Sei sulla pagina 1di 5

Study of routing protocols in mobile ad-hoc networks

regarding rescue scenarios


Dara Khajavi

James Foot

Jean-Marie Mellet

Grad. School of Science Engineering


and Technology
Istanbul Technical University
Istanbul, Turkey
khajavi@itu.edu.tr

Grad. School of Science Engineering


and Technology
Istanbul Technical University
Istanbul, Turkey
foot@itu.edu.tr

Grad. School of Science Engineering


and Technology
Istanbul Technical University
Istanbul, Turkey
mellet@itu.edu.tr

these devices could create a network between themselves. This


type of network is called a MANET.
A Mobile Ad Hoc Network, also known as MANET, is a
network that consists of an autonomous system of mobile hosts
(that also serve as routers) interconnected with each other
through wireless links. A MANET would solve all the
limitations that were mentioned above. Because with a
MANET all the hosts are connected with each other, the person
you want to send information to doesn't necessarily need to be
in the proximity. So that solves the problem of range. A Mobile
Ad Hoc Network work with IP packages, allowing the users to
transmit and receive any type of data, and it also allows users
to receive and transmit data at the same time. So MANET
solves the main limitations of two-way radios, but it also
introduces some limitations of its own, such as, limited
bandwidth, battery life and processing power of mobile
devices.

AbstractThe mobile devices of today present a great


opportunity for building large-scale mobile sensing and
information sharing systems that could be used for emergency
situations and rescue scenarios. Usually in large catastrophes,
one of the first things that gets effected is the mobile cellular
network, either by the loss or damage of its infrastructure, or by
the high levels of congestion in the cellular network. In this case
the use of the sensory and communication capabilities of a
smartphones, tablet or any other mobile device that requires a
connection to the cellular network would be limited or even
pointless. Using mobile ad-hoc networks (MANET), we can get
around this problem, by using smartphones and all the other
mobile devices to create a dynamic temporary network that can
share information between them selves without the need for any
additional infrastructure.
Index Terms Mobile Ad hoc Networks (MANET), routing
protocols, Optimized Link State Routing Protocol (OLSR), Ad
hoc On-demand Distance Vector (AODV), OMNET++.

I. INTRODUCTION
When a city or a country get hit by a major disaster, more
often than not, the communication infrastructure get damaged,
completely destroyed or its just too congested to be used as a
reliable communication network. At the moment, emergency
responders use two-way radios, although this is a reliable
communication system, it has got some limitations. The first
one is range, an average two-way radios has a range of 48
meters, which isn't bad if the person you want to talk to is in
the neighborhood, but if you want to communicate across a
city, it can be a bit more problematic. The second limitation is
that a two-way radio can only be used to transmit and receive
voice, that isnt bad but it doesnt take advantage of all the new
technologies we have at our disposal, such as the global
positioning system (GPS) and Unmanned aerial vehicles
(UAV) or any other kind of useful robotics. And the last
limitation is traffic, two-way radios can't use simultaneously
more than one channel, and only one person can communicate
on that channel at a time. A possible solution is to continue on
using the smart devices we use every day, like smartphones,
tablets, but instead of connecting them to the carrier network,

Fig. 1. Representation of a MANET in an emergency scenario.

In a MANET we deal with hosts aren't always in the same


geographical position, meaning that when a host is moving its
neighbors will change and because of that new routing routes
become available or the host my move so far, that it can
establish a link to another user in a network. This is the
challenge, how do you make sure a host is still in the network,
and how do you check which is the shortest path from one
mobile host to another. In this paper were going to focus on

two of the most common MANET routing protocols, Ad hoc


On-demand Distance Vector also known as AODV and
Optimized Link State Routing protocol or OLSR, and test them
out to see how they would perform in a MANET design for
emergency scenarios.
To test these routing protocols, were going to run a several
test scenarios on the OMNET++ simulator.
OMNET++ is a C++ based simulator library and
framework, used to run network simulations.

network topology. This can causes a constant overhead of


routing traffic, but no initial delay, solving the main problem of
AODV. To avoid massive overhead traffic, OLSR uses a
technique called Multipoint Relay (MPR). Instead of all the
nodes broadcasting there LS information, 2-hop neighborhoods
are formed, each one with one MPR. MPR is a node
responsible for transmitting the LS information about the nodes
in its neighborhood to the rest of the network, and also to
transmit the LS information of all the other nodes in the ad hoc
network to the nodes in its neighborhood.

II. OVERVIEW OF AODV PROTOCOL

A. OLSR does this by sending three types of messages:

The Ad hoc On-demand Distance Vector is a routing


protocol developed for MANETs, it operates like a reactive
protocol, meaning that the information about the topology of
the network only get transmitted by the nodes when needed
(On-demand). If a node needs to transmit data to another node,
to which theres no established route, it generates a route
request (RREQ) message, and broadcasts it throw the network.
When the RREQ message reaches the destination node or
another node that know the route to the destination node, the
route is establish and the node can begin transmitting data. For
as long as there exists a route between the nodes, AODV
remains passive. When the route becomes unavailable, AODV
will broadcast another RREQ message.

