International Journal of Advanced Smart Sensor Network Systems (IJASSN)

International Journal Of Advanced Smart Sensor Network Systems ( IJASSN ), Vol 2, No.2, April 2012
DOI: 10.5121/ijassn.2012.2205 45
Base Station and Zone Based Performance
Evaluation of Optimal Clustering in Wireless
Sensor Network
S. Taruna1
Jain Kusum Lata2
,
Purohit G.N.
Computer Science Department, Banasthali University, India
1
staruna71@yahoo.com
2
kusum_2000@rediffmail.com
3
gn_purohitjaipur@yahoo.co.in
Abstract –
The placement of base stations in wireless sensor networks affect the energy consumption for
communication between sensor node and base station. In this paper we analyzed the performance of the
zone based clustering protocol [2] under varying position of base stations, different zone sizes and the
effect on network life time with multiple base stations. While evaluating the communication overhead of
various cluster sizes, we observed that the optimal cluster size for a given network is complex, depending
on a range of parameters. Simulation results show that communication overhead decreases as we increase
the number of zone in the network. We show that placing multiple base stations in place of single base
station in zone based routing protocol enhance the network life time.
Keywords-
zone based clustering protocol, Zone, Wireless sensor network, network life time.
1. Introduction
Sensor network is composed of a large number of tiny autonomous devices, called sensor nodes.
Each sensor node has four basic components: a sensing unit, a processing unit, a radio unit, and a
power unit. WSN is an emerging technology that can be deployed in such situation where human
interaction is not possible like border area tracking enemy moment or fire detection system.
Sensor network is composed of a large number of tiny autonomous devices, called sensor nodes.
Each sensor node has four basic components: a sensing unit, a processing unit, a radio unit, and a
power unit. Figure 1 shows an overview of WSN. Sensor are deployed in the environment which
can be fire area, border or open environment. These tiny devices sense the area of interest and
then communicate with Base Station (BS). On BS the gathered information is analyzed.

46
Figure 1. Overview of a wireless sensor network.
Clustering in wireless sensor networks (WSNs) is the process of dividing the nodes of the WSN
into groups, where each group agrees on a central node, called the cluster head, which is
responsible for gathering the sensory data of all group members, aggregating it and sending it to
the base station(s).
Despite the wealth and variety of protocols and their individual evaluations through simulation,
experimentation, and comparison to existing approaches, little help is available for the WSN
application developer. We define the optimal cluster as the one sized such that routing data from
the cluster member to cluster head and subsequently to base stations incurs the minimal
communication overhead. We confronted this question of optimal clusters for zone based
clustering algorithm ZBCA[2], a low energy, low overhead clustering protocol that divided the
area into zone and create a cluster head per zone. While evaluating the communication overhead
of various cluster sizes, we observed that the optimal cluster size for a given network is complex,
depending on a range of parameters.
Therefore, in this paper, we explore the problem of optimal clustering in detail, through
experimentation using MATLAB. Section 2 sketches the current state of the art of clustering
algorithms, Section 3 Identifying the optimal cluster, Section 4 Energy model and Simulation
parameters, Section 5 Experiments and Simulation Result, Section 6 conclusion.
2 Related Work
Routing is a process of selecting a path in the network between source and destination along
which the data can be transmitted. Various protocols are available to route the data from node to
base station in WSN in an energy efficient way, enhancing the network survivability. Battery
power being limited in the sensor nodes let the node to expire on its full consumption. The
expired node is called the dead node which results in its no contribution in the network further.
Grouping sensor nodes into clusters has been widely pursued by the research community in order
to achieve the network scalability objective.[1] Sensors organize themselves into clusters and
each cluster has a leader called as cluster head(CH), i.e. sensor nodes form clusters where the low
energy nodes called cluster members (CM) are used to perform the sensing in the proximity of the
phenomenon. For the cluster based wireless sensor network, the cluster information and cluster
head selection are the basic issues. The cluster head coordinates the communication among the
cluster members and manages their data.[3]
Low-energy adaptive clustering hierarchy (LEACH) is a popular energy-efficient clustering
algorithm for sensor networks. It involves distributed cluster formation. LEACH randomly selects
a few sensor nodes as CHs and rotate this role to evenly load among the sensors in the network in

