ThesisPDF Available

MVC: Modified VIKOR Model based Clustering Protocol for WSNs

Authors:

Abstract and Figures

Stability and lifetime of Wireless Sensor Networks (WSNs) mainly depend on energy of each node in the network. Hence, it is necessary for WSNs to be energy efferent. There are different methods to preserve energy in WSNs and clustering is one of those methods. It is a technique to divide whole network into small blocks, each having a managing node, called cluster head (CH) and rest act as members. CH is responsible to provide communication bridge between members and the base station. In this paper, we propose a distributed clustering scheme that uses multiple criteria i.e. residual energy, node density, distance to the Base Station (BS) and average distance between a node and its neighbors, to select a CH. Modified VIKOR method is used to outrank the potential nodes as CHs. The realistic multi-hoping communication model is used, instead of single hop as in previous schemes. Simulation results show that our purposed technique performs much better than those previous methods in terms of energy efficiency and network life time. Our proposed scheme has less CH deformation and control overhead.
Content may be subject to copyright.
MVC: Modified VIKOR Model based
Clustering Protocol for WSNs
By
Tauseef Shah
Registration Number: -REE-049/ISB
MS Thesis
In
Electrical Engineering
COMSATS Institute of Information Technology
Islamabad – Pakistan
Spring, 2013
CIIT/FA11
MVC: Modified VIKOR Model based
Clustering Protocol for WSNs
 
!"
  !
 #! 
$%%&
'"

()**+%%+,-.('
/ 0,*1
COMSATS Institute of Information Technology
ii
Engineering).
MVC: Modified VIKOR Model based
Clustering Protocol for WSNs
A post Graduate Thesis submitted to Department of Electrical Engineering as
partial fulfillment of the requirement for the award of Degree of M.S
(Electrical
Name Registration Number
Tauseef Shah CIIT/FA11-REE-049/ISB
Supervisor
Dr. Safdar H. Bouk
Assistant Professor, Department of Electrical Engineering
Islamabad Campus
COMSATS Institute of Information and Technology (CIIT)
Islamabad
Co-Supervisor
Dr. Nadeem Javaid
Assistant Professor,
Center for Advanced Studies in Telecommunications (CAST),
COMSATS Institute of Information and Technology (CIIT)
Islamabad
Spring, 2013
iii
Final Approval
This thesis titled
MVC: Modified VIKOR Model based
Clustering Protocol for WSNs
By
Tauseef Shah
CIIT/FA11-REE-049/ISB
Has been approved
For the COMSATS Institute of Information Technology, Islamabad
External Examiner: __________________________________
(To be decided)
Supervisor: ________________________
Dr. Safdar H. Bouk/Assistant Professor,
Department of Electrical Engineering CIIT, Islamabad
Co-Supervisor: ________________________
Dr. Nadeem Javaid/Assistant professor,
Center for Advanced Studies in Telecommunications (CAST),
CIIT, Islamabad.
Head of Department:________________________
Dr. Raja Ali Riaz / Associate professor,
Department of Electrical Engineering,
CIIT, Islamabad.
iv
Declaration
I Tauseef Shah, CIIT/FA11-REE-049/ISB hereby declare that I have produced the work
presented in this thesis, during the schedule period of study. I also declare that I have
not taken any material from any source except ref erred to wherever due that amount of
plagiarism is within acceptable range. If a violation of HEC rules on research has
occurred in this thesis, I shall be liable to punishable action under the plagiarism rule of
the HEC.
Date: ___________
_____________
Tauseef Shah
CIIT/FA11-REE-049/ISB
v
Certificate
Date: _________________
Supervisor: ________________________
Dr. Safdar H. Bouk/Assistant Professor,
Department of Electrical Engineering CIIT, Islamabad
Co-Supervisor: ________________________
Dr. Nadeem Javaid/Assistant professor,
Center for Advanced Studies in Telecommunications (CAST),
CIIT, Islamabad.
Head of Department:________________________
Dr. Raja Ali Riaz / Associate professor,
Department of Electrical Engineering,
CIIT, Islamabad.
vi
DEDICATION
Dedicated
To
My Family
Teachers and Friends
Whom prayers and attention enable me to reach this
extent
.
vii
ACKNOWLEDGMENT
Tauseef Shah
CIIT/FA11-REE-049/ISB
I tank to Almighty ALLAH for His Blessing and Guidance in completing this project
in time.
.
I would like to admit the great and unconditional academic support and
encouragement from my family. This success is all because of their prayers and
help in my university career.
I would also like to specially thank my highly regarded teacher and supervisor of
thesis Dr. Safdear Hussain Bouk and co-supervisor Dr. Nadeem Javaid for their
utmost help and precious guidance and expert advices during execution of the
work and throughout my studies at COMSATS Institute of Information Technology,
Islamabad.
I am extremely obliged to my research fellows , Imran Israr , Aziz-ur-
Rehman, Mohammad Mateen Yaqoob and Haad Akmal , who helped me in some
way or the other in completing my thesis.
At the end my gratitude is for my parents and friends who supported me morally
to reach this stage.
.
Mansoor Mustafa,
viii
ABSTRACT
MVC: Modified VIKOR Model based Clustering Technique
for WSNs
Stability and lifetime of Wireless Sensor Networks (WSNs) mainly depend on energy
of each node in the network. Hence, it is necessary for WSNs to be energy efferent.
There are different methods to preserve energy in WSNs and clustering is one of those
methods. It is a technique to divide whole network into small blocks, each having a
managing node, called cluster head (CH) and rest act as members. CH is responsible to
provide communication bridge between members and the base station. In this paper, we
propose a distributed clustering scheme that uses multiple criteria i.e. residual energy,
node density, distance to the Base Station (BS) and average distance between a node
and its neighbors, to select a CH. Modified VIKOR method is used to outrank the
potential nodes as CHs. The realistic multi-hoping communication model is used,
instead of single hop as in previous schemes. Simulation results show that our
purposed technique performs much better than those previous methods in terms of
energy efficiency and network life time. Our proposed scheme has less CH deformation
and control overhead.
.
ix
Table of Contents
1 Introduction 2
1.1 Motivation................................ 3
1.2 ProblemStatement ........................... 4
1.3 Research Methodology . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Thesis Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Evaluation of WSNs 8
2.1 Brief History of WSNs . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Introduction of WSNs . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 Challenges in WSNs . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.3.1 Energy constrained . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.2 Network Deployment . . . . . . . . . . . . . . . . . . . . . . 11
2.3.3 Self-Management . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3.4 Wireless Networking and Communication Channel . . . . . . 12
2.3.5 Hardware Limitations . . . . . . . . . . . . . . . . . . . . . 12
2.4 Difference between Traditional Networks and WSNs . . . . . . . . . 12
2.5 Sensor Node architecture . . . . . . . . . . . . . . . . . . . . . . . . 13
2.5.1 Sensing Subsystem . . . . . . . . . . . . . . . . . . . . . . . 14
2.5.2 Processing Subsystem . . . . . . . . . . . . . . . . . . . . . 14
2.5.3 Communication Subsystem . . . . . . . . . . . . . . . . . . . 14
2.5.4 Power Subsystem . . . . . . . . . . . . . . . . . . . . . . . . 15
2.6 Application of WSNs . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.6.1 Military Application . . . . . . . . . . . . . . . . . . . . . . 16
2.6.2 Environmental Application . . . . . . . . . . . . . . . . . . . 16
2.6.3 Health-Care Applications . . . . . . . . . . . . . . . . . . . . 17
2.6.4 Home applications . . . . . . . . . . . . . . . . . . . . . . . 18
2.6.5 Traccontrol.......................... 19
3 Overview of Routing Techniques and Related Work 21
3.1 Overview of Routing Techniques . . . . . . . . . . . . . . . . . . . 21
x
3.1.1 Location Based Routing Protocols . . . . . . . . . . . . . . . 22
3.1.2 Flat and Data Centric Routing Protocols . . . . . . . . . . . 22
3.1.3 Hierarchical Routing Protocols . . . . . . . . . . . . . . . . 23
3.2 Existing Clustering Protocols for WSNs . . . . . . . . . . . . . . . . 23
3.2.1 LEACH ............................. 23
3.2.2 Multi-Hop LEACH . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.3 CEEC .............................. 25
3.2.4 Cluster Head Selection in Wireless Sensor Networks under
Fuzzy Environment . . . . . . . . . . . . . . . . . . . . . . . 27
4 Proposed Clustering Protocol 30
4.1 Energy Model for Wireless Sensor Node . . . . . . . . . . . . . . . . 30
4.2 ProposedScheme ............................ 31
4.2.1 Neighbor discovery . . . . . . . . . . . . . . . . . . . . . . . 32
4.2.2 Cluster Head Selection . . . . . . . . . . . . . . . . . . . . . 32
4.2.3 Sensor Nodes Communication . . . . . . . . . . . . . . . . . 36
5 Simulation and Results 41
5.1 Simulation Environment . . . . . . . . . . . . . . . . . . . . . . . . 41
5.2 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.2.1 Network Lifetime . . . . . . . . . . . . . . . . . . . . . . . . 42
5.2.2 Network Stability . . . . . . . . . . . . . . . . . . . . . . . . 43
5.2.3 Energy Consumption . . . . . . . . . . . . . . . . . . . . . . 43
5.2.4 Number of CHs per round . . . . . . . . . . . . . . . . . . . 44
5.2.5 Control Overhead (Hello) Packets . . . . . . . . . . . . . . . 44
5.2.6 Packets Sent to Base Station . . . . . . . . . . . . . . . . . . 45
6 Conclusion 48
References 48
xi
List of Figures
2.1 Crossbow Sensor Node . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Basic Architecture of WSNs [35] . . . . . . . . . . . . . . . . . . . . 10
2.3 Architecture of WSN node [36] . . . . . . . . . . . . . . . . . . . . 13
2.4 Architecture of WSN node . . . . . . . . . . . . . . . . . . . . . . . 16
2.5 Architecture of WSN node . . . . . . . . . . . . . . . . . . . . . . . 17
2.6 Environmental Application of WSNs [37] . . . . . . . . . . . . . . . 18
2.7 Health-care Applications [37] . . . . . . . . . . . . . . . . . . . . . . 18
3.1 Network Layout of LEACH . . . . . . . . . . . . . . . . . . . . . . 25
3.2 Network Layout of Multi-hop LEACH . . . . . . . . . . . . . . . . 26
3.3 Network Layout of CEEC . . . . . . . . . . . . . . . . . . . . . . . 27
4.1 RadioModel[5]............................. 31
4.2 Neighbor Discovery Process . . . . . . . . . . . . . . . . . . . . . . 33
4.3 CH Selection Process . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.4 Communication Model . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.5 Flow chart of Proposed Protocol . . . . . . . . . . . . . . . . . . . . 39
5.1 Network Stability and Lifetime . . . . . . . . . . . . . . . . . . . . 42
5.2 Network Stability and Lifetime . . . . . . . . . . . . . . . . . . . . 43
5.3 Energy consumption per round . . . . . . . . . . . . . . . . . . . . 44
5.4 Total No. of Cluster heads per round . . . . . . . . . . . . . . . . . 45
5.5 Number of hello packets . . . . . . . . . . . . . . . . . . . . . . . . 46
5.6 Number of packets to base station . . . . . . . . . . . . . . . . . . . 46
xii
List of Tables
4.1 Radio Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.1 Simulation Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 41
xiii
Chapter 1
Introduction
1
Chapter 1
Introduction
In recent years WSNs have got interest due to the development in micro-electronics.
