Académique Documents
Professionnel Documents
Culture Documents
I.
I NTRODUCTION
II.
978-1-5090-2494-0
R ELATED W ORK
Fig. 1.
III.
978-1-5090-2494-0
P ROPOSED A LGORITHM
P tLevel
P tV alue
P tRange
High
Medium
Low
0.28183815
0.24169726
0.20191908
250 m
200 m
170 m
Fig. 2.
A. Control Packets
To determine whether the node in the network is a
Gateway or not, we need to exchange packets and for this
we modify the existing control packets in AODV protocol
[19] to broadcast the P tLevel (High/M edium/Low) and
Status(Gateway/N onGateway) instead of adding new control packets.
HELLO packet: All nodes in the proposed protocol make use
of the HELLO packet in AODV to exchange the information
of their transmission-power level and status among each other.
In order to accumulate such information, a new fields called
Transmission-Power Level (P tLevel ) and Status are added
to the HELLO packet. With the help of the HELLO packet,
every node creates a neighbour table which store the P tLevel
and Status values of all 1-hop neighbours.
B. Route Discovery
The route discovery process has operations on all nodes
which include source node, intermediate nodes and destination node. The route discovery process of proposed scheme
operates as follows:
Source Node: The source process is same as used in AODV
that whenever a source needs a route to the destination. When a
source node is going to transfer packets with a destination node
978-1-5090-2494-0
Fig. 3.
Fig. 5.
Fig. 6. Effect of the proposed sending and forwarding RREP process on the
number of data packets drop
IV.
Fig. 4.
Send RREP:
Forward RREP:
978-1-5090-2494-0
TABLE II.
SIMULATION PARAMETERS
Mac Layer
Bandwidth
Propagation model
Antenna model
Simulation area
Simulation time
Number of nodes
Node speed
Pause time
Traffic type
Number of connections
IEEE 802.11
2 Mb/s
Two-ray ground
Omni-Antenna
1000 m x 1000 m
100 (sec)
75, 100, 150, 200
10, 20, 30 (m/s)
40 (sec)
CBR
40
40
35
P ERFORMANCE R ESULTS
25
20
15
10
PF-AODV
DP-AODV
Proposed Algorithm
To analyse the performance of our scheme, we compare our Proposed Algorithm with Pure-Flooding (PF-AODV)
which use blind flooding to disseminate RREQ packets, and
Dynamic-Power (DP-AODV) which use Transmission-Power
control mechanism to reduce the signal collision. We study
the effect of node density on the performance of each protocol.
Our simulation results are obtained from 5 different movement
scenarios and 5 different traffic scenarios which means that
each metric value is the mean of the 25 runs. Generally, 95%
of the confidence interval was small compared with the value
that was being reported.
5
50
100
150
200
Number of Nodes
Fig. 8.
300000
PF-AODV
DP-AODV
Proposed Algorithm
250000
30
200000
150000
220
200
100000
180
Throughput (bps)
50000
0
50
100
150
200
Number of Nodes
160
140
120
100
Fig. 7.
Routing Overhead
80
PF-AODV
DP-AODV
Proposed Algorithm
60
978-1-5090-2494-0
40
50
100
150
Number of Nodes
Fig. 9.
Throughput
200
3.6
3.4
End-End-Delay (second)
3.2
VI.
2.6
PF-AODV
DP-AODV
Proposed Algorithm
2.4
2.2
1.8
1.6
1.4
50
100
150
200
Number of Nodes
Fig. 10.
End-to-End Delay
PF-AODV
DP-AODV
Proposed Algorithm
2.5
In our future work, more optimization and further improvement for reducing congestion and energy consumption will be
planned. For instance, once gateway node set is determined, the
problem of how to schedule those gateway nodes to efficiently
forward RREQ packets requires further study.
2.4
2.3
2.2
2.1
R EFERENCES
2
1.9
1.8
50
100
150
200
Number of Nodes
Fig. 11.
2.8
Energy Consumption
978-1-5090-2494-0
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
978-1-5090-2494-0