Depending on the number of parameters mobility based on statistics to decide the best network candidate,
and (ii) a scheme based on GPS location information. Finally, the Nevertheless, the gathered information is often imprecise; to han-
decision is made based on the mobile device’s speed and mobility dle this issue, fuzzy control theory and neural networks techniques
pattern regardless of the scheme used to determine the pattern. are applied [26,58]. Usually, these algorithms are applied first in
Similarly, the work in [23] takes into account the device’s location order to convert imprecise data into precise data. Afterwards, a
information based on the packet arrival time and the RSS, a typical Multiple Criteria Decision-Making (MCDM) algorithm is fed with
metric for horizontal handover. Once the information is collected, these data to determine the best choice. Kassar et al. [24,26,25]
it is processed by a multiple criteria algorithm. A framework for combine fuzzy logic procedures with classical MCDM procedures.
location-aware VHO based on GPS location management and Similarly, the work in [30] combines fuzzyfication processes with
wake-up procedures is presented in [14]. This framework also al- Grey Relational Analysis (GRA) techniques. A fuzzy-based algo-
lows to evaluate cross-layer VHA combining location-aware and rithm is presented in [77] which combines fuzzy logic and a cost
power-based algorithms or dwell-timer based algorithms. The function-based algorithm in order to make the best decision. The
work in [39] presents an adaptive VHO technique based on predic- work in [59] presents an evaluation and review of fuzzy-based
tive RSS patterns. Finally, Kwon et al. [17] present a classification of algorithms taking into account parameters such as RSS, Cost and
location-aware algorithms based on RSS, movement extrapolation, Bandwidth. In order to improve the QoS of a real-time application,
history data, mobility pattern and distance to the access point. the proposal in [8] combines a fuzzy-based algorithm with a rule-
base algorithm to make an accurate decision. Finally, Horrich et al.
5.2. Parameter processing algorithms [31] presents a neural networks knowledge based algorithm that
fuzzyfies imprecise information in order to score the alternatives.
The parameter processing algorithms are in charge of the pro- These methods may be helpful at combining the different
cessing of the chosen parameters and of the provision of the input sources of information to extract relevant information, since in
information to the parameter aggregation algorithms. The evalu- mobile environments, such as VNs, high speeds make the gathered
ated works use diverse functions to process the information information to have low reliability. However, the applicability of
depending on the its nature; therefore, the functions used could this type of algorithms can be reduced if computational times in-
vary from pure mathematical to computational algorithms. volved become too high.
5.2.1. Mathematical algorithms 5.3. Parameter aggregation algorithms