Applications of WSNs include atmospheric sensing, like temperature, pressure,
and humidity etc, natural disasters, like earthquake monitoring and military ap-
plications like battle field monitoring [1]. WSNs consist of tiny battery operated
electronic devices called sensor nodes. Besides sensing, these nodes have memory,
processor and possess communication capabilities. Sensor nodes are deployed in
the field either strategically or randomly. In case of random deployment these
sensor nodes have to be self organized in ad-hoc fashion. Replacement or recharge
of battery is not possible once sensor nodes are dispersed in field [2][3].
Network lifetime and stability of WSNs depend on energy consumption of sensor
nodes. Lesser the energy consumption, longer will be the network lifetime. Sensor
nodes consume energy in processing and communication with BS or other sensor
nodes. For prolonging network lifetime and stability, WSNs should be energy
efficient. Energy efficiency can be achieved by different means, like intelligently
designing of MAC and routing protocols. Routing protocols can be flat, Minimum
Transmission Energy (MTE) or hierarchical protocols. In flat routing protocols
sensor nodes directly send data to BS. In MTE each sensor node sends data to its
neighbor node, therefore load at sensor nodes near to base station is much greater
than other sensor nodes, resulting in shorter lifetime [4]. In hierarchical protocols,
whole network is divided in small number of blocks, called clusters. Each cluster
consists of multiple numbers of sensor nodes. Inside each cluster a CH is selected
to perform management and routing tasks for that cluster [5].
Flat and MTE transmission protocols perform better in small networks. If network
size is large, then these types of protocols are not efficient. To overcome this
deficiency, hierarchical or clustering protocols are designed for large networks [5].
2
In recent years researchers are focusing on design of energy efficient hierarchical
protocols. Suitable CH selection is a major constituent of hierarchal protocols. CH
performs a major role in network stability and lifetime. Many routing protocols
are designed in this regard. Most of the previous designed clustering protocols
consider single criterion for CH selection [6] or CH is selected randomly based on
probabilistic model [5]. Single criterion includes residual energy, distance from BS
or density of sensor node etc. Considering only one criterion may not be sufficient
in many specific cases, for example in a clustering protocol which considers residual
energy as CH selection parameter, worst case arises when a sensor node with high
residual energy located far away from BS, selected as CH. In this case this selected
CH require large amount of energy to forward data to BS, resulting in shorter
network lifetime. Similar kind of situation arises when considering distance form
BS as a CH selection parameter and a sensor node near to BS selected as CH having
very small amount of residual energy. Hence by analyzing these bottlenecks it is
observed that single criterion is not sufficient enough to prolong network stability
and lifetime.
In this thesis we design a distributed multi-hop energy efficient clustering proto-
col based on four criteria, including residual energy, number of neighbor sensor
nodes, distance form BS and average distance between neighbor sensor nodes.
Proposed protocol is based on modified VIKOR model [7]. VIKOR is an abbrevi-
ation in bosnian language for a technique called VIeKriterijumska Optimizacija I
Kompromisno Resenje, whose English meaning is Multicriteria optimization and
compromise solution. The modified VIKOR method was developed for multi-
criteria optimization of complex systems. The compromise solution, compromise
ranking-list and weight stability intervals is determined by this method for prefer-
ence consistency of the compromise solution obtained with the given weights. The
main aim of this method is to focus on ranking and selecting one alternative from
a set of alternatives, if there is some contradiction between different alternatives.
It introduces the multi-criteria ranking index based on the specific measure of
closeness to the ideal solution [9].
1.1 Motivation
Earlier research in WSNs clustering, including LEACH and M-LEACH are based
on some random value or single criteria for CH selection [5][8]. In most of the
previously proposed protocols for example CEEC, Centralized algorithm is used, in
which BS elects the CHs [6]. Also in previous schemes, cluster heads are changing
3
in every round. The goal of the thesis is to use Multi-criteria to select CHs which
is better than single criteria in terms of performance. Centralized schemes and
frequent CH change increase processing overhead. Hence previous schemes are
not much energy efficient. The VIKOR method is a popular method applied in
multi-criteria analysis. Although it is a famous method but in some particular
situations error would occur in calculation, for example when a criterion has same
value for all sensor nodes. To overcome the calculations errors occurring in VIKOR
method, modified VIKOR is introduced in [7]. In our proposed scheme we will
use modified VIKOR method for the selection of cluster head.
To improve communication model, we introduce two level hierarchal communica-
tion. In first level CH collects data from the sensor nodes. In second level, CHs
either sends that data directly to BS or transmit to the nearby CH. The decision
of transmitting data either to BS or other CH is taken on the basis of distance
between CH and BS.
1.2 Problem Statement
In most of the existing clustering techniques CH is selected randomly and based
on single criterion. They use centralized scheme, means BS performs CH selection
process. Also they use single hop communication model. In previous protocols CH
is changing in every round. The bottlenecks of previous protocols are discussed
below.
Problems with single criterion:
Residual energy is considered mostly
Other information, like sensor nodes density, sensor nodes location etc,
are not considered
So optimal CH is not selected.
Problems with centralize algorithm:
Increased CH selection time
Increased control overhead
Increased processing overhead, maximize energy consumption
Problems with single hop communication:
4
CHs that are far away from BS consumes more energy in communication
with BS
Shorter network lifetime and stability period
Problems with frequent CH change:
increased processing over head. Results in shorter lifetime of sensor
nodes.
In this thesis we improve the deficiencies in existing clustering protocols. We de-
sign and simulate a new clustering protocol which is based on four criteria, includ-
ing residual energy of sensor nodes, number of neighbors of sensor nodes, distance
form BS and average distance of nodes form their neighbors. For this purpose we
use modified VIKOR model for CH selection. In order to reduce control overhead,
we avoid quick deformation of CHs. Reduction in control overhead minimizes en-
ergy consumption due to which network life increases and stability improves. In
our proposed scheme we use realistic communication model by introducing multi-
hop communication model in inter-cluster communication mechanism.
1.3 Research Methodology
In this work we comprehensively analyse the bottlenecks in existing clustering pro-
tocols, and design a new clustering scheme based on multiple criteria. Different
weights are assigned to each criteria according to their importance. Each node
share all four criteria value with all its neighbors, then rank index value is calcu-
lated for each node. The node with highest rank, elect itself as CH. The elected
CHs remain CHs until their rank index value differ from any other node value
by 0.1. (threshold value). In intra cluster, nodes with 5mrange of CH directly
communicate to CH, other nodes perform multi-hoping with neighbor nodes. Sim-
ilarly, in inter cluster CHs within 20mrange of BS directly communicate with BS,
remaining CHs perform multi-hoping with neighbor CHs.
1.4 Thesis Organization
Remaining thesis is organized as follows:
Evaluation of WSNs is presented in Chapter 2, including history of WSNs, detailed
introduction of WSNs, basic WSN architecture and common applications. This
chapter gives basic overview of field.
5
Chapter 3 discusses overview of general routing techniques, types of routing pro-
tocols. Related work about our proposed protocol is also presented in this chapter.
At the end we discuss some well known routing protocols in the field of WSNs.
In Chapter 4, our propose protocol is discussed in detail. We describe all phase
along with mathematical formulae and equations. Detailed flow charts of proposed
protocol are also given in this chapter.
Chapter 5 discusses simulation parameters and simulation results of our proposed
scheme. In this chapter we compare the results of proposed scheme with previous
clustering techniques, and analyze the improvements in our proposed scheme in
terms of network lifetime and stability.
Chapter 6 concludes the manuscript of our thesis report.
6
Chapter 2
Evaluation of WSNs
7
Chapter 2
Evaluation of WSNs
2.1 Brief History of WSNs
Military is the major reasons for the development of the WSNs. It is started in
1978, when Defense Advance Research Projects Agency (DARPA) organized the
distributed sensor nets workshop [10]. The major focus in that workshop was to
cop the research challenges in sensor networks, including network technologies,
signal processing techniques and distributed algorithms. Then in 1980s DARPA
started Distributed Sensors Networks program (DSN). This was then followed by
Sensor Information Technology (SensIT) program. The concept of Wireless In-
tegrated Network Sensors (WINS) [11] started by University of California at Los
Angeles with the collaboration of Rockwell Science Center. In 1996 WINS pro-
duced a smart sensing system which was integrated with multiple sensor nodes,
interface circuits, DSPs, wireless radio and micro-controller on a single CMOS
based chip. Another project which focused on the design on extremely small sen-
sor nodes called motes. This project was design by University of California and
knows as Smart Dust project [12]. The main purpose of this project was to show
that integrated sensor devices can be as small as dust particles. Energy con-
sumption minimization was the main focus of the PicoRadio [13] project, which
was developed Berkeley Wireless Research Center (BWRC). Which produce such
low-power sensing devices with very small energy consumption that the can be
power from environmental energy sources for example vibrational or solar energy.
