Sunteți pe pagina 1din 4

Candidate Neighbours to Rebroadcast the RREQ for

Efficient flooding in Mobile Ad hoc Network


Sofian Hamad, Nazar Radhi, Hamed Al-Raweshidy - IEEE Senior Member
Department of Electronic and Computer Engineering
Brunel University, London, UK
Sofian.Hamad@brunel.ac.uk

Abstract- Most of the Routing protocol used flooding as the process to network topological view. Additionally, there are hybrid
disseminate the update packets in the network. The basic flooding is protocols that combine the best features of both proactive and
cause high retransmissions which lead to packet collisions and media reactive routing protocols. The periodic routing information
congestion that can significantly degrade the network performance. updates due to broken links in the proactive routing protocols
Knowing the geographical position of the mobile nodes can assist the can lead to a heavy control overhead during high mobility.
protocol to reduce the number of retransmissions, therefore enhancing
the protocol performance. In this paper, an Efficient Flooding
Hence, these protocols surfer from excessive routing control,
Algorithm has been proposed that makes use of the nodes' position to overhead and therefore are not scalable in MANETs, which have
rebroadcast the packets and efficiently spread the control traffic in the limited bandwidth and whose topologies are highly dynamic.
network. The proposed algorithm is applied on the route discovery
process of Ad-hoc On Demand Distance Vector (AODV) protocol to In conventional on-demand routing protocols [2-4], a node
reduce the number of propagating Route Request (RREQ) messages. discovers routes to a particular destination, by broadcasting a
The RREQ has been modified by assigning a list to the RREQ contain Route Request packet (RREQ). Upon receiving the RREQ, the
fourth Candidate Neighbours to Rebroadcast the RREQ (CNRR). The node checks whether or not the packet has been previously
simulation results shows that our scheme reduces the routing overhead received. In case of packet has been received previously the
of AODV protocol up to 38%. node will drop the packet, otherwise the node will checks
whether it has a route to the destination, if yes, the node will
Keywords: Position Based; Efficient Flooding; RREQ. send back Route Reply (RREP) to the source node; otherwise
the node will rebroadcast the RREQ to its immediate neighbors
I. Introduction until the destination is found. This method of route discovery is
There has been a growing research activity on wireless referred as blind flooding. Every mobile node rebroadcasts one
Mobile Ad-hoc Networks (MANETs) over the recent years due copy of received RREQ, so the maximum number of
to their potential effectiveness in civilian and military rebroadcasts is equal to N – 2, where N is the number of nodes
applications. MANETs are formed dynamically by an in the network. This can potentially lead to excessive redundant
autonomous system of mobile nodes that are connected via retransmissions therefore high channel contention and causing
wireless links without using an existing fixed network excessive packet collisions in dense networks. Such a
infrastructure or centralized administration [1]. The nodes are phenomenon is refried to as broadcast storm problem [10],
free to move randomly and organize themselves arbitrarily; thus, which significantly increases network communication overhead
the network’s wireless topology may change rapidly and and end-to-end delay [10, 11]. To reduce the impact of blind
unpredictably. Nodes in MANETs act as end points as well as flooding, a number of broadcasting techniques have been
relay routers to forward packets for other nodes in a wireless suggested in [10-13].
multi-hop environment. One of the fundamental challenges in Many approaches are proposed to improve flooding
MANETs and especially in a multi-hop scenario is the design of performances which are based on reducing the number of
dynamic routing protocol that can efficiently establish routes to redundant messages. However, reducing the number of
deliver data packets between mobile nodes with minimum redundant messages leads to a low degree of coverage and
communication overhead while ensuring high throughput and connectivity. This interdependency between two phenomenons
low end-to-end delay. poses a challenge between the balance of overhead message (i.e.,
the level of redundancy) and coverage.
Many routing protocols have been proposed for MANETs
over the past few years [2-7]. In general, the routing protocols II. Related Work
for MANETs fall into two categories [8] i-e proactive and
reactive routing protocol. Proactive routing protocols, such as Several schemes have been proposed to decrease the effect of the
DSDV [5] and OLSR [9], attempt to maintain consistent and up- broadcast storm caused by simple flooding [12, 13]. Those
schemes can be classified into three categories: probability-based
to-date routing information tables on each node to every possible
methods, area based methods and neighbor knowledge methods.
destination in the network by periodically exchanging routing
table information. On the other hand, in reactive routing 1) Probability-based methods:
protocols, such as AODV [2] and DSR [3]; routes are only
discovered when needed. Each node maintains a route for a Reference [10] proposed two methods to reduce the number
of rebroadcasts: the probabilistic scheme and the counter-based
destination without periodic routing table exchanges or full