HELLO this message is sent to all the neighbors, for


neighborhood sensing and MPR calculations.
TC topology control, is a message sent by the nodes
with the link state information.
MID Multiple Interface Declaration is a message
sent by node running OLSR on more than one
interface.

Fig. 3. The first figure represents the broadcast by the nodes of their
link state information (without MPRs). The second figure
represents the broadcast of the link state information using MPRs
(nodes in black).

The draw back of OLSR, because its a proactive routing


protocol, its always sharing link state information, using up
power and network resources. This is okay for a fixed network,
but for a mobile network with devices that want to use their
processor as little as possible to conserve battery, OLSR could
be.

Fig. 2. Application of the AODV routing protocol in finding a route


from node A to node H.

A. Basic AODV control messages:


RREQ A route request is a message broadcasted by a
node requesting a route to another node.
RREP A route response is a message sent back to the
originator if the node is the desired destination node or
has a route to the destination node.
RERR Is a message transmitted to all the other
nodes, when a breakage in the link is detected.
The downside of AODV is that, when a node wants to
communicate with another, it first has to find the route and
only then, can it transmit traffic. This procedure causes an
initial delay.

IV. TEST SCENARIO

The goal of this paper is to see how the MANET routing


protocols would perform in an emergency scenario. As
mention in the introduction, the primary purpose of the
MANET is for communication, meaning transmitting and
receiving real time traffic, such as VoIP. This real time traffic
consists of UDP packages. Although there could also be other
type of data, like GPS coordinates, text messages, all using
TCP packages. Our intention is to run both routing protocols
throw 3 test scenarios in the simulator, and afterwards analyze
and compare the performance of the two routing protocols.

III. OVERVIEW OF OLSR PROTOCOL

A. Variables
The goal of these simulations is to understand and analyze
the difference between the two major ad-hoc protocols OLSR

The Optimized Link State Routing protocol is a tabledriven, link state (LS) proactive approach to MANET routing,
as all the nodes always have an updated knowledge of the

and AODV. We will perform simulations with the following


variables:
Node density: Our first variable will be the node
density. We can either change the number of nodes in a
fixed area, or change the size of the area with a fixed
number of nodes. As the range of the 802.11b is
around 300m (in theory), we will calculate a middle
density for filling the area with nodes at 300m from
each other.
Traffic: There are several ways to tune the traffic in
the network. First, we will define some node sources,
which will send packets to the other nodes. The more
sources we have, the more heavy the traffic will be.
The other way to do it, is the keep the same number of
sources but increasing the packet sending ratio.
Traffic type: As our case study is emergency
situation, some traffic type wouldnt be used, or at
least are not the priority. We can imagine that well
need real time traffic so that people in the damaged
area can inform the others.
Mobility: Using the mass mobility model in
OMNET++, we can change the parameters of the
mobility. For instance, the speed of the users, the angle
of the new direction from the previous one, and the
time for a new decision to be taken can be tuned.

These parameters are related to the mean fluctuation and


help us to determine which of the different protocols would be
the most appropriate at the time to emergency and rescue
scenario.
V. RESULTS AND DISCUSSION
As explained before, we are going to run to distinct test, the
first one, where we change the number of nodes in a network
located within an area of 1 km2. The node density changes as
follows. We start with a low density of 15 nodes, then we go
up to a medium density of 25 nodes, and finally to a high
density of 50 nodes in 1km2. Now, as discussed before, we
want to observe the throughput, End-to-End delay, jitter, the
amount of dropped packets and packet loss. As we can observe
in figures 5 to 8, the delay, jitter, packets dropped and loss rise
as the density of nodes increases, what is understandable,
because with more nodes the network has to manage more
traffic created by those nodes. When we examine the figure 5,
6 and 8, OLSR is performing better than AODV. But when we
analyze the throughput and the dropped packets AODV out
performs OLSR. OLSR has a lot of overhead because each
node is constantly sharing its link state information to the other
nodes, explaining the difference in between the throughput of
the network using AODV and OLSR. In figure 7, we can
observe that OLSR can cope as well as AODV, with the
increase in the number of nodes.

TABLE I. VARIABLES
Variable name:
Routing protocol

Value
AODV
OLSR

Node density

Low (15 nodes)


Medium (25 nodes)
High (50 nodes)

Area size

Packet Loss: This is the amount of packets dropped


by intermediate nodes due to effects produced by
mobility of these nodes, timer expires, unreachable or
erased destination.
Jitter: Is the time difference in packet inter-arrival
time to their destination, meaning if packets arrive at
their destination at a constant rate, then theres no
jitter. But if the theres a variation in the packet rate,
then you can notice some jitter.

Small (500 m2)


Medium (1 km2)
Large (4 km2)
Very Large (9 km2)

Traffic type

UDP

Mobility model

Random way point

B. Parameters
OMNET++ simulator is used to determine the different
routing protocols behavior using following indicators for
comparison:
Average Delay: This is very significant to measure for
our purpose because there is a need to send and receive
network management information as fast as possible in
emergency cases. We will measure this parameter in
different number of nodes in the network and different
number if connections in between due to an emergency
case network for different routing protocols.
Throughput: The ratio obtained by the number of
packets sent to the number of packets received.

