You are on page 1of 6

2018 First International Conference on Secure Cyber Computing and Communication (ICSCCC)

Q-AODV: A Flood control Ad-Hoc on Demand


Distance Vector Routing Protocol
Bhagyalakshmi Amit Kumar Dogra
Ph.D. Scholar,Department of Computer Science & IT Ph.D. Scholar, School of Computer Science & Engineering
Central University of Jammu National Institute of Technology
Jammu, India Jalandhar, Punjab, India
bhagyalakshmi.magotra@gmail.com amitkumar62003@gmail.com

Abstract – Mobile ad hoc network (MANET) is a self- In Pure or blind flooding [5] the flooding message is
organized and self-configurable infrastructure less network transmitted again in the network by the receiving node. It
where the mobile nodes move arbitrarily. One of the major also assures that the flooded messages reaches every other
characteristic that differentiate mobile ad-hoc networks from node in the network provided the connections in the network
other types of networks is the ability of the mobile nodes to remain connected and also the network is collision free. But
receive and forward packets as a router. The focus of the work at the same time, huge amount of energy of the mobile
is to devise a strategy to control the flooding of control packets nodes is consumed and congestion is caused due to the
through the network in a way to improve the QoS parameters generation of redundant messages causing huge traffic. The
associated with MANETs. The proposed strategy tries to
congestion in the network increases the chances of collision
reduce the number of the intermediate nodes that participate
which further leads to retransmission of messages and also,
in the route discovery process thereby, reducing the total
number of control packets that are forwarded by the nodes in some nodes may fail to receive the messages. Sinha et al. [6]
the network. This is achieved by controlling the route request suggested that the number of nodes in the network that
(RREQ) broadcast storm using the node’s queue length. The receive a broadcast message is 80% low in moderately
source appends a random number with RREQ which is sparse graphs as compared to the dense networks. However
compared with the queue vacancy proportion at each in dense networks, this results into wastage of bandwidth
intermediate node. The intermediate node relays the RREQ and scarce node resources.
packet if the random number generated is less than the queue
The pure flooding, thus, suffers from the problem of
vacancy proportion. This reduces the number of congested
nodes forwarding the RREQ packets thereby improving QoS
contention, collision and redundant broadcast. Together,
parameters, preserving the energy and increasing the overall these are referred to as the broadcast storm problem [7]. To
network lifetime. The proposed algorithm Q-AODV is counter this problem many efficient flooding techniques
advancement over AODV that tries to find a less congested have been proposed. These techniques may be classified
route based on queue vacancy. The proposed algorithm Q- into three categories:
AODV improves average end to end delay, throughput and
x Probabilistic Techniques: The nodes that use
jitter, to some extent, as compared to AODV. The simulation
has been carried out on Qualnet.
probability based techniques for flooding decide to
transmit the message depending upon a specific
calculated probability or upon the occurrence of an
Keywords— AODV, Q-AODV, Queue, Queue length, Cross
event that related to the calculation of the probability
Layer, QoS, Improved QoS, Energy Efficient, Routing Protocol which with message can be transmitted further to the
neighboring nodes. Probabilistic based techniques
I. INTRODUCTION are usually proposed so as to reduce the amount of
redundant rebroadcasts. In order to avoid collisions,
this method also differentiates the timings of the
Broadcasting is the most elementary and the basic broadcasting of the messages. The probabilistic
operation in MANETs and plays an important role. It simply scheme includes probability-based and counter-based
means that a message is diffused to all nodes of the network methods.
from a given source node. The simplest and the commonly
used method for broadcasting in MANETs is flooding, x Area-Based Techniques: Area-Based Techniques use
where every uniquely received message is retransmitted knowledge of sender node locations to estimate
exactly once by each node. Various routing protocols, like whether a transmission would reach a significant
DSR (Dynamic Source Routing) [1], AODV (Ad hoc On amount of additional coverage area. In case of Area
Demand Distance Vector Routing) [2], ZRP (Zone Routing Based techniques, if the distance from the previously
Protocol) [3], LAR (Location-Aided Routing) [4], etc., broadcasting node is large then there is a higher
transmit route discovery packets, route maintenance packets probability of reaching more nodes. The relative
and topology update packets using the flooding technique. distance between neighbor nodes is estimated
This flooding function is very often raised in MANETs and either based on the received signal strength
thus, implementation of an efficient and effective flooding (distance-based scheme) or based on pre-acquired
scheme is crucial and important for optimizing the location information of neighbors by location
throughput of the network and for reducing the overhead of tracking devices such as the Global Position System
the routing protocols. (location-based scheme). The area based techniques
include distance-based and location-based schemes.