The micro-Adaptive Multi-domain Power-aware Sensors (AMPS) project, devel-
oped by MIT, also focused on low power software and hardware components for
sensor nodes. This included micro-controllers capable of dynamic voltage scaling
and some techniques for redesign of data processing algorithms to reduce energy
consumptions requirements at software level [14]. Most of the previous research in
8
the field of wireless sensor nodes was conducted by educational institutes, however
in last few years a number of commercial efforts have also appeared. The main
contributors of this commercial research are Worldsens [15], Crossbow [16], Senso-
ria [17], Dust Networks [18] and Ember Corporation [19]. Crossbow sensor node
is shown in fig. 2.1.Varieties of sensor devices produced by these companies are
available to be purchased and deployed in different application scenarios. Differ-
ent tools for programming, management maintenance, Sensors data visualization
and analysis are also provided along with the sensor node nodes devices by the
companies.
Figure 2.1: Crossbow Sensor Node
2.2 Introduction of WSNs
Wireless sensor networks is a growing technology in recent years which has a large
number of application, including environmental sensing, industrial sensing and
health care diagnostic etc. The advancements of these areas help the researchers
to enhance the performance of WSNs. Large number of sensor nodes, deployed
in a certain are for a specific purpose, bring themselves together to form a WSN.
WSNs are usually deployed in an environment to monitor static or dynamic events.
The measurement of static events (such as temperature, humidity etc) is very easy
to carry out. On the other hand, dynamic events are typically non-cooperative
event is the movement of an unwanted vehicle in a battle field and the movement
of whales in the ocean. They are not easy to monitor and they are not stable
as they go up and down. Therefore, it is highly difficult to study energy saving
schemes for sensing of dynamic event. For example, a forest monitoring appli-
cation involves static monitoring approach whereas a target tracking application
involves a dynamic monitoring approach [20]. After sensing the monitored data
these sensor nodes send the data to a central controlling device. Besides, some
of the sensor nodes are directly connected to controlling devises, in most of the
situations sensor nodes are connected remotely and tirelessly to centralized pro-
cessing stations. Since in most of the applications sensor nodes are deployed in
9
remote locations without human interference, they are usually inaccessible after
deployment. Hence, a wireless sensor node not only sense data but must also be
capable of processing of data and communication with coordinator. So it must
have processing, communication and storage components. Sensor node not only
sense its own data but also collects and aggregates data from other sensor nodes
in the network, therefore sensor nodes not only communicate with the BS but also
with other sensor nodes in there communication range, which allows the sensor
nodes to cover large geographical area. Figure 2.2 shows the basic architecture of
WSNs.
Figure 2.2: Basic Architecture of WSNs [35]
The sensing, processing and communication require a specific amount of energy
which is provided by limited power supply unit of sensor. Since power supply unit
provide small amount of power as they are mostly equipped with AA or AAA
battery. Hence efficient power utilization in sensor nodes is very difficult because
sensor once deployed in a field cannot be recharged. Therefore energy is the major
issue in WSN. Therefore different techniques are used for efficient utilization of
energy to increase the lifetime of sensor and WSN.
2.3 Challenges in WSNs
Due to small size of sensor nodes, types of communication channel, deployment
scenarios and environmental conditions, WSNs have many constrains. These con-
straints impact the design of a WSN, leading to protocols and algorithms that
10
differ from their counterparts, like ad-hoc networks, in other distributed systems.
This section describes the most important challenges in a WSNs.
2.3.1 Energy constrained
As illustrated above, the energy is a major concern in WSNs, because they are
powered through batteries, which must be either replaced or re-energized (e.g.,
using solar power) when exhausted. For most of WSNs applications, none of the
options is suitable as they will simply be superfluous, once their energy source is
washed-out because sensor nodes are deployed in areas which are not accessible.
Battery recharge is notably a matter of concern as the strategy applied to energy
consumption is affected. In case non rechargeable batteries are employed, a sensor
node should be able to maneuver till its mission time is over or the battery is
replaced. The life of the mission depends on the type of application, as scientists
monitoring significant movements may require sensor nodes that have a life span
of several years whereas a sensor in a battlefield scenario may only be needed for
a few hours or days. As a consequence, the first and often most important design
challenge for a WSN is energy efficiency.
2.3.2 Network Deployment
Predetermined deployment of sensor is not required in several WSNs applications.
Networks deployed in remote or inaccessible areas like, sensor nodes serving the
evaluation of battlefield or disaster areas could be thrown from airplanes in the re-
quired places. While some of the sensor nodes might not carry on because of being
dropped from height might get damaged, so they may not be able to start their
sensing activities. On the contrary, surviving sensor nodes will autonomously start
functioning a variety of setup and configuration steps, including the establishment
of link with neighboring sensor nodes, locating their positions, and initialization
of their sensing responsibilities [21]. There can be difference in mode of operation
of sensor nodes based on information such as, a sensor nodes location and the
number of identities of its neighbors. All this may determine the amount and type
of information it will produce and forward on behalf of other sensor nodes.
2.3.3 Self-Management
In most of situations, inaccessible places that are out of human reach require the
deployment of wireless sensor nodes, hence these sensor nodes cannot be repaired
11
or managed once deployed, as in battlefield. Support and maintenance is not
possible in such scenario. So, sensor nodes should have the ability of self-managing
so that they can configure themselves, operate and cooperate with each other, and
are adaptable to environmental changes [22], can withstand topological changes
and failures without any human interaction [23].
2.3.4 Wireless Networking and Communication Channel
Sensor network designer faces number of challenges due to the dependence on
wireless networks and communications. As the power of radio frequency signals
weakens while it propagates through the hindrance filled medium. As a fact,
relationship between the power consumption and the distance between BS and
sensor nodes has nonlinear traits. So, for efficient transmission it is better to
divide longer distances into several shorter distances that results in providing
better strategy in handling the challenge of supporting multi-hop communications
and routing.
2.3.5 Hardware Limitations
Creating smaller, low cost and more efficient devices is the main purpose of sensor
design. It is needed to execute a dedicated application with more efficient energy
consumption by the sensor nodes. Therefore typical sensor nodes have relatively
low processing speeds and less storage capacities. Integration of many desirable
components like GPS receivers is excluded just because of the need for small
form factor and efficient resource utilization. The design of many protocols and
algorithms executed in WSNs is affected by such limitations of hardware. For
example routing tables with entries of each potential destination may be too large
for small memory of a sensor. So, sensor nodes memory can store only a small
amount of data, such as a list of neighbors. Limited hardware resources compel
to design algorithms and protocols in such a way that they operate efficiently.
2.4 Difference between Traditional Networks and
WSNs
It becomes clear from the previous section that many design choices in a WSN
differ from the design choices of other systems and networks. Some of the main
difference between traditional networks and WSNs are discussed in this section.
12
Traditional networks are general purpose networks, which serve many applica-
tions. Whereas WSNs are single purpose networks and serving one application,
like WSNs are used to sense any environmental condition.
In traditional networks the primary concern is network performance and latencies.
However in WSNs, main constrains is energy. Sensor nodes are designed and
deployed in such manner that it consumes minimum amount of energy.
Traditional networks are mostly designed and engineered according to plans. Whereas
WSNs are mostly deployed in ad-hoc fashion i-e without any proper planning.
In traditional networks devices and networks are operated in controlled and mild
environment, whereas WSNs are usually operated in harsh conditions.
Maintenance and repair of traditional networks is much easier than WSNs, because
in WSNs physical access to sensor nodes is very difficult or even impossible. Hence
component failure in traditional networks is addressed through maintenance and
repair, where as in WSNs it is addressed in design of network.
In traditional networks, it is possible to have centralized management, and it
is possible to obtain global information about network, whereas in WSNs most
decisions are made localized without support of central management.
2.5 Sensor Node architecture
A typical wireless network consists of a small battery powered device called sen-
sor nodes. The four basic components of sensor nodes are Sensing subsystem,
power subsystem, communication subsystem and processing subsystem [24]. Fig-
ure 2.3 given below explains the single node architecture. The explanation of each
subsystem is given as follows:
Figure 2.3: Architecture of WSN node [36]
13
2.5.1 Sensing Subsystem
Sensors are the components that are responsible for creating a link between com-
putational environment and the real world. Therefore, sensors are considered a
crucial component of WSNs sensor node architecture. Sensors are small hardware
devices that are capable of reading data of a certain type and reacting to a certain
changes in the environment in which they are designed to work. The sensor reads
the data in a particular environment, and then sends this data to be processed
further. Transducers are used in the sensors subsystem to read this signal. The
data is sent in the form of analog signals, which is converted into digital form
using Analog to digital convertors (ADC) [25]. A sensor node consists of ADC
and micro-controllers for processing. Each sensor is designed separately according
to the application for which it is used. An important factor of consideration when
choosing a sensor node is the consumption and size of the sensor node battery.
2.5.2 Processing Subsystem
Along with a memory unit and convertors, another important component of sensor
node subsystem is the processing unit. The data is processed by the communica-
tion interface. This data is used to study the networks performance. The purpose
of this unit here in a WSN architecture is to adapt the routing information and it
can also help in aligning the topology [26]. Data gathering, data acquisition and
processing of the retrieved data are also responsibilities of this subsystem. Data
fusion also occurs in this subsystem as data packets arrive from different sensor
nodes are all combined to make a single data packet. This subsystem reduces the
energy used for transmission between the user and sensor.