Most mathematical algorithms are self-designed algorithms.
Some authors propose their own self-design decision algorithms Since VHO mechanisms take into account diverse metrics and
in order to satisfy their VHO needs based on the information avail- parameters to evaluate the best candidate networks, there is a
able in their systems. need for algorithms that are able to jointly handle multiple param-
Stevens et al. [58] present an algorithm based on a Markov deci- eters and metrics, as shown in Table 2. Consequently, Multiple
sion process using two types of functions: there is a link-reward Criteria Decision-Making (MCDM) algorithms are adopted to fulfill
function associated with the QoS received by the connection, and this requirement by aggregating all these processed parameters.
a signaling-cost function associated with the signaling overhead MCDM algorithms include algorithms based on multiple attri-
and latency when the handover execution is performed. They have butes or multiple goals. Multiple Attribute Decision-Making (MADM)
evaluated the performance of the algorithm using voice and data algorithms evaluate the different alternatives based on their attri-
traffic. This algorithm can work in conjunction with the 802.21 butes, while Multiple Objective Decision-Making (MODM) algorithms
framework. Similarly, the work in [40] proposes two Markov deci- focus on diverse objectives that can not be reached simultaneously
sion approaches based on rank aggregation, where the top [26]. Usually, both types of algorithms are generally called MCDM.
weighted network is selected. This selection process is similar to We now proceed to briefly describe the most popular algorithms:
the MCDM technique for order preference by similarity to ideal
solution (TOPSIS) [58], but obviates the ideal network comparison. 5.3.1. Analysis hierarchy process (AHP)
This proposal also works in conjunction with IEEE 802.21. Hasswa This type of algorithms is based on the divide-and-win para-
et al. [75] present a decision function algorithm which allows users digm. The main decision problem is divided into sub-problems,
to prioritize the available networks based on network perfor- where each sub-problem is evaluated as a decision factor. From
mance, user preference and monetary costs. Based on routing is- the set of alternative solutions, AHP finds the most optimal solu-
sues, Liao et al. [3] present a VHO cost-function in order to tion [26,23,78].
optimize the VHO latency. Works such as [39,76] also present a
cost-based optimal network selection function based on a cross-layer 5.3.2. Grey relational analysis (GRA)
adaptive scheme, that takes into account parameters from different This mathematical algorithm builds a grey relationship be-
layers. A framework with a normalized function in its decision tween elements (networks), one of them with the ideal quality val-
module is presented in [4], which normalizes the incoming values ues. So, the rest of the elements are compared and evaluated
in terms of the RSS, QoS, and user preferences to select the best against the ideal solution. The option that better approaches this
candidate network. A score function is used in a smart decision ideal solution receives the highest score [26,23,38,44].
model to decide the best network interface and the best time to
perform the handoff. This decision-making model performs a 5.3.3. Technique for order preference by similarity to ideal solution
smart decision based on various factors, such as the properties of (TOPSIS)
available network interfaces, the system information, and the user Similarly to GRA algorithms, TOPSIS algorithms consider and
preferences [60,1]. ideal solution for performance comparison, considering as the best
alternative the one nearest to the ideal solution, and as worst the
5.2.2. Computational algorithms one furthest from such solution [58,38,26].
Computational algorithms use fuzzy logic and neural networks
techniques to interpret imprecise information, although authors 5.3.4. Simple additive weighting (SAW)
usually apply their own self-designed algorithms to perform the SAW algorithms are frequently used when MCDM is applied.
interpretation task. The above algorithms require precise data in This technique consists in scoring each alternative by adding the
order to weight attributes and to perform an accurate decision. attribute values multiplied by its weight, in order to score the
overall alternative, being the highest score the most optimal choice 6.2. Mobility management
[44,38,26,79].
One of the key issues of the seamless handover concept is
mobility management. In IP-based networks, the standard proto-
5.3.5. Multiplicative weighting exponent (MWE)
cols designed for mobility become an important solution to main-
MWE works similarly to SAW algorithms. To score the overall
tain the session alive when targeting a seamless handover. These
alternative, it uses the weighted product of all attributes. Since this
protocols usually work on the intermediate layers of the TCP/IP
product does not have an upper-bound, it is advisable to compare
protocol stack. The most common protocols used for mobility in
the score against an ideal solution [44,38].
VHO are the Mobility support for Internet Protocol v.4 (MIPv4)
An interesting comparison and evaluation of MCDM algorithms
[84], Mobility support for Internet Protocol v.6 (MIPv6) [85], Ses-
and their performance is presented in [44]. This evaluation takes
sion Initiated Protocol (SIP) [86], Network Mobility Basic Support
into account basic network parameters such as BER, delay, jitter
Protocol (NEMO) [87] and Host Identity Protocol (HIP) [88].
and bandwidth, suggesting that most of the algorithms have a sim-
A study of the different techniques for mobility management in
ilar performance, depending heavily on the weight assigned to
IP-based wireless networks is presented in [89,90]. These works
each parameter or attribute.
cover different cross-layer methods, as well as particular solutions
depending on the layer and the wireless technology.
5.4. Comments on algorithm applicability to VNs The work in [4] evaluates a VHO architecture covering Wi-Fi
and WiBro wireless technologies achieving low handover latencies
In the context of VNs, information about location and mobility through an improved implementation of the Mobility for IP (MIP)
patterns are of utmost importance due to the high speeds involved. protocol. Similarly, the proposal in [20] appeals to MIP to manage
Thus, any handover decision scheme proposed should account for the mobility in a new VHO proposal in boundary conditions for
this information. UMTS and Wi-Fi. The handover management among different IP
The most appropriate VHA scheme for vehicular networks must versions is studied in [91], presenting a scheme to support IPv4–
include advanced context information and particularly location- IPv6 traversal between domains considering routing issues as well
aware information. It is important to consider location and also as IP-related details. In contrast, Andersson et al. [92] present a
location predictions in order to choose the best network candidate mobility management scheme which avoids tunneling by main-
in the area. Nevertheless, context information adds valuable infor- taining the same IP while moving across different subnets.
mation to the VHA. To combine the different sources of informa- A proactive seamless handover scheme based on 802.21
tion a fast MCDM algorithm is recommended, being in charge of features and SIP is evaluated in [48]. This scheme offers a proactive
weighting the set of parameters and providing a fast decision on authentication in order to early obtain an IP and, thanks to SIP
how to perform the VHO at the best time and place. Nevertheless, management, maintain the session during a seamless handover.
a self-designed algorithm, defined according to the special charac- Work [13] combines the use of SIP and the stream control trans-
teristics of VNs, can also be used to make decisions. mission protocol (SCTP) in order to maintain the session in a mul-
tihoming environment, taking advantage of the multithreading
features of the SCTP, despite of the redundant retransmission that
6. Handover execution phase
the latter implies [93,15,94]. Howie et al. [95] combines the use of
SIP at the application layer and MIP at the network layer to provide
The third phase of the VHO process focuses on execution. This
a supernetworking mobile application able to perform fast hand-
