D.Rajini Girinath#1, Dr.S.Selvan*2
                                            Dept of CSE, MNM Jain Engg College,
                                             Thoraipakkam, Chennai-97,India.
                                                Principal, St. Peters Engg College,
                                                      Avadi, Chennai, India.

   Abstract--This paper presents a comparative simulation study      Especially in the VANET, the mobility model is an important
of Grid Mobility model and Reference Point Group mobility            factor that creates realistic moving behavior of VNs.
models on the performance study of VANET that uses Ad-hoc
On-Demand Distance Vector (AODV) as the routing protocol. It           Grid mobility model is simple and is widely used to
is a crucial part in the performance evaluation of VANET.
Reference Point Group mobility model (RPGM) is the only model
                                                                     evaluate the performance of VANET. Most previous studies
that has been widely used in the simulation study of VANET           on mobility modeling and analysis considered protocol case
despite some unrealistic movement behaviors such as sudden stop      studies for on-demand and table-driven ad hoc routing
and sharp turn. Whilst Grid mobility model has been proved           protocols. Many other MANET protocols and services were
that it can solve both of these problems. The results show that      not evaluated over a rich set of mobility models. There is a
both mobility models are not different in case each MN is moving     need to re-visit MANET protocols and service architectures
at human running speed. Therefore, it is suggested to use RPGM       and study their performance over various mobility models.
mobility model because of its less computational overhead            Recent case studies [4], [5] considered mobility effects on
comparing to Grid mobility model. At the time of using realistic     geographic routing protocols.
traces, the performance result using           mobility model is
significant different from RPGM mobility model. Therefore,
RPGM mobility model should be used instead. Moreover,                  In this paper, the performance of VANET using Ad-hoc
different levels of randomness setting have no effect on the         On-Demand Distance Vector (AODV) routing protocol is
accuracy of throughput and end-to-end delay.                         evaluated by comparing the use of Grid and RPGM mobility
                                                                     models. RPGM mobility model has been proved more realistic
                    I. INTRODUCTION                                  movement pattern of VNs.

  A Vehicular ad hoc network (VANET) is a network                                        II. MOBILITY MODELS
consisting of a set of wireless mobile nodes that communicate
with each other without centralized control or established              In order to thoroughly simulate a new protocol for an ad hoc
infrastructure. The mobility model represents the moving             network, it is imperative to use a mobility model that
behavior of each mobile node (MN) in the VANET that                  accurately represents the vehicle nodes (VNs) that will
should be realistic. A Vehicular Ad hoc Network is a network         eventually utilize the given protocol. Only in this type of
comprising wireless vehicle nodes (VN) that communicate              scenario is it possible to determine whether or not the
with each other without centralized control or established           proposed protocol will be useful when implemented. Currently
infrastructure. VNs that are within each other's radio range can     there are two types of mobility models used in the simulation
communicate directly, while distant VNs rely on their                of networks: traces and synthetic models. Traces are those
neighboring VNs to forward packets. Each MN acts as either a         mobility patterns that are observed in real life systems. Traces
host or a router. In VANET environment, VNs are free to join         provide accurate information, especially when they involve a
or leave the network at any point of time, resulting in a highly     large number of participants and an appropriately long
dynamic network environment compared to wired network                observation period. However, new network environments (e.g.
[1]. The considerations about developing routing protocols for       ad hoc networks) are not easily modeled if traces have not yet
VANET are computation-restricted, bandwidth constrained,             been created. In this type of situation it is necessary to use
and energy-constrained. For a new protocol development,              synthetic models. A mobility model should attempt to mimic
performance evaluation is important and essential because the        the movements of real VNs [2]. Changes in speed and
result can be used in many applications. Its performance can         direction must occur and they must occur in reasonable time
be evaluated by two typical techniques: simulation and               slots. For example, we would not want VNs to travel in
analysis. Simulation is used in many research works.                 straight lines at constant speeds throughout the course of the
                                                                     entire simulation because real VNs would not travel in such a
restricted manner. In Section 2, we discuss seven different          traveled by a logical center for the group as shown in Fig 2.
synthetic entity mobility models for ad hoc networks:                The logical center for the group is used to calculate group
                                                                     motion via a group motion vector. The motion of the group
1. The Random Walk Mobility Model with a small input                 center completely characterizes the movement of its
parameter produces Brownian motion and, therefore, basically         corresponding group of MNs, including their direction and
evaluates a static network when used in a performance                speed. Individual MNs randomly move about their own pre-
investigation. A large input parameter is similar to the             defined reference points, whose movements depend on the
Random Waypoint Mobility Model without pause times. The              group movement. As the individual reference points move
main difference between these two mobility models is that            from time t to t+1, their locations are updated according to the
MNs are more likely to cluster in the center of the simulation       group’s logical center.
area with the Random Waypoint Mobility Model.

