Professional Documents
Culture Documents
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
ABSTRACT
Simulation of mobile networks requires reliable
movement
generation.
Evaluation
of
the
performance of the existing random movement
generator used in ns2 shows a bias for placing the
mobile nodes in the center of the experimental area.
Applying the quadrat count and the VMR tests
confirms the tendency for clustering of nodes. To
remedy this problem we propose, and implement a
different method for random movement generation
for use with the ns2 simulator, and show that our
movement generator improves the randomness of the
node distribution during the simulation.
Keywords
Movement generator; network simulation; setdest
utility; MANET; VMR; Quadrats Count.
INTRODUCTION
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
performance, the errors in network operation
can be detected and corrected before the
network is put to use. There are three
approaches that can be followed to investigate
the network performance: empirical study,
analytical modeling, and simulation.
In an empirical study, the actual network is
used under a controlled environment. Test users
that behave like real users are needed.
Performance is measured by observing the
actual network in use. The main inconvenience
of this approach is that the entire network or a
significant portion thereof must be built, which
can be both difficult and expensive. In an
empirical study, the actual network is used
under a controlled environment. Test users that
behave like real users are needed. Performance
is measured by observing the actual network in
use. The main inconvenience of this approach is
that the entire network or a significant portion
thereof must be built, which can be both
difficult and expensive.
Analytical modeling is a tool that uses
analytical techniques to predict network
performance. Using mathematical formulae to
calculate interactions that might happen
between different network entities allows us to
predict network performance. Analytical models
help to detect errors that might occur while the
network is operational so that they can be solved
before the actual network is constructed.
Analytical modeling should be done using onedimensional and two-dimensional tests with
both static and dynamic situations for topology.
Simulation is testing using a computer
program to track the behaviour of a network.
Simulation includes protocol simulation,
movement simulation, and traffic generators.
Network simulation helps determine how a real
network would function and helps predict the
networks behaviour. Several programming
languages and simulators are used for network
simulation, such as Java, C, or ns2; in this
report, however, we focus on ns2 and Java only.
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
Walk, Random Waypoint and Random
Direction. Each mobile node in Random Walk
has a randomly generated starting position. The
nodes travel from their starting position to a
randomly generated new location by generating
random direction and velocity. The velocity has
to be bounded by minspeed and maxspeed
values and the direction also has to be in range
between 0 and 2 [8]. A node changes direction
and speed either at the end of a time interval t or
if it traveled a distance d. In Random Waypoint,
the movement is not constant whereas pause
times are introduced. The nodes start at
randomly chosen positions, then "pause" for
some time and then start moving in a uniformly
distributed velocity between minspeed and
maxspeed towards a chosen destination. The
nodes in this model have to "pause" for some
time before they change direction or speed [8].
The drawback in Random Waypoint is
clustering near the center (i.e. having the nodes
near each other near the center of the
experimental area). Random Direction Mobility
model was introduced to overcome this
drawback in the Random Waypoint model. In
Random Direction model, a node travels at a
chosen velocity in a chosen direction until it
reaches the boundaries of the area rather that
until it reaches a randomly chosen location.
Once a node hits the boundaries it pauses for a
time t and chooses a new direction and starts
moving in this new direction again, and so on
[6].
ns2 Setdest Utility
Setdest is a tool used to generate nodes
movements for the mobile nodes in the network
simulation ns2 by positioning network nodes in
a bounded area and setting the movement in a
random direction [9, 10]. Setdest tool uses the
random waypoint mobility model algorithm to
create the random movements for the mobile
nodes [11, 9, 20, 10]. There are two versions of
setdest tool can be used in the ns2 known as v
1 and v 2 where version 2 is the latest[21].
Setdest version v1 is using the parameters,
number of nodes, pause time, maximum speed
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
and counted the total number of bytes delivered
at the destination node. The higher the delivery
ratio for transmission along the edges, the better
the movement generator.
5 CUSTOM JAVA MOVEMENT
GENERATOR
The setdest utility uses the random waypoint
mobility model algorithm. The waypoint
algorithm is known to have the border effect,
which creates a kind of clustering at the center
of the simulation area [6]. The border effect
problem can be avoided by following a different
algorithm called the boundless simulation area
mobility model. The main idea of the boundless
model is to allow going over the edges thus
avoiding the influence of the simulation areas
edges. Moreover, going over the edges in the
boundless simulation means that once the
mobile node reaches the boundary from any side
of the simulation area, it does not bounce the
same as in the other models, but it disappears
from the side and appears from the other side
continuing with the same direction, which
makes the simulation area look more like a tube
than a plane. Figure 1 shows the simulation area
that looks like this.
