See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/278405374

Analyzing Optimal Setting Of Reference Point Group Mobility Model Using

DSR Protocol In MANETS

Conference Paper · May 2015

DOI: 10.15849/icit.2015.0032

CITATIONS READS

0 130

2 authors:

Naseer Ali Husieen Mohammad M. Rasheed

Wasit University UNIVERSITY OF INFORMATION TECHNOLOGY & COMMUNICATIONS
25 PUBLICATIONS   15 CITATIONS    26 PUBLICATIONS   28 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Prevalence of Toxoplasma gondii in free range eggs detected by smear impression and PCR. View project

IoT-Blockchain Digital Trust View project

All content following this page was uploaded by Mohammad M. Rasheed on 12 September 2015.

The user has requested enhancement of the downloaded file.

ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

Analyzing Optimal Setting Of Reference Point

Group Mobility Model Using DSR Protocol In
MANETS
Nasser Ali Husieen
Department of Computer Science, College of Education
Wasit University
Kut, Iraq
Email: [email protected]

Mohammad M. Rasheed
Information Technology Directorate
Ministry of Science and Technology
Baghdad, Iraq
Email: [email protected]

Abstract— Mobile Ad-hoc Network (MANET) is a self-configuring wireless network. The nodes can configure themselves to be in
any arbitrary topology. It is essential that mobility models used in simulating different scenarios must emulate closely the real scenario
in order to assess the MANET performance as accurately as possible. In this paper, the authors have studied the effect of different
maximum pause times and maximum node speeds on different performance metrics in order to arrive at optimal settings for these two
attributes under the Reference Point Group Mobility model for the DSR protocol. In addition, this work is part of an ongoing research
on link failures in DSR protocol. Thus, the performance of the DSR protocol under the Reference Point Group Mobility (RPGM) model
in terms of different pause times, node speeds, number of nodes and number of source connections were evaluated. The simulation
results show that the maximum pause time and the maximum speed have direct impacts on the performance parameters such as packet
delivery ratio, routing overhead, average end-to-end delay, normalized routing load and packet drop under the Reference Point Group
Mobility model.

Keywords— MANET, Mobility Models, Routing Protocols, DSR Protocol, RPGM

the data packets between those nodes in an efficient manner.
I. INTRODUCTION In mobile ad hoc networks, high mobility, limited computing
Mobile ad hoc network is a collection of wireless nodes capability and low bandwidth associated with nodes make
that allows devices to communicate with each other without routing of data one of the most challenging tasks. Researchers
help of an existing infrastructure [1]. A MANET is a self have proposed different routing protocols to address these
configuring and a self-organizing such that it can create an issues. These protocols can be categorized into two main
arbitrary topology temporarily for short while in which mobile groups. They are namely on demand or reactive protocols and
nodes work as both routers and end nodes. MANET based periodic or proactive protocols. Protocols such as Dynamic
applications range from military applications connecting Source Routing (DSR) [2], Ad-hoc On-Demand Distance
soldiers in a battlefield, social networks for communication Vector Routing ( AODV) [3] and Temporally ordered routing
during emergencies such as natural disasters to personal area algorithm (TORA) [4] coming under the reactive protocols
networks. New applications such as telemedicine, weather establish the route only prior to sending the packets and
reporting and dissemination and disaster information maintain the route only when it is needed. The proactive
dissemination are emerging in the recent times exploiting the protocols such as Destination-Sequenced Distance Vector
new developments and the advantages of MANETs compared (DSDV) [5] on other hand periodically exchange routing
to the traditional infrastructure based networks. The above information to maintain the routes continuously. In general,
mentioned developments will increase the size and reach of reactive protocols have been shown to outperform the
MANETs to thousands of nodes which is hitherto limited to proactive protocols due to their reduced overhead and ability
few nodes in both military and civil application domains. to react quickly when routes change [6-9]. This motivated us
to investigate more into on demand protocols. One of the most
One of the main challenges in setting up and managing a important and difficult tasks in simulating mobile ad hoc
dynamic mobile ad hoc network is routing. Routing protocols networks is the definition of a mobility model to represent real
detect the optimum path between the source node and world scenarios. In the real world, mobile networks encounter
destination node in a complex network of nodes and deliver various situations including intense mobility creating heavy