2.5.3 Communication Subsystem
The transmission of data is the responsibility of this subsystem. The signals are
carried from sensors to end users through BS and sensor nodes use radio frequen-
cies for this purpose. BS here is responsible for communication between user and
WSNs. The number of BS can vary for each network depending on the area,
number of sensors nodes and requirements of the network. In any network, each
sensor node can coordinate and communicate with all the other sensor nodes.
The two types of communications between sensor nodes are infrastructure and
application. Communication for building, maintaining and optimizing a network
is called infrastructure communications. Even a static sensor node network needs
14
infrastructure communication, as due to some environmental changes such situa-
tions can occur when sensor nodes fail and topologies vary. In such a situation,
the topology needs reconfiguration for which communication infrastructure is re-
quired. The data gathered by the sensor nodes, which is sent to the monitoring
end for further processing is called application communications. The amount of
energy required for a single packet to transmit a packet is fixed and it depends on
the distance. For longer distance, the energy required will be higher. Therefore,
to reduce the energy required, shortest path is selected for transmission of packets.
This communication is called application-based communication [27].
2.5.4 Power Subsystem
The power subsystem is composed batteries. All of the above subsystems need
power to perform their activities and tasks. The power consume by these sub-
system is restricted and carefully managed by power subsystem because of energy
constrains in sensor nodes. The power subsystems are normally equipped with
AA batteries or AAA batteries. This subsystem also contains power regulator. In
most of the cases the power consumption is minimized by turning the sensor in
sleep and awake mode. Power subsystem performs the task of the turning mode
of sensor nodes.
2.6 Application of WSNs
WSNs may consist of a several number of sensor nodes and these are application
specific. The various type of sensors include infrared, thermal, acoustic infrared,
seismic, radar and visual. Due to these various types of sensors WSNs can be used
for various applications such as monitoring of physical and atmospheric conditions
like temperature, humidity, pressure, noise level and movement. It can also be used
to measure characteristic like size and speed of object. The implementation of
wireless and communication and micro-sensing combines gives life to a whole new
line of applications. These applications can be termed into two main categories
[28] and four further sub-categories shown in fig.2.4.
The two main categories of applications of WSNs are Tracking applications and
monitoring applications. Tracking applications include tracking of objects, hu-
mans and animals. While the monitoring applications include monitoring of envi-
ronment, power, health, etc.
15
There are some other commercial and non-commercial applications as well, which
are discussed here in detail.
Figure 2.4: Architecture of WSN node
2.6.1 Military Application
Sensor networks have a very wide scope when it comes to military applications
due to their self-organization, ease of deployment and fault tolerance. The wire-
less networks are used in many areas such as military command, intelligence,
communications, computing, surveillance, targeting and control system. Military
applications of wireless networks are far spread. These networks can be used to
gather intelligence and valuable information about the movement and strategy of
the enemy in a battlefield. It can also be used in surveillance and spying. More-
over, it can also be used to detect biological and nuclear attacks in a war. For
instance, a sensor based system is developed that helps to locate the location of a
hidden sniper in an area [29]. The sensor network acquires information from the
muzzle blast and shot waves to identify the location of a shooter.
2.6.2 Environmental Application
A lot of environmental applications have been developed using wireless networks.
WSN can be used to track and detect various environmental conditions such as
environmental conditions that can affect the crop and livestock. Moreover, it is
also used to track animals. WSN are capable of collecting data in a large area over
a certain amount of time. There are other more important biological and chem-
ical applications of WSNs that can help the forestry and agricultural systems.It
can be used for biological and chemical precision detection, fire detection in fire,
16
Figure 2.5: Architecture of WSN node
detecting volcano eruption and flood detection. Microscope of Redwood [30] is a
known example of environmental application of WSNs. Here, the wireless sensor
network records and monitors trees in an area in California. The sensor nodes are
used in this system to collect information such as, photo synthetically-active solar
radiation, humidity and temperature in air. These sensors are placed at different
positions on the tree at different heights. Biologists study this collected data and
read the environmental changes to prove their theories. Similarly, another envi-
ronmental application of WSNs is a system that is used to track the migration of
animals called ZebraNet system [31].
2.6.3 Health-Care Applications
With an aging population and increasing health care cost, there is an immense
need of cost effective health-care solutions to solve the challenges faced by humans.
For this purpose, wireless sensor networks have been programmed and put used
to a number of health-care applications. Body Area sensor network (BAN) is
an application of WSNs and it is used to collect physiological data from a human
body. This is used for monitoring of disables and it is used in developing interfaces
for disables as well. Another application of body sensor network is to detect the
movement of elderly people and see if they have fallen. It helps the patients to
move with freedom, while also allowing the doctor to be aware of the problem
as soon as it occurs, and sometimes even before it by reading the symptoms.
Similarly, you can also provide sensitive patients with a small sensor node, which
17
Figure 2.6: Environmental Application of WSNs [37]
will keep track of their heart-rate and blood pressure.
Figure 2.7: Health-care Applications [37]
2.6.4 Home applications
With the advance of technology, the tiny sensor nodes can be embedded into fur-
niture and appliances, such as vacuum. The sensor nodes are tiny devices that
can be used to collect tons of different type of data. These devices can be at-
tached to or placed inside your daily usage appliances such as vacuum cleaner,
refrigerators and microwave ovens. These sensors allow these devices to commu-
nicate in a room server. Using this data, you can program these devices to be
18
self-organized and self-regulated and adapted. It will help you to turn your home
server environment into a smart computing environment. ZigBee is an example of
such applications. In these application, WSNs are used to form a Personal Area
Network which is capable of monitoring and controlling of several functions such
as electricity switching control, safety control, surveillance and monitoring.
2.6.5 Traffic control
Ground transportation plays a key role in socio-economic infrastructure. It pro-
vides a link between many systems such as emergency response, supply-chain and
public health. Due to increase traffic, congestion, which cause billion gallons of
fuel and billion hours per annum. This can be avoided by building new roads,
however this solution is not feasible because of lack of free space and high cost.
Another solution is to place distributed systems which reduce the congestions.
One the most suitable distributed system for traffic control is WSN. The sensor
nodes in WSNs gather information about size, speed and densities of vehicles. If
detect congestion from these information and suggest alternative and emergency
exits.
19
Chapter 3
Literature Review and
Related Work
20
Chapter 3
Overview of Routing Techniques
and Related Work
3.1 Overview of Routing Techniques
Limited bandwidth and energy supply in sensors creates many challenges in WSNs.
To overcome these challenges, energy aware protocols are required at all levels of
protocol stack. To offer efficient power management in WSN, researchers have
been focus on areas such as system-level power awareness like radio communi-
cation hardware, low duty cycle issues and energy-aware MAC protocols. Also,
it was observed that the network layer offers a better means through which reli-
able relaying of data and energy-efficient route setup within a network can help
to maximize the network lifetime. It should be noted that routing in WSN has
much distinguishable features compare to contemporary communication and ad
hoc networks [20]. These features are as follows:
I. WSN cannot be built with global addressing such as Internet scheme Protocol
(IP) for the enormous number of sensor nodes as the maintenance of ID overhead
is high;
II. There is significant redundancy in generated data because several sensors may
gather the same data within a particular field. These redundancy needs to be
removed to increase the bandwidth utilization and also reduce energy consumption
in the network.
III. Transmission power, processing capacity and storage are constraint factors to
be considered when managing a WSN.
Due to these differences, new protocols are being researched and fashioned to
21
eliminate the problem faced in WSN. These routing protocols have been fashion
on sensor nodes characteristics alongside it application and architectural require-
ment. The various protocols can be classified as location-based, data-centric or
hierarchical.
3.1.1 Location Based Routing Protocols
Most routing technique for WSN depends on location information of sensor nodes
for estimation of distance between two specific nodes to deduce energy consump-
tion. For example, to sense a known region, through the use of location sensor,
a specified query can be sent to that known region and this will significantly re-
duce transmitted data compare to a broadcast request being sent to the entire
network [20]. In other words, the location-based protocol utilizes the position in-
formation to relay the data to the desired regions rather than the whole network.
An example of location based routing protocol is, Geographic Adaptive Fidelity
(GAF) algorithm. GAF is mainly designed for energy conservation. Here, the
sensor network is divided into grids and each sensor is equipped with GPS, for its
location information in a particular grid. There is a switching between the states
which means that the sensors which are not active are turned off maintaining the
constant routing fidelity simultaneously.
3.1.2 Flat and Data Centric Routing Protocols
In most sensor networks, the sensor nodes themselves are less important than the
information they generate. Therefore, in data-centric routing techniques, the fo-
cus is on the retrieval and dissemination of information of a particular type or
described by certain attributes, as opposed to the data collection from particular
sensor nodes. Since assigning global identifiers to every sensor nodes in a WSN
may appear not visible (due to sheer number) in some randomly deployed ap-
plication, data transmitted by every sensor node within a particular region has
significant redundancy with it. To reduce this redundancy, data centric protocols
were developed to select a set of sensor nodes and also utilize data aggregation
during relaying of data. An example of data centric is Flooding routing protocol
in which each sensor node receives data and then sends them to the neighbors by
broadcasting, unless a maximum number of hops for the packet are reached or
the destination of packet is achieved. Another type of Flat based routing protocol
is Sensor Protocol for Information via Negotiation (SPIN) [32]. SPINs data are
named using meta-data that highly describes the characteristics of the data which
22
is the key feature of SPIN.
The advantage of SPIN is that the topological changes are localized since each
of the sensor nodes needs to know only its single-hop neighbors. However, it
has a disadvantages of scalability (not scalable) and also, the nodes around the
base station could deplete their energy if the BS is interested in too many event.
Moreover, SPINs data advertisement mechanism cannot guarantee the delivery of
data. For instance, if the sensor nodes which are interested in the data are far
away from the source node and the nodes between source and destination are not
interested in that data, such a data will not be transmitted to the destination at
all.
