Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No.

1, March 2013 1


ISSN 2320 - 5377 | 2013 Bonfring
AbstractWireless Sensor Networks have attracted much
attention in recent years. The potential applications of WSNs
are immense. They are used for sensing, measuring,
collecting, storing, and sharing the sensed data. Wireless
Sensor Networkshave been used for various applications
including habitat monitoring, agriculture, nuclear reactor
control, security and tactical surveillance.In Wireless Sensor
Network the most power hungry part is the communication.
Communication is the most energy conserving factor in
Wireless Sensor Network. Different routing protocols exist on
the basis of network architecture and applications. Wireless
Sensor Network is a dynamic distributed system where sensor
nodes perform the complex tasks collectively. Wireless Sensor
Networkhas many constraints associated with it. These are
limited computational capacity, battery power, and adequate
storage capacity etc.So management of resources is the utmost
important issues in Wireless Sensor Network while computing
the path from source to the base station. Most of the recent
research has focused on the sensor networks where the sensor
nodes are static only. We have focused our work in a dynamic
sensor environment where the base station is mobile to collect
the data from the cluster head. This paper presents a
distributed leveling and clustering based Quality of Service
routing protocol which can increase the network lifetime by
saving some battery power. In the same time,it increases the
overall network performance by decreasing the packet loss
and bit error probability. Also, the proposed protocol
introduces the cognitive radio technology where it utilizes the
unutilized assigned radio spectrum effectively and efficiently
by using the dynamic Time Division Multiple Accesses
technique.
Keywords--- Cognitive Network, Quality of Service, Super
Cluster Head, Wireless Sensor Networks

I. INTRODUCTION
WIRELESS SENSOR NETWORK (WSN) consists of a
large no. of tiny, low power, cheap sensor nodes have
sensing, data processing, and limited computing capabilities.
As sensor nodes usually operate in unattended, harsh

M. Sujeethnanda, MS Student, Communication Research Center, IIITH,
Hyderabad, India.,E-mail: sujeeth@research.iiit.ac.in
PadmalayaNayak, Professor, Dept.of IT, GRIET, Hyderabad, India. E-
mail : padma_nayak@yahoo.com
G. Ramamurthy, Professor, Communication Research Center, IIITH,
Hyderabad, India. E-mail: rammurthy@iiit.ac.in

DOI: 10.9756/BIJNTA.3089
environment they are prone to failure and may run out of the
battery power. The energy consumption can be managed
efficiently by proper selection of low power devices and
proper software programming. Advances in sensor technology
and computer networks have enabled Distributed Sensor
Networks (DSNs) to evolve from small clusters of large
sensors to large swarms of micro-sensors, from fixed sensor
nodes to mobile nodes, from wired communication to wireless
communication, from static network topology to dynamic
network topology [1].
To be effective, a wireless sensor network must implement
some procedures such as neighborhood discovery, self-
organization, localization, routing, fusion etc. From the
literature survey it is concluded that todays wireless networks
utilize only 15%-85% of spectrum of assigned spectrum
indicating that the remaining spectrum is unutilized. So, the
cognitive radio technique utilizes or shares the spectrum in a
more efficient manner. Dynamic TDMA technique allows the
utilization of cognitive radio technology, thus making the
channel more available.
Communication as well as Connectivity is a crucial issue
in the WSN Network. Most of the recent research has focused
on finding a path from source to the base station where the
senor nodes are static. In such environments sensor nodes do
not move after deployment. But there are many applications in
which nodes can be mobile. For instance, agriculture field,
hospitals etc.Rapidly changing topology and frequent link
failure makes the sensor network more challenging.Mobility
and rapid changing topology can cause frequent path breakage
which results packet loss.
In this paper, we have approached a novel routing protocol
based on Clustering, and path planning to find a path from
source to BS. Next the whole sensor network is divided into
two regions such as high interest region and low interest
region. The data can be sent from source node to the Cluster
Head (CH), from cluster head to Super Cluster Head (SCH),
and finally Super Cluster Head delivers the message to the
base station. The proposed algorithm is described in details in
section III.
The rest of the paper is organized as follows, Section II,
discusses the related aspects of Quality of Service (QoS) and
its effecton the performance of the wireless sensor networks.
Section III presents about the related work, Section IV
describes about the proposed algorithm, and Section V
indicates our future work followed by the concluding remark.
QoS Related Parameters
QoS is the ability to provide different performance levels
Quality of Service Routing Algorithm for Cognitive
Wireless Sensor Network