978-1-5386-6373-8/18/$31.00 ©2018 IEEE 294


2018 First International Conference on Secure Cyber Computing and Communication (ICSCCC)

x Neighbor knowledge techniques: Neighbor forwarder node or not, if so, it repeats the process of
Knowledge Techniques require the use of “Hello” selecting the forwarder nodes and attaches it to the RREQ
type packets so that nodes have explicit data before rebroadcasting it; otherwise it will drop the RREQ.
regarding their neighboring nodes and depending The routing overhead of the AODV protocol has been
upon this knowledge further broadcast of the packets reduced by 38% by using this protocol.
is decided.
R. Vinod Kumar et al in [10] have proposed an energy
II. LITERATURE REVIEW efficient routing scheme (EEAODV) based on node’s
energy and congestion level. In their work every
Broadcasting is the fundamental communication intermediate node on receiving a RREQ, calculates a
operation in Mobile Ad hoc Networks where one node sends threshold value essentially based on its queue occupancy
a message to all other nodes in the network. Broadcasting is and energy level. The threshold value is then compared with
also used for topology updates, for network maintenance, or interface queue occupancy, and the node rebroadcasts the
simply for sending a control or warning message. Flooding RREQ only if its interface queue occupancy is lower than
is a popular broadcast scheme used during the discovery the threshold value. In this way overloaded nodes do not
phase of most MANET routing protocols, where every take part in the formation of new routes, and traffic load is
message that is received for the first time, is broadcasted to distributed evenly across network. The proposed scheme has
all the neighboring nodes. Using flooding, each node been shown by the authors to have better energy efficiency
receives the message from all its neighbors in a collision- and end to end delay.
free network. A lot of research has been done for the
reduction of surplus transmissions from neighboring nodes. K. Muthumayil et al in [11] have proposed a dynamic
Some have used the probabilistic approach for controlling and energy conscious routing algorithm ECL-AODV. The
the flooding of control packets while others have used the proposed algorithm works by avoiding the nodes having low
threshold value of remaining battery. Few have taken into residual energy, and tries to establish a route with high
consideration the threshold on the node’s queue length along residual energy. The proposed routing protocol promotes
with the hop count. balanced energy consumption across the network thereby
increasing the lifetime of the network.
Daniel Reina et al in [8] have proposed broadcast
scheme based on probabilistic flooding technique for mobile Radhika D. Joshi et al in [12] have proposed a flood
wireless networks. In addition to addressing the broadcast control algorithm DEEAR, in which every intermediate
storm problem, the protocol also extends the lifetime of the node receiving a RREQ packet delays the rebroadcast of the
network. Authors have suggested the use of simple RREQ packet by rebroadcast time proportional to the ratio
probabilistic broadcast scheme based upon battery level of of mean residual battery power of the entire network to its
the nodes. All the nodes exchange their battery levels by residual battery power, and hence the nodes having low
