OPTIMUM NEIGHBORS FOR RESOURCECONSTRAINED MOBILE AD HOC NETWORKS

International Journal of Ad hoc, Sensor & Ubiquitous Computing (IJASUC) Vol.12, No.2, April 2021
DOI : 10.5121/ijasuc.2021.12201 1
OPTIMUM NEIGHBORS FOR RESOURCE-
CONSTRAINED MOBILE AD HOC NETWORKS
Mohammed Tarique1
and Rumana Islam2
1
Department of Electrical Engineering, University of Science and Technology of Fujairah
(USTF), P.O. Box 2202, Fujairah, United Arab Emirates
2
Department of Electrical and Computer Engineering, University of Windsor, 401 Sunset
Avenue, ON N9B 3P4, Canada
ABSTRACT
This paper presents an investigation on the optimum number of neighbors for mobile ad hoc networks
(MANETs). The MANETs are self-configuring and self-organizing networks. In such a network, energy-
constrained mobile nodes share limited bandwidth to send their packets to the destinations. The mobile nodes
have a limited transmission range and they rely on their neighbors to deliver their packets. Hence, the mobile
nodes must be associated with the required (i.e., optimum) number of neighbors. As the number of neighbors
is varied, a trade-off exists between the network connectivity and available bandwidth per mobile node. To
investigate this issue, we consider Dynamic Source Routing (DSR) as the routing protocol and IEEE 802.11
as the MAC layer protocol in this work. We consider both static and dynamic scenarios in this work. We
simulated the ad hoc networks via network simulator (NS-2) and the simulation results show that there exists
an optimum number of neighbors for the static case. We also show that mobility has a grave impact on the
performance of the MANETs in terms of network throughput, end-to-end delay, energy consumption, and
packet loss. Hence, we need to increase the number of neighbors under mobility conditions. However, there
is no global optimum number of neighbors for the mobility case.
KEYWORDS
MANETs, Wireless, MAC, Protocol, End-to-end Delay, Energy Consumption, Routing Overhead,
Throughput
1. INTRODUCTION
Wireless networking has been an active research focus since the early days of the packet radio
network introduced by the Defense Advanced Research Project Agency (DARPA) [1]. Recent
developments in wireless devices have made laptop computers, personal digital assistants (PDA),
pagers, and cellular telephones portable. Now, users can carry these devices to any place at any
time. Hence, there is a need for a network that can be deployed at any place at any time without
any infrastructure support. In some cases, an infrastructure-based network is hard to build.
Networks used by the soldiers on the battlefield are worthwhile to mention here. In some cases,
infrastructures may not exist due to natural calamities such as cyclones, tsunami, and tornados.
Hence, there is always a need for setting up a temporary network among a group of users without
any pre-existing infrastructure and centralized administration. Mobile Ad hoc Networks
(MANETs) are considered suitable solutions for these kinds of temporary networks. MANETs
consist of a group of mobile nodes, which have limited battery and limited processing power.
MANETs are self-organizing and self-configuring networks and they can be deployed without any
infrastructure support. Numerous groundbreaking applications have been proposed based on
MANETs. These applications include disaster management, search and recovery, remote