M. Sujeethnanda, PadmalayaNayak and G. Ramamurthy
A
Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 2


ISSN 2320 - 5377 | 2013 Bonfring
to different applications and users or to guarantee certain
performance measures for delay, jitter, throughput, bit error
probability, error rate, and packet dropping ratio etc.QoS is
not only dependent on above mentioned performance
measures but also depends on some non functional parameters.
The non functional parameters include Energy, Reliability,
Robustness, Security, Scalability, Mobility, Cost-
effectiveness, delay etc.The major challenges in
implementingQoS in WSNs are resource constraints such as
energy, bandwidth, and transmission capacity of each sensor
node. Along with the resource constraints, the utilization of
energy in an efficient manner is another major considerable
factor. Moreover, the time varying capacity of wireless links,
limited resources, and node mobility make very difficult (if
not impossible) to maintain correct routing information in ad-
hoc WSNs environment. Thus, providing QoS during path
computation from source to BS is a tremendous challenge in
WSNs.
As WSNs are always deployed in a harsh environment,
recharging/ replacement of battery is highly expensive and
difficult. Usually, QoS gets affected due to the mobility of
sensor nodes as well as dynamic nature of network topology
due to thefrequent link failure and node failure.Various
applications have various requirements. The QoS associated
with such applications differ from each application. For
example, in environmental monitoring case,it is expected
thataccurate data should be received with less redundancy. So
in this case, bandwidth and redundancy are the key parameters
for QoS measures. For inhabitant monitoring, where the nodes
are not easily accessible, battery power of each node is the key
QoS parameter to enhance the network life time.Similarly, for
search and surveillance applications,security plays the major
role. Unlike traditional networks,QoS parameters are not fixed
in WSN applications, but vary with respect to the
requirements of the applications. The various factors which
play a major role to enhance the network performance in WSN
are listed below.
Limited Battery Power
Bandwidth
Data Redundancy
Security
Delay
Mobility
Low Battery Power: Sensor Networks are equipped with
very limited power sources. The replacement of power source
is not possible (even impossible) as these networks are
deployed in inaccessible environments. The power constraints
in sensor networks are much more stringent than other ad hoc
networks. The life time of the whole network depends on the
life time of each sensor node. Normally, the major power
drain takes place in sensor network during transmitting the
packets or forwarding the packets to the sink or BS. In static
wireless sensor networks, the nearby nodes of BS drain fast as
compared to other nodes because these nodes have to transmit
their own messages and also forward all the messages coming
from higher levels. Thus if these nodes die, the entire network
will be disconnected from the sink node/BS. This problem can
be solved by making sink node/ BS mobile. As BSwill be
mobile, the battery can be replaced for the sink node/ BS. In
our proposed protocol, the cluster head and super cluster heads
are chosen randomly among the nodes to distribute the load
evenly to all the nodes in the network. The BS or sink node
will collect the data from such randomly selected super cluster
heads. Thus our proposed protocol ensures the longitivity of
the network.
Bandwidth: The limited licensed frequency band is always
a resource constraint in wireless medium. Each node broadcast
the message into the wireless medium. The lower level nodes
do the same in the network. Due to the limited bandwidth, the
network suffers from congestion and as a result loss of packet
occurs in the network. There must be some efficient
bandwidth management schemes must be implemented to
overcome this problem.
Data Redundancy: Normally, sensor nodes are densely
deployed in a sensor field to sense the required phenomenon.
There is a high probability that more than one node detect or
sense the same phenomenon. Thus, the redundancy of data
increases proportional to the additional number of nodes in the
network. So efficient data fusion and normalization algorithms
are needed to reduce the redundant data in the network. In
turn, this will help to overcome the bandwidth problem too. In
our proposed protocol, we adapt a simple mechanism to avoid
these two problems. The sensor nodes transmit the data to the
selected cluster head and these cluster heads implement simple
data average technique to reduce the redundant data. These
cluster heads will in turn transmit the data to super cluster
heads. These super cluster heads again use data fusion
algorithm to reduce the data. Thus the local fusion and global
fusion helps to avoid redundant data. The data transmitted by
sensor nodes is designated to reach a particular cluster head.
Thus our proposed algorithm saves bandwidth by discarding
the unwanted packets from the other clusters.
Security: In wireless network, the data is transmitted
through wireless links. These wireless links are highly
insecure, and error prone. If these networks are used for
specific applications such as surveillance or tactical
applications, security is a major issue to be considered.
Different types of attacks such as eavesdropping, message
distortion are the common phenomena in wireless sensor
networks. As resources are major constraints in WSN, it is
very hard to provide secure communication in Wireless sensor
networks.
Delay: If sensor networks are deployed to monitor the
hazardous environments, they have to sense and send data
very fast. There should be minimum delay for the data to
reach the sink node or the base station. So delay is one of the
main QoS metric in wireless sensor network. As most of the
times these sensor network are deployed to monitor the
activities in inaccessible and hazardous environments.
Mobility: Mobility of sensor nodes is not mandatory
requirement in sensor networks. When the small sensor nodes
are deployed in a particular field to monitor a phenomenon,
the unwanted mobility caused by natural activities like wind,
animal, human etc affects the performance of the network.
This unwanted mobility changes the network topology
Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 3


