Article

Routing Load of Route Discovery and Route Maintenance in Wireless Reactive Routing Protocols

Abstract and Figures

In this paper, we present analytical study of routing overhead of reactive routing protocols for Wireless Multihop Networks (WMhNs). To accomplish the framework of generalized routing overhead, we choose Ad-Hoc on Demand Distance Vector (AODV), Dynamic Source Routing (DSR) and Dynamic MANET on Demand (DYMO). Considering basic themes of these protocols, we enhance the generalized network models by adding route monitoring overhead. Later, we take different network parameters and produce framework discussing the impact of variations of these parameters in network and routing performance. In the second part of our work, we simulate above mentioned routing protocols and give a brief discussion and comparison about the environments where these routing protocols perform better.
Content may be subject to copyright.
arXiv:1208.2405v1 [cs.NI] 12 Aug 2012
Routing Load of Route Discovery and Route
Maintenance in Wireless Reactive Routing Protocols
D. Mahmood, N. Javaid, U. Qasim, Z. A. Khan$
University of Alberta, Alberta, Canada.
Department of Electrical Engineering, COMSATS
Institute of Information Technology, Islamabad, Pakistan.
$Faculty of Engineering, Dalhousie University, Halifax, Canada.
Abstract—In this paper, we present analytical study of routing
overhead of reactive routing protocols for Wireless Multihop
Networks (WMhNs). To accomplish the framework of general-
ized routing overhead, we choose Ad-Hoc on Demand Distance
Vector (AODV), Dynamic Source Routing (DSR) and Dynamic
MANET on Demand (DYMO). Considering basic themes of
these protocols, we enhance the generalized network models
by adding route monitoring overhead. Later, we take different
network parameters and produce framework discussing the
impact of variations of these parameters in network and routing
performance. In the second part of our work, we simulate above
mentioned routing protocols and give a brief discussion and
comparison about the environments where these routing protocols
perform better.
Index Terms—Overhead, Routing, Reactive, Protocols, Route,
Discovery, Maintenance.
I. INTRODUCTION
Recent demands of communication make infrastructure
communications replaced with infrastructure-less communica-
tions. Along with other technologies, WMhNs are promising
to provide freedom of communications as they offer supple
structures, low costs and ability to cope with ever growing
needs of bandwidth. In WMhNs, one node can be out of range
with another. Hence to communicate between such nodes,
there must be some node/s working as a bridge between them.
In other words, intermediate nodes act as router to receive
and transmit routing and data packets. This is the reason that
each node must work as a routing device. Such networks are
gaining popularity day by day. So it is a challenge to maintain
and improve quality of WMhNs.
This communication is possible with the help of numerous
protocols that are functioning on different layers. Besides
other protocols, routing layer protocols play vital role in
contributing smoothness and better functionality of a network.
A routing protocol creates, maintains and synchronizes a
routing table and relevant routing information for a node. More
efficient routing layer protocol results in more efficient net-
work performance. Numerous protocols have been developed
and can be categorized into two major classes i.e., reactive
routing protocols and proactive routing protocols. Reactive
routing is based on the immediate response phenomena [1].
In our work we emphasis on reactive routing only. It is a
common observation that lots of simulated work has been
contributed with respect to analytical modeling of above men-
tioned routing protocols. This was the basic motivation to work
on mathematical framework that gives and discusses precise
information about the behaviors of reactive routing protocols
in different environments and parameters. For this purpose, we
choose three main routing protocols from reactive routing i.e.,
DYnamic MANET On-demand (DY M O)[2], Ad-hoc On-
demand Distance Vector (AODV )[3] and Dynamic Source
Routing (DSR)[4]) for our studies.
II. RELATED WORK AND MOTIVATION
Authors in [5] provide analytical framework for calculating
routing overhead of reactive protocols. They quantify route
discovery process, i.e., overhead due to route REQuest packets
and route REPly packets of any network underlying a reactive
routing protocol. However, link monitoring overhead is not
considered in their work. [6] gives a combined framework of
reactive and proactive routing protocols. The proposed models
express scalability issues of a network considering both classes
of routing protocols i.e., reactive and proactive. In [7], authors
propose analytical model which presents the effect of traffic
on routing overhead whereas, [8] presents a survey of routing
overhead on both reactive and proactive protocols and discuss
cost of energy as routing metric. I.D Aron et.al [9] present
link repairing modeling, both in local repairing and source
to destination repairing of two routing protocols, which were
DSR and WRP. They compare these two routing protocols,
though aggregate routing overhead is not considered in [9].
In [10], authors present brief understanding of scalability
issues of network, however, impact of topology change is not
sufficiently addressed.