Page | 213
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

uncertainties especially in disaster situations. Hence taking

measurements under real world conditions is an almost
impossibility. The alternative to real world measurements is to
simulate the different situations using a computer software
represent the real world situation as closely as possible and
take the measurements. Presently in the mobile ad hoc Fig. 2 shows the route maintenance mechanism of DSR
environments, the following mobility models have been used; protocol. In this mechanism, DSR used two types of packets:
Random Way Point, Manhattan Grid, Gauss Markov mobility ACKs packet is used for correction and verification operation
model and Reference Point Group Mobility (RPGM) model of the routes, and second type of packets is called Route Error
[10]. In this paper, we use the RPGM model implemented in Packet (RRER). RRER is generated when there is link failure
the Network Simulator 2 (NS2) version 2.34 under Linux to occurs between intermediate nodes and next hop due to battery
study the effect of pause time and node speed on the DSR usage, hard medium contention, and the node mobility that is
protocol. Different scenarios were simulated and the leading to loss the packets as shows in the figure below when
performance of the network was critically evaluated with the link fails between the intermediate node I and destination
special reference to these parameters. The main objective of node D.
the analysis was to determine the optimal settings for these
two parameters.
The rest of the paper is organized as follows: Section 1 Fig. 1 shows the route discovery mechanisms. In this figure,
provides an introduction to the study. Sections 2 and 3 discuss when a source node (A) wants to send data packets to the
RPGM model in details and related work respectively. Details destination node (J). First, it checks if there is a route to the
of the simulation environment setup and results are presented destination in the route cache, route request will be discarded.
in Sections 4 and 5. Finally Sections 6 concludes the paper Otherwise, source node (A) broadcasts Route Request Packet
along with suggestions for future work. (RREQ) to the neighbour nodes (B, C, D) within wireless
transmission range of a node (A). The route request identifies
II. DYNAMIC SOURCE ROUTING PROTOCOL the destination node J to which a route in needed. If the
neighbour nodes (B, C, D) are not the target, locally these
Dynamic Source Routing (DSR) protocol is a simple reactive
neighbour nodes rebroadcast RREQ to the next hop after
routing protocol developed by Johnson at the Carnegie Mellon
adding its own address to a list of nodes in the route cache.
University in 1996. The key feature of the DSR protocol is
Each RREQ packet contains source address, destination
based on the source routing where full or partial route is
address, request ID, and route record. When a route request
specified by the sender [11]. In mobile ad hoc networks, the
packet arrives at the target node (J), the destination node (J)
source routing technique provides several advantages,
returns a Route Reply Packet (RREP) along with the reverse
including flexibility, simplicity and correctness [12-14]. When
of a recorded path to the source node (A), which is (A, D, G,
a node wants to send data packets to another node, the
J). When a source node (A) receives more than one RREP for
intermediate nodes forwarding the packets towards the
a given destination, it selects the first route that receives
detonation need not maintain up to date routing information as
RREQ in order to reduce the time for route discovery packet.
the data packet will contain information on how to forward the
DSR protocol is made up of two mechanisms namely, Route
packet. DSR protocol is made up of two mechanisms namely,
Discovery and Route Maintenance. DSR uses route cache to
Route Discovery and Route Maintenance. DSR uses route
store the routes to other nodes. The main advantage of using
cache to store the routes to other nodes. The main advantage
route cache is it speeds up route discovery and reduces the
of using route cache is it speeds up route discovery and
propagation of route request packets.
reduces the propagation of route request packets. DSR can
also provide interconnection between wireless devices with
multiple network interfaces. This is vital in strategic
communications as nodes in the military require
communicating with different devices at different ranges