2. The Random Waypoint Mobility Model is used in many
prominent simulation studies of ad hoc network protocols. It is
flexible, and it appears to create realistic mobility patterns for
the way people might move in, for example, a conference
setting or museum. One concern with this model is the straight
movement pattern created by the MN to the next chosen

3. The Gauss-Markov Mobility Model also provides
movement patterns that one might expect in the real-world, if
appropriate parameters are chosen. In addition, the method
used to force MNs away from the edges of the simulation area
is of note.

4. The City Section Mobility Model appears to create realistic                       Fig. 1. RPGM-Vanet Implementation
movements for a section of a city, since it severely restricts the
traveling behavior of MNs; MNs do not have the ability to              Once the updated reference points, RP (t+1), are calculated,
roam freely without regard to obstacles and other traffic            they are combined with a random motion vector, RV, to
regulations.                                                         represent the random motion of each MN about its individual
                                                                     reference point. Both the movement of the logical center for
Regarding the five synthetic group mobility models for ad hoc        each group, and the random motion of each individual MN
networks, the following list summarizes our conclusions.             within the group, are implemented via the Random Waypoint
                                                                     Mobility Model. One difference, however, is that individual
1. The Column, Nomadic Community, and Pursue Mobility                MNs do not use pause times while the group is moving.
Models are useful group mobility models for specific realistic
scenarios. The movement patterns provided by these three                                          TABLE 1
mobility models can be obtained by changing the parameters                           Groups specified in the RPGM model
associated with the Reference Point Group Mobility Model.
                                                                              Number        of   Total nodes        Percent    of
                                                                              Clusters                              total
2. The Reference Point Group Mobility Model (RPGM) is a
generic method for handling group mobility. An entity                                       9                  20             32%
mobility model (or models) needs to be specified to handle                                  5                  15             26%
both the movement of a group of MNs and the movement of
                                                                                            7                  18             30%
the individual MNs within the group. The input parameters of
the RPGM model allow the flexibility to implement the                                       4                  12             11%
Column, Nomadic Community, and Pursue Mobility Models.                                   Total                 65             99%

III. REFERENCE POINT GROUP MOBILITY MODEL                              The RPGM model was designed to depict scenarios such as
                                                                     an avalanche rescue. During an avalanche rescue, the
  The Reference Point Group Mobility (RPGM) model                    responding team consisting of human and canine members
represents the random motion of a group of MNs as well as            work cooperatively. The human guides tend to set a general
the random motion of each individual MN within the group as          path for the dogs to follow, since they usually know the
shown in Fig 1. Group movements are based upon the path              approximate location of victims. The dogs each create their
own “random” paths around the general area chosen by their
human counterparts.

                                                                                Fig. 3. Grid mobility model-Ns2 Implementation

                 Fig. 2. RPGM –Ns2 implementation.