phase is in charge of committing the VHO itself. So, once the infor-
overs. An identification middleware layer is proposed by Wong
mation is collected in phase one and processed in phase two by
et al. [96] decoupling the host identification from the location.
selecting the network candidate, the execution phase will trigger
Thus, allowing the host to move across domains not interrupting
a network binding update. With this purpose this phase is con-
the sessions.
cerned with control, security, session and mobility, among other is-
Concerning security, the work in [97] performs an analysis of
sues in order to perform a seamless handover.
HIP based mobility and triggering using a hardware/software test-
We now briefly describe the main processes participating in this
bed to evaluate VHO techniques; the authors also study how
phase:
mobility triggers affect latency. Latvakoski et al. [32] evaluate the
VHO focusing on VoIP in an Ad-hoc environment under HIP and
6.1. Handover management SIP protocols to provide continuously secure communication.
When operating in vehicular environments both latency and
In the handover process there must be an entity in-charge of security issues must be addressed. Due to the characteristics of
controlling the VHO process. Usually, the handover can be either VNs in terms of speed and node reliability, the execution of the
network controlled or mobile controlled. In the former case it is ini- VHO must not add excessive latency. Although solutions based
tiated and controlled by the network, a solution that is typically on Mobile IP seem to be the most natural and adequate, further
adopted by operators to achieve load balancing duties and traffic scrutiny is required to assess which solution is able to offer the
management, among others [80]. In the latter case (mobile con- best trade-off.
trolled) the VHO is initiated and controlled by the mobile device.
This type of control is the most common case, usually based on
user preferences [73,18,81]. Also, the VHO can be network assisted, 7. VHO evaluation
referring to the VHO initiated by the mobile device but assisted by
the network making use of the information services, or mobile In this section we briefly describe the most representative VHO
assisted when it is initiated by the network but assisted by the mo- evaluation methodologies found in the literature. Moreover, simu-
bile device [61,82]. lators and testbeds frequently used in the different proposals are
During the handover process, when a mobile device reaches a also described. As mentioned before, most proposals do not evalu-
new point of access, the system might execute procedures to man- ate VHO techniques from a VN’s perspective; therefore, most of the
age the connections. These procedures usually perform Registra- following works and tools evaluate the VHO in environments other
tion, Association, Re-association, and Dissociation tasks [82–85]. than VN ones. Nevertheless, in this section we also point out how
some of them are able to evaluate the VHO in a VN environment by that are dropped during the VHO process. This metric is described
modifying or adapting their components to VN characteristics. by Eq. (3):
1 Packetsreceived
7.1. Evaluation methodology Packet LossVHO ¼ : ð3Þ
Packetssent
The VHO evaluation done by the different authors follows their Notice that packet loss should be calculated solely for the VHO time
specific research objectives, and so the input parameters, method- period.
ologies and metrics adopted by the different authors are quite het-
erogeneous. Such drawback impedes a fair comparison between
7.1.4. VHA cost
proposals, thus limiting their usefulness and hindering research
The metrics shown above can be used when performing simula-
in this field. As an evidence of the heterogeneity of metrics found
tions or executing implementations on real frameworks. Thus, they
in the literature, Table 3 shows a collection of metrics gathered
can be applied a posteriori in order to evaluate a proposal. How-
from different works in the VHO-related literature. As can be ob-
ever, a priori metrics are also required in order to pre-evaluate or
served, the metrics and parameters are classified by information
compare proposals. To do so, we propose the use of metrics based
source (i.e. network, user preferences, mobile terminal and VHO it-
on the degree of complexity for memory and time where complex-
self) and also by the domain of usage (e.g. coverage, security, loca-
ity is represented in the O(f(n)) format (e.g. O(log2n), O(n), O(n2),
tion, etc.). However, some metrics are used not only as an
etc.).
evaluation metric, but also as input parameters. Moreover, differ-
ent metrics are used for the evaluation of the same parameters,
or simply the same parameters receive different names. After an 7.2. Simulation tools
extensive review of the current literature, we conclude that there
is a lack of evaluation methodologies considering homogeneous For evaluation purposes, the proposals use diverse simulation/
metrics and methods. As far as we know, only Lassoued et al. emulation tools, as well as specific and self-designed testbeds
[38] present an attempt of an evaluation methodology. Their meth- and hardware/software implementations. Most of the authors rely
odology takes into account parameters and metrics related to both on widely used simulators, especially network simulators, in order
upper and lower layers of the vertical handover architecture, such to reproduce network conditions and schemes. Nevertheless, there
as the current context, the user mobility models, the user prefer- are no simulators specifically designed for VHO evaluation;
ences, along with wireless access technologies specificities and therefore, most proposals make some adjustments or develop
QoS parameters. Nevertheless, in order to avoid all the undesired modules to mitigate this shortage. According to Table 4, the
issues, and to unify and clarify concepts, we now propose a set of research community is adopting several tools to aid them in their
metrics to harmonize evaluations in this field. endeavors.
The aforementioned proposals evaluate the VHO by using
7.1.1. End-to-end latency network simulators indirectly. Consequently, simulators such as
The end-to-end latency refers to the time taken to deliver a data network simulator (ns-2) [98] and OPNET [99] are used to simulate
packet from the source to the destination. This latency implies the the network environment, although none of them offers tools to
network latency and the VHO handover latency, as shown in Eq. (1). evaluate the VHO itself. To do so, authors modify the code or con-
Through this metric we are able to compare proposals considering figure the simulators to reach their goals. Referring to VHO models,
not only the decision algorithm’s performance, but also the condi- MATLAB [100] is used to evaluate mathematical models that de-
tions and performance of the chosen candidate network. scribe the VHO process, usually VHA performance. Finally, some
authors build their own framework to fit their needs and interests,
Latencye2e ¼ LatencyNetwork þ LatencyVHO : ð1Þ as shown in Table 4.
Nevertheless, in order to provide a heterogeneous environment
to study the VHO process, only a few modules, testbeds and archi-
7.1.2. VHO latency
tectures have been developed. We briefly describe the most popu-
This metric only takes into account the time associated with the
lar ones:
VHO process itself, considering the three phases of the vertical
Seamless and secure mobility [101]. This simulation frame-
handover (VHO): information gathering, decision and execution
work is being developed for ns-2 in order to provide modules for
phases. It is defined as follows:
several wireless and wired technologies such as IEEE 802.3, IEEE
LatencyVHO ¼ LatencyGathering þ LatencyDecision þ LatencyExecution : 802.11b, IEEE 802.15.1, and UMTS to simulate heterogeneous envi-
ð2Þ ronments. This free and opensource tool is a powerful solution to
evaluate the VHO in VN environments since ns-2 is able to model
multiple network topologies and mobility protocols. This module
7.1.3. VHO packet loss provides VHO management based on 802.21, as well as wide area
Merely comparing the time consumed by the VHO is not en- wireless technologies and protocols. When combined with ns-2,
ough. In fact, it also important to compare the amount of packets the seamless andsecure mobility solution, in addition to the flexibility
Table 4
Evaluation tools used in the literature.
Type Tool/scale % Proposals