periodically sending HELLO messages. The initiation of the battery level will rebroadcast RREQ packets later than the
process of route discovery is done by a node with the help of nodes with relatively higher battery level.
flooding. The RREQ packets are flooded in the network Sunho Lim et al in [13] have proposed a novel
which includes the information about the maximum and communication mechanism, called Random Cast, in which a
minimum battery levels in the neighborhood of the sender node defines a threshold value for overhearing, that
originator node. Every intermediate node on reception of a maintains a balance between energy and routing
RREQ packet calculates its forwarding probability performance. It also reduces superfluous rebroadcasts for
depending upon its own battery level and the maximum and broadcast packets, thereby saving more energy.
minimum battery levels. The proposed simple probabilistic RandomCast allows a node to decide whether to overhear (a
based algorithm extends network lifetime, and has higher unicast message) and whether to forward a broadcast
throughput and improved fault tolerance. message or not. Every node maintains an overhearing
Sofian Hamad et al [9] has proposed a proficient (rebroadcast) probability based on its mobility, the number
flooding CNRR (Candidate Neighbors to Rebroadcast the of neighbors and sender ID. Random Cast has been shown
RREQ) algorithm in which the packets are rebroadcasted to improve the energy good put by as much as 56 percent.
based upon the node’s position in the network thereby In literature, many researchers have provided the
efficiently spreading the control packets across the network. overview [14][15][16][17][18][19] of various routing
The proposed algorithm is applied on AODV’s route protocols while focusing on the advantages and
discovery process to reduce the number of Route Request shortcomings of each of them. However, the broadcast
(RREQ) messages that propagate through the network. The storm is one of the key concerns of the Mobile Ad hoc
proposed algorithm improves the performance of the Networks [20][21][22]. Flooding also results in wastage of
network by decreasing the retransmission that are redundant considerable amount of the sacred energy resource. In past
in nature, thus helps in minimizing the chances of collision few years’, lot of researchers have focused in the area of
among the nodes and also reduces the chances of contention. flooding [23][24][25][26][27][28][29]. Some have used the
The algorithm does its work by partitioning the radio probabilistic approach to control the broadcast storm, some
transmission range of the nodes into four zones such that have used the area based or neighborhood based knowledge
only one node per zone is selected to forward the RREQ. [30][31], few have used the dominating set theory
The selection is done on the basis of the closeness of the [32][33][34] while various others have used the hybrid or
nodes towards the edge of the zone such that more area can the adaptive approaches. However most of them, using the
be covered. The node broadcasting the RREQ attaches the probabilistic approach, have considered mainly the battery
address of the Candidate Neighbors to Rebroadcast the life as the metric to avoid flooding, for example choosing
RREQ (CNRR) into RREQ. Whenever a node receives the the nodes having high remaining battery as the relaying
RREQ packet, it checks if the sender node has selected it as