Fig.2. Route Maintenance Mechanism

Fig.1. Route Discovery Mechanism

Page | 214
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

III. REFERENCE POINT GROUP MOBILITY MODEL Fig.3. Node Movement in RPGM Model

The Reference Point Group Mobility (RPGM) model was

proposed by Hong [15]. In this model, all nodes work as a
With appropriate selection of the predefined paths for the
group and the nodes of the group moves as a single entity to
group leader and other parameters, the RPGM model can
achieve different tasks. Each group has the logical centre
emulate a variety of mobility behaviours. The RPGM model is
called the group leader. The path of the group in its entirety is
thus able to represent various mobility scenarios, such as:
represented by the locus of the centre. Each node in the group
has its own reference point for communicating with other
nodes. The reference point of the node follows the group  In-Place Mobility Model: In this model, the whole field is
movement; the real location of the node can be determined by divided into adjacent regions. Each region is totally
its reference point plus a random motion vector that denotes occupied by a single group. An example of this model is
its stability from the reference point. Reference point Group battlefield communication.
mobility is adapted for several applications such as a
battlefield situation where a number of soldiers move together  Overlap Mobility Model: In this model, different groups
in a group, disaster recovery and convention scenarios. with different tasks move on the same field in an
According to Hong's report, the RPGM outperforms Random overlapping style. An example of this model is Disaster
Way Point model in case of link failures due to the inherent relief.
characteristic of spatial dependency between nodes. The
RPGM model incurs fewer link breakages and achieves better  Convention Mobility Model: In this model, the area is
performance for various routing protocols compared to divided into few regions and several groups are permitted
Random Way Point model. The functions of the group leader to move between the regions. An example of this model is
and group members are as follows: a conference.
In RPGM model, the vector RMi indirectly determines how
A. Group Leader far the group members deviate from their leader. The
𝑡 movement can be characterized as follows:
𝑉𝑔𝑟𝑜𝑢𝑝 It provides the general motion movement of the
whole group. Each member of this group moves away from
𝑡
this group motion. The motion vector 𝑉𝑔𝑟𝑜𝑢𝑝 can be arbitrarily
selected or carefully designed based on several predefined
paths.
(2)
B. Group members
The group members’ movement is heavily affected by its
group leader’s movement. Each mobile node is assigned with Where 0 < SDR, ADR < 1,
a reference point that follows the group movement. With SDR is the Speed Deviation Ratio and
respect to this predefined reference point, every mobile node
might be arbitrarily located in the neighborhood. Formally, the ADR is the Angle Deviation Ratio
motion vector of the group members i, at the time t, Vit can
be defined as:
SDR and ADR are used to control the deviation of the
Vit = Vgroup
t
+ RMit (1) velocity in terms of both magnitude and direction of group
Where 𝑅𝑀𝑖𝑡 is the random motion vector representing the members from that of the leader. Different mobility scenarios
deviation of the group member i from its reference point. The can be generated by adjusting these two parameters.
vector 𝑅𝑀𝑖𝑡 is free identically distributed random procedure
whose duration is uniformly distributed in the interval IV. RELATED WORK
[0, 𝑟𝑚𝑎𝑥 ], where 𝑟𝑚𝑎𝑥 is maximum acceptable distance and Different mobility models have different impacts on the
whose path is uniformly distributed in the interval [0, 2π]. performance of mobile ad hoc routing protocols [16].
Fig. 3 illustrates the Reference Point Group Mobility Different protocols have different metrics that capture
model with the group leader represented in green and the interesting mobility characteristics like spatial and temporal
t
members represented in red and yellow respectively. Vgroup dependence and geographic restrictions. Hong et al [15]
introduced a mobility model called the Mobility Vector (MV)
is the motion vector of the group leader and the whole group.
model and compared the performance of different routing
protocols such as DSR, AODV and FSR against other
mobility models including Random Way Point, RPGM and
Random Walk. Packet delivery ratio and link Up/Down were
measured using simulation varying the average speed and
transmission range respectively. Hence it is important to select
the right mobility model to represent the real scenario under