Simulator ns-2 14.3 [37,20,33,6,36,106]
OPNET 4.1 [31,7,13]
MATLab 11.9 [28,107,29,11,18]
Self design 2.4 [43]
Others 7.1 Qualnet [17], seamless and secure mobility [36], CanuMobiSim [38]
Testbed Short scale 31.3 [44,58,23,30,40,19,8,4,15,16,5,48,76,108,109]
Medium scale 25.7 [22,26,17,9,32,21,34,41,26,91,12]
and freedom of open source, provides a wide set of simulation pos- Car-2-Car communication. This proposal consists of two main
sibilities within VN environments. components: tower component and car component, where both
ns2-Miracle [102]. As its name shows, this simulation frame- components are based on the current GeoNET implementation,
work is also designed to work with the ns-2. The Muli-InteRfAce although enhanced with the features offered by the IEEE 802.21
Cross Layer Extension (MIRACLE) is a set of dynamic libraries which, for heterogeneous networks.
in conjunction with the ns-2, offer support for multi-technologies
and cross-layering. Although the first implementation only consid-
8. Open research issues
ered UMTS and Wi-Fi, currently a set of plugins have been devel-
oped to offer WiMAX support.
Although VHO techniques have been significantly improved in
SHINE [103] is a simulation framework able to simulate heter-
the last years, there are still several issues requiring further
ogeneous networks and perform vertical handovers within over-
scrutiny concerning VHO in VNs environment. Summarizing, we
lapped coverage areas. This framework uses a main server which
suggest the main issues to be addressed.
acts as the upper layers in charge of the mobility and users manage-
ment, transport protocols, etc. it also uses clients for each simu-
lated access technology. These clients represent the lower layers. 8.1. Quality of service (QoS)
The framework is able to simulate GPRS, UMTS, Wi-Fi and WiMAX.
The main features of this framework are that it allows reproducing The applications being executed on the mobile device require
the transport protocol behavior and start an instance for each new different levels of QoS. However, providing the required QoS across
traffic session originated by the client. the different wireless networks is a major issue. In addition, high
Universal seamless handoff architecture (USHA) [60,1] is a levels of mobility introduce more challenging issues. In order to
simple, fast and reliable handover solution to perform seamless guarantee the QoS, VHO techniques must carefully consider vehic-
vertical handover (VHO). It is based on a two part software compo- ular mobility and network conditions in order to choose the best
nent (USHA client and USHA server) that uses MIP tunneling to candidate network and perform a fast handover.
maintain the session open. This solution was fully tested in a real
implementation. 8.2. Quality of Experience (QoE)
Y-Comm [104] presents a new architecture for mobile hetero-
geneous networking. This architecture is based on two frame- Guaranteeing the QoS is not enough in order to provide the best
works: the first one is called Peripheral framework, and the possible service to users. Quality of Experience (QoE) is a concept
second one Core framework. The Peripheral framework interacts related to the users’ satisfaction, and QoE assessment sometimes
with the peripheral wireless network, and is implemented in the evidences that good networking performance is not a synonym
mobile devices. It is in charge of controlling the VHO and doing of total satisfaction to the end users. Hence, issues related with
policy management, among other tasks. The second framework the user preferences and mobile equipments must be taken into
runs in a distributed manner throughout the infrastructure provid- account in order to design VHO strategies and techniques.
ing configuration services, as well as network management and
QoS. The Y-Comm testbed is being developed for the Android mo-
8.3. Security
bile operating system [105].
Table 4 presents the use of the different evaluation tools,
Security is one of the most important issues in many technolog-
including the references to the proposals that have used each tool.
ical areas. In networking, security has always been a primary con-
We have surveyed the main conferences and journals on VNs and
cern. Unfortunately, the IEEE 802.21 standard does not consider
VHO research areas, mostly in the last decade. This table clearly
security in the main standard leaving this area to be handled by
shows the trend towards testbed usage. Short scale testbeds are
other protocols such as MIP, which places a strong emphasis on
typically used to evaluate the performance of VHA, while medium
security. So, in order to improve the VHO techniques, robust secu-
scale ones are used to evaluate the whole VHO process within net-
rity solutions must be adopted that can be used in heterogeneous
works with several mobile devices.
networks, thus allowing security anytime and anywhere.
To the best of our knowledge, there are very few proposals of
VHO techniques designed and implemented in VN environments.
Lee et al. [21] present VHA controllers to optimize the overall 8.4. Industry standardization effort and alliances
VHO performance; these controllers take into account a route-
selection algorithm considering wireless Ad-hoc networks, as well The design and implementation of the IEEE 802.21 standard is a
as Vehicular Ad-hoc Network (VANET) environments. Zhu et al. great step towards seamless VHO among heterogeneous wireless
[50] present a recompilation of different solutions for mobility networks. However, issues related to the operators are not solved
management combining simultaneous Vehicle-to-Vehicle (V2V) by this standard. The billing and pricing for the use of the networks
and Vehicle-to-Infrastructure (V2V) routing in order to perform owned by the operators, the management of these issues among
VHO using multi-homed mobile routers. A specific proposal for them and between operators and users, must be solved at least
VHO in a VANET scenario is presented in [108]. The authors pro- partially in order to facilitate the VHO and to guarantee the QoE
pose a simple VHO architecture – VANET Backup Communications as well as the QoS.
(VANBA) – as an alternative to the IEEE 802.21 standard and the
Communication Access for Land Mobiles (CALM) architecture. 8.5. Homogeneous VHO evaluation
The key issue is the simplicity offered through its 2.5 layer and
its function-based decision (monetary cost and bandwidth avail- Aforementioned, there is a lack of an homogeneous evaluation
able). This architecture has been evaluated over the DRIVE testbed methodology. This issue is a barrier when comparing different pro-
[110], achieving low latencies in the execution of the VHO among posals. Therefore, a common methodology is required in order to
Wi-Fi and UMTS in a VN environment. Considering the IEEE 802.21 permit researchers, developers and users easily compare and eval-
standard along the GeoNET project’s V2V communication system, uate the diverse VHO techniques found in the literature. Hence, the
Widhiasi et al. [67] present an interesting VHO scheme for VHO-related research community, should join efforts to release an
V2V communication called FANSCI: Fast handover scheme for standard or guidelines on good practices for VHO evaluation.