47
each round. In LEACH, the cluster head (CH) nodes compress data arriving from nodes that
belong to the respective cluster, and send an aggregated packet to the base station. A
predetermined fraction of nodes, p, elect themselves as CHs in the following manner
Another cluster selection zone based scheme Zone Based Clustering Algorithm (ZBCA) [2,4]
involves dividing the area into rectangle grids called zones. In this algorithm, communication
between the sensor nodes for the CH selection is reduced, so that the energy consumption for CH
selection can be reduced. . The probability of a node to become CH is p, where p is the maximum
number of CHs in network. CHs organize and control the cluster members in their cluster. All
nodes have to establish a connection with a CH to join its cluster. The CH of each zone sends the
join request to nodes of its neighbouring zones only. On the basis of distance the nodes select one
CH, which is most near to it. Sensor nodes send data to CHs and CHs further to base station.
Cluster head is responsible for data aggregation and compression that reduces the energy
consumed for data transmission.
3 Identifying the optimal cluster
Our analysis considers the effect of key clustering parameters on the communication overhead.
We follow an experimental approach for two reasons. First, it is hard if not impossible to derive
generally valid formulas for the network communication overhead that consider all parameters,
especially random topologies. Our motivation is to make our results applicable on a real
application and the most relevant scenarios from our experiments directly derive the optimal
clustering parameters. We performed our simulations in MATLAB. In the scenario, nodes have
uniform random distribution in area. The network area is divided into equal-size zone and each
node has a cluster. Results are presented for a variety of cluster sizes that allow the area to be
precisely divided into such equal-size zone. The shortest distance in terms of ETX, expected
number of transmissions, is computed between each node and its corresponding cluster head and
between all cluster heads and the base stations. The energy expenditure is calculated as the sum
of ETX and ERX (expected number of receivers) for one round of data gathering. Overhearing
costs are included in ERX. Cluster formation energy is used Each of the reported experiments is
the mean of 100 independent random topologies with random nodes selected as base stations Our
goal is always to identify the optimal cluster size for scenarios, first studying the placement of
base stations and then the varying size of zone for the clustering in Zone based routing
Algorithm.
4 Energy model and Simulation parameters
We use the energy model for simulation as [3,5]. Following simulation parameters is used for the
simulation.

48
Table 1: Simulation Parameters
Parameter Values
Simulation Round 2000
Topology Size 200 X 200
Number of nodes 100
CH probability 0.5
Initial node power 0.5 Joule
Nodes Distribution Nodes are uniformly
randomly
distributed
BS position Located at (100,250)
Energy for Transmission
(ETX)
50*0.000000001
Energy for Reception (ERX) 50*0.000000001
Energy for Data
Aggregation (EDA)
5*0.000000001
5 Experiments and Simulation Result
Simulation is performed for two scenarios.
5.1. Multiple Base stations
In this experiment we change the location of the base station and number of base station. In the
first experiment there is a single base station located at the boundary of the wireless sensor area.
Table 2 show the distance from the first cluster head to base station. In each round cluster head
send the aggregated data to the base station. According to energy model transmission energy
depend on the distance. As the distance increases node required the more energy for transmission.
Figure 2: Zone based clustering with single base station
BS

49
Table 2 : Energy consumption for a single transmission from cluster head to base station in single base
station scenario
Zone
ID
Distance
from cluster
to Base
Station
Energy consumption for a single
transmission
E=ETX*PL+Emp*PL*(dist*dist*dist))
Z1 117.47 5.21073E-05
Z2 100.54 5.13212E-05
Z3 110.89 5.17726E-05
Z4 231.48 6.61244E-05
Z5 226.88 6.51821E-05
Z6 234.09 6.6676E-05
If we use 2 base station located at the boundary of the network area in opposite direction.
Figure 3: Zone based clustering with two base station
Following table shows the energy consumption for the single transmission from cluster head to
base station.
BS
BS

50
Table 3: Energy consumption for a single transmission from cluster head to base station in two base station
scenarios
Zone ID Distance from the
Cluster head to base
station near to it.
Energy consumption for a single
transmission
E=ETX*PL+Emp*PL*(dist*dist*dist))
Z1 182.44 5.78941E-05
Z2 177.13 5.72247E-05
Z3 187.88 5.86215E-05
Z4 176.92 5.7199E-05
Z5 150.36 5.44192E-05
Z6 180.49 5.76437E-05
If we use base station located at the center the network area.
Figure 3: Zone based clustering with base station at center
Following table shows the energy consumption for the single transmission from cluster head to
base station.
Table 3: Energy consumption for a single transmission from cluster head to base station in base station at
center scenarios
Zone ID Distance
from
cluster to
Base
Station
Energy consumption
for a single
transmission
z1 87.658 5.87562E-05
z2 90.789 5.97284E-05
z3 106.3 6.5615E-05
z4 96.438 6.16597E-05
z5 51.07 5.17316E-05
z6 117.46 7.10675E-05
BS

51
Figure-4 : Comparison in energy consumption for a single transmission from cluster head to base station in
one and two base station scenarios
5.1.1. Network performance can also be defined in terms of number of round in which first node
dead. In above scenarios we find the number of round in which first node dead for 10 simulation
runs.
Table 4 : Number of round in which first node dead in single base station scenario and two base station
scenario and base station at center
Simulation
run
Number of
round in
which first
node dead in
case of single
Base station
located at
boundary
Number of round in
which first node
dead in case of 2
base stations located
at boundary in
opposite direction
Number of
round in which
first node dead
in case of base
stations located
at center
Simulation
run 1
667
669 722
Simulation
run 2
682
862 608
Simulation
run 3
754
912 704
0.0000000E+00
1.0000000E-05
2.0000000E-05
3.0000000E-05
4.0000000E-05
5.0000000E-05
6.0000000E-05
7.0000000E-05
8.0000000E-05
Z1 Z2 Z3 Z4 Z5 Z6
Single base station
Scenario
Two base station
Scenario
Single base station
at center