3.1.3 Hierarchical Routing Protocols
Hierarchical routing protocols are based on the grouping of nodes into clusters
to address some weaknesses of at routing protocols, most notably scalability and
efciency. The main idea behind hierarchical routing is that sensor nodes commu-
nicate only directly with a leader node in their own cluster, typically referred to as
CH. These CHs, which may be more powerful and less energy-constrained devices
than regular sensor nodes, are then responsible for propagating the sensor data
to the sink. This approach can signicantly reduce the communication and energy
burdens on sensor nodes, while CH will experience signicantly more trafc than
regular sensor nodes.
3.2 Existing Clustering Protocols for WSNs
Energy consumption is the major issue in wireless sensors network. In order to
balance the energy consumption and prolong the lifetime of the WSNs many asym-
metrical clustering algorithms have been proposed for WSNs. Here we mention
some of the most relevant clustering algorithms.
3.2.1 LEACH
Low-Energy Adaptive Hierarchy (LEACH) [5] is very popular clustering protocol
designed in year 2000. In LEACH protocol the network is divided into clusters.
In each cluster one sensor node as act as CH and the remaining sensor nodes act
as member nodes. The CH in a cluster is responsible to collect, aggregate, and
forward the data from member to BS. The member nodes cannot communicate
23
directly with the BS, as CH act as router between member nodes and BS. Because
of extra responsibilities, CH consumes more energy than the member nodes and
can die early, if it remains permanent as a CH for the entire life time of network.
This issue is resolved by using randomized rotation of CH positions among nodes
in network. Hence load of energy consumption is distributed among all nodes. The
overall process of LEACH protocol is divided in rounds. Each round in LEACH
has two phases; setup phase and steady state phase. In setup phase all sensors
nodes organize themselves into clusters, where as in steady state phase normal
nodes send data to their respective CHs. Each node elects itself as CH in the
setup phase by generating a random value between (0-1). If the random is less
than a threshold T given in equation 3.1 for a sensor node it elected itself as CH
and broadcast an advertisement packet to sensor nodes in its range, otherwise
sensor node waits for an advertisement packet from other sensor node which is
elected as CH [5].
T(n) = (p
1p(rmod(1/p)) if n ǫG
0 otherwise (3.1)
Where Gis set of nodes that have not been cluster-head in previous 1/p rounds,
Psuggested percentage of cluster-head, ris current round.
For proper distribution of CH responsibilities, sensor nodes which are elected as
CHs in last 1/P are not eligible for CH election process. Sensors nodes, which
are selected as CHs before 1/P rounds are eligible for CH election process. The
uniform service of each node as a CH prevents the uneven energy consumption.
The CSMA/CA protocol is used by the CH to broadcast its status. The non-
cluster nodes select their CH on basis of Received Signal Strength Indication
(RSSI). Time Division Multiple Access (TDMA) slots are assigned by each CH to
its member nodes. In steady state phase member nodes communicates with the
CH in their associated TDMA slot. The CH collects the data from its member
nodes, aggregates and compresses the collected data and transmits it directly to
the BS. Network layout of LEACH protocol is shown if fig. 3.1.
3.2.2 Multi-Hop LEACH
It is difficult for all CHs to communicate directly with the base station when the
sensor nodes deployed area is very large. High transmission power is required to
send the data from CHs to the base station, if the base station is far away from
the CHs. LEACH only assumes that all CHs are at single hop distance from the
24
0 10 20 30 40 50 60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
Figure 3.1: Network Layout of LEACH
base station, which is not a suitable approach for large geographical area. An
approach, Multi-Hop LEACH, addresses this issue [8]. Multi-Hop LEACH is an
enhancement of LEACH, which reduces the energy consumption of the CHs in
large WSNs. Like LEACH protocol, the formation of clusters and selection of the
CHs is done in the setup phase. However, During steady state phase, the non-
cluster head nodes send their data to the CHs. The CHs aggregate and transmit
that data towards through other CHs.
The most feasible and energy efficient path is selected for that CH which is far
from the base station. The method used for the selection of intermediate CH is the
distance between the CH and the base station. The CH closer to the base station
receives the data from the other CH which is far from base station. This helps to
save energy of those CHs which belong to the clusters with larger distance from the
base station as higher transmission energy cost is required for communication with
larger distances. Figure 3.2 represents the network layout of multi-hop LEACH.
3.2.3 CEEC
Centralized Energy Efficient Clustering (CEEC) [6] routing protocol is centralized
algorithm for cluster-head selection and its performance is much better in het-
erogeneous environment as compared to homogeneous environment. An advance
heterogeneous network model is proposed for CEEC, in which sensor nodes with
different energy level are deployed in separate regions. In CEEC, BS performs
25
0 20 40 60 80 100 120 140 160 180 200
0
50
100
150
200
250
Figure 3.2: Network Layout of Multi-hop LEACH
central clustering formation in network, with help of central control algorithm of
CEEC. Advance central control algorithm considers four factors for selection of
cluster-heads, initial energy of sensor nodes, residual energy of sensor nodes, and
average energy of each region and location of sensor nodes. Operation of CEEC
is based on rounds, with adjustable duration. Each round is divided into Network
Settling Time (NST) and Network Transmission Time (NTT).
During Network Settling Time (NST) suitable cluster heads are selected by BS,
with the help of central control algorithm. In central control algorithm, BS calcu-
lates three different average energies for normal, advance and super sensor nodes
to obtain separate cluster-heads for all regions. BS knows the initial energy of all
sensor nodes for the first round and it can simply calculate the average energies
for first round. After first round, sensor nodes provide their residual energy infor-
mation to BS. Another significance of the protocol is that sensor nodes provide
their residual energy information along data packets transmitted in NTT. In this
way centralized algorithm calculate weight-age of nodes for selection of CH. Best
selected CHs in CEEC are called First Selected Cluster Heads (FSCHs).
During Network Transmission Time (NTT) real communication take place towards
the BS. In NTT all sensor nodes send their data to their Current Cluster Heads
(CCHs), in assigned time slots. Cluster-heads receive the data from its cluster
and aggregate the data. Figure 3.3 represents the network layout of multi-hop
LEACH.
26
0 50 100 150 200 250 300
0
50
100
150
200
250
300
Figure 3.3: Network Layout of CEEC
3.2.4 Cluster Head Selection in Wireless Sensor Networks
under Fuzzy Environment
In [34] authors have proposed CH selection technique which much is based on
multiple criteria. In this technique, Fuzzy -TOPSIS system of MCDM, is used.
Three criteria, which are residual energy, number of neighbor and distance form
BS, are consider in this technique. In this paper centralized algorithm is used,
means BS is taking decision for CH selection on this basis of sensor node values
of all criteria. They assign different weights to each criterion. This technique uses
the concept of fuzzy membership function. Fuzzy membership functions are used
in fuzzy systems to assign relative importance to each criterion. CH selection is
divided into five steps. In step 1, values obtained from each sensor node about its
every neighbor is converted into normalize values. Note that there are some criteria
whose larger value is suitable for a sensor node to be selected as a CH e.g. residual
energy and number of sensor nodes. These criteria are called Positive criteria, and
are normalized using formula for Positive Ideal Solution (PIS). On the other hand,
the criteria with smaller values are appropriate for a sensor node to be selected
as a CH, e.g. distance from BS. This type of criteria are called Negative criteria,
and normalized by using formula for Negative Ideal Solution (NIS). In step 2,
respective weights are assigned to each criteria and fuzzy membership function
is determined. On this basis of weights and fuzzy membership function, weight
matrix is calculated. In step 3, PIS and NIS is calculated form weight matrix.
Then in step 4, separation measure is calculated also from weight matrix. Finally
27
in step 5, rank index is calculated. Then the sensor nodes with higher rank value
from their neighbor are elected as CHs, and other sensor nodes join these CHs.
Because of the centralized algorithm processing time and hello overhead increases,
as every sensor node has to send its information to BS in each round. Hello
overhead also increases because of CH changing in every round.
28
Chapter 4
Proposed Clustering Protocol
29
Chapter 4
Proposed Clustering Protocol
In this chapter we discuss our proposed protocol in detail along with mathematical
equations, explanations and flow diagrams. Next section describes the first order
radio model used in wireless sensors.
4.1 Energy Model for Wireless Sensor Node
First order radio model is assumed by mostly energy efficient routing protocols
as given in [5]. We also adopt this radio model to analyze realistically the pro-
posed protocol with other clustering protocols. Energy dissipation values indicate
the hardware energy consumptions during transmission, reception and aggrega-
tion of data. EeleT X and EeleRX denote the energy consumptions values to run
transmitter and receiver circuitry per bit. Radio dissipates ǫamp for transmission
amplifier in order to obtain suitable Eb/N0.
Energy values used in selection of suitable Eamp and Eafs are given in Table 4.1.
These values are extensively adopted in previous research works.
Table 4.1: Radio Characteristics
Parameter value
Data Aggregation Energy cost 50pj/bit j
Transmitter Electronics EeleT x 50 nj/bit
Receiver Electronics EeleRx 50 nj/bit
Transmit amplifier Eamp 100 pj/bit/m4
Transmit amplifier Eafs 100 pj/bit/m2
We are using radio model of sensor node is shown in Fig.4.1, used in [5]. Energy
dissipation of an individual sensor node depends upon the number of transmissions,
30
L bit packet Transmit
Electronics Tx Amlifier
Receiver
Electronics
L bit packet
ETx(d)
EeleTX
*L Eamp
*L*d2
EeleRX
*L
d
Figure 4.1: Radio Model [5]
number of receptions, amount of data to transmit or receive, and distance between
transmitter and receiver. But in most of cases, only transmission energy cost is
considered during performance analysis of protocol. Energy dissipation cost for
transmitting and receiving given in [5] is given in equation 4.1 and 4.2 respectively.
ET X (L, d) =
L×EeleT X +L×Eaf sd2If d < do
L×EeleT X +L×Eamp d4If d do
(4.1)
ERX =L×EeleRX (4.2)
ET X (L, d) is transmitting energy cost, ERX is receiving energy cost, Ldata bits
to be transmitted, Ef s is free space transmission Model, Emp is Multi-path trans-
mission Model.