Page | 215
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

consideration. Geetha and Gopinath compared the was to select a suitable model for AODV protocol. The
performance of two on demand routing protocols, namely metrics used by them were packet delivery ratio and delay.
AODV and DSR under different mobility models such as Their simulation result shows that the AODV performs well
Random Way Point and Reference Point Group Mobility in with RPGM model in terms of packet delivery ratio and end to
order to characterize the two routing protocols under the end delay. The simulation parameters used were fixed pause
different mobility models [17]. Simulation parameters time of 10 s, different velocities of 5, 10, 15, 20, 25 m/s and
computed included fixed pause time at 25 s, five low traffic increased network load of 4, 8, 12 and 16 packets per second.
source and speeds up to 20 m/s. In this study, the authors The work presented by this paper focuses on velocities up to
concentrate on different pause times and maximum speed with 80 m/s with different pause times up to 40 with the objective
increased number of source connection along with increased of selecting the average optimal setting for these two
node density in order to find the optimal setting of pause time parameters.
and node speed for the DSR protocol in the Reference Point
Group Mobility. V. SIMULATION ENVIRONMENT
Kioumourtzis evaluated the performance of OLSR, AODV The Network Simulator 2 (NS2) version 2.34 installed on
and DSR three reactive protocols under Manhattan-Grid, and Centos Linux operation system was used as the simulation tool
Reference Point Group mobility models [18]. The main in this work. In NS2 the node movement has to be defined in a
objective of his work was to compare the performance of each OTCL script or be imported from an external file. In this
protocol under the two models in order to understand the project, the mobility scenarios were created using the Bonn
limitations of the protocols. The simulations were carried out Motion version 1.4 a Java software tool specifically designed
under different conditions. The number of nodes was varied for this purpose [20]. Bonn Motion was developed by the
up to 90, the number of connections was increased up to 40, Communication Systems Group at the Institute of Computer
the packet rate was also increased and the nodes’ speeds of Science 4 of the University of Bonn, Germany. This tool can
movement were fixed at 5, 10, 15 and 20 m/s along with a generate the most common mobility models such as Random
pause time of 5 s. However, this paper presents the Way Point, Gauss-Markov and Manhattan-Grid models. In
performance of the DSR protocol with increased node density, this project, the Reference Point Group mobility (RPGM)
increased CBR traffic and various nodes’ mobility velocities model has been used as a movement model and the cbrgen.tcl
up to 80 m/s with different pause times up to 40 s in order to generator tool which is located under the directory indep-
select the optimal setting for the DSR protocol under utils/cmu-scen-gen has been used to generate random source
Reference Point Group Mobility model. traffic. The traffic pattern used is CBR. Fig. 4 shows the
Agrawal, Tiwari and Vyas have evaluated the AODV simulation methodology adopted in this work. Each scenario
protocol under four different mobility models, namely the was simulated for 200 seconds within a simulated rectangular
Manhattan-Grid, Markov-Grid, Random Way Point and geographical area of 1000 m x 500 m. Tables 1 and 2 lists the
Reference Point Group Mobility model [19]. Their objective rest of simulation parameters.

Page | 216
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

Fig 4. Simulation Methodology

TABLE 1: Simulation Parameters Standard deviation 2.0

Parameters Value MAC Type 802.11
Simulation Time 200 s
Number of nodes 10, 20, 40, 80 According to the data listed in Table 1, there are eight
Number of connections 4, 8, 30, 40 main scenarios comprising four main scenarios of different
speeds and another four main scenarios of different pause
Maximum Pause Time 5, 10, 20, 30, 40 times. In the case of first main scenarios, for each speed was
tested under four sub-scenarios by changing different metrics
Simulation Area 1000 x 500 m at a time making the total number of scenarios under this
Minimum node speed 0 m/s category to be 16. Under each pause time, five sub-scenarios
were created using different additional parameters making the
Maximum node speed 20,40,60,80 m/s total number of scenarios under this category to be 20. Thus,
Mobility Model RPGM the total number of scenarios tested in the experiment is 36.
Since the performance of ad hoc routing protocols is sensitive
Routing Protocol DSR to the movement pattern, scenario files were generated with 50
different movement patterns representing 10 movement
Traffic Type Constant Bit Rate
(CBR)
./bm –f scenatio1 –b RPGM –n 80 –d 200 –x 1000 –y
Packet size 512 bytes 500 –h 20.0 –l 0.0 –p 20.0 –a 5.0 –c 0.05 –r 100 –s
Distribution of nodes 5 groups 2.0
Probability of group change 0.05
patterns per pause time.
Maximum distance to group 100 m
centre

Page | 217
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

Table 2 lists the parameters used in generating the RPGM centre

mobility model in Bonn Motion.
-s Group size standard
The following command will generate the Reference Point deviation
Group Mobility scenario.

