Global mobility and Handover management for heterogeneous network in vanet

International Journal of Next-Generation Networks (IJNGN) Vol.5, No.3, September 2013
DOI : 10.5121/ijngn.2013.5302 23
GLOBAL MOBILITY AND HANDOVER MANAGEMENT
FOR HETEROGENEOUS NETWORK IN VANET
Ravi Shankar Shukla1
, Neeraj Tyagi2
1
Department of Computer Science and Engineering & IT, Invertis University, Bareilly,
India
2
Department of Computer Science & Engineering, MNNIT, Allahabad, India
ABSTRACT
Now a day’s Vehicular Ad Hoc Network (VANET) is an emerging technology. Mobility management is one
of the most challenging research issues for VANETs to support variety of intelligent transportation system
(ITS) applications. VANETs are getting importance for inter-vehicle communication, because they allow
the communication among vehicles without any infrastructure, configuration effort, and without the high
costs of cellular networks. Besides local data exchange, vehicular applications may be used to accessing
Internet services. The access is provided by Internet gateways located on the site of roadside. However, the
Internet integration requires a respective mobility support of the vehicular ad hoc network. In this paper
we will study about the network mobility approach in vehicular ad hoc network; the model will describe the
movement of vehicles from one network to other network. The proposed handover scheme reduces the
handover latency, packet loss signaling overhead.
Keywords: VANET, handoff, NEMO, FMIPv6.
1. INTRODUCTION
The improvement of the network technologies has provided the use of them in several different
fields. One of the most emergent applications of them is the development of the Vehicular Ad-
hoc Networks (VANETs), one special kind of Mobile Ad-hoc Networks (MANETs) in which the
communications are among the nearby vehicles. Now a day’s vehicular Ad Hoc Network
Communication is the wide area of research topic for Wireless technologies in education
environment as well as automobile industry. Basically Vehicular Ad Hoc Network (VANET’s)
technically based upon the Intelligent Transportation Systems. Today mostly used mobility model
are basically based on the simple random patterns model that cannot describe the vehicular
mobility in the realistic way. Vehicular to vehicular (V2V) communication is the efficient due to
various reasons like that short range, cheapest Communication and better bandwidth. VANET
that is subclass of the MANET consist of the number of the vehicle travelling and communicating
with each other without the fixed infrastructure this is the big benefit of the Ad Hoc Network that
it is not required any fixed infrastructure. We can also characterized VANET on some other
important basis i.e. high mobile node, potential large scale network and variable network density.
VANET also provide facility for wide variety of applications for road safety, driving comfort,
internet access and multimedia (Audio, Video). The ITS applications mostly on infrastructure

24
based communications and used for the Internet Access by the Vehicles. In NEMO there ate
multihop communication is not supported as they designed for the direct communication (single
hop) with the Access Point. In this paper we present a Network Mobility Model for vehicular Ad
Hoc Network and it is the scheme basically used wired cum Wireless Scenario. By the analysis of
simulation, we show that our scheme provides better robust and seamless handover compared to
other scheme.
2. RELATED WORK
Before the proposed handover scheme there are many model are introduced to reduce the packet
loss and for reduce the latency. In order to support the real-time applications in Vehicular Ad-hoc
Networks (VANETs), the proposal of a Leader-based scheme that exploits the topology of
VANETs and a automatic configuration protocol like DHCP ensure the fast and stable address
configuration. However, it assumes the use of a DHCP server and suffers from the control
message overhead problem since it is a proactive protocol. Also, it needs the Duplicate Address
Detection (DAD) when a vehicle changes its leader are describe in [4]. The optimized handover
procedure of FMIPv6 by using Media Independent Handover (MIH) services in VANETs [6].
They introduced the information table to store the static and dynamic information of neighboring
access networks and proposed to use a special cache maintained by the vehicle or by Access
Router (AR) to reduce the anticipation time in FMIPv6, thus increasing the probability of the
predictive mode of FMIPv6. The Global Mobility Management (GMM) was proposed for the
inter-VANET handover of vehicles [10]. The proposed method supports the fast handover
process using the L2 initiating and the route optimization for packet transmission. The Packet
Forwarding Control (PFC) method was proposed in VANETs to select a Common Ahead Point
(CAP) to forward packets [3]. The MMIP6, a communication protocol that integrates multi-hop
IPv6-based vehicular ad hoc networks into the Internet [8]. MMIP6 is highly optimized for
scalability and efficiency based on the principles of Mobile IPv4 and does not provide the
interoperability with IPv6, FMIPv6 and HMIPv6 mobility schemes. A NEMO protocol proposed
for VANETs [9]. The major problem happen in MIPv6 is the handover latency, the time interval
in which Vehicular Node (VN) cannot send or receive packets during handover.
3. Proposed Network Architecture
In the proposed architecture, heterogeneous networks are considered where a mobile network is
provided with mobility through a number of ISPs (Internet Service Provider) as shown in the
figure 1. The MRs (Mobile Routers) are associated with ARs (Access Routers) of different access
network e.g. WLAN network and cellular network. The ISPs have been assigned different home
addresses and home agents to the vehicle equipped with MRs and each MR configures a global
IPv6 address through router advertisement from the AR. During handover process, flow
redirection may create a problem when traffic is redirected between MRs in different
administrative domains. In such cases ingress filtering prevents MR from communicating with
home network of other MRs because MRs belong to different HAs. Thus, the MR is now forced
to use a different home address which can cause the termination of ongoing communication
sessions.