4.2 Proposed Scheme
In this section, we describe the details of our propose scheme for CH selection
using distributed algorithm. In WSNs CHs are selected using one or more criteria.
These criteria include residual energy, distance of node from BS and node density
31
etcetera. However considering all of these criteria is very difficult job because a
node might have high residual energy but its node density will be low, another
node might have low residual energy but might have high node density. Therefore
we adopt modified VIKOR [7] method, which is multi-criteria technique to select
a suitable CH. According to this method each node evaluates itself with respect
to its neighbors and decides to become CH or not. Four criteria to be consider by
a node for CH selection are residual energy, node density, distance from BS and
average distance from its neighbor nodes. Our proposed scheme consists of three
phases, which are described below in detail:
4.2.1 Neighbor discovery
Here we explain neighborhood discover technique of a node. Initially when nodes
are randomly deployed, they don’t have any information about their surroundings
and therefore cannot take part in CH selection process. In order to discover neigh-
bors, every node broadcast its ID in a HELLO message in its intra-communication
range and listens to medium for short period of time for similar messages. each
time when a sensor node receives this HELLO message, it creates and updates its
neighbor table by storing neighbor IDs from this HELLO message. The sensor
nodes in our proposed scheme are static, therefore no need of neighbor discovery
in every communication cycle. After completing neighbor discovery, sensor nodes
calculate average distance from its neighbor and also calculate its distance from
the BS. The sensor node then broadcast an information packet which contains in-
formation about sensor nodes energy, number of neighbors, average distance from
its neighbor and sensor nodes distance from BS. In our proposed scheme neighbor
discovery process is performed after few communication cycles in order to check
for dead neighbors and to update routing table accordingly. Performing neighbor
discovery process after few communication cycles decreases HELLO over head, re-
sulting in lower energy consumption. Overall neighbor discovery process is shown
in fig.4.2. In the next section, we present detail description of CH selection in our
proposed scheme.
4.2.2 Cluster Head Selection
Comprehensive explanation of CH selection process of our proposed scheme is
given in this section. Based on information packet received, sensor node calcu-
lates its CH value (CH V al) and share this (CH V al) with all of its neighbors.
Following are the steps involve in calculating (CH V al) of a sensor node.
32
Start
Start
Round = 1
Round = 1
Broadcast
Hello packet
Broadcast
Hello packet
If Round =
Neighboring
round
If Round =
Neighboring
round
Listen to the
medium for
Hello packets
Listen to the
medium for
Hello packets
Create and update
neighbor table
Create and update
neighbor table End
End
NoYes
Yes
No
Figure 4.2: Neighbor Discovery Process
Step 1: Every sensor node compares its each criterion value Vwith every neighbor
and determines the node with maximum value and the node with minimum value
for that criteria. We represent maximum value with Positive Ideal Candidate
(PIC), and minimum value with Negative Ideal Candidate (NIC). Formulae for
calculating PIC and NIC are taken form [7] and modified according to our model
are shown in equation 4.3 and 4.4 respectively.
P I Ckj(k= 1,2, ..., N ) = max[Vij |(j= 1,2, ..., c)
i= 1,2, ...(n+ 1)] (4.3)
NICkj(k= 1,2, ..., N) = min[Vij |(j= 1,2, ..., c)
i= 1,2, ...(n+ 1)] (4.4)
Where kis the number of sensor node, Nis the total number of sensor nodes,
iis number of neighbor (it includes the sensor node it self), jis the number of
criteria, nis the total number of neighbor, cis the total number of criteria and
max valkand min valkare the maximum and minimum values of kth sensor node’s
33
jth neighbor.
Step 2: We assign weight wto all four criteria according to their importance.
Since residual energy is major criterion for selection of CH, so we assign 0.4 weight
to it and 0.2 to each other three criteria; i.e. number of neighbor sensor nodes,
distance from BS and average distance between sensor node and its neighbors.
Step 3: In this step each sensor node calculates the distance of each criterion (j)
to the ideal solution (MAX/MIN) and then computes sum of all the distances
to obtain the final value. For some criteria like, residual energy, greater values
provides better chances for a sensor node to become CH therefore its distance is
calculated to the positive ideal solution by equation 4.5 while criterion like distance
to BS low value provides better chance for a sensor node to become CH, so its
distance is calculated from negative ideal solution by equation 4.6. Equation 4.7 is
used to calculate regret measure. These equations are taken form [7] and modified
according to our model.
S
k=
c
X
j=1
[(P I CjVkj)/(P I CjNICj)] (4.5)
S
k=
c
X
j=1
[(Vkj N ICj)/(P ICjN I Cj)] (4.6)
Rk=maxj[wj((P I Cj)/
(P I CjNI Cj))|j= 1,2, .., c] (4.7)
Where Skrepresents the distance of sensor node kto the ideal solution and Rk
shows the regret measure. After computing these Skand Rkvalues, sensor node
broadcast this information to its neighbors and listens to channel for similar in-
formation. The ranking of a sensor node with its neighbor depends on the above
information.
Step 4: Each node calculate its CH value (CH val) and its neighbor CH val from
the broadcast information using following equations, which are taken and modified
from [7].
34
CH valk=
v(SiSmin)/(Smax Smin)+
(1 v)(RiRmin)/(Rmax Rmin)Smax 6=Smin &Rmax 6=Rmin
(RiRmin)/(Rmax Rmin)Smax =Smin &Rmax 6=Rmin
(SiSmin)/(Smax Smin)Rmax =Rmin &Smax 6=Smin
constant Rmax =Rmin &Smax =Smin
Smin =min[(Si)] (4.8)
Smax =max[(Si)] (4.9)
Rmin =min[(Ri)] (4.10)
Rmax =max[(Ri)] (4.11)
where i= 1,2,3.......(n+ 1) and vis the weight for strategy of maximum group
utility and 1 vis the weight of individual regret. Value of vis set to 0.5.
Step 5: After calculating CH val, node will then compare its CH val with its
neighbors. If C H val of the nodes is less then its any neighbor, it elect itself as
CH and broadcast advertisement packet to all of its neighbors. If the node CH val
is greater then at least one of its neighbors, it will wait for advertisement packet
from its neighbor node which has lowest CH val.
Every sensor node checks that whether its is the round immediately after de-
ployment or not. If its not the round immediately after deployment, sensor node
predicting CH’s CH val by calculates the remaining energy of neighbor’s and CH’s
by calculating their energy consumed on transmission. The remaining energy of
neighbor’s and CH’s is calculated from their last know energy and distance from
BS. All other criteria remain same as all the sensor nodes are static. Sensor node
compare its own CH val with the CH’s CH val. If difference between sensor
node’s and CH’s CH val is greater than threshold, the sensor node broadcast a
notification message with all criteria information and and waits for its neighbor’s
information. After receiving neighbor’s information, sensor node perform all of the
above steps. If difference between sensor node’s and CH’s CH val is not greater
than the threshold, sensor node wait for a specific period of time for notification
from other node. The sensor node broadcast its criteria information if receive
35
notification from other sensor node and get associated with new CH otherwise it
remain associated with the current CH as there no new CH elected. The flowchart
of CH election process is show in fig. 4.3.When CHs are elected they send an
advertisement packet to their entire neighbors. Nodes which receive this adver-
tisement packet will send a join request message to the CH in order to join that
respective cluster. CH, in response assign TDMA slots for node communication,
and sends an acknowledgment packet to node.
4.2.3 Sensor Nodes Communication
After CH selection and nodes association with their CHs, nodes communication
starts. Nodes send their sensed data to their associated CHs in assigned TDMA
slots. The data received from the nodes are collected, aggregated and amplified
by the CH. CH then forward this collected data to the BS as shown in fig. 4.4. In
our proposed scheme there are two types of communication operations. These are
inter-cluster communication and intra-cluster communication. In Multi-hop inter-
cluster communication, when whole network is divided into multiple clusters each
cluster has one CH. This CH is responsible for communication for all nodes in the
cluster. CH receive data from all nodes at single-hop and aggregate and transmit
directly to BS or through intermediate CH. In Multi-hop inter-cluster communi-
cation when distance between CH and BS is larger than 10 meter then CH use
intermediate cluster-head to communicate to BS.Through multi-hop communica-
tion, energy consumption is minimized which in turn increases overall network
lifetime and stability. Flowchart of proposed protocol functionality is show in fig-
ure 4.5. Packets to BS and other simulation results are discussed in next section.
36
start
If round= 1
Predict
CH_val
Diff. b/w CH and
any sensor node
CH_val
> 0.1
Wait for
notification
Notification
with all
criteria
information
Notification
Received
Remain
Associated
Broadcast
criteria
Information
Wait for
neighbor
Information
Calculate
CH_val
Less then its
neighbour
Become CH
Wait for CH
broadcast
message
Broadcast CH
information
Broadcast
Information
End
Yes
Yes
Yes
No
Yes
No
No
No
Figure 4.3: CH Selection Process
37
Figure 4.4: Communication Model
38
Node Deployment
Neighbor Discovery
Sharing of Information
(four criteria values
among neighbors)
CH
Send Join Request
Broadcast CH
advertisement
Receive CH
advertisement
Receive Join
Request
CH selection based on
Modified VIKOR
Yes No
Start
Receive TDMA slot
from CH
CHs assign and broadcast
TDMA slots to member nodes
Receive data
from nodes
Receive data
from nodes Send sensed data in assigned
slot
Send sensed data in assigned
slot
Aggregate received data
from member nodes
Aggregate received data
from member nodes
Base Station
Base Station
End
Communication
Phase
Distance >
10 m
Intermediate
CH
Intermediate
CH
Aggregate all
data and
transmit to
BS
Aggregate all
data and
transmit to
BS
Non CH node
Figure 4.5: Flow chart of Proposed Protocol
39
Chapter 5
Simulation and Results
40
Chapter 5
Simulation and Results
The performance of our proposed scheme is compared and analyzed in this Chap-
ter. We compare and analyze our proposed scheme with fuzzy-TOPSIS based
[34] routing protocol. Simulation of previously proposed scheme and our proposed
protocol is done in MATLAB.