295
2018 First International Conference on Secure Cyber Computing and Communication (ICSCCC)

node while nodes having low energy silently dropping the


request. This approach although provides better results but C. Message Formats
at the same time can make the links, to these high energy
nodes, congested which may lead to dropping of packets and The proposed algorithm, Q-AODV is based upon the
hence retransmission. AODV routing protocol. The number of control messages
used in Q-AODV is same as in AODV with few
The proposed solution considers the level of congestion modifications. The control messages used in the whole
at a node as the metric to decide if the route request packet process of path discovery are: Route Request Packet
can be forwarded or not. The proposed system takes a (RREQ), Route Reply Packet (RREP), Route Error Packet
probability-based approach, where the probability of a node (RERR) and Route Reply Acknowledgement (RREP-ACK).
forwarding the route request is directly proportional to the The message formats of the control packets that have been
queue vacancy. used for Q-AODV are described in the following section.
III. SYSTEM MODEL FOR QUEUE AODV (Q- 1) Route Request (RREQ) Packet format
AODV) The RREQ packet in Q-AODV is similar to the one used
by AODV, except for an extra field “Random Probability”.
A. Focused Parameters
This field contains the random number generated by the
There are number of parameters that can be used to source. The random number generated lies within the range
calculate the rebroadcast probability of a node, for example
[0.1, 1]. The lower limit so defined imposes the condition of
number of neighboring nodes, bandwidth available,
remaining battery life, congestion etc. In our work we have threshold. The intermediate node uses this number along
focused on the congestion as the metric to calculate with their calculated Rebroadcast probability to decide
rebroadcast probability. In simple terms congestion can be whether to rebroadcast the RREQ or not.
defined as the number of packets waiting for being 2) Route Reply (RREP) Packet Format
processed at that node. These packets can be destined for the The message format for RREQ packet in Q-AODV is
node itself or the node may just have to forward/ rebroadcast same as that in AODV.
them to neighboring node. Route Error (RERR) and Route Ack. (RREP-ACK)
When a node receives a packet, the packet is first stored packet format in Q-AODV are same as in AODV.
in the Input Queue before being processed, the packet waits
in the queue until it is processed, and if after processing the IV. PROPOSED ALGORITHM
packet needs to be transmitted to some other node, then the
packet is inserted into the output queue from where it is In AODV the whole process of route discovery relies
fetched and then transmitted properly. So it is evident that heavily on flooding, where in any node requiring a route to
number of packets in these Queues is good indicator of other nodes broadcasts a route request (RREQ) packet. The
congestion at that node. In our proposed system, we have neighboring nodes (intermediate nodes) on receiving an
used the size of the output queue as an indicator of amount RREQ processes it, makes necessary changes in the RREQ
of congestion at any node. packet, its route table and then rebroadcasts the RREQ
packet. This process continues until the RREQ reaches the
B. Parameters Appended on each Node
destination or the TTL (time to live) of the RREQ packet
Ever node maintains a number of parameters and has expired. If any node receives an already processed
variables required for proper functioning of the nodes. RREQ, it silently discards it. Moreover even if destination
These parameters are essential as they may store important node has received the RREQ packet, other nodes still keep
information related to each layer of communication model, rebroadcasting the RREQ, until it has been disseminated
such as propagation model used, transmission power, throughout the network. In last few years a lot of research
receiver sensitivity, routing protocol, and many other has been done to restrict the broadcast storm, one of the way
parameter related to hardware and software specifications of to achieve this is probabilistic scheme. The probabilistic
the node. In order to implement the proposed algorithm we scheme is one of the alternative approaches that aim at
need each node to store few more parameters, these are: reducing redundancy through rebroadcast control in an
attempt to alleviate the broadcast storm problem. In the
a) Output Queue Max Size
probabilistic scheme, a mobile node rebroadcasts packets
This parameter is used by the node to store the value according to a certain probability. We intend to propose an
of maximum queue size. algorithm Q-AODV where the node decides whether to
b) Current Output Queue Size rebroadcast the received broadcast packet (RREQ) or not
This parameter is used to store the value of current during the route discovery process. The proposed algorithm
occupancy of the queue i.e. how much the output queue of focuses on probabilistic broadcast. The algorithm aims to
the node is full. improve the QoS parameters, namely End to End Delay,
Throughput, and Jitter by providing a less congested route
c) Rebroadcast Probability between the source and destination pair by considering the
It indicates the probability of the node with which it queue occupancy of the node. The algorithm has been
can rebroadcast the RREQ packet to the neighboring nodes. implemented on AODV and tends to reduce the
When a node receives a RREQ packet not destined for it, the communication through congested nodes thereby reducing
node calculates ‘Rebroadcast Probability’ based on the the probability of packet dropping and thus, packet
‘current output queue size’. Rebroadcast probability is then retransmission.
used by the node to decide whether it should rebroadcast the
A. Q-AODV Route Discovery Process
RREQ or not.

296
2018 First International Conference on Secure Cyber Computing and Communication (ICSCCC)

Whenever a node prepares itself to send some probability)


