International Journal of Technical Research and Applications e-ISSN: 2320-8163,
www.ijtra.com Volume 5, Issue 4, (July-August) 2017), PP. 53-58
ENERGY PROFICIENT CLUSTER BASED
ROUTING PROTOCOL FOR WSN
1
Shriom Maurya, 2Shish Ahmad, 3Mohd Haroon
Dept. of Computer Sc. & Engg., Integral University Lucknow, India
1
Shriommaurya87@gmail.com, 2shish@iul.ac.in , 3mharoon@iul.ac.in
Abstract— Wireless sensor networks (WSNs) have become
more and more popular and have been widely used recently.
WSNs usually consist of a large number of sensors for different
applications of sensing that includes Military, medical, civil,
disaster management, environmental, and commercial
applications.
In this research we aim to design a cluster based routing
approach for heterogeneous environment that increase the
lifetime of wireless sensor network by reducing the energy
consumption.
Keywords— Routing protocols, Energy proficient routing,
WSN,MATLAB.
I. INTRODUCTION
The main aspire of energy efficient routing is to minimize
the energy required to transmit or receive packets also called
as active communication energy. Inactive energy is the energy
which not only tries to reduce the energy consumed when a
mobile node stays idle but also listens to the wireless medium
for any possible communication requests from other nodes.
Transmission power control method and load distribution
method are the two methods which decreases active
communication energy[1]. The sleep or power-down mode
method decreases in- active energy. Both the protocol has
specific benefits and drawbacks and therefore is applicable for
certain situations. Thus it is not clear that which particular
algorithm or a class of algorithms is the most excellent for all
scenarios. To conserve energy, many energy efficient routing
protocols have been proposed. Many re- searches are being
made to carry out to develop energy aware routing protocols.
Some are designed to search for the most energy efficient path
from the source to the destination while some attempt to
balance the remaining battery-power at each node when
searching for the energy efficient path.
fundamental, for instance, in armed forces environments
where node cooperation is assumed. The goals can be
achieved either by developing better batteries, or by making
the network terminals operation more energy competent. The
first method is likely to give a 40% increase in batter y life in
the near future (with Li-Polymer batteries). As to the device
power utilization, the primary aspect are achieving energy
savings through the low power hardware development using
techniques such as variable clock speed CPUs, flash memory,
and disk spin down. Nevertheless, from the networking point
of view, our attention naturally focuses on the device's
network interface, which is often the single largest consumer
of power. Energy effectiveness at the network interface can be
improved by developing transmission/ reception technologies
on the physical layer.
B. Sensor Network Communication Architecture
According to [4], the sensor network is composed of the
number of sensor devices or nodes. Each node has the capacity
to gather information and then send these useful information
to the sink and the end users. With the aid of multi-hop
infrastructure and less architecture the information gathered is
routed back to the final user through sink as shown in figure 1.
II. LITERATURE REVIEW
A. Energy Conservation
Energy conservative networks [2][3] are becoming
extremely popular within the Ad hoc networking research.
Energy preservation is presently being addressed in every
layer of the protocol stack. There are two chief research topics
which are almost identical: maximization of lifetime of a
single battery and maximization of the lifetime of the whole
network. The previous is related to commercial applications
and node cooperation issues whereas the latter is more
Figure 1: Wireless Sensor Network.
Here, in this network the sink send commands or queries to
other sensor nodes in sensing area, on other hand sensor node
work in a group to achieve the sensing task and send sensed
information to sink. In the meantime, sink act as gateway to
the outer networks. Further, sink gather information from
53 | P a g e
International Journal of Technical Research and Applications e-ISSN: 2320-8163,
www.ijtra.com Volume 5, Issue 4, (July-August) 2017), PP. 53-58
sensor nodes, and performs simple processing on these
gathered information and then finally, sends appropriate data
to the end user through internet. Each of the sensor nodes in
the network uses single-hop long-distance transmission to
send information to the sink. Both sink and nodes uses
protocol stack where it combines power and routing
awareness, merges information with networking protocols,
communicates power efficiently by means of wireless medium
and promotes joint efforts of sensor nodes.
However, this method is expensive in terms of energy
consumption for long-distance transmission [5]. Therefore,
from the above context it can be declared that sensor network
consists of large number of small nodes with computation,
sensing and wireless communication capabilities. Apart from
these the network still produces high-quality data due to its
coordination of sensor nodes.