356
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
the previous
equations:
coordinates
following
the
357
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
___________________________________________________________________________________
Algorithm 1 Pseudocode (Custom Generator)
__________________________________________________________________________________________
(maxX, maxY) = simulation area dimensions
maxTime = max duration of nodes movement
maxMovement = max distance of nodes movement
currentTime = 0
(currentX, currentY) = random position
while Simulating do
angle = rand(0..1) * 360_
destX = rand(0..1) * cos(angle) * maxMovement //set the movement destination
destY = rand(0..1) * sin(angle) * maxMovement
MovementTime = rand(0..1) * maxTime
speed = geometric distance (currentX, currentY) to (destX, destY)/movementTime
if path does not cross the boundary then
writeMovement ($ns at current Time $nodeX setdest destX, destY, speed)
else
if path crosses the boundary then
(interceptX, interceptY) = boundary intercept
distToBound = geometric distance to the boundary in the direction of the movement path
writeMovement ($ns at currentTime $nodeX setdest interceptX, interceptY, speed)
timeAtBoundary = currentTime + distToBound/speed
writeMovement ($ns at timeAtBoundary $nodeX setpos interceptX, interceptY)
destX = destX % maxX
destY = destX % maxY
writeMovement ($ns at timeAtBoundary $nodeX setdest destX, destY, speed)
(currentX, currentY) = (destX, destY)
currentTime = currentTime + movementTime
end if
end if
end while
___________________________________________________________________________________
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
Movement
type
either
Regular
Movement (M) or Jump (J)
Node Number
Time
Current xCoordinate
Current yCoordinate
Next xCoordinate
Next yCoordinate
6 RESULTS
In this paper, we have investigated the
randomness performance and the packets
transmission of both generators the setdest
utility and the custom generator. The
randomness performance of both generators is
shown in the Figure 3.
359
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
Table 1. The confidence interval result of the
delivery ratio through the center for both
generators
Setdest Utility
Custom
Generator
60 Nodes
9459.907005
TO
9731.892885
9371.082322
TO
9610.517678
30 Nodes
6724.788306
TO
8182.211694
5150.647877
TO
6050.352123
20 Nodes
2945.226379
TO
4811.973621
1325.714676
TO
2448.285324
Setdest Utility
Custom
Generator
60 Nodes
3664.285794
TO
4802.914206
7933.676792
TO
8238.223208
30 Nodes
478.0674426
TO
955.5325574
1712.377421
TO
2351.922579
20 Nodes
193.966817
TO
576.433183
197.3408355
TO
468.9591645
7 CONCLUSION
We investigated the performance of the
popular setdest utility used in the ns2 network
simulator. The movement generated with setdest
utility tends to cluster the mobile nodes in the
center of the experimental area. This has an
effect on the VMR coefficient used to measure
randomness of the positions of points in the
simulation. We proposed and demonstrated the
advantage of new random motion generator for
use in ns2 simulator. Testing the new generator
shows a marked advantage over the standard
setdest utility.
ACKNOWLEDGEMENT
This work is sponsored and funded by the
Ministry of Higher Education of Saudi Arabia
through the Saudi Arabian Cultural Bureau in
Canada.
8 REFERENCES
1.
2.
3.
360
International Journal of Digital Information and Wireless Communications (IJDIWC) 2(4): 353-361
The Society of Digital Information and Wireless Communications, 2012 (ISSN: 2225-658X)
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Personal
Communications
Record,
1997.
Conference Record. 1997 IEEE 6th International
Conference on 12-16 Oct 1997, vol.2, pp. 562-566.
Wikipedia, the free encyclopedia (November 2010),
Wireless Ad-hoc Network,
http://en.wikipedia.org/wiki/Wireless_adhoc_network (accessed Nov 2011)
N. S. Dalton and R. C. Dalton, The theory of
natural movement and its application to the
simulation of mobile ad hoc networks (MANET),
in Proc. the Fifth Annual Conference on
Communication Networks and Services Research,
2007, pp. 359363.
Y. Zhang and W. Li, An integrated environment
for testing mobile ad-hoc networks, in Proc.the
Third ACM International Symposium on Mobile Ad
Hoc Networking and Computing (MobiHoc '02),
Lausanne, Switzerland, June 2002, pp. 104-111.
V. Davies, Evaluating Mobility Models Within an
Ad Hoc Network, Masters thesis, Colorado School
of Mines, Colorado, 2000.
T. Camp, J. Boleng, and V. Davies, A survey of
mobility models for ad hoc network research,
Wireless Communications & Mobile Computing,
Special Issue on Mobile Ad Hoc Networking:
Research,Trends and Applications, vol. 2, no. 5, pp.
483502, 2002.
B. J. Culpepper and H. C. Tseng, Sinkhole
intrusion indicators in DSR MANETs, in Proc.
First International Conference on Broad band
Networks, 2004, pp. 681 688.
N.
Frangiadakis,
M.
Kyriakakos,
S.
Hadjiefthymiades, and L. Merakos, "Realistic
mobility pattern generator: design and application in
path prediction algorithm evaluation," in Proc.
Personal,
Indoor
and
Mobile
Radio
Communications,
2002.
The
13th
IEEE
International Symposium on 15-18 Sept. 2002,
vol.2, pp. 765- 769.
P. Caballero-Gil, C. Caballero-Gil, J. Molina-Gil,
and C. Hernandez-Goya, A simulation study of
new security schemes in mobile Ad-hoc
NETworks, in EUROCAST 2007. LNCS, vol. 4739,
R. Moreno Daz, F. Pichler, A. Quesada Arencibia,
Ed., Heidelberg: Springer, 2007, pp. 7381.
J. Hu and R. Marculescu, DyAD: smart routing for
networks-on-chip, in Proc. the 41st annual
conference on Design automation, San Diego, CA,
USA, June 07-11, 2004, pp. 260-263.
A. L. Cavilla, MANET extensions to ns2.
Available:
http://www.cs.toronto.edu/andreslc/software/MAN
ET extensions.tgz.
B. Carbone, Routing Protocols for Interconnecting
Cellular and Ad Hoc Networks, Masters thesis,
15.
16.
17.
18.
19.
20.
21.
361