25
Figure 1: Proposed Network Architecture
The assumption made in proposed architecture assumes the following parameters:
(1) Range of AR (ARi) is fixed (i.e. 100 m radius) for better communication.
(2) Distance between two AR’s (dAR) is assumed to be 150 m so the maximum overlapping
region between two AR’s is 50 m on the road as shown in figure 1.
Figure 2:
(3) It is also assumed that the road topology & location of AR’s on roadside is fixed, such
that the minimum and maximum length of overlapping region is 10 m and 50 m
respectively. It is justified by the argument that the width of roads is in the range of 40-
50m in highway as shown in figure by point a, b, c, d.
(4) Another assumption which is also considered is that the traffic is on the road is sufficient
enough that communication between two vehicles with MR’s is always possible with
fixed range of MR’s, it also implies that vehicles are running nearly in the same speed.
The proposed architecture can support any number of vehicles and every vehicle must equipped
with Mobile Router (MR). For vehicular scenario, we can assume any number of vehicles, but for
the alliterations purpose, we have considered two vehicles that are deployed on the road and
every vehicle maintains an ad-hoc network communication. The figure 2 represented a details

26
view of handover region as shown in figure 1. In figure 3, MR1 and MR2 are mounted on first
and second vehicles respectively. When vehicles move from the subnet of AR1 to AR2, vehicle
equipped with MR1 leaves the AR1 and vehicle equipped with MR2 is still in the range of AR1,
so that vehicle connectivity of MR1 to the internet is maintained through by MR2 until MR1
completes handover process. During handover period the traffic addressed to MR1 is transported
via MR2. The vehicle gets internet connectivity through different ARs (AR1, AR2….,ARn) via
MR2 and MR1 respectively for some time until MR2 leaves AR1. During MR2’s handover
period the traffic addressed to MR2 is transported by MR1.
Let A(x1, y1) or B(x2,y2) be any point on the boundary of overlapping region of AR1 or AR2 such
that MR1 enters or exit from point A or B. (See figure 2), for simplicity, we write A(x1, y1) as A
and B(x2,y2) as B.
According to the assumption number (3) the distance travelled by any vehicle in the overlapping
region lies between 10 m ≤ d (A, B) ≤50 m.
Figure 3:
When MR1 comes into the overlapping region then it gets weaker signal and hence disconnection
problem with the internet may arise. But it can be overcome by the fact that enough traffic exists
on the road for communication among the vehicles. This implies that MR1 remains connected to
AR1 since MR2 is still in the range of AR1 and there is ad hoc communication between MR1 and
MR2. At the same time, the MR1 sends message to AR2 for registration. As long as the
registration process is not completed, MR1 gets the message from MR2. The time duration for
MR1 to complete handover process is given by:
=
Or thp =distance travelled in handover region / vehicle speed
where d(A,B) is the distance between overlapping region at point A & B which lies at boundary
region & Vh is the speed of vehicle (MR1)
Formally,
thp is minimum when d(A,B)=10 m &
thp is maximum when d(A,B)=50 m
assuming Vh to be constant in both cases