5.1 Simulation Environment
In our simulation 100 number of sensor nodes is randomly dispersed in a field of
100m×100m. The BS is located at the top of the field. In our simulation we
made some assumptions. First is that sensor nodes are continuously monitoring
the environment and always have data to be sent to the BS. Second is that wireless
channel is free of signal collision and interference. Simulation parameters are given
in Table 5.1.
Table 5.1: Simulation Parameters
Parameter Value
Network Area 100m x 100m
Number of Nodes n100
Base Station Position (50,100)
Initial Energy 0.5 J
Data Aggregation Energy 50pj/bit/report
Data Packet Size 4000 bits
Hello Packet Size 200 bits
41
5.2 Simulation Results
We evaluate fuzzy-TOPSIS based protocol and and our proposed protocol by the
network lifetime, stability of the network, packets to BS, network energy and
hello overhead packets. Following graphs show simulation results of our purposed
scheme compared with previous schemes.
5.2.1 Network Lifetime
In fig. 5.1, the network lifetime is shown which depicts that our proposed schemes
outstrip the previous protocol. This is because of minimizing hello overhead by
restricting the CH selection to some threshold. Furthermore, the multi-hope com-
munication model that we used in our protocol, also play a vital role in increasing
the network lifetime. Figure 5.1 shows that previous fuzzy-TOPSIS based proto-
col dies out around 1100 rounds. Where as in our protocol all sensor nodes die
put around 2400 rounds. This proves that our protocol perform much better that
others.
0 500 1000 1500 2000 2500
0
10
20
30
40
50
60
70
80
90
100
Number of Rounds
Number of Allive nodes
Fuzzy Based Model
Proposed Model
Figure 5.1: Network Stability and Lifetime
42
5.2.2 Network Stability
Number of dead sensor nodes per round in the network are shown in fig.5.2 .It is
clear from graph that in fuzzy-TOPSIS based protocol, first node dies around 750
rounds where as our proposed scheme first sensor node dies around 1400 rounds.
This show that our protocol surpass previously proposed protocol in terms of sta-
bility as well. Reason for this is that in previous fuzzy-TOPSIS based method, BS
is performing CH selection process, which does not depends on geographical con-
ditions of nodes, where as in our purposed scheme every node itself take decision
for CH, considering the knowledge of neighbor nodes.
0 500 1000 1500 2000 2500
0
10
20
30
40
50
60
70
80
90
100
Number of Rounds
sgg
Fuzzy Based Model
Proposed Model
Figure 5.2: Network Stability and Lifetime
5.2.3 Energy Consumption
Energy consumption of network per round is shown in Fig. 5.3. It is observed
that our proposed scheme consumes less energy than previous scheme as residual
energy per round is much higher in our protocol as shown in fig. 5.3. A major
constituent of energy consumption is communication process. Almost 70 percent
of whole network’s energy is consumed in communication. So a proper communi-
cation model is very much necessary for any energy efficient clustering protocol.
In our proposed scheme we use multi-hop communication in both inter and intra-
cluster communication, this is the main reason of lower energy consumption in
43
our proposed scheme is much less than previous fuzzy-TOPSIS based scheme. In
previous schemes, energy variation at different points is observed and it due to
re-selection of CHs in each round. This CH re-selection is avoided in our proposed
scheme by introducing the CH changing threshold and that is also the main reason
of less energy consumption in our proposed scheme.
0 500 1000 1500 2000 2500
−5
0
5
10
15
20
25
30
35
40
45
50
Number of Rounds
Residual Energy per Round
Fuzzy Based Model
Proposed Model
Figure 5.3: Energy consumption per round
5.2.4 Number of CHs per round
Figure 5.4 shows the number of CHs per round. It can be analyzed from the
fig. 5.4 that in previous fuzzy-TOPSIS based protocol, number of CHs are almost
12 percent of the total number of sensor nodes. This is because of centralized
algorithm used in this protocol. In our protocol, however distributed algorithm is
used but still the number of CHs remain constant. This is because of the efficient
CH selection technique.
5.2.5 Control Overhead (Hello) Packets
The control overhead or Hello packets are the control signals required for any type
of data processing in WSNs. Lager the control overhead packets, greater will be
44
0 500 1000 1500 2000 2500
0
2
4
6
8
10
12
14
16
Number of Rounds
Number of CHs Per round
Fuzzy Based Model
Proposed Model
Figure 5.4: Total No. of Cluster heads per round
the energy consumption, Hence for any energy efficient clustering protocol it is
necessary that these packets should be minimized. In our proposed scheme CH
change is very rare and we use distributed algorithm for selection of CH. These
are main reasons for minimization of control overhead packets. Figure 5.5 shows
comparison of our proposed scheme with previous fuzzy-TOPSIS model. It can
clearly observed that control overhead packets in our proposed scheme are very
small as compared to previous fuzzy-TOPSIS model. This is because less changes
in CHs per round, which decreases the control overhead tremendously as shown
in fig. 5.5 and increases overall network lifetime.
5.2.6 Packets Sent to Base Station
Number of packets to BS is are shown in fig. 5.6. Total number of packets sent to
BS in our protocol is much higher than other protocols as dipicts in fig. 5.6. This
is because of extended lifetime and stability of network.
45
0 500 1000 1500 2000 2500
0
2
4
6
8
10
12 x 104
Number of Rounds
Number of Hello Packets per Round
Fuzzy Based Model
Proposed Model
Figure 5.5: Number of hello packets
0 500 1000 1500 2000 2500
0
0.5
1
1.5
2
2.5
3x 104
NUmber of Rounds
Number of Packets to BS
Fuzzy Based Model
Proposed Model
Figure 5.6: Number of packets to base station
46
Chapter 6
Conclusion
47
Chapter 6
Conclusion
In this thesis we design a new routing protocol, which overcomes the deficiencies
in the existing routing protocols. Our protocol is based on four criteria, including
residual energy, number of nodes, distance from BS, and average number of nodes.
WE apply modified VIKOR method for selection of cluster head, which overcomes
the bottlenecks in simple VIKOR method. Our proposed protocol is based on
distributed algorithm. We control frequent change of CH in our proposed scheme.If
the index value of any node is greater than index value of CH plus specic threshold,
then the CH will no more be eligible to act as CH, and nodes will perform re-
election process within the cluster. Due to this number of control overhead packets
in proposed protocol is very small. To improve communication model, we will
introduce two level hierarchy for cluster head selection. In first level CH will be
selected on the basis of modified VIKOR method, while in next level CH will be
selected on basis of its location and residual energy.
We perform MATLAB simulation to compare the results of our proposed scheme
with LEACH and previously proposed fuzzy based centralized clustering model.
All results show that the network performance of our proposed scheme is much bet-
ter than perviously proposed clustering techniques. This improvement is achieved
using distributed algorithm, using multi criteria for CH selection and using two
level hierarchy model.
48
References
[1] Shah, T., N. Javaid, and T. N. Qureshi. “Energy Efficient Sleep Awake Aware
(EESAA) Intelligent Sensor Network Routing Protocol.” In Proceedings of the
15th IEEE International Multi Topic Conference (INMIC12), 2012.
[2] J.ElsonandD.Estrin. “An Address-Free Architecture for Dynamic Sensor Net-
works”. Technical Report 00-724, Computer Science Department, USC, Jan-
uary 2000.
[3] D. Estrin, R. Govindan, J. Heidemann, and S. Kumar. “Next Century Chal-
lenges: Scalable Coordination in Wireless Networks” In Proceedings of the
5th Annual ACM/IEEE International Conference on Mobile Computing and
Networking (MOBICOM), pp. 263270, 1999.
[4] S.Muruganathan, Daniel C.F.MA,R.Bhasin, A.Fapojuwo “A Centralized
Energy-Efficient Routing Protocol for Wireless Sensor Networks” IEEE Com-
munications Magazine, 2005
[5] W. Heinzelman, A. Chandrakasan, and H. Balakrishnan, “Energy-efficient
routing protocols for wireless microsensor networks,” in Proc.33rd Hawaii Int.
Conf. System Sciences(HICSS), Maui, HI,Jan. 2000.
[6] Aslam, M., et al. ”CEEC: Centralized energy efficient clustering a new routing
protocol for WSNs.” Sensor, Mesh and Ad Hoc Communications and Networks
(SECON), 2012 9th Annual IEEE Communications Society Conference on.
IEEE, 2012.
[7] Chang, Chia-Ling. ”A modified VIKOR method for multiple criteria analysis.”
Environmental monitoring and assessment 168, no. 1-4 (2010): 339-344.
[8] Xiangning, Fan, and Song Yulin. ”Improvement on LEACH protocol of wireless
sensor network.” Sensor Technologies and Applications, 2007. SensorComm
2007. International Conference on. IEEE, 2007.
49
[9] Opricovic, Serafim, and Gwo-Hshiung Tzeng. ”Compromise solution by
MCDM methods: A comparative analysis of VIKOR and TOPSIS.” European
Journal of Operational Research 156, no. 2 (2004): 445-455
[10] Chong, Chee-Yee, and Srikanta P. Kumar. Sensor networks: evolution, op-
portunities, and challenges. Proceedings of the IEEE 91.8 (2003): 1247-1256.
[11] Pottie, Gregory J. Wireless integrated network sensors (wins): the web gets
physical. Frontiers of Engineering: Reports on Leading-Edge Engineering from
the 2001 NAE Symposium on Frontiers of Engineering. National Academies
Press, 2002.
[12] Kahn, Joseph M., Randy H. Katz, and Kristofer SJ Pister. Next century
challenges: mobile networking for Smart Dust. Proceedings of the 5th annual
ACM/IEEE international conference on Mobile computing and networking.