52
Simulation
run 4
860
862 546
Simulation
run 5
594
820 606
Simulation
run 6
822
970 590
Simulation
run 7
710
685 716
Simulation
run 8
781
750 847
Simulation
run 9
745
970 641
Simulation
run 10
611
770 683
As above the range of first node dead in single base station is from 594 to 860 and in the two base
station scenarios it range from 669 to 970. This shows the enhancement in network life time.
Figure-5: Number of round in which first node dead in single base station scenario and two base station
scenario and single base station at center
5.3. Using different zone size
We also perform simulation with the same simulation parameter but with different zone size.
0
200
400
600
800
1000
1200
Simulationrun1
Simulationrun2
Simulationrun3
Simulationrun4
Simulationrun5
Simulationrun6
Simulationrun7
Simulationrun8
Simulationrun1
Simulationrun1
Number of round in
which first node
dead in case of
single Base station
located at boundary
Number of round in
which first node
dead in case of 2
base stations
located at boundary
in opposite direction

53
Simulation is done with 2 zone , 4 zone , 6 zone, 16 and 64 zone of network area. Following table
shows the number of round in which first node dead.
Table 5:Number of rounds in which first dead node occurs with different zone size.
Simulation
Run
with 1
zone
with 4
zones
with 6
zones
with 16
zones
with 64
zones
1 516 608 745 1054 1205
2 552 591 792 1070 1340
3 541 650 683 970 1190
4 494 710 782 1140 1423
5 511 704 743 1005 1322
6 520 622 740 1280 1306
7 535 698 692 750 1155
8 549 598 699 1153 1465
9 562 550 763 1295 1537
10 501 601 635 929 1211
Figure-6 No. of round in which first dead node occurs with different zone size
As shown in above table as we reduce the area of zone the round number in which the first node
dead occur is increased. For a large network area the area can be divided in the smaller zone to
increase the network lifetime.
Conclusion:
In wireless sensor area network clustering is used to increase network lifetime. As the simulation
result show if we increase the number of base station than the energy required for communication
will be decrease. And if the base station is in center than it will also reduce the communication
0
200
400
600
800
1000
1200
1400
1600
1800
1 2 3 4 5 6 7 8 9 10
with 1 zone
with 4 zones
with 6 zones
with 16 zones
with 64 zones

54
cost. Transmission energy depends on the distance and distance for the communication is less in
multiple base station scenarios. In zone based algorithm the size of the zone will also affect the
residual energy of the node. Less zone area will increase the life time of network. Although this is
a theoretical simulation result for homogeneous network, but it can be easily used for the live
application.
References
1. Ameer Ahmed Abbasi and Mohamed Younis “A survey on clustering algorithms for wireless
sensor networks”,Computer Communications, Volume 30, Issues 14-15, 15 October 2007.
2. S. Taruna ,Jain Kusum Lata and Purohit G.N. ,” Performance Analysis of Energy Efficient
Routing Protocol for Homogeneous Wireless Sensor Network”, in proceedings of international
Conferences on “Trends in Networks and Communication”, NeCoM ,Chennai,India,July2011.
3. J. Al-Karaki, and A. Kamal, “Routing Technique s in Wireless Sensor Net-works: A Survey“,
IEEE Communications Magazine, vol 11, no. 6, Dec. 2004, pp. 6-28.
4. S. Taruna ,Jain Kusum Lata, Purohit G.N. “Power Efficient Clustering Protocol (PECP)-
Heterogeneous Wireless Sensor Netwrok”, International Journal of Wireless & Mobile
Network(IJWMN) Vol.3, No.3,June2011,pp 55-67.
5. C.R. Lin, and M. Gerla, “Adaptive Clustering for Mobile Wireless Net-works”, IEEE Journal on
Selected Areas in Communication, vol. 15, no. 7, Sept. 1997, pp. 11265-1275.
6. W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan. “Energy efficient communication
protocol for wireless microsensor networks”. In Proceedings of the Hawaii International
Conference on System Sciences,2000.
7. Yu, J.Y.; Chong, P.H.J. A Survey of Clustering Schemes for Mobile Ad Hoc Networks. IEEE
Commun. Surv. Tutorials 2005, 7, 32– 48.
8. Lin, C.R.; Gerla, M. Adaptive Clustering for Mobile Wireless Networks. IEEE J. Sel. Areas
Commun. 1997, 15, 1265–1275.

International Journal of Advanced Smart Sensor Network Systems (IJASSN)

More Related Content

What's hot (18)

Similar to International Journal of Advanced Smart Sensor Network Systems (IJASSN) (20)

Recently uploaded (20)

International Journal of Advanced Smart Sensor Network Systems (IJASSN)