978-1-4577-0109-2/11/$26.00 ©2018 IEEE 26


scheme. The probabilistic scheme is similar to the simple selection process is performed by determining the closest node to
flooding, except that the nodes rebroadcast the RREQ with a the edge of the zone as shown in Figure1 to provide more
predetermined probability p. In the counter-based scheme, upon coverage area. The sender attaches the address of the Candidate
receiving a previously unseen broadcast message, the mobile Neighbor to Rebroadcast the RREQ (CNRR) into the RREQ
node initializes a counter with a value of one and set a random field. Any neighbors when received the RREQ it will check if the
defer time. During this deferring time; the counter is incremented sender select it as forwarder node or not, if so, it will partition its
by one for each redundant message received. If the counter is less transmission range and select a new set forwarder nodes and
than a predetermined threshold, when the deferring time expires, attach them into the RREQ and rebroadcast the RREQ, otherwise
the message will be relayed. Otherwise, it is simply discarded. it will drop the RREQ. We assume in this work that every node
The probability-based and the counter-based methods are simple, is able to obtain its geographical position. Besides that, every
but their performance depends on the variation of network node shares its position information with its direct neighbor
density. This is due to the values of the probability and the through the HELLO message mechanism.
counter threshold that are defined regardless of the variation on
the network environment.
2) Area-based methods:
The area-based methods consist of the distance-based
scheme and the location-based scheme. It emphasize on how
much more additional area (than the area covered by the previous
broadcast) can a node offer if it rebroadcasts the message.)
Within the node transmission range, the longer distance from the
previous broadcasting node, the more additional coverage can be
acquired resulting in more opportunity to reach more nodes. The
relative distance between neighboring nodes is estimated either
by received signal strength in distance-based scheme or pre-
acquired location information of neighbor by location tracking
devices such as the Global Position System (GPS) in location-
based scheme. EFPA [15], the author proposed an efficient
flooding algorithm, which generates a small number of packet Figure 1: Divide the transmission range and locate each neighbor in the
transmissions during a short time. EFPA [15] allocates a priority right Zone.
of packet transmission or a waiting time to every node For instance, Figure. 1 illustrates an example of how to partition
considering the distance from a sender node and the direction of the transmission range into four zones. To locate each neighbor
packet transmission, so every node in a network can receive in the right zone we follow the bellow equation:
packets rapidly.
S = Sender Node
Our proposed method is based on the area based methods
where the accurate position of node is determined using GPS. If: Ss ≤ As and Sy ≤ Ay ) (1)
3) Neighbor knowledge methods: Then we can locate node A inside Zone1 of node S.
The neighbor knowledge methods usually utilize the one-hop Else if: Ss > Bs and Sy ≤ By (2)
or two-hop neighborhood information to reduce redundant Then we can locate node B inside Zone 2 of node S.
transmissions. The neighborhood information is obtained by Else if: Ss ≥ Cs and Sy > Cy (3)
periodically exchanging "HELLO" messages among neighbor Then we can say node C inside Zone 3 of node S.
nodes. In [16], T.D. Le and H. Choo proposed the concept of 2- Else: Ss < Ds and Sy < Dy (4)
hop backward information to be used for minimizing the number Then we can say node D inside Zone 4 of node S.
of forwarding nodes and reducing the collisions in the network. After locate each neighbor in the right zone, then we calculate
The rest of the paper is organized as follows: In Section III, the distance from the sender node to each neighbor according to
we present in detail our proposed Efficient Flooding Based on bellow equation:
Node Position Location. We evaluate our approach and present
the simulation results in Section IV. Section V. includes the Distance (S, N) = ƒ(Ss − Ns )2 + (Sy − Ny )2 (5)
concluding remarks of this paper. According to equation 5, node S will be able to know the
distance from each neighbor.
III. Proposed Method So now node S locates each neighbor in the right zone from its
The aim of this work is to design an efficient flooding perspective in addition the distance from each neighbor. To
algorithm for mobile ad hoc network to improve the network choose the candidate neighbor in each zone, node S will choose
performance by eliminating the redundant retransmission, the farther node in each zone. The final step is to insert the four
therefore reducing the chances of contention and collision among candidates neighbors into CNRR (Candidate Neighbors to
the neighboring nodes. This can be achieved by involving a Rebroadcast RREQ) inside the RREQ field which has been
specific set of nodes in the dissemination process of the RREQ. modified as illustrate in figure2. At the end, node S will send the
The key concept of the proposed algorithm is to partition the RREQ which now contains four candidates’ neighbors (A, B, C
radio transmission range of the mobile node into four zones. and D).
Then, one node per zone is selected to forward the RREQ. The