ISSN 2320 - 5377 | 2013 Bonfring
dynamically and also may cause path breaks to the sink node
or BS. The delay incurred in re-establishing paths may cause
some packet drops, and thus affects the QoS of the network.

Figure 1: The Parameters Affecting QoS
II. RELATED WORK
There are many studies based on QoS provisioning in
Wireless Sensor Networks [6]-[8], [13].Both the functional
and non functional parameters related to QoS are shown in
Figure1. The QoS parameters are affected mainly due to the
mobility of the sensor nodes and also with limited energy
resources. Some of the recent research works based on energy
efficient MAC protocols are discussed in the literature [9].
Some of the energy efficient packet transmissions scheduling
schemes in the data link layer for WSNs have been proposed
in the literature [2]-[5]. In [2] the author has developed a
round robin scheduling scheme to evenly discharge multiple
battery cells. In [3]a MAC protocol has been proposed which
explores the battery dynamics to increase the network life
time. In [4] a cluster based TDMA protocol has been
proposed. In [5] a TDMA based multiple hop protocol for
large scale networks has been proposed. However, most of
them do not consider the software and hardware integration,
effect of battery dynamics, which has immediate and
significant effects on the life span of sensor nodes.
In this paper, we propose an energy aware routing protocol
which minimizes the energy consumption by proper design of
software and hard ware integration. Specifically our aim is to
minimize the energy consumption at the hardware devices,
which are used at the sensor nodes. These devices are
microprocessors/micro controllers, transceivers, memories,
ADC/DAC etc. This can be achieved by selecting low power
devices and by implementing sleep cycles/modes on the
devices or energy harvesting. Our further aim is to reduce the
power consumption by implementing energy aware routing
protocol using proper data aggregation, distributed processing
and scheduling mechanisms.
III. OUR PROPOSED ALGORITHMS
Our proposed solution tries to solve the most of the above
discussed issues. Thus we are concentrating on the paradigm
where the sensor nodes in the network are stationery and the
sink node or Base station is mobile. The following discussion
shows how the mobile sink node or base station solves the
above discussed problems.
The following are the basic assumptions in our proposed
algorithm.