Fig 2: Clustering and Forming Hierarchies
C. Hierarchical State Routing (HSR)
Hierarchical State Routing (HSR) employs a multilevel
clustering and logical partitioning scheme. The network is
partitioned into clusters and a cluster-head is elected as in a
cluster-based algorithm. Cluster heads again organize
themselves into clusters up to any desired clustering level as
shown in Fig 2. Within a cluster, nodes broadcast their link
information to one another. A cluster head summarizes its
cluster information and sends it to neighboring clusters
through a gateway node. A gateway node is one, which is
adjacent to one or more cluster heads. Here cluster heads are
members of a higher- level cluster[7]. At each level,
summarization and link information exchanges are executed.
The manner in which the information is exchanged in this
hierarchy is, first information is collected among the nodes
forming the base level cluster, it is then passed on to the
cluster head which in turn passes to its next hierarchical
cluster head and from there on the information is disseminated
into other cluster heads and thus the information traverses
down the hierarchy. Here every node has a hierarchical
address, which may be obtained by assigning numbers from
the top root to the bottom node. But as a gateway can be
reached from the root from more than one path, so a gateway
can have more than one hierarchical address [6].
Also, each subnet contains a location management server
(LMS). All nodes in the subnet are registered with the local
LMS. LMS has to inform upper levels, and upper level
information comes to local LMS server. When two nodes wish
to communicate, they send their initial data to the LMS, and
the LMS then forwards it to the destination. But if the source
and destination know each other’s hierarchical addresses, they
communicate directly. The protocol is highly adaptive to
network changes.
The cluster head can monitor all the traffic with in the
cluster and provide QoS service to real time applications
simply by appending bandwidth and channel quality
information to the link state information. The control traffic in
HSR can be comparable to that of in on-demand protocols.
The latency for access to non-frequently used destinations is
low. But, the average number of hops the packets take,
protocol complexity, packets dropped because of invalid
routes is more in HSR when compared to that of in on-demand
protocols.
D. Clustered Gateway Switch Routing protocol (CGSR)
In this protocol, nodes are aggregated into clusters
controlled by a cluster head elected using a distributed
algorithm as shown in Fig 3. All nodes within the transmission
range of the cluster-head belong to this cluster [8]
54 | P a g e
International Journal of Technical Research and Applications e-ISSN: 2320-8163,
www.ijtra.com Volume 5, Issue 4, (July-August) 2017), PP. 53-58
•
Chief Precinct 2: Half of advance nodes are deployed
randomly in Chief Precinct 2, lying between 80<Y<=100.
The reason behind this type of deployment is that advance
nodes have high energy than normal nodes. As corners are
most distant places in the field, so if a node is at corner then it
requires more energy to communicate with base station so we
have deployed high energy nodes (advance nodes) in Chief
Precinct 1 and Chief Precinct 2.
Fig 3: Clustering
Clustering provides framework for the development of
important features such as code separation (among clusters),
effective channel allocation and spatial reuse, routing and
bandwidth allocation. But the selection of the cluster heads
may cause complexity and overhead, thus degrading
performance. Also, there are traffic bottleneck and single point
failures at the cluster heads and gateways.
III. PROPOSED METHODOLOGY
It is cluster based routing protocol, in which cluster head is
elected randomly according to the election probability.
We divide all the nodes of WSN in two categories on the
basis of their energy.
•
Normal node
•
Advanced Node
Advance nodes have high energy than normal nodes. We
do not form cluster of normal nodes as energy of normal node
is less than advance node, and cluster head consumes more
energy than cluster members in receiving data from cluster
members. If we allow normal nodes to become cluster head
they die soon resulting in the shortening of stability period.
A. Proposed Precinct-Based Steady Choice Protocol [PSCP]
In most routing protocols, nodes are deployed randomly in
network field and energy of nodes in network is not utilized
efficiently. We modified this theme: network field is divided
in three precincts: precinct 0, chief precinct 1 and chief
precinct 2, on the basis of energy levels and Y co-ordinate of
network field. We assume that a fraction of the total nodes are
equipped with more energy. Let m be fraction of the total
nodes n, which are equipped with α time more energy than the
other nodes. We refer these nodes as advance nodes, (1-m)×n
are normal nodes.
•
Precinct 0: Normal nodes are deployed randomly in
Precinct 0, lying between 20<Y<=80.
•
Chief Precinct 1: Half of advance nodes are deployed
randomly in this region, lying between 0<Y<=20.
Fig.4 Network Architecture
1) PSCP Operation
PSCP uses two techniques to transmit data to base station.
Techniques are:
•
Direct communication.
•
Transmission via Cluster head.
Direct Communication:
Nodes in Precinct 0 send their data directly to base station.