We enhance the framework produced by [5] by adding link
monitoring overhead and trigger message overhead. Hence,
we present a general routing overhead framework. After cal-
culating the aggregate routing overhead, we calculate rate of
change in different network and protocol parameters. Besides
modeling, we simulate above mentioned three reactive routing
protocols and discuss their behaviors according to different
environments and scenrios.
Source
Destination
RREQ
RREP1
RREP2
RREQ Propagation
Route described Source
to destination
mobile node NETWORK ASSUMPTIONS
No. of hops between Source and
Destination = 4
No. of neighbors at first Tier = 4
No. of neighbors at 2nd and higher
tiers = 3
Fig. 1. Propagation of RREQ and RREP in Network
III. MODELING ROUTING OPERATIONS
In our work, we assume that nodes of network are placed in
grid environment while nodes have different life times. Certain
sections of the grid are vulnerable to failure due to some
reasons. That can be the power failure or radio jamming. If one
section of grid fails for certain time, it surely gives variations
in number of nodes, number of hops and obviously in link
monitoring overhead.
A. Reactive Route Discovery Overhead
Route discovery overhead can further split into two parts
i.e., Overhead due to Route REQuest (RREQ) dissemination
and Overhead due to generation of Route REPly (RREP). Con-
sidering RREQ dissemination Overhead, it mainly depends
upon the number of hops a packet travels and number of
neighbors of nodes at each hop.
Route Request is propagated in entire network till destina-
tion is find. If we consider that source and destination nodes
of a network are placed at opposite corners of network, then
they bear highest number of hops [11]. As shown in Fig.
1, when a RREQ is generated on the originator node, there
are four neighbors, though, from second tier up to ntiers,
number of effective neighbors are three. Moreover, at each
intermediate node, there is some coverage index where the
packet is processed [12]. As, RREQ packet is blind flooded
that allows a node to process RREQ control packet once
and discard itself if receives second time. [5] presents the
mathematical model for route discovery overhead i.e., RREQ
overhead and RREP overhead of reactive routing. According
to this mathematical modeling, overhead of RREQ is:
RRREQ =
H
X
n1
(4)3H14
X
i=2
[(n1i)
H1
X
j=1
Nj]pCi(1)
Ciis the additional coverage index of a node having inodes
as its neighbors. Ciis described and presented in [12], His the
expected number of hops of network and Njis the expected
number of neighbors at th hop.
Once RREQ reaches destination, RREP is generated. It
follows the reverse path. For simplifying the concept, in Fig. 1,
Source
Destination
ACK
HELLO
HELLO
HELLO
HELLO
ACK
ACK
ACK
Route Monitoring Overhead
No. of Links = 4
No. of Periodic Messages per Link = 2
Route Monitoring Overhead = 4 x 2 (Route Life Time /
Periodic Message Interval Time)
Fig. 2. Link Monitoring Overhead of a Route
two route RREP’s are generated i.e., two paths to destination
are found. Considering the same point of view, [5] has given
the expected overhead generated by RREP packets as:
RRREP =H+H
2(nh2)p(2)
Overall routing overhead due to route discovery can be repre-
sented as:
RDIS COV E RY =RREQ +RREP (3)
Placing values in Eq.3we get
RDIS COV ERY =
H
X
n1
(4)3H14
X
i=2
[(n1i)
H1
X
j=1
Nj]pCi+H
2
(nh2)p
(4)
B. Reactive Route Maintenance Overhead
Link monitoring overhead in reactive protocols is minor
with respect to route discovery overhead, however, it also
contributes in aggregate routing overhead. H ello messages
are periodically propagated when a route is established till
expiration of route in AODV. In DSR AC K message works
almost in the same fashion [3-4]. As we know that life time
of any link in a route is a random variable. It can be defined
as the time when two nodes create a pair with each other till
the time, this pair is prone to breaks due to any reason mostly
due to radio problems [13].