information to another node, it checks if there is any
available valid route for that destination in its route table, if x The process illustrated above continues until the
a valid route is available the node fetches the next hop for RREQ reaches the destination. Once the RREQ
that route and unicast the data packets to the next hop. reaches the destination node, the destination initiates
However if a valid route is not available, the route discovery route reply phase wherein, an RREP packet is
process starts. Q-AODV follows the same process for route unicasted to the next hop toward the originator node.
discovery as AODV but with some variations. Q-AODV In Q-AODV we have also incorporated the concept
brings in the concept of probability into the route discovery of threshold, by implying a condition that node in
process. The process of route discovery in Q-AODV is which the output queue is more than 90 percent full
divided into two phases namely route request phase and should not rebroadcast the RREQ packet and hence
route reply phase. Q-AODV has modified route request do not take part in route formation process. This
phase, to bring in the concept of probability to restrict the threshold has been implemented by restricting the
broadcast storm. The route reply phase, and route values taken by random probability generated by the
maintenance works same as in AODV. The section below originator node to be within the range [0.1, 1]. So if
describes in detail the route request phase of Q-AODV. the intermediate node has output queue more than 90%
full, then the rebroadcast probability calculated by it
1) Route Request Phase would always be less than 0.1, which in turn would
x Whenever a node needs to find out a route to any always be less than the random probability as it
other node, it generates a RREQ packet with all its ranges between [0.1, 1], so the node will not
fields properly initiated. It also generates a random rebroadcast the RREQ packet.
number within the interval [0.1, 1], and appends it
2) Route Reply Phase
with the RREQ packet as random probability. Finally
the node broadcasts the RREQ packet. The route Reply phase of Q-AODV works exactly in the
x Each intermediate node receiving the RREQ packet, same way as in AODV.
processes it and makes route table entries for the V. SIMULATION AND RESULT
originator node and node from which the RREQ was
received.
x The node then checks if it itself is the destination The simulation for the proposed algorithm is carried out
node. If it is, then the node initiates the route reply in Qualnet 5.1 and the results have been evaluated in
process. comparison to the performance of AODV. Comparisons are
x If the node is the not destination node, then it must made on the basis of various simulations that have been
rebroadcast the RREQ packet. In AODV every made with the purpose of analyzing the working and the
intermediate node unconditionally rebroadcasts the performance of the proposed algorithm Q-AODV. The
RREQ which leads to broadcast storm. However in following section describes the simulation environment that
our proposed algorithm each intermediate node has been used and various parameters that have been taken
calculates its own rebroadcasting probability based into consideration. Here we have considered a Mobile Ad
on its current queue length and maximum queue hoc Network having 22 nodes and Q-AODV as its routing
length and compares it with the random probability protocol. As Q-AODV has been designed for congested
stored in the RREQ packet. The node calculates the networks, so to simulate such a situation we have added 5
following parameter: CBRs between various nodes. Table below shows the
parameters and their values.
ሺை௨௧௣௨௧ொ௠௔௫௦௜௭௘ି஼௨௥௥௘௡௧௢௨௧௣௨௧ொ௦௜௭௘ሻ
‫ ܾ݋ݎܲݐݏܾܽܿ݁ݎ‬ൌ (1)
ை௨௧௣௨௧ொ௠௔௫௦௜௭௘

TABLE I P ARAMETERS OF SCENARIO


Above equation gives the proportion of how much the
queue is vacant. More the queue is vacant, more will Parameter Value
be Rebroadcast Probability and vice versa.
Simulator Qualnet 5.1
x The calculated Rebroadcast Probability is compared
with the random probability. If: No. of nodes 22
Simulation Time 200s
a) Random probability < Rebroadcast Probability; the
node will rebroadcast the RREQ packet. Environment Size 1500 x 1500 m2
b) Random probability > = Rebroadcast Probability; Traffic Type CBR (Constant Bit Rate)
the node will not rebroadcast RREQ packet and hence will Maximum Output Queue Size 10000 b
not take further part in the route formation process. Packet Size 512 b
x The basic concept behind the above condition is that
Packet Rate 10 packets/s
more the queue is vacant more will be the
rebroadcast probability, and more are the chances Mobility Model Random Waypoint
that the random number stored in the RREQ is Speed 1m/s
smaller than rebroadcast probability. Now as a result, Pause Time 1s
the less congested nodes (i.e. nodes having larger
rebroadcast probability) are more likely to Antenna Type Omni directional
rebroadcast the received RREQ than the nodes that
are congested (i.e. nodes having smaller rebroadcast

297
2018 First International Conference on Secure Cyber Computing and Communication (ICSCCC)

Throughput

83205

81681
81120
80784
Throughput (bits/s)