Normal nodes sense environment, gathers data of interest and
send it data directly to base station.
Transmission via Cluster head:
Nodes in Chief Precinct 1 and Chief Precinct 2 transmit
data to base station through clustering algorithm. Cluster head
is selected among nodes in Chief Precinct 1 and Head zone 2.
Cluster head collect data from member nodes, aggregate it and
transmit it to base station. Cluster head selection is most
important. As shown in Fig.4 advance nodes are deployed
randomly in Chief Precinct 1 and Chief Precinct 2. Cluster is
formed only in advance nodes. Assume an optimal number of
clusters Kopt and n is the number of advance nodes.
Every node decides whether to become cluster head in
current round or not. Every node has optimal probability
(Popt) to be cluster head which is calculated as follows.
55 | P a g e
International Journal of Technical Research and Applications e-ISSN: 2320-8163,
www.ijtra.com Volume 5, Issue 4, (July-August) 2017), PP. 53-58
Popt
K opt
n
(1)
A random number between 0 and 1 is generated for node. If
this random number is less than or equal threshold T(n) for
node then it is selected as cluster head. Threshold T(n) is given
by as follows.
(2)
Where G is the set of nodes which have not been cluster
heads in the last 1/Popt rounds.
Probability for advance nodes to become cluster head is
proposed as
Fig.5 Nodes sending data to cluster head
When data is received from nodes, Cluster head then
aggregates this data and send it to the base station this phase is
called as transmission phase. Fig.6 illustrates this phase.
(3)
Accordingly the threshold for advance nodes is
(4)
G' is the set of advance nodes that have not been cluster
head in the last 1/Padv rounds.
Once the cluster head is selected then the cluster head
broadcasts an advertisement message to the nodes. The nodes
receive the message and decide to which cluster head it will
belong for the current round. This phase is called as cluster
configuration phase.
On the basis of received signal strength, nodes respond to
cluster head and become member of cluster head. Cluster head
then assign a TDMA schedule for the nodes during which
nodes can send data to cluster head. After the clusters
configuration, every node data and sends it to the cluster head
in the time slot allocated by the cluster head to the node. This
phase is shown in Fig. 5.
Fig.6 Cluster head transmitting data to base station
IV. EXPERIMENTATION EVALUATION
A. Simulations
We simulate our proposed protocol in a field with
dimensions 100m×100m and 100 nodes deployed in specific
zones with respect to their energy. Base station is placed in the
center of the network field. We are using the first order radio
56 | P a g e
International Journal of Technical Research and Applications e-ISSN: 2320-8163,
www.ijtra.com Volume 5, Issue 4, (July-August) 2017), PP. 53-58
model as used in SEP. MATLAB is used to implement the
simulations.
Specifically, we have following settings.
Let 20% of nodes are advance nodes and half of them are
deployed in Chief Precinct 1 and half in Chief Precinct 2.
Since Popt is 0.1 so we have 2 cluster heads per round. One
cluster head in Chief Precinct 1 and one in Chief Precinct 2
per round.
Other simulation parameters are shown in Table 1.
Table 1: Simulation parameters
Parameters
Value
Initial energy Eo
0.5 J
Initial
nodes
energy
of
advance Eo(1+α) α
Denoted by a
In Graph
weighted probability for selection of cluster head for both
normal nodes and advance nodes. PSCP performs better than
LEACH and SEP, because nodes in Precinct 0 (normal nodes)
communicates directly to base station while nodes in Chief
Precinct 1 and Chief Precinct 2 communicates via cluster head
to base station: As in clustering technique, cluster head
consumes energy in the form of data aggregation and also by
receiving data from nodes in the cluster. So this energy is
conserved in normal nodes as they do not have to aggregate
data and receive data from other nodes, so energy is not
dissipated as that of cluster head, resulting the increase of
stability period. In Fig.7, we can see that network lifetime is
also increased because of the advance node. Advance nodes
have α time more energy than normal nodes so advance nodes
die later than normal nodes. So this increases the instability
period.
Energy for data aggregation 5 nJ/bit/signal
EDA
Transmitting and receiving 5 nJ/bit
energy Eelec
Amplification energy for short 10 Pj/bit/m2
distance Efs
Amplification energy for long 0.013 pJ/bit/m4
distance Eamp
Probability Popt
0.1
B. Result and Discussion
Here, we compare the results of our protocol with SEP and
LEACH. We have introduced heterogeneity in LEACH, with
the same setting as in our proposed protocol, so as to access
the performance of all the protocol in presence of
heterogeneity. Our goals in conducting simulation are
•
To examine the stability period of LEACH, SEP and
PSCP.