   If appropriate group paths are chosen, along with proper
initial locations for various groups, many different mobility
applications may be represented with the RPGM model. In [6],
three applications for the RPGM model are defined. First, the
In-place Mobility Model partitions a given geographical area
such that each subset of the original area is assigned to a
specific group; the specified group then operates only within
that geographic subset. Second, the Overlap Mobility Model
simulates several different groups, each of which has a
different purpose, working in the same geographic region;
each group within this model may have different
characteristics than other groups within the same geographical
boundary. For example, in disaster recovery of a geographical
                                                                                      Fig. 4. Grid-Vanet implementation
area, one might encounter a rescue personnel team, a medical
team, and a psychologist team, each of which have unique
                                                                                      V. SIMULATION MODEL
traveling patterns, speeds, and behaviors.
                                                                        The simulation model was based on the Network Simulation
                IV. GRID MOBILITY MODEL
                                                                     (NS2) and VANET. An unslotted carrier sense multiple access
                                                                     with collision avoidance (CSMA/CD) is used for data
   We introduce the model to emulate the movement pattern
                                                                     transmission in MAC layer. The radio model uses
of mobile nodes on streets defined by maps. It can be useful in
                                                                     characteristics similar to a commercial radio interface. In the
modeling movement in an urban area where a pervasive
                                                                     simulation study, the Ad-hoc On- Demand Distance Vector
computing service between portable devices is provided. Maps
                                                                     (AODV) was used as the routing protocol. Table 2 provides all
are used in this model too. However, the map is composed of a
                                                                     the simulation parameters of      Grid and RPGM mobility
number of horizontal and vertical streets as shown in Fig 4.
The mobile node is allowed to move along the grid of                                             TABLE 2
horizontal and vertical streets on the map. At an intersection of                        Simulation parameter values
a horizontal and a vertical street, the mobile node can turn left,
right or go straight with certain probability. Except the above            Time of simulation         10.0
difference, the inter-node and intra-node relationships                    Routing protocol           AODV
involved in the model are the same as in the Freeway model                 Number of nodes            100
[3]. Thus, the mobility model is also expected to have high                Network Interface          Phy/WirelessPhy
spatial dependence and high temporal dependence. It too                    Bandwidth                  11 Mb
                                                                           Traffic Type               CBR
imposes geographic restrictions on node mobility. However, it
                                                                           Max Packet in queue        50
differs from the Freeway model in giving a node some
                                                                           MAC protocol type          MAC/802.11
freedom to change its direction as shown in Fig 3.
                                                                           Packet size                1500 bytes
                                                                           Area size                  1000*1500
                VI. SIMULATION RESULTS

6.1. Throughput

  The throughput of VANET using the RPGM mobility model
is equal to the arrival packet rate if all data packets are
successfully transmitted. When the number of MNs is 50
nodes, the arrival packet rate is 40 pps that is equal to the
packet rate sent from source MNs (99%). However, when the
number of MNs is 100 nodes, the arrival packet rate are 75
pps, but the simulation result shows the received packet rate of
approximately 60-70 pps (72.5-78.5%). When the number of
MNs increases, the network congestion and packet loss occur
[6]. When pause time is closed to 0, the value of throughput is
independent of the number of MNs but depends on pause time
and speed. When pause time is high and speed is low, the                         Fig. 6. End- End delay for RPGM model.
throughput increases. Fig 5 shows the effect of a on
throughput of VANET using the Grid mobility model. It has
no effect on the level of throughput. Therefore, the value of a
can be chosen to meet the requirement of a particular scenario.

                                                                                  Fig. 7. End-End delay for Grid model.