2
healthcare, tele-geoprocessing, education, traffic management, process control, and security [2].
These applications impose diversified design and performance constraints on the MANETs.
However, MANETs have many limitations too. They have many unique characteristics that are not
present in wired networks. Dynamic topology is one of them. Mobile nodes can join and leave the
network at any time. Hence, route ‘breakage’ is a very frequent phenomenon in the MANETs. The
wireless is the communication medium for the MANETs. The wireless medium is always
unreliable and random due to high interference and noise. The channel randomness can cause a
mobile node to receive a packet at a signal level that is below a minimum threshold level. Hence,
a mobile node may fail to detect a packet that it receives from other mobile nodes. In MANETs,
mobile nodes share wireless channel among themselves. Hence, each mobile node can enjoy only
a very limited bandwidth. Another major limitation of the MANETs is that the mobile nodes have
a very limited transmission range. Hence, they communicate with each other in a multi-hop fashion.
It means that they deliver packets to their destinations through other mobile nodes. Hence, network
connectivity is a very challenging issue in MANETs. The mobile nodes must be associated with a
required number of neighbors. However, the random movement of the neighbors worsens the
network-partitioning problem.
Routing protocol is the most important element of MANETs that provides them with self-
organizing and self-configuring capabilities [3]. Researchers have proposed many routing
protocols for the MANETs. Broadly, these routing protocols can be classified as proactive and
reactive. In proactive routing protocols, like Destination Sequence Distance Vector (DSDV) [3],
mobile nodes periodically exchange routing information among themselves. This kind of periodic
routing information exchange among nodes generates a huge number of overhead packets in the
network. Hence, the proactive routing protocols are not considered suitable for the MANETs.
Reactive routing protocols like Ad hoc On-demand Distance Vector (AODV) [4] and Dynamic
Source Routing (DSR) [5] work on-demand. It means that a route is discovered when it is required.
Hence, reactive routing protocols generate fewer overhead packets compared to proactive routing
protocols. However, the proactive routing protocols use a global search procedure called ’flooding’
to discover routes to the destinations and the neighbors of a mobile node helps them to find these
routes. This ‘flooding’ procedure also generates a huge number of overhead packets in the network,
especially when the network size is large.
To ensure efficient routing operation, mobile nodes must be associated with enough neighbors. In
a lightly populated network, a mobile node may not find the required neighbors to assist in the
route discovery process. Hence, network partitioning will occur. On the other hand, in a densely
populated network, a mobile node finds enough neighbors to assist them during the route discovery
operation. However, as the number of neighbors increases, per-node throughput decreases because
more nodes contend with each other to access the medium. Hence, each node will be associated
with an optimum number of neighbors. We investigate this issue in this paper. We choose DSR
protocol as the routing protocol and Carrier Sense Multiple Access with Collision Avoidance
(CSMA/CA) (i.e., IEEE 802.11) [6] as the underlying medium access protocol.
The rest of the paper is organized as follows. Some related works are presented in section 2. A brief
description of the DSR protocol has been presented in section 3 to provide the readers with enough
background to understand the importance of the optimum number of neighbors. The need for
optimum node density is explained in section 4. The simulation results are presented in section 5.
The paper is concluded with section 6.

3
2. RELATED WORKS
Optimum node density has been an active research area for a long time. One of the initial well-
known works on this topic can be found in [7]. The paper presents an analytical model that shows
that there is a trade-off between the transmission range and available bandwidth per node in the
packet radio network. The work shows that the optimum number of neighbors for a given node
must be six. The major limitation of the work is that it considers only static networks. One of the
earliest works that investigate the optimum number of neighbors under mobility conditions is
presented in [8]. The authors introduce the critical transmission range in ad hoc networks. The
authors define the critical transmission range as the transmission range that is good enough to
maintain the connectivity in ad hoc networks.
Transmission power adjustment, to maintain connectivity and to conserve energy, has been
addressed in [9]. The authors suggest that mobile nodes should adjust transmission power
depending on the network topology and hence transmission power adjustment must be adaptive. A
cluster-based power management technique for ad hoc networks has been presented in [10]. In this
work, mobile nodes form a group called a cluster and they adjust transmission power dynamically
so that they can reach other mobile nodes located at the farthest distance in the cluster. The authors
show that this type of power adjustment saves energy of the mobile nodes and increases network
throughput.
In [11], the optimal transmission range of a mobile node is determined by the progress of a packet
in a desired direction. The authors choose slotted ALOHA [12] as the multiple access technique.
They conclude that mobile nodes should be associated with eight neighbors to maximize network
throughput. However, they did not consider other important performance factors including energy
consumption, end-to-end delay, and routing overhead in their work.
In a similar work [13], the mobile nodes are allowed to adjust their transmission range
independently. The authors conclude that mobile nodes should use the lowest possible transmission
power to maximize the network throughput. Another similar work, presented in [14], shows
appropriate transmission range can conserve energy in ad hoc networks. The authors present a
position-based technique to determine the transmission range of the mobile node.
The optimum transmission range, for a low mobility and high node density network, has been
presented in [15]. The authors argue that optimum transmission range must be considered as a
system design issue. The authors suggest that the optimum transmission range should depend on
the propagation environment as well as the type of radio transceiver. In the work, it is assumed that
mobile nodes are associated with a large number of neighbors that are readily available to forward
the packet in a given direction.
To minimize energy consumption, a bit-meter-per-joule metric has been used in [16]. It is shown
in the work that the energy consumption in a network depends on several factors including
transceiver characteristics, node density, and traffic distribution. A relay architecture has been
presented in the work. According to this architecture, a source chooses a relay neighbor based on
the lowest bit-meter-per-joule metric. In a recent work [17], the authors present a comparative
analysis of three different routing protocols namely DSR, Zone Routing Protocol (ZRP) [18], and
Optimized Link State Routing (OLSR) [19]. In the work, the authors investigate the effects of
transmission range, mobility, and node density on these three routing protocols. They conclude that
the DSR protocol outperforms ZRP and OLSR in terms of delay and packet delivery ratio.