27
So
4. Proposed Handover Scheme and Tunneling Establishment:
To achieve seamless handover of vehicular network, the proposed scheme is to redirect the traffic
flow of the MRs (which is mounted on the vehicles) to one another via different HAs during
handover period. When the mobile network moves along with the vehicle, the Mobile Router1
(which is installed on first vehicle V1) detects AR2 through Router advertisement message from
AR2. As it moves from the subnet Access Router1, it loses the connection from the global
Internet. The traffic of MR1 should be transported through HA1-MR2 (which is installed on
vehicle V2) using the bi-directional tunnel until it completes handover to AR2. When MR1
moves out of the subnet of AR1and undergoes handover process, it re-establishes the tunnel with
HA1 and executes the following functions:
(i) Tunneling with AR1:
MR1 informs HA1 by Binding Update (BU) message via MR2 to tunnel its packet to MR2,
which is still in the subnet of AR1, as MR1and MR2 are both in the ad-hoc network region.
Figure 4 shows the tunnel established of MR2 during its handover period. The home address,
destination and alternate care-of-address option field of the binding update message contains
MR1 previous care-of-address (IP2) and MR2’s previous care-of-address (IP1) respectively. On
receiving BU message, HA1 creates a mapping between IP1 and IP2 which is used to tunnel the
packets addressed to IP2 through AR1. All the packets routed to MR1 are intercepted by HA1
and the packets are encapsulated and forwarded to IP1. The encapsulated packets routed to MR1
and decapsulated to take IP1 off. The destined address is IP2 which is send by the Correspondent
Node (CN).
Figure 4: Tunnel establishment of MR1 during its handover

28
(ii) New CoA configuration and HA2 registration:
MR1 sends a route solicitation message for configuring a new CoA (IP3) to AR2. MR1 receives
an route advertisement message broadcasted by AR2 and configures IP3. MR1 verifies the
uniqueness of IP3 by sending NA message to AR2 and simultaneously receiving the NAck
message. After IP3 configuration, MR1 deletes the previous Home Agent (HA1) record and
perform the registration of a new home agent (HA2) by sending a new binding update (HA2-BU)
message to HA2. When HA2 replies with a BAck message, it tunnels packet to IP3 through AR2.
After completing the handover process, MR1 reaches in the service area of different service
provider (ISP2) using different access technology than ISP1. ISP2 assigns different home address
and home agent to MR1. The mobile network is now connected with different home networks
through MR2 and MR1 respectively, i.e. MR2 and MR1 are advertising different prefixes on their
ingress interfaces. The prefix advertised by MR2 is registered to HA1 and prefix advertised by
MR1 is registered to HA2. Once MR1 has completed the handover process and started receiving
packets through AR2, now Vehicle equipped with MR2 experience that it has to perform
handover to AR2 in the near time. MR2 can start establishing a new tunnel using alternative
route, this action causes MR2 to perform handover with a little disruption as possible. MR2 then
inform HA1 to tunnel its packets through HA2-MR1 bi-directional tunnel to MR1 which is
located in the subnet of AR2. In this case, if HA2 performs ingress filtering packets with a source
address prefix of MR2 may be discarded.
(iii) Tunneling with AR2:
MR2 can prevent ingress filtering from dropping the packets when the two tunnels end at
different HAs by re-establishing the tunnel with HA1 through CoA configured from the MNP
advertised by MR1. Figure 5 shows the bi-directional tunnel established by the new binding and
for doing so MR2 executes the following functions:
1. MR2 gets its own ingress interface along with a new CoA with the prefix announced by
MR1. Then, it sends BU message to HA1 using new CoA. On receiving BU message,
HA1 can encapsulate incoming packets through the bi-directional tunnel via MR1.
2. All the packets routed to MR2 are intercepted by HA1. The new binding causes HA1 to
tunnel the packets with new CoA of binding cache in HA1 to the MR2. Thus the packets
are encapsulated and forwarded to the new CoA of MR2 via HA2-MR1 bi-directional
tunnel.
After successful BU, all the traffic between CN and MR will experience the two levels of
encapsulation. The first level of tunneling is done between MR2 and HA1 and the second level of
tunneling is done between MR1 and HA2. This mechanism allows MR1 to send and receive
packets using MR2’s bi-directional tunnel and there is no need for MR2 to change HoA and
ongoing sessions do not cause disruption during handover.