Fig. 4. This graph shows the effect that AODV and OLSR have on the
throughput of a MANET network, when the number of nodes
changes.

Fig. 8. This graph shows the effect that AODV and OLSR have on the
loss of packages in a MANET network, when the number of
nodes changes.

The second test we ran consisted of maintaining a constant


number of nodes (50 nodes), but changing the size of the
network area. Using the same parameters as before, we ran the
network simulation for 4 different Area sized, first we tried it
on a small area (500 m2), then we moved up to a medium sized
area (1 km2) and from there on increasing it another two times
by 1 km2, achieving a large area of (4 km2) and (9 km2).
Analyzing the Figures 9 to 13, starting with the throughput,
we can observe that it behaves similarly to the throughput on
the first test, AODV outperforming OLSR. We can also
observe that as the distance in between the nodes increases,
which makes the number of neighbors decrease, because of
link failures, this explains the increasing loss rate observed in
figure 13.
OLSR tends to have a steadier trend in comparison to
AODV, but in our case (emergency scenarios) dropped packets
is the most important parameter, and out of the two routing
protocols, AODV is the better one for large networks.

Fig. 5. This graph shows the effect that AODV and OLSR have on the
End-to-End delay of packages in a MANET network, when the
number of nodes changes.

Fig. 6. This graph shows the effect that AODV and OLSR have on the
jitter in a MANET network, when the number of nodes changes.

Fig. 9. This graph shows the effect that AODV and OLSR have on the
throughput in a MANET network, where the size (area) of the
network changes.

Fig. 7. This graph shows the effect that AODV and OLSR have on the
number of packages that get dropped in a MANET network, when
the number of nodes changes.

Fig. 10. This graph shows the effect that AODV and OLSR have on
the end-to-end delay in a MANET network, where the size (area)
of the network changes.

Fig. 13. This graph shows the effect that AODV and OLSR have on
the loss in a MANET network, where the size (area) of the
network changes.

REFERENCES
[1] B. Brownlee, Y. Lee, Mobile Ad Hoc Networks An Evaluation
of Smartphone Technologies, Royal Military College of
Canada, 2011.
[2] I. Akyildiz, X. Wang, W. Wang, Wireless mesh networks: a
survey, 20th of December 2004.
[3] M. Ikeda, E. Kulla, L. Barolli, M. Takizawa, R. Miho,
Performance Evaluation of Wireless Mobile Ad-hoc Networks
via NS-3 Simulator,, International Conference on NetworkBased Information Systems, 2011.
[4] S. Baxla, Prof. R. Nema, Performance Analysis of AODV,
OLSR, DSR and GRP Routing protocols of Mobile Ad Hoc
Networks,, International Journal of Computer Science and
Mobile Computing, 2013.
[5] S. Faint, Impact of AODV and OLSR routing protocols on the
network cost of the Radio Environment Map,, Defence R&D
Canada, March 2011.
[6] Varga, A., OMNeT++ users manual: OMNeT++ version 4.0.
Available
at
http://www.omnetpp.org/doc/omnetpp40/manual/usman.html.
[7] INET Framework for OMNeT++/OMNEST. Available at
http://inet.omnetpp.org/doc/INET/neddoc/index.html.
[8] T. Clausen, P. Jacquet, Optimized Link State Routing Protocol
(OLSR), documentation, Project Hipercom, INRIA, The
Internet
Society
October
2003.
Available
at
http://www.ietf.org/rfc/rfc3626.txt.
[9] C. Perkins, E. Belding-Royer, S. Das, Ad hoc On-Demand
Distance Vector (AODV) Routing, documentation, The
Internet
Society,
July
2003.
Available
at
http://www.ietf.org/rfc/rfc3561.txt.
[10] Perkins, Charles E. and Royer, Elizabeth M. (1999), Ad hoc OnDemand Distance Vector Routing, In Proc IEEE WMCSA 1999,
pp. 90100.
[11] W. Kiess and M. Mauve, A survey on real-world implementations of mobile ad-hoc networks, Ad Hoc Networks, vol. 5,
no. 3, pp. 324339, 2007.
[12] L. Quispe, R. Tandazo, L. Galan, Analysis of Ad Hoc Routing
Protocols for emergency and rescue scenarios, Survey of
clustering algorithms for MANET, 2009.
[13] P. Sondi, D. Gantsou, S. Lecomte, Mobile Ad Hoc Networkbased Monitoring of Battlefields or Rescue Operations in Urban
Scenarios, UKSim Fourth European Modelling Symposium on
Computer Modelling and Simulation, 2009.

Fig. 11. This graph shows the effect that AODV and OLSR have on
the jitter in a MANET network, where the size (area) of the
network changes.

Fig. 12. This graph shows the effect that AODV and OLSR have on
the number of packages dropped in a MANET network, where the
size (area) of the network changes.

Potrebbero piacerti anche