9. Conclusion [6] J. Jo, J. Cho, A cross-layer vertical handover between mobile WiMAX and 3G
networks, in: IEEE Wireless Communications and Mobile Computing
Conference, 2008, pp. 644–649. <http://dx.doi.org/10.1109/IWCMC.
Nowadays users demand high service availability to access the 2008.111>.
network infrastructure not only under the Anywhere, Anytime, [7] I. Joe, S. Hong, A mobility-based prediction algorithm for vertical handover in
hybrid wireless networks, in: 2nd IEEE/IFIP International Workshop on
but also under the Always best connected paradigms.
Broadband Convergence Networks, 2007, pp. 1–5. <http://dx.doi.org/
In this paper we described the VHO process, the different 10.1109/BCN.2007.372748>.
phases that are part of this process, as well as the procedures trig- [8] H. Attaullah, F. Iqbal, M.Y. Javed, Intelligent vertical handover decision model
to improve QoS, in: IEEE Third International Conference on Digital
gered at each phase. Additionally the Media Independent Handover
Information Management, 2008, pp. 119–124. <http://dx.doi.org/
Services Protocol (IEEE 802.21) designed for VHO is briefly de- 10.1109/ICDIM.2008.4746775>.
scribed. Moreover, we analyzed the VHO from the Vehicular Net- [9] R. Corvaja, QoS analysis in overlay Bluetooth–WiFi networks with profile-
work (VN)’s perspective to evaluate the branch of solutions that based vertical handover, IEEE Transactions on Mobile Computing 5 (12)
(2006) 1679–1690. <http://dx.doi.org/10.1109/TMC.2006.187>.
the conjunction of wireless technologies and VHO techniques of- [10] T. Buburuzan, G. May, T. Melia, J. Modeker, M. Wetterwald, Integration of
fers to satisfy the connectivity needs on the road. Furthermore, broadcast technologies with heterogeneous networks – an IEEE 802.21
we classified the most widely used vertical handover decision algo- centric approach, in: IEEE International Conference on Consumer Electronics,
2007, pp. 1–2. <http://dx.doi.org/10.1109/ICCE.2007.341356>.
rithm (VHA) in the literature into different sets of algorithms [11] P. Nay, C. Zhou, Vertical handoff decision algorithm for integrated UMTS and
depending primarily on the information used to make decisions, LEO satellite networks, in: IEEE International Conference on Communications
and the techniques employed. Concerning VNs, we consider that and Mobile Computing, vol. 2, 2009, pp. 180–184. <http://dx.doi.org/
10.1109/CMC.2009.135>.
MIHS protocol must be taken into account when designing the [12] C. Liu, C. Zhou, An improved interworking architecture for UMTS–WLAN tight
VHO architecture. Multiple-parameters-based algorithms should coupling, in: IEEE Wireless Communications and Networking Conference, vol.
be used to select the information, in order to the take the most 3, 2005, pp. 1690–1695. <http://dx.doi.org/10.1109/WCNC.2005.
1424767>.
of the context environment.
[13] S. Asadullah, A. Mahmoud, A.M. Amara, Vertical handoff characterization for
Different tools are presented in order to compare the multiple sip and mSCTP based UMTS–WLAN integration solutions, in: IEEE–GCC 2007,
VHO possibilities and to evaluate the most appropriate VHO tech- 2007. <https://eprints.kfupm.edu.sa/1451/>.
[14] M. Ylianttila, J. Makela, K. Pahlavan, Analysis of handoff in a location-
nique to be applied in VN environments. Moreover, the most suit-
aware vertical multi-access network, Elsevier Computer Networks, 47 (2)
able tools for the study of VHO and VNs have been presented. (2005). <http://www.sciencedirect.com/science/article/B6VRG-
When evaluating a VHO proposal in a testbed, DRIVE offers a fully 4D1MTXK-5/2/ed3a50786e229946a88e7c5abc1b1efa>.
function car-framework. Whereas, simulating VHO in VNs, diverse [15] I.-C. Lin, C.-S. Shieh, W.K. Lai, Avoidance of redundant retransmission in
vertical handover by modified stream control transmission protocol, in: 3rd
tools might be used; nevertheless, we consider ns-2 in conjunction IEEE International Conference on Innovative Computing Information and
with seamless and secure mobility as fast and flexible tools able to Control, 2008, p. 374. <http://dx.doi.org/10.1109/ICICIC.2008.182>.
fulfill the VHO in VNs requirements. [16] K. Tsukamoto, T. Yamaguchi, S. Kashihara, Y. Oie, Implementation design of
handover management scheme for efficient multimodal communication, in:
When gathering information in VNs we must take into account IEEE International Conference on Wireless and Mobile Computing,
the vehicular mobility prediction and the vehicular speed in order Networking and Communications, 2008, pp. 684–689. <http://
to improve the selection effectiveness in the presence of highly dy- dx.doi.org/10.1109/WiMob.2008.36>.
[17] H. Kwon, M.-J. Yang, A.-S. Park, S. Venkatesan, Handover prediction strategy
namic channel conditions. In addition, the decision process must for 3G–WLAN overlay networks, in: IEEE Network Operations and
rely on a fast MCDM algorithm to perform an accurate decision, Management Symposium, 2008, pp. 819–822. <http://dx.doi.org/
and to switch to the best candidate network within a very short 10.1109/NOMS.2008.4575222>.
[18] T. Inzerilli, A.M. Vegni, A. Neri, R. Cusani, A location-based vertical handover
period of time. Finally, when executing a seamless VHO, low laten-
algorithm for limitation of the ping-pong effect, in: IEEE International
cies and high security levels must be achieved. To do so, we con- Conference on Wireless and Mobile Computing, Networking and
sider Mobile IP as one of the candidate solution offering the best Communications, 2008, pp. 385–389. <http://dx.doi.org/10.1109/
WiMob.2008.64>.
trade-off.
[19] D.P. Kim, S.J. Koh, V. Leung, On the packet reordering of mSCTP for vertical
handover in heterogeneous wireless networks, in: IEEE 68th Vehicular
Acknowledgments Technology Conference, 2008, pp. 1–5. <http://dx.doi.org/10.1109/
VETECF.2008.273>.
[20] M. Fouladian, F. Hendessi, A. Shafieinejad, M. Rahimi, M.M. Bayat, A new
This work was partially supported by the Ministerio de method for vertical handoff between WLANs and UMTS in boundary
Educación y Ciencia, Spain, under Grant TIN2008-06441-C02-01, conditions, in: IEEE International Conference on Computer Engineering and
Technology, vol. 1, 2009, pp. 166–171. <http://dx.doi.org/10.1109/
and by the Ayudas complementarias para proyectos de I+D para ICCET.2009.166>.
grupos de calidad de la Generalitat Valenciana (ACOMP/2010/005). [21] S. Lee, K. Sriram, K. Kim, Y.H. Kim, N. Golmie, Vertical handoff decision
algorithms for providing optimized performance in heterogeneous wireless
networks, IEEE Transactions on Vehicular Technology 58 (2) (2009) 865–881.
References <http://dx.doi.org/10.1109/TVT.2008.925301>.
[22] A. Sur, D.C. Sicker, Multi layer rules based framework for vertical handoff, in:
[1] L.J. Chen, T. Sun, G. Yang, M. Gerla, USHA: a simple and practical seamless 2nd IEEE International Conference on Broadband Networks, vol. 1, 2005, pp.
vertical handoff solution, in: IEEE Consumer Communications and 571–580. <http://dx.doi.org/10.1109/ICBN.2005.1589662>.
Networking Conference, vol. 2, 2006. [23] M. Li, S. Chen, D. Xie, A multi-step vertical handoff mechanism for cellular
[2] M. Stemm, R.H. Katz, Vertical handoffs in wireless overlay networks, Springer multi-hop networks, in: 2nd ACM Workshop on Performance Monitoring and