29
Figure 5: Tunnel re-establishment of MR2 during its handover
(iv) Depending upon the availability and strength of the radio signal from AR2, MR1 decides
whether to stay in AR1 or move to the AR2. During handover to AR2, MR2 executes the same
sequence of functions as performed by MR1.
The same process will perform in other mobile vehicles which are on the road and try to get
global internet connectivity, through different ARs (AR1, AR2….,ARn).
5. SIMULATION PARAMETER AND RESULTS ANALYSIS
Table 1. Simulation Parameter

30
To evaluate the performance of our proposed scheme and FMIPv6, simulation was performed at
various parameters as shown in Table 1. We use several abstract function in simulation
topologysuch as HA, CN, FA, AR and Vehicular node (VN).
Figure 6:Handover Latency
5.1 Handover latency
Figure 6 shows the handover latency experienced during the simulation of proposed scheme and
is compared with FMIPv6. Handover latency of a mobile network is defined as the complete
handover time from one access router (AR1) to another access router (AR2) and it is proportional
to the distance between MR and HA for all the protocols. However, the latency of the proposed
scheme is less than FMIPv6 because FMIPv6 requires a number of signaling messages in
establishing tunnel between AR1 and AR2 before performing HA-BU. Another reason of less
handover latency is that the second vehicle MR (MR2) can configure its new CoA in advance by
getting Rt Adv information from MR1 for actual handover.
5.2 Service disruption time
In FMIPv6 service disruption time during handover can be defined as the time between the
receptions of last packet from AR1 until the first packet is received from AR2 Via the tunnel
established between AR1 and AR2. In the proposed scheme, service disruption time during
handover can be defined as the time between the reception of the last packet through the MR
which is about to undergo handover process until the first packet is received through another MR
of the mobile network. Figure 7 compares the service disruption time between the proposed
scheme and FMIPv6 with respect to MR-HA latency. The disruption time of FMIPv6 is about 0.9
second (because of the signaling and time it takes in establishing tunnel between AR1 and AR2),
while the service disruption time of the proposed scheme is close to 0.2 second (because only one
signaling message is required to reestablish the bi-directional tunnel). This enhances that the
proposed scheme can support seamless network mobility.

31
Figure 7: Service disruption time
5.3 Packet loss
Figure 8 shows the packet loss during the handover from AR1 to AR2. It is evident from the
graph that lowest number of packet loss is experienced in the case of proposed scheme because of
the cooperative packet reception of MRs. The packets addressed to the MR which is undergoing
handover process are received via another MR by reestablishing the bi-directional tunnel and the
packet loss is independent of the distance between MR and it’s HA. FMIPv6 also establishes a
tunnel between AR1 and AR2 and the packet loss is independent of MR-HA latency. Due to the
number of signaling messages involved in establishing the tunnel, the number of lost packets is
higher than that of the proposed scheme.
Figure 8: Packet loss

32
5.4 Signaling overhead
Figure 9 compares the signaling overhead ratio between FMIPv6 and the proposed scheme in
multiple MRs based mobile network. Signaling overhead involves the exchange of signaling
messages to manage handover process effectively, such as the messages required for re-
establishing the bi-directional tunnel to achieve flow redirection of MRs via one another. The
proposed scheme has some overhead because the cast involves in maintaining signaling messages
of MRs. However, it has less overhead in comparison to FMIPv6, because FMIPv6 involves a lot
of signaling messages to establish a tunnel between AR1 and AR2. The proposed scheme
increases the signaling overhead in the network and signaling overhead of these protocols
increases with increasing MR-HA latency.
Figure 9: Signaling Overhead ratio
6. CONCLUSION
The proposed architecture support seamless mobility of mobile network across heterogeneous
network. A vehicular scenario is considered where a vehicle is provided with mobility through
different Internet Server Protocol’s (ISPs) and purpose to use multiple Mobile Router s (MR)
based handover schema in vehicle. The multiple Mobile Router (MR) based handover schema
where MRs cooperatively receive packet destined for each other can provide no service
disruption and significantly reduce packet loss during handover. It also makes the packet loss
independent to handover latency. Moreover Multiple Mobile Router’s (MR’s) architecture is
extended to include Multiple Home Agent’s, where each Home Agent belongs to different
administrative domains. This endows the vehicle to be able to do smooth handover over
heterogeneous network where mobility is provided through different Internet Server Protocols
(ISP’s). An extensive simulation study is carried out to show the comparative performance
evolutions of the proposed handover architecture in terms of throughput, handover latency service
disruption time, packet loss and signaling overhead. The simulation results shows that the
proposed architecture provide a mobile network with seamless mobility across heterogeneous