As depicted in Fig. 2 , when a route is established having
llinks, a link monitoring message is propagated periodically
till the route expires. Hence, total number of hello messages
broadcasted for monitoring of one route whose route expiry
time is Tand periodic interval time is tcan be represented
mathematically as:
RHE LLO(e)= 2( T
t)l(5)
And number of periodic link monitoring messages for nroutes
can simply be represented as:
Source to Destination
No. of links = 4
Active Nodes of Network = 20
Source
Destination
ABC
PATH = Source-A- B- C- Destination
Fig. 3. Source to Destination Route(Grid Environment)
RHE LLO =
n
X
i=1
2( T
t)l(6)
C. Aggregate Reactive Overhead
Overall routing overhead due to route discovery and route
monitoring is be stated as:
RO =RDI SCOV E RY +RHE LLO (7)
Putting values from Eq.4 and Eq.6, we get:
RO =
H
X
n1
(4)3H14
X
i=2
[(n1i)
H1
X
j=1
Nj]p(i) + H+H
2
(nh2)p+
n
X
i=1
2( T
t
)l(8)
One can not make bricks without clay. In proposed network
model, nodes are placed in a grid that works as clay for
calculating reactive routing overhead. However, if we consider
that these nodes have different life times within the same
network, or have power failure at certain sections of grid then
number of hops as well as number of nodes, at that instance,
can be varied. To calculate overhead during such scenarios,
we can take partial derivatives with respect to number of hops
and number of nodes of network. Besides number of hops and
number of nodes of network, we take parameters of routing
protocols as route life time and periodic messages interval time
which can also be varied. Hence, we have two more metrics
while modeling routing overhead.
Considering Fig. 3, E q.8gives the routing overhead of
reactive routing protocol. However, if there is some change
in the number of nodes of network due to any reason like
power failure or radio jamming, the source requires a new
route to its destination, as illustrated in Fig. 4. That new route
bears different number of intermediate nodes as well as hops
or links. To calculate such variations in the network, we have
to analyze rate of change in routing overhead with respect to
different parameters.
From Eq.8, function y(i.e. RO) we get different parameters
as n,H,Tand t. As discussed earlier, nstands for total
S
D
DEAD NODES OF NETWORK
U
VW X Y
PATH = S- U-V- W- X- Y- D
If Three Nodes Die due to any Reason
Source to Destination
No. of Links in Route = 6
No. of Active nodes in Network = 17
Fig. 4. Source to Destination Route(Portion of Grid Black Out)
number of nodes in a network, Hrepresents number of hops of
a network, Tis the route life time while tis periodic interval
time for link monitoring. If we take partial derivative with
respect to number of nodes of network, periodic interval time,
route life time and number of hops of network, we get to
know the overall rate of change in the network with respect
to routing overhead.
y(n, H, T , t) =
H
X
n1
(4)3H14
X
i=2
[(n1i)
H1
X
j=1
Nj]p.Ci) +
H+H
2
(nh2)p(9)
Considering the parameters of function y, it is interesting to
know that number of hops or links are dependent on the
number of nodes of a network. Whereas, number of link
monitoring messages are dependent upon number of links of
route, link life time and periodic message life time. Their
relationships are discussed further in the coming equations.
Considering the variation in number of nodes, we take partial
derivative of ywith respect to nand we get:
∂y/∂ n =
H
X
n1
(4)3H1[
4
X
i=2
[(i)
H1
X
j=1
Nj]p.Ci] +
H+H
2
(h2)p(10)
Varying number of nodes of a network certainly effects the
number of hops of some routes. Such variation can result in
change in routing overhead. Number of hops are in relationship
with the number of links per route and number of links is a
vital parameter for link monitoring overhead. When number
of nodes vary, there is a possibility of change in number of
hops. To calculate this change, we take partial derivative of
function ywith respect to H.
∂y/∂ H =
H
X
n1
(4)3H1+H1(3H1)[
4
X
i=2
[(n1i)
H1
X
j=1
Nj]p.Ci] +
1 + 1
2
(n3)p(11)
Number of nodes and number of hops play a vital role
in routing overhead. We can use chain rule to calculate
overall routing overhead, assuming route life time and periodic
link monitoring message interval constant. For this purpose,
we have our Eq.4discussing routing overhead due to route
discovery process. Considering Eq.4as a function xwe get
our partial derivatives of nand Has Eq.10 and E q.11.
Routing overhead due to varying number of nodes and number
of hops can be calculated as:
dx = ( ∂x
∂n )dn + ( ∂x
∂H )dH (12)
placing values in Eq.12, we get::
dx = (
H
X
n1
(4)3H1[
4
X
i=2
[(n1i)
H1
X
j=1
Nj]p(Ci)] + H+H
2
(nh2)p)dn +
(
H
X
n1
(4)3H1[
4
X
i=2
[(n1i)
H1
X
j=1
Nj]p(Ci)] + 1 + 1
2
(n3)p)dH +
(13)
Considering rate of change in route life time:
∂y/∂T =
n
X
i=1
(2
t)li(14)
And to analyze variation in periodic interval time for link
monitoring:
∂y/∂t =
n
X
i=1
2( T
t2)li(15)
Considering Eq.14 and Eq.15, we can conclude that if there
are different route life times active in a network along with
different periodic message intervals, than overall routing over-
head of link monitoring of a Reactive Routing Protocol is the
total derivative with respect to route life time and periodic
message interval. To calculate so, we consider RHELLO
(expressed in Eq.6)as a function zwhose partial derivatives
are expressed in Eq.14 and Eq.15. Taking total derivative we
get:
dz = ( z
∂T )dT + ( ∂z
∂t )dt (16)
Placing the values we get the routing overhead due to varying
routing protocol parameters of route life time and periodic
message update time:
dz = (
n
X
i=1
(2
t)li)dT + (
n
X
i=1
2( T
t2)li)dt (17)
Applying chain rule on the function yto get the total deriva-
tive which is actually the sum of all the partial derivatives of
a function, we get the optimum model for route discovery and
route monitoring overhead in reactive routing protocols.
