22-05-2012, 11:43 AM
Performance Analysis of AODV and DSR Routing Protocols for IEEE 802.15.4
[attachment=22671]
INTRODUCTION
WSNs are a new technology that appeared in the nineties
and were considered as the successor of Ad hoc networks.
They are wild spread thanks to their easy deployment and their
low cost. In the past, getting data transferred from a unit to another
necessitates expensive wiring. Nowadays, we can place
nodes wherever we want to, having no wiring constraints,
thanks to the tiny batteries feeding the nodes. Additionally,
the sensors’ small dimensions give them the possibility to
be used any where even discreetly. These advantages allowed
WSNs to be deployed in all fields: agriculture, medicine, home
automation, military, etc.
Reactive Protocols
Reactive protocols create routes only when needed. As a
result, they help consume less energy comparing to proactive
protocols. In this work, we studied AODV and DSR.
1) AODV : AODV, as illustrated in [6], consists on holding
a routing table in each node. Whenever a route is needed, the
source emits route discovery requests (RREQ) in the whole
network. Once reached, the destination sends a route reply
(RREP) towards the source. Thus, each node has a clear
view on the network. This ensures a quick update when the
network’s topology changes. As a result, the packets loss is
reduced, the communication is fastened and the network’s
throughput is improved [7][8].
Simulation Parameters
Many parameters can be used to study the network performances.
1) Packet Loss and Packet Delivery Ratio : Packet loss
is the percentage of packets that couldn’t make it to the
destination. As for the packet delivery ratio, it indicates the
percentage of the transmitted data packets that are successfully
received. They are two important metrics which can be used
as indicators to a congested network. The number of packets
dropped does not take into account retransmissions. If the
packet is successfully received by the destination after several
retransmissions, the drops are not considered. This would
effectively make the number of transmitted packets equal to
the sum of the number of received packets and the number of
dropped packets.
Energy Consumption
To study the energy consumption, we have considered the
topology containing 201 nodes with three different traffics:
node1!node41, node9!node11 and node14!node38, each
one is repeated six times. Our aim from repeating the same
traffics is being as close as possible from having the same
routes. But, they are not totally the same. The nodes are tested
under the same conditions.
CONCLUSION
We have evaluated the WSNs performances through altering
nodes number and traffic density. We discovered that AODV,
mostly, performs better than DSR. In fact, AODV maintains
a routing table in every node. Therefore, packets should be
able to find their way quickly and updates should be obtained
fast. As a result, the packet loss will be decreased and the
throughput will be increased. For a small network, a censor is
very active since the possible routes are limited. That’s why,
with such topology AODV doesn’t perform as good as for
a medium topology. The higher traffics we have, the most
updates we get and this should fasten the communication.
As AODV makes more updates than DSR, it consumes more
energy and this is a real handicap for this protocol. As future
work, we think about of creating a new routing protocol
which guarantees AODV’s performances and DSR’s energy
consumption.
[attachment=22671]
INTRODUCTION
WSNs are a new technology that appeared in the nineties
and were considered as the successor of Ad hoc networks.
They are wild spread thanks to their easy deployment and their
low cost. In the past, getting data transferred from a unit to another
necessitates expensive wiring. Nowadays, we can place
nodes wherever we want to, having no wiring constraints,
thanks to the tiny batteries feeding the nodes. Additionally,
the sensors’ small dimensions give them the possibility to
be used any where even discreetly. These advantages allowed
WSNs to be deployed in all fields: agriculture, medicine, home
automation, military, etc.
Reactive Protocols
Reactive protocols create routes only when needed. As a
result, they help consume less energy comparing to proactive
protocols. In this work, we studied AODV and DSR.
1) AODV : AODV, as illustrated in [6], consists on holding
a routing table in each node. Whenever a route is needed, the
source emits route discovery requests (RREQ) in the whole
network. Once reached, the destination sends a route reply
(RREP) towards the source. Thus, each node has a clear
view on the network. This ensures a quick update when the
network’s topology changes. As a result, the packets loss is
reduced, the communication is fastened and the network’s
throughput is improved [7][8].
Simulation Parameters
Many parameters can be used to study the network performances.
1) Packet Loss and Packet Delivery Ratio : Packet loss
is the percentage of packets that couldn’t make it to the
destination. As for the packet delivery ratio, it indicates the
percentage of the transmitted data packets that are successfully
received. They are two important metrics which can be used
as indicators to a congested network. The number of packets
dropped does not take into account retransmissions. If the
packet is successfully received by the destination after several
retransmissions, the drops are not considered. This would
effectively make the number of transmitted packets equal to
the sum of the number of received packets and the number of
dropped packets.
Energy Consumption
To study the energy consumption, we have considered the
topology containing 201 nodes with three different traffics:
node1!node41, node9!node11 and node14!node38, each
one is repeated six times. Our aim from repeating the same
traffics is being as close as possible from having the same
routes. But, they are not totally the same. The nodes are tested
under the same conditions.
CONCLUSION
We have evaluated the WSNs performances through altering
nodes number and traffic density. We discovered that AODV,
mostly, performs better than DSR. In fact, AODV maintains
a routing table in every node. Therefore, packets should be
able to find their way quickly and updates should be obtained
fast. As a result, the packet loss will be decreased and the
throughput will be increased. For a small network, a censor is
very active since the possible routes are limited. That’s why,
with such topology AODV doesn’t perform as good as for
a medium topology. The higher traffics we have, the most
updates we get and this should fasten the communication.
As AODV makes more updates than DSR, it consumes more
energy and this is a real handicap for this protocol. As future
work, we think about of creating a new routing protocol
which guarantees AODV’s performances and DSR’s energy
consumption.