Assumptions:
1. All the nodes are similar and have equal
computational capabilities.
2. The sensor network is densely deployed.
3. All the nodes are static except the base station i.e
base station is mobile.
4. More than one base station is available to collect the
data from various corner of the sensor field.
5. Cluster heads are 5-10% of the sensor nodes.
6. Super Cluster Heads are 2-5% of the sensor nodes.
The steps involved in proposed Algorithm:
1. Leveling
2. Sectoring
3. Clustering
4. Path planning
Leveling:
The entire sensor field is divided into number of levels.
Leveling can be done by using number of hop counts required
to transmit the data from source to the base station.
Sectoring:
Base station (BS) using a directional antenna, will transmit
the signals with maximum power and divides the sensor field
into equiangular sectors with an angle of (Let =45
0
). So
that each node knows its level ID and sector ID.
Clustering:
Cluster can be formed based on the signal strength and
cluster head will be decided by using round robin
technique[11]. The entire sensor field is divided into levels
and sectors as shown in Figure2.
Path Planning:
There will be many CHs available in an entire sensor field.
When an event occurs at a particular cluster, all the nodes will
send the data to the CH. Then the CHs will forward the packet
to a Super Cluster Head (SCH). In the static BS case, the
Leveling, Clustering, Sectoring of sensor field enables routing,
fusion, and Localization in an efficient manner [12]. Our
proposed protocol aims to includemobility at the BS. It
includes local data fusion technique inside the cluster and
global data fusion technique at the SCH.
Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 4


ISSN 2320 - 5377 | 2013 Bonfring


Figure 2: Sensor Field after Leveling and Clustering
QoS Provisioning with Mobile BS:
As the BS is mobile, the BS will collect the data from SCH
along the path followed by it over the sensor field. Due to the
fact that BS is made mobile and locally collects the data from
SCH, the reliability of reporting data to the BS is increased.
This is due to the fact that relaying data through multiple
wireless links is less reliable than dumping data to the mobile
BS, when it is in close neighborhood of BS. There can be two
trajectories involved with BS. One is proactive monitoring and
another is reactive monitoring. In proactive monitoring the BS
moves throughout the sensor field either with a constant
velocity or it can move with a different velocity. But in case
of reactive monitoring the BS reacts according to the events to
be occurred.
The following figures explain the proposed paths for base
station to cover the entire sensor field. The BS will start at a
particular location on sensor field and travel along thepossible
pathsas shown in Figure 3, 4 and 5given below.

Figure 3: The Spiral Path


Figure 4: The Concentric Circle Path


Figure 5: Star Like Path
If the BS takes the concentric circle path it covers one
level and it will go to the next higher level and so on. So total
distance covered by BS is given in equation (1).
(1)

T
d
= Total distance covered by BS
K = Random variable
r =radius of the circle
n = No. of circular paths

If BSmoves with a uniform velocity u m/s, then time
taken to cover the total field or delay to reach the final point
D
max
can be measured by using equation (2).

(2)

When BS is at the center of the sensor field and moves in a
particular direction, and covers a particular sector with respect





BS
Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 5


ISSN 2320 - 5377 | 2013 Bonfring
to 360
0
angle, then Arc length can be calculated by using
equation (3).

(3)

When BS covers the total field, then total length will be
equal to the radius of the level and Arc length. So, it can be
expressed in mathematical form, asgiven in equation (4).
(4)
A. Adaptive Path Planning
We include adaptive path planning to find a path for the
BS based on the probability of occurrence of a critical event.
The trajectory of BS varies based on any EVENT reported at
any location on the sensor field.

B. Reactive Path Planningbased on Mobile BS:
Distance between two Super Cluster Head (SCH) = d
Assuming 5 SCHs per area of 500 500 mts,
Total distance travelled by BS= 5d mts
Assuming uniform velocity for BS i.e m/sec,
Time taken to cover the region = 5d/ sec.
No of SCHs in that distance travelled will be= nd
Then time taken to travel several area of such size
T
t
= No. of region total distance= knd
Where k= No. of region
C. Proactive path planning based on Mobile BS
When the BS is mobile, the reliability of packet delivery
from SCH to BS will increase. It decreases the packet
dropping ratio. It can be proved by equation (5) given below.
(5)