dy = ( ∂y
∂n )dn + ( ∂y
∂H )dH + ( y
∂T )dT + ( ∂y
∂t )dt (18)
Placing the values, we get:
dy = (
H
X
n1
(4)3H1[
4
X
i=2
[(n1i)
H1
X
j=1
Nj]p(Ci)] + H+H
2
(nh2)p)dn +
(
H
X
n1
(4)3H1[
4
X
i=2
[(n1i)
H1
X
j=1
Nj]p(Ci)] + 1 + 1
2
(n3)p)dH +
(
n
X
i=1
(2
t
)li)dT + (
n
X
i=1
2( T
t2)li)dt (19)
0
0.2
0.1
5
0.1
0.05
700
500300
100
End – End Delay (Sec.)
Mobility
(2m/s)
900
DSRàAODVàDYMOà
0
4
3
2
1
700
500300
100
Routing Load
Mobility
(2m/s)
900
70
120
110
100
90
80
700
500300
100
Pause Time
Throughput (kbps)
Mobility
(2m/s)
900
0
10
7.5
5
2.5
70
5030
10
Routing load
90
15
12.5
0
100
75
50
25
70
5030
10
Throughput (kbps)
90
150
125
0
0.4
0.3
0.2
0.1
70
5030
10
Number of Nodes
End – End Dealy (Sec)
90
0.6
0.5
Fig. 5. Simulation Results of Reactive Protocols: AODV, DSR, DYMO
IV. SIMULATION RESULTS AND DISCUSSIONS
We use NS-2 as our simulation tool. AODV [15] coding
was developed by CMU/MONARCH group while it was
optimized by Samir Das and Mahesh Marina (University of
Cincinnati). Coding of DYMOUM by MASIMUM [17] is used
for DYMO. We use NS-2.34 for simulating AODV and DSR
while, DYMOUM is simulated in NS-2.29. We focus on the
mobility and scalability factors of Ad Hoc networks in our
work.
We considered a network of 50 nodes where nodes are ran-
domly located and are mobile. These nodes have a bandwidth
of 2Mbps each. Mobility is set as 2m/s which is average
walking speed. Packet size is defined as 512 bytes, while
simulation setup runs on Continues Bit Rates (CBR). The size
of network is defined as 1000 m2. Given these parameters, we
have confined our experiments to following three metrics.
1. Throughput
2. End to End Delay
3. Normalized Routing Load.
A. Throughput of Reactive Protocols
In general sense, throughput refers to the amount of data
that has successfully reached its destination. Mathematically
it can be stated as:
T hrougpu t =messagesRecievedS uccessf ully
T ime (20)
Mobility Factor: Considering graphs for throughput, DSR
attains the maximum throughput with respect to AODV
and DYMO. If we consider AODV, than it, surely have a
T I ME OU T factor involved. AODV waits for a specified
time, then route is termed invalid and finally erased from
routing table. “HELLO” messages (used for link monitoring)
in AODV also works very well for mobile environment.
Overall considering mobility factor, DSR gives stable
throughput, as no unnecessary packets are generated by this
routing protocol. In link breakages, DSR have multiple routes
while, in AODV, routing table keeps the best chosen path
only. Hence, within the environment where links are immune
to breaks, DSR supersedes AODV and DYMO. DYMO
proves to be the worst amongst the other two protocols.
Scalability Factor: According to experiments performed,
AODV converges at almost all data rates with salabilities.
While DSR proves itself to be scalable but only during high
data traffic, it can not converge the network. DYMO performs
worst among these studied routing protocols. As the number
of nodes increases or data traffic increases, its performance de-
grades dramatically. According to [14], a network of multiple
thousands of nodes with different traffic loads can be handled
by AODV. The reason that AODV supersedes DSR and DYMO
is lower packet loss ratio and propagation of information
regarding distant vector which practically consume minimum
bandwidth. This feature gives AODV a room for scalability.