•
We also examine the throughput of LEACH, SEP and
PSCP.
Fig.7 and Fig.8 shows result for the case when m=0.1 and
α=1.This means that there are 100 advance nodes out of total
nodes which are 100. According to our proposed protocol 5
advance nodes will be deployed randomly in Chief Precinct 1
and 5 advance nodes will be placed in Chief Precinct 2.
Fig.7 shows the number of alive nodes against rounds. Fig.7
clearly shows that our protocol is enhanced from SEP and
LEACH in terms of steadiness. As LEACH is very sensitive to
heterogeneity so nodes die at a faster rate. SEP performs better
than LEACH in two level heterogeneity, because SEP has
Fig.7 Alive nodes in LEACH, SEP and PSCP
Fig.8 Throughput in LEACH, SEP and PSCP
Fig. 8 shows the throughput of LEACH, SEP and PSCP.
Throughput of PSCP is greater than LEACH and SEP.
57 | P a g e
International Journal of Technical Research and Applications e-ISSN: 2320-8163,
www.ijtra.com Volume 5, Issue 4, (July-August) 2017), PP. 53-58
CONCLUSION
In this research we proposed we proposed a approach for
heterogeneous environment of wireless sensor network. We
divide all the nodes of WSN in two categories Normal node
and Advanced Node.
Advance nodes have high energy than normal nodes. We
do not form cluster of normal nodes as energy of normal node
is less than advance node, and cluster head consumes more
energy than cluster members in receiving data from cluster
members. We divided in three precincts: precinct 0, chief
precinct 1 and chief precinct 2, on the basis of energy levels
and Y co-ordinate of network field. Nodes in Precinct 0 send
their data directly to base station. Nodes in Chief Precinct 1
and Chief Precinct 2 transmit data to base station through
clustering algorithm. Cluster head is selected among nodes in
Chief Precinct 1 and Head zone 2. Cluster head collect data
from member nodes, aggregate it and transmit it to base
station.
We simulate our proposed protocol in a field with
dimensions 100m×100m and 100 nodes deployed in specific
zones with respect to their energy. Base station is placed in the
center of the network field. We are using the first order radio
model as used in SEP. MATLAB is used to implement the
simulations.
We have compared the average results for LEACH, SEP
and our proposed approach PSCP. Approximately 100%
stability period of our proposed protocol is increased from
LEACH and SEP, however network lifetime is increased little
bit when compared with LEACH. When compared with SEP,
PSCP network life time is increased due to advance nodes
which die slower than normal nodes. Network lifetime of SEP
is short because of the weighted probability for normal and
advance nodes in the field.
.
Wireless
Sensor
Networks
through
Zone
Partitioning." Indonesian Journal of Electrical
Engineering and Computer Science 2.2 pp 390-395,
2016
[7] Ouafaa, Ibrihich, et al. "The comparison study of
hierarchical routing protocols for ad-hoc and wireless
sensor networks: A literature survey." Proceedings of
the The International Conference on Engineering &
MIS 2015. ACM, 2015.
[8] Devarajan, K., and V. Padmathilagam. "An Enhanced
Cluster Gateway Switch Routing Protocol (ECGSR)
for Congestion Control using AODV Algorithm in
MANET." International Journal of Computer
Applications 123.3, 2015.
REFERENCES
[1] Saraiya, Smit D., et al. "Algorithm for Energy
Preservation in Wireless Sensor Network."
International Journal of Computer Applications
124.12 , 2015.
[2] C. Petrioli, R. Rao, and J. Redi, “Guest
editorial:
Energy conserving protocols”, ACM
Mobile Networks and Applications (MONET), 6(3):
pp 207-209, June 2001.
[3] C. Jones, K. Sivalingam, P. Agrawal, and J.
Chen, “A survey of energy efficient network
protocols for wireless networks”,
Wireless
Networks, 7(4): pp 343-358, September 2001.
[4] Mooi Choo Chuah and Qinqing Zhang, “Design
and Performance of 3G Wireless Networks and
Wireless Lans”, Springer US, ISBN:978-0-38724152-4 (Print) 978-0-387-24153-1 (Online).
[5] P. Jacquet,P.
Muhlethaler,
and A. Qayyum,
“Optimized Link State Routing Protocol”,IETF
Internet Draft, draft-ietf-manet-olsr-10.txt, June 2002.
[6] Hassan, Shahzad, Muhammad S. Nisar, and Hongbo
Jiang. "Energy Preservation in Heterogeneous
58 | P a g e