67298
67417
35830
34163
33916

29338
100000
50000
AODV
0
1 2 3 4 5 Q-AODV
CBR

Figure 3: Comparison of Throughput achieved using AODV and Q-AODV

Figure 4 presents the comparison of Average throughput


of CBRs while using AODV and Q-AODV as their routing
Protocol. From the figure, it is evident that throughput for
all the CBRs has increased while using Q-AODV.
Simulation results have shown that overall the throughput of
Figure 1: Screen Shot of Scenario the network has increased by 5%.
VI. CONCLUSION AND FUTURE WORK
Average End to End Delay
Routing is a significant issue in Mobile Ad hoc
Average End To End Delay

5.67617

6.5608

Networks. We have proposed an AODV based routing


4.01423
4.9913
3.57642
3.09787

2.68552

10
2.13175
1.52715
1.04429

algorithm “Q-AODV”. It has been designed to be used in


5 highly congested networks. Our proposed algorithm
(ms)

AODV outperforms the existing AODV protocol by controlling the


0
Q-AODV broadcast storm and reducing the total number of RREQ
1 2 3 4 5 relayed during the route discovery process i.e., the route
CBR discovery phase has been redefined by relaxing the
restriction of forwarding the RREQ to all the neighbors by
each node. To achieve this, Q-AODV follows a probabilistic
Figure 2: Comparison of Average End to End Delay achieved using AODV approach to find out least congested route. Q-AODV also
and Q-AODV enforces a threshold condition in the network to avoid the
nodes having their respective queues more that 90% full.
Figure 2 presents the comparison of Average End to End
Delay incurred by CBRs while using AODV and Q-AODV Simulation results have shown that for congested
as their routing Protocol. From the above figure it is clear MANETs, Q-AODV performs better than AODV. The
that end to end delay has decreased for CBRs while using proposed algorithm reduces the total number of RREQ
Q-AODV. Simulation results have shown that overall the packets relayed during the route formation process as
End to End delay for all the CBRs has reduced by 23%. compared to AODV. Using Q-AODV improves QoS
parameters, namely average jitter, throughput and average
Average Jitter end to end delay. The simulations have been performed
with the finite number of nodes and the simulations show
3.844

2.7983
2.15975

6 that by using Q-AODV, average jitter decreases by 7%,


0.894286
0.864828
3.22

0.673681
0.588419
1.3955
Jitter (ms)

1.1287

4 average End to End delay has reduced by 23%, and


2 AODV throughput has increased by 5%.
0 Q-AODV Although Q-AODV significantly improves specified
1 2 3 4 5 QoS parameters, but at the same time it faces certain
CBR challenges which are yet to be mitigated.
One of the major challenge being that Q-AODV may
sometimes lead to selection of route that may not be good
Figure 3: Comparison of Average Jitter achieved using AODV and Q- enough in terms of total number of nodes in the selected
AODV route i.e., Q-AODV may not lead to the shortest route
selection between the source and the destination.
Figure 3 presents the comparison of Average Jitter
incurred by CBRs while using AODV and Q-AODV as their Second challenge arises from the core concept of
routing Protocol. From the above figure it is clear that Probability, as it may happen that originator node generates
average jitter has decreased for CBRs while using Q-AODV. a large number (closer to 1) as random probability for
Simulation results have shown that overall the jitter for all RREQ which in turn may result in, most of the intermediate
the CBRs has reduced by 7%. nodes not rebroadcasting RREQ thereby reducing the
chances of route formation. So work can be done to
mitigate these challenges so as to make Q-AODV even
more efficient.

298
2018 First International Conference on Secure Cyber Computing and Communication (ICSCCC)