After the generation of scenario1, the following command

should be typed to transform the scenario1 into a file that can
TABLE 2: RPGM Parameters in Bonn Motion be read by ns-2. Fig. 5 shows the creation of clusters as shown
in the NAM console at completion of the simulation.
Parameters Explanation
-n Number of mobile nodes ./bm NSFile –f scenario1
-d Simulation duration time
-x Simulation area width
VI. SIMULATION RESULTS
-y Simulation area height
Four different scenarios were considered with different
-c Group change Probability pause times. Each scenario contained five sub scenarios along
with four different scenarios for different node speeds
-l Lowest velocity containing four sub scenarios each. The simulation for five
times for each scenario and the average of the results were
-h Highest velocity computed. The packets delivery ratio, average end to end
delay, routing overhead, normalized routing load and packet
-p Pause time
drop were used as metrics in evaluating performance.
-a Average number of nodes
per group
-r Maximum distance to group

Fig. 5. Creation of Clusters with the RPGM model

Page | 218
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

A. Packet Delivery Ratio

Packet Delivery Ratio (PDR) is defined as the fraction of
the number packets received at the destination to the number
of packets originated by the source application. PDR
describes the loss rate seen by the transport layer protocols.
This will affect the overall network throughput.
The PDR as a percentage is given by Formula (3).

(3)

Fig. 6. (A) shows the PDR for different pause times. The
investigation was carried at pause times 5, 10, 20, 30 and 40 s Fig. 6 (B). PDF vs. Node speed
and the PDR at these pause times were measured in order to
select the best pause time. From Fig. 6 (A), it can be seen that B. Average End-to-End Delay
the DSR protocol performs well when the pause time is 20 s
under all the four scenarios. The results show that the highest Average end-to-end delay is defined as the average time
PDR with the maximum number of packets sent and received taken by a packet to reach the destination from the source. The
at the pause time 20 s, is 100.00%, 100.00%, 99.35% and average end-to-end delay includes all the types of delays such
99.96%. After selecting the pause time to be fixed at 20 s, as transmission delay, propagation delay, processing delay and
different maximum speed of 20, 40, 60 and 80 m/s were interface queuing delay. The average end-to-end delay
investigated in order to select the optimal node speed in the between two nodes is computed using Formula (4).
RPGM model.
(4)

The delay is affected by higher CBR packet generation rate at

the source as well. When packets are generated at high rates,
the buffer at the source becomes full resulting in longer
queuing delays at the source node. Figures 8(A) and (B) show
the effect of maximum pause time and speed on average end-
to-end delay respectively. Fig. 7(A) shows that the average
end to end delay is reduced when the pause time is 20 s in the
2nd and 4th scenarios. The average end-to-end delays were
30.56 s and 15.82 s respectively. The average end to end
delays in the 1st and 3rd cases at 10 s pause time were 6.62 s
and 11.66 s respectively. The reason for this discrepancy is the
increased time consumption for route discovery at 20 s pause
time due to buffer overflow in the 1st and 3rd scenarios.

Fig. 6 (A). PDF vs. Pause time

Figure 6(B) shows the effect of node speed on the DSR

protocol. The results show that the highest PDR is achieved
when maximum speed is 20 m/s or 72 km/h. The PDR at 20
m/s were 100.00%, 99.96%, 99.97% and 99.88% respectively.
Whenever the node speed was increased, the PDR dropped
due to the reason that the group leader and the members were
moving very fast during the packet delivery process. This
causes packets to drop leading to reduced PDR. It was also
observed that when a mobile node moves fast and the pause
time is small; the topology is likely to be in a highly dynamic
condition. Hence, the optimal setting for node speed and the
pause time are 20 m/s and 20 s respectively.

Fig 7(A). Delay vs. Pause Time

Fig. 7(B) shows the effect of different maximum speeds on the

average of end-to-end delay. From Fig. 7(B) it can be seen that

Page | 219
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

the average end-to-end delay is decreased when maximum Fig. 8(B) shows that different maximum speeds affect the
speed is 20 m/s. An increase in maximum speed also results in routing overhead. It can be seen that the routing overhead is
an increase in the average end-to-end delay. This is due to the the minimum at the maximum speed is at 20 m/s in all
reason that when a node moves fast, it causes link failures scenarios. Whenever the maximum speed of a node is
leading to new route discovery processes. New route increased, the overhead is also increased. Hence, it can be
discovery processes would create additional delays increasing concluded that the average optimal settings for maximum
the overall end-to-end delay. Hence it can be concluded that pause time and speed are 20 s and 20 m/s for minimizing the
the optimum settings for the maximum pause time and the routing overhead.
speed are 20 s and 20 m/s respectively.