Mobile Networks and Applications 3 (4) (1998) 335–350. <http:// Measurement of Heterogeneous Wireless and Wired Networks, New York,
dx.doi.org/10.1023/A:1019197320544>. NY, USA, 2007, pp. 119–123. <http://dx.doi.org/10.1145/1298275.
[3] W.K. Liao, Y.C. Chen, Supporting vertical handover between universal mobile 1298301>.
telecommunications system and wireless LAN for real-time services, IEEE [24] M. Kassar, B. Kervella, G. Pujolle, Architecture of an intelligent inter-system
Institution of Engineering and Technology Communications 2 (1) (2008) 75– handover management scheme, in: IEEE Future Generation Communication
81. <http://dx.doi.org/10.1049/iet-com:20050368>. and Networking Conference, vol. 1, 2007, pp. 332–337. <http://
[4] D.-H. Kim, W.-T. Kim, H.-G. Lee, S.-J. Kim, C.-H. Lee, A performance evaluation dx.doi.org/10.1109/FGCN.2007.70>.
of vertical handover architecture with low latency handover, in: IEEE [25] M. Kassar, B. Kervella, G. Pujolle, Autonomic-oriented architecture for an
International Conference on Convergence and Hybrid Information intelligent handover management scheme, in: 6th IEEE Annual
Technology, 2008, pp. 66–69. <http://dx.doi.org/10.1109/ICHIT. Communication Networks and Services Research Conference, 2008, pp.
2008.218>. 139–146. <http://dx.doi.org/10.1109/CNSR.2008.30>.
[5] D.J. Wright, Maintaining QoS during handover among multiple wireless [26] M. Kassar, B. Kervella, G. Pujolle, An overview of vertical handover decision
access technologies, in: IEEE/ACM International Conference on the strategies in heterogeneous wireless networks, Elsevier Computer Comm
Management of Mobile Business, Washington, DC, USA, 2007. <http:// unications, 31 (10) (2008). <http://www.sciencedirect.com/science/
dx.doi.org/10.1109/ICMB.2007.40>. article/B6TYP-4RR1NT4-7/2/d1e1be26ca3741a23f0cfe44213f448e>.
[27] D. Kwon, A. Park, Methods for seamless vertical handoff between UMTS and [49] I. Broustis, M. Faloutsos, Routing in vehicular networks: feasibility, modeling,
WLAN, in: 9th IEEE International Conference on Advanced Communication and security, International Journal of Vehicular Technology, 2008. <http://
Technology, vol. 2, 2007, pp. 1286–1289. <http://dx.doi.org/10.1109/ www.hindawi.com/GetArticle.aspx?doi=10.1155/2008/267513&
ICACT.2007.358592>. e=cta>.
[28] K. Yang, I. Gondal, B. Qiu, L.S. Dooley, Combined SINR based vertical handoff [50] K. Zhu, D. Niyato, P. Wang, E. Hossain, D.I. Kim, Mobility and handoff
algorithm for next generation heterogeneous wireless networks, in: IEEE management in vehicular networks: a survey, Wireless Communications and
Global Telecommunications Conference, 2007, pp. 4483–4487. <http:// Mobile Computing (2009). <http://dx.doi.org/10.1002/wcm.853>.
dx.doi.org/10.1109/GLOCOM.2007.852>. [51] J. Jakubiak, Y. Koucheryavy, State of the art and research challenges for
[29] A. Mehbodniya, J. Chitizadeh, An intelligent vertical handoff algorithm for VANETs, in: 5th IEEE Consumer Communications and Networking
next generation wireless networks, in: 2nd IFIP/IEEE International Conference Conference, 2008, pp. 912–916. <http://dx.doi.org/10.1109/ccnc08.
on Wireless and Optical Communications Networks, 2005, pp. 244–249. 2007.212>.
<http://dx.doi.org/10.1109/WOCN.2005.1436027>. [52] IEEE P1609, Standard for Wireless Access in Vehicular Environments (WAVE)
[30] L. Xia, J. Ling-Ge, H. Chen, L. Hong-Wei, An intelligent vertical handoff Committee SCC32, IEEE Intelligent Transportation Systems Council, 2006.
algorithm in heterogeneous wireless networks, in: IEEE International [53] US Department of Transportation, ITS Intelligent Transport Communication
Conference on Neural Networks and Signal Processing, 2008, pp. 550–555. Program, Dedicated Short Range Communications (DSRC), ITS Standards
<http://dx.doi.org/10.1109/ICNNSP.2008.4590411>. Advisory, 2003.
[31] S. Horrich, Ben, P. Godlewski, Neural networks for adaptive vertical handover [54] IEEE 802.11, IEEE P802.11p: Wireless Access in Vehicular Environments
decision, in: 5th IEEE International Symposium on Modeling and (WAVE), IEEE Computer Society, 2006.
Optimization in Mobile, Ad Hoc and Wireless Networks, 2007, pp. 1–7. [55] S. Eichler, Performance evaluation of the IEEE 802.11p WAVE communication
<http://dx.doi.org/10.1109/WIOPT.2007.4480068>. standard, in: 66th IEEE Conference on Vehicular Technology, 2007, pp. 2199–
[32] J. Latvakoski, P. Valitalo, T. Vaisanen, Vertical handover during a VoIP call in 2203. <http://dx.doi.org/10.1109/VETECF.2007.461>.
hybrid mobile ad hoc networks, in: IEEE Wireless Telecommunications [56] M. Hayashi, S. Fukuzawa, H. Ichikawa, T. Kawato, J. Yamada, T. Tsuboi, S.
Symposium, 2008, pp. 38–45. <http://dx.doi.org/10.1109/WTS.2008. Matsui, T. Maruyama, Development of vehicular communication (WAVE)
4547541>. system for safety applications, in: 7th IEEE International Conference on ITS
[33] L.-J. Chen, G. Yang, T. Sun, M.Y. Sanadidi, M. Gerla, Enhancing QoS support for Telecommunications, 2007, pp. 1–5. http://dx.doi.org/10.1109/ITST.
vertical handoffs using implicit/explicit handoff notifications, in: 2nd IEEE 2007.4295841.
International Conference on Quality of Service in Heterogeneous Wired/ [57] J. Santa, Arquitectura de una plataforma telematica integral para el
Wireless Networks, 2005, pp. 8-37. <http://dx.doi.org/10.1109/ despliegue de servicios ubicuos en el ambito de los sistemas inteligentes de
QSHINE.2005.22>. transporte, Ph.D. Thesis, Universidad de Murcia. Departamento de Ingenieria
[34] E.K. Paik, S.Y. Heo, H. Kim, J.S. Jin, S.-C. Lee, S.H. Lee, Seamless vertical de Informacion y Comunicaciones, 2009.
handover using multi-homed mobile access point, in: IEEE Global [58] E. Stevens-Navarro, Y. Lin, V.W.S. Wong, An MDP-based vertical handoff
Telecommunications Conference, 2008, pp. 1–4. <http://dx.doi.org/ decision algorithm for heterogeneous wireless networks, IEEE Transactions
10.1109/GLOCOM.2008.ECP.1081>. on Vehicular Technology 57 (2) (2008) 1243–1254. <http://dx.doi.org/
[35] Fredson Phiri, M. Murthy, WLAN–GPRS tight coupling based interworking 10.1109/TVT.2007.907072>.
architecture with vertical handoff support, Springer Wireless Personal [59] A. Singhrova, N. Prakash, A review of vertical handoff decision algorithm in
Communications 40 (2) (2007) 137–144. <http://dx.doi.org/10.1007/ heterogeneous networks, in: 4th ACM International Conference on Mobile
s11277-006-9079-9>. Technology, Applications, and Systems, New York, NY, USA, 2007, pp. 68–71.
[36] Y.C. Chen, J.H. Hsia, Y.J. Liao, Advanced seamless vertical handoff architecture http://dx.doi.org/10.1145/1378063.1378075.
for WiMAX and WiFi heterogeneous networks with qos guarantees, Elsevier [60] L.J. Chen, T. Sun, B. Chen, V. Rajendran, M. Gerla, A smart decision model for
Computer Communications 32 (2) (2009) 281–293. <http://dx.doi.org/ vertical handoff, in: 4th International Workshop on Wireless Internet and
10.1016/j.comcom.2008.10.014>. Reconfigurability, Athens, Greece, 2004.
[37] T. Buburuzan, L.N. Nyamen, Performance evaluation of an enhanced IEEE [61] V. Gupta, IEEE 802.21 media independent handover tutorial, in: IEEE 802.21
802.21 handover model, in: 1st Workshop on Wireless Broadband Access for Working Group Meeting, July 2006.