REFERENCES [23] B. Williams and T. Camp, “Comparison of Broadcasting Techniques


for Mobile Ad hoc Networks”, Proc. MOBIHOC ’02, pp. 914-205,
[1] D. Johnson and D.A. Maltz, “Dynamic Source Routing in Ad hoc 2002.
Wireless Networks”, Mobile Computing, Imielinski and H.F. Korth,
Kluwer Academic Publishers, pp. 153-181, 1996. [24] Shih-Hsien Wang, Ming-Chieh Chan, and Ting-Chao Hou, “Zone-
based Controlled Flooding in Mobile Ad hoc Networks”,
[2] Charles E. Perkins, Elizabeth M. Royer, “Ad hoc On Demand International Conference on Wireless Networks, Communications and
Distance Vector Routing”, In the Proceedings of Second IEEE Mobile Computing, 2005.
Workshop on Mobile Computing Systems and Applications,
(WMCSA '99), pp. 90-100, 25-26 Feb 1999. [25] ToshimitsuFujii, Mitsuru Kaji, Yuya Sasaki, Takahiro Hara ,
ShojiroNishio, “A Flooding Control Method with Ack-carry for
[3] Z.J. Haas and M.R. Pearlman, “The Zone Routing Protocol (ZRP) for Location-based Information Dissemination in Mobile Ad hoc
Ad hoc Networks”, Mobile Ad hoc Networking (MANET) Working Networks”, IEEE/IPSJ International Symposium on Applications and
Group of the Internet Eng. Task Force (IETF), Nov. 1997. the Internet, 2011.
[4] Y. Ko and N. Yaidya, “Location-Aided Routing (LAR) in Mobile Ad [26] H. Lim and C. Kim, “Multicast Tree Construction and Flooding in
hoc Networks”, Proc. MOBICOM ’98, pp. 66-75, 1998. Wireless Ad hoc Networks,” In Proc. of the ACM Int’l Workshop on
[5] C. Ho, K. Obraczka, G. Tsudik, and K. Viswanath, “Flooding for Modeling, Analysis and Simulation of Wireless and Mobile System
Reliable Multicast in Multi-Hop Ad hoc Networks”, Proc. Int’l (MSWIM), pp. 61-68, Aug. 2000.
Workshop Discrete Algorithms and Methods for Mobile Computing [27] Xinxin Liu, XiaohuaJia, Hai Liu, and Li Feng, “A Location
and Comm., pp. 64-71, 1999. Aided Flooding Protocol for Wireless Ad hoc Networks”, pp.302-
[6] P. Sinha, R. Sivakumar and V. Bharghavan, “Enhancing ad-hoc 313, 2007.
routing with dynamic virtual infrastructures”, IEEE INFOCOM 2001, [28] D. G. Reina, S.L.Toral, P. Jonhson and F. Barrero, “Hybrid Flooding
pp. 1763-1772, 2001. Scheme for Mobile Ad hoc Networks”, IEEE Communications letters,
[7] S. Ni, Y. Tseng, Y. Chen, and J. Sheu, “The broadcast storm problem vol. 17, no. 3, March 2013.
in a Mobile Ad hoc Network”, Proc. of ACM/IEEE MOBICOM’99, [29] Yunjung Yi and Mario Gerla, Taek Jin Kwon, “Efficient Flooding in
pp. 151-162, Aug. 1999. Ad hoc Networks: a Comparative Performance Study”, IEEE, pp.
[8] Daniel Reina, Princy Johnson, Federico Barrero and Sergio Toral 1059-1063, 2003.
“Optimization of network lifetime through energy-efficient broadcast [30] A. Qayyum, L. Viennot, and A. Laouiti, “Multipoint Relaying for
scheme using dynamic random walk”, pp. 1-5, 4-6 Sept. 2012. Flooding Broadcast Messages in Mobile Wireless Networks”, Proc.
[9] Sofian Hamad, Nazar Radhi, Hamed Al-Raweshidy, “Candidate 35th Hawaii Int’l Conf. System Sciences,2002.
Neighbours to Rebroadcast the RREQ for Efficient flooding in [31] Hai Liu, Pengjun Wan, Xiaohua Jia, Xinxin Liu and Frances
Mobile Ad hoc Network”, Wireless Advanced-2011 (WiAd’11), 20- Yao, “Efficient Flooding Scheme Based on 1-hop Information in
22 June 2011. Ad Hoc Networks”, in proceedings IEEE infocom,
[10] R. Vinod Kumar, R.S.D.WahidaBanu, “A New Approach For Load- Communications Society subject, 2006.