Fig. 7 (B). Delay vs. Node Speed

Fig. 8 (B). Overhead vs. Node speed

C. Routing Overhead
D. Normalized Routing Load (NRL)
Overhead is an important issue as higher overheads reduce
The Normalized Routing Load (NRL) is an estimate of
the overall network resources utilization. Hence it is important
how efficient a routing protocol is. The number of routing
to reduce the overhead in a network as much as possible. Fig.
packets sent per data packet is an indication of how well the
8 (A) and (B) show the routing overhead for different pause
protocol maintains the routing information updated. The
times. From Fig. 9(A), it can be seen that different pause times
higher the normalized routing load, the higher the overhead of
affect the routing overhead under all the four scenarios. Fig.
routing packets is and consequently the lower the efficiency of
8(A) shows that the overhead is the minimum when maximum
the protocol. Fig. 9(A) and (B) show the results for the NRL
pause time is 20 s, except for the 3rd scenario with total
against different maximum pause times and maximum node
number of nodes to be equal to 40 with 30 source connections
speed respectively. Fig. 9(A) shows that the DSR protocol
and the maximum pause time to be 10 s. The 3rd scenario has
performs best when maximum pause time is 20 s except for
the optimum setup at the maximum pause time equal to 10 s
the 3rd scenario, where the pause time is 10 s. This is due to
due to the lower number of packets transmitted. In addition,
the reason that the NRL is directly proportional to the
DSR protocol uses a route cache to reduce the number of route
overhead and the packets sent. As shown in Fig. 10(A), the
discovery processes during the establishment and transmission
NRL is reduced when the maximum pause time is 10 s.
of packets.