27
On the reception of RREQ, each node checks CNRR field inside 2 Mbps and the transmission range is set to 250 m. The
RREQ and the decision of rebroadcast is taken based on the evaluations are conducted with a total of 150 nodes that are
inclusion of its network address in the list. If the node finds its randomly distributed in an area of 750m x 750m. We use
address inside this field that means rebroadcast the RREQ Random Waypoint to model node mobility. The speed was
otherwise discard it. constant to 10 m/s in all the simulation time. In each test, the
Nodes A,B,C and D when receive the RREQ, they check CNRR simulation lasts for 800 seconds. The size of each Constant Bit
field and since they find themselves inside the RREQ therefore Rate (CBR) packet is 1000 bytes and packets are generated at
they do the same as node S and rebrodcast the RREQ. the fixed interval rate of 5 packets per second. 25 flows (5, 10,
15, 20, and 25) were configured to choose a random source and
Type J R G D U Reserved Hop Count
destination during the simulation.
RREQ ID
Desdtination IP Address B. Performance Metrics
Four metrics were used to evaluate our CNRR and AODV
Desdtination Sequance Number
protocol:
Originator IP Address
 Packet Delivery Ratio: The ratio of the data packets
Originator Sequance Number
succeed to deliver at the destinations to those generated
CNRR IP Address by the CBR sources.
Figure 2: Modified RREQ  Data Drop: This includes all deleted packets in the
network.
We noticed that in MANETs all the nodes move randomly
 Total Overhead: The number of control packets
with high mobility, the farthest neighbours may move out of the
transmitted in the network.
communication range with a high probability. Also, due to the
collisions, interference and decrease of the channel capacity with  Total Throughout: The total number of application
high distance between the sender and receiver; some farthest layer data bits successfully transmitted in the network
per second.
neighbours in the candidates list may fail to receive the broadcast
RREQ successfully. We deal with these problems by C. Results and Discussion
mechanisms in which the candidate nodes are selected based on Here we examine five performances of CNRR and compare it
their distance from the source node. A source node can select with AODV protocol. The results of simulation experiments
only a candidate node among the neighbours if the distance demonstrate that CNRR indeed has more advantages and
between them is less than 80% of the source transmission range. improves the performances of classic AODV.

Figure 4(a): Throughput (b/s) vs Flows Number


Figure 3: Flowchart of processing a new RREQ packet on the Sender side Throughput
and Reciver side.
Throughput is a vital metric that measures the transmission
IV. Performance Evaluation ability of a network. It is defined as the number of bits
In order to evaluate the performance of our CNRR protocol, we transmitted within a unit time. In Fig. 4(a), we compare the
simulate the proposed mechanisms using NS2 Simulator [14]. network throughput for different flows settings. The figure
The simulation environment, performance metrics and results shows that CNRR outperforms AODV. The throughput
are discussed in the subsequent sections. improvement of CNRR is due to its reduction of rebroadcasting.
The fewer rebroadcasts, the smaller bandwidth consumption by
A. Simulation Environment control messages is. This also results in lower degree of
In our simulations, the MAC layer runs on the IEEE 802.11 contentions and collisions, which leads to relatively higher
Distributed Coordination Function (DCF). The bandwidth set to throughput.

28
Data Drop:
Fig. 4(d) shows that CNRR has viewer data drop than AODV that
because in AODV case, any node try to send more data
(RREQs), so packet collision more likely to happen. For this
reason AODV curve go very high when the flows become more
than 10 while stay in same level in case of CNRR protocol.