In AODV, routing packet contains only one hop information
while in DSR, packet size is larger as it keeps the information
of whole route. This is another reason that AODV outperforms
DSR.
1) End to End Delay of Reactive Routing: Time which a
packet takes in reaching destined node from the originator
node can be termed as end to end delay. Mathematically we
can express it as:
ED =(N umbero f T ransmittedP ack ets)(RT T )
N umberof RecievedP ackets
Mobility Factor: As shown in the graphs, AODV gives
lowest performance as, link breakages may lead to longer
routes. DYMO, though works worst in throughput case but
here it works best amongst DSR and AODV. It is so because,
DYMO does not check the routes in memory as DSR looks
into route cache and AODV in to its routing table, instead it
starts Expanding Ring Search (ERS)algorithm whenever a
route is required.
Scalability Factor: The concept of gratitous RREP is
used both in DSR and AODV. This is the reason that DYMO
results in lowest End to End delay, irrelevant of number of
nodes in the network. Gratitous RREP though results in
lower delay at normal traffic rates though, DSR checks the
route cache before starting Expanding Ring Search (ERS)
algorithm in the same way as AODV search route in its
routing table before starting a route request using ERS
Algorithm. DYMO does not use such stored information
TABLE I
COMPARISON: REACTIVE ROUTING PROTOCOLS
Feature AODV DSR DYMO
Protocol
type Distance
Vector Source rout-
ing Source rout-
ing
Route main-
tained in Routing
table Route Cache Routing
table
Multiple
route
discovery
No Yes No
Update des-
tination Source Source Source
Broadcast Full Full Full
Reuse of
routing
information
No Yes No
Route selec-
tion Only
searched
route
Hop count Only
searched
route
Route recon-
figuration Erase route
notify source Erase route
notify source Erase route
notify source
Route
discovery
packets
using RREQ
and RREP
packets
using RREQ
and RREP
packets
using RREQ
and RREP
packets
Limiting
overhead,
collision
avoidance,
network
congestion
Expanding
Ring Search
Algorithm
Expanding
Ring Search
Algorithm
Expanding
Ring Search
Algorithm
Limiting
overhead,
collision
avoidance,
network
congestion
Binary
Exponential
Back off
Time
Binary
Exponential
Back off
Time
Binary
Exponential
Back off
Time
Update
information By RERR
message By RERR
message By RERR
message
rather it simply initiates ERS. AODV also have a link repair
feature that makes it bear the highest end to end delay with
respect to any scalability among DYMO and DSR.
2) Routing Load of Reactive Routing: When a single
data packet is to be sent from one node to another within a
network, a number of routing packets are involved in sending
this data packet. The numbers of these routing packets which
are sent just to transfer one data packet are termed as Routing
Load or Normalized Routing Load. Mathematically, we can
state:
RoutingLoad = (Routing +DataLoad)
(N umberof DataP acketssent)
Mobility Factor: AODV and DSR use the concept of grat.
RREP, i.e. when a RREQ reaches any node that has a valid
route stored in its route cache or routing table, it generates a
RREP by itself to the original source node. This RREP
contains the full information up to the destined node and
overhead of finding route beyond that node limits. DYMO does
not use this grat. RREP. That’s why it suffers from greater
routing overhead with respect to the other two protocols.
AODV also works well in the context of normalized routing
overhead however, there is a concept of local link repair and
above all, use of HELLO message for link monitoring, makes
it performance lower then DSR. A node with underlying DSR
protocol use promiscuous mode and this is the reason that it
bears lowest overhead.
A common observation with respect to increase in mobility
of nodes in the network is that all the three routing protocols
bear gradually higher overhead. The reason is propagation of
route error packets. As the mobility increases, chances of link
breaks also increase in the same proportion which results in
increase of routing overhead.
Scalability Factor: Routing overhead of DYMO is lower
than that of AODV and DSR. AODV bears high routing
overhead in dense networks. Periodic link sensing packets
involved in local link repair mechanism and grat.RREP
results in high routing overhead. Whereas promiscuous mode
utilized by DSR reduces the routing overhead in not so dense
environment.
B. Discussion
The protocol that uses minimum resources of bandwidth by
its control packets can provide better data flow. Hence, the
environments where traffic load is very high, protocols having
low routing overhead survive. If we consider scalability, than
AODV stands at top of rest of studied routing protocols. It uses
distance vector distribution that minimize network resource
consumption. The network underlying AODV protocol bears
low routing overhead as control packets of AODV contains a
very small part of information in them where as if we compare
it with DSR, control packet of DSR carries whole routing
information in it. Hence we can say that DSR has higher
routing overhead in terms of bytes or size. If we consider
number of control packets than DSR broadcast less number of
packets than that of AODV. AODV use periodic hello packet
for link sensing and also bear local repair routing overhead.