Fig. 9 (A). NRL vs. Pause time

Fig. 8 (A). Overhead vs. Pause time

Page | 220
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

Fig. 9(B) shows the effect of the different maximum node REFERENSES
speeds on the NRL. Fig. 9(B) shows that the DSR protocol [1] J. Jain, M. Fatima, and R. Gupta, "Overview Challenges of Routing
performs best when maximum node speed is 20 m/s resulting Protocol and Mac Layer in Mobile Ad-hoc Network," Journal of
in the lowest NRL value. Whenever the maximum node speed Theoretical and Applied Information Technology vol. 8, pp. 6-12,
is increased, the NRL is also increased leading to lower 2009.
efficiency. Hence, it can be concluded that the optimal setting [2] C. E. Perkins and E. M. Royer, "Ad-Hoc on-Demand Distance Vector
for the maximum pause time and the maximum node speed are Routing," in The Second IEEE Workshop on Mobile Computing
Systems and Applications, 1999, pp. 90–100.
20 s and 20 m/s respectively for efficient NRLs.
[3] M. S. Corson and A. Ephremides, "A Distributed Routing Algorithm
for Mobile Wireless Networks," Journal of Wireless Networks, vol. 1,
pp. 61–81, 1995.
[4] C. E. Perkins and P. Bhagwat, "Highly Dynamic Destination-
Sequenced Distance-Vector Routing (DSDV) for Mobile Computers,"
in SIGCOMM ’94 Conference on Communications Architectures,
Protocols and Applications, 1994, pp. 234–244.
[5] J. Broch, D. A. Maltz, D. B. Johnson, Y. Hu, and J. G. Jetcheva, "A
Performance Comparison of Multi-Hop Wireless Ad Hoc Network
Routing Protocols," in Fourth Annual ACM/IEEE International
Conference on Mobile Computing and Networking 1998, pp. 85–97.
[6] P. Johansson, T. Larsson, N. Hedman, B. Mielczarek, and M.
Degermark, "Scenario-based Performance Analysis of Routing
Protocols for Mobile Ad-hoc Networks," Fifth Annual ACM/IEEE
International Conference on Mobile Computing and Networking, pp.
195–206, 1999.
[7] D. A. Maltz, J. Broch, J. Jetcheva, and D. B. Johnson, "The Effects of
On-Demand Behaviour in Routing Protocols for Multi-Hop Wireless
Fig. 9( B). NRL vs. Node speed Ad Hoc Networks," IEEE Journal on Selected Areas in
Communications, Special Issue on “Wireless Ad Hoc Networks, vol.
1439–1453, pp. 1439–1453, 1999.
VII. CONCLUSIONS
[8] S. R. Das, C. E. Perkins, and E. M.Royer, "Performance Comparison of
In this paper, the authors have presented the work of Two On-demand Routing Protocols for Ad Hoc Networks," in
evaluating the performance of the DSR protocol under Nineteenth Annual Joint Conference of the IEEE Computer and
Reference Point Group Mobility model with respect to the Communications Societies 2000, pp. 3–12.
effects of the maximum pause time and maximum speed under [9] P. Prabhakaran and R. Sankar, "Impact of Realistic Mobility Models
different scenarios. The simulation parameters included 36 on Wireless Networks Performance," in Wireless and Mobile
different scenarios of which 20 scenarios are for different Computing, Networking and Communications, 2006, pp. 329 - 334
maximum pause times and 16 scenarios are for different [10] D. B. Johnson, "Routing in Ad Hoc Networks of Mobile Hosts," in
maximum node speeds. Simulation results show that the IEEE Workshop on Mobile Computing Systems and Applications,
maximum pause time and speed have a direct impact on the 1994, pp. 158–163.
performance of the DSR protocol. The Reference Point Group [11] D. B. Johnson and D. A. Maltz, "Dynamic Source Routing in Ad Hoc
Mobility model has two constraints. One is that the movement Wireless Networks," in Mobile Computing, ed, 1996, pp. 153–181.
of nodes should follow that of the group leader and the other [12] D. B. Johnson, D. A. Maltz, and Y. Hu. (2004). Dynamic Source
one is that each group leader has the group movement limited Routing Protocol for Mobile Ad-hoc Networks (DSR). Available:
to certain speed limits. The simulation results show that when http://www.ietf.org/internet-drafts/draft-ietf-manet-dsr-10.txt
the mobility is high, the possibility of route caches becoming [13] D. B. Johnson, D. A. Maltz, Y.-C. Hu, and J. G. Jetcheva. (2001). The
stale is also high resulting in frequent link failures. When a Dynamic Source Routing Protocol for Mobile Ad Hoc Networks, .
route discovery is initiated, the large number of replies Available: Internet- Draft, draft-ietf-manet-dsr-05.tx
received in response is associated with high MAC overhead [14] D. B. Johnson, D. A. Maltz, and J. Broch, "The Dynamic Source
causing increased interference with data traffic. Hence, the Routing Protocol for Multi-hop Wireless Ad Hoc Networks," in Ad
cache staleness and high MAC overhead together result in Hoc Networking, ed, 2001, pp. 139–172.
significant degradation of performance in DSR in high [15] X. Hong, M. Gerla, G. Pei, and C. C. Chiang, "A Group Mobility
mobility scenarios. Furthermore, simulation results show that Model for Ad Hoc Wireless Networks," in 2nd ACM International
there are four main factors that can affect the DSR protocol. Workshop on Modelling, Analysis and Simulation of Wireless and
Mobile Systems, 1999, pp. 53–60.
They are namely, the maximum node speed, the maximum
pause time, the number of connections, and the number of [16] F. Bai, N. Sadagopan, and A. Helmy, "Important: a Framework to
nodes. From the simulation results, it was observed that the Systematically Analyze the Impact of Mobility on Performance of
Routing Protocols for Ad Hoc Networks," in 22nd Annual Joint
optimal settings for the maximum pause time and the Conference on the IEEE Computer and Communications Societies
maximum node speed are 20 s and 20 m/s respectively. USA, 2003, pp. 825–835.
[17] G. Jayakumar and G. Gopinath, "Performance Comparison of Two On-
demand Routing Protocols for Ad-hoc Networks based on Random

Page | 221
ICIT 2015 The 7th International Conference on Information Technology
doi:10.15849/icit.2015.0032 © ICIT 2015 (http://icit.zuj.edu.jo/ICIT15)

Waypoint Mobility Model," Journal of Applied Sciences American, [20] Bonn Motion Mobility Generator
vol. 6, pp. 659-664, 2008. www.informatik.unibonnde/IV/BonnMotion.
[18] G. Kioumourtzis, "Simulation and Evaluation of Routing Protocols for
Mobile Ad Hoc Networks," MSc. Thesis, Naval Postgraduate School,
California, 2005.
[19] C. P. Agrawal, M. K Tiwari, and O. P. Vyas, “Evaluation of AODV
Protocol for Varying Mobility Models of MANET for Ubiquitous
Computing,” in Proceeding of the Third International Conference on
Convergence and Hybrid Information Technology, 2005.

Page | 222

View publication stats