4
The effects of signal fading and shadowing have been investigated in [20]. The authors have shown
that signal fading and shadowing affect the network performances. Hence, the authors determine
the optimal transmission range in the presence of fading and shadowing.
In [21], the authors present a position-based algorithm that minimizes the energy consumption in
ad hoc networks. The authors argue that the optimum transmission range needs to be fixed in a way
so that the energy consumption in the network becomes minimum.
In our work, we also consider the DSR protocol. We consider three network conditions namely (a)
without mobility (i.e., static network), (b) with low mobility (i.e., maximum node velocity of 10
m/sec), and (c) with high mobility (i.e., maximum node velocity of 30 m/sec) to find the optimum
number of neighbors. We also consider the two-ray ground reflection model [22] as the propagation
model and the random waypoint [23] as the mobility model. In contrast to other related works, we
consider four performance parameters namely end-to-end delay, throughput, energy consumption,
and routing overhead to find the optimum number of neighbors, instead of the throughput.
3. THE DSR PROTOCOL
The DSR protocol consists of two basic mechanisms: (1) route discovery and (2) route
maintenance. By route discovery, a mobile node discovers a route to a destination, and by route
maintenance; a mobile node detects a route ‘breakage’.
A source node initiates the route discovery process when it wants to send some data packets to a
destination. It first searches its route cache to find a route to that destination. If a source cannot find
a route in the route cache, it initiates a route discovery mechanism by broadcasting a request packet
to its neighbors. When the neighbors of a source receive the request packet, they first check whether
the request packet is intended for it or not. If a neighbor discovers that it is the destination, it sends
a reply back to the source after copying the accumulated routing information contained in the route
request packet into a route reply packet. Otherwise, it forwards the request to its neighbors. When
a destination receives a route request, it replies to the source through unicast transmission.
When a source node receives a route reply packet, it starts sending data packets using the route
indicated in the reply packet. If multiple paths are discovered, it chooses a path that is the shortest
one. It continues using this path until the path ‘breaks’. When the shortest path breaks, a source
uses the available alternative paths.
A route discovery mechanism is illustrated in Figure 1. Here, the source ‘S’ discovers two routes
to the destination ‘D’ and it uses the shortest route ‘S-E-F-J-D’ to send the data packet. One of the
major limitations of the DSR protocol is that it uses a source routing technique. It means a data
packet carries complete routing information with it. Hence, the packet size is large in the DSR
protocol compared to other protocols. A mobile node needs to spend higher energy for transmitting
this kind of longer packet.
Route maintenance is the mechanism by which a node can detect any change in the network
topology. When a node detects a broken link by using the underlying MAC layer acknowledgment
[24], it removes the link from its route cache. It also creates a special packet called route error
message and sends this route error message to each node that has previously sent the packets over
that broken link. The route maintenance of the DSR protocol is illustrated in Figure 2. In this figure,
the node “F” detects that the link between the node itself and node ‘J’ is broken. The node ‘F’
generates a route error message and sends it to the source. The source marks this route ‘invalid’ in
the route cache and starts using the alternative path (for example ‘S-C-G-K-W-D’) to send the
packet.

5
Figure 1. The route discovery mechanism of DSR protocol.
Figure 2. The route maintenance of DSR protocol
4. NEED FOR OPTIMUM NEIGHBORS
To ensure the correct operation of the routing protocol (like DSR), the mobile nodes must be
associated with the optimum number of neighbors. We will consider the following two cases in
this work namely (i) lightly populated network: where mobile nodes are associated with a very few
number of neighbors, and (ii) densely populated network: where mobile nodes are associated with
a large number of neighbors.
4.1. Low Populated Network
When the mobile nodes are not associated with enough neighbors, the following problems will
occur: (a) node isolation, (b) inefficient path discovery, (c) unsuccessful route maintenance, and
(d) multipath routing problems.
A. Node Isolation
To ensure successful route discovery, a source node needs to find suitable neighbors to spread the
request messages throughout the network and they must reach the destination. Let us consider the
illustration shown in Figure 3. The scenario, shown in Figure 3, is similar to the scenario shown in
Figure 1; however, nodes “B”, “C”, and “E” are outside the range of source “S” or these nodes are
‘dead’ due to battery exhaustion. In this case, source “S” is isolated from the network and hence
it will not be able to discover a path to the destination node “D”. In this case, the source will initiate
route discovery for a certain number of times and it will give up.