Hence if we compare both of these routing protocols (AODV
and DSR) considering mobility and speed factors, we can
conclude that both of these protocols give more or less same
performance.
Concluding all the routing protocols, our study suggest that,
AODV can be selected for denser environments where lower
routing overhead is required, DSR should be used within a
network having limited number of hops but it is better for
highly mobile environment. DYMO routing protocol can be
used in networks where delay is in tolerable. As like other
reactive protocols, DYMO does not look for any stored route
as DSR looks into its cache and AODV in its routing table.
It initializes binary exponential back off and ERS algorithm
immediately.
V. CONCLUSION
In our work, we have discussed and presented generalized
routing overhead framework of reactive routing protocols in
WMhNs. We analyze route discovery overhead plus route
monitoring overhead. Furthermore, we presented framework
with variation in number of nodes of a network, number
of hops of a network, route life time of a route, periodic
update interval time and frequency of trigger updates to better
understand the behavior and functionality of routing protocol.
We have discussed overhead due to RREQ packets, RREP
packets and link monitoring. In future, we will analyze the
routing overhead due to link repair processes as well. Later
experiments are conducted on three routing protocols keeping
parameters of throughput, End to End delay and routing
overhead in emphasis with respect to mobility and scalability
factors. These experiments shows that AODV performs better
in highly scalable environment where as DSR works almost in
same fashion as of AODV in mobility scenario. Though DSR
performs better if there are less number of hops with in the
network.
REFERENCES
[1] Ph.D. Thesis of Nadeem Javaid, Analysis and De-
sign of Link Metrics for Quality Routing in Wireless
Multi-hop Networks”, University of Paris-Est, 2010.
http://hal.archives-ouvertes.fr/docs/00/58/77/65/PDF/TH2010PEST1028
complete
[2] I. Chakeres, C. Perkins, “Dynamic MANET On-demand (DYMO) Rout-
ing draft-ietf-manet-dymo-21”,July 2010.
[3] C. Perkins, E. Belding-Royer, and S. Das, “RFC 3561, Ad-hoc on-demand
distance vector (AODV) routing”, July 2003.
[4] D. B. Johnson and D. A. Maltz, “RFC 4728, The Dynamic Source Routing
Protocol (DSR) for Mobile Ad Hoc Networks for IPv4” Feb 2007.
[5] Mohammad Naserian, Kemal E. Tepe, Mohammed Tarique, “Routing
Overhead Analysis for Reactive Routing Protocols in Wireless Ad Hoc
Networks”, Wireless And Mobile Computing, Networking And Commu-
nications, 2005. (WiMob’2005), IEEE International Conference on, vol.3,
no., pp. 87- 92 Vol. 3, 22-24 Aug. 2005.
[6] Hui Xu, et.al,“ A Unified Analysis of Routing Protocols in MANETs”,
IEEE Transactions on Communications, VOL. 58, NO. 3, March 2010.
[7] Nianjun Zhou, et.al.,“The Impact of Traffic Patterns on the Overhead
of Reactive Routing Protocols”, IEEE Journal on Selected Areas in
Communications, VOL. 23, NO. 3, March 2005
[8] Jacquet, P. and Viennot, L., “Overhead in mobile ad-hoc network
protocols”, RAPPORT DE RECHERCHE-INSTITUT NATIONAL DE
RECHERCHE EN INFORMATIQUE ET EN AUTOMATIQUE, 2000.
[9] I.D. Aron and S. Gupta, “Analytical Comparison of Local and End-to-
End Error Recovery in Reactive Routing Protocols for Mobile Ad Hoc
Networks”, Proc. ACM Workshop Modeling, Analysis, and Simulation
of Wireless and Mobile Systems (MSWiM), 2000.
[10] N. Zhou and A.A. Abouzeid,“ Routing in ad hoc networks: A theoretical
framework with practical implications”,INFOCOM 2005, pages 1240-
1251, Miami, 2005.
[11] Javaid. N, et al., “Modeling Routing Overhead Generated by Wireless
Reactive Routing Protocols”, 17th IEEE APCC 2011.
[12] Szw-Yao Ni et al, “The Broadcast Storm Problem in a Mobile Ad Hoc
Networks”, in Proc. of the MobiCom 99, pp. 151-163.
[13] Ming Zhao, Yujin Li and Wenye Wang, “Modeling and Analytical Study
of Link Properties in Multihop Wireless Networks”, IEEE Transections
on Communications, Vol. 60, NO. 2, February 2012.