                                                                     Fig 7 shows the effect of degree of random on the end-to-
                                                                   end delay of VANET using the Grid mobility model. The level
                                                                   of randomness almost gives the same results for both end-to-
                        Fig. 5. Throughput                         end delay and throughput.
6.2. End-to-End Delay
                                                                   6.3. Energy Efficient Routing
  Fig 6 shows the effect of end-to-end delay of VANET using
                                                                     Ideally, we would like the Vanet to perform its functionality
the RPGM mobility model. When the number of MNs is 50
                                                                   as long as possible. Optimal routing in energy constrained
nodes, the end-to- end delay is very small since data packets
                                                                   networks is not practically feasible [7]. However, we can
can be sent to the destination immediately. When the number
                                                                   soften our requirements towards a statistically optimal
of MNs is increased to 100 nodes, the end-to-end delay
                                                                   scheme, which maximizes the network functionality
increases because of the time consumed for route discovery
                                                                   considered over all possible future activity.
and the increasing number of packets in the buffer. However,
when the pause time is increased, the network is stable and the      In most practical surveillance or monitoring applications,
end-to-end delay decreases. With normal speed, the end-to-         we do not want any coverage gaps to develop. We therefore
end delay is low because the network is not congested. If          define the lifetime we want to maximize as the worst-case
pause time is closed to 0, the end-to-end delay is minimized       time until a node breaks down, instead of the average time
and the throughput is maximized since there is a small amount      over all scenarios. However, taking into account all possible
of packets in the buffer.                                          future scenarios is too computationally intensive, even for
                                                                   simulations as shown in Fig 8. It is therefore certainly
unworkable as a guideline to base practical schemes on.            model have no effect on the accuracy of throughput and end-
When a node detects that its energy reserve has dropped below      to-end delay.
a certain threshold (50% in our simulations), it discourages
others from sending data to it by increasing its height. This                                 REFERENCES
may change a neighbor’s height (since a node’s height is one
more than that of its lowest neighbor). It in turn informs other   [1] T.D.C. Little and A. Agarwal, ”An Information Propagation Scheme for
nodes and these updates are propagated as far as is needed to          VANETs”, Department of Electrical and Computer Engineering, Boston
                                                                       University, Boston, Massachusetts 02215, USA.
keep all the gradients consistent as shown in Fig 9.               [2] Andre E. Bergh, Student Member, IEEE Neco Ventura, Member, IEEE,”
                                                                       Movement Prediction Assisted Fast Handovers for Seamless IP
                                                                       Mobility”, 2007 IEEE.
                                                                   [3] Babak Pazand, Chris McDonald,” A Critique of Mobility Models for
                                                                       Wireless Network Simulation”, University of Western Australia, 6th
                                                                       IEEE/ACS International Conference on Computer and Information
                                                                       Science (ICIS 2007).
                                                                   [4] Fan Bai and Ahmed Helmy,” A SURVEY OF MOBILITY MODELS in
                                                                       Wireless Adhoc Networks, University of Southern California, U.S.A.
                                                                   [5] Fan Bai and Ahmed Helmy,” THE IMPORTANT FRAMEWORK for
                                                                       Analyzing and Modeling the Impact of Mobility in Wireless Adhoc
                                                                       Networks, University of Southern California, U.S.A.
                                                                   [6] Yufang Zhu and Thomas Kunz,” MAODV Implementation for NS-
                                                                       2.26”, Systems and Computing Engineering, Carleton University,
                                                                       Technical Report SCE-04-01, January 2004,Carleton University
                                                                   [7] Kevin Fall ,Kannan Varadhan,” The ns Manual (formerly ns Notes and
                                                                       Documentation)” ,January 25, 2005.

             Fig. 8. Energy consumption for RPGM model

              Fig. 9. Energy consumption for Grid model

                      VII. CONCLUSION

   This paper presents a comparative simulation study of
Grid and RPGM mobility models on the performance study of
VANET that uses Ad-hoc On-Demand Distance Vector
(AODV) routing protocol. The results show that both mobility
models are not different in the normal real time scenario. In
this case, it is suggested to use RPGM mobility model because
of its less energy consumption compared to Grid mobility
model. When the speed of MNs is as high as fast automobile,
the performance result using RPGM mobility model is
significant different from Grid mobility model. Moreover,
different levels of randomness setting in the RPGM mobility

To top