Communities and Rural Developing Regions, Karlstad, 2008. [62] F. Siddiqui, S. Zeadally, Mobility management across hybrid wireless
[38] I. Lassoued, J.M. Bonnin, B.Z. Hamouda, A. Belghith, A methodology for networks: trends and challenges, Computer Communications 29 (9) (2006)
evaluating vertical handoff decision mechanisms, in: 7th IEEE International 1363–1385. <http://dx.doi.org/10.1016/j.comcom.2005.09.003>.
Conference on Networking, 2008, pp. 377–384. <http://dx.doi.org/ [63] E. Gustafsson, A. Jonsson, Always best connected, IEEE Wireless
10.1109/ICN.2008.71>. Communications 10 (1) (2003) 49–55. <http://dx.doi.org/10.1109/
[39] B.-J. Chang, J.-F. Chen, Cross-layer-based adaptive vertical handoff with MWC.2003.1182111>.
predictive RSS in heterogeneous wireless networks, IEEE Transactions on [64] S.-C. Oh, C.-Y. Lee, Y.-S. Shin, J.-U. Kim, A vertical handover based on software
Vehicular Technology 57 (6) (2008) 3679–3692. <http://dx.doi.org/ communications architectures, in: 60th IEEE Vehicular Technology
10.1109/TVT.2008.921619>. Conference, vol. 3, 2004, pp. 1973–1977. <http://dx.doi.org/10.1109/
[40] W. Ying, Y. Jun, Z. Yun, L. Gen, Z. Ping, Vertical handover decision in an VETECF.2004.1400383>.
enhanced media independent handover framework, in: IEEE Wireless [65] J.M. Seigneur, X. Titi, T.E. Maliki, Towards mobile/wearable device
Communications and Networking Conference, 2008, pp. 2693–2698. electrosmog reduction through careful network selection, in: Proceedings of
<http://dx.doi.org/10.1109/WCNC.2008.472>. the 1st Augmented Human International ACM Conference, New York, USA,
[41] D. Niyato, E. Hossain, Dynamics of network selection in heterogeneous 2010, pp. 1–5. <http://dx.doi.org/10.1145/1785455.1785476>.
wireless networks: an evolutionary game approach, IEEE Transactions on [66] I. Joe, M. Shin, A mobility-based prediction algorithm with dynamic lgd
Vehicular Technology 58 (4) (2009) 2008–2017. <http://dx.doi.org/ triggering for vertical handover, in: 7th IEEE Consumer Communications and
10.1109/TVT.2008.2004588>. Networking Conference, 2010, pp. 1–2. <http://dx.doi.org/10.1109/
[42] A. Salkintzis, M. Hammer, I. Tanaka, C. Wong, Voice call handover CCNC.2010.5421627>.
mechanisms in next-generation 3GPP systems, IEEE Communications [67] A. Widhiasi, V. Mohanan, M.F. Pasha, R. Budiarto, Vertical handover scheme
Magazine 47 (2) (2009) 46–56. <http://dx.doi.org/10.1109/MCOM. for car-to-car communication based on IEEE 802.21 standard, in: Second
2009.4785380>. International IEEE Conference on Computer Engineering and Applications,
[43] P. Goyal, S.K. Saxena, A dynamic decision model for vertical handoffs across 2010, pp. 143–147. <http://dx.doi.org/10.1109/ICCEA.2010.36>.
heterogeneous wireless networks, in: World Academy of Science, Engineering [68] S. Mansor, T.-C. Wan, Mobility management in heterogeneous wireless access
and Technology, vol. 31, 2008. network with ieee 802.21 services, in: Second International IEEE Conference
[44] E. Stevens-Navarro, V.W.S. Wong, Comparison between vertical handoff on Computer and Network Technology, 2010. <http://dx.doi.org/
decision algorithms for heterogeneous wireless networks, in: 63rd IEEE 10.1109/ICCNT.2010.91>.
Vehicular Technology Conference, vol. 2, 2006, pp. 947–951. <http:// [69] I.-S. Jang, W.-T. Kim, J.-M. Park, Y.-J. Park, Mobile multicast mechanism based
dx.doi.org/10.1109/VETECS.2006.1682964>. MIH for efficient network resource usage in heterogeneous networks, in: 12th
[45] M. Bernaschi, F. Cacace, G. Iannello, Vertical handoff performance in International IEEE Conference on Advanced Communication Technology,
heterogeneous networks, in: IEEE International Conference on Parallel 2010, pp. 850–854. <http://ieeexplore.ieee.org/search/srchabs-
Processing, 2004, pp. 100–107. <http://ieeexplore.ieee.org/xpls/ tract.jsp?tp=& #38;arnumber= 5440151>.
abs_all.jsp?arnumber=1328002>. [70] B.Y. Jung, M.S. Choi, H.Y. Youn, O. Song, Vertical handover based on the
[46] IEEE standard for local and metropolitan area networks – part 21: media prediction of mobility of mobile node, in: 8th IEEE International Conference
independent handover, Technical report, 2009. <http://dx.doi.org/ on Pervasive Computing and Communications Workshops (PERCOM
10.1109/IEEESTD.2009.4769367>. Workshops), 2010, pp. 534–539. <http://dx.doi.org/10.1109/
[47] A. De-La-Oliva, T. Melia, A. Vidal, C.J. Bernardos, I. Soto, A. Banchs, IEEE 802.21 PERCOMW.2010.5470496>.
enabled mobile terminals for optimized WLAN/3G handovers: a case study, [71] Y. Wang, P. Zhang, Y. Zhou, J. Yuan, F. Liu, G. Li, Handover management in
ACM Mobile Computing and Communications Review 11 (2) (2007) 29–40. enhanced MIH framework for heterogeneous wireless networks
<http://dx.doi.org/10.1145/1282221.1282224>. environment, Wireless Personal Communications 52 (3) (2010) 615–636.
[48] A. Dutta, S. Das, D. Famolari, Y. Ohba, K. Taniuchi, V. Fajardo, R.M. Lopez, T. <http://dx.doi.org/10.1007/s11277-008-9628-5>.
Kodama, H. Schulzrinne, Seamless proactive handover across heterogeneous [72] D. Griffith, R. Rouil, N. Golmie, Performance metrics for IEEE 802.21 media
access networks, Springer Wireless Personal Communications 43 (3) (2007) independent handover (MIH) signaling, Wireless Personal Communications 52
837–855. <http://dx.doi.org/10.1007/s11277-007-9266-3>. (3) (2010) 537–567. <http://dx.doi.org/10.1007/s11277-008-9629-4>.
[73] P. Pawar, K. Wac, B.J. van Beijnum, P. Maret, A. van Halteren, H. Hermens, Consumer Communications and Networking Conference, 2010, pp. 1–6.
Context-aware middleware architecture for vertical handover support to <http://dx.doi.org/10.1109/CCNC.2010.5421742>.
multi-homed nomadic mobile services, in: ACM Symposium on Applied [93] L. Ma, F. Yu, V. Leung, T. Randhawa, A new method to support UMTS/WLAN
Computing, Fortaleza, Ceara, Brazil, 2008, pp. 481–488. <http://dx.doi. vertical handover using SCTP, in: IEEE 58th Vehicular Technology Conference,
org/10.1145/1363686.1363806>. 2003, pp. 1788–1792. <http://dx.doi.org/10.1109/VETECF.2003.
[74] C. Gu, M. Song, Y. Zhang, J. Song, Access network selection strategy using 1285333>.
position prediction in heterogeneous wireless networks, Frontiers of [94] D. Kim, S. Koh, Adaptive congestion control of mSCTP for vertical handover
Electrical and Electronic Engineering in China 5 (1) (2010) 23–28. based on bandwidth estimation in heterogeneous wireless networks,
<http://dx.doi.org/10.1007/s11460-009-0064-4>. Wireless Personal Communications, 2009. <http://dx.doi.org/10.1007/
[75] A. Hasswa, N. Nasser, H. Hossanein, Generic vertical handoff decision function s11277-009-9872-3>.
for heterogeneous wireless, in: 2nd IFIP/IEEE International Conference on [95] D. Howie, M. Ylianttila, E. Harjula, J. Sauvola, State-of-the-art SIP for mobile
Wireless and Optical Communications Networks, 2005, pp. 239–243. application supernetworking, in: Nordic Radio Symposium, Oulu, Finland,
<http://dx.doi.org/10.1109/WOCN.2005.1436026>. 2004. <http://www.mediateam.oulu.fi/publications/show/548/
[76] M. Mani, N. Crespi, WLC14-5: handover criteria considerations in future ?lang=fi>.
convergent networks, in: IEEE Global Telecommunications Conference, 2006, [96] W. Wong, M. Giraldi, M. a. Mauricio, F. Verdi, An identifier-based architecture