[14] Atsushi Iwata, Ching-Chuan Chiang, Guangyu Pei, Mario Gerla, and
Tsu wei Chen, Scalable routing strategies for ad hoc wireless networks”
Tech. Rep., Department of Computer Science University of California,
Los Angeles, 1999.
[15] C. M. University, “AODV Copyright (c) 1997, 1998 Carnegie Mellon
University, http://www.cmu.edu/”, 1998.
[16] CMU/Monarch, “DSR Copyright (C) 2000 by the University of Southern
California. http://www.usc.edu/”, 2000
[17] University of Murcia,“MASIMUM (MANET Simulation and Implemen-
tation at the University of Murcia)”, 2009.
... With the increasing number of overhead in the network, the network resources will be used up and causing severe unfairness in the network [7,16]. As shown in Figure 8, there are no significant differences in the fairness index trend between AODV and DSDV routing protocols. ...
... The researchers in [7] have presented a routing overhead of the reactive routing protocol study based on Wireless Multihop Networks (WMhNs). The network performance has been improved by implementing routing overhead monitoring methods to Dynamic Source Routing (DSR), AODV, and Dynamic MANET On-demand (DYMO) routing protocol. ...
Article
Full-text available
In the past few decades, sensed data such as wind, vibration, temperature, and humidity were sent through the costly wire cable. With recent advancement and innovation in the wireless field, wireless sensor network (WSN) has emerged to benefit many industries in different aspects primarily in term of cost. WSN has been widely used in various applications due to its wide coverage area, particularly in oil and gas facility monitoring. Hence, the usage of grid WSN fits the wide coverage area mainly in the pipeline of downstream of oil and gas process. The main objective of this paper is to give a better understanding of the impact of different packet sizes, routing protocols, and the number of nodes towards a grid WSN as well as to provide open research issues in this field. Two types of routing protocols have been presented in this paper; reactive (AODV) and proactive (DSDV) with different packet sizes using a grid node arrangement. The simulated results have shown that with the increasing size of the network, more performance degradation occurred regardless of the packet size used. This issue can be seen through the packet loss and throughput loss in the network prominently after the deployment of 80 nodes and above. Apart from that, DSDV has shown a better delivery ratio, less number of passive nodes existence and less routing overhead produced in the network as compared to the AODV routing protocol.
... MANETs consist of wireless hosts that communicate with each other in the absence of a fixed infrastructure. They are used in disaster relief, conference and battlefield environments, and received significant attention in recent years [11], [12], [13]. The positionbased routing protocols are the type that use node location information instead of linking information to routing. ...
... Each mobile host becomes a potential router, and it is possible to dynamically establish routes between them, as well as nodes to an existing route. The network topology changes rapidly and randomly at unpredictable times [13], [14]. In multihop capability, there must be a router in each node. ...
Article
Full-text available
The goal of this paper is to show the differences between ad hoc networks and mobile ad hoc networks (MANETs). The paper will show the routing in ad hoc networks and compare it to that of MANETs, after providing a brief introduction to MANETs. In addition, the types of routing protocols will be described as proactive and reactive protocols, where proactive routing attempts to maintain optimal routes to all destinations at all times, whether it is needed or not. In contrast, reactive protocols determine routes to given destinations only when there is data to be sent to those destinations. The paper will also provide a summary comparing uni-path and multi-path routing, by considering different parameters. Multi-path routing often has advantages more than uni-path routing, and Multi-path routing can also result in packet reordering. However, with uni-path routing, traffic allocation is not an issue since only one route is used.
... [27] Gives a multiple quality of service selection mechanism considering Ad-Hoc networks. A detailed framework of route discovery and route maintenance is produced by [33]. Authors give a generalized model for reactive control overhead. ...
Article
To ensure seamless communication in wireless multi-hop networks, certain classes of routing protocols are defined. This vary paper, is based upon proactive routing protocols for Wireless multihop networks. Initially, we discuss Destination Sequence Distance Vector (DSDV), Fish-eye State Routing (FSR) and Optimized Link State Routing (OLSR), precisely followed by mathematical frame work of control overhead regarding proactive natured routing protocols. Finally, extensive simulations are done using NS 2 respecting above mentioned routing protocols covering mobility and scalability issues. Said protocols are compared under mobile and dense environments to conclude our performance analysis.
... However, DYMO though has a bit less throughput however, there is no delay. Table 3 discusses the different mobility sceneries and categorize these three routing protocols as best, average and worst with respect to throughput, delay and routing load [32]. ...