ACM, 1999.
[13] Rabaey, J., et al. PicoRadio: Ad-hoc wireless networking of ubiquitous low-
energy sensor/monitor nodes. VLSI, 2000. Proceedings. IEEE Computer Soci-
ety Workshop on. IEEE, 2000.
[14] Shih, Eugene, et al. Design considerations for energy-ecient radios in wireless
microsensor networks. Journal of VLSI signal processing systems for signal,
image and video technology 37.1 (2004): 77-94.
[15] http://www.xbow.com
[16] http://www.sensoria.com
[17] http://worldsens.citi.insa-lyon.fr
[18] http://www.dustnetworks.com
[19] http://www.ember.com
[20] Kemal Akkaya and Mohammed Younis (May,2005.). A survey on routing
protocols for wireless sensor networks and Ad hoc networks, pp325 -349.
[21] Qureshi, T. N., et al. ”BEENISH: Balanced Energy Efficient Network Inte-
grated Super Heterogeneous Protocol for Wireless Sensor Networks.” arXiv
preprint arXiv:1303.5285 (2013).
[22] Mills, K.L. (2007) A brief survey of self-organization in wireless sensor net-
works. Wireless Communications and Mobile Computing 7 (7), 823834
[23] Cerpa, A., and Estrin, D. (2004) Ascent: Adaptive self-conguring sensor net-
work topologies. IEEE Transactions on Mobile Computing 3 (3), 272285.
50
[24] M. N. Elshakankiri, M. N. Moustafa, and Y. H. Dakroury, Energy Efficient
Routing protocol For Wireless Sensor Networks, in International Conference on
Intelligent Sensors, Sensor Networks and Information Processing, 2008, pp.393
398.
[25] J. Feng, F. Koushanfar, and M. Potkonjak, Sensor Network Architecture,
Handbook of Sensor Networks, 2004.
[26] Evaluation of Routing Protocols in Wireless Sensor Networks - Arkiv EX -
Blekinge Tekniska Hgskola.
[27] L. B. Ruiz, J. M. S. Nogueira, and A. A. F. Loureiro, Sensor Network Man-
agement, Handbook of Sensor Networks: Compact Wireless and Wired Sensing
Systems, M. Ilyas and I. Mahgoub, Eds., Ch. 3. CRC Press, 2005.
[28] Jennifer Yick, Biswanath Mukherjee, Dipak Ghosal, ”Wireless sensor network
survey,” Computer Networks,Elsevier, vol. 52, pp. 2292-2330, 2008.
[29] G. Simon, M. Maroti, ”Sensor network-based counter sniper system,” in Pro-
ceedings of the 2nd international conference on Embedded networked sensor
systems, Baltimore, MD, 2004, pp. 1-12.
[30] G. Tolle, D. Culler, W. Hong, et al., ”A macroscope in the redwoods,” in
Proceedings of the 3rd international conference on Embedded networked sensor
systems, San Diego, CA, 2005, pp. 51-63.
[31] P. Zhang, C.M. Sadler, S.A. Lyon, M. Martonosi, ”Hardware design experi-
ences in ZebraNet,” in Proceedings of the SenSys04, Baltimore, MD, 2004.
[32] C.F.Chissseenni, P.Monti and A.Nucci, Energy Efficient Design of Wireless
Ad Hoc Networks, in Proceedings of European Wireless, February 2002.
[33] S. Dai, X. Jing and L. Li Research and analysis on routing protocols for
wireless sensor networks, in Proceeding IEEE, pp 407-41, 2005
[34] Azada, Puneet, and Vidushi Sharmab. Cluster Head Selection in Wireless
Sensor Networks under Fuzzy Environment. ISRN Sensor Networks, vol. 2013,
Article ID 909086, 8 pages, 2013. doi:10.1155/2013/909086.
[35] Islam, Md Motaharul, and Eui-Nam Huh. ”A novel addressing scheme for
PMIPv6 based global IP-WSNs.” Sensors 11.9 (2011): 8430-8455.
[36] Eugen Coca and Valentin Popa (2011). Third Generation Active RFID
from the Locating Applications Perspective, Current Trends and Challenges
in RFID, Prof. Cornel Turcu (Ed.), ISBN: 978-953-307-356-9, InTech, DOI:
10.5772/17534. Available from: http://www.intechopen.com/books/current-
51
trends-and-challenges-in-rfid/third-generation-active-rfid-from-the-locating-
applications-perspective
[37] ”COMSYS - Communications and Distributed systems : Wireless
Sensors Network labs” comsys.rwth-aachen.de. n.d. Web. 25 May.
2013. ¡http://www.comsys.rwth-aachen.de/teaching/ss-12/wireless-sensor-
networks-lab/¿.
52
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Clustering is one of the important methods for prolonging the network lifetime in wireless sensor networks (WSNs). It involves grouping of sensor nodes into clusters and electing cluster heads (CHs) for all the clusters. CHs collect the data from respective cluster's nodes and forward the aggregated data to base station. A major challenge in WSNs is to select appropriate cluster heads. In this paper, we present a fuzzy decision-making approach for the selection of cluster heads. Fuzzy multiple attribute decision-making (MADM) approach is used to select CHs using three criteria including residual energy, number of neighbors, and the distance from the base station of the nodes. The simulation results demonstrate that this approach is more effective in prolonging the network lifetime than the distributed hierarchical agglomerative clustering (DHAC) protocol in homogeneous environments.
Article
Full-text available
In past years there has been increasing interest in field of Wireless Sensor Networks (WSNs). One of the major issue of WSNs is development of energy efficient routing protocols. Clustering is an effective way to increase energy efficiency. Mostly, heterogenous protocols consider two or three energy level of nodes. In reality, heterogonous WSNs contain large range of energy levels. By analyzing communication energy consumption of the clusters and large range of energy levels in heterogenous WSN, we propose BEENISH (Balanced Energy Efficient Network Integrated Super Heterogenous) Protocol. It assumes WSN containing four energy levels of nodes. Here, Cluster Heads (CHs) are elected on the bases of residual energy level of nodes. Simulation results show that it performs better than existing clustering protocols in heterogeneous WSNs. Our protocol achieve longer stability, lifetime and more effective messages than Distributed Energy Efficient Clustering (DEEC), Developed DEEC (DDEEC) and Enhanced DEEC (EDEEC).
Conference Paper
In order to increase the network lifetime, scalable and energy-aware routing protocols are very essential for Wireless Sensor Networks (WSNs). In this paper, we propose a heterogeneity-aware Multi-hop Centralized Energy Efficient Clustering (MCEEC) protocol for routing in WSNs. Operation of MCEEC is based upon the advanced central control algorithm, in which Base Station (BS) is responsible for the selection of Cluster-Heads (CHs) which are selected on the bases of wireless sensors' (nodes') residual energy, average energy of the network, and average of the relative distance between nodes and BS. We adopt multi-hop inter-cluster communication for MCEEC. The advanced heterogeneous network model of the proposed protocol divides the network area into three equally spaced rectangular regions such that nodes of the same energy level are deployed in the respective region. Furthermore, nodes can only associate with their own region's CHs. In MCEEC, deployment of nodes in the network area is in descending order of energy level w.r.t BS's position. Simulation results show that MCEEC yields maximum scalability, network lifetime, stability period and throughput as compared to the selected routing protocols.
Article
One of the most compelling challenges of the next decade is the solution of the "last meter" problem, which extends the network into the end-user data-collection and monitoring devices. This paper discusses the challenges and opportunities of a "PicoRadio Network" that supports the assembly of an ad-hoc wireless network of meso-scale, low-cost and low-energy sensor and monitor nodes.
Conference Paper
In the last few years, wireless sensor networks have gained a lot of interest in the research field. WSNs consist of small nodes with sensing, computation, and wireless communications capabilities. Many routing protocols have been proposed for WSNs. Most of the hierarchical algorithms proposed for WSNs concentrate mainly on maximizing the lifetime of the network by trying to minimize the energy consumption, but delay is also an important metric that should be considered. In this paper, we propose pairs energy efficient routing protocol (PEER), a new routing protocol for WSNs that uses dual power management and focuses on minimizing both the energy dissipated and the delay cost. We have evaluated the performance of our protocol for two different cases: for all nodes having the same energy level at startup and for nodes starting with different energy levels. In terms of energy consumption, network lifetime, and average delay, our protocol has performed better than LEACH (a well-known hierarchical sensor network protocol) by more than 200%, 200%, and 400% respectively.
Article
A wireless sensor network (WSN) has important applications such as remote environmental monitoring and target tracking. This has been enabled by the availability, particularly in recent years, of sensors that are smaller, cheaper, and intelligent. These sensors are equipped with wireless interfaces with which they can communicate with one another to form a network. The design of a WSN depends significantly on the application, and it must consider factors such as the environment, the application’s design objectives, cost, hardware, and system constraints. The goal of our survey is to present a comprehensive review of the recent literature since the publication of [I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, E. Cayirci, A survey on sensor networks, IEEE Communications Magazine, 2002]. Following a top-down approach, we give an overview of several new applications and then review the literature on various aspects of WSNs. We classify the problems into three different categories: (1) internal platform and underlying operating system, (2) communication protocol stack, and (3) network services, provisioning, and deployment. We review the major development in these three categories and outline new challenges.
Article
The multiple criteria decision making (MCDM) methods VIKOR and TOPSIS are based on an aggregating function representing “closeness to the ideal”, which originated in the compromise programming method. In VIKOR linear normalization and in TOPSIS vector normalization is used to eliminate the units of criterion functions. The VIKOR method of compromise ranking determines a compromise solution, providing a maximum “group utility” for the “majority” and a minimum of an individual regret for the “opponent”. The TOPSIS method determines a solution with the shortest distance to the ideal solution and the greatest distance from the negative-ideal solution, but it does not consider the relative importance of these distances. A comparative analysis of these two methods is illustrated with a numerical example, showing their similarity and some differences.