6
B. Inefficient path discovery
It is very essential in ad hoc networks that a source should discover an efficient path. Let us assume
that the most efficient path is the one that has the fewest number of hops (i.e., the lowest cost path).
Without enough number of neighbors, a source may discover a path that may not be the optimum
one (i.e., the shortest path). For example, the shortest path discovered by source “S” is “S-E-F-J-
D” as shown in Figure 1 and it has four hops. It is assumed that the neighbors “E” and “C” do not
exist due to node movement or battery exhaustion. In this case, the source discovers the paths “S-
B-H-I-G-K-W-D” or “S-B-H-I-G-F-J-D” as shown in Figure 4. These paths have seven hops. Since
these paths have more hops, it is more likely that a packet travels along these paths will suffer from
larger delay. Routing through these paths will also increase the interference level in the network.
Figure 3. The node isolation problem.
Figure 4. Inefficient route discovery
C. Unsuccessful Route Maintenance
To ensure correct operation of the route maintenance, a source must have more than one path to a
destination. According to the route maintenance algorithm, a source uses alternative paths as soon
as it detects a route breakage. Without enough number of neighbors, a source may fail to find
alternative paths. In this case, the source needs to initiate another route discovery and the network
will be flooded with new route request again. This type of route discovery generates a huge number
of overhead packets in the network and increases the interference level in the network.
D. Multipath Routing Problem
In multipath routing protocols, the mobile nodes must discover multiple paths to a destination and
they select the optimum path or a set of paths among these paths [25]. If mobile nodes are not
associated with enough neighbors, a few paths will be discovered. Hence, multipath routing
protocols may not work properly and the routing protocols may behave like a single path routing
protocol.

7
4.2. Densely Populated Networks
When mobile nodes are associated with a large number of neighbors (more than that are required
to maintain connectivity), the following problems will arise (a) broadcast storm, (b) high
interference, and (c) larger delay.
A. Broadcast Storm
Routing protocols, like DSR, use broadcast techniques during the route discovery phase. In ad hoc
networks, mobile nodes are obligated to forward request message that it receives for the first time.
This forwarding process is done directly or indirectly. In the first case, a single-hop routing will
occur. The packets, originated by a source, are delivered to the destination in a single-hop because
the source and destination are within the range of each other. In the latter case, a multi-hop scenario
will occur, where several intermediate nodes relay the request messages, originated by a source,
before reaching the destination. This process is called “flooding”. In a CSMA/CA based ad hoc
network, “flooding” causes redundant rebroadcasting in the network because all mobile nodes
receive multiple copies of a request [26]. The flooding may cause high ‘contention’ and packet
collisions in the network when many neighbors try to rebroadcast the request messages at the same
time.
B. High Interference
In a densely populated network, the number of mobile nodes participating in packet transmission
increases. Although some underlying MAC protocol like CSMA/CA uses some techniques (i.e.,
exponential back-off) [24] to reduce interference level in the network, mobile nodes still generate
high interference because they are obligated to transmit a packet that they receive from their
neighbors.
C. Larger Delay
In a densely populated network, many mobile nodes exist in a given region and these mobile nodes
contend with each other to get access to the medium. When a mobile node tries to get access to the
medium, it may fail to do so because of the other mobile nodes. Hence, they often go for the back-
off period for a long time, which consequently increases the delay in the network.
5. THE SIMULATION RESULTS
To investigate the optimum number of nodes we conduct simulations by using NS-2.35 network
simulator [27]. In these simulations, we create an ad hoc network consisting of 63 mobile nodes
and distribute them over an area of 3000m by 1000m. In the network, a mobile node’s location is
determined by (x, y) coordinates, where x and y are uniformly distributed random variables in the
range of 0-3000m and 0-1000m respectively. The average number of neighbors is determined by
n =
N
A
πR2
(1)
, where n is the number of neighbors, A is the network area (in sq. m), N is the total mobile nodes
in the network, and R is the transmission range. We set the transmission range to 250m in the
simulations so that each node will be associated with four neighbors under the two-ray propagation
model. Then, we increase the number of nodes to 73, 92, 107, 122, 137, 152, 162, and 182 without
increasing the network area so that the number of neighbors will increase to 5, 6, 7, 8, 9, 10, 11,

8
and 12 respectively. Ten Unigram Data Protocol (UDP) connections are set up during each
simulation. These connections start randomly during the simulation time by using the Constant Bit
Rate (CBR) agent. The source and destination for each connection are chosen randomly. Each
simulation is run by using ten different topologies for a given number of neighbors. The total
simulation time is 250 seconds. The transmission range of a mobile node, is determined based on
the two-ray model [20], is expressed by
Pr = PtGtGr (
hthr
d2 )
2
(2)
, where Pt is the transmission power, Gt is the transmission antenna gain, Gr is the receiving antenna
gain, d is the distance between the transmitter and receiver, and ht is the transmission antenna
height, hris the receiving antenna height. By using the simulation parameter mentioned in Table 1,
the transmission range (i.e., d=250 m) is calculated.
Table 1: Simulation Parameters
Parameters Value
No. of nodes 73,92,107,122, 137, 152, 162, and 182
Node distribution Uniform
Network areas 1000m x 3000m
No. of connections 10
Antenna height (transmitter and
receiver)
1.5 m
Transmission power 0.2818 watts
Received threshold power 3.631 x 10-10
watts
Application Constant Bit Rate (CBR)
Transport layer protocol Unigram Data Protocol (UDP)
Simulation time 250 sec
Propagation model Two-ray ground reflection
Medium Access Control IEEE 802.11
Number of simulations 10
Packet rate 1 packet/sec
Packet size 512 bytes.
The simulation results for the static network (i.e., without mobility) are shown in Figure 5-8. The
network throughput is plotted in Figure 5. The staircase pattern plot shows that the network
throughput is very low for a few neighbors. The reason is that not all sources could find a neighbor
to discover paths because of the low number of neighbors. It is also shown that the throughput
increases with the number of neighbors and it reaches maximum (i.e., 290 kbps) at 10 number of
neighbors.
The end-to-end delay per packet is shown in Figure 6, which shows that the delay slightly increases
with the number of neighbors. However, it increases abruptly when the number of neighbors
exceeds 10. A similar conclusion can be drawn from Figure 7, which shows the packet loss in the
network. This figure shows that the packet loss is low until 10 number of neighbors. Later, it
increases to a very high level with the number of neighbors. Based on the simulation results for the
static network, we can conclude 10 is the optimum number of neighbors for MANETs under static
conditions.

9
Figure 5. The network throughput
Figure 6. The end-to-end delay per packet
Figure 7. The lost packets

10
Figure 8. The energy consumption per packet
Energy consumption is an important issue in the energy-constrained MANETs. To find the
optimum number of neighbors, we also consider energy consumption in the network. We measure
the total energy consumed by all nodes and we divide the total consumed energy by the total packets
delivered to the destination to find the energy consumption per packet, which is plotted in Figure
8. This figure shows that the energy consumption, expressed in Joules/packet increases linearly
with the number of neighbors in the network. This conclusion is contrary to another claim [9] that
the optimum transmission range minimizes energy consumption.
The simulation results for the mobility cases are shown in Figure 9-12. In this investigation, we
consider two cases (a) low mobility (i.e., node velocity 10m/sec) and (b) high mobility (i.e., node
velocity 30m/sec). We use a random waypoint model [18] to generate node movement in the
network. The network throughput is plotted in Figure 9, which shows that the optimum node
density is six for the low mobility case. However, the same is eight for the high mobility case. We
can conclude that the optimum number of neighbors need to increase when mobility increases in
the network. However, comparing Figure 5 and Figure 9, we also conclude that the network
throughput is significantly low for mobility cases compared to the same as the static network.
Figure 9. The network throughput for mobility case

11
Figure 10. The end-to-end delay per packet for mobility cases
The end-to-end delay per packet for the mobility cases is presented in Figure 10. This figure shows
that the end-to-end delay per packet decreases with the number of neighbors. This figure also shows
that the delay-per-packet is always higher for high mobility case. The energy consumed by the data
packet under the mobility case is shown in Figure 11, which depicts energy consumption per packet
increases with the number of neighbors. However, the energy consumed under the high mobility
case is much higher compared to the same in the low mobility case.
Figure 11. The energy consumption for mobility cases

12
Figure 12. The lost packets for mobility cases
Finally, the lost packets are plotted in Figure 12. This figure shows that the lost packet decreases
with the number of neighbors for both mobility cases. This is contrary to the static networks, where
the packet loss increases sharply after the optimum number of neighbors. However, the lost packets
are always higher for high mobility cases compared to its low mobility counterpart. The reason is
that more route breakages occur as the mobility increases in the network.
6. CONCLUSION
In this paper, we investigate the optimum number of neighbors for MANETs. It is shown that the
optimum node density depends on the network conditions. It is shown that the optimum number of
neighbors is 10 for a static MANET. However, the same is very difficult to determine in dynamic
networks. The simulation results show that the number of end-to-end delays is very high in dynamic
networks and hence network throughput is less compared to static networks. The main reason for
this is that there occur more unsuccessful route discovery, more breakages, and unsuccessful data
transmission in dynamic networks compared to their static counterparts. All these ultimately affect
network throughput. It is also shown that the maximum network throughput in mobility cases is
much lower compared to the same in the static case. As the mobility increases, the maximum
network throughput decreases. However, there is no global optimum number of neighbors for
dynamic networks. It is also shown in the paper that increasing the number of neighbors can only
improve the network throughput to some extent, but the network throughput cannot reach the same
level as that in static networks.
In this work, we consider two-ray propagation model in our simulations. However, the two-ray
model may not represent a realistic propagation model for the ad hoc networks that are deployed
in hazardous areas. Hence, a more practical propagation models (like probabilistic models,
shadowing models etc.) need to be considered in the simulations. Another limitation of this paper
is that we consider only two values for the velocity of the mobile nodes. We arbitrarily select these
two values. A further investigation is required to choose a realistic velocity for the mobile nodes.
To find the energy consumption by the mobile nodes, we use the built in energy consumption model
provided with the NS-2.35 network simulator. However, many practical energy consumption
models have been presented in the literature. These models also need to be included in the
simulations for accurately determining the energy consumption by the mobile nodes. These all are
left for the future investigations.

13
REFERENCES
[1] Jubin, J. and Torrow, J. “The DARPA Packet Radio Network Protocol,” Proceedings of the IEEE, vol.
75, no. 1, January 1987, pp. 21-32
[2] Goldsmith A.J. and Wicker S.B., “Design Challenges for Energy-Constrained Ad hoc Wireless
Networks”, IEEE Wireless Communication, vol. 9, no. 4, August 2002, pp. 8-27
[3] Perkins, C.E. and Bhagwat, P., “Highly dynamic destination-sequence distance-vector routing (DSDV)
for mobile computers,” Proceedings of ACM SIGCOMM, London, UK, August 1994, pp. 234-244
[4] Perkins, C.E., “Ad Hoc On-Demand Distance Vector (AODV) Routing,” Internet-Draft, draft-ietf-
manet-aodv-00.txt, November 1997
[5] Broch, J., Johnson, D.B., and Maltz, D.A., “The Dynamic Source Routing Protocol for Mobile Ad Hoc
Networks”, IETF Internet-Draft, draft-ietf-manet-dsr-00.txt March, 1998
[6] Crow, B. P., Widjaja, I., Kim, , J, G. and Sakai, P. T., “IEEE 802.11 Wireless Local Area Network,”
IEEE Communication Magazine, vol. 35, no. 9, September 1997, pp.116-127
[7] Kleinrock, L. and Silvester, J. “Optimum Transmission Radii for Packet Radio Networks or Why six
is a magic number”, Proceedings of the IEEE International Conference on Communication, New
Orleans, LA, June 2000, pp. 4.3.1-4.3.5
[8] Chansez, M., Manzoni, P., and Haas, Z.J., “ Determination of Critical Transmission Range in Ad hoc
networks”, Proceedings of Multi-access, Mobility, and Tele-traffic for Wireless Communications
Conference, Venice, Italy, 2000
[9] Ramanathan, R. and Rosales, H. R., “Topology Control of Multihop Wireless Networks using
Transmission Power Adjustment”, Proceedings of IEEE Conference on Computer Communication, Tel
Aviv, Israel, March 2000
[10] ElBatt,T.A. , Krishnamurthy S.V., Conners, D., and Dao, S. “Power Management for Throughput
Enhancement in Wireless Ad hoc Networks”, Proceedings of the IEEE Conference on Communication,
New Orleans, LA, June 2000, pp. 1503-1513
[11] Takagi, H., and Kleinrock, L. “Optimal transmission ranges for randomly distributed packet radio
terminals,” IEEE Transaction on Communication, vol. COM-32, no. 3, March 1984, pp. 246–257
[12] Farouzan, B.A. “Data Communications and Networking”, Fourth Edition, McGraw-Hill, NY, pp. 364-
370
[13] Hou, T.A. and Li, V. O. K. “Transmission range control in multihop packet radio networks,” IEEE
Transaction on Communication, vol. COM-34, no. 1, January 1986, pp. 38–44
[14] Rodoplu, V. and Meng, T.H., “Minimum energy mobile wireless networks,” IEEE Journals on Selected
Areas of Communications, vol. 17, August 1999, pp. 1333–1344
[15] Chen, P., O’Dea, B., and Callaway, E. “Energy-efficient system design with optimum transmission
range for wireless ad hoc networks,” Proceedings of the IEEE ICC, 2002, pp. 945–952.
[16] Gao,J.L., “Analysis of energy consumption for ad hoc wireless sensor networks using a bit- meter-per-
joule metric,” IPN Progress Report 42-150, August 2002, pp. 189–198.
[17] Kumar, D. and Gupta, S.C., “Transmission Range, Density & Speed based Performance Analysis of
Ad Hoc Networks”, African Journal of Computing and ICT, vol. 8, no. 1, March 2015, pp. 173-178

14
[18] Zygmunt, J. H., Marc, R. P., and Samar, P. “The Zone Routing Protocol for Adhoc networks,” IETF
Draft available at draft-ietf-manet-zone-zrp-04 - The Zone Routing Protocol (ZRP) for Ad Hoc
Networks
[19] Jacque, P. et.al., “Optimized Link State Routing Protocol for ad hoc network,” Proceedings of the
IEEE International Multi-Topic Conference: Technology for the 21st Century, December 30, Lahore,
Pakistan, pp. 62-68
[20] Zorzi, M. and Pupolin, S., “Optimum transmission ranges in multihop packet radio networks in the
presence of fading,” IEEE Transaction on Communication, vol. COM-43, no. 7, July 1995, pp. 2201–
2205
[21] Rodoplu, V., and Meng, T.H., “Minimum energy mobile wireless networks,” IEEE Journals on
Selected Areas of Communications, vol. 17, August 1999, pp. 1333–1344
[22] Rappaport, T.S., “Wireless Communication: Principles and Practices,” Second Edition, Prentice-Hall,
NJ, USA, pp. 120-124
[23] Broch, J., Maltz, D.A, and Johnson, D. Y.-C. Hu, and Jetcheva, J. “A Performance Comparison of
multi-hop Wireless Ad Hoc Network Routing Protocols,” Proceedings of the 4th Annual ACM/IEEE
International Conference on Mobile Computing and Networking (Mobicom), Dallas, Texas, October
1998, pp. 85-97
[24] EEE Standards Department, “Wireless LAN Medium Access Control (MAC) and Physical Layer
(PHY) Specifications”, IEEE Standard 802.11-1997, 1994
[25] Tarique, M., Tepe, K., Abadi,S., and Erfani, S. “Survey of multipath routing protocol for Mobile Ad
hoc Network,” International Journal of Networks and Computer Applications, vol. 32, no. 6, pp. 1125-
1143
[26] Tseng, YC., Ni, S,Y., Chen, Y,S. et al., “The Broadcast Storm Problem in a Mobile Ad Hoc Network.”
Wireless Networks, vol. 8, March 2002, pp. 151–162
[27] The Network Simulator (NS-2) available at https://www.isi.edu/nsnam/ns/

OPTIMUM NEIGHBORS FOR RESOURCECONSTRAINED MOBILE AD HOC NETWORKS

More Related Content

What's hot (13)

Similar to OPTIMUM NEIGHBORS FOR RESOURCECONSTRAINED MOBILE AD HOC NETWORKS (20)

Recently uploaded (20)

OPTIMUM NEIGHBORS FOR RESOURCECONSTRAINED MOBILE AD HOC NETWORKS