Article
Reactive routing protocols are gaining popularity due to their event driven nature day by day. In this vary paper, reactive routing is studied precisely. Route request, route reply and route maintenance phases are modeled with respect to control overhead. Control overhead varies with respect to change in various parameters. Our model calculates these variations as well. Besides modeling, we chose three most favored reactive routing protocols as Ad-Hoc on Demand Distance Vector (AODV), Dynamic Source Routing (DSR) and Dynamic MANET on Demand (DYMO) for our experiments. We simulated these protocols using ns-2 for a detailed comparison and performance analysis with respect to mobility and scalability issues keeping metrics of throughput, route delay and control over head. Their performances and comparisons are extensively presented in last part of our work.
Article
Multi-hop cellular network (MCN) is a wireless communication architecture that combines the benefits of conventional single-hop cellular networks and multi-hop ad hoc relaying networks. The route selection in MCN depends on the availability of intermediate nodes and their neighborhood connectivity. Cognitive radio (CR) is an emerging communication paradigm that exploits the available radio frequencies opportunistically for the effective utilization of the radio frequency spectrum. The incorporation of CR and mobile ad hoc network routing protocols in MCN could potentially improve the spectrum utilization and the routing performance of MCN. This paper firstly presents the proposed model for the multi-interface CR mobile node with transceiver synchronization and then investigates its opportunistic spectrum utilization and routing performance in MCN.
Article
Full-text available
The present note proposes a survey of protocol overheads in mobile ad-hoc networks. An analysis is proposed to estimate overhead due to control packets. An analysis and simulations are proposed to estimate overhead due to non-optimality of the routes constructed by some protocols.
Article
Full-text available
An ad hoc network is a collection of wireless mobile nodes dynamically forming a temporary network without the use of any existing network infrastructure or centralized administration. Routing protocols used inside ad hoc networks must be prepared to automatically adjust to an environment that can vary between the extremes of high mobility with low bandwidth, and low mobility with high bandwidth. Over the past 5 years, I have developed a routing protocol called Dynamic Source Routing (DSR) that shows extremely good performance in the environment of an ad hoc network. As a result of its unique design, the protocol adapts quickly to routing changes when host movement is frequent, yet requires little or no overhead during periods in which hosts move less frequently. My thesis will describe the DSR protocol, and the trade-offs encountered in its design. By presenting a detailed analysis of DSR's behavior in a variety of situations, I will generalize the lessons learned from DSR so they can be directly applied to the many other new routing protocols that have adopted the basic DSR framework.
Article
Full-text available
This paper presents a mathematical framework for the evaluation of the performance of proactive and reactive routing protocols in mobile ad hoc networks (MANETs). This unified framework provides a parametric view of protocol performance, which in turn provides a deeper insight into protocol operations and reveals the compounding and interacting effects of protocol logic and network parameters. The parametric model comes from a combinatorial model, where the routing logic is synthesized along with the characterization of MAC performance. Each wireless node is seen independently as a two-customer queue without priority, where the two types of customers are unicast and broadcast packets. The model captures the essential behavior and scalability limits in network size of both classes of routing protocols, and provides valuable guidance on the performance of reactive or proactive routing protocols under various network configurations and mobility conditions. The analytical results obtained with the proposed model are in close agreement with simulation results obtained from discrete-event Qualnet simulations.
Article
The radio link between a pair of wireless nodes is determined by radio channels, transmission range, node mobility and node-pair distance, which form a set of random factors in multihop wireless networks. The properties of such radio links can be characterized by link lifetime, residual link lifetime and link change rate, which, in fact, have been widely used for network design and performance evaluation. In this paper, we take a new modeling approach that captures the dynamics of radio channels and node movements in small-scale. More specifically, distance transition probability matrix is designed in order to describe the joint effects of dynamic transmission range due to radio channel fading and relative distance of a node-pair resulting from random movements. We find that the PDF of link lifetime can be approximated by an exponential distribution with parameter characterized by the ratio of average node speed bar{V} to effective transmission range R_e. To further understand the implication of link properties, analytical results are used to investigate the upper bound of network connectivity and the associated network performance is evaluated by extensive simulations.
Conference Paper
In this paper we investigate the effect of local error recovery vs. end-to-end error recovery in reactive protocols. For this purpose, we analyze and compare the performance of two protocols: the Dynamic Source Routing protocol (DSR[2]), which does end-to-end error recovery when a route fails and the Witness Aided Routing protocol (WAR[1]), which uses local correction mechanisms to recover from route failures. We show that the performance of DSR degrades extremely fast as the route length increases (that is, DSR is not scalable), while WAR maintains both low latency and low resource consumption regardless of the route length. Keywords: Mobile Ad Hoc Network, Routing Protocol, Performance Analysis 1