V. Conclusion

In this paper, CNRR has been proposed to improve the


flooding process for MANET. Knowing the geographical
position of the mobile nodes can assist the protocol to reduce the
number of retransmissions, therefore enhancing the protocol
performance. CNRR protocol selects four neighbors as
Figure 4(b): Total Overhead vs Flows Number
candidates to rebroadcast the RREQ in case there aren’t
Overhead available route on those nodes rather than all the neighbors
Fig. 4(b) shows the overhead increases when the number of rebroadcast the RREQ. We succeed to maintain the level of
flows becomes high in AODV while stay in same level with connectively among the network and in the same time reduce the
CNRR this is due to the reason that AODV sends RREQ without overhead.
knowledge of which the best neighbors shall rebroadcast the
RREQ. Also the figure shows that CNRR performs better that References
AODV due to less rebroadcast RREQs. [1] http://www3.ietf.org/proceedings/02mar/179.htm .
[2] C. E. Perkins, E. M. Belding-Royer, and S. R. Das, "Ad hoc On-Demand
Distance Vector (AODV) Routing," IETF Mobile Ad Hoc Networking
Working Group INTERNET DRAFT, 19 January 2002.
[3] D. B. Johnson, D. A. Maltz, and Y.-C. Hu, "The Dynamic Source Routing
Protocol for Mobile Ad Hoc Networks (DSR)," IETF MANET Working
Group INTERNET-DRAFT, 19 July 2004.
[4] http://www.ietf.org/internet-drafts/draft-ietf-manettora.
[5] C. E. Perkins and P. Bhagwat, "Highly dynamic destination-sequenced
distance vector routing (DSDV) for mobile computers," Proceedings of
ACM SIGCOMM'94, pp. 234-244, September 1994.
[6] Y.-B. Ko and N. H. Vaidya, "Location-aided routing (LAR) in mobile ad
hoc networks," Proceedings of the 4th annual ACM/IEEE international
conference on Mobile computing and networking (Mobicom '98), pp. 66-
75, October 25-30,1998.
[7] Z. J. Haas, M. R. Pearlman, and P. Samar, "The zone routing protocol
(ZRP) for ad hoc networks," IETF Internet Draft, draft-ietf-manet-zonezrp-
Figure 4(c): Packet Delivery Ratio vs Flows Number 04.txt., July 2002.
[8] C. K. Toh, Ad-hoc Mobile Wireless Networks: Protocols and Systems:
Packet Delivery Ratio: Prentice Hall, Inc., 2002.
Fig. 4(c) represents the comparison of Packet Delivery Ratio and [9] www.ietf.org/rfc/rfc3626.txt.
we can see that CNRR outperforms AODV. The reason is that [10] S.-Y. Ni, Y.-C. Tseng, Y.-S. Chen, and J.-P. Sheu, "The broadcast storm
when the number of flows becomes very high, CNRR will be problem in a mobile ad hoc networks," Proceedings of the Fifth Annual
able to maintain the level of stability, because any nodes in the ACM/IEEE International Conference on Mobile Computing and
Networking, pp. 152-162, August 1999.
network will not forward the RREQ if the sender doesn’t select
[11] Y.-C. Tseng, S.-Y. Ni, and E.-Y. Shih, "Adaptive approaches to relieving
it to that. broadcast storms in a wireless multihop mobile ad hoc networks," Proc. of
IEEE Transactions on Computers, vol. 52, pp. 545--557, May 2003.
[12] B. Williams and T. Camp, "Comparison of broadcasting techniques for
mobile ad hoc networks," Proc. of the 3rd ACM international symposium
on MANET & computing, MOBIHOC, pp. 194 - 205, June 2002.
[13] W. L. a. J. Wu, "A reliable broadcast algorithm with selected
acknowledgements in mobile ad hoc networks," Proceedings of IEEE 2003
Global Communications Conference (GLOBECOM 2003), San Francisco,
Califonia, 2003.
[14] http://www.isi.edu/nsnam/ns.
[15] Chan Jaegal,“An Efficient Flooding Algorithm for Position-based Wireless
Ad hoc Networks” Proceedings of the Third International Conference on
Convergence and Hybrid Information Technology, 2008.
[16] T.D. Le and H. Choo, ―Efficient Flooding Scheme Based on 2-Hop
Backward Information in Ad Hoc Networks□, In Proceeding of the IEEE
International Conference on Communications, pages 2443-2447, 2008
Figure 4(d): Data Drop vs Flows Number

29