balancing in EEAODV Protocol”, in the proceedings of International [32] Y. Cai, K.A. Hua, and A. Phillips, “Leveraging 1-Hop Neighborhood
Conference on Advances in Engineering, Science and Management Knowledge for Efficient Flooding in Wireless Ad hoc Networks”,
2012 (ICAESM’ 12), pp. 485-488, 30-31 March 2012. Proc. 24th IEEE Int’l Performance Computing and Comm. Conf.
[11] K.Muthumayi , V.Rajamani , S.Manikandan, “A Novel Cross Layered (IPCCC), Phoenix, Arizona, Apr. 2005.
Energy based Ad hoc On-Demand Routing Protocol for MANETs”, [33] J. Wu and H. Li, “On Calculating Connected Dominating Set for
In the proceedings of Third International Conference on Advanced Efficient Routing in Ad hoc Wireless Networks”, Proc. Third Int’l
Computing-2011 (ICoAC’11), pp. 276-281, 14-16 Dec. 2011. Workshop Discrete Algorithms and Methods for Mobile Computing
[12] Radhika D. Joshi, Priti P. Rege, “Distributed Energy Efficient and Comm., pp. 7-14, 1999.
Routing in Ad hoc Networks”, Proceedings of Fourth International [34] I. Stojmenovic, M. Seddigh, and J. Zunic, “Dominating Sets and
Conference on Wireless Communication and Sensor Networks, Neighbor Elimination Based Broadcasting Algorithms in Wireless
2008(WCSN 2008), pp. 16 – 21, 27-29 Dec. 2008. Networks”, IEEE Trans. Parallel and Distributed Systems, vol. 13, no.
[13] Sunho Lim, Chansu Yu, Chita R. Das, “RandomCast: An Energy- 1, pp. 14-25, Jan. 2002.
Efficient Communication Scheme for Mobile Ad hoc Networks”, [35] F. Dai and J. Wu, “An Extended Localized Algorithm for Connected
IEEE Transactions on Mobile Computing, vol. 8, no. 8, pp. 1039- Dominating Set Formation in Ad hoc Wireless Networks”, IEEE
1051, 2009. Trans. Parallel and Distributed Systems, vol. 15, no. 10, pp. 908-920,
[14] Atul Eiman Alotaibi, Biswanath Mukherjee, “A survey on routing 2004.
algorithms for wireless Ad-Hoc and mesh networks”, Computer
Networks, pp. 940–965, 2012.
[15] Azzedine Boukerche, Begumhan Turgut, Nevin Aydin, Mohammad Z.
Ahmad, Ladislau Bölöni, Damla Turgut, “Routing protocols in Ad
hoc Networks: A survey”, Computer Networks, pp. 3032–3080, 2011.
[16] S. Hamma, E. Cizeron, H. Issaka, J.-P. Guedon, “Performance
evaluation of reactive and proactive routing protocol in IEEE 802.11
Ad hoc Network” , in the Proceedings of the SPIE Ad hoc and Sensor
Networks, Next-Generation Communication and Sensor Networks,
vol. 6387, no. 9, 2006.
[17] C. Liu, J. Kaiser, “A Survey of Mobile Ad-Hoc network Routing
Protocols”, Tech. Report, University of Ulm, pp. 2003-08, 2005.
[18] Mehran , Tadeusz Wysocki, Eryk Dutkiewicz, “A review of routing
protocols for Mobile Ad hoc Networks”, Ad hoc Networks, pp. 1–22,
2004
[19] C. Liu, J. Kaiser, “A Survey of Mobile Ad-Hoc network Routing
Protocols”, Tech. Report, University of Ulm, no. 2003-08, 2005.
[20] Martha Steenstrup, “Routing in communication networks”. New
Jersey, Prentice Hall. ISBN 0-13-010752-2.
[21] Majid Khabbazian and Vijay K. Bhargava, “Localized Broadcasting
with Guaranteed Delivery and Bounded Transmission Redundancy”,
IEEE Transactions on computers, vol. 57, no. 8, August 2008.
[22] Majid Khabbazian and Vijay K. Bhargava, “Efficient Broadcasting in
Mobile Ad hoc Networks”, IEEE Transactions on mobile computing,
vol. 8, no. 2, February 2009.

299

You might also like