33
networks. The overlapped reception of packet from different Access Router’s (AR’s) significantly
minimizes packet losses during handover even without reducing handover latency.
REFERENCES
[1] Himanshu Tyagi1, A. K. Vatsa1 (2011) “Seamless Handoff through Information Retrieval in VANET
Using Mobile Agent” Vol. 8.
[2] Muhammad Nawaz Khan,Ishtiaq Wahid & Gulzar Mehmood (2012) “A Global Mobility
Management Scheme for Reducing Overall Handover Latency in IP based VANETs” in IJASUC)
Vol.3.
[3] E. Karamad and F. Ashtiani, “A modified 802.11based MAC scheme to assure fair access for vehicle-
to-roadside communications,” Computer communications, Vol. 31, 2008.
[4] M. Bechler and L. Wolf, “ Mobility management for vehicular ad hoc networks,” in proceeding of
IEEE vehicular technology, 2005.
[5] Roberto Baldessari , Andress Festag and JulienAbeille,(2007)“Nemo Meets VANET : A
Deployability analysis of Network Mobility in Vehicular Communication” In ,IEEE.
[6] Y. S. Chen, C. H. Cheng, C. S. Hsu and G. M. Chiu, “ Network Mobility protocol for vehicular ad
hoc,” in proceeding of IEEE WCNC, 2009.
[7] Wu Ming Yang Lin-tao ,Li Cheng-yi and Jiang Hao (2008) “Capacity ,collision and interference of
VANET with IEEE 802.11 MAC” in IEEE.
[8] T. Arnold, W. Lloyd, J. Zhao and G. Cao, “ IP address passing for VANETs,” in proceedings of IEEE
percom, 2008.
[9] J. M. Lee, M. J. Yu, Y. H. Yoo and S. G. Choi, “ A new scheme for global mobility management for
inter-VANETs handover of vehicles in V2V and V2I network environments,” in proceeding of IEEE
NCM, 2008.
[10] S. Balasubramaniam and J. Indulska, “Handovers between heterogeneous networks in pervasive
systems,” in proceedings of ICCT, Vol.2, 2003
[11] V. Devarapalli, R. Wakikawa, A. Petrescu, P. Thubert (2005) “Network Mobility (NEMO) Basic
Support Protocol”.
[12] Ana Gainaru, Ciprian Dobre, Valentin Cristea (2009) “A Realistic Mobility Model based on Social
Networks for the Simulation of VANETs” in IEEE.
[13] Mrs. Vaishali D. Khairnar, Dr. S.N. Pradhan (2011) “Mobility Models for Vehicular Ad-hoc
Network Simulation” in IEEE.
[14] Kun-chan Lan and Chien-Ming Chou (2008) “Realistic Mobility Models for Vehicular Ad hoc
Network (VANET) Simulations” in IEEE.
[15] Md. Asif Nashiry1, Md. Abdullah-Al-Mahmud1, Md. Mahbubur Rahman1, Md. Nashid Anjum2
(2008) “Evaluation of TCP Performance over Mobile IP Wired-cum-Wireless Networks” in IEEE.
[16] Gustavo Marfia,Giovanni Pau,Eugenio Giordanotl,Enzo De Senat, Mario Gerla (2007) “VANET: On
Mobility Scenarios and Urban Infrastructure” in IEEE.
[17] Reinaldo Bezerra Braga and Herve Martin (2011) “Understanding Geographic Routing in Vehicular
Ad Hoc Network” in IARIA.
[18] Ines Ben Jemaa, Manabu Tsukada,Hamid Menourar and Thierry Ernst, ( 2010)” Validation and
evolution of NEMO in VANET using geographic routing” ,in ITST.
[19] FleetNetDocument www.et2.tuharburg.de/fleetnet/english/documet.html.
[20] Sherali Zeadally • Ray Hunt • Yuh-Shyan Chen •Angela Irwin • Aamir Hassan (2010) “Vehicular ad
hoc networks (VANETS): status, results, and challenges” in Springer Science+Business Media, LLC.

34
Authors
Ravi S Shukla obtained a M.Tech. Degree in Software Engineering in 2002 from MNNIT,
Allahabad, Deemed University. Currently, he is working as an Associate Professor in the
Dept. of CSE & IT and research scholar of MNNIT, Allahabad in the dept. of CSE. His
research interest includes Computer Networking, TCP/IP, Mobile Computing and Adhoc
networking.
Dr Neeraj Tyagi obtained a Ph.D. degree in 2008 in Computer Science & Engineering from
MNNIT, Allahabad, Deemed University. Currently, he is working as an Associate Professor
in the Dept. of CSE. His research interest includes Computer Networking, TCP/IP, Mobile
Computing and Adhoc networking.

Global mobility and Handover management for heterogeneous network in vanet

More Related Content

Similar to Global mobility and Handover management for heterogeneous network in vanet (20)

More from josephjonse (20)

Recently uploaded (20)

Global mobility and Handover management for heterogeneous network in vanet