pp. 1–5. <http://dx.doi.org/10.1109/GLOCOM.2006.696>. for native vertical handover support, in: 24th IEEE International Conference
[77] H. Liao, L. Tie, Z. Du, A vertical handover decision algorithm based on fuzzy on Advanced Information Networking and Applications, Perth, Australia,
control theory, in: IEEE International Multi-Symposiums on Computer and 2010. <http://dx.doi.org/10.1109/AINA.2010.21>.
Computational Sciences, Los Alamitos, CA, USA, vol. 2, 2006, pp. 309–313. [97] P. Paakkonen, P. Salmela, R. Aguero, J. Choque, Performance analysis of HIP-
<http://dx.doi.org/10.1109/IMSCCS.2006.173>. based mobility and triggering, in: IEEE International Symposium on a World
[78] J.-Z. Sun, A review of vertical handoff algorithms for cross-domain mobility, of Wireless, Mobile and Multimedia Networks, 2008, pp. 1–9. <http://
in: IEEE International Conference on Wireless Communications, Networking dx.doi.org/10.1109/WOWMOM.2008.4594827>.
and Mobile Computing, 2007, pp. 3156–3159. <http://dx.doi.org/ [98] K. Fall, K. Varadhan, Ns Notes and Documents. The VINT Project. UC Berkeley,
10.1109/WICOM.2007.782>. LBL, USC/ISI, and Xerox PARC, June 2009. <http://www.isi.edu/nsnam/
[79] T. Kim, S.-w. Han, Y. Han, A QoS-aware vertical handoff algorithm based on ns/ns-documentation.html>.
service history information, IEEE Communications Letters 14 (6) (2010) 527– [99] OPNET Technologies, OPNET Modeler. <http://www.opnet.com/
529. <http://dx.doi.org/10.1109/LCOMM.2010.06.100176>. solutions/network_rd/modeler.html>.
[80] K.Y. Kim, D.G. Yun, J.C. Shim, K.S. Cho, S.-G. Choi, QoS-aware seamless [100] MathWorks, Mathlab and Simulink for Technical Computing. <http://
handover schemes between heterogeneous networks in NGN, in: 12th www.mathworks.com/>.
International IEEE Conference on Advanced Communication Technology, [101] Advanced Network Technology Division, National Institute of Standards and
2010. <http://http://ieeexplore.ieee.org/stamp/stamp.jsp?tp= Technology, Seamless and Secure Mobility. <http://www.antd.nist.gov/
&arnumber=5440480>. seamlessandsecure/>.
[81] J. Sen, A. Ukil, A QoS-aware end-to-end connectivity management algorithm [102] N. Baldo, F. Maguolo, M. Miozzo, M. Rossi, M. Zorzi, ns2-MIRACLE: a modular
for mobile applications, in: Proceedings of the Third Annual ACM Bangalore framework for multi-technology and cross-layer support in network
Conference, New York, NY, USA, 2010, pp. 1–9. <http://dx.doi.org/ simulator 2, in: Proceedings of the 2nd International Conference on
10.1145/1754288.1754292>. Performance Evaluation Methodologies and Tools, ICST, Brussels, Belgium,
[82] M. Ylianttila, Vertical handoff and mobility – system architecture and 2007, pp. 1–8. <http://portal.acm.org/citation.cfm?id=1345284>.
transition analysis, Ph.D. Thesis, Teknillinen tiedekunta, Oulun yliopisto [103] A. Bazzi, G. Pasolini, C. Gambetti, Shine: simulation platform for
Sahko-ja tietotekniikan osasto, Oulun yliopisto Infotech Oulu, Oulun heterogeneous interworking networks, in: IEEE International Conference on
yliopisto, 2005. Communications, 2006, pp. 5534–5539. <http://dx.doi.org/10.1109/
[83] H.H. Choi, O. Song, D.H. Cho, Seamless handoff scheme based on pre- ICC.2006.255543>.
registration and pre-authentication for UMTS–WLAN interworking, Springer [104] G.E. Mapp, F. Shaikh, D. Cottingham, J. Crowcroft, J. Baliosian, Y-Comm: a
Wireless Personal Communications 41 (3) (2007) 345–364. <http:// global architecture for heterogeneous networking, in: 3rd ICST/ACM
dx.doi.org/10.1007/s11277-006-9146-2>. International Conference on Wireless Internet, Belgium, 2007, pp. 1–5.
[84] C. Perkins, IP Mobility Support for IPv4, August 2002. <http://portal.acm.org/citation.cfm?id=1460047.1460075>.
[85] D. Johnson, C. Perkins, J. Arkko, Mobility Support in IPv6, 2004. [105] Open Handset Alliance, Android/ <http://www.android.com>.
[86] J. Rosenberg, H. Schulzrinne, G. Camarillo, A. Johnston, J. Peterson, Sparksand, [106] S.-H. Kwon, S.-J. Koh, T.-W. Um, W. Ryu, Mobile SCTP with bicasting for
E. Schooler, SIP: Session Initiation Protocol, 2002. vertical handover, in: 3rd IEEE International Conference on Convergence and
[87] V. Devarapalli, R. Wakikawa, A. Petrescu, P. Thubert, RFC3963 – Network Hybrid Information Technology, vol. 1, 2008, pp. 123–125. <http://
Mobility (NEMO) Basic Support Protocol, June 2005. dx.doi.org/10.1109/ICCIT.2008.94>.
[88] R. Moskowitz, P. Nikander, Jokela, T. Henderson, RFC 5201 – Host Identity [107] A.E.M. Taha, H.S. Hassanein, H.T. Mouftah, Exploiting vertical handoffs in next
Protocol, April 2008. <http://tools.ietf.org/html/rfc5201>. generation radio resource management, Proceedings of the International
[89] I.F. Akyildiz, J. Xie, S. Mohanty, A survey of mobility management in next- Conference on Communications, vol. 5, IEEE, 2006, pp. 2083–2088.
generation all–IP-based wireless systems, IEEE Wireless Communications 11 <http://dx.doi.org/10.1109/ICC.2006.255077>.
(4) (2004) 16–28. <http://dx.doi.org/10.1109/MWC.2004.1325888>. [108] J.A. Olivera, I. Cortázar, C. Pinart, Los, I. Lequerica, VANBA: a simple handover
[90] I. Ali, A. Casati, K. Chowdhury, K. Nishida, E. Parsons, S. Schmid, R. Vaidya, mechanism for transparent, always-on v2v communications, in: 69th IEEE
Network-based mobility management in the evolved 3GPP core network, Vehicular Technology Conference, Barcelona, 2009.
IEEE Communications Magazine 47 (2) (2009) 58–66. <http:// [109] A.S. Reaz, S. Fu, S. Ferdousi, M. Atiquzzaman, mSIGMA: an efficient handoff
dx.doi.org/10.1109/MCOM.2009.4785381>. scheme for multiclass networks, IEEE Systems Journal, 2010. <http://
[91] S. Jeong, M.K. Shin, H.J. Kim, A handover scheme supporting IPv4–IPv6 dx.doi.org/10.1109/JSYST.2010.2047295>.
traversal over network-based mobility management domains, in: 9th IEEE [110] C. Pinart, P. Sanz, I. Lequerica, D. Garcı́a, I. Barona, D.S. Aparisi, Drive: a
International Conference on Advanced Communication Technology, vol. 2, reconfigurable testbed for advanced vehicular services and communications,
2007, pp. 1033–1036. <http://dx.doi.org/10.1109/ICACT.2007. in: Proceedings of the 4th International Conference on Testbeds and Research
358534>. Infrastructures for the Development of Networks and Communities, ICST,
[92] K. Andersson, D. Granlund, M. Elkotob, C. Ahlund, Bandwidth efficient ICST, Brussels, Belgium, 2008, pp. 1–8. <http://portal.acm.org/
mobility management for heterogeneous wireless networks, in: 7th IEEE citation.cfm?id=1390576.1390595>.

Barja

Uploaded by

Copyright:

Available Formats

Barja

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Barja

Uploaded by

Copyright:

Available Formats

Computer Communications 34 (2011) 985–997

Contents lists available at ScienceDirect

An overview of vertical handover techniques: Algorithms, protocols and tools

The research community has been making signiﬁcant efforts

Proposals % Networking technologies

Upper Layers (L3 and above)

SIP MIPv4 MIPv6 HIP ....

Mobility Management Handover Management

MIH MIH Information

Smart Handover Information

Link Link Information

Lower Layers (L2 and below)

Protocol and Device Hardware

802.3 802.11 3GPP 802.16 ....

5.2.1. Mathematical algorithms 5.3. Parameter aggregation algorithms

Type Tool/scale % Proposals

You might also like