Where
N = Number of bits
K= Number of error bits
q= Number of hops
Bit error probability varies with respect to the number of
hops.
The following section discusses some of the QoS
parameters which can enhance the network performance w.r.t.
mobile BS.
Battery power: The bottle neck of the static network is the
battery of the nodes near to Sink node or Base station will
drain fast due to high traffic i.e. the nodes near to sink
node has to send their own messages as well as forward
all the messages from higher levels to the sink node thus
reducing the lifetime of the network. This problem is
solved by making sink node or Base station mobile. As
Base station is mobile, we can replace the battery on the
sink node or Base station. In our proposed network
architecture the cluster head and super cluster heads are
chosen randomly among the nodes to distribute the load
evenly on to all nodes in the network. The base station or
sink node will collect the data from such randomly
selected super cluster heads. Thus ensures the longevity
of the network.
Bandwidth and Redundancy:Due to the broadcasting
nature of the transmitting medium the bandwidth problem
and redundant data takes place. We propose a simple
mechanism to avoid these two problems. The sensor
nodes transmit the data to the selected cluster head and
these cluster heads implement simple data average
technique to reduce the redundant data. These cluster
heads will in turn transmit the data super cluster heads.
These super cluster heads again use data fusion algorithm
to reduce the data. Thus the local fusion and global fusion
helps in avoiding redundant data. The data transmitted by
sensor nodes is designated to reach a particular cluster
head. The other cluster heads nearby may receive the
message but it will discard the message because it not
designated to it. Thus we save bandwidth by discarding
the unwanted packets from the other clusters.
Delay: As the sensor networks are deployed for military
surveillance and hazardous environment monitoring, the
delay should be very minimum from the source to
destination. In our proposed architecture the sink node or
Base station is mobile. There is a limit on minimum
velocity and maximum velocity assigned for the sink node
or BS. The BS or sink node moves with a certain velocity
cover the entire sensor field and again repeats the
procedure. If the area of the sensor field is too large,
multiple sink nodes are deployed in the sensor field to
cover a particular region as shown in Figure 6.

Figure 6: Huge Sensor Field with Multiple Coverage Regions
These multiple sink nodes collect the data from the
respective regions and will transfer the data to the base station.
Security: As discussed above the error rate and message
distortion will increase with the number of nodes in the
network and also with the hop count. In the proposed
architecture the data is transmitted to its respective cluster

Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 6


ISSN 2320 - 5377 | 2013 Bonfring
head and from cluster head to the respective super cluster
head. We are limiting the message hop count to two hops
only there by reducing the error rate and distortion. This
can be proved by the above equation 5.
Thus with the help of our proposed algorithm we are
providing QoS in-terms of longevity of the network by
reducing battery consumption at various stages and also by
implementing local and global data fusion algorithms in-terms
of data redundancy and bandwidth of the network.
IV. PRACTICAL APPLICATIONS WITH MOBILE BS
Recent studies conclude that there are tremendous interest
have been increased for wireless sensor network applications
in various fields. Starting from military applications to
medical applications, everywhere sensor network application
is taking a huge space in day today life. Our proposed protocol
mainly aims at the mobile base station to collect the data in
order to reduce the delay and to enhance the network life time
by consuming less energy. The focus is to find its application
in different areas like precision agricultures and healthcare
centers which are described in the next paragraph
For instance, in healthcare applications all the patient
terminals can be considered as staticsensor nodes equipped
with various smart sensors to measure the vital parameters of
the patients. The doctor terminal can be considered as mobile
sink node or base station. In this paradigm, whenthe BS is
mobile such as the doctor moves from its original location to
the ward to check the patients health condition, the doctor
terminal willcollect the information from the patients by using
Wireless Body Area Networking (WBAN) and the doctor
terminal can be updated with the new available data.
Similarly, in precision agriculture applications the similar
idea can be employed to monitor the soil parameters and to
check the soil conditions. The entire sensor field is divided
into small regions and can be equipped with various sensors to
get the data from the entire sensor field. If the farm is too huge
or farm has more than one variety of crops, then each crop or
each small region may have different values/parameters to be
measured. In this scenario, multiple mobile sink nodes or Base
stations can be deployed to cover the entire region to collect
the data from respective fields. Then the collected data by
these mobile sink nodes can be dumped onto one Base station.
V. SIMULATIONAND RESULTS
The entire sensor network has been simulated using
Network Simulatorversion 2.34 (NS2). The sensor network
w.r.t Static Base station has been simulated using IEEE
802.15.4 Mac and WirelessPhy. The simulation has been
carried out on 500 nodes in the area of 1000 sq meters (1000 x
1000). Energy Model has been used for tracing the energy
levels of the nodes before and after the simulation. Each
simulation has been carried out for 100 seconds and it has
been run for 20 times. Thus total simulation time to analysis
the result took 2000 seconds. The results have been analyzed
and printed by using plotting tool called XGRAPH. The
simulation environment and the results have been illustrated in
the figures 7, 8, 9, and 10 which are given below.


Figure 7: Simulation Environment with BS at the Center
Figure 7 explains about the simulation environment with
static BS at the center. The figure 8 and 9 shows the energy
levels of the nearest node and farthest node from the sink node
respectively. It can beeasily seenfrom the Fig. 10 that there is
a huge difference in the energy levels of both the nodes. In
other words, the energy consumption is more at the nearest
node to the BS compared to the farthest node from the BS. As
the nearest node has to transmit its own message along with
the other messages passingfrom the upper level nodes to the
sink node/BS, its energy drains very fast compared to other
nodesas shown in the figure. The x-axis on the graph
represents time in seconds and y-axis represents energy in
joules. The initial energy of all the nodes is chosen as 50
joules.

Figure 8: Energy Level of nearest node w.r.t static BS
Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 7


ISSN 2320 - 5377 | 2013 Bonfring


Figure 9: Energy Level of Farthest Node w.r.t Static BS




Figure 10: Comparison of Energy Levels between Nearest and
Farthest Node w.r.t Static BS
Again, we tried to measure the energy consumption at
some nodes in a typical topology as shown in Figure 11. We
selected 3 random nodes to measure the energy consumption.
In the simulation one node has been selectedas cluster head
randomly during one of the cycles of the simulation and other
two nodes do not get a chance to bethe cluster head.After
conducting extensive simulations for 2000 Seconds in the area
of (1000x 500) meters, it has been concluded that the energy
level of node being selected as cluster head decreases very fast
compared to the other nodes as shown in Figure 12. The green
line indicates the dropping energy level of the node which has
been selected as the Cluster Head. In this simulation scenario
the sink node has been taken as mobile sink. It is also evident
from the Figure 12 that the total energy consumption by nodes
in the simulation is almost equal but the energy consumption
levels of the nodes in the static base station or sink node case
shown in figure 10 makes a lot of difference in the energy
levels after the simulation period.


Figure 11: Example Topology of WSN



Figure 12: Comparison of Energy Levels between Three
Random Nodes
Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 8


ISSN 2320 - 5377 | 2013 Bonfring


Figure 13: Comparison between Static Sink Node/Base Station
w.r.t Our Proposed Work
The figure 13 shows the comparison results of the energy
levels of the static sink/base station node with respect to
mobile sink/base station node. Thus from the graph it is
concluded that our proposed model consumes less energy
compared to static base station case.

Figure 14: No. of Packets Transmitted Vs Received
Further, we tried to measure the efficiency of the network
in terms of packet loss in both the cases i.e. when the base
station is static as well as it is mobile. We have transmitted
9794 number of packets to the base station. After simulating
for 1000 seconds, it is concluded that in case of static BS, the
number of dropped packets are 437 and number of received
packets are 9357. Similarly, when the base station is mobile,
the amount of dropped packets is 168 and the amount of
received packets is 9626. It has been plotted in Figure 14 for
static base station and for mobile base station. From this
statistics we have calculated the efficiency which is 98.2% for
mobile BS and 95.5% for static base station which is shown in
figure 15.


Figure 15: Efficiency Percentage


Thus, from the simulations and the results we can proof
that our proposed algorithm performs better and improves the
lifetime of the network with respect to static sink node
scenario.
VI. FUTURE WORK
Our future work includes implementing the proposed
protocol w.r.t mobile BS in cognitive radio WSNs. Now a
days Cognitive radio is becoming hotresearch topic to use
unutilized spectrum inan efficient and effective manner. Even
though a lot of research has been carried out on cognitive
radio, still more research is required to ensure the efficient use
of spectrum for spectrum aware communication.
VII. CONCLUSION
In this paper, we have proposed routing protocol based on
dynamic slot reservation protocol for Mobile Wireless Sensor
Networks. In this proposal we have proposed a sensor network
model consisting of two regions of interest. We have
considered clustering and path planning techniques to find an
optimal path for base station to collect the data from the entire
sensor field. Our proposed protocol also includes local data
fusion algorithm inside the cluster and global data fusion
algorithms at the base station. Each Super Cluster Head is
assigned with a predefined slot. In our proposal, Base Station
has been considered as mobile. As Base station is mobile, it
can collect the information from the Super Cluster Head while
finding thepath. We analyze the performance measures using
NS-2 simulator in a mobile wireless environments.
REFERENCES
[1] H. Qi, S. S. Iyengar, K. Chakrabarty, Distributed Sensor Networks-A
Review of Recent Research, Journal of the Franklin Institute, 338,
Pp. 655-668. 2001
[2] C. Chiasserini and R. Rao, Energy efficient battery management,in
Proc. of IEEE INFOCOM, Vol. 2, Pp. 396403, 2630 March 2000.
Bonfring International Journal of Networking Technologies and Applications, Vol. 2, No. 1, March 2013 9


ISSN 2320 - 5377 | 2013 Bonfring
[3] S. Jayashree, B. S. Manoj, and C. S. R. Murthy, On using battery state
for medium access control in ad hoc wireless networks, in Proc. of
ACM MobiCom, Philadelphia, PA, USA, Pp. 360373, 2004.
[4] G. Pei and C. Chien, Low power TDMA in large wireless sensor
networks, in Proc. IEEE MILCOM , 2001.
[5] V. Rajendran, K. Obraczka, and J. J. Garcia-Luna-Aceves, Energy-
efficient collision-free medium access control for wireless sensor
networks, in Proc. of 1st international conference on Embedded
networked sensor systems (inSenSys), Pp. 181192, 2003.
[6] Ameen M.A, Nessa A, Kyung Sup Kwak, "QoS Issues with Focus on
Wireless Body Area Networks", in Proc. of Third International
Conference on Convergence and Hybrid Information Technology
(ICCIT) , Vol.1, Pp.801-807, 11-13 Nov. 2008.
[7] Kusy B., HyungJune Lee, Wicke M., Milosavljevic N, Guibas L,
"Predictive QoS routing to mobile sinks in wireless sensor networks" , in
Proc. ofInternational Conference onInformation Processing in Sensor
Networks (IPSN) , Pp.109-120, 13-16 April 2009.
[8] Long Zhaohua; Su Meijun; , "Research on Quality of Service in
Wireless Sensor Networks", in Proc. of2nd International Conference on
Information Engineering and Computer Science (ICIECS),pp.1-4,25-26
Dec. 2010
[9] Rajesh yadav, Shirshu varma, and N. Malaviya, A Survey of MAC
Protocols for Wireless Sensor Networks, in UbiCC Journal, Vol. 4,
No. 3, August 2009.
[10] J. Sa silva, B.Krishnamachari and F.Boavida (Eds): Wireless Sensor
Networks 7th European Conference EWSN 2010 Coimbra Portugal
LNCS 5970, Pp. 81-96, February 2010
[11] Do-hyun Nam, hong-ki min, "An Energy-Efficient Clustering Using a
Round-Robin Method in a Wireless Sensor Network," in Proc. of 5th
International Conference Software Engineering Research, Management
& Applications (SERA ), Pp.54-60, 20-22 August 2007
[12] Lakshmi Phani, G. Venkat Sayeesh, K Vinod Kumar, Rama Murthy G ,
"ERFLA: Energy efficient combined routing, fusion, localization
algorithm in cognitive WSN", in Proc. ofSeventh International
Conference On Wireless And Optical Communications Networks
(WOCN), Pp.1-5, 6-8 September 2010.
[13] Ehsan, S.; Hamdaoui, B.; , "A Survey on Energy-Efficient Routing
Techniques with QoS Assurances for Wireless Multimedia Sensor
Networks," Communications Surveys & Tutorials, IEEE , vol.14, no.2,
Pp.265-278, Second Quarter 2012
[14] Yun Wang, Peizhong Shi, Kai Li, Zhikui Chen, "D-TDMA: An
Approach of Dynamic TDMA Scheduling for Target Tracking in
Wireless Sensor Networks," in Proc. of Green Computing and
Communications (GreenCom), 2010 IEEE/ACM Int'l Conference on &
Int'l Conference on Cyber, Physical and Social Computing (CPSCom) ,
Pp.546-553, 18-20 December 2010
[15] Kamruzzaman, S.M. Alam, M.S., "Dynamic TDMA slot eservation
protocol for cognitive radio ad hoc networks," in Proc. of 13th
InternationalConference on Computer and
Information Technology (ICCIT), Pp.142-147, 23-25 December 2010.
[16] Tian Yan-xia, Wu Ke-jun, "A novel dynamic TDMA protocol for ad hoc
networks using directional antennas",in Proc. of Electric Information
and Control Engineering (ICEICE), Pp. 65-69, 15-17 April 2011.
[17] Abbes Y., Abdelkefi F., Besbes H., "Novel power and time
allocation algorithm for a Dynamic TDMA slot assignment multiuser
transmission schemes", in Proc of GLOBECOM Workshops (GC
Wkshps), IEEE, Pp.824-828, 6-10 December 2010.
[18] Iwanicki K, Azim T, Experimentally studying the sensornet point-to-
point routing techniques spectrum, in Proceedings of the 7
th

International Conference on Networked Sensing Systems (INSS), Pp. 6-
13, 2010.
[19] Iwanicki K, Van Steen, On hierarchical routing in wireless sensor
networks in Proceedings of 8
th
ACM/IEEE Conference on Information
Processing in Sensor Networks (IPSN), Pp. 133-144, 2009.
[20] Bose P., et. al. routing with guaranteed delivery in ad hoc wireless
networks in Proceedings of the 3
rd
International workshop on Discreat
Algorithms and Methods for Mobile Computing and Communications
(DIALM), Pp. 48-55, 1999.






M. Sujeethnanda (Research student, CRC, IIIT-
Hyderabad)obtained his B.Tech degree in Electronics
and Communication engineering from JNTU
Hyderabad. Currently Pursuing M.S. in Embedded
systems in IIIT-Hyderabad. He is a member of CSI.
He has about 5 research publication and a book
chapter to his credit.His research interests include
Wireless Sensor network, Embedded system Design,
Alternate energy sources.

PadmalayaNayak received the engineering degree in
Electronics and Telecommunication Engineering (ECE)
from Institution of Electronics and Telecommunication
Engineers (IETE), New Delhi in 1997. She received
the M.E degree with gold medal in Computer Science
and Engineering (CSE) from University of Madras in
2002. She also received her Doctorate degree in
Computer Science and Engineering (CSE) from
National Institute of Engineering Technology (NIT),
Tiruchirappalli, India, in 2010. Currently, she is working as a Professor in
GokarajuRangaraju Institute of Engineering Technology, under JNTU,
Hyderabad. Previously she was working as Assistant Professorunder Anna
University, Chennai. She is the member of IE and IETE Professional bodies.
She has published many International/National Conference and Journal
papersin the area of Ad hoc and Sensor Networks. Her current research
interest includes Ad hoc and Sensor networks, Mobile communication, and
Ethical hackings etc.
G. RamaMurthy (Associate Professor,CRC,IIIT-
Hyderabad) obtained his B.Tech degree in Electronics
and Communication Engineering from Sri
Venkateswara University, Tirupati, India, M.S. in
Electrical Engg. from the Louisiana State University,
Baton Rouge U.S.A and Ph.D. degree in computer
engineering from Purdue University, West Lafayette,
U.S.A. His research interests include Wireless Sensor
Networks, Adhoc Wireless Networks, Computer
Architecture, Multi Dimensional Neural Networks, Performance Evaluation,
Cybernetics and Optimal Filtering. He is currently an associate professor at
the International Institute of Information Technology (IIIT), Hyderabad, India.
He is a member of Eta Kappa Nu, Phi Kappa Phi, IEEE and senior member of
ACM. He has about 140 referred research publications and about 40 Journal
publications to his credit. He is currently the editor of International Journal
of Systemic, Cybernetics and Informatics and Research Journal of
Telecommunication